diff options
Diffstat (limited to 'drivers/gpu/drm/amd')
162 files changed, 6845 insertions, 2654 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/Makefile b/drivers/gpu/drm/amd/amdgpu/Makefile index c56320e78c0e..f089794bbdd5 100644 --- a/drivers/gpu/drm/amd/amdgpu/Makefile +++ b/drivers/gpu/drm/amd/amdgpu/Makefile @@ -57,7 +57,8 @@ amdgpu-y += amdgpu_device.o amdgpu_kms.o \ amdgpu_xgmi.o amdgpu_csa.o amdgpu_ras.o amdgpu_vm_cpu.o \ amdgpu_vm_sdma.o amdgpu_discovery.o amdgpu_ras_eeprom.o amdgpu_nbio.o \ amdgpu_umc.o smu_v11_0_i2c.o amdgpu_fru_eeprom.o amdgpu_rap.o \ - amdgpu_fw_attestation.o amdgpu_securedisplay.o amdgpu_hdp.o + amdgpu_fw_attestation.o amdgpu_securedisplay.o amdgpu_hdp.o \ + amdgpu_eeprom.o amdgpu-$(CONFIG_PROC_FS) += amdgpu_fdinfo.o @@ -75,7 +76,7 @@ amdgpu-y += \ vega20_reg_init.o nbio_v7_4.o nbio_v2_3.o nv.o navi10_reg_init.o navi14_reg_init.o \ arct_reg_init.o navi12_reg_init.o mxgpu_nv.o sienna_cichlid_reg_init.o vangogh_reg_init.o \ nbio_v7_2.o dimgrey_cavefish_reg_init.o hdp_v4_0.o hdp_v5_0.o aldebaran_reg_init.o aldebaran.o \ - beige_goby_reg_init.o yellow_carp_reg_init.o + beige_goby_reg_init.o yellow_carp_reg_init.o cyan_skillfish_reg_init.o # add DF block amdgpu-y += \ @@ -111,6 +112,7 @@ amdgpu-y += \ psp_v3_1.o \ psp_v10_0.o \ psp_v11_0.o \ + psp_v11_0_8.o \ psp_v12_0.o \ psp_v13_0.o diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index 8ac6eb9f1fdb..d10baa3338bc 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -1108,8 +1108,13 @@ void amdgpu_device_fini_sw(struct amdgpu_device *adev); int amdgpu_gpu_wait_for_idle(struct amdgpu_device *adev); +void amdgpu_device_mm_access(struct amdgpu_device *adev, loff_t pos, + void *buf, size_t size, bool write); +size_t amdgpu_device_aper_access(struct amdgpu_device *adev, loff_t pos, + void *buf, size_t size, bool write); + void amdgpu_device_vram_access(struct amdgpu_device *adev, loff_t pos, - uint32_t *buf, size_t size, bool write); + void *buf, size_t size, bool write); uint32_t amdgpu_device_rreg(struct amdgpu_device *adev, uint32_t reg, uint32_t acc_flags); void amdgpu_device_wreg(struct amdgpu_device *adev, @@ -1385,12 +1390,12 @@ int amdgpu_acpi_smart_shift_update(struct drm_device *dev, enum amdgpu_ss ss_sta int amdgpu_acpi_pcie_notify_device_ready(struct amdgpu_device *adev); void amdgpu_acpi_get_backlight_caps(struct amdgpu_dm_backlight_caps *caps); -bool amdgpu_acpi_is_s0ix_supported(struct amdgpu_device *adev); +bool amdgpu_acpi_is_s0ix_active(struct amdgpu_device *adev); void amdgpu_acpi_detect(void); #else static inline int amdgpu_acpi_init(struct amdgpu_device *adev) { return 0; } static inline void amdgpu_acpi_fini(struct amdgpu_device *adev) { } -static inline bool amdgpu_acpi_is_s0ix_supported(struct amdgpu_device *adev) { return false; } +static inline bool amdgpu_acpi_is_s0ix_active(struct amdgpu_device *adev) { return false; } static inline void amdgpu_acpi_detect(void) { } static inline bool amdgpu_acpi_is_power_shift_control_supported(void) { return false; } static inline int amdgpu_acpi_power_shift_control(struct amdgpu_device *adev, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c index b8655ff73a65..cc9c9f8b23b2 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acp.c @@ -160,17 +160,28 @@ static int acp_poweron(struct generic_pm_domain *genpd) return 0; } -static struct device *get_mfd_cell_dev(const char *device_name, int r) +static int acp_genpd_add_device(struct device *dev, void *data) { - char auto_dev_name[25]; - struct device *dev; + struct generic_pm_domain *gpd = data; + int ret; - snprintf(auto_dev_name, sizeof(auto_dev_name), - "%s.%d.auto", device_name, r); - dev = bus_find_device_by_name(&platform_bus_type, NULL, auto_dev_name); - dev_info(dev, "device %s added to pm domain\n", auto_dev_name); + ret = pm_genpd_add_device(gpd, dev); + if (ret) + dev_err(dev, "Failed to add dev to genpd %d\n", ret); - return dev; + return ret; +} + +static int acp_genpd_remove_device(struct device *dev, void *data) +{ + int ret; + + ret = pm_genpd_remove_device(dev); + if (ret) + dev_err(dev, "Failed to remove dev from genpd %d\n", ret); + + /* Continue to remove */ + return 0; } /** @@ -181,11 +192,10 @@ static struct device *get_mfd_cell_dev(const char *device_name, int r) */ static int acp_hw_init(void *handle) { - int r, i; + int r; uint64_t acp_base; u32 val = 0; u32 count = 0; - struct device *dev; struct i2s_platform_data *i2s_pdata = NULL; struct amdgpu_device *adev = (struct amdgpu_device *)handle; @@ -341,15 +351,10 @@ static int acp_hw_init(void *handle) if (r) goto failure; - for (i = 0; i < ACP_DEVS ; i++) { - dev = get_mfd_cell_dev(adev->acp.acp_cell[i].name, i); - r = pm_genpd_add_device(&adev->acp.acp_genpd->gpd, dev); - if (r) { - dev_err(dev, "Failed to add dev to genpd\n"); - goto failure; - } - } - + r = device_for_each_child(adev->acp.parent, &adev->acp.acp_genpd->gpd, + acp_genpd_add_device); + if (r) + goto failure; /* Assert Soft reset of ACP */ val = cgs_read_register(adev->acp.cgs_device, mmACP_SOFT_RESET); @@ -410,10 +415,8 @@ failure: */ static int acp_hw_fini(void *handle) { - int i, ret; u32 val = 0; u32 count = 0; - struct device *dev; struct amdgpu_device *adev = (struct amdgpu_device *)handle; /* return early if no ACP */ @@ -458,13 +461,8 @@ static int acp_hw_fini(void *handle) udelay(100); } - for (i = 0; i < ACP_DEVS ; i++) { - dev = get_mfd_cell_dev(adev->acp.acp_cell[i].name, i); - ret = pm_genpd_remove_device(dev); - /* If removal fails, dont giveup and try rest */ - if (ret) - dev_err(dev, "remove dev from genpd failed\n"); - } + device_for_each_child(adev->acp.parent, NULL, + acp_genpd_remove_device); mfd_remove_devices(adev->acp.parent); kfree(adev->acp.acp_res); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c index 84a1b4bc9bb4..58520b0ad0bf 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c @@ -26,6 +26,7 @@ #include <linux/slab.h> #include <linux/power_supply.h> #include <linux/pm_runtime.h> +#include <linux/suspend.h> #include <acpi/video.h> #include <acpi/actbl.h> @@ -853,8 +854,8 @@ int amdgpu_acpi_init(struct amdgpu_device *adev) if (amdgpu_device_has_dc_support(adev)) { #if defined(CONFIG_DRM_AMD_DC) struct amdgpu_display_manager *dm = &adev->dm; - if (dm->backlight_dev) - atif->bd = dm->backlight_dev; + if (dm->backlight_dev[0]) + atif->bd = dm->backlight_dev[0]; #endif } else { struct drm_encoder *tmp; @@ -1031,18 +1032,18 @@ void amdgpu_acpi_detect(void) } /** - * amdgpu_acpi_is_s0ix_supported + * amdgpu_acpi_is_s0ix_active * * @adev: amdgpu_device_pointer * * returns true if supported, false if not. */ -bool amdgpu_acpi_is_s0ix_supported(struct amdgpu_device *adev) +bool amdgpu_acpi_is_s0ix_active(struct amdgpu_device *adev) { #if defined(CONFIG_AMD_PMC) || defined(CONFIG_AMD_PMC_MODULE) if (acpi_gbl_FADT.flags & ACPI_FADT_LOW_POWER_S0) { if (adev->flags & AMD_IS_APU) - return true; + return pm_suspend_target_state == PM_SUSPEND_TO_IDLE; } #endif return false; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c index f9c01bdc3d4c..7b46ba551cb2 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c @@ -21,6 +21,7 @@ */ #include "amdgpu_amdkfd.h" +#include "amd_pcie.h" #include "amd_shared.h" #include "amdgpu.h" @@ -553,6 +554,88 @@ uint8_t amdgpu_amdkfd_get_xgmi_hops_count(struct kgd_dev *dst, struct kgd_dev *s return (uint8_t)ret; } +int amdgpu_amdkfd_get_xgmi_bandwidth_mbytes(struct kgd_dev *dst, struct kgd_dev *src, bool is_min) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)dst, *peer_adev; + int num_links; + + if (adev->asic_type != CHIP_ALDEBARAN) + return 0; + + if (src) + peer_adev = (struct amdgpu_device *)src; + + /* num links returns 0 for indirect peers since indirect route is unknown. */ + num_links = is_min ? 1 : amdgpu_xgmi_get_num_links(adev, peer_adev); + if (num_links < 0) { + DRM_ERROR("amdgpu: failed to get xgmi num links between node %d and %d. ret = %d\n", + adev->gmc.xgmi.physical_node_id, + peer_adev->gmc.xgmi.physical_node_id, num_links); + num_links = 0; + } + + /* Aldebaran xGMI DPM is defeatured so assume x16 x 25Gbps for bandwidth. */ + return (num_links * 16 * 25000)/BITS_PER_BYTE; +} + +int amdgpu_amdkfd_get_pcie_bandwidth_mbytes(struct kgd_dev *dev, bool is_min) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)dev; + int num_lanes_shift = (is_min ? ffs(adev->pm.pcie_mlw_mask) : + fls(adev->pm.pcie_mlw_mask)) - 1; + int gen_speed_shift = (is_min ? ffs(adev->pm.pcie_gen_mask & + CAIL_PCIE_LINK_SPEED_SUPPORT_MASK) : + fls(adev->pm.pcie_gen_mask & + CAIL_PCIE_LINK_SPEED_SUPPORT_MASK)) - 1; + uint32_t num_lanes_mask = 1 << num_lanes_shift; + uint32_t gen_speed_mask = 1 << gen_speed_shift; + int num_lanes_factor = 0, gen_speed_mbits_factor = 0; + + switch (num_lanes_mask) { + case CAIL_PCIE_LINK_WIDTH_SUPPORT_X1: + num_lanes_factor = 1; + break; + case CAIL_PCIE_LINK_WIDTH_SUPPORT_X2: + num_lanes_factor = 2; + break; + case CAIL_PCIE_LINK_WIDTH_SUPPORT_X4: + num_lanes_factor = 4; + break; + case CAIL_PCIE_LINK_WIDTH_SUPPORT_X8: + num_lanes_factor = 8; + break; + case CAIL_PCIE_LINK_WIDTH_SUPPORT_X12: + num_lanes_factor = 12; + break; + case CAIL_PCIE_LINK_WIDTH_SUPPORT_X16: + num_lanes_factor = 16; + break; + case CAIL_PCIE_LINK_WIDTH_SUPPORT_X32: + num_lanes_factor = 32; + break; + } + + switch (gen_speed_mask) { + case CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1: + gen_speed_mbits_factor = 2500; + break; + case CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2: + gen_speed_mbits_factor = 5000; + break; + case CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3: + gen_speed_mbits_factor = 8000; + break; + case CAIL_PCIE_LINK_SPEED_SUPPORT_GEN4: + gen_speed_mbits_factor = 16000; + break; + case CAIL_PCIE_LINK_SPEED_SUPPORT_GEN5: + gen_speed_mbits_factor = 32000; + break; + } + + return (num_lanes_factor * gen_speed_mbits_factor)/BITS_PER_BYTE; +} + uint64_t amdgpu_amdkfd_get_mmio_remap_phys_addr(struct kgd_dev *kgd) { struct amdgpu_device *adev = (struct amdgpu_device *)kgd; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h index cf62f43a03da..0462d4aceab7 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h @@ -226,6 +226,8 @@ uint32_t amdgpu_amdkfd_get_num_gws(struct kgd_dev *kgd); uint32_t amdgpu_amdkfd_get_asic_rev_id(struct kgd_dev *kgd); int amdgpu_amdkfd_get_noretry(struct kgd_dev *kgd); uint8_t amdgpu_amdkfd_get_xgmi_hops_count(struct kgd_dev *dst, struct kgd_dev *src); +int amdgpu_amdkfd_get_xgmi_bandwidth_mbytes(struct kgd_dev *dst, struct kgd_dev *src, bool is_min); +int amdgpu_amdkfd_get_pcie_bandwidth_mbytes(struct kgd_dev *dev, bool is_min); /* Read user wptr from a specified user address space with page fault * disabled. The memory must be pinned and mapped to the hardware when @@ -330,7 +332,7 @@ int kgd2kfd_pre_reset(struct kfd_dev *kfd); int kgd2kfd_post_reset(struct kfd_dev *kfd); void kgd2kfd_interrupt(struct kfd_dev *kfd, const void *ih_ring_entry); void kgd2kfd_set_sram_ecc_flag(struct kfd_dev *kfd); -void kgd2kfd_smi_event_throttle(struct kfd_dev *kfd, uint32_t throttle_bitmask); +void kgd2kfd_smi_event_throttle(struct kfd_dev *kfd, uint64_t throttle_bitmask); #else static inline int kgd2kfd_init(void) { @@ -389,7 +391,7 @@ void kgd2kfd_set_sram_ecc_flag(struct kfd_dev *kfd) } static inline -void kgd2kfd_smi_event_throttle(struct kfd_dev *kfd, uint32_t throttle_bitmask) +void kgd2kfd_smi_event_throttle(struct kfd_dev *kfd, uint64_t throttle_bitmask) { } #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c index 4fb15750b9bb..c4b3eeeab0a1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c @@ -1285,11 +1285,22 @@ int amdgpu_amdkfd_gpuvm_acquire_process_vm(struct kgd_dev *kgd, if (avm->process_info) return -EINVAL; + /* Free the original amdgpu allocated pasid, + * will be replaced with kfd allocated pasid. + */ + if (avm->pasid) { + amdgpu_pasid_free(avm->pasid); + amdgpu_vm_set_pasid(adev, avm, 0); + } + /* Convert VM into a compute VM */ - ret = amdgpu_vm_make_compute(adev, avm, pasid); + ret = amdgpu_vm_make_compute(adev, avm); if (ret) return ret; + ret = amdgpu_vm_set_pasid(adev, avm, pasid); + if (ret) + return ret; /* Initialize KFD part of the VM and process info */ ret = init_kfd_vm(avm, process_info, ef); if (ret) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 8398daa0c06a..6ca9fe54ee0a 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -116,6 +116,7 @@ const char *amdgpu_asic_name[] = { "RENOIR", "ALDEBARAN", "NAVI10", + "CYAN_SKILLFISH", "NAVI14", "NAVI12", "SIENNA_CICHLID", @@ -287,7 +288,7 @@ bool amdgpu_device_supports_smart_shift(struct drm_device *dev) */ /** - * amdgpu_device_vram_access - read/write a buffer in vram + * amdgpu_device_mm_access - access vram by MM_INDEX/MM_DATA * * @adev: amdgpu_device pointer * @pos: offset of the buffer in vram @@ -295,22 +296,65 @@ bool amdgpu_device_supports_smart_shift(struct drm_device *dev) * @size: read/write size, sizeof(@buf) must > @size * @write: true - write to vram, otherwise - read from vram */ -void amdgpu_device_vram_access(struct amdgpu_device *adev, loff_t pos, - uint32_t *buf, size_t size, bool write) +void amdgpu_device_mm_access(struct amdgpu_device *adev, loff_t pos, + void *buf, size_t size, bool write) { unsigned long flags; - uint32_t hi = ~0; + uint32_t hi = ~0, tmp = 0; + uint32_t *data = buf; uint64_t last; int idx; if (!drm_dev_enter(&adev->ddev, &idx)) return; + BUG_ON(!IS_ALIGNED(pos, 4) || !IS_ALIGNED(size, 4)); + + spin_lock_irqsave(&adev->mmio_idx_lock, flags); + for (last = pos + size; pos < last; pos += 4) { + tmp = pos >> 31; + + WREG32_NO_KIQ(mmMM_INDEX, ((uint32_t)pos) | 0x80000000); + if (tmp != hi) { + WREG32_NO_KIQ(mmMM_INDEX_HI, tmp); + hi = tmp; + } + if (write) + WREG32_NO_KIQ(mmMM_DATA, *data++); + else + *data++ = RREG32_NO_KIQ(mmMM_DATA); + } + + spin_unlock_irqrestore(&adev->mmio_idx_lock, flags); + drm_dev_exit(idx); +} + +/** + * amdgpu_device_vram_access - access vram by vram aperature + * + * @adev: amdgpu_device pointer + * @pos: offset of the buffer in vram + * @buf: virtual address of the buffer in system memory + * @size: read/write size, sizeof(@buf) must > @size + * @write: true - write to vram, otherwise - read from vram + * + * The return value means how many bytes have been transferred. + */ +size_t amdgpu_device_aper_access(struct amdgpu_device *adev, loff_t pos, + void *buf, size_t size, bool write) +{ #ifdef CONFIG_64BIT + void __iomem *addr; + size_t count = 0; + uint64_t last; + + if (!adev->mman.aper_base_kaddr) + return 0; + last = min(pos + size, adev->gmc.visible_vram_size); if (last > pos) { - void __iomem *addr = adev->mman.aper_base_kaddr + pos; - size_t count = last - pos; + addr = adev->mman.aper_base_kaddr + pos; + count = last - pos; if (write) { memcpy_toio(addr, buf, count); @@ -322,35 +366,37 @@ void amdgpu_device_vram_access(struct amdgpu_device *adev, loff_t pos, memcpy_fromio(buf, addr, count); } - if (count == size) - goto exit; - - pos += count; - buf += count / 4; - size -= count; } + + return count; +#else + return 0; #endif +} - spin_lock_irqsave(&adev->mmio_idx_lock, flags); - for (last = pos + size; pos < last; pos += 4) { - uint32_t tmp = pos >> 31; +/** + * amdgpu_device_vram_access - read/write a buffer in vram + * + * @adev: amdgpu_device pointer + * @pos: offset of the buffer in vram + * @buf: virtual address of the buffer in system memory + * @size: read/write size, sizeof(@buf) must > @size + * @write: true - write to vram, otherwise - read from vram + */ +void amdgpu_device_vram_access(struct amdgpu_device *adev, loff_t pos, + void *buf, size_t size, bool write) +{ + size_t count; - WREG32_NO_KIQ(mmMM_INDEX, ((uint32_t)pos) | 0x80000000); - if (tmp != hi) { - WREG32_NO_KIQ(mmMM_INDEX_HI, tmp); - hi = tmp; - } - if (write) - WREG32_NO_KIQ(mmMM_DATA, *buf++); - else - *buf++ = RREG32_NO_KIQ(mmMM_DATA); + /* try to using vram apreature to access vram first */ + count = amdgpu_device_aper_access(adev, pos, buf, size, write); + size -= count; + if (size) { + /* using MM to access rest vram */ + pos += count; + buf += count; + amdgpu_device_mm_access(adev, pos, buf, size, write); } - spin_unlock_irqrestore(&adev->mmio_idx_lock, flags); - -#ifdef CONFIG_64BIT -exit: -#endif - drm_dev_exit(idx); } /* @@ -518,7 +564,7 @@ void amdgpu_mm_wreg_mmio_rlc(struct amdgpu_device *adev, adev->gfx.rlc.funcs && adev->gfx.rlc.funcs->is_rlcg_access_range) { if (adev->gfx.rlc.funcs->is_rlcg_access_range(adev, reg)) - return adev->gfx.rlc.funcs->rlcg_wreg(adev, reg, v, 0, 0); + return adev->gfx.rlc.funcs->sriov_wreg(adev, reg, v, 0, 0); } else { writel(v, ((void __iomem *)adev->rmmio) + (reg * 4)); } @@ -1395,6 +1441,10 @@ static int amdgpu_device_init_apu_flags(struct amdgpu_device *adev) break; case CHIP_YELLOW_CARP: break; + case CHIP_CYAN_SKILLFISH: + if (adev->pdev->device == 0x13FE) + adev->apu_flags |= AMD_APU_IS_CYAN_SKILLFISH2; + break; default: return -EINVAL; } @@ -2101,6 +2151,7 @@ static int amdgpu_device_ip_early_init(struct amdgpu_device *adev) case CHIP_BEIGE_GOBY: case CHIP_VANGOGH: case CHIP_YELLOW_CARP: + case CHIP_CYAN_SKILLFISH: if (adev->asic_type == CHIP_VANGOGH) adev->family = AMDGPU_FAMILY_VGH; else if (adev->asic_type == CHIP_YELLOW_CARP) @@ -3505,13 +3556,13 @@ int amdgpu_device_init(struct amdgpu_device *adev, r = amdgpu_device_get_job_timeout_settings(adev); if (r) { dev_err(adev->dev, "invalid lockup_timeout parameter syntax\n"); - goto failed_unmap; + return r; } /* early init functions */ r = amdgpu_device_ip_early_init(adev); if (r) - goto failed_unmap; + return r; /* doorbell bar mapping and doorbell index init*/ amdgpu_device_doorbell_init(adev); @@ -3625,6 +3676,8 @@ fence_driver_init: goto release_ras_con; } + amdgpu_fence_driver_hw_init(adev); + dev_info(adev->dev, "SE %d, SH per SE %d, CU per SH %d, active_cu_number %d\n", adev->gfx.config.max_shader_engines, @@ -3737,10 +3790,6 @@ release_ras_con: failed: amdgpu_vf_error_trans_all(adev); -failed_unmap: - iounmap(adev->rmmio); - adev->rmmio = NULL; - return r; } @@ -3796,7 +3845,7 @@ void amdgpu_device_fini_hw(struct amdgpu_device *adev) else drm_atomic_helper_shutdown(adev_to_drm(adev)); } - amdgpu_fence_driver_fini_hw(adev); + amdgpu_fence_driver_hw_fini(adev); if (adev->pm_sysfs_en) amdgpu_pm_sysfs_fini(adev); @@ -3818,7 +3867,7 @@ void amdgpu_device_fini_hw(struct amdgpu_device *adev) void amdgpu_device_fini_sw(struct amdgpu_device *adev) { amdgpu_device_ip_fini(adev); - amdgpu_fence_driver_fini_sw(adev); + amdgpu_fence_driver_sw_fini(adev); release_firmware(adev->firmware.gpu_info_fw); adev->firmware.gpu_info_fw = NULL; adev->accel_working = false; @@ -3893,7 +3942,7 @@ int amdgpu_device_suspend(struct drm_device *dev, bool fbcon) /* evict vram memory */ amdgpu_bo_evict_vram(adev); - amdgpu_fence_driver_suspend(adev); + amdgpu_fence_driver_hw_fini(adev); amdgpu_device_ip_suspend_phase2(adev); /* evict remaining vram memory @@ -3938,7 +3987,7 @@ int amdgpu_device_resume(struct drm_device *dev, bool fbcon) dev_err(adev->dev, "amdgpu_device_ip_resume failed (%d).\n", r); return r; } - amdgpu_fence_driver_resume(adev); + amdgpu_fence_driver_hw_init(adev); r = amdgpu_device_ip_late_init(adev); @@ -4428,7 +4477,7 @@ int amdgpu_device_pre_asic_reset(struct amdgpu_device *adev, amdgpu_fence_driver_force_completion(ring); } - if(job) + if (job && job->vm) drm_sched_increase_karma(&job->base); r = amdgpu_reset_prepare_hwcontext(adev, reset_context); @@ -4892,7 +4941,7 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev, DRM_INFO("Bailing on TDR for s_job:%llx, hive: %llx as another already in progress", job ? job->base.id : -1, hive->hive_id); amdgpu_put_xgmi_hive(hive); - if (job) + if (job && job->vm) drm_sched_increase_karma(&job->base); return 0; } @@ -4916,7 +4965,7 @@ int amdgpu_device_gpu_recover(struct amdgpu_device *adev, job ? job->base.id : -1); /* even we skipped this reset, still need to set the job to guilty */ - if (job) + if (job && job->vm) drm_sched_increase_karma(&job->base); goto skip_recovery; } @@ -5283,6 +5332,10 @@ int amdgpu_device_baco_exit(struct drm_device *dev) adev->nbio.funcs->enable_doorbell_interrupt) adev->nbio.funcs->enable_doorbell_interrupt(adev, true); + if (amdgpu_passthrough(adev) && + adev->nbio.funcs->clear_doorbell_interrupt) + adev->nbio.funcs->clear_doorbell_interrupt(adev); + return 0; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index 2bd13fc2541a..d637b0536f84 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -870,11 +870,10 @@ MODULE_PARM_DESC(reset_method, "GPU reset method (-1 = auto (default), 0 = legac module_param_named(reset_method, amdgpu_reset_method, int, 0444); /** - * DOC: bad_page_threshold (int) - * Bad page threshold is to specify the threshold value of faulty pages - * detected by RAS ECC, that may result in GPU entering bad status if total - * faulty pages by ECC exceed threshold value and leave it for user's further - * check. + * DOC: bad_page_threshold (int) Bad page threshold is specifies the + * threshold value of faulty pages detected by RAS ECC, which may + * result in the GPU entering bad status when the number of total + * faulty pages by ECC exceeds the threshold value. */ MODULE_PARM_DESC(bad_page_threshold, "Bad page threshold(-1 = auto(default value), 0 = disable bad page retirement)"); module_param_named(bad_page_threshold, amdgpu_bad_page_threshold, int, 0444); @@ -1213,6 +1212,9 @@ static const struct pci_device_id pciidlist[] = { {0x1002, 0x740F, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_ALDEBARAN|AMD_EXP_HW_SUPPORT}, {0x1002, 0x7410, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_ALDEBARAN|AMD_EXP_HW_SUPPORT}, + /* CYAN_SKILLFISH */ + {0x1002, 0x13FE, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_CYAN_SKILLFISH|AMD_IS_APU}, + {0, 0, 0} }; @@ -1467,7 +1469,7 @@ static int amdgpu_pmops_suspend(struct device *dev) struct amdgpu_device *adev = drm_to_adev(drm_dev); int r; - if (amdgpu_acpi_is_s0ix_supported(adev)) + if (amdgpu_acpi_is_s0ix_active(adev)) adev->in_s0ix = true; adev->in_s3 = true; r = amdgpu_device_suspend(drm_dev, true); @@ -1483,7 +1485,7 @@ static int amdgpu_pmops_resume(struct device *dev) int r; r = amdgpu_device_resume(drm_dev, true); - if (amdgpu_acpi_is_s0ix_supported(adev)) + if (amdgpu_acpi_is_s0ix_active(adev)) adev->in_s0ix = false; return r; } diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_eeprom.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_eeprom.c new file mode 100644 index 000000000000..4d9eb0137f8c --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_eeprom.c @@ -0,0 +1,239 @@ +/* + * Copyright 2021 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#include "amdgpu_eeprom.h" +#include "amdgpu.h" + +/* AT24CM02 and M24M02-R have a 256-byte write page size. + */ +#define EEPROM_PAGE_BITS 8 +#define EEPROM_PAGE_SIZE (1U << EEPROM_PAGE_BITS) +#define EEPROM_PAGE_MASK (EEPROM_PAGE_SIZE - 1) + +#define EEPROM_OFFSET_SIZE 2 + +/* EEPROM memory addresses are 19-bits long, which can + * be partitioned into 3, 8, 8 bits, for a total of 19. + * The upper 3 bits are sent as part of the 7-bit + * "Device Type Identifier"--an I2C concept, which for EEPROM devices + * is hard-coded as 1010b, indicating that it is an EEPROM + * device--this is the wire format, followed by the upper + * 3 bits of the 19-bit address, followed by the direction, + * followed by two bytes holding the rest of the 16-bits of + * the EEPROM memory address. The format on the wire for EEPROM + * devices is: 1010XYZD, A15:A8, A7:A0, + * Where D is the direction and sequenced out by the hardware. + * Bits XYZ are memory address bits 18, 17 and 16. + * These bits are compared to how pins 1-3 of the part are connected, + * depending on the size of the part, more on that later. + * + * Note that of this wire format, a client is in control + * of, and needs to specify only XYZ, A15:A8, A7:0, bits, + * which is exactly the EEPROM memory address, or offset, + * in order to address up to 8 EEPROM devices on the I2C bus. + * + * For instance, a 2-Mbit I2C EEPROM part, addresses all its bytes, + * using an 18-bit address, bit 17 to 0 and thus would use all but one bit of + * the 19 bits previously mentioned. The designer would then not connect + * pins 1 and 2, and pin 3 usually named "A_2" or "E2", would be connected to + * either Vcc or GND. This would allow for up to two 2-Mbit parts on + * the same bus, where one would be addressable with bit 18 as 1, and + * the other with bit 18 of the address as 0. + * + * For a 2-Mbit part, bit 18 is usually known as the "Chip Enable" or + * "Hardware Address Bit". This bit is compared to the load on pin 3 + * of the device, described above, and if there is a match, then this + * device responds to the command. This way, you can connect two + * 2-Mbit EEPROM devices on the same bus, but see one contiguous + * memory from 0 to 7FFFFh, where address 0 to 3FFFF is in the device + * whose pin 3 is connected to GND, and address 40000 to 7FFFFh is in + * the 2nd device, whose pin 3 is connected to Vcc. + * + * This addressing you encode in the 32-bit "eeprom_addr" below, + * namely the 19-bits "XYZ,A15:A0", as a single 19-bit address. For + * instance, eeprom_addr = 0x6DA01, is 110_1101_1010_0000_0001, where + * XYZ=110b, and A15:A0=DA01h. The XYZ bits become part of the device + * address, and the rest of the address bits are sent as the memory + * address bytes. + * + * That is, for an I2C EEPROM driver everything is controlled by + * the "eeprom_addr". + * + * P.S. If you need to write, lock and read the Identification Page, + * (M24M02-DR device only, which we do not use), change the "7" to + * "0xF" in the macro below, and let the client set bit 20 to 1 in + * "eeprom_addr", and set A10 to 0 to write into it, and A10 and A1 to + * 1 to lock it permanently. + */ +#define MAKE_I2C_ADDR(_aa) ((0xA << 3) | (((_aa) >> 16) & 7)) + +static int __amdgpu_eeprom_xfer(struct i2c_adapter *i2c_adap, u32 eeprom_addr, + u8 *eeprom_buf, u16 buf_size, bool read) +{ + u8 eeprom_offset_buf[EEPROM_OFFSET_SIZE]; + struct i2c_msg msgs[] = { + { + .flags = 0, + .len = EEPROM_OFFSET_SIZE, + .buf = eeprom_offset_buf, + }, + { + .flags = read ? I2C_M_RD : 0, + }, + }; + const u8 *p = eeprom_buf; + int r; + u16 len; + + for (r = 0; buf_size > 0; + buf_size -= len, eeprom_addr += len, eeprom_buf += len) { + /* Set the EEPROM address we want to write to/read from. + */ + msgs[0].addr = MAKE_I2C_ADDR(eeprom_addr); + msgs[1].addr = msgs[0].addr; + msgs[0].buf[0] = (eeprom_addr >> 8) & 0xff; + msgs[0].buf[1] = eeprom_addr & 0xff; + + if (!read) { + /* Write the maximum amount of data, without + * crossing the device's page boundary, as per + * its spec. Partial page writes are allowed, + * starting at any location within the page, + * so long as the page boundary isn't crossed + * over (actually the page pointer rolls + * over). + * + * As per the AT24CM02 EEPROM spec, after + * writing into a page, the I2C driver should + * terminate the transfer, i.e. in + * "i2c_transfer()" below, with a STOP + * condition, so that the self-timed write + * cycle begins. This is implied for the + * "i2c_transfer()" abstraction. + */ + len = min(EEPROM_PAGE_SIZE - (eeprom_addr & + EEPROM_PAGE_MASK), + (u32)buf_size); + } else { + /* Reading from the EEPROM has no limitation + * on the number of bytes read from the EEPROM + * device--they are simply sequenced out. + */ + len = buf_size; + } + msgs[1].len = len; + msgs[1].buf = eeprom_buf; + + /* This constitutes a START-STOP transaction. + */ + r = i2c_transfer(i2c_adap, msgs, ARRAY_SIZE(msgs)); + if (r != ARRAY_SIZE(msgs)) + break; + + if (!read) { + /* According to EEPROM specs the length of the + * self-writing cycle, tWR (tW), is 10 ms. + * + * TODO: Use polling on ACK, aka Acknowledge + * Polling, to minimize waiting for the + * internal write cycle to complete, as it is + * usually smaller than tWR (tW). + */ + msleep(10); + } + } + + return r < 0 ? r : eeprom_buf - p; +} + +/** + * amdgpu_eeprom_xfer -- Read/write from/to an I2C EEPROM device + * @i2c_adap: pointer to the I2C adapter to use + * @eeprom_addr: EEPROM address from which to read/write + * @eeprom_buf: pointer to data buffer to read into/write from + * @buf_size: the size of @eeprom_buf + * @read: True if reading from the EEPROM, false if writing + * + * Returns the number of bytes read/written; -errno on error. + */ +static int amdgpu_eeprom_xfer(struct i2c_adapter *i2c_adap, u32 eeprom_addr, + u8 *eeprom_buf, u16 buf_size, bool read) +{ + const struct i2c_adapter_quirks *quirks = i2c_adap->quirks; + u16 limit; + + if (!quirks) + limit = 0; + else if (read) + limit = quirks->max_read_len; + else + limit = quirks->max_write_len; + + if (limit == 0) { + return __amdgpu_eeprom_xfer(i2c_adap, eeprom_addr, + eeprom_buf, buf_size, read); + } else if (limit <= EEPROM_OFFSET_SIZE) { + dev_err_ratelimited(&i2c_adap->dev, + "maddr:0x%04X size:0x%02X:quirk max_%s_len must be > %d", + eeprom_addr, buf_size, + read ? "read" : "write", EEPROM_OFFSET_SIZE); + return -EINVAL; + } else { + u16 ps; /* Partial size */ + int res = 0, r; + + /* The "limit" includes all data bytes sent/received, + * which would include the EEPROM_OFFSET_SIZE bytes. + * Account for them here. + */ + limit -= EEPROM_OFFSET_SIZE; + for ( ; buf_size > 0; + buf_size -= ps, eeprom_addr += ps, eeprom_buf += ps) { + ps = min(limit, buf_size); + + r = __amdgpu_eeprom_xfer(i2c_adap, eeprom_addr, + eeprom_buf, ps, read); + if (r < 0) + return r; + res += r; + } + + return res; + } +} + +int amdgpu_eeprom_read(struct i2c_adapter *i2c_adap, + u32 eeprom_addr, u8 *eeprom_buf, + u16 bytes) +{ + return amdgpu_eeprom_xfer(i2c_adap, eeprom_addr, eeprom_buf, bytes, + true); +} + +int amdgpu_eeprom_write(struct i2c_adapter *i2c_adap, + u32 eeprom_addr, u8 *eeprom_buf, + u16 bytes) +{ + return amdgpu_eeprom_xfer(i2c_adap, eeprom_addr, eeprom_buf, bytes, + false); +} diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_eeprom.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_eeprom.h new file mode 100644 index 000000000000..6935adb2be1f --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_eeprom.h @@ -0,0 +1,37 @@ +/* + * Copyright 2021 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#ifndef _AMDGPU_EEPROM_H +#define _AMDGPU_EEPROM_H + +#include <linux/i2c.h> + +int amdgpu_eeprom_read(struct i2c_adapter *i2c_adap, + u32 eeprom_addr, u8 *eeprom_buf, + u16 bytes); + +int amdgpu_eeprom_write(struct i2c_adapter *i2c_adap, + u32 eeprom_addr, u8 *eeprom_buf, + u16 bytes); + +#endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c index d4547d195173..0b1c48590c43 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c @@ -417,9 +417,6 @@ int amdgpu_fence_driver_start_ring(struct amdgpu_ring *ring, } amdgpu_fence_write(ring, atomic_read(&ring->fence_drv.last_seq)); - if (irq_src) - amdgpu_irq_get(adev, irq_src, irq_type); - ring->fence_drv.irq_src = irq_src; ring->fence_drv.irq_type = irq_type; ring->fence_drv.initialized = true; @@ -525,7 +522,7 @@ int amdgpu_fence_driver_init(struct amdgpu_device *adev) * * Tear down the fence driver for all possible rings (all asics). */ -void amdgpu_fence_driver_fini_hw(struct amdgpu_device *adev) +void amdgpu_fence_driver_hw_fini(struct amdgpu_device *adev) { int i, r; @@ -553,7 +550,7 @@ void amdgpu_fence_driver_fini_hw(struct amdgpu_device *adev) } } -void amdgpu_fence_driver_fini_sw(struct amdgpu_device *adev) +void amdgpu_fence_driver_sw_fini(struct amdgpu_device *adev) { unsigned int i, j; @@ -572,49 +569,18 @@ void amdgpu_fence_driver_fini_sw(struct amdgpu_device *adev) } /** - * amdgpu_fence_driver_suspend - suspend the fence driver - * for all possible rings. - * - * @adev: amdgpu device pointer - * - * Suspend the fence driver for all possible rings (all asics). - */ -void amdgpu_fence_driver_suspend(struct amdgpu_device *adev) -{ - int i, r; - - for (i = 0; i < AMDGPU_MAX_RINGS; i++) { - struct amdgpu_ring *ring = adev->rings[i]; - if (!ring || !ring->fence_drv.initialized) - continue; - - /* wait for gpu to finish processing current batch */ - r = amdgpu_fence_wait_empty(ring); - if (r) { - /* delay GPU reset to resume */ - amdgpu_fence_driver_force_completion(ring); - } - - /* disable the interrupt */ - if (ring->fence_drv.irq_src) - amdgpu_irq_put(adev, ring->fence_drv.irq_src, - ring->fence_drv.irq_type); - } -} - -/** - * amdgpu_fence_driver_resume - resume the fence driver + * amdgpu_fence_driver_hw_init - enable the fence driver * for all possible rings. * * @adev: amdgpu device pointer * - * Resume the fence driver for all possible rings (all asics). + * Enable the fence driver for all possible rings (all asics). * Not all asics have all rings, so each asic will only * start the fence driver on the rings it has using * amdgpu_fence_driver_start_ring(). * Returns 0 for success. */ -void amdgpu_fence_driver_resume(struct amdgpu_device *adev) +void amdgpu_fence_driver_hw_init(struct amdgpu_device *adev) { int i; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_fru_eeprom.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_fru_eeprom.c index 39b6c6bfab45..7709caeb233d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_fru_eeprom.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_fru_eeprom.c @@ -27,10 +27,10 @@ #include "smu_v11_0_i2c.h" #include "atom.h" #include "amdgpu_fru_eeprom.h" +#include "amdgpu_eeprom.h" -#define I2C_PRODUCT_INFO_ADDR 0xAC -#define I2C_PRODUCT_INFO_ADDR_SIZE 0x2 -#define I2C_PRODUCT_INFO_OFFSET 0xC0 +#define FRU_EEPROM_MADDR 0x60000 +#define I2C_PRODUCT_INFO_OFFSET 0xC0 static bool is_fru_eeprom_supported(struct amdgpu_device *adev) { @@ -62,19 +62,11 @@ static bool is_fru_eeprom_supported(struct amdgpu_device *adev) } static int amdgpu_fru_read_eeprom(struct amdgpu_device *adev, uint32_t addrptr, - unsigned char *buff) + unsigned char *buff) { int ret, size; - struct i2c_msg msg = { - .addr = I2C_PRODUCT_INFO_ADDR, - .flags = I2C_M_RD, - .buf = buff, - }; - buff[0] = 0; - buff[1] = addrptr; - msg.len = I2C_PRODUCT_INFO_ADDR_SIZE + 1; - ret = i2c_transfer(&adev->pm.smu_i2c, &msg, 1); + ret = amdgpu_eeprom_read(&adev->pm.smu_i2c, addrptr, buff, 1); if (ret < 1) { DRM_WARN("FRU: Failed to get size field"); return ret; @@ -83,13 +75,9 @@ static int amdgpu_fru_read_eeprom(struct amdgpu_device *adev, uint32_t addrptr, /* The size returned by the i2c requires subtraction of 0xC0 since the * size apparently always reports as 0xC0+actual size. */ - size = buff[2] - I2C_PRODUCT_INFO_OFFSET; - /* Add 1 since address field was 1 byte */ - buff[1] = addrptr + 1; - - msg.len = I2C_PRODUCT_INFO_ADDR_SIZE + size; - ret = i2c_transfer(&adev->pm.smu_i2c, &msg, 1); + size = buff[0] - I2C_PRODUCT_INFO_OFFSET; + ret = amdgpu_eeprom_read(&adev->pm.smu_i2c, addrptr + 1, buff, size); if (ret < 1) { DRM_WARN("FRU: Failed to get data field"); return ret; @@ -101,8 +89,8 @@ static int amdgpu_fru_read_eeprom(struct amdgpu_device *adev, uint32_t addrptr, int amdgpu_fru_get_product_info(struct amdgpu_device *adev) { unsigned char buff[34]; - int addrptr, size; - int len; + u32 addrptr; + int size, len; if (!is_fru_eeprom_supported(adev)) return 0; @@ -125,7 +113,7 @@ int amdgpu_fru_get_product_info(struct amdgpu_device *adev) * Bytes 8-a are all 1-byte and refer to the size of the entire struct, * and the language field, so just start from 0xb, manufacturer size */ - addrptr = 0xb; + addrptr = FRU_EEPROM_MADDR + 0xb; size = amdgpu_fru_read_eeprom(adev, addrptr, buff); if (size < 1) { DRM_ERROR("Failed to read FRU Manufacturer, ret:%d", size); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c index 0d01cfaca77e..7dfdabe1cdf9 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_irq.c @@ -584,7 +584,7 @@ void amdgpu_irq_gpu_reset_resume_helper(struct amdgpu_device *adev) { int i, j, k; - if (amdgpu_sriov_vf(adev)) + if (amdgpu_sriov_vf(adev) || amdgpu_passthrough(adev)) amdgpu_restore_msix(adev); for (i = 0; i < AMDGPU_IRQ_CLIENTID_MAX; ++i) { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c index 96ef3f1051d8..20b049ad61c1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c @@ -374,8 +374,8 @@ static int amdgpu_firmware_info(struct drm_amdgpu_info_firmware *fw_info, fw_info->feature = adev->sdma.instance[query_fw->index].feature_version; break; case AMDGPU_INFO_FW_SOS: - fw_info->ver = adev->psp.sos_fw_version; - fw_info->feature = adev->psp.sos_feature_version; + fw_info->ver = adev->psp.sos.fw_version; + fw_info->feature = adev->psp.sos.feature_version; break; case AMDGPU_INFO_FW_ASD: fw_info->ver = adev->psp.asd_fw_version; @@ -390,8 +390,8 @@ static int amdgpu_firmware_info(struct drm_amdgpu_info_firmware *fw_info, fw_info->feature = 0; break; case AMDGPU_INFO_FW_TOC: - fw_info->ver = adev->psp.toc_fw_version; - fw_info->feature = adev->psp.toc_feature_version; + fw_info->ver = adev->psp.toc.fw_version; + fw_info->feature = adev->psp.toc.feature_version; break; default: return -EINVAL; @@ -1179,10 +1179,14 @@ int amdgpu_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv) pasid = 0; } - r = amdgpu_vm_init(adev, &fpriv->vm, pasid); + r = amdgpu_vm_init(adev, &fpriv->vm); if (r) goto error_pasid; + r = amdgpu_vm_set_pasid(adev, &fpriv->vm, pasid); + if (r) + goto error_vm; + fpriv->prt_va = amdgpu_vm_bo_add(adev, &fpriv->vm, NULL); if (!fpriv->prt_va) { r = -ENOMEM; @@ -1210,8 +1214,10 @@ error_vm: amdgpu_vm_fini(adev, &fpriv->vm); error_pasid: - if (pasid) + if (pasid) { amdgpu_pasid_free(pasid); + amdgpu_vm_set_pasid(adev, &fpriv->vm, 0); + } kfree(fpriv); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_nbio.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_nbio.h index 45295dce5c3e..843052205bd5 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_nbio.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_nbio.h @@ -95,6 +95,7 @@ struct amdgpu_nbio_funcs { void (*program_aspm)(struct amdgpu_device *adev); void (*apply_lc_spc_mode_wa)(struct amdgpu_device *adev); void (*apply_l1_link_width_reconfig_wa)(struct amdgpu_device *adev); + void (*clear_doorbell_interrupt)(struct amdgpu_device *adev); }; struct amdgpu_nbio { diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c index 261283f5b424..d15eee98204d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c @@ -731,7 +731,7 @@ retry: /** * amdgpu_bo_add_to_shadow_list - add a BO to the shadow list * - * @bo: BO that will be inserted into the shadow list + * @vmbo: BO that will be inserted into the shadow list * * Insert a BO to the shadow list. */ diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c index 3ec5099ffeb6..10b046147190 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c @@ -24,7 +24,6 @@ */ #include <linux/firmware.h> -#include <linux/dma-mapping.h> #include <drm/drm_drv.h> #include "amdgpu.h" @@ -34,6 +33,7 @@ #include "psp_v3_1.h" #include "psp_v10_0.h" #include "psp_v11_0.h" +#include "psp_v11_0_8.h" #include "psp_v12_0.h" #include "psp_v13_0.h" @@ -122,6 +122,12 @@ static int psp_early_init(void *handle) psp_v13_0_set_psp_funcs(psp); psp->autoload_supported = true; break; + case CHIP_CYAN_SKILLFISH: + if (adev->apu_flags & AMD_APU_IS_CYAN_SKILLFISH2) { + psp_v11_0_8_set_psp_funcs(psp); + psp->autoload_supported = false; + } + break; default: return -EINVAL; } @@ -356,6 +362,44 @@ int psp_wait_for(struct psp_context *psp, uint32_t reg_index, return -ETIME; } +static const char *psp_gfx_cmd_name(enum psp_gfx_cmd_id cmd_id) +{ + switch (cmd_id) { + case GFX_CMD_ID_LOAD_TA: + return "LOAD_TA"; + case GFX_CMD_ID_UNLOAD_TA: + return "UNLOAD_TA"; + case GFX_CMD_ID_INVOKE_CMD: + return "INVOKE_CMD"; + case GFX_CMD_ID_LOAD_ASD: + return "LOAD_ASD"; + case GFX_CMD_ID_SETUP_TMR: + return "SETUP_TMR"; + case GFX_CMD_ID_LOAD_IP_FW: + return "LOAD_IP_FW"; + case GFX_CMD_ID_DESTROY_TMR: + return "DESTROY_TMR"; + case GFX_CMD_ID_SAVE_RESTORE: + return "SAVE_RESTORE_IP_FW"; + case GFX_CMD_ID_SETUP_VMR: + return "SETUP_VMR"; + case GFX_CMD_ID_DESTROY_VMR: + return "DESTROY_VMR"; + case GFX_CMD_ID_PROG_REG: + return "PROG_REG"; + case GFX_CMD_ID_GET_FW_ATTESTATION: + return "GET_FW_ATTESTATION"; + case GFX_CMD_ID_LOAD_TOC: + return "ID_LOAD_TOC"; + case GFX_CMD_ID_AUTOLOAD_RLC: + return "AUTOLOAD_RLC"; + case GFX_CMD_ID_BOOT_CFG: + return "BOOT_CFG"; + default: + return "UNKNOWN CMD"; + } +} + static int psp_cmd_submit_buf(struct psp_context *psp, struct amdgpu_firmware_info *ucode, @@ -417,10 +461,10 @@ psp_cmd_submit_buf(struct psp_context *psp, */ if (!skip_unsupport && (psp->cmd_buf_mem->resp.status || !timeout) && !ras_intr) { if (ucode) - DRM_WARN("failed to load ucode id (%d) ", - ucode->ucode_id); - DRM_WARN("psp command (0x%X) failed and response status is (0x%X)\n", - psp->cmd_buf_mem->cmd_id, + DRM_WARN("failed to load ucode (%s) ", + amdgpu_ucode_name(ucode->ucode_id)); + DRM_WARN("psp gfx command (%s) failed and response status is (0x%X)\n", + psp_gfx_cmd_name(psp->cmd_buf_mem->cmd_id), psp->cmd_buf_mem->resp.status); if (!timeout) { ret = -EINVAL; @@ -479,9 +523,9 @@ static int psp_load_toc(struct psp_context *psp, if (!cmd) return -ENOMEM; /* Copy toc to psp firmware private buffer */ - psp_copy_fw(psp, psp->toc_start_addr, psp->toc_bin_size); + psp_copy_fw(psp, psp->toc.start_addr, psp->toc.size_bytes); - psp_prep_load_toc_cmd_buf(cmd, psp->fw_pri_mc_addr, psp->toc_bin_size); + psp_prep_load_toc_cmd_buf(cmd, psp->fw_pri_mc_addr, psp->toc.size_bytes); ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr); @@ -511,8 +555,8 @@ static int psp_tmr_init(struct psp_context *psp) /* For ASICs support RLC autoload, psp will parse the toc * and calculate the total size of TMR needed */ if (!amdgpu_sriov_vf(psp->adev) && - psp->toc_start_addr && - psp->toc_bin_size && + psp->toc.start_addr && + psp->toc.size_bytes && psp->fw_pri_buf) { ret = psp_load_toc(psp, &tmr_size); if (ret) { @@ -691,18 +735,18 @@ static int psp_rl_load(struct amdgpu_device *adev) struct psp_context *psp = &adev->psp; struct psp_gfx_cmd_resp *cmd = psp->cmd; - if (psp->rl_bin_size == 0) + if (!is_psp_fw_valid(psp->rl)) return 0; memset(psp->fw_pri_buf, 0, PSP_1_MEG); - memcpy(psp->fw_pri_buf, psp->rl_start_addr, psp->rl_bin_size); + memcpy(psp->fw_pri_buf, psp->rl.start_addr, psp->rl.size_bytes); memset(cmd, 0, sizeof(struct psp_gfx_cmd_resp)); cmd->cmd_id = GFX_CMD_ID_LOAD_IP_FW; cmd->cmd.cmd_load_ip_fw.fw_phy_addr_lo = lower_32_bits(psp->fw_pri_mc_addr); cmd->cmd.cmd_load_ip_fw.fw_phy_addr_hi = upper_32_bits(psp->fw_pri_mc_addr); - cmd->cmd.cmd_load_ip_fw.fw_size = psp->rl_bin_size; + cmd->cmd.cmd_load_ip_fw.fw_size = psp->rl.size_bytes; cmd->cmd.cmd_load_ip_fw.fw_type = GFX_FW_TYPE_REG_LIST; return psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr); @@ -1041,6 +1085,12 @@ int psp_xgmi_get_node_id(struct psp_context *psp, uint64_t *node_id) return 0; } +static bool psp_xgmi_peer_link_info_supported(struct psp_context *psp) +{ + return psp->adev->asic_type == CHIP_ALDEBARAN && + psp->ta_xgmi_ucode_version >= 0x2000000b; +} + int psp_xgmi_get_topology_info(struct psp_context *psp, int number_devices, struct psp_xgmi_topology_info *topology) @@ -1084,6 +1134,23 @@ int psp_xgmi_get_topology_info(struct psp_context *psp, topology->nodes[i].sdma_engine = topology_info_output->nodes[i].sdma_engine; } + /* Invoke xgmi ta again to get the link information */ + if (psp_xgmi_peer_link_info_supported(psp)) { + struct ta_xgmi_cmd_get_peer_link_info_output *link_info_output; + + xgmi_cmd->cmd_id = TA_COMMAND_XGMI__GET_PEER_LINKS; + + ret = psp_xgmi_invoke(psp, TA_COMMAND_XGMI__GET_PEER_LINKS); + + if (ret) + return ret; + + link_info_output = &xgmi_cmd->xgmi_out_message.get_link_info; + for (i = 0; i < topology->num_nodes; i++) + topology->nodes[i].num_links = + link_info_output->nodes[i].num_links; + } + return 0; } @@ -2069,7 +2136,7 @@ static int psp_hw_start(struct psp_context *psp) int ret; if (!amdgpu_sriov_vf(adev)) { - if (psp->kdb_bin_size && + if ((is_psp_fw_valid(psp->kdb)) && (psp->funcs->bootloader_load_kdb != NULL)) { ret = psp_bootloader_load_kdb(psp); if (ret) { @@ -2078,7 +2145,8 @@ static int psp_hw_start(struct psp_context *psp) } } - if (psp->spl_bin_size) { + if ((is_psp_fw_valid(psp->spl)) && + (psp->funcs->bootloader_load_spl != NULL)) { ret = psp_bootloader_load_spl(psp); if (ret) { DRM_ERROR("PSP load spl failed!\n"); @@ -2086,16 +2154,49 @@ static int psp_hw_start(struct psp_context *psp) } } - ret = psp_bootloader_load_sysdrv(psp); - if (ret) { - DRM_ERROR("PSP load sysdrv failed!\n"); - return ret; + if ((is_psp_fw_valid(psp->sys)) && + (psp->funcs->bootloader_load_sysdrv != NULL)) { + ret = psp_bootloader_load_sysdrv(psp); + if (ret) { + DRM_ERROR("PSP load sys drv failed!\n"); + return ret; + } } - ret = psp_bootloader_load_sos(psp); - if (ret) { - DRM_ERROR("PSP load sos failed!\n"); - return ret; + if ((is_psp_fw_valid(psp->soc_drv)) && + (psp->funcs->bootloader_load_soc_drv != NULL)) { + ret = psp_bootloader_load_soc_drv(psp); + if (ret) { + DRM_ERROR("PSP load soc drv failed!\n"); + return ret; + } + } + + if ((is_psp_fw_valid(psp->intf_drv)) && + (psp->funcs->bootloader_load_intf_drv != NULL)) { + ret = psp_bootloader_load_intf_drv(psp); + if (ret) { + DRM_ERROR("PSP load intf drv failed!\n"); + return ret; + } + } + + if ((is_psp_fw_valid(psp->dbg_drv)) && + (psp->funcs->bootloader_load_dbg_drv != NULL)) { + ret = psp_bootloader_load_dbg_drv(psp); + if (ret) { + DRM_ERROR("PSP load dbg drv failed!\n"); + return ret; + } + } + + if ((is_psp_fw_valid(psp->sos)) && + (psp->funcs->bootloader_load_sos != NULL)) { + ret = psp_bootloader_load_sos(psp); + if (ret) { + DRM_ERROR("PSP load sos failed!\n"); + return ret; + } } } @@ -2313,7 +2414,7 @@ static int psp_prep_load_ip_fw_cmd_buf(struct amdgpu_firmware_info *ucode, return ret; } -static int psp_execute_np_fw_load(struct psp_context *psp, +static int psp_execute_non_psp_fw_load(struct psp_context *psp, struct amdgpu_firmware_info *ucode) { int ret = 0; @@ -2349,7 +2450,7 @@ static int psp_load_smu_fw(struct psp_context *psp) } } - ret = psp_execute_np_fw_load(psp, ucode); + ret = psp_execute_non_psp_fw_load(psp, ucode); if (ret) DRM_ERROR("PSP load smu failed!\n"); @@ -2404,14 +2505,14 @@ int psp_load_fw_list(struct psp_context *psp, for (i = 0; i < ucode_count; ++i) { ucode = ucode_list[i]; psp_print_fw_hdr(psp, ucode); - ret = psp_execute_np_fw_load(psp, ucode); + ret = psp_execute_non_psp_fw_load(psp, ucode); if (ret) return ret; } return ret; } -static int psp_np_fw_load(struct psp_context *psp) +static int psp_load_non_psp_fw(struct psp_context *psp) { int i, ret; struct amdgpu_firmware_info *ucode; @@ -2450,7 +2551,7 @@ static int psp_np_fw_load(struct psp_context *psp) psp_print_fw_hdr(psp, ucode); - ret = psp_execute_np_fw_load(psp, ucode); + ret = psp_execute_non_psp_fw_load(psp, ucode); if (ret) return ret; @@ -2527,7 +2628,7 @@ skip_memalloc: if (ret) goto failed; - ret = psp_np_fw_load(psp); + ret = psp_load_non_psp_fw(psp); if (ret) goto failed; @@ -2727,7 +2828,7 @@ static int psp_resume(void *handle) if (ret) goto failed; - ret = psp_np_fw_load(psp); + ret = psp_load_non_psp_fw(psp); if (ret) goto failed; @@ -2825,7 +2926,7 @@ int psp_update_vcn_sram(struct amdgpu_device *adev, int inst_idx, ucode.mc_addr = cmd_gpu_addr; ucode.ucode_size = cmd_size; - return psp_execute_np_fw_load(&adev->psp, &ucode); + return psp_execute_non_psp_fw_load(&adev->psp, &ucode); } int psp_ring_cmd_submit(struct psp_context *psp, @@ -2918,7 +3019,7 @@ int psp_init_toc_microcode(struct psp_context *psp, const char *chip_name) { struct amdgpu_device *adev = psp->adev; - char fw_name[30]; + char fw_name[PSP_FW_NAME_LEN]; const struct psp_firmware_header_v1_0 *toc_hdr; int err = 0; @@ -2937,10 +3038,10 @@ int psp_init_toc_microcode(struct psp_context *psp, goto out; toc_hdr = (const struct psp_firmware_header_v1_0 *)adev->psp.toc_fw->data; - adev->psp.toc_fw_version = le32_to_cpu(toc_hdr->header.ucode_version); - adev->psp.toc_feature_version = le32_to_cpu(toc_hdr->sos.fw_version); - adev->psp.toc_bin_size = le32_to_cpu(toc_hdr->header.ucode_size_bytes); - adev->psp.toc_start_addr = (uint8_t *)toc_hdr + + adev->psp.toc.fw_version = le32_to_cpu(toc_hdr->header.ucode_version); + adev->psp.toc.feature_version = le32_to_cpu(toc_hdr->sos.fw_version); + adev->psp.toc.size_bytes = le32_to_cpu(toc_hdr->header.ucode_size_bytes); + adev->psp.toc.start_addr = (uint8_t *)toc_hdr + le32_to_cpu(toc_hdr->header.ucode_array_offset_bytes); return 0; out: @@ -2950,6 +3051,82 @@ out: return err; } +static int parse_sos_bin_descriptor(struct psp_context *psp, + const struct psp_fw_bin_desc *desc, + const struct psp_firmware_header_v2_0 *sos_hdr) +{ + uint8_t *ucode_start_addr = NULL; + + if (!psp || !desc || !sos_hdr) + return -EINVAL; + + ucode_start_addr = (uint8_t *)sos_hdr + + le32_to_cpu(desc->offset_bytes) + + le32_to_cpu(sos_hdr->header.ucode_array_offset_bytes); + + switch (desc->fw_type) { + case PSP_FW_TYPE_PSP_SOS: + psp->sos.fw_version = le32_to_cpu(desc->fw_version); + psp->sos.feature_version = le32_to_cpu(desc->fw_version); + psp->sos.size_bytes = le32_to_cpu(desc->size_bytes); + psp->sos.start_addr = ucode_start_addr; + break; + case PSP_FW_TYPE_PSP_SYS_DRV: + psp->sys.fw_version = le32_to_cpu(desc->fw_version); + psp->sys.feature_version = le32_to_cpu(desc->fw_version); + psp->sys.size_bytes = le32_to_cpu(desc->size_bytes); + psp->sys.start_addr = ucode_start_addr; + break; + case PSP_FW_TYPE_PSP_KDB: + psp->kdb.fw_version = le32_to_cpu(desc->fw_version); + psp->kdb.feature_version = le32_to_cpu(desc->fw_version); + psp->kdb.size_bytes = le32_to_cpu(desc->size_bytes); + psp->kdb.start_addr = ucode_start_addr; + break; + case PSP_FW_TYPE_PSP_TOC: + psp->toc.fw_version = le32_to_cpu(desc->fw_version); + psp->toc.feature_version = le32_to_cpu(desc->fw_version); + psp->toc.size_bytes = le32_to_cpu(desc->size_bytes); + psp->toc.start_addr = ucode_start_addr; + break; + case PSP_FW_TYPE_PSP_SPL: + psp->spl.fw_version = le32_to_cpu(desc->fw_version); + psp->spl.feature_version = le32_to_cpu(desc->fw_version); + psp->spl.size_bytes = le32_to_cpu(desc->size_bytes); + psp->spl.start_addr = ucode_start_addr; + break; + case PSP_FW_TYPE_PSP_RL: + psp->rl.fw_version = le32_to_cpu(desc->fw_version); + psp->rl.feature_version = le32_to_cpu(desc->fw_version); + psp->rl.size_bytes = le32_to_cpu(desc->size_bytes); + psp->rl.start_addr = ucode_start_addr; + break; + case PSP_FW_TYPE_PSP_SOC_DRV: + psp->soc_drv.fw_version = le32_to_cpu(desc->fw_version); + psp->soc_drv.feature_version = le32_to_cpu(desc->fw_version); + psp->soc_drv.size_bytes = le32_to_cpu(desc->size_bytes); + psp->soc_drv.start_addr = ucode_start_addr; + break; + case PSP_FW_TYPE_PSP_INTF_DRV: + psp->intf_drv.fw_version = le32_to_cpu(desc->fw_version); + psp->intf_drv.feature_version = le32_to_cpu(desc->fw_version); + psp->intf_drv.size_bytes = le32_to_cpu(desc->size_bytes); + psp->intf_drv.start_addr = ucode_start_addr; + break; + case PSP_FW_TYPE_PSP_DBG_DRV: + psp->dbg_drv.fw_version = le32_to_cpu(desc->fw_version); + psp->dbg_drv.feature_version = le32_to_cpu(desc->fw_version); + psp->dbg_drv.size_bytes = le32_to_cpu(desc->size_bytes); + psp->dbg_drv.start_addr = ucode_start_addr; + break; + default: + dev_warn(psp->adev->dev, "Unsupported PSP FW type: %d\n", desc->fw_type); + break; + } + + return 0; +} + static int psp_init_sos_base_fw(struct amdgpu_device *adev) { const struct psp_firmware_header_v1_0 *sos_hdr; @@ -2961,32 +3138,32 @@ static int psp_init_sos_base_fw(struct amdgpu_device *adev) le32_to_cpu(sos_hdr->header.ucode_array_offset_bytes); if (adev->gmc.xgmi.connected_to_cpu || (adev->asic_type != CHIP_ALDEBARAN)) { - adev->psp.sos_fw_version = le32_to_cpu(sos_hdr->header.ucode_version); - adev->psp.sos_feature_version = le32_to_cpu(sos_hdr->sos.fw_version); + adev->psp.sos.fw_version = le32_to_cpu(sos_hdr->header.ucode_version); + adev->psp.sos.feature_version = le32_to_cpu(sos_hdr->sos.fw_version); - adev->psp.sys_bin_size = le32_to_cpu(sos_hdr->sos.offset_bytes); - adev->psp.sys_start_addr = ucode_array_start_addr; + adev->psp.sys.size_bytes = le32_to_cpu(sos_hdr->sos.offset_bytes); + adev->psp.sys.start_addr = ucode_array_start_addr; - adev->psp.sos_bin_size = le32_to_cpu(sos_hdr->sos.size_bytes); - adev->psp.sos_start_addr = ucode_array_start_addr + + adev->psp.sos.size_bytes = le32_to_cpu(sos_hdr->sos.size_bytes); + adev->psp.sos.start_addr = ucode_array_start_addr + le32_to_cpu(sos_hdr->sos.offset_bytes); } else { /* Load alternate PSP SOS FW */ sos_hdr_v1_3 = (const struct psp_firmware_header_v1_3 *)adev->psp.sos_fw->data; - adev->psp.sos_fw_version = le32_to_cpu(sos_hdr_v1_3->sos_aux.fw_version); - adev->psp.sos_feature_version = le32_to_cpu(sos_hdr_v1_3->sos_aux.fw_version); + adev->psp.sos.fw_version = le32_to_cpu(sos_hdr_v1_3->sos_aux.fw_version); + adev->psp.sos.feature_version = le32_to_cpu(sos_hdr_v1_3->sos_aux.fw_version); - adev->psp.sys_bin_size = le32_to_cpu(sos_hdr_v1_3->sys_drv_aux.size_bytes); - adev->psp.sys_start_addr = ucode_array_start_addr + + adev->psp.sys.size_bytes = le32_to_cpu(sos_hdr_v1_3->sys_drv_aux.size_bytes); + adev->psp.sys.start_addr = ucode_array_start_addr + le32_to_cpu(sos_hdr_v1_3->sys_drv_aux.offset_bytes); - adev->psp.sos_bin_size = le32_to_cpu(sos_hdr_v1_3->sos_aux.size_bytes); - adev->psp.sos_start_addr = ucode_array_start_addr + + adev->psp.sos.size_bytes = le32_to_cpu(sos_hdr_v1_3->sos_aux.size_bytes); + adev->psp.sos.start_addr = ucode_array_start_addr + le32_to_cpu(sos_hdr_v1_3->sos_aux.offset_bytes); } - if ((adev->psp.sys_bin_size == 0) || (adev->psp.sos_bin_size == 0)) { + if ((adev->psp.sys.size_bytes == 0) || (adev->psp.sos.size_bytes == 0)) { dev_warn(adev->dev, "PSP SOS FW not available"); return -EINVAL; } @@ -3003,8 +3180,10 @@ int psp_init_sos_microcode(struct psp_context *psp, const struct psp_firmware_header_v1_1 *sos_hdr_v1_1; const struct psp_firmware_header_v1_2 *sos_hdr_v1_2; const struct psp_firmware_header_v1_3 *sos_hdr_v1_3; + const struct psp_firmware_header_v2_0 *sos_hdr_v2_0; int err = 0; uint8_t *ucode_array_start_addr; + int fw_index = 0; if (!chip_name) { dev_err(adev->dev, "invalid chip name for sos microcode\n"); @@ -3033,35 +3212,52 @@ int psp_init_sos_microcode(struct psp_context *psp, if (sos_hdr->header.header_version_minor == 1) { sos_hdr_v1_1 = (const struct psp_firmware_header_v1_1 *)adev->psp.sos_fw->data; - adev->psp.toc_bin_size = le32_to_cpu(sos_hdr_v1_1->toc.size_bytes); - adev->psp.toc_start_addr = (uint8_t *)adev->psp.sys_start_addr + + adev->psp.toc.size_bytes = le32_to_cpu(sos_hdr_v1_1->toc.size_bytes); + adev->psp.toc.start_addr = (uint8_t *)adev->psp.sys.start_addr + le32_to_cpu(sos_hdr_v1_1->toc.offset_bytes); - adev->psp.kdb_bin_size = le32_to_cpu(sos_hdr_v1_1->kdb.size_bytes); - adev->psp.kdb_start_addr = (uint8_t *)adev->psp.sys_start_addr + + adev->psp.kdb.size_bytes = le32_to_cpu(sos_hdr_v1_1->kdb.size_bytes); + adev->psp.kdb.start_addr = (uint8_t *)adev->psp.sys.start_addr + le32_to_cpu(sos_hdr_v1_1->kdb.offset_bytes); } if (sos_hdr->header.header_version_minor == 2) { sos_hdr_v1_2 = (const struct psp_firmware_header_v1_2 *)adev->psp.sos_fw->data; - adev->psp.kdb_bin_size = le32_to_cpu(sos_hdr_v1_2->kdb.size_bytes); - adev->psp.kdb_start_addr = (uint8_t *)adev->psp.sys_start_addr + + adev->psp.kdb.size_bytes = le32_to_cpu(sos_hdr_v1_2->kdb.size_bytes); + adev->psp.kdb.start_addr = (uint8_t *)adev->psp.sys.start_addr + le32_to_cpu(sos_hdr_v1_2->kdb.offset_bytes); } if (sos_hdr->header.header_version_minor == 3) { sos_hdr_v1_3 = (const struct psp_firmware_header_v1_3 *)adev->psp.sos_fw->data; - adev->psp.toc_bin_size = le32_to_cpu(sos_hdr_v1_3->v1_1.toc.size_bytes); - adev->psp.toc_start_addr = ucode_array_start_addr + + adev->psp.toc.size_bytes = le32_to_cpu(sos_hdr_v1_3->v1_1.toc.size_bytes); + adev->psp.toc.start_addr = ucode_array_start_addr + le32_to_cpu(sos_hdr_v1_3->v1_1.toc.offset_bytes); - adev->psp.kdb_bin_size = le32_to_cpu(sos_hdr_v1_3->v1_1.kdb.size_bytes); - adev->psp.kdb_start_addr = ucode_array_start_addr + + adev->psp.kdb.size_bytes = le32_to_cpu(sos_hdr_v1_3->v1_1.kdb.size_bytes); + adev->psp.kdb.start_addr = ucode_array_start_addr + le32_to_cpu(sos_hdr_v1_3->v1_1.kdb.offset_bytes); - adev->psp.spl_bin_size = le32_to_cpu(sos_hdr_v1_3->spl.size_bytes); - adev->psp.spl_start_addr = ucode_array_start_addr + + adev->psp.spl.size_bytes = le32_to_cpu(sos_hdr_v1_3->spl.size_bytes); + adev->psp.spl.start_addr = ucode_array_start_addr + le32_to_cpu(sos_hdr_v1_3->spl.offset_bytes); - adev->psp.rl_bin_size = le32_to_cpu(sos_hdr_v1_3->rl.size_bytes); - adev->psp.rl_start_addr = ucode_array_start_addr + + adev->psp.rl.size_bytes = le32_to_cpu(sos_hdr_v1_3->rl.size_bytes); + adev->psp.rl.start_addr = ucode_array_start_addr + le32_to_cpu(sos_hdr_v1_3->rl.offset_bytes); } break; + case 2: + sos_hdr_v2_0 = (const struct psp_firmware_header_v2_0 *)adev->psp.sos_fw->data; + + if (le32_to_cpu(sos_hdr_v2_0->psp_fw_bin_count) >= UCODE_MAX_PSP_PACKAGING) { + dev_err(adev->dev, "packed SOS count exceeds maximum limit\n"); + err = -EINVAL; + goto out; + } + + for (fw_index = 0; fw_index < le32_to_cpu(sos_hdr_v2_0->psp_fw_bin_count); fw_index++) { + err = parse_sos_bin_descriptor(psp, + &sos_hdr_v2_0->psp_fw_bin[fw_index], + sos_hdr_v2_0); + if (err) + goto out; + } + break; default: dev_err(adev->dev, "unsupported psp sos firmware\n"); @@ -3080,7 +3276,7 @@ out: } static int parse_ta_bin_descriptor(struct psp_context *psp, - const struct ta_fw_bin_desc *desc, + const struct psp_fw_bin_desc *desc, const struct ta_firmware_header_v2_0 *ta_hdr) { uint8_t *ucode_start_addr = NULL; @@ -3168,7 +3364,7 @@ int psp_init_ta_microcode(struct psp_context *psp, goto out; } - if (le32_to_cpu(ta_hdr->ta_fw_bin_count) >= UCODE_MAX_TA_PACKAGING) { + if (le32_to_cpu(ta_hdr->ta_fw_bin_count) >= UCODE_MAX_PSP_PACKAGING) { dev_err(adev->dev, "packed TA count exceeds maximum limit\n"); err = -EINVAL; goto out; @@ -3235,11 +3431,12 @@ static ssize_t psp_usbc_pd_fw_sysfs_write(struct device *dev, { struct drm_device *ddev = dev_get_drvdata(dev); struct amdgpu_device *adev = drm_to_adev(ddev); - void *cpu_addr; - dma_addr_t dma_addr; int ret, idx; char fw_name[100]; const struct firmware *usbc_pd_fw; + struct amdgpu_bo *fw_buf_bo = NULL; + uint64_t fw_pri_mc_addr; + void *fw_pri_cpu_addr; if (!adev->ip_blocks[AMD_IP_BLOCK_TYPE_PSP].status.late_initialized) { DRM_INFO("PSP block is not ready yet."); @@ -3254,31 +3451,24 @@ static ssize_t psp_usbc_pd_fw_sysfs_write(struct device *dev, if (ret) goto fail; - /* We need contiguous physical mem to place the FW for psp to access */ - cpu_addr = dma_alloc_coherent(adev->dev, usbc_pd_fw->size, &dma_addr, GFP_KERNEL); - - ret = dma_mapping_error(adev->dev, dma_addr); + /* LFB address which is aligned to 1MB boundary per PSP request */ + ret = amdgpu_bo_create_kernel(adev, usbc_pd_fw->size, 0x100000, + AMDGPU_GEM_DOMAIN_VRAM, + &fw_buf_bo, + &fw_pri_mc_addr, + &fw_pri_cpu_addr); if (ret) goto rel_buf; - memcpy_toio(cpu_addr, usbc_pd_fw->data, usbc_pd_fw->size); - - /* - * x86 specific workaround. - * Without it the buffer is invisible in PSP. - * - * TODO Remove once PSP starts snooping CPU cache - */ -#ifdef CONFIG_X86 - clflush_cache_range(cpu_addr, (usbc_pd_fw->size & ~(L1_CACHE_BYTES - 1))); -#endif + memcpy_toio(fw_pri_cpu_addr, usbc_pd_fw->data, usbc_pd_fw->size); mutex_lock(&adev->psp.mutex); - ret = psp_load_usbc_pd_fw(&adev->psp, dma_addr); + ret = psp_load_usbc_pd_fw(&adev->psp, fw_pri_mc_addr); mutex_unlock(&adev->psp.mutex); + amdgpu_bo_free_kernel(&fw_buf_bo, &fw_pri_mc_addr, &fw_pri_cpu_addr); + rel_buf: - dma_free_coherent(adev->dev, usbc_pd_fw->size, cpu_addr, dma_addr); release_firmware(usbc_pd_fw); fail: if (ret) { @@ -3307,7 +3497,10 @@ static DEVICE_ATTR(usbc_pd_fw, S_IRUGO | S_IWUSR, psp_usbc_pd_fw_sysfs_read, psp_usbc_pd_fw_sysfs_write); - +int is_psp_fw_valid(struct psp_bin_desc bin) +{ + return bin.size_bytes; +} const struct amd_ip_funcs psp_ip_funcs = { .name = "psp", @@ -3369,6 +3562,14 @@ const struct amdgpu_ip_block_version psp_v11_0_ip_block = .funcs = &psp_ip_funcs, }; +const struct amdgpu_ip_block_version psp_v11_0_8_ip_block = { + .type = AMD_IP_BLOCK_TYPE_PSP, + .major = 11, + .minor = 0, + .rev = 8, + .funcs = &psp_ip_funcs, +}; + const struct amdgpu_ip_block_version psp_v12_0_ip_block = { .type = AMD_IP_BLOCK_TYPE_PSP, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h index 3030ec24eb3b..4435b180ace0 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h @@ -48,11 +48,15 @@ struct psp_context; struct psp_xgmi_node_info; struct psp_xgmi_topology_info; +struct psp_bin_desc; enum psp_bootloader_cmd { PSP_BL__LOAD_SYSDRV = 0x10000, PSP_BL__LOAD_SOSDRV = 0x20000, PSP_BL__LOAD_KEY_DATABASE = 0x80000, + PSP_BL__LOAD_SOCDRV = 0x90000, + PSP_BL__LOAD_INTFDRV = 0xA0000, + PSP_BL__LOAD_DBGDRV = 0xB0000, PSP_BL__DRAM_LONG_TRAIN = 0x100000, PSP_BL__DRAM_SHORT_TRAIN = 0x200000, PSP_BL__LOAD_TOS_SPL_TABLE = 0x10000000, @@ -93,6 +97,9 @@ struct psp_funcs int (*bootloader_load_kdb)(struct psp_context *psp); int (*bootloader_load_spl)(struct psp_context *psp); int (*bootloader_load_sysdrv)(struct psp_context *psp); + int (*bootloader_load_soc_drv)(struct psp_context *psp); + int (*bootloader_load_intf_drv)(struct psp_context *psp); + int (*bootloader_load_dbg_drv)(struct psp_context *psp); int (*bootloader_load_sos)(struct psp_context *psp); int (*ring_init)(struct psp_context *psp, enum psp_ring_type ring_type); int (*ring_create)(struct psp_context *psp, @@ -106,7 +113,7 @@ struct psp_funcs int (*mem_training)(struct psp_context *psp, uint32_t ops); uint32_t (*ring_get_wptr)(struct psp_context *psp); void (*ring_set_wptr)(struct psp_context *psp, uint32_t value); - int (*load_usbc_pd_fw)(struct psp_context *psp, dma_addr_t dma_addr); + int (*load_usbc_pd_fw)(struct psp_context *psp, uint64_t fw_pri_mc_addr); int (*read_usbc_pd_fw)(struct psp_context *psp, uint32_t *fw_ver); }; @@ -116,6 +123,7 @@ struct psp_xgmi_node_info { uint8_t num_hops; uint8_t is_sharing_enabled; enum ta_xgmi_assigned_sdma_engine sdma_engine; + uint8_t num_links; }; struct psp_xgmi_topology_info { @@ -282,6 +290,13 @@ struct psp_runtime_boot_cfg_entry { uint32_t reserved; }; +struct psp_bin_desc { + uint32_t fw_version; + uint32_t feature_version; + uint32_t size_bytes; + uint8_t *start_addr; +}; + struct psp_context { struct amdgpu_device *adev; @@ -297,20 +312,15 @@ struct psp_context /* sos firmware */ const struct firmware *sos_fw; - uint32_t sos_fw_version; - uint32_t sos_feature_version; - uint32_t sys_bin_size; - uint32_t sos_bin_size; - uint32_t toc_bin_size; - uint32_t kdb_bin_size; - uint32_t spl_bin_size; - uint32_t rl_bin_size; - uint8_t *sys_start_addr; - uint8_t *sos_start_addr; - uint8_t *toc_start_addr; - uint8_t *kdb_start_addr; - uint8_t *spl_start_addr; - uint8_t *rl_start_addr; + struct psp_bin_desc sys; + struct psp_bin_desc sos; + struct psp_bin_desc toc; + struct psp_bin_desc kdb; + struct psp_bin_desc spl; + struct psp_bin_desc rl; + struct psp_bin_desc soc_drv; + struct psp_bin_desc intf_drv; + struct psp_bin_desc dbg_drv; /* tmr buffer */ struct amdgpu_bo *tmr_bo; @@ -325,8 +335,6 @@ struct psp_context /* toc firmware */ const struct firmware *toc_fw; - uint32_t toc_fw_version; - uint32_t toc_feature_version; /* fence buffer */ struct amdgpu_bo *fence_buf_bo; @@ -402,6 +410,12 @@ struct amdgpu_psp_funcs { ((psp)->funcs->bootloader_load_spl ? (psp)->funcs->bootloader_load_spl((psp)) : 0) #define psp_bootloader_load_sysdrv(psp) \ ((psp)->funcs->bootloader_load_sysdrv ? (psp)->funcs->bootloader_load_sysdrv((psp)) : 0) +#define psp_bootloader_load_soc_drv(psp) \ + ((psp)->funcs->bootloader_load_soc_drv ? (psp)->funcs->bootloader_load_soc_drv((psp)) : 0) +#define psp_bootloader_load_intf_drv(psp) \ + ((psp)->funcs->bootloader_load_intf_drv ? (psp)->funcs->bootloader_load_intf_drv((psp)) : 0) +#define psp_bootloader_load_dbg_drv(psp) \ + ((psp)->funcs->bootloader_load_dbg_drv ? (psp)->funcs->bootloader_load_dbg_drv((psp)) : 0) #define psp_bootloader_load_sos(psp) \ ((psp)->funcs->bootloader_load_sos ? (psp)->funcs->bootloader_load_sos((psp)) : 0) #define psp_smu_reload_quirk(psp) \ @@ -414,9 +428,9 @@ struct amdgpu_psp_funcs { #define psp_ring_get_wptr(psp) (psp)->funcs->ring_get_wptr((psp)) #define psp_ring_set_wptr(psp, value) (psp)->funcs->ring_set_wptr((psp), (value)) -#define psp_load_usbc_pd_fw(psp, dma_addr) \ +#define psp_load_usbc_pd_fw(psp, fw_pri_mc_addr) \ ((psp)->funcs->load_usbc_pd_fw ? \ - (psp)->funcs->load_usbc_pd_fw((psp), (dma_addr)) : -EINVAL) + (psp)->funcs->load_usbc_pd_fw((psp), (fw_pri_mc_addr)) : -EINVAL) #define psp_read_usbc_pd_fw(psp, fw_ver) \ ((psp)->funcs->read_usbc_pd_fw ? \ @@ -427,6 +441,7 @@ extern const struct amd_ip_funcs psp_ip_funcs; extern const struct amdgpu_ip_block_version psp_v3_1_ip_block; extern const struct amdgpu_ip_block_version psp_v10_0_ip_block; extern const struct amdgpu_ip_block_version psp_v11_0_ip_block; +extern const struct amdgpu_ip_block_version psp_v11_0_8_ip_block; extern const struct amdgpu_ip_block_version psp_v12_0_ip_block; extern const struct amdgpu_ip_block_version psp_v13_0_ip_block; @@ -483,4 +498,5 @@ int psp_load_fw_list(struct psp_context *psp, struct amdgpu_firmware_info **ucode_list, int ucode_count); void psp_copy_fw(struct psp_context *psp, uint8_t *start_addr, uint32_t bin_size); +int is_psp_fw_valid(struct psp_bin_desc bin); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c index fc66aca28594..194f7ccfbf94 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c @@ -71,8 +71,8 @@ const char *ras_block_string[] = { /* inject address is 52 bits */ #define RAS_UMC_INJECT_ADDR_LIMIT (0x1ULL << 52) -/* typical ECC bad page rate(1 bad page per 100MB VRAM) */ -#define RAS_BAD_PAGE_RATE (100 * 1024 * 1024ULL) +/* typical ECC bad page rate is 1 bad page per 100MB VRAM */ +#define RAS_BAD_PAGE_COVER (100 * 1024 * 1024ULL) enum amdgpu_ras_retire_page_reservation { AMDGPU_RAS_RETIRE_PAGE_RESERVED, @@ -355,8 +355,9 @@ static int amdgpu_ras_debugfs_ctrl_parse_data(struct file *f, * to see which blocks support RAS on a particular asic. * */ -static ssize_t amdgpu_ras_debugfs_ctrl_write(struct file *f, const char __user *buf, - size_t size, loff_t *pos) +static ssize_t amdgpu_ras_debugfs_ctrl_write(struct file *f, + const char __user *buf, + size_t size, loff_t *pos) { struct amdgpu_device *adev = (struct amdgpu_device *)file_inode(f)->i_private; struct ras_debug_if data; @@ -370,7 +371,7 @@ static ssize_t amdgpu_ras_debugfs_ctrl_write(struct file *f, const char __user * ret = amdgpu_ras_debugfs_ctrl_parse_data(f, buf, size, pos, &data); if (ret) - return -EINVAL; + return ret; if (data.op == 3) { ret = amdgpu_reserve_page_direct(adev, data.inject.address); @@ -403,9 +404,9 @@ static ssize_t amdgpu_ras_debugfs_ctrl_write(struct file *f, const char __user * /* umc ce/ue error injection for a bad page is not allowed */ if ((data.head.block == AMDGPU_RAS_BLOCK__UMC) && amdgpu_ras_check_bad_page(adev, data.inject.address)) { - dev_warn(adev->dev, "RAS WARN: 0x%llx has been marked " - "as bad before error injection!\n", - data.inject.address); + dev_warn(adev->dev, "RAS WARN: inject: 0x%llx has " + "already been marked as bad!\n", + data.inject.address); break; } @@ -439,21 +440,24 @@ static ssize_t amdgpu_ras_debugfs_ctrl_write(struct file *f, const char __user * * will reset EEPROM table to 0 entries. * */ -static ssize_t amdgpu_ras_debugfs_eeprom_write(struct file *f, const char __user *buf, - size_t size, loff_t *pos) +static ssize_t amdgpu_ras_debugfs_eeprom_write(struct file *f, + const char __user *buf, + size_t size, loff_t *pos) { struct amdgpu_device *adev = (struct amdgpu_device *)file_inode(f)->i_private; int ret; ret = amdgpu_ras_eeprom_reset_table( - &(amdgpu_ras_get_context(adev)->eeprom_control)); + &(amdgpu_ras_get_context(adev)->eeprom_control)); - if (ret == 1) { + if (!ret) { + /* Something was written to EEPROM. + */ amdgpu_ras_get_context(adev)->flags = RAS_DEFAULT_FLAGS; return size; } else { - return -EIO; + return ret; } } @@ -1316,6 +1320,12 @@ static struct dentry *amdgpu_ras_debugfs_create_ctrl_node(struct amdgpu_device * &con->bad_page_cnt_threshold); debugfs_create_x32("ras_hw_enabled", 0444, dir, &adev->ras_hw_enabled); debugfs_create_x32("ras_enabled", 0444, dir, &adev->ras_enabled); + debugfs_create_file("ras_eeprom_size", S_IRUGO, dir, adev, + &amdgpu_ras_debugfs_eeprom_size_ops); + con->de_ras_eeprom_table = debugfs_create_file("ras_eeprom_table", + S_IRUGO, dir, adev, + &amdgpu_ras_debugfs_eeprom_table_ops); + amdgpu_ras_debugfs_set_ret_size(&con->eeprom_control); /* * After one uncorrectable error happens, usually GPU recovery will @@ -1833,13 +1843,12 @@ int amdgpu_ras_save_bad_pages(struct amdgpu_device *adev) control = &con->eeprom_control; data = con->eh_data; - save_count = data->count - control->num_recs; + save_count = data->count - control->ras_num_recs; /* only new entries are saved */ if (save_count > 0) { - if (amdgpu_ras_eeprom_process_recods(control, - &data->bps[control->num_recs], - true, - save_count)) { + if (amdgpu_ras_eeprom_append(control, + &data->bps[control->ras_num_recs], + save_count)) { dev_err(adev->dev, "Failed to save EEPROM table data!"); return -EIO; } @@ -1857,28 +1866,24 @@ int amdgpu_ras_save_bad_pages(struct amdgpu_device *adev) static int amdgpu_ras_load_bad_pages(struct amdgpu_device *adev) { struct amdgpu_ras_eeprom_control *control = - &adev->psp.ras.ras->eeprom_control; - struct eeprom_table_record *bps = NULL; - int ret = 0; + &adev->psp.ras.ras->eeprom_control; + struct eeprom_table_record *bps; + int ret; /* no bad page record, skip eeprom access */ - if (!control->num_recs || (amdgpu_bad_page_threshold == 0)) - return ret; + if (control->ras_num_recs == 0 || amdgpu_bad_page_threshold == 0) + return 0; - bps = kcalloc(control->num_recs, sizeof(*bps), GFP_KERNEL); + bps = kcalloc(control->ras_num_recs, sizeof(*bps), GFP_KERNEL); if (!bps) return -ENOMEM; - if (amdgpu_ras_eeprom_process_recods(control, bps, false, - control->num_recs)) { + ret = amdgpu_ras_eeprom_read(control, bps, control->ras_num_recs); + if (ret) dev_err(adev->dev, "Failed to load EEPROM table records!"); - ret = -EIO; - goto out; - } - - ret = amdgpu_ras_add_bad_pages(adev, bps, control->num_recs); + else + ret = amdgpu_ras_add_bad_pages(adev, bps, control->ras_num_recs); -out: kfree(bps); return ret; } @@ -1918,11 +1923,9 @@ static bool amdgpu_ras_check_bad_page(struct amdgpu_device *adev, } static void amdgpu_ras_validate_threshold(struct amdgpu_device *adev, - uint32_t max_length) + uint32_t max_count) { struct amdgpu_ras *con = amdgpu_ras_get_context(adev); - int tmp_threshold = amdgpu_bad_page_threshold; - u64 val; /* * Justification of value bad_page_cnt_threshold in ras structure @@ -1943,18 +1946,15 @@ static void amdgpu_ras_validate_threshold(struct amdgpu_device *adev, * take no effect. */ - if (tmp_threshold < -1) - tmp_threshold = -1; - else if (tmp_threshold > max_length) - tmp_threshold = max_length; + if (amdgpu_bad_page_threshold < 0) { + u64 val = adev->gmc.mc_vram_size; - if (tmp_threshold == -1) { - val = adev->gmc.mc_vram_size; - do_div(val, RAS_BAD_PAGE_RATE); + do_div(val, RAS_BAD_PAGE_COVER); con->bad_page_cnt_threshold = min(lower_32_bits(val), - max_length); + max_count); } else { - con->bad_page_cnt_threshold = tmp_threshold; + con->bad_page_cnt_threshold = min_t(int, max_count, + amdgpu_bad_page_threshold); } } @@ -1962,15 +1962,24 @@ int amdgpu_ras_recovery_init(struct amdgpu_device *adev) { struct amdgpu_ras *con = amdgpu_ras_get_context(adev); struct ras_err_handler_data **data; - uint32_t max_eeprom_records_len = 0; + u32 max_eeprom_records_count = 0; bool exc_err_limit = false; int ret; - if (adev->ras_enabled && con) - data = &con->eh_data; - else + if (!con) return 0; + /* Allow access to RAS EEPROM via debugfs, when the ASIC + * supports RAS and debugfs is enabled, but when + * adev->ras_enabled is unset, i.e. when "ras_enable" + * module parameter is set to 0. + */ + con->adev = adev; + + if (!adev->ras_enabled) + return 0; + + data = &con->eh_data; *data = kmalloc(sizeof(**data), GFP_KERNEL | __GFP_ZERO); if (!*data) { ret = -ENOMEM; @@ -1980,10 +1989,9 @@ int amdgpu_ras_recovery_init(struct amdgpu_device *adev) mutex_init(&con->recovery_lock); INIT_WORK(&con->recovery_work, amdgpu_ras_do_recovery); atomic_set(&con->in_recovery, 0); - con->adev = adev; - max_eeprom_records_len = amdgpu_ras_eeprom_get_record_max_length(); - amdgpu_ras_validate_threshold(adev, max_eeprom_records_len); + max_eeprom_records_count = amdgpu_ras_eeprom_max_record_count(); + amdgpu_ras_validate_threshold(adev, max_eeprom_records_count); /* Todo: During test the SMU might fail to read the eeprom through I2C * when the GPU is pending on XGMI reset during probe time @@ -1999,13 +2007,13 @@ int amdgpu_ras_recovery_init(struct amdgpu_device *adev) if (exc_err_limit || ret) goto free; - if (con->eeprom_control.num_recs) { + if (con->eeprom_control.ras_num_recs) { ret = amdgpu_ras_load_bad_pages(adev); if (ret) goto free; if (adev->smu.ppt_funcs && adev->smu.ppt_funcs->send_hbm_bad_pages_num) - adev->smu.ppt_funcs->send_hbm_bad_pages_num(&adev->smu, con->eeprom_control.num_recs); + adev->smu.ppt_funcs->send_hbm_bad_pages_num(&adev->smu, con->eeprom_control.ras_num_recs); } return 0; @@ -2015,7 +2023,7 @@ free: kfree(*data); con->eh_data = NULL; out: - dev_warn(adev->dev, "Failed to initialize ras recovery!\n"); + dev_warn(adev->dev, "Failed to initialize ras recovery! (%d)\n", ret); /* * Except error threshold exceeding case, other failure cases in this diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h index b504ed8c9b50..4d9c63f2f377 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.h @@ -318,6 +318,7 @@ struct amdgpu_ras { /* sysfs */ struct device_attribute features_attr; struct bin_attribute badpages_attr; + struct dentry *de_ras_eeprom_table; /* block array */ struct ras_manager *objs; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c index f40c871da0c6..d2e5b2567bc1 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.c @@ -26,90 +26,110 @@ #include "amdgpu_ras.h" #include <linux/bits.h> #include "atom.h" +#include "amdgpu_eeprom.h" +#include <linux/debugfs.h> +#include <linux/uaccess.h> -#define EEPROM_I2C_TARGET_ADDR_VEGA20 0xA0 -#define EEPROM_I2C_TARGET_ADDR_ARCTURUS 0xA8 -#define EEPROM_I2C_TARGET_ADDR_ARCTURUS_D342 0xA0 -#define EEPROM_I2C_TARGET_ADDR_SIENNA_CICHLID 0xA0 -#define EEPROM_I2C_TARGET_ADDR_ALDEBARAN 0xA0 +#define EEPROM_I2C_MADDR_VEGA20 0x0 +#define EEPROM_I2C_MADDR_ARCTURUS 0x40000 +#define EEPROM_I2C_MADDR_ARCTURUS_D342 0x0 +#define EEPROM_I2C_MADDR_SIENNA_CICHLID 0x0 +#define EEPROM_I2C_MADDR_ALDEBARAN 0x0 /* * The 2 macros bellow represent the actual size in bytes that * those entities occupy in the EEPROM memory. - * EEPROM_TABLE_RECORD_SIZE is different than sizeof(eeprom_table_record) which + * RAS_TABLE_RECORD_SIZE is different than sizeof(eeprom_table_record) which * uses uint64 to store 6b fields such as retired_page. */ -#define EEPROM_TABLE_HEADER_SIZE 20 -#define EEPROM_TABLE_RECORD_SIZE 24 - -#define EEPROM_ADDRESS_SIZE 0x2 +#define RAS_TABLE_HEADER_SIZE 20 +#define RAS_TABLE_RECORD_SIZE 24 /* Table hdr is 'AMDR' */ -#define EEPROM_TABLE_HDR_VAL 0x414d4452 -#define EEPROM_TABLE_VER 0x00010000 +#define RAS_TABLE_HDR_VAL 0x414d4452 +#define RAS_TABLE_VER 0x00010000 /* Bad GPU tag ‘BADG’ */ -#define EEPROM_TABLE_HDR_BAD 0x42414447 +#define RAS_TABLE_HDR_BAD 0x42414447 + +/* Assume 2-Mbit size EEPROM and take up the whole space. */ +#define RAS_TBL_SIZE_BYTES (256 * 1024) +#define RAS_TABLE_START 0 +#define RAS_HDR_START RAS_TABLE_START +#define RAS_RECORD_START (RAS_HDR_START + RAS_TABLE_HEADER_SIZE) +#define RAS_MAX_RECORD_COUNT ((RAS_TBL_SIZE_BYTES - RAS_TABLE_HEADER_SIZE) \ + / RAS_TABLE_RECORD_SIZE) + +/* Given a zero-based index of an EEPROM RAS record, yields the EEPROM + * offset off of RAS_TABLE_START. That is, this is something you can + * add to control->i2c_address, and then tell I2C layer to read + * from/write to there. _N is the so called absolute index, + * because it starts right after the table header. + */ +#define RAS_INDEX_TO_OFFSET(_C, _N) ((_C)->ras_record_offset + \ + (_N) * RAS_TABLE_RECORD_SIZE) + +#define RAS_OFFSET_TO_INDEX(_C, _O) (((_O) - \ + (_C)->ras_record_offset) / RAS_TABLE_RECORD_SIZE) + +/* Given a 0-based relative record index, 0, 1, 2, ..., etc., off + * of "fri", return the absolute record index off of the end of + * the table header. + */ +#define RAS_RI_TO_AI(_C, _I) (((_I) + (_C)->ras_fri) % \ + (_C)->ras_max_record_count) -/* Assume 2 Mbit size */ -#define EEPROM_SIZE_BYTES 256000 -#define EEPROM_PAGE__SIZE_BYTES 256 -#define EEPROM_HDR_START 0 -#define EEPROM_RECORD_START (EEPROM_HDR_START + EEPROM_TABLE_HEADER_SIZE) -#define EEPROM_MAX_RECORD_NUM ((EEPROM_SIZE_BYTES - EEPROM_TABLE_HEADER_SIZE) / EEPROM_TABLE_RECORD_SIZE) -#define EEPROM_ADDR_MSB_MASK GENMASK(17, 8) +#define RAS_NUM_RECS(_tbl_hdr) (((_tbl_hdr)->tbl_size - \ + RAS_TABLE_HEADER_SIZE) / RAS_TABLE_RECORD_SIZE) #define to_amdgpu_device(x) (container_of(x, struct amdgpu_ras, eeprom_control))->adev static bool __is_ras_eeprom_supported(struct amdgpu_device *adev) { - if ((adev->asic_type == CHIP_VEGA20) || - (adev->asic_type == CHIP_ARCTURUS) || - (adev->asic_type == CHIP_SIENNA_CICHLID) || - (adev->asic_type == CHIP_ALDEBARAN)) - return true; - - return false; + return adev->asic_type == CHIP_VEGA20 || + adev->asic_type == CHIP_ARCTURUS || + adev->asic_type == CHIP_SIENNA_CICHLID || + adev->asic_type == CHIP_ALDEBARAN; } static bool __get_eeprom_i2c_addr_arct(struct amdgpu_device *adev, - uint16_t *i2c_addr) + struct amdgpu_ras_eeprom_control *control) { struct atom_context *atom_ctx = adev->mode_info.atom_context; - if (!i2c_addr || !atom_ctx) + if (!control || !atom_ctx) return false; if (strnstr(atom_ctx->vbios_version, "D342", sizeof(atom_ctx->vbios_version))) - *i2c_addr = EEPROM_I2C_TARGET_ADDR_ARCTURUS_D342; + control->i2c_address = EEPROM_I2C_MADDR_ARCTURUS_D342; else - *i2c_addr = EEPROM_I2C_TARGET_ADDR_ARCTURUS; + control->i2c_address = EEPROM_I2C_MADDR_ARCTURUS; return true; } static bool __get_eeprom_i2c_addr(struct amdgpu_device *adev, - uint16_t *i2c_addr) + struct amdgpu_ras_eeprom_control *control) { - if (!i2c_addr) + if (!control) return false; switch (adev->asic_type) { case CHIP_VEGA20: - *i2c_addr = EEPROM_I2C_TARGET_ADDR_VEGA20; + control->i2c_address = EEPROM_I2C_MADDR_VEGA20; break; case CHIP_ARCTURUS: - return __get_eeprom_i2c_addr_arct(adev, i2c_addr); + return __get_eeprom_i2c_addr_arct(adev, control); case CHIP_SIENNA_CICHLID: - *i2c_addr = EEPROM_I2C_TARGET_ADDR_SIENNA_CICHLID; + control->i2c_address = EEPROM_I2C_MADDR_SIENNA_CICHLID; break; case CHIP_ALDEBARAN: - *i2c_addr = EEPROM_I2C_TARGET_ADDR_ALDEBARAN; + control->i2c_address = EEPROM_I2C_MADDR_ALDEBARAN; break; default: @@ -119,10 +139,11 @@ static bool __get_eeprom_i2c_addr(struct amdgpu_device *adev, return true; } -static void __encode_table_header_to_buff(struct amdgpu_ras_eeprom_table_header *hdr, - unsigned char *buff) +static void +__encode_table_header_to_buf(struct amdgpu_ras_eeprom_table_header *hdr, + unsigned char *buf) { - uint32_t *pp = (uint32_t *) buff; + u32 *pp = (uint32_t *)buf; pp[0] = cpu_to_le32(hdr->header); pp[1] = cpu_to_le32(hdr->version); @@ -131,10 +152,11 @@ static void __encode_table_header_to_buff(struct amdgpu_ras_eeprom_table_header pp[4] = cpu_to_le32(hdr->checksum); } -static void __decode_table_header_from_buff(struct amdgpu_ras_eeprom_table_header *hdr, - unsigned char *buff) +static void +__decode_table_header_from_buf(struct amdgpu_ras_eeprom_table_header *hdr, + unsigned char *buf) { - uint32_t *pp = (uint32_t *)buff; + u32 *pp = (uint32_t *)buf; hdr->header = le32_to_cpu(pp[0]); hdr->version = le32_to_cpu(pp[1]); @@ -143,303 +165,168 @@ static void __decode_table_header_from_buff(struct amdgpu_ras_eeprom_table_heade hdr->checksum = le32_to_cpu(pp[4]); } -static int __update_table_header(struct amdgpu_ras_eeprom_control *control, - unsigned char *buff) +static int __write_table_header(struct amdgpu_ras_eeprom_control *control) { - int ret = 0; + u8 buf[RAS_TABLE_HEADER_SIZE]; struct amdgpu_device *adev = to_amdgpu_device(control); - struct i2c_msg msg = { - .addr = 0, - .flags = 0, - .len = EEPROM_ADDRESS_SIZE + EEPROM_TABLE_HEADER_SIZE, - .buf = buff, - }; + int res; - - *(uint16_t *)buff = EEPROM_HDR_START; - __encode_table_header_to_buff(&control->tbl_hdr, buff + EEPROM_ADDRESS_SIZE); - - msg.addr = control->i2c_address; + memset(buf, 0, sizeof(buf)); + __encode_table_header_to_buf(&control->tbl_hdr, buf); /* i2c may be unstable in gpu reset */ down_read(&adev->reset_sem); - ret = i2c_transfer(&adev->pm.smu_i2c, &msg, 1); + res = amdgpu_eeprom_write(&adev->pm.smu_i2c, + control->i2c_address + + control->ras_header_offset, + buf, RAS_TABLE_HEADER_SIZE); up_read(&adev->reset_sem); - if (ret < 1) - DRM_ERROR("Failed to write EEPROM table header, ret:%d", ret); - - return ret; -} - -static uint32_t __calc_hdr_byte_sum(struct amdgpu_ras_eeprom_control *control) -{ - int i; - uint32_t tbl_sum = 0; - - /* Header checksum, skip checksum field in the calculation */ - for (i = 0; i < sizeof(control->tbl_hdr) - sizeof(control->tbl_hdr.checksum); i++) - tbl_sum += *(((unsigned char *)&control->tbl_hdr) + i); - - return tbl_sum; -} - -static uint32_t __calc_recs_byte_sum(struct eeprom_table_record *records, - int num) -{ - int i, j; - uint32_t tbl_sum = 0; - - /* Records checksum */ - for (i = 0; i < num; i++) { - struct eeprom_table_record *record = &records[i]; - - for (j = 0; j < sizeof(*record); j++) { - tbl_sum += *(((unsigned char *)record) + j); - } + if (res < 0) { + DRM_ERROR("Failed to write EEPROM table header:%d", res); + } else if (res < RAS_TABLE_HEADER_SIZE) { + DRM_ERROR("Short write:%d out of %d\n", + res, RAS_TABLE_HEADER_SIZE); + res = -EIO; + } else { + res = 0; } - return tbl_sum; + return res; } -static inline uint32_t __calc_tbl_byte_sum(struct amdgpu_ras_eeprom_control *control, - struct eeprom_table_record *records, int num) +static u8 __calc_hdr_byte_sum(const struct amdgpu_ras_eeprom_control *control) { - return __calc_hdr_byte_sum(control) + __calc_recs_byte_sum(records, num); -} + int ii; + u8 *pp, csum; + size_t sz; -/* Checksum = 256 -((sum of all table entries) mod 256) */ -static void __update_tbl_checksum(struct amdgpu_ras_eeprom_control *control, - struct eeprom_table_record *records, int num, - uint32_t old_hdr_byte_sum) -{ - /* - * This will update the table sum with new records. - * - * TODO: What happens when the EEPROM table is to be wrapped around - * and old records from start will get overridden. - */ - - /* need to recalculate updated header byte sum */ - control->tbl_byte_sum -= old_hdr_byte_sum; - control->tbl_byte_sum += __calc_tbl_byte_sum(control, records, num); - - control->tbl_hdr.checksum = 256 - (control->tbl_byte_sum % 256); -} - -/* table sum mod 256 + checksum must equals 256 */ -static bool __validate_tbl_checksum(struct amdgpu_ras_eeprom_control *control, - struct eeprom_table_record *records, int num) -{ - control->tbl_byte_sum = __calc_tbl_byte_sum(control, records, num); - - if (control->tbl_hdr.checksum + (control->tbl_byte_sum % 256) != 256) { - DRM_WARN("Checksum mismatch, checksum: %u ", control->tbl_hdr.checksum); - return false; - } + /* Header checksum, skip checksum field in the calculation */ + sz = sizeof(control->tbl_hdr) - sizeof(control->tbl_hdr.checksum); + pp = (u8 *) &control->tbl_hdr; + csum = 0; + for (ii = 0; ii < sz; ii++, pp++) + csum += *pp; - return true; + return csum; } static int amdgpu_ras_eeprom_correct_header_tag( - struct amdgpu_ras_eeprom_control *control, - uint32_t header) + struct amdgpu_ras_eeprom_control *control, + uint32_t header) { - unsigned char buff[EEPROM_ADDRESS_SIZE + EEPROM_TABLE_HEADER_SIZE]; struct amdgpu_ras_eeprom_table_header *hdr = &control->tbl_hdr; - int ret = 0; - - memset(buff, 0, EEPROM_ADDRESS_SIZE + EEPROM_TABLE_HEADER_SIZE); - - mutex_lock(&control->tbl_mutex); + u8 *hh; + int res; + u8 csum; + + csum = -hdr->checksum; + + hh = (void *) &hdr->header; + csum -= (hh[0] + hh[1] + hh[2] + hh[3]); + hh = (void *) &header; + csum += hh[0] + hh[1] + hh[2] + hh[3]; + csum = -csum; + mutex_lock(&control->ras_tbl_mutex); hdr->header = header; - ret = __update_table_header(control, buff); - mutex_unlock(&control->tbl_mutex); + hdr->checksum = csum; + res = __write_table_header(control); + mutex_unlock(&control->ras_tbl_mutex); - return ret; + return res; } +/** + * amdgpu_ras_eeprom_reset_table -- Reset the RAS EEPROM table + * @control: pointer to control structure + * + * Reset the contents of the header of the RAS EEPROM table. + * Return 0 on success, -errno on error. + */ int amdgpu_ras_eeprom_reset_table(struct amdgpu_ras_eeprom_control *control) { - unsigned char buff[EEPROM_ADDRESS_SIZE + EEPROM_TABLE_HEADER_SIZE] = { 0 }; struct amdgpu_ras_eeprom_table_header *hdr = &control->tbl_hdr; - int ret = 0; + u8 csum; + int res; - mutex_lock(&control->tbl_mutex); + mutex_lock(&control->ras_tbl_mutex); - hdr->header = EEPROM_TABLE_HDR_VAL; - hdr->version = EEPROM_TABLE_VER; - hdr->first_rec_offset = EEPROM_RECORD_START; - hdr->tbl_size = EEPROM_TABLE_HEADER_SIZE; + hdr->header = RAS_TABLE_HDR_VAL; + hdr->version = RAS_TABLE_VER; + hdr->first_rec_offset = RAS_RECORD_START; + hdr->tbl_size = RAS_TABLE_HEADER_SIZE; - control->tbl_byte_sum = 0; - __update_tbl_checksum(control, NULL, 0, 0); - control->next_addr = EEPROM_RECORD_START; + csum = __calc_hdr_byte_sum(control); + csum = -csum; + hdr->checksum = csum; + res = __write_table_header(control); - ret = __update_table_header(control, buff); + control->ras_num_recs = 0; + control->ras_fri = 0; - mutex_unlock(&control->tbl_mutex); + amdgpu_ras_debugfs_set_ret_size(control); - return ret; + mutex_unlock(&control->ras_tbl_mutex); + return res; } -int amdgpu_ras_eeprom_init(struct amdgpu_ras_eeprom_control *control, - bool *exceed_err_limit) -{ - int ret = 0; - struct amdgpu_device *adev = to_amdgpu_device(control); - unsigned char buff[EEPROM_ADDRESS_SIZE + EEPROM_TABLE_HEADER_SIZE] = { 0 }; - struct amdgpu_ras_eeprom_table_header *hdr = &control->tbl_hdr; - struct amdgpu_ras *ras = amdgpu_ras_get_context(adev); - struct i2c_msg msg = { - .addr = 0, - .flags = I2C_M_RD, - .len = EEPROM_ADDRESS_SIZE + EEPROM_TABLE_HEADER_SIZE, - .buf = buff, - }; - - *exceed_err_limit = false; - - if (!__is_ras_eeprom_supported(adev)) - return 0; - - /* Verify i2c adapter is initialized */ - if (!adev->pm.smu_i2c.algo) - return -ENOENT; - - if (!__get_eeprom_i2c_addr(adev, &control->i2c_address)) - return -EINVAL; - - mutex_init(&control->tbl_mutex); - - msg.addr = control->i2c_address; - /* Read/Create table header from EEPROM address 0 */ - ret = i2c_transfer(&adev->pm.smu_i2c, &msg, 1); - if (ret < 1) { - DRM_ERROR("Failed to read EEPROM table header, ret:%d", ret); - return ret; - } - - __decode_table_header_from_buff(hdr, &buff[2]); - - if (hdr->header == EEPROM_TABLE_HDR_VAL) { - control->num_recs = (hdr->tbl_size - EEPROM_TABLE_HEADER_SIZE) / - EEPROM_TABLE_RECORD_SIZE; - control->tbl_byte_sum = __calc_hdr_byte_sum(control); - control->next_addr = EEPROM_RECORD_START; - - DRM_DEBUG_DRIVER("Found existing EEPROM table with %d records", - control->num_recs); - - } else if ((hdr->header == EEPROM_TABLE_HDR_BAD) && - (amdgpu_bad_page_threshold != 0)) { - if (ras->bad_page_cnt_threshold > control->num_recs) { - dev_info(adev->dev, "Using one valid bigger bad page " - "threshold and correcting eeprom header tag.\n"); - ret = amdgpu_ras_eeprom_correct_header_tag(control, - EEPROM_TABLE_HDR_VAL); - } else { - *exceed_err_limit = true; - dev_err(adev->dev, "Exceeding the bad_page_threshold parameter, " - "disabling the GPU.\n"); - } - } else { - DRM_INFO("Creating new EEPROM table"); - - ret = amdgpu_ras_eeprom_reset_table(control); - } - - return ret == 1 ? 0 : -EIO; -} - -static void __encode_table_record_to_buff(struct amdgpu_ras_eeprom_control *control, - struct eeprom_table_record *record, - unsigned char *buff) +static void +__encode_table_record_to_buf(struct amdgpu_ras_eeprom_control *control, + struct eeprom_table_record *record, + unsigned char *buf) { __le64 tmp = 0; int i = 0; /* Next are all record fields according to EEPROM page spec in LE foramt */ - buff[i++] = record->err_type; + buf[i++] = record->err_type; - buff[i++] = record->bank; + buf[i++] = record->bank; tmp = cpu_to_le64(record->ts); - memcpy(buff + i, &tmp, 8); + memcpy(buf + i, &tmp, 8); i += 8; tmp = cpu_to_le64((record->offset & 0xffffffffffff)); - memcpy(buff + i, &tmp, 6); + memcpy(buf + i, &tmp, 6); i += 6; - buff[i++] = record->mem_channel; - buff[i++] = record->mcumc_id; + buf[i++] = record->mem_channel; + buf[i++] = record->mcumc_id; tmp = cpu_to_le64((record->retired_page & 0xffffffffffff)); - memcpy(buff + i, &tmp, 6); + memcpy(buf + i, &tmp, 6); } -static void __decode_table_record_from_buff(struct amdgpu_ras_eeprom_control *control, - struct eeprom_table_record *record, - unsigned char *buff) +static void +__decode_table_record_from_buf(struct amdgpu_ras_eeprom_control *control, + struct eeprom_table_record *record, + unsigned char *buf) { __le64 tmp = 0; int i = 0; /* Next are all record fields according to EEPROM page spec in LE foramt */ - record->err_type = buff[i++]; + record->err_type = buf[i++]; - record->bank = buff[i++]; + record->bank = buf[i++]; - memcpy(&tmp, buff + i, 8); + memcpy(&tmp, buf + i, 8); record->ts = le64_to_cpu(tmp); i += 8; - memcpy(&tmp, buff + i, 6); + memcpy(&tmp, buf + i, 6); record->offset = (le64_to_cpu(tmp) & 0xffffffffffff); i += 6; - record->mem_channel = buff[i++]; - record->mcumc_id = buff[i++]; + record->mem_channel = buf[i++]; + record->mcumc_id = buf[i++]; - memcpy(&tmp, buff + i, 6); + memcpy(&tmp, buf + i, 6); record->retired_page = (le64_to_cpu(tmp) & 0xffffffffffff); } -/* - * When reaching end of EEPROM memory jump back to 0 record address - * When next record access will go beyond EEPROM page boundary modify bits A17/A8 - * in I2C selector to go to next page - */ -static uint32_t __correct_eeprom_dest_address(uint32_t curr_address) -{ - uint32_t next_address = curr_address + EEPROM_TABLE_RECORD_SIZE; - - /* When all EEPROM memory used jump back to 0 address */ - if (next_address > EEPROM_SIZE_BYTES) { - DRM_INFO("Reached end of EEPROM memory, jumping to 0 " - "and overriding old record"); - return EEPROM_RECORD_START; - } - - /* - * To check if we overflow page boundary compare next address with - * current and see if bits 17/8 of the EEPROM address will change - * If they do start from the next 256b page - * - * https://www.st.com/resource/en/datasheet/m24m02-dr.pdf sec. 5.1.2 - */ - if ((curr_address & EEPROM_ADDR_MSB_MASK) != (next_address & EEPROM_ADDR_MSB_MASK)) { - DRM_DEBUG_DRIVER("Reached end of EEPROM memory page, jumping to next: %lx", - (next_address & EEPROM_ADDR_MSB_MASK)); - - return (next_address & EEPROM_ADDR_MSB_MASK); - } - - return curr_address; -} - bool amdgpu_ras_eeprom_check_err_threshold(struct amdgpu_device *adev) { struct amdgpu_ras *con = amdgpu_ras_get_context(adev); @@ -454,197 +341,756 @@ bool amdgpu_ras_eeprom_check_err_threshold(struct amdgpu_device *adev) if (!(con->features & BIT(AMDGPU_RAS_BLOCK__UMC))) return false; - if (con->eeprom_control.tbl_hdr.header == EEPROM_TABLE_HDR_BAD) { + if (con->eeprom_control.tbl_hdr.header == RAS_TABLE_HDR_BAD) { dev_warn(adev->dev, "This GPU is in BAD status."); - dev_warn(adev->dev, "Please retire it or setting one bigger " - "threshold value when reloading driver.\n"); + dev_warn(adev->dev, "Please retire it or set a larger " + "threshold value when reloading driver.\n"); return true; } return false; } -int amdgpu_ras_eeprom_process_recods(struct amdgpu_ras_eeprom_control *control, - struct eeprom_table_record *records, - bool write, - int num) +/** + * __amdgpu_ras_eeprom_write -- write indexed from buffer to EEPROM + * @control: pointer to control structure + * @buf: pointer to buffer containing data to write + * @fri: start writing at this index + * @num: number of records to write + * + * The caller must hold the table mutex in @control. + * Return 0 on success, -errno otherwise. + */ +static int __amdgpu_ras_eeprom_write(struct amdgpu_ras_eeprom_control *control, + u8 *buf, const u32 fri, const u32 num) { - int i, ret = 0; - struct i2c_msg *msgs, *msg; - unsigned char *buffs, *buff; - struct eeprom_table_record *record; struct amdgpu_device *adev = to_amdgpu_device(control); - struct amdgpu_ras *ras = amdgpu_ras_get_context(adev); + u32 buf_size; + int res; - if (!__is_ras_eeprom_supported(adev)) - return 0; + /* i2c may be unstable in gpu reset */ + down_read(&adev->reset_sem); + buf_size = num * RAS_TABLE_RECORD_SIZE; + res = amdgpu_eeprom_write(&adev->pm.smu_i2c, + control->i2c_address + + RAS_INDEX_TO_OFFSET(control, fri), + buf, buf_size); + up_read(&adev->reset_sem); + if (res < 0) { + DRM_ERROR("Writing %d EEPROM table records error:%d", + num, res); + } else if (res < buf_size) { + /* Short write, return error. + */ + DRM_ERROR("Wrote %d records out of %d", + res / RAS_TABLE_RECORD_SIZE, num); + res = -EIO; + } else { + res = 0; + } - buffs = kcalloc(num, EEPROM_ADDRESS_SIZE + EEPROM_TABLE_RECORD_SIZE, - GFP_KERNEL); - if (!buffs) - return -ENOMEM; + return res; +} + +static int +amdgpu_ras_eeprom_append_table(struct amdgpu_ras_eeprom_control *control, + struct eeprom_table_record *record, + const u32 num) +{ + u32 a, b, i; + u8 *buf, *pp; + int res; - mutex_lock(&control->tbl_mutex); + buf = kcalloc(num, RAS_TABLE_RECORD_SIZE, GFP_KERNEL); + if (!buf) + return -ENOMEM; - msgs = kcalloc(num, sizeof(*msgs), GFP_KERNEL); - if (!msgs) { - ret = -ENOMEM; - goto free_buff; + /* Encode all of them in one go. + */ + pp = buf; + for (i = 0; i < num; i++, pp += RAS_TABLE_RECORD_SIZE) + __encode_table_record_to_buf(control, &record[i], pp); + + /* a, first record index to write into. + * b, last record index to write into. + * a = first index to read (fri) + number of records in the table, + * b = a + @num - 1. + * Let N = control->ras_max_num_record_count, then we have, + * case 0: 0 <= a <= b < N, + * just append @num records starting at a; + * case 1: 0 <= a < N <= b, + * append (N - a) records starting at a, and + * append the remainder, b % N + 1, starting at 0. + * case 2: 0 <= fri < N <= a <= b, then modulo N we get two subcases, + * case 2a: 0 <= a <= b < N + * append num records starting at a; and fix fri if b overwrote it, + * and since a <= b, if b overwrote it then a must've also, + * and if b didn't overwrite it, then a didn't also. + * case 2b: 0 <= b < a < N + * write num records starting at a, which wraps around 0=N + * and overwrite fri unconditionally. Now from case 2a, + * this means that b eclipsed fri to overwrite it and wrap + * around 0 again, i.e. b = 2N+r pre modulo N, so we unconditionally + * set fri = b + 1 (mod N). + * Now, since fri is updated in every case, except the trivial case 0, + * the number of records present in the table after writing, is, + * num_recs - 1 = b - fri (mod N), and we take the positive value, + * by adding an arbitrary multiple of N before taking the modulo N + * as shown below. + */ + a = control->ras_fri + control->ras_num_recs; + b = a + num - 1; + if (b < control->ras_max_record_count) { + res = __amdgpu_ras_eeprom_write(control, buf, a, num); + } else if (a < control->ras_max_record_count) { + u32 g0, g1; + + g0 = control->ras_max_record_count - a; + g1 = b % control->ras_max_record_count + 1; + res = __amdgpu_ras_eeprom_write(control, buf, a, g0); + if (res) + goto Out; + res = __amdgpu_ras_eeprom_write(control, + buf + g0 * RAS_TABLE_RECORD_SIZE, + 0, g1); + if (res) + goto Out; + if (g1 > control->ras_fri) + control->ras_fri = g1 % control->ras_max_record_count; + } else { + a %= control->ras_max_record_count; + b %= control->ras_max_record_count; + + if (a <= b) { + /* Note that, b - a + 1 = num. */ + res = __amdgpu_ras_eeprom_write(control, buf, a, num); + if (res) + goto Out; + if (b >= control->ras_fri) + control->ras_fri = (b + 1) % control->ras_max_record_count; + } else { + u32 g0, g1; + + /* b < a, which means, we write from + * a to the end of the table, and from + * the start of the table to b. + */ + g0 = control->ras_max_record_count - a; + g1 = b + 1; + res = __amdgpu_ras_eeprom_write(control, buf, a, g0); + if (res) + goto Out; + res = __amdgpu_ras_eeprom_write(control, + buf + g0 * RAS_TABLE_RECORD_SIZE, + 0, g1); + if (res) + goto Out; + control->ras_fri = g1 % control->ras_max_record_count; + } } + control->ras_num_recs = 1 + (control->ras_max_record_count + b + - control->ras_fri) + % control->ras_max_record_count; +Out: + kfree(buf); + return res; +} - /* - * If saved bad pages number exceeds the bad page threshold for - * the whole VRAM, update table header to mark the BAD GPU tag - * and schedule one ras recovery after eeprom write is done, - * this can avoid the missing for latest records. - * - * This new header will be picked up and checked in the bootup - * by ras recovery, which may break bootup process to notify - * user this GPU is in bad state and to retire such GPU for - * further check. +static int +amdgpu_ras_eeprom_update_header(struct amdgpu_ras_eeprom_control *control) +{ + struct amdgpu_device *adev = to_amdgpu_device(control); + struct amdgpu_ras *ras = amdgpu_ras_get_context(adev); + u8 *buf, *pp, csum; + u32 buf_size; + int res; + + /* Modify the header if it exceeds. */ - if (write && (amdgpu_bad_page_threshold != 0) && - ((control->num_recs + num) >= ras->bad_page_cnt_threshold)) { + if (amdgpu_bad_page_threshold != 0 && + control->ras_num_recs >= ras->bad_page_cnt_threshold) { dev_warn(adev->dev, - "Saved bad pages(%d) reaches threshold value(%d).\n", - control->num_recs + num, ras->bad_page_cnt_threshold); - control->tbl_hdr.header = EEPROM_TABLE_HDR_BAD; + "Saved bad pages %d reaches threshold value %d\n", + control->ras_num_recs, ras->bad_page_cnt_threshold); + control->tbl_hdr.header = RAS_TABLE_HDR_BAD; } - /* In case of overflow just start from beginning to not lose newest records */ - if (write && (control->next_addr + EEPROM_TABLE_RECORD_SIZE * num > EEPROM_SIZE_BYTES)) - control->next_addr = EEPROM_RECORD_START; + control->tbl_hdr.version = RAS_TABLE_VER; + control->tbl_hdr.first_rec_offset = RAS_INDEX_TO_OFFSET(control, control->ras_fri); + control->tbl_hdr.tbl_size = RAS_TABLE_HEADER_SIZE + control->ras_num_recs * RAS_TABLE_RECORD_SIZE; + control->tbl_hdr.checksum = 0; + + buf_size = control->ras_num_recs * RAS_TABLE_RECORD_SIZE; + buf = kcalloc(control->ras_num_recs, RAS_TABLE_RECORD_SIZE, GFP_KERNEL); + if (!buf) { + DRM_ERROR("allocating memory for table of size %d bytes failed\n", + control->tbl_hdr.tbl_size); + res = -ENOMEM; + goto Out; + } - /* - * TODO Currently makes EEPROM writes for each record, this creates - * internal fragmentation. Optimized the code to do full page write of - * 256b + down_read(&adev->reset_sem); + res = amdgpu_eeprom_read(&adev->pm.smu_i2c, + control->i2c_address + + control->ras_record_offset, + buf, buf_size); + up_read(&adev->reset_sem); + if (res < 0) { + DRM_ERROR("EEPROM failed reading records:%d\n", + res); + goto Out; + } else if (res < buf_size) { + DRM_ERROR("EEPROM read %d out of %d bytes\n", + res, buf_size); + res = -EIO; + goto Out; + } + + /* Recalc the checksum. */ - for (i = 0; i < num; i++) { - buff = &buffs[i * (EEPROM_ADDRESS_SIZE + EEPROM_TABLE_RECORD_SIZE)]; - record = &records[i]; - msg = &msgs[i]; + csum = 0; + for (pp = buf; pp < buf + buf_size; pp++) + csum += *pp; + + csum += __calc_hdr_byte_sum(control); + /* avoid sign extension when assigning to "checksum" */ + csum = -csum; + control->tbl_hdr.checksum = csum; + res = __write_table_header(control); +Out: + kfree(buf); + return res; +} + +/** + * amdgpu_ras_eeprom_append -- append records to the EEPROM RAS table + * @control: pointer to control structure + * @record: array of records to append + * @num: number of records in @record array + * + * Append @num records to the table, calculate the checksum and write + * the table back to EEPROM. The maximum number of records that + * can be appended is between 1 and control->ras_max_record_count, + * regardless of how many records are already stored in the table. + * + * Return 0 on success or if EEPROM is not supported, -errno on error. + */ +int amdgpu_ras_eeprom_append(struct amdgpu_ras_eeprom_control *control, + struct eeprom_table_record *record, + const u32 num) +{ + struct amdgpu_device *adev = to_amdgpu_device(control); + int res; - control->next_addr = __correct_eeprom_dest_address(control->next_addr); + if (!__is_ras_eeprom_supported(adev)) + return 0; - /* - * Update bits 16,17 of EEPROM address in I2C address by setting them - * to bits 1,2 of Device address byte - */ - msg->addr = control->i2c_address | - ((control->next_addr & EEPROM_ADDR_MSB_MASK) >> 15); - msg->flags = write ? 0 : I2C_M_RD; - msg->len = EEPROM_ADDRESS_SIZE + EEPROM_TABLE_RECORD_SIZE; - msg->buf = buff; - - /* Insert the EEPROM dest addess, bits 0-15 */ - buff[0] = ((control->next_addr >> 8) & 0xff); - buff[1] = (control->next_addr & 0xff); - - /* EEPROM table content is stored in LE format */ - if (write) - __encode_table_record_to_buff(control, record, buff + EEPROM_ADDRESS_SIZE); - - /* - * The destination EEPROM address might need to be corrected to account - * for page or entire memory wrapping - */ - control->next_addr += EEPROM_TABLE_RECORD_SIZE; + if (num == 0) { + DRM_ERROR("will not append 0 records\n"); + return -EINVAL; + } else if (num > control->ras_max_record_count) { + DRM_ERROR("cannot append %d records than the size of table %d\n", + num, control->ras_max_record_count); + return -EINVAL; } + mutex_lock(&control->ras_tbl_mutex); + + res = amdgpu_ras_eeprom_append_table(control, record, num); + if (!res) + res = amdgpu_ras_eeprom_update_header(control); + if (!res) + amdgpu_ras_debugfs_set_ret_size(control); + + mutex_unlock(&control->ras_tbl_mutex); + return res; +} + +/** + * __amdgpu_ras_eeprom_read -- read indexed from EEPROM into buffer + * @control: pointer to control structure + * @buf: pointer to buffer to read into + * @fri: first record index, start reading at this index, absolute index + * @num: number of records to read + * + * The caller must hold the table mutex in @control. + * Return 0 on success, -errno otherwise. + */ +static int __amdgpu_ras_eeprom_read(struct amdgpu_ras_eeprom_control *control, + u8 *buf, const u32 fri, const u32 num) +{ + struct amdgpu_device *adev = to_amdgpu_device(control); + u32 buf_size; + int res; + /* i2c may be unstable in gpu reset */ down_read(&adev->reset_sem); - ret = i2c_transfer(&adev->pm.smu_i2c, msgs, num); + buf_size = num * RAS_TABLE_RECORD_SIZE; + res = amdgpu_eeprom_read(&adev->pm.smu_i2c, + control->i2c_address + + RAS_INDEX_TO_OFFSET(control, fri), + buf, buf_size); up_read(&adev->reset_sem); + if (res < 0) { + DRM_ERROR("Reading %d EEPROM table records error:%d", + num, res); + } else if (res < buf_size) { + /* Short read, return error. + */ + DRM_ERROR("Read %d records out of %d", + res / RAS_TABLE_RECORD_SIZE, num); + res = -EIO; + } else { + res = 0; + } - if (ret < 1) { - DRM_ERROR("Failed to process EEPROM table records, ret:%d", ret); + return res; +} - /* TODO Restore prev next EEPROM address ? */ - goto free_msgs; +/** + * amdgpu_ras_eeprom_read -- read EEPROM + * @control: pointer to control structure + * @record: array of records to read into + * @num: number of records in @record + * + * Reads num records from the RAS table in EEPROM and + * writes the data into @record array. + * + * Returns 0 on success, -errno on error. + */ +int amdgpu_ras_eeprom_read(struct amdgpu_ras_eeprom_control *control, + struct eeprom_table_record *record, + const u32 num) +{ + struct amdgpu_device *adev = to_amdgpu_device(control); + int i, res; + u8 *buf, *pp; + u32 g0, g1; + + if (!__is_ras_eeprom_supported(adev)) + return 0; + + if (num == 0) { + DRM_ERROR("will not read 0 records\n"); + return -EINVAL; + } else if (num > control->ras_num_recs) { + DRM_ERROR("too many records to read:%d available:%d\n", + num, control->ras_num_recs); + return -EINVAL; } + buf = kcalloc(num, RAS_TABLE_RECORD_SIZE, GFP_KERNEL); + if (!buf) + return -ENOMEM; - if (!write) { - for (i = 0; i < num; i++) { - buff = &buffs[i*(EEPROM_ADDRESS_SIZE + EEPROM_TABLE_RECORD_SIZE)]; - record = &records[i]; + /* Determine how many records to read, from the first record + * index, fri, to the end of the table, and from the beginning + * of the table, such that the total number of records is + * @num, and we handle wrap around when fri > 0 and + * fri + num > RAS_MAX_RECORD_COUNT. + * + * First we compute the index of the last element + * which would be fetched from each region, + * g0 is in [fri, fri + num - 1], and + * g1 is in [0, RAS_MAX_RECORD_COUNT - 1]. + * Then, if g0 < RAS_MAX_RECORD_COUNT, the index of + * the last element to fetch, we set g0 to _the number_ + * of elements to fetch, @num, since we know that the last + * indexed to be fetched does not exceed the table. + * + * If, however, g0 >= RAS_MAX_RECORD_COUNT, then + * we set g0 to the number of elements to read + * until the end of the table, and g1 to the number of + * elements to read from the beginning of the table. + */ + g0 = control->ras_fri + num - 1; + g1 = g0 % control->ras_max_record_count; + if (g0 < control->ras_max_record_count) { + g0 = num; + g1 = 0; + } else { + g0 = control->ras_max_record_count - control->ras_fri; + g1 += 1; + } - __decode_table_record_from_buff(control, record, buff + EEPROM_ADDRESS_SIZE); - } + mutex_lock(&control->ras_tbl_mutex); + res = __amdgpu_ras_eeprom_read(control, buf, control->ras_fri, g0); + if (res) + goto Out; + if (g1) { + res = __amdgpu_ras_eeprom_read(control, + buf + g0 * RAS_TABLE_RECORD_SIZE, + 0, g1); + if (res) + goto Out; + } + + res = 0; + + /* Read up everything? Then transform. + */ + pp = buf; + for (i = 0; i < num; i++, pp += RAS_TABLE_RECORD_SIZE) + __decode_table_record_from_buf(control, &record[i], pp); +Out: + kfree(buf); + mutex_unlock(&control->ras_tbl_mutex); + + return res; +} + +inline uint32_t amdgpu_ras_eeprom_max_record_count(void) +{ + return RAS_MAX_RECORD_COUNT; +} + +static ssize_t +amdgpu_ras_debugfs_eeprom_size_read(struct file *f, char __user *buf, + size_t size, loff_t *pos) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)file_inode(f)->i_private; + struct amdgpu_ras *ras = amdgpu_ras_get_context(adev); + struct amdgpu_ras_eeprom_control *control = ras ? &ras->eeprom_control : NULL; + u8 data[50]; + int res; + + if (!size) + return size; + + if (!ras || !control) { + res = snprintf(data, sizeof(data), "Not supported\n"); + } else { + res = snprintf(data, sizeof(data), "%d bytes or %d records\n", + RAS_TBL_SIZE_BYTES, control->ras_max_record_count); } - if (write) { - uint32_t old_hdr_byte_sum = __calc_hdr_byte_sum(control); + if (*pos >= res) + return 0; + + res -= *pos; + res = min_t(size_t, res, size); - /* - * Update table header with size and CRC and account for table - * wrap around where the assumption is that we treat it as empty - * table - * - * TODO - Check the assumption is correct + if (copy_to_user(buf, &data[*pos], res)) + return -EFAULT; + + *pos += res; + + return res; +} + +const struct file_operations amdgpu_ras_debugfs_eeprom_size_ops = { + .owner = THIS_MODULE, + .read = amdgpu_ras_debugfs_eeprom_size_read, + .write = NULL, + .llseek = default_llseek, +}; + +static const char *tbl_hdr_str = " Signature Version FirstOffs Size Checksum\n"; +static const char *tbl_hdr_fmt = "0x%08X 0x%08X 0x%08X 0x%08X 0x%08X\n"; +#define tbl_hdr_fmt_size (5 * (2+8) + 4 + 1) +static const char *rec_hdr_str = "Index Offset ErrType Bank/CU TimeStamp Offs/Addr MemChl MCUMCID RetiredPage\n"; +static const char *rec_hdr_fmt = "%5d 0x%05X %7s 0x%02X 0x%016llX 0x%012llX 0x%02X 0x%02X 0x%012llX\n"; +#define rec_hdr_fmt_size (5 + 1 + 7 + 1 + 7 + 1 + 7 + 1 + 18 + 1 + 14 + 1 + 6 + 1 + 7 + 1 + 14 + 1) + +static const char *record_err_type_str[AMDGPU_RAS_EEPROM_ERR_COUNT] = { + "ignore", + "re", + "ue", +}; + +static loff_t amdgpu_ras_debugfs_table_size(struct amdgpu_ras_eeprom_control *control) +{ + return strlen(tbl_hdr_str) + tbl_hdr_fmt_size + + strlen(rec_hdr_str) + rec_hdr_fmt_size * control->ras_num_recs; +} + +void amdgpu_ras_debugfs_set_ret_size(struct amdgpu_ras_eeprom_control *control) +{ + struct amdgpu_ras *ras = container_of(control, struct amdgpu_ras, + eeprom_control); + struct dentry *de = ras->de_ras_eeprom_table; + + if (de) + d_inode(de)->i_size = amdgpu_ras_debugfs_table_size(control); +} + +static ssize_t amdgpu_ras_debugfs_table_read(struct file *f, char __user *buf, + size_t size, loff_t *pos) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)file_inode(f)->i_private; + struct amdgpu_ras *ras = amdgpu_ras_get_context(adev); + struct amdgpu_ras_eeprom_control *control = &ras->eeprom_control; + const size_t orig_size = size; + int res = -EFAULT; + size_t data_len; + + mutex_lock(&control->ras_tbl_mutex); + + /* We want *pos - data_len > 0, which means there's + * bytes to be printed from data. + */ + data_len = strlen(tbl_hdr_str); + if (*pos < data_len) { + data_len -= *pos; + data_len = min_t(size_t, data_len, size); + if (copy_to_user(buf, &tbl_hdr_str[*pos], data_len)) + goto Out; + buf += data_len; + size -= data_len; + *pos += data_len; + } + + data_len = strlen(tbl_hdr_str) + tbl_hdr_fmt_size; + if (*pos < data_len && size > 0) { + u8 data[tbl_hdr_fmt_size + 1]; + loff_t lpos; + + snprintf(data, sizeof(data), tbl_hdr_fmt, + control->tbl_hdr.header, + control->tbl_hdr.version, + control->tbl_hdr.first_rec_offset, + control->tbl_hdr.tbl_size, + control->tbl_hdr.checksum); + + data_len -= *pos; + data_len = min_t(size_t, data_len, size); + lpos = *pos - strlen(tbl_hdr_str); + if (copy_to_user(buf, &data[lpos], data_len)) + goto Out; + buf += data_len; + size -= data_len; + *pos += data_len; + } + + data_len = strlen(tbl_hdr_str) + tbl_hdr_fmt_size + strlen(rec_hdr_str); + if (*pos < data_len && size > 0) { + loff_t lpos; + + data_len -= *pos; + data_len = min_t(size_t, data_len, size); + lpos = *pos - strlen(tbl_hdr_str) - tbl_hdr_fmt_size; + if (copy_to_user(buf, &rec_hdr_str[lpos], data_len)) + goto Out; + buf += data_len; + size -= data_len; + *pos += data_len; + } + + data_len = amdgpu_ras_debugfs_table_size(control); + if (*pos < data_len && size > 0) { + u8 dare[RAS_TABLE_RECORD_SIZE]; + u8 data[rec_hdr_fmt_size + 1]; + struct eeprom_table_record record; + int s, r; + + /* Find the starting record index */ - control->num_recs += num; - control->num_recs %= EEPROM_MAX_RECORD_NUM; - control->tbl_hdr.tbl_size += EEPROM_TABLE_RECORD_SIZE * num; - if (control->tbl_hdr.tbl_size > EEPROM_SIZE_BYTES) - control->tbl_hdr.tbl_size = EEPROM_TABLE_HEADER_SIZE + - control->num_recs * EEPROM_TABLE_RECORD_SIZE; - - __update_tbl_checksum(control, records, num, old_hdr_byte_sum); - - __update_table_header(control, buffs); - } else if (!__validate_tbl_checksum(control, records, num)) { - DRM_WARN("EEPROM Table checksum mismatch!"); - /* TODO Uncomment when EEPROM read/write is relliable */ - /* ret = -EIO; */ + s = *pos - strlen(tbl_hdr_str) - tbl_hdr_fmt_size - + strlen(rec_hdr_str); + s = s / rec_hdr_fmt_size; + r = *pos - strlen(tbl_hdr_str) - tbl_hdr_fmt_size - + strlen(rec_hdr_str); + r = r % rec_hdr_fmt_size; + + for ( ; size > 0 && s < control->ras_num_recs; s++) { + u32 ai = RAS_RI_TO_AI(control, s); + /* Read a single record + */ + res = __amdgpu_ras_eeprom_read(control, dare, ai, 1); + if (res) + goto Out; + __decode_table_record_from_buf(control, &record, dare); + snprintf(data, sizeof(data), rec_hdr_fmt, + s, + RAS_INDEX_TO_OFFSET(control, ai), + record_err_type_str[record.err_type], + record.bank, + record.ts, + record.offset, + record.mem_channel, + record.mcumc_id, + record.retired_page); + + data_len = min_t(size_t, rec_hdr_fmt_size - r, size); + if (copy_to_user(buf, &data[r], data_len)) { + res = -EFAULT; + goto Out; + } + buf += data_len; + size -= data_len; + *pos += data_len; + r = 0; + } } + res = 0; +Out: + mutex_unlock(&control->ras_tbl_mutex); + return res < 0 ? res : orig_size - size; +} -free_msgs: - kfree(msgs); +static ssize_t +amdgpu_ras_debugfs_eeprom_table_read(struct file *f, char __user *buf, + size_t size, loff_t *pos) +{ + struct amdgpu_device *adev = (struct amdgpu_device *)file_inode(f)->i_private; + struct amdgpu_ras *ras = amdgpu_ras_get_context(adev); + struct amdgpu_ras_eeprom_control *control = ras ? &ras->eeprom_control : NULL; + u8 data[81]; + int res; + + if (!size) + return size; + + if (!ras || !control) { + res = snprintf(data, sizeof(data), "Not supported\n"); + if (*pos >= res) + return 0; + + res -= *pos; + res = min_t(size_t, res, size); -free_buff: - kfree(buffs); + if (copy_to_user(buf, &data[*pos], res)) + return -EFAULT; - mutex_unlock(&control->tbl_mutex); + *pos += res; - return ret == num ? 0 : -EIO; + return res; + } else { + return amdgpu_ras_debugfs_table_read(f, buf, size, pos); + } } -inline uint32_t amdgpu_ras_eeprom_get_record_max_length(void) +const struct file_operations amdgpu_ras_debugfs_eeprom_table_ops = { + .owner = THIS_MODULE, + .read = amdgpu_ras_debugfs_eeprom_table_read, + .write = NULL, + .llseek = default_llseek, +}; + +/** + * __verify_ras_table_checksum -- verify the RAS EEPROM table checksum + * @control: pointer to control structure + * + * Check the checksum of the stored in EEPROM RAS table. + * + * Return 0 if the checksum is correct, + * positive if it is not correct, and + * -errno on I/O error. + */ +static int __verify_ras_table_checksum(struct amdgpu_ras_eeprom_control *control) { - return EEPROM_MAX_RECORD_NUM; + struct amdgpu_device *adev = to_amdgpu_device(control); + int buf_size, res; + u8 csum, *buf, *pp; + + buf_size = RAS_TABLE_HEADER_SIZE + + control->ras_num_recs * RAS_TABLE_RECORD_SIZE; + buf = kzalloc(buf_size, GFP_KERNEL); + if (!buf) { + DRM_ERROR("Out of memory checking RAS table checksum.\n"); + return -ENOMEM; + } + + res = amdgpu_eeprom_read(&adev->pm.smu_i2c, + control->i2c_address + + control->ras_header_offset, + buf, buf_size); + if (res < buf_size) { + DRM_ERROR("Partial read for checksum, res:%d\n", res); + /* On partial reads, return -EIO. + */ + if (res >= 0) + res = -EIO; + goto Out; + } + + csum = 0; + for (pp = buf; pp < buf + buf_size; pp++) + csum += *pp; +Out: + kfree(buf); + return res < 0 ? res : csum; } -/* Used for testing if bugs encountered */ -#if 0 -void amdgpu_ras_eeprom_test(struct amdgpu_ras_eeprom_control *control) +int amdgpu_ras_eeprom_init(struct amdgpu_ras_eeprom_control *control, + bool *exceed_err_limit) { - int i; - struct eeprom_table_record *recs = kcalloc(1, sizeof(*recs), GFP_KERNEL); + struct amdgpu_device *adev = to_amdgpu_device(control); + unsigned char buf[RAS_TABLE_HEADER_SIZE] = { 0 }; + struct amdgpu_ras_eeprom_table_header *hdr = &control->tbl_hdr; + struct amdgpu_ras *ras = amdgpu_ras_get_context(adev); + int res; + + *exceed_err_limit = false; + + if (!__is_ras_eeprom_supported(adev)) + return 0; - if (!recs) - return; + /* Verify i2c adapter is initialized */ + if (!adev->pm.smu_i2c.algo) + return -ENOENT; - for (i = 0; i < 1 ; i++) { - recs[i].address = 0xdeadbeef; - recs[i].retired_page = i; + if (!__get_eeprom_i2c_addr(adev, control)) + return -EINVAL; + + control->ras_header_offset = RAS_HDR_START; + control->ras_record_offset = RAS_RECORD_START; + control->ras_max_record_count = RAS_MAX_RECORD_COUNT; + mutex_init(&control->ras_tbl_mutex); + + /* Read the table header from EEPROM address */ + res = amdgpu_eeprom_read(&adev->pm.smu_i2c, + control->i2c_address + control->ras_header_offset, + buf, RAS_TABLE_HEADER_SIZE); + if (res < RAS_TABLE_HEADER_SIZE) { + DRM_ERROR("Failed to read EEPROM table header, res:%d", res); + return res >= 0 ? -EIO : res; } - if (!amdgpu_ras_eeprom_process_recods(control, recs, true, 1)) { + __decode_table_header_from_buf(hdr, buf); - memset(recs, 0, sizeof(*recs) * 1); + control->ras_num_recs = RAS_NUM_RECS(hdr); + control->ras_fri = RAS_OFFSET_TO_INDEX(control, hdr->first_rec_offset); - control->next_addr = EEPROM_RECORD_START; + if (hdr->header == RAS_TABLE_HDR_VAL) { + DRM_DEBUG_DRIVER("Found existing EEPROM table with %d records", + control->ras_num_recs); + res = __verify_ras_table_checksum(control); + if (res) + DRM_ERROR("RAS table incorrect checksum or error:%d\n", + res); + } else if (hdr->header == RAS_TABLE_HDR_BAD && + amdgpu_bad_page_threshold != 0) { + res = __verify_ras_table_checksum(control); + if (res) + DRM_ERROR("RAS Table incorrect checksum or error:%d\n", + res); + if (ras->bad_page_cnt_threshold > control->ras_num_recs) { + /* This means that, the threshold was increased since + * the last time the system was booted, and now, + * ras->bad_page_cnt_threshold - control->num_recs > 0, + * so that at least one more record can be saved, + * before the page count threshold is reached. + */ + dev_info(adev->dev, + "records:%d threshold:%d, resetting " + "RAS table header signature", + control->ras_num_recs, + ras->bad_page_cnt_threshold); + res = amdgpu_ras_eeprom_correct_header_tag(control, + RAS_TABLE_HDR_VAL); + } else { + *exceed_err_limit = true; + dev_err(adev->dev, + "RAS records:%d exceed threshold:%d, " + "maybe retire this GPU?", + control->ras_num_recs, ras->bad_page_cnt_threshold); + } + } else { + DRM_INFO("Creating a new EEPROM table"); - if (!amdgpu_ras_eeprom_process_recods(control, recs, false, 1)) { - for (i = 0; i < 1; i++) - DRM_INFO("rec.address :0x%llx, rec.retired_page :%llu", - recs[i].address, recs[i].retired_page); - } else - DRM_ERROR("Failed in reading from table"); + res = amdgpu_ras_eeprom_reset_table(control); + } - } else - DRM_ERROR("Failed in writing to table"); + return res < 0 ? res : 0; } -#endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.h index 178721170974..f95fc61b3021 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras_eeprom.h @@ -28,10 +28,11 @@ struct amdgpu_device; -enum amdgpu_ras_eeprom_err_type{ - AMDGPU_RAS_EEPROM_ERR_PLACE_HOLDER, +enum amdgpu_ras_eeprom_err_type { + AMDGPU_RAS_EEPROM_ERR_NA, AMDGPU_RAS_EEPROM_ERR_RECOVERABLE, - AMDGPU_RAS_EEPROM_ERR_NON_RECOVERABLE + AMDGPU_RAS_EEPROM_ERR_NON_RECOVERABLE, + AMDGPU_RAS_EEPROM_ERR_COUNT, }; struct amdgpu_ras_eeprom_table_header { @@ -40,15 +41,45 @@ struct amdgpu_ras_eeprom_table_header { uint32_t first_rec_offset; uint32_t tbl_size; uint32_t checksum; -}__attribute__((__packed__)); +} __packed; struct amdgpu_ras_eeprom_control { struct amdgpu_ras_eeprom_table_header tbl_hdr; - uint32_t next_addr; - unsigned int num_recs; - struct mutex tbl_mutex; - uint32_t tbl_byte_sum; - uint16_t i2c_address; // 8-bit represented address + + /* Base I2C EEPPROM 19-bit memory address, + * where the table is located. For more information, + * see top of amdgpu_eeprom.c. + */ + u32 i2c_address; + + /* The byte offset off of @i2c_address + * where the table header is found, + * and where the records start--always + * right after the header. + */ + u32 ras_header_offset; + u32 ras_record_offset; + + /* Number of records in the table. + */ + u32 ras_num_recs; + + /* First record index to read, 0-based. + * Range is [0, num_recs-1]. This is + * an absolute index, starting right after + * the table header. + */ + u32 ras_fri; + + /* Maximum possible number of records + * we could store, i.e. the maximum capacity + * of the table. + */ + u32 ras_max_record_count; + + /* Protect table access via this mutex. + */ + struct mutex ras_tbl_mutex; }; /* @@ -74,21 +105,26 @@ struct eeprom_table_record { unsigned char mem_channel; unsigned char mcumc_id; -}__attribute__((__packed__)); +} __packed; int amdgpu_ras_eeprom_init(struct amdgpu_ras_eeprom_control *control, - bool *exceed_err_limit); + bool *exceed_err_limit); + int amdgpu_ras_eeprom_reset_table(struct amdgpu_ras_eeprom_control *control); bool amdgpu_ras_eeprom_check_err_threshold(struct amdgpu_device *adev); -int amdgpu_ras_eeprom_process_recods(struct amdgpu_ras_eeprom_control *control, - struct eeprom_table_record *records, - bool write, - int num); +int amdgpu_ras_eeprom_read(struct amdgpu_ras_eeprom_control *control, + struct eeprom_table_record *records, const u32 num); + +int amdgpu_ras_eeprom_append(struct amdgpu_ras_eeprom_control *control, + struct eeprom_table_record *records, const u32 num); + +inline uint32_t amdgpu_ras_eeprom_max_record_count(void); -inline uint32_t amdgpu_ras_eeprom_get_record_max_length(void); +void amdgpu_ras_debugfs_set_ret_size(struct amdgpu_ras_eeprom_control *control); -void amdgpu_ras_eeprom_test(struct amdgpu_ras_eeprom_control *control); +extern const struct file_operations amdgpu_ras_debugfs_eeprom_size_ops; +extern const struct file_operations amdgpu_ras_debugfs_eeprom_table_ops; #endif // _AMDGPU_RAS_EEPROM_H diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h index e7d3d0dbdd96..27adffa7658d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ring.h @@ -107,8 +107,6 @@ struct amdgpu_fence_driver { }; int amdgpu_fence_driver_init(struct amdgpu_device *adev); -void amdgpu_fence_driver_fini_hw(struct amdgpu_device *adev); -void amdgpu_fence_driver_fini_sw(struct amdgpu_device *adev); void amdgpu_fence_driver_force_completion(struct amdgpu_ring *ring); int amdgpu_fence_driver_init_ring(struct amdgpu_ring *ring, @@ -117,8 +115,9 @@ int amdgpu_fence_driver_init_ring(struct amdgpu_ring *ring, int amdgpu_fence_driver_start_ring(struct amdgpu_ring *ring, struct amdgpu_irq_src *irq_src, unsigned irq_type); -void amdgpu_fence_driver_suspend(struct amdgpu_device *adev); -void amdgpu_fence_driver_resume(struct amdgpu_device *adev); +void amdgpu_fence_driver_hw_fini(struct amdgpu_device *adev); +void amdgpu_fence_driver_sw_fini(struct amdgpu_device *adev); +void amdgpu_fence_driver_hw_init(struct amdgpu_device *adev); int amdgpu_fence_emit(struct amdgpu_ring *ring, struct dma_fence **fence, unsigned flags); int amdgpu_fence_emit_polling(struct amdgpu_ring *ring, uint32_t *s, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_rlc.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_rlc.h index 7a4775ab6804..00afd0dcae86 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_rlc.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_rlc.h @@ -127,8 +127,8 @@ struct amdgpu_rlc_funcs { void (*reset)(struct amdgpu_device *adev); void (*start)(struct amdgpu_device *adev); void (*update_spm_vmid)(struct amdgpu_device *adev, unsigned vmid); - void (*rlcg_wreg)(struct amdgpu_device *adev, u32 offset, u32 v, u32 acc_flags, u32 hwip); - u32 (*rlcg_rreg)(struct amdgpu_device *adev, u32 offset, u32 acc_flags, u32 hwip); + void (*sriov_wreg)(struct amdgpu_device *adev, u32 offset, u32 v, u32 acc_flags, u32 hwip); + u32 (*sriov_rreg)(struct amdgpu_device *adev, u32 offset, u32 acc_flags, u32 hwip); bool (*is_rlcg_access_range)(struct amdgpu_device *adev, uint32_t reg); }; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c index acd95d3a4434..9fd5db58067d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -1396,6 +1396,41 @@ static bool amdgpu_ttm_bo_eviction_valuable(struct ttm_buffer_object *bo, return ttm_bo_eviction_valuable(bo, place); } +static void amdgpu_ttm_vram_mm_access(struct amdgpu_device *adev, loff_t pos, + void *buf, size_t size, bool write) +{ + while (size) { + uint64_t aligned_pos = ALIGN_DOWN(pos, 4); + uint64_t bytes = 4 - (pos & 0x3); + uint32_t shift = (pos & 0x3) * 8; + uint32_t mask = 0xffffffff << shift; + uint32_t value = 0; + + if (size < bytes) { + mask &= 0xffffffff >> (bytes - size) * 8; + bytes = size; + } + + if (mask != 0xffffffff) { + amdgpu_device_mm_access(adev, aligned_pos, &value, 4, false); + if (write) { + value &= ~mask; + value |= (*(uint32_t *)buf << shift) & mask; + amdgpu_device_mm_access(adev, aligned_pos, &value, 4, true); + } else { + value = (value & mask) >> shift; + memcpy(buf, &value, bytes); + } + } else { + amdgpu_device_mm_access(adev, aligned_pos, buf, 4, write); + } + + pos += bytes; + buf += bytes; + size -= bytes; + } +} + /** * amdgpu_ttm_access_memory - Read or Write memory that backs a buffer object. * @@ -1415,8 +1450,6 @@ static int amdgpu_ttm_access_memory(struct ttm_buffer_object *bo, struct amdgpu_bo *abo = ttm_to_amdgpu_bo(bo); struct amdgpu_device *adev = amdgpu_ttm_adev(abo->tbo.bdev); struct amdgpu_res_cursor cursor; - unsigned long flags; - uint32_t value = 0; int ret = 0; if (bo->resource->mem_type != TTM_PL_VRAM) @@ -1424,41 +1457,21 @@ static int amdgpu_ttm_access_memory(struct ttm_buffer_object *bo, amdgpu_res_first(bo->resource, offset, len, &cursor); while (cursor.remaining) { - uint64_t aligned_pos = cursor.start & ~(uint64_t)3; - uint64_t bytes = 4 - (cursor.start & 3); - uint32_t shift = (cursor.start & 3) * 8; - uint32_t mask = 0xffffffff << shift; - - if (cursor.size < bytes) { - mask &= 0xffffffff >> (bytes - cursor.size) * 8; - bytes = cursor.size; + size_t count, size = cursor.size; + loff_t pos = cursor.start; + + count = amdgpu_device_aper_access(adev, pos, buf, size, write); + size -= count; + if (size) { + /* using MM to access rest vram and handle un-aligned address */ + pos += count; + buf += count; + amdgpu_ttm_vram_mm_access(adev, pos, buf, size, write); } - if (mask != 0xffffffff) { - spin_lock_irqsave(&adev->mmio_idx_lock, flags); - WREG32_NO_KIQ(mmMM_INDEX, ((uint32_t)aligned_pos) | 0x80000000); - WREG32_NO_KIQ(mmMM_INDEX_HI, aligned_pos >> 31); - value = RREG32_NO_KIQ(mmMM_DATA); - if (write) { - value &= ~mask; - value |= (*(uint32_t *)buf << shift) & mask; - WREG32_NO_KIQ(mmMM_DATA, value); - } - spin_unlock_irqrestore(&adev->mmio_idx_lock, flags); - if (!write) { - value = (value & mask) >> shift; - memcpy(buf, &value, bytes); - } - } else { - bytes = cursor.size & ~0x3ULL; - amdgpu_device_vram_access(adev, cursor.start, - (uint32_t *)buf, bytes, - write); - } - - ret += bytes; - buf = (uint8_t *)buf + bytes; - amdgpu_res_next(&cursor, bytes); + ret += cursor.size; + buf += cursor.size; + amdgpu_res_next(&cursor, cursor.size); } return ret; @@ -2148,7 +2161,6 @@ static ssize_t amdgpu_ttm_vram_write(struct file *f, const char __user *buf, return -ENXIO; while (size) { - unsigned long flags; uint32_t value; if (*pos >= adev->gmc.mc_vram_size) @@ -2158,11 +2170,7 @@ static ssize_t amdgpu_ttm_vram_write(struct file *f, const char __user *buf, if (r) return r; - spin_lock_irqsave(&adev->mmio_idx_lock, flags); - WREG32_NO_KIQ(mmMM_INDEX, ((uint32_t)*pos) | 0x80000000); - WREG32_NO_KIQ(mmMM_INDEX_HI, *pos >> 31); - WREG32_NO_KIQ(mmMM_DATA, value); - spin_unlock_irqrestore(&adev->mmio_idx_lock, flags); + amdgpu_device_mm_access(adev, *pos, &value, 4, true); result += 4; buf += 4; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c index 2834981f8c08..5fdeceaa979f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c @@ -409,6 +409,12 @@ amdgpu_ucode_get_load_type(struct amdgpu_device *adev, int load_type) return AMDGPU_FW_LOAD_DIRECT; else return AMDGPU_FW_LOAD_PSP; + case CHIP_CYAN_SKILLFISH: + if (!(load_type && + adev->apu_flags & AMD_APU_IS_CYAN_SKILLFISH2)) + return AMDGPU_FW_LOAD_DIRECT; + else + return AMDGPU_FW_LOAD_PSP; default: DRM_ERROR("Unknown firmware load type\n"); } @@ -416,6 +422,84 @@ amdgpu_ucode_get_load_type(struct amdgpu_device *adev, int load_type) return AMDGPU_FW_LOAD_DIRECT; } +const char *amdgpu_ucode_name(enum AMDGPU_UCODE_ID ucode_id) +{ + switch (ucode_id) { + case AMDGPU_UCODE_ID_SDMA0: + return "SDMA0"; + case AMDGPU_UCODE_ID_SDMA1: + return "SDMA1"; + case AMDGPU_UCODE_ID_SDMA2: + return "SDMA2"; + case AMDGPU_UCODE_ID_SDMA3: + return "SDMA3"; + case AMDGPU_UCODE_ID_SDMA4: + return "SDMA4"; + case AMDGPU_UCODE_ID_SDMA5: + return "SDMA5"; + case AMDGPU_UCODE_ID_SDMA6: + return "SDMA6"; + case AMDGPU_UCODE_ID_SDMA7: + return "SDMA7"; + case AMDGPU_UCODE_ID_CP_CE: + return "CP_CE"; + case AMDGPU_UCODE_ID_CP_PFP: + return "CP_PFP"; + case AMDGPU_UCODE_ID_CP_ME: + return "CP_ME"; + case AMDGPU_UCODE_ID_CP_MEC1: + return "CP_MEC1"; + case AMDGPU_UCODE_ID_CP_MEC1_JT: + return "CP_MEC1_JT"; + case AMDGPU_UCODE_ID_CP_MEC2: + return "CP_MEC2"; + case AMDGPU_UCODE_ID_CP_MEC2_JT: + return "CP_MEC2_JT"; + case AMDGPU_UCODE_ID_CP_MES: + return "CP_MES"; + case AMDGPU_UCODE_ID_CP_MES_DATA: + return "CP_MES_DATA"; + case AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL: + return "RLC_RESTORE_LIST_CNTL"; + case AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM: + return "RLC_RESTORE_LIST_GPM_MEM"; + case AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM: + return "RLC_RESTORE_LIST_SRM_MEM"; + case AMDGPU_UCODE_ID_RLC_IRAM: + return "RLC_IRAM"; + case AMDGPU_UCODE_ID_RLC_DRAM: + return "RLC_DRAM"; + case AMDGPU_UCODE_ID_RLC_G: + return "RLC_G"; + case AMDGPU_UCODE_ID_STORAGE: + return "STORAGE"; + case AMDGPU_UCODE_ID_SMC: + return "SMC"; + case AMDGPU_UCODE_ID_UVD: + return "UVD"; + case AMDGPU_UCODE_ID_UVD1: + return "UVD1"; + case AMDGPU_UCODE_ID_VCE: + return "VCE"; + case AMDGPU_UCODE_ID_VCN: + return "VCN"; + case AMDGPU_UCODE_ID_VCN1: + return "VCN1"; + case AMDGPU_UCODE_ID_DMCU_ERAM: + return "DMCU_ERAM"; + case AMDGPU_UCODE_ID_DMCU_INTV: + return "DMCU_INTV"; + case AMDGPU_UCODE_ID_VCN0_RAM: + return "VCN0_RAM"; + case AMDGPU_UCODE_ID_VCN1_RAM: + return "VCN1_RAM"; + case AMDGPU_UCODE_ID_DMCUB: + return "DMCUB"; + default: + return "UNKNOWN UCODE"; + } +} + #define FW_VERSION_ATTR(name, mode, field) \ static ssize_t show_##name(struct device *dev, \ struct device_attribute *attr, \ @@ -440,7 +524,7 @@ FW_VERSION_ATTR(rlc_srlg_fw_version, 0444, gfx.rlc_srlg_fw_version); FW_VERSION_ATTR(rlc_srls_fw_version, 0444, gfx.rlc_srls_fw_version); FW_VERSION_ATTR(mec_fw_version, 0444, gfx.mec_fw_version); FW_VERSION_ATTR(mec2_fw_version, 0444, gfx.mec2_fw_version); -FW_VERSION_ATTR(sos_fw_version, 0444, psp.sos_fw_version); +FW_VERSION_ATTR(sos_fw_version, 0444, psp.sos.fw_version); FW_VERSION_ATTR(asd_fw_version, 0444, psp.asd_fw_version); FW_VERSION_ATTR(ta_ras_fw_version, 0444, psp.ta_ras_ucode_version); FW_VERSION_ATTR(ta_xgmi_fw_version, 0444, psp.ta_xgmi_ucode_version); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h index 270309e7f5f5..e2e2624ac653 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.h @@ -71,7 +71,7 @@ struct smc_firmware_header_v2_1 { uint32_t pptable_entry_offset; }; -struct psp_fw_bin_desc { +struct psp_fw_legacy_bin_desc { uint32_t fw_version; uint32_t offset_bytes; uint32_t size_bytes; @@ -80,30 +80,57 @@ struct psp_fw_bin_desc { /* version_major=1, version_minor=0 */ struct psp_firmware_header_v1_0 { struct common_firmware_header header; - struct psp_fw_bin_desc sos; + struct psp_fw_legacy_bin_desc sos; }; /* version_major=1, version_minor=1 */ struct psp_firmware_header_v1_1 { struct psp_firmware_header_v1_0 v1_0; - struct psp_fw_bin_desc toc; - struct psp_fw_bin_desc kdb; + struct psp_fw_legacy_bin_desc toc; + struct psp_fw_legacy_bin_desc kdb; }; /* version_major=1, version_minor=2 */ struct psp_firmware_header_v1_2 { struct psp_firmware_header_v1_0 v1_0; - struct psp_fw_bin_desc res; - struct psp_fw_bin_desc kdb; + struct psp_fw_legacy_bin_desc res; + struct psp_fw_legacy_bin_desc kdb; }; /* version_major=1, version_minor=3 */ struct psp_firmware_header_v1_3 { struct psp_firmware_header_v1_1 v1_1; - struct psp_fw_bin_desc spl; - struct psp_fw_bin_desc rl; - struct psp_fw_bin_desc sys_drv_aux; - struct psp_fw_bin_desc sos_aux; + struct psp_fw_legacy_bin_desc spl; + struct psp_fw_legacy_bin_desc rl; + struct psp_fw_legacy_bin_desc sys_drv_aux; + struct psp_fw_legacy_bin_desc sos_aux; +}; + +struct psp_fw_bin_desc { + uint32_t fw_type; + uint32_t fw_version; + uint32_t offset_bytes; + uint32_t size_bytes; +}; + +enum psp_fw_type { + PSP_FW_TYPE_UNKOWN, + PSP_FW_TYPE_PSP_SOS, + PSP_FW_TYPE_PSP_SYS_DRV, + PSP_FW_TYPE_PSP_KDB, + PSP_FW_TYPE_PSP_TOC, + PSP_FW_TYPE_PSP_SPL, + PSP_FW_TYPE_PSP_RL, + PSP_FW_TYPE_PSP_SOC_DRV, + PSP_FW_TYPE_PSP_INTF_DRV, + PSP_FW_TYPE_PSP_DBG_DRV, +}; + +/* version_major=2, version_minor=0 */ +struct psp_firmware_header_v2_0 { + struct common_firmware_header header; + uint32_t psp_fw_bin_count; + struct psp_fw_bin_desc psp_fw_bin[]; }; /* version_major=1, version_minor=0 */ @@ -138,18 +165,11 @@ enum ta_fw_type { TA_FW_TYPE_MAX_INDEX, }; -struct ta_fw_bin_desc { - uint32_t fw_type; - uint32_t fw_version; - uint32_t offset_bytes; - uint32_t size_bytes; -}; - /* version_major=2, version_minor=0 */ struct ta_firmware_header_v2_0 { struct common_firmware_header header; uint32_t ta_fw_bin_count; - struct ta_fw_bin_desc ta_fw_bin[]; + struct psp_fw_bin_desc ta_fw_bin[]; }; /* version_major=1, version_minor=0 */ @@ -312,6 +332,7 @@ union amdgpu_firmware_header { struct psp_firmware_header_v1_0 psp; struct psp_firmware_header_v1_1 psp_v1_1; struct psp_firmware_header_v1_3 psp_v1_3; + struct psp_firmware_header_v2_0 psp_v2_0; struct ta_firmware_header_v1_0 ta; struct ta_firmware_header_v2_0 ta_v2_0; struct gfx_firmware_header_v1_0 gfx; @@ -326,7 +347,7 @@ union amdgpu_firmware_header { uint8_t raw[0x100]; }; -#define UCODE_MAX_TA_PACKAGING ((sizeof(union amdgpu_firmware_header) - sizeof(struct common_firmware_header) - 4) / sizeof(struct ta_fw_bin_desc)) +#define UCODE_MAX_PSP_PACKAGING ((sizeof(union amdgpu_firmware_header) - sizeof(struct common_firmware_header) - 4) / sizeof(struct psp_fw_bin_desc)) /* * fw loading support @@ -449,4 +470,6 @@ void amdgpu_ucode_sysfs_fini(struct amdgpu_device *adev); enum amdgpu_firmware_load_type amdgpu_ucode_get_load_type(struct amdgpu_device *adev, int load_type); +const char *amdgpu_ucode_name(enum AMDGPU_UCODE_ID ucode_id); + #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c index f4489773715e..0c7c56a91b25 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c @@ -134,7 +134,7 @@ int amdgpu_umc_process_ras_data_cb(struct amdgpu_device *adev, amdgpu_ras_save_bad_pages(adev); if (adev->smu.ppt_funcs && adev->smu.ppt_funcs->send_hbm_bad_pages_num) - adev->smu.ppt_funcs->send_hbm_bad_pages_num(&adev->smu, con->eeprom_control.num_recs); + adev->smu.ppt_funcs->send_hbm_bad_pages_num(&adev->smu, con->eeprom_control.ras_num_recs); } amdgpu_ras_reset_gpu(adev); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c index b71dd1deeb2d..12a7cc2f01cd 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c @@ -531,7 +531,7 @@ static void amdgpu_virt_populate_vf2pf_ucode_info(struct amdgpu_device *adev) POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_RLC_SRLS, adev->gfx.rlc_srls_fw_version); POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_MEC, adev->gfx.mec_fw_version); POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_MEC2, adev->gfx.mec2_fw_version); - POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_SOS, adev->psp.sos_fw_version); + POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_SOS, adev->psp.sos.fw_version); POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_ASD, adev->psp.asd_fw_version); POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_TA_RAS, adev->psp.ta_ras_ucode_version); POPULATE_UCODE_INFO(vf2pf_info, AMD_SRIOV_UCODE_ID_TA_XGMI, adev->psp.ta_xgmi_ucode_version); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c index 078c068937fe..6362e861a3f5 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c @@ -88,6 +88,46 @@ struct amdgpu_prt_cb { struct dma_fence_cb cb; }; +/** + * amdgpu_vm_set_pasid - manage pasid and vm ptr mapping + * + * @adev: amdgpu_device pointer + * @vm: amdgpu_vm pointer + * @pasid: the pasid the VM is using on this GPU + * + * Set the pasid this VM is using on this GPU, can also be used to remove the + * pasid by passing in zero. + * + */ +int amdgpu_vm_set_pasid(struct amdgpu_device *adev, struct amdgpu_vm *vm, + u32 pasid) +{ + int r; + + if (vm->pasid == pasid) + return 0; + + if (vm->pasid) { + r = xa_err(xa_erase_irq(&adev->vm_manager.pasids, vm->pasid)); + if (r < 0) + return r; + + vm->pasid = 0; + } + + if (pasid) { + r = xa_err(xa_store_irq(&adev->vm_manager.pasids, pasid, vm, + GFP_KERNEL)); + if (r < 0) + return r; + + vm->pasid = pasid; + } + + + return 0; +} + /* * vm eviction_lock can be taken in MMU notifiers. Make sure no reclaim-FS * happens while holding this lock anywhere to prevent deadlocks when @@ -2863,14 +2903,13 @@ long amdgpu_vm_wait_idle(struct amdgpu_vm *vm, long timeout) * * @adev: amdgpu_device pointer * @vm: requested vm - * @pasid: Process address space identifier * * Init @vm fields. * * Returns: * 0 for success, error for failure. */ -int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, u32 pasid) +int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm) { struct amdgpu_bo *root_bo; struct amdgpu_bo_vm *root; @@ -2944,19 +2983,6 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, u32 pasid) amdgpu_bo_unreserve(vm->root.bo); - if (pasid) { - unsigned long flags; - - spin_lock_irqsave(&adev->vm_manager.pasid_lock, flags); - r = idr_alloc(&adev->vm_manager.pasid_idr, vm, pasid, pasid + 1, - GFP_ATOMIC); - spin_unlock_irqrestore(&adev->vm_manager.pasid_lock, flags); - if (r < 0) - goto error_free_root; - - vm->pasid = pasid; - } - INIT_KFIFO(vm->faults); return 0; @@ -3012,7 +3038,6 @@ static int amdgpu_vm_check_clean_reserved(struct amdgpu_device *adev, * * @adev: amdgpu_device pointer * @vm: requested vm - * @pasid: pasid to use * * This only works on GFX VMs that don't have any BOs added and no * page tables allocated yet. @@ -3020,7 +3045,6 @@ static int amdgpu_vm_check_clean_reserved(struct amdgpu_device *adev, * Changes the following VM parameters: * - use_cpu_for_update * - pte_supports_ats - * - pasid (old PASID is released, because compute manages its own PASIDs) * * Reinitializes the page directory to reflect the changed ATS * setting. @@ -3028,8 +3052,7 @@ static int amdgpu_vm_check_clean_reserved(struct amdgpu_device *adev, * Returns: * 0 for success, -errno for errors. */ -int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm, - u32 pasid) +int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm) { bool pte_support_ats = (adev->asic_type == CHIP_RAVEN); int r; @@ -3043,19 +3066,6 @@ int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm, if (r) goto unreserve_bo; - if (pasid) { - unsigned long flags; - - spin_lock_irqsave(&adev->vm_manager.pasid_lock, flags); - r = idr_alloc(&adev->vm_manager.pasid_idr, vm, pasid, pasid + 1, - GFP_ATOMIC); - spin_unlock_irqrestore(&adev->vm_manager.pasid_lock, flags); - - if (r == -ENOSPC) - goto unreserve_bo; - r = 0; - } - /* Check if PD needs to be reinitialized and do it before * changing any other state, in case it fails. */ @@ -3065,7 +3075,7 @@ int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm, to_amdgpu_bo_vm(vm->root.bo), false); if (r) - goto free_idr; + goto unreserve_bo; } /* Update VM state */ @@ -3082,7 +3092,7 @@ int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm, r = amdgpu_bo_sync_wait(vm->root.bo, AMDGPU_FENCE_OWNER_UNDEFINED, true); if (r) - goto free_idr; + goto unreserve_bo; vm->update_funcs = &amdgpu_vm_cpu_funcs; } else { @@ -3092,36 +3102,11 @@ int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm, vm->last_update = NULL; vm->is_compute_context = true; - if (vm->pasid) { - unsigned long flags; - - spin_lock_irqsave(&adev->vm_manager.pasid_lock, flags); - idr_remove(&adev->vm_manager.pasid_idr, vm->pasid); - spin_unlock_irqrestore(&adev->vm_manager.pasid_lock, flags); - - /* Free the original amdgpu allocated pasid - * Will be replaced with kfd allocated pasid - */ - amdgpu_pasid_free(vm->pasid); - vm->pasid = 0; - } - /* Free the shadow bo for compute VM */ amdgpu_bo_unref(&to_amdgpu_bo_vm(vm->root.bo)->shadow); - if (pasid) - vm->pasid = pasid; - goto unreserve_bo; -free_idr: - if (pasid) { - unsigned long flags; - - spin_lock_irqsave(&adev->vm_manager.pasid_lock, flags); - idr_remove(&adev->vm_manager.pasid_idr, pasid); - spin_unlock_irqrestore(&adev->vm_manager.pasid_lock, flags); - } unreserve_bo: amdgpu_bo_unreserve(vm->root.bo); return r; @@ -3137,14 +3122,7 @@ unreserve_bo: */ void amdgpu_vm_release_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm) { - if (vm->pasid) { - unsigned long flags; - - spin_lock_irqsave(&adev->vm_manager.pasid_lock, flags); - idr_remove(&adev->vm_manager.pasid_idr, vm->pasid); - spin_unlock_irqrestore(&adev->vm_manager.pasid_lock, flags); - } - vm->pasid = 0; + amdgpu_vm_set_pasid(adev, vm, 0); vm->is_compute_context = false; } @@ -3168,15 +3146,7 @@ void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm) root = amdgpu_bo_ref(vm->root.bo); amdgpu_bo_reserve(root, true); - if (vm->pasid) { - unsigned long flags; - - spin_lock_irqsave(&adev->vm_manager.pasid_lock, flags); - idr_remove(&adev->vm_manager.pasid_idr, vm->pasid); - spin_unlock_irqrestore(&adev->vm_manager.pasid_lock, flags); - vm->pasid = 0; - } - + amdgpu_vm_set_pasid(adev, vm, 0); dma_fence_wait(vm->last_unlocked, false); dma_fence_put(vm->last_unlocked); @@ -3258,8 +3228,7 @@ void amdgpu_vm_manager_init(struct amdgpu_device *adev) adev->vm_manager.vm_update_mode = 0; #endif - idr_init(&adev->vm_manager.pasid_idr); - spin_lock_init(&adev->vm_manager.pasid_lock); + xa_init_flags(&adev->vm_manager.pasids, XA_FLAGS_LOCK_IRQ); } /** @@ -3271,8 +3240,8 @@ void amdgpu_vm_manager_init(struct amdgpu_device *adev) */ void amdgpu_vm_manager_fini(struct amdgpu_device *adev) { - WARN_ON(!idr_is_empty(&adev->vm_manager.pasid_idr)); - idr_destroy(&adev->vm_manager.pasid_idr); + WARN_ON(!xa_empty(&adev->vm_manager.pasids)); + xa_destroy(&adev->vm_manager.pasids); amdgpu_vmid_mgr_fini(adev); } @@ -3341,13 +3310,13 @@ void amdgpu_vm_get_task_info(struct amdgpu_device *adev, u32 pasid, struct amdgpu_vm *vm; unsigned long flags; - spin_lock_irqsave(&adev->vm_manager.pasid_lock, flags); + xa_lock_irqsave(&adev->vm_manager.pasids, flags); - vm = idr_find(&adev->vm_manager.pasid_idr, pasid); + vm = xa_load(&adev->vm_manager.pasids, pasid); if (vm) *task_info = vm->task_info; - spin_unlock_irqrestore(&adev->vm_manager.pasid_lock, flags); + xa_unlock_irqrestore(&adev->vm_manager.pasids, flags); } /** @@ -3389,15 +3358,15 @@ bool amdgpu_vm_handle_fault(struct amdgpu_device *adev, u32 pasid, struct amdgpu_vm *vm; int r; - spin_lock_irqsave(&adev->vm_manager.pasid_lock, irqflags); - vm = idr_find(&adev->vm_manager.pasid_idr, pasid); + xa_lock_irqsave(&adev->vm_manager.pasids, irqflags); + vm = xa_load(&adev->vm_manager.pasids, pasid); if (vm) { root = amdgpu_bo_ref(vm->root.bo); is_compute_context = vm->is_compute_context; } else { root = NULL; } - spin_unlock_irqrestore(&adev->vm_manager.pasid_lock, irqflags); + xa_unlock_irqrestore(&adev->vm_manager.pasids, irqflags); if (!root) return false; @@ -3415,11 +3384,11 @@ bool amdgpu_vm_handle_fault(struct amdgpu_device *adev, u32 pasid, goto error_unref; /* Double check that the VM still exists */ - spin_lock_irqsave(&adev->vm_manager.pasid_lock, irqflags); - vm = idr_find(&adev->vm_manager.pasid_idr, pasid); + xa_lock_irqsave(&adev->vm_manager.pasids, irqflags); + vm = xa_load(&adev->vm_manager.pasids, pasid); if (vm && vm->root.bo != root) vm = NULL; - spin_unlock_irqrestore(&adev->vm_manager.pasid_lock, irqflags); + xa_unlock_irqrestore(&adev->vm_manager.pasids, irqflags); if (!vm) goto error_unlock; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h index f8fa653d4da7..ee0028ec2c4f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vm.h @@ -359,8 +359,7 @@ struct amdgpu_vm_manager { /* PASID to VM mapping, will be used in interrupt context to * look up VM of a page fault */ - struct idr pasid_idr; - spinlock_t pasid_lock; + struct xarray pasids; }; struct amdgpu_bo_va_mapping; @@ -375,9 +374,12 @@ extern const struct amdgpu_vm_update_funcs amdgpu_vm_sdma_funcs; void amdgpu_vm_manager_init(struct amdgpu_device *adev); void amdgpu_vm_manager_fini(struct amdgpu_device *adev); +int amdgpu_vm_set_pasid(struct amdgpu_device *adev, struct amdgpu_vm *vm, + u32 pasid); + long amdgpu_vm_wait_idle(struct amdgpu_vm *vm, long timeout); -int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, u32 pasid); -int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm, u32 pasid); +int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm); +int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm); void amdgpu_vm_release_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm); void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm); void amdgpu_vm_get_pd_bo(struct amdgpu_vm *vm, diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c index 8567d5d77346..258cf86b32f6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c @@ -486,6 +486,18 @@ int amdgpu_xgmi_get_hops_count(struct amdgpu_device *adev, return -EINVAL; } +int amdgpu_xgmi_get_num_links(struct amdgpu_device *adev, + struct amdgpu_device *peer_adev) +{ + struct psp_xgmi_topology_info *top = &adev->psp.xgmi_context.top_info; + int i; + + for (i = 0 ; i < top->num_nodes; ++i) + if (top->nodes[i].node_id == peer_adev->gmc.xgmi.node_id) + return top->nodes[i].num_links; + return -EINVAL; +} + int amdgpu_xgmi_add_device(struct amdgpu_device *adev) { struct psp_xgmi_topology_info *top_info; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h index 12969c0830d5..d2189bf7d428 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.h @@ -59,6 +59,8 @@ int amdgpu_xgmi_remove_device(struct amdgpu_device *adev); int amdgpu_xgmi_set_pstate(struct amdgpu_device *adev, int pstate); int amdgpu_xgmi_get_hops_count(struct amdgpu_device *adev, struct amdgpu_device *peer_adev); +int amdgpu_xgmi_get_num_links(struct amdgpu_device *adev, + struct amdgpu_device *peer_adev); uint64_t amdgpu_xgmi_get_relative_phy_addr(struct amdgpu_device *adev, uint64_t addr); static inline bool amdgpu_xgmi_same_hive(struct amdgpu_device *adev, diff --git a/drivers/gpu/drm/amd/amdgpu/cyan_skillfish_reg_init.c b/drivers/gpu/drm/amd/amdgpu/cyan_skillfish_reg_init.c new file mode 100644 index 000000000000..58808814d8fb --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/cyan_skillfish_reg_init.c @@ -0,0 +1,51 @@ +/* + * Copyright 2018 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#include "amdgpu.h" +#include "nv.h" + +#include "soc15_common.h" +#include "soc15_hw_ip.h" +#include "cyan_skillfish_ip_offset.h" + +int cyan_skillfish_reg_base_init(struct amdgpu_device *adev) +{ + /* HW has more IP blocks, only initialized the blocke needed by driver */ + uint32_t i; + for (i = 0 ; i < MAX_INSTANCE ; ++i) { + adev->reg_offset[GC_HWIP][i] = (uint32_t *)(&(GC_BASE.instance[i])); + adev->reg_offset[HDP_HWIP][i] = (uint32_t *)(&(HDP_BASE.instance[i])); + adev->reg_offset[MMHUB_HWIP][i] = (uint32_t *)(&(MMHUB_BASE.instance[i])); + adev->reg_offset[ATHUB_HWIP][i] = (uint32_t *)(&(ATHUB_BASE.instance[i])); + adev->reg_offset[NBIO_HWIP][i] = (uint32_t *)(&(NBIO_BASE.instance[i])); + adev->reg_offset[MP0_HWIP][i] = (uint32_t *)(&(MP0_BASE.instance[i])); + adev->reg_offset[MP1_HWIP][i] = (uint32_t *)(&(MP1_BASE.instance[i])); + adev->reg_offset[VCN_HWIP][i] = (uint32_t *)(&(UVD0_BASE.instance[i])); + adev->reg_offset[DF_HWIP][i] = (uint32_t *)(&(DF_BASE.instance[i])); + adev->reg_offset[DCE_HWIP][i] = (uint32_t *)(&(DMU_BASE.instance[i])); + adev->reg_offset[OSSSYS_HWIP][i] = (uint32_t *)(&(OSSSYS_BASE.instance[i])); + adev->reg_offset[SDMA0_HWIP][i] = (uint32_t *)(&(GC_BASE.instance[i])); + adev->reg_offset[SDMA1_HWIP][i] = (uint32_t *)(&(GC_BASE.instance[i])); + adev->reg_offset[SMUIO_HWIP][i] = (uint32_t *)(&(SMUIO_BASE.instance[i])); + } + return 0; +} diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c index a64b2c706090..24b0195fe665 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c @@ -56,6 +56,10 @@ #define GFX10_NUM_GFX_RINGS_Sienna_Cichlid 1 #define GFX10_MEC_HPD_SIZE 2048 +#define RLCG_VFGATE_DISABLED 0x4000000 +#define RLCG_WRONG_OPERATION_TYPE 0x2000000 +#define RLCG_NOT_IN_RANGE 0x1000000 + #define F32_CE_PROGRAM_RAM_SIZE 65536 #define RLCG_UCODE_LOADING_START_ADDRESS 0x00002000L @@ -181,6 +185,9 @@ #define GFX_RLCG_GC_READ (0x1 << 28) #define GFX_RLCG_MMHUB_WRITE (0x2 << 28) +#define RLCG_ERROR_REPORT_ENABLED(adev) \ + (amdgpu_sriov_reg_indirect_mmhub(adev) || amdgpu_sriov_reg_indirect_gc(adev)) + MODULE_FIRMWARE("amdgpu/navi10_ce.bin"); MODULE_FIRMWARE("amdgpu/navi10_pfp.bin"); MODULE_FIRMWARE("amdgpu/navi10_me.bin"); @@ -249,6 +256,39 @@ MODULE_FIRMWARE("amdgpu/yellow_carp_mec.bin"); MODULE_FIRMWARE("amdgpu/yellow_carp_mec2.bin"); MODULE_FIRMWARE("amdgpu/yellow_carp_rlc.bin"); +MODULE_FIRMWARE("amdgpu/cyan_skillfish_ce.bin"); +MODULE_FIRMWARE("amdgpu/cyan_skillfish_pfp.bin"); +MODULE_FIRMWARE("amdgpu/cyan_skillfish_me.bin"); +MODULE_FIRMWARE("amdgpu/cyan_skillfish_mec.bin"); +MODULE_FIRMWARE("amdgpu/cyan_skillfish_mec2.bin"); +MODULE_FIRMWARE("amdgpu/cyan_skillfish_rlc.bin"); + +MODULE_FIRMWARE("amdgpu/cyan_skillfish2_ce.bin"); +MODULE_FIRMWARE("amdgpu/cyan_skillfish2_pfp.bin"); +MODULE_FIRMWARE("amdgpu/cyan_skillfish2_me.bin"); +MODULE_FIRMWARE("amdgpu/cyan_skillfish2_mec.bin"); +MODULE_FIRMWARE("amdgpu/cyan_skillfish2_mec2.bin"); +MODULE_FIRMWARE("amdgpu/cyan_skillfish2_rlc.bin"); + +static const struct soc15_reg_golden golden_settings_gc_10_0[] = +{ + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGRBM_CAM_INDEX, 0xffffffff, 0x00000000), + /* TA_GRAD_ADJ_UCONFIG -> TA_GRAD_ADJ */ + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGRBM_CAM_DATA, 0xffffffff, 0x2544c382), + /* VGT_TF_RING_SIZE_UMD -> VGT_TF_RING_SIZE */ + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGRBM_CAM_DATA, 0xffffffff, 0x2262c24e), + /* VGT_HS_OFFCHIP_PARAM_UMD -> VGT_HS_OFFCHIP_PARAM */ + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGRBM_CAM_DATA, 0xffffffff, 0x226cc24f), + /* VGT_TF_MEMORY_BASE_UMD -> VGT_TF_MEMORY_BASE */ + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGRBM_CAM_DATA, 0xffffffff, 0x226ec250), + /* VGT_TF_MEMORY_BASE_HI_UMD -> VGT_TF_MEMORY_BASE_HI */ + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGRBM_CAM_DATA, 0xffffffff, 0x2278c261), + /* VGT_ESGS_RING_SIZE_UMD -> VGT_ESGS_RING_SIZE */ + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGRBM_CAM_DATA, 0xffffffff, 0x2232c240), + /* VGT_GSVS_RING_SIZE_UMD -> VGT_GSVS_RING_SIZE */ + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGRBM_CAM_DATA, 0xffffffff, 0x2233c241), +}; + static const struct soc15_reg_golden golden_settings_gc_10_1[] = { SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_HW_CONTROL_4, 0xffffffff, 0x00400014), @@ -1486,6 +1526,7 @@ static u32 gfx_v10_rlcg_rw(struct amdgpu_device *adev, u32 offset, u32 v, uint32 uint32_t i = 0; uint32_t retries = 50000; u32 ret = 0; + u32 tmp; scratch_reg0 = adev->rmmio + (adev->reg_offset[GC_HWIP][0][mmSCRATCH_REG0_BASE_IDX] + mmSCRATCH_REG0) * 4; @@ -1519,9 +1560,8 @@ static u32 gfx_v10_rlcg_rw(struct amdgpu_device *adev, u32 offset, u32 v, uint32 writel(v, scratch_reg0); writel(offset | flag, scratch_reg1); writel(1, spare_int); - for (i = 0; i < retries; i++) { - u32 tmp; + for (i = 0; i < retries; i++) { tmp = readl(scratch_reg1); if (!(tmp & flag)) break; @@ -1529,8 +1569,19 @@ static u32 gfx_v10_rlcg_rw(struct amdgpu_device *adev, u32 offset, u32 v, uint32 udelay(10); } - if (i >= retries) - pr_err("timeout: rlcg program reg:0x%05x failed !\n", offset); + if (i >= retries) { + if (RLCG_ERROR_REPORT_ENABLED(adev)) { + if (tmp & RLCG_VFGATE_DISABLED) + pr_err("The vfgate is disabled, program reg:0x%05x failed!\n", offset); + else if (tmp & RLCG_WRONG_OPERATION_TYPE) + pr_err("Wrong operation type, program reg:0x%05x failed!\n", offset); + else if (tmp & RLCG_NOT_IN_RANGE) + pr_err("The register is not in range, program reg:0x%05x failed!\n", offset); + else + pr_err("Unknown error type, program reg:0x%05x failed!\n", offset); + } else + pr_err("timeout: rlcg program reg:0x%05x failed!\n", offset); + } } ret = readl(scratch_reg0); @@ -1538,7 +1589,7 @@ static u32 gfx_v10_rlcg_rw(struct amdgpu_device *adev, u32 offset, u32 v, uint32 return ret; } -static void gfx_v10_rlcg_wreg(struct amdgpu_device *adev, u32 offset, u32 value, u32 acc_flags, u32 hwip) +static void gfx_v10_sriov_wreg(struct amdgpu_device *adev, u32 offset, u32 value, u32 acc_flags, u32 hwip) { u32 rlcg_flag; @@ -1554,7 +1605,7 @@ static void gfx_v10_rlcg_wreg(struct amdgpu_device *adev, u32 offset, u32 value, WREG32(offset, value); } -static u32 gfx_v10_rlcg_rreg(struct amdgpu_device *adev, u32 offset, u32 acc_flags, u32 hwip) +static u32 gfx_v10_sriov_rreg(struct amdgpu_device *adev, u32 offset, u32 acc_flags, u32 hwip) { u32 rlcg_flag; @@ -3488,12 +3539,51 @@ static const struct soc15_reg_golden golden_settings_gc_10_3_5[] = { SOC15_REG_GOLDEN_VALUE(GC, 0, mmUTCL1_CTRL, 0xffbfffff, 0x00a00000) }; +static const struct soc15_reg_golden golden_settings_gc_10_0_cyan_skillfish[] = { + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGRBM_GFX_INDEX, 0xffffffff, 0xe0000000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGE_FAST_CLKS, 0x3fffffff, 0x0000493e), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_CPF_CLK_CTRL, 0xfcff8fff, 0xf8000100), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmCGTT_SPI_CLK_CTRL, 0xff7f0fff, 0x3c000100), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_HW_CONTROL_3, 0xa0000000, 0xa0000000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmCB_HW_CONTROL_4, 0x00008000, 0x003c8014), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmCH_DRAM_BURST_CTRL, 0x00000010, 0x00000017), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmCH_PIPE_STEER, 0xffffffff, 0xd8d8d8d8), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmCH_VC5_ENABLE, 0x00000003, 0x00000003), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmCP_SD_CNTL, 0x800007ff, 0x000005ff), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmDB_DEBUG, 0xffffffff, 0x20000000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmDB_DEBUG3, 0xffffffff, 0x00000200), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmDB_DEBUG4, 0xffffffff, 0x04800000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmDB_LAST_OF_BURST_CONFIG, 0xffffffff, 0x03860210), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGB_ADDR_CONFIG, 0x0c1800ff, 0x00000044), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGCR_GENERAL_CNTL, 0x00009d00, 0x00008500), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGCMC_VM_CACHEABLE_DRAM_ADDRESS_END, 0xffffffff, 0x000fffff), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL1_DRAM_BURST_CTRL, 0x00000010, 0x00000017), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL1_PIPE_STEER, 0xfcfcfcfc, 0xd8d8d8d8), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2_PIPE_STEER_0, 0x77707770, 0x21302130), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2_PIPE_STEER_1, 0x77707770, 0x21302130), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2A_ADDR_MATCH_MASK, 0xffffffff, 0xffffffcf), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2C_ADDR_MATCH_MASK, 0xffffffff, 0xffffffcf), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2C_CGTT_SCLK_CTRL, 0x10000000, 0x10000100), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2C_CTRL2, 0xfc02002f, 0x9402002f), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmGL2C_CTRL3, 0x00002188, 0x00000188), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_ENHANCE, 0x08000009, 0x08000009), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_BINNER_EVENT_CNTL_0, 0xcc3fcc03, 0x842a4c02), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmPA_SC_LINE_STIPPLE_STATE, 0x0000000f, 0x00000000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmRMI_SPARE, 0xffff3109, 0xffff3101), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSQ_ARB_CONFIG, 0x00000100, 0x00000130), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSQ_LDS_CLK_CTRL, 0xffffffff, 0xffffffff), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmTA_CNTL_AUX, 0x00030008, 0x01030000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmUTCL1_CTRL, 0x00800000, 0x00800000) +}; + #define DEFAULT_SH_MEM_CONFIG \ ((SH_MEM_ADDRESS_MODE_64 << SH_MEM_CONFIG__ADDRESS_MODE__SHIFT) | \ (SH_MEM_ALIGNMENT_MODE_UNALIGNED << SH_MEM_CONFIG__ALIGNMENT_MODE__SHIFT) | \ (SH_MEM_RETRY_MODE_ALL << SH_MEM_CONFIG__RETRY_MODE__SHIFT) | \ (3 << SH_MEM_CONFIG__INITIAL_INST_PREFETCH__SHIFT)) +/* TODO: pending on golden setting value of gb address config */ +#define CYAN_SKILLFISH_GB_ADDR_CONFIG_GOLDEN 0x00100044 static void gfx_v10_0_set_ring_funcs(struct amdgpu_device *adev); static void gfx_v10_0_set_irq_funcs(struct amdgpu_device *adev); @@ -3718,6 +3808,14 @@ static void gfx_v10_0_init_golden_registers(struct amdgpu_device *adev) golden_settings_gc_10_3_5, (const u32)ARRAY_SIZE(golden_settings_gc_10_3_5)); break; + case CHIP_CYAN_SKILLFISH: + soc15_program_register_sequence(adev, + golden_settings_gc_10_0, + (const u32)ARRAY_SIZE(golden_settings_gc_10_0)); + soc15_program_register_sequence(adev, + golden_settings_gc_10_0_cyan_skillfish, + (const u32)ARRAY_SIZE(golden_settings_gc_10_0_cyan_skillfish)); + break; default: break; } @@ -3891,6 +3989,7 @@ static void gfx_v10_0_check_fw_write_wait(struct amdgpu_device *adev) case CHIP_NAVI10: case CHIP_NAVI12: case CHIP_NAVI14: + case CHIP_CYAN_SKILLFISH: if ((adev->gfx.me_fw_version >= 0x00000046) && (adev->gfx.me_feature_version >= 27) && (adev->gfx.pfp_fw_version >= 0x00000068) && @@ -4025,6 +4124,12 @@ static int gfx_v10_0_init_microcode(struct amdgpu_device *adev) case CHIP_YELLOW_CARP: chip_name = "yellow_carp"; break; + case CHIP_CYAN_SKILLFISH: + if (adev->apu_flags & AMD_APU_IS_CYAN_SKILLFISH2) + chip_name = "cyan_skillfish2"; + else + chip_name = "cyan_skillfish"; + break; default: BUG(); } @@ -4604,6 +4709,14 @@ static void gfx_v10_0_gpu_early_init(struct amdgpu_device *adev) adev->gfx.config.gb_addr_config_fields.num_pkrs = 1 << REG_GET_FIELD(gb_addr_config, GB_ADDR_CONFIG, NUM_PKRS); break; + case CHIP_CYAN_SKILLFISH: + adev->gfx.config.max_hw_contexts = 8; + adev->gfx.config.sc_prim_fifo_size_frontend = 0x20; + adev->gfx.config.sc_prim_fifo_size_backend = 0x100; + adev->gfx.config.sc_hiz_tile_fifo_size = 0x30; + adev->gfx.config.sc_earlyz_tile_fifo_size = 0x4C0; + gb_addr_config = CYAN_SKILLFISH_GB_ADDR_CONFIG_GOLDEN; + break; default: BUG(); break; @@ -4708,6 +4821,7 @@ static int gfx_v10_0_sw_init(void *handle) case CHIP_NAVI10: case CHIP_NAVI14: case CHIP_NAVI12: + case CHIP_CYAN_SKILLFISH: adev->gfx.me.num_me = 1; adev->gfx.me.num_pipe_per_me = 1; adev->gfx.me.num_queue_per_pipe = 1; @@ -5319,7 +5433,8 @@ static int gfx_v10_0_rlc_resume(struct amdgpu_device *adev) { int r; - if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP) { + if (adev->firmware.load_type == AMDGPU_FW_LOAD_PSP && + adev->psp.autoload_supported) { r = gfx_v10_0_wait_for_rlc_autoload_complete(adev); if (r) @@ -5379,7 +5494,7 @@ static int gfx_v10_0_parse_rlc_toc(struct amdgpu_device *adev) int ret; RLC_TABLE_OF_CONTENT *rlc_toc; - ret = amdgpu_bo_create_reserved(adev, adev->psp.toc_bin_size, PAGE_SIZE, + ret = amdgpu_bo_create_reserved(adev, adev->psp.toc.size_bytes, PAGE_SIZE, AMDGPU_GEM_DOMAIN_GTT, &adev->gfx.rlc.rlc_toc_bo, &adev->gfx.rlc.rlc_toc_gpu_addr, @@ -5390,7 +5505,7 @@ static int gfx_v10_0_parse_rlc_toc(struct amdgpu_device *adev) } /* Copy toc from psp sos fw to rlc toc buffer */ - memcpy(adev->gfx.rlc.rlc_toc_buf, adev->psp.toc_start_addr, adev->psp.toc_bin_size); + memcpy(adev->gfx.rlc.rlc_toc_buf, adev->psp.toc.start_addr, adev->psp.toc.size_bytes); rlc_toc = (RLC_TABLE_OF_CONTENT *)adev->gfx.rlc.rlc_toc_buf; while (rlc_toc && (rlc_toc->id > FIRMWARE_ID_INVALID) && @@ -7608,10 +7723,8 @@ static int gfx_v10_0_soft_reset(void *handle) static uint64_t gfx_v10_0_get_gpu_clock_counter(struct amdgpu_device *adev) { - uint64_t clock; + uint64_t clock, clock_lo, clock_hi, hi_check; - amdgpu_gfx_off_ctrl(adev, false); - mutex_lock(&adev->gfx.gpu_clock_mutex); switch (adev->asic_type) { case CHIP_VANGOGH: case CHIP_YELLOW_CARP: @@ -7619,12 +7732,21 @@ static uint64_t gfx_v10_0_get_gpu_clock_counter(struct amdgpu_device *adev) ((uint64_t)RREG32_SOC15(SMUIO, 0, mmGOLDEN_TSC_COUNT_UPPER_Vangogh) << 32ULL); break; default: - clock = (uint64_t)RREG32_SOC15(SMUIO, 0, mmGOLDEN_TSC_COUNT_LOWER) | - ((uint64_t)RREG32_SOC15(SMUIO, 0, mmGOLDEN_TSC_COUNT_UPPER) << 32ULL); + preempt_disable(); + clock_hi = RREG32_SOC15_NO_KIQ(SMUIO, 0, mmGOLDEN_TSC_COUNT_UPPER); + clock_lo = RREG32_SOC15_NO_KIQ(SMUIO, 0, mmGOLDEN_TSC_COUNT_LOWER); + hi_check = RREG32_SOC15_NO_KIQ(SMUIO, 0, mmGOLDEN_TSC_COUNT_UPPER); + /* The SMUIO TSC clock frequency is 100MHz, which sets 32-bit carry over + * roughly every 42 seconds. + */ + if (hi_check != clock_hi) { + clock_lo = RREG32_SOC15_NO_KIQ(SMUIO, 0, mmGOLDEN_TSC_COUNT_LOWER); + clock_hi = hi_check; + } + preempt_enable(); + clock = clock_lo | (clock_hi << 32ULL); break; } - mutex_unlock(&adev->gfx.gpu_clock_mutex); - amdgpu_gfx_off_ctrl(adev, true); return clock; } @@ -7665,6 +7787,7 @@ static int gfx_v10_0_early_init(void *handle) case CHIP_NAVI10: case CHIP_NAVI14: case CHIP_NAVI12: + case CHIP_CYAN_SKILLFISH: adev->gfx.num_gfx_rings = GFX10_NUM_GFX_RINGS_NV1X; break; case CHIP_SIENNA_CICHLID: @@ -8261,8 +8384,8 @@ static const struct amdgpu_rlc_funcs gfx_v10_0_rlc_funcs_sriov = { .reset = gfx_v10_0_rlc_reset, .start = gfx_v10_0_rlc_start, .update_spm_vmid = gfx_v10_0_update_spm_vmid, - .rlcg_wreg = gfx_v10_rlcg_wreg, - .rlcg_rreg = gfx_v10_rlcg_rreg, + .sriov_wreg = gfx_v10_sriov_wreg, + .sriov_rreg = gfx_v10_sriov_rreg, .is_rlcg_access_range = gfx_v10_0_is_rlcg_access_range, }; @@ -9425,6 +9548,7 @@ static void gfx_v10_0_set_rlc_funcs(struct amdgpu_device *adev) case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: case CHIP_YELLOW_CARP: + case CHIP_CYAN_SKILLFISH: adev->gfx.rlc.funcs = &gfx_v10_0_rlc_funcs; break; case CHIP_NAVI12: diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c index 044076ec1d03..03acc777adf7 100644 --- a/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v9_0.c @@ -787,7 +787,7 @@ static void gfx_v9_0_rlcg_w(struct amdgpu_device *adev, u32 offset, u32 v, u32 f } -static void gfx_v9_0_rlcg_wreg(struct amdgpu_device *adev, u32 offset, +static void gfx_v9_0_sriov_wreg(struct amdgpu_device *adev, u32 offset, u32 v, u32 acc_flags, u32 hwip) { if ((acc_flags & AMDGPU_REGS_RLC) && @@ -5131,7 +5131,7 @@ static const struct amdgpu_rlc_funcs gfx_v9_0_rlc_funcs = { .reset = gfx_v9_0_rlc_reset, .start = gfx_v9_0_rlc_start, .update_spm_vmid = gfx_v9_0_update_spm_vmid, - .rlcg_wreg = gfx_v9_0_rlcg_wreg, + .sriov_wreg = gfx_v9_0_sriov_wreg, .is_rlcg_access_range = gfx_v9_0_is_rlcg_access_range, }; diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c index 4523df2785d6..24b781e90bef 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v10_0.c @@ -810,6 +810,7 @@ static int gmc_v10_0_mc_init(struct amdgpu_device *adev) case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: case CHIP_YELLOW_CARP: + case CHIP_CYAN_SKILLFISH: default: adev->gmc.gart_size = 512ULL << 20; break; @@ -879,6 +880,7 @@ static int gmc_v10_0_sw_init(void *handle) case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: case CHIP_YELLOW_CARP: + case CHIP_CYAN_SKILLFISH: adev->num_vmhubs = 2; /* * To fulfill 4-level page support, @@ -996,6 +998,7 @@ static void gmc_v10_0_init_golden_registers(struct amdgpu_device *adev) case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: case CHIP_YELLOW_CARP: + case CHIP_CYAN_SKILLFISH: break; default: break; diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c index 7eb70d69f760..7cf653f9e9a7 100644 --- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c +++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c @@ -53,6 +53,7 @@ #include "mmhub_v1_7.h" #include "umc_v6_1.h" #include "umc_v6_0.h" +#include "umc_v6_7.h" #include "hdp_v4_0.h" #include "ivsrcid/vmc/irqsrcs_vmc_1_0.h" @@ -1168,6 +1169,18 @@ static void gmc_v9_0_set_umc_funcs(struct amdgpu_device *adev) adev->umc.channel_idx_tbl = &umc_v6_1_channel_idx_tbl[0][0]; adev->umc.ras_funcs = &umc_v6_1_ras_funcs; break; + case CHIP_ALDEBARAN: + adev->umc.max_ras_err_cnt_per_query = UMC_V6_7_TOTAL_CHANNEL_NUM; + adev->umc.channel_inst_num = UMC_V6_7_UMC_INSTANCE_NUM; + adev->umc.umc_inst_num = UMC_V6_7_CHANNEL_INSTANCE_NUM; + adev->umc.channel_offs = UMC_V6_7_PER_CHANNEL_OFFSET; + if (!adev->gmc.xgmi.connected_to_cpu) + adev->umc.ras_funcs = &umc_v6_7_ras_funcs; + if (1 & adev->smuio.funcs->get_die_id(adev)) + adev->umc.channel_idx_tbl = &umc_v6_7_channel_idx_tbl_first[0][0]; + else + adev->umc.channel_idx_tbl = &umc_v6_7_channel_idx_tbl_second[0][0]; + break; default: break; } diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v2_3.c b/drivers/gpu/drm/amd/amdgpu/nbio_v2_3.c index 7b79eeaa88aa..b184b656b9b6 100644 --- a/drivers/gpu/drm/amd/amdgpu/nbio_v2_3.c +++ b/drivers/gpu/drm/amd/amdgpu/nbio_v2_3.c @@ -508,6 +508,26 @@ static void nbio_v2_3_apply_l1_link_width_reconfig_wa(struct amdgpu_device *adev WREG32_PCIE(smnPCIE_LC_LINK_WIDTH_CNTL, reg_data); } +static void nbio_v2_3_clear_doorbell_interrupt(struct amdgpu_device *adev) +{ + uint32_t reg, reg_data; + + if (adev->asic_type != CHIP_SIENNA_CICHLID) + return; + + reg = RREG32_SOC15(NBIO, 0, mmBIF_RB_CNTL); + + /* Clear Interrupt Status + */ + if ((reg & BIF_RB_CNTL__RB_ENABLE_MASK) == 0) { + reg = RREG32_SOC15(NBIO, 0, mmBIF_DOORBELL_INT_CNTL); + if (reg & BIF_DOORBELL_INT_CNTL__DOORBELL_INTERRUPT_STATUS_MASK) { + reg_data = 1 << BIF_DOORBELL_INT_CNTL__DOORBELL_INTERRUPT_CLEAR__SHIFT; + WREG32_SOC15(NBIO, 0, mmBIF_DOORBELL_INT_CNTL, reg_data); + } + } +} + const struct amdgpu_nbio_funcs nbio_v2_3_funcs = { .get_hdp_flush_req_offset = nbio_v2_3_get_hdp_flush_req_offset, .get_hdp_flush_done_offset = nbio_v2_3_get_hdp_flush_done_offset, @@ -531,4 +551,5 @@ const struct amdgpu_nbio_funcs nbio_v2_3_funcs = { .program_aspm = nbio_v2_3_program_aspm, .apply_lc_spc_mode_wa = nbio_v2_3_apply_lc_spc_mode_wa, .apply_l1_link_width_reconfig_wa = nbio_v2_3_apply_l1_link_width_reconfig_wa, + .clear_doorbell_interrupt = nbio_v2_3_clear_doorbell_interrupt, }; diff --git a/drivers/gpu/drm/amd/amdgpu/nv.c b/drivers/gpu/drm/amd/amdgpu/nv.c index 94d029dbf30d..436fb13e32f0 100644 --- a/drivers/gpu/drm/amd/amdgpu/nv.c +++ b/drivers/gpu/drm/amd/amdgpu/nv.c @@ -666,6 +666,9 @@ legacy_init: case CHIP_YELLOW_CARP: yellow_carp_reg_base_init(adev); break; + case CHIP_CYAN_SKILLFISH: + cyan_skillfish_reg_base_init(adev); + break; default: return -EINVAL; } @@ -682,7 +685,10 @@ int nv_set_ip_blocks(struct amdgpu_device *adev) { int r; - if (adev->flags & AMD_IS_APU) { + if (adev->asic_type == CHIP_CYAN_SKILLFISH) { + adev->nbio.funcs = &nbio_v2_3_funcs; + adev->nbio.hdp_flush_reg = &nbio_v2_3_hdp_flush_reg; + } else if (adev->flags & AMD_IS_APU) { adev->nbio.funcs = &nbio_v7_2_funcs; adev->nbio.hdp_flush_reg = &nbio_v7_2_hdp_flush_reg; } else { @@ -889,6 +895,20 @@ int nv_set_ip_blocks(struct amdgpu_device *adev) amdgpu_device_ip_block_add(adev, &vcn_v3_0_ip_block); amdgpu_device_ip_block_add(adev, &jpeg_v3_0_ip_block); break; + case CHIP_CYAN_SKILLFISH: + amdgpu_device_ip_block_add(adev, &nv_common_ip_block); + amdgpu_device_ip_block_add(adev, &gmc_v10_0_ip_block); + amdgpu_device_ip_block_add(adev, &navi10_ih_ip_block); + if (adev->apu_flags & AMD_APU_IS_CYAN_SKILLFISH2) { + if (likely(adev->firmware.load_type == AMDGPU_FW_LOAD_PSP)) + amdgpu_device_ip_block_add(adev, &psp_v11_0_8_ip_block); + amdgpu_device_ip_block_add(adev, &smu_v11_0_ip_block); + } + if (adev->enable_virtual_display || amdgpu_sriov_vf(adev)) + amdgpu_device_ip_block_add(adev, &dce_virtual_ip_block); + amdgpu_device_ip_block_add(adev, &gfx_v10_0_ip_block); + amdgpu_device_ip_block_add(adev, &sdma_v5_0_ip_block); + break; default: return -EINVAL; } @@ -1241,6 +1261,11 @@ static int nv_common_early_init(void *handle) else adev->external_rev_id = adev->rev_id + 0x01; break; + case CHIP_CYAN_SKILLFISH: + adev->cg_flags = 0; + adev->pg_flags = 0; + adev->external_rev_id = adev->rev_id + 0x82; + break; default: /* FIXME: not supported yet */ return -EINVAL; diff --git a/drivers/gpu/drm/amd/amdgpu/nv.h b/drivers/gpu/drm/amd/amdgpu/nv.h index eb9aff1e7253..1f40ba3b0460 100644 --- a/drivers/gpu/drm/amd/amdgpu/nv.h +++ b/drivers/gpu/drm/amd/amdgpu/nv.h @@ -38,5 +38,6 @@ void vangogh_reg_base_init(struct amdgpu_device *adev); int dimgrey_cavefish_reg_base_init(struct amdgpu_device *adev); int beige_goby_reg_base_init(struct amdgpu_device *adev); int yellow_carp_reg_base_init(struct amdgpu_device *adev); +int cyan_skillfish_reg_base_init(struct amdgpu_device *adev); #endif diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c index bc133db2d538..8862684f8b43 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/psp_v11_0.c @@ -80,6 +80,9 @@ MODULE_FIRMWARE("amdgpu/beige_goby_ta.bin"); /* For large FW files the time to complete can be very long */ #define USBC_PD_POLLING_LIMIT_S 240 +/* Read USB-PD from LFB */ +#define GFX_CMD_USB_PD_USE_LFB 0x480 + static int psp_v11_0_init_microcode(struct psp_context *psp) { struct amdgpu_device *adev = psp->adev; @@ -284,7 +287,7 @@ static int psp_v11_0_bootloader_load_kdb(struct psp_context *psp) return ret; /* Copy PSP KDB binary to memory */ - psp_copy_fw(psp, psp->kdb_start_addr, psp->kdb_bin_size); + psp_copy_fw(psp, psp->kdb.start_addr, psp->kdb.size_bytes); /* Provide the PSP KDB to bootloader */ WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_36, @@ -315,7 +318,7 @@ static int psp_v11_0_bootloader_load_spl(struct psp_context *psp) return ret; /* Copy PSP SPL binary to memory */ - psp_copy_fw(psp, psp->spl_start_addr, psp->spl_bin_size); + psp_copy_fw(psp, psp->spl.start_addr, psp->spl.size_bytes); /* Provide the PSP SPL to bootloader */ WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_36, @@ -346,7 +349,7 @@ static int psp_v11_0_bootloader_load_sysdrv(struct psp_context *psp) return ret; /* Copy PSP System Driver binary to memory */ - psp_copy_fw(psp, psp->sys_start_addr, psp->sys_bin_size); + psp_copy_fw(psp, psp->sys.start_addr, psp->sys.size_bytes); /* Provide the sys driver to bootloader */ WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_36, @@ -380,7 +383,7 @@ static int psp_v11_0_bootloader_load_sos(struct psp_context *psp) return ret; /* Copy Secure OS binary to PSP memory */ - psp_copy_fw(psp, psp->sos_start_addr, psp->sos_bin_size); + psp_copy_fw(psp, psp->sos.start_addr, psp->sos.size_bytes); /* Provide the PSP secure OS to bootloader */ WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_36, @@ -753,44 +756,26 @@ static void psp_v11_0_ring_set_wptr(struct psp_context *psp, uint32_t value) WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_67, value); } -static int psp_v11_0_load_usbc_pd_fw(struct psp_context *psp, dma_addr_t dma_addr) +static int psp_v11_0_load_usbc_pd_fw(struct psp_context *psp, uint64_t fw_pri_mc_addr) { struct amdgpu_device *adev = psp->adev; uint32_t reg_status; int ret, i = 0; - /* Write lower 32-bit address of the PD Controller FW */ - WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_36, lower_32_bits(dma_addr)); - ret = psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_35), - 0x80000000, 0x80000000, false); - if (ret) - return ret; - - /* Fireup interrupt so PSP can pick up the lower address */ - WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_35, 0x800000); - ret = psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_35), - 0x80000000, 0x80000000, false); - if (ret) - return ret; - - reg_status = RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_35); - - if ((reg_status & 0xFFFF) != 0) { - DRM_ERROR("Lower address load failed - MP0_SMN_C2PMSG_35.Bits [15:0] = %02x...\n", - reg_status & 0xFFFF); - return -EIO; - } - - /* Write upper 32-bit address of the PD Controller FW */ - WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_36, upper_32_bits(dma_addr)); + /* + * LFB address which is aligned to 1MB address and has to be + * right-shifted by 20 so that LFB address can be passed on a 32-bit C2P + * register + */ + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_36, (fw_pri_mc_addr >> 20)); ret = psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_35), 0x80000000, 0x80000000, false); if (ret) return ret; - /* Fireup interrupt so PSP can pick up the upper address */ - WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_35, 0x4000000); + /* Fireup interrupt so PSP can pick up the address */ + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_35, (GFX_CMD_USB_PD_USE_LFB << 16)); /* FW load takes very long time */ do { @@ -806,7 +791,7 @@ static int psp_v11_0_load_usbc_pd_fw(struct psp_context *psp, dma_addr_t dma_add done: if ((reg_status & 0xFFFF) != 0) { - DRM_ERROR("Upper address load failed - MP0_SMN_C2PMSG_35.Bits [15:0] = x%04x\n", + DRM_ERROR("Address load failed - MP0_SMN_C2PMSG_35.Bits [15:0] = 0x%04x\n", reg_status & 0xFFFF); return -EIO; } diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v11_0_8.c b/drivers/gpu/drm/amd/amdgpu/psp_v11_0_8.c new file mode 100644 index 000000000000..ff13e1beb49b --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/psp_v11_0_8.c @@ -0,0 +1,208 @@ +/* + * Copyright 2021 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#include "amdgpu.h" +#include "amdgpu_psp.h" +#include "amdgpu_ucode.h" +#include "soc15_common.h" +#include "psp_v11_0_8.h" + +#include "mp/mp_11_0_8_offset.h" + +static int psp_v11_0_8_ring_init(struct psp_context *psp, + enum psp_ring_type ring_type) +{ + int ret = 0; + struct psp_ring *ring; + struct amdgpu_device *adev = psp->adev; + + ring = &psp->km_ring; + + ring->ring_type = ring_type; + + /* allocate 4k Page of Local Frame Buffer memory for ring */ + ring->ring_size = 0x1000; + ret = amdgpu_bo_create_kernel(adev, ring->ring_size, PAGE_SIZE, + AMDGPU_GEM_DOMAIN_VRAM, + &adev->firmware.rbuf, + &ring->ring_mem_mc_addr, + (void **)&ring->ring_mem); + if (ret) { + ring->ring_size = 0; + return ret; + } + + return 0; +} + +static int psp_v11_0_8_ring_stop(struct psp_context *psp, + enum psp_ring_type ring_type) +{ + int ret = 0; + struct amdgpu_device *adev = psp->adev; + + if (amdgpu_sriov_vf(adev)) { + /* Write the ring destroy command*/ + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_101, + GFX_CTRL_CMD_ID_DESTROY_GPCOM_RING); + /* there might be handshake issue with hardware which needs delay */ + mdelay(20); + /* Wait for response flag (bit 31) */ + ret = psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_101), + 0x80000000, 0x80000000, false); + } else { + /* Write the ring destroy command*/ + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_64, + GFX_CTRL_CMD_ID_DESTROY_RINGS); + /* there might be handshake issue with hardware which needs delay */ + mdelay(20); + /* Wait for response flag (bit 31) */ + ret = psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_64), + 0x80000000, 0x80000000, false); + } + + return ret; +} + +static int psp_v11_0_8_ring_create(struct psp_context *psp, + enum psp_ring_type ring_type) +{ + int ret = 0; + unsigned int psp_ring_reg = 0; + struct psp_ring *ring = &psp->km_ring; + struct amdgpu_device *adev = psp->adev; + + if (amdgpu_sriov_vf(adev)) { + ret = psp_v11_0_8_ring_stop(psp, ring_type); + if (ret) { + DRM_ERROR("psp_v11_0_8_ring_stop_sriov failed!\n"); + return ret; + } + + /* Write low address of the ring to C2PMSG_102 */ + psp_ring_reg = lower_32_bits(ring->ring_mem_mc_addr); + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_102, psp_ring_reg); + /* Write high address of the ring to C2PMSG_103 */ + psp_ring_reg = upper_32_bits(ring->ring_mem_mc_addr); + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_103, psp_ring_reg); + + /* Write the ring initialization command to C2PMSG_101 */ + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_101, + GFX_CTRL_CMD_ID_INIT_GPCOM_RING); + + /* there might be handshake issue with hardware which needs delay */ + mdelay(20); + + /* Wait for response flag (bit 31) in C2PMSG_101 */ + ret = psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_101), + 0x80000000, 0x8000FFFF, false); + + } else { + /* Wait for sOS ready for ring creation */ + ret = psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_64), + 0x80000000, 0x80000000, false); + if (ret) { + DRM_ERROR("Failed to wait for trust OS ready for ring creation\n"); + return ret; + } + + /* Write low address of the ring to C2PMSG_69 */ + psp_ring_reg = lower_32_bits(ring->ring_mem_mc_addr); + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_69, psp_ring_reg); + /* Write high address of the ring to C2PMSG_70 */ + psp_ring_reg = upper_32_bits(ring->ring_mem_mc_addr); + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_70, psp_ring_reg); + /* Write size of ring to C2PMSG_71 */ + psp_ring_reg = ring->ring_size; + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_71, psp_ring_reg); + /* Write the ring initialization command to C2PMSG_64 */ + psp_ring_reg = ring_type; + psp_ring_reg = psp_ring_reg << 16; + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_64, psp_ring_reg); + + /* there might be handshake issue with hardware which needs delay */ + mdelay(20); + + /* Wait for response flag (bit 31) in C2PMSG_64 */ + ret = psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, mmMP0_SMN_C2PMSG_64), + 0x80000000, 0x8000FFFF, false); + } + + return ret; +} + +static int psp_v11_0_8_ring_destroy(struct psp_context *psp, + enum psp_ring_type ring_type) +{ + int ret = 0; + struct psp_ring *ring = &psp->km_ring; + struct amdgpu_device *adev = psp->adev; + + ret = psp_v11_0_8_ring_stop(psp, ring_type); + if (ret) + DRM_ERROR("Fail to stop psp ring\n"); + + amdgpu_bo_free_kernel(&adev->firmware.rbuf, + &ring->ring_mem_mc_addr, + (void **)&ring->ring_mem); + + return ret; +} + +static uint32_t psp_v11_0_8_ring_get_wptr(struct psp_context *psp) +{ + uint32_t data; + struct amdgpu_device *adev = psp->adev; + + if (amdgpu_sriov_vf(adev)) + data = RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_102); + else + data = RREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_67); + + return data; +} + +static void psp_v11_0_8_ring_set_wptr(struct psp_context *psp, uint32_t value) +{ + struct amdgpu_device *adev = psp->adev; + + if (amdgpu_sriov_vf(adev)) { + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_102, value); + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_101, + GFX_CTRL_CMD_ID_CONSUME_CMD); + } else + WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_67, value); +} + +static const struct psp_funcs psp_v11_0_8_funcs = { + .ring_init = psp_v11_0_8_ring_init, + .ring_create = psp_v11_0_8_ring_create, + .ring_stop = psp_v11_0_8_ring_stop, + .ring_destroy = psp_v11_0_8_ring_destroy, + .ring_get_wptr = psp_v11_0_8_ring_get_wptr, + .ring_set_wptr = psp_v11_0_8_ring_set_wptr, +}; + +void psp_v11_0_8_set_psp_funcs(struct psp_context *psp) +{ + psp->funcs = &psp_v11_0_8_funcs; +} diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v11_0_8.h b/drivers/gpu/drm/amd/amdgpu/psp_v11_0_8.h new file mode 100644 index 000000000000..890377a5afe0 --- /dev/null +++ b/drivers/gpu/drm/amd/amdgpu/psp_v11_0_8.h @@ -0,0 +1,30 @@ +/* + * Copyright 2021 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#ifndef __PSP_V11_0_8_H__ +#define __PSP_V11_0_8_H__ + +#include "amdgpu_psp.h" + +void psp_v11_0_8_set_psp_funcs(struct psp_context *psp); + +#endif diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v12_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v12_0.c index 618e5b6b85d9..0c908d4566e8 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v12_0.c +++ b/drivers/gpu/drm/amd/amdgpu/psp_v12_0.c @@ -67,7 +67,7 @@ static int psp_v12_0_init_microcode(struct psp_context *psp) err = psp_init_asd_microcode(psp, chip_name); if (err) - goto out; + return err; snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_ta.bin", chip_name); err = request_firmware(&adev->psp.ta_fw, fw_name, adev->dev); @@ -80,7 +80,7 @@ static int psp_v12_0_init_microcode(struct psp_context *psp) } else { err = amdgpu_ucode_validate(adev->psp.ta_fw); if (err) - goto out2; + goto out; ta_hdr = (const struct ta_firmware_header_v1_0 *) adev->psp.ta_fw->data; @@ -105,10 +105,9 @@ static int psp_v12_0_init_microcode(struct psp_context *psp) return 0; -out2: +out: release_firmware(adev->psp.ta_fw); adev->psp.ta_fw = NULL; -out: if (err) { dev_err(adev->dev, "psp v12.0: Failed to load firmware \"%s\"\n", @@ -139,7 +138,7 @@ static int psp_v12_0_bootloader_load_sysdrv(struct psp_context *psp) return ret; /* Copy PSP System Driver binary to memory */ - psp_copy_fw(psp, psp->sys_start_addr, psp->sys_bin_size); + psp_copy_fw(psp, psp->sys.start_addr, psp->sys.size_bytes); /* Provide the sys driver to bootloader */ WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_36, @@ -178,7 +177,7 @@ static int psp_v12_0_bootloader_load_sos(struct psp_context *psp) return ret; /* Copy Secure OS binary to PSP memory */ - psp_copy_fw(psp, psp->sos_start_addr, psp->sos_bin_size); + psp_copy_fw(psp, psp->sos.start_addr, psp->sos.size_bytes); /* Provide the PSP secure OS to bootloader */ WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_36, diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v13_0.c b/drivers/gpu/drm/amd/amdgpu/psp_v13_0.c index c8949add88f3..47a500f64db2 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v13_0.c +++ b/drivers/gpu/drm/amd/amdgpu/psp_v13_0.c @@ -35,6 +35,12 @@ MODULE_FIRMWARE("amdgpu/yellow_carp_asd.bin"); MODULE_FIRMWARE("amdgpu/yellow_carp_toc.bin"); MODULE_FIRMWARE("amdgpu/yellow_carp_ta.bin"); +/* For large FW files the time to complete can be very long */ +#define USBC_PD_POLLING_LIMIT_S 240 + +/* Read USB-PD from LFB */ +#define GFX_CMD_USB_PD_USE_LFB 0x480 + static int psp_v13_0_init_microcode(struct psp_context *psp) { struct amdgpu_device *adev = psp->adev; @@ -111,7 +117,9 @@ static int psp_v13_0_wait_for_bootloader(struct psp_context *psp) return ret; } -static int psp_v13_0_bootloader_load_kdb(struct psp_context *psp) +static int psp_v13_0_bootloader_load_component(struct psp_context *psp, + struct psp_bin_desc *bin_desc, + enum psp_bootloader_cmd bl_cmd) { int ret; uint32_t psp_gfxdrv_command_reg = 0; @@ -130,12 +138,12 @@ static int psp_v13_0_bootloader_load_kdb(struct psp_context *psp) memset(psp->fw_pri_buf, 0, PSP_1_MEG); /* Copy PSP KDB binary to memory */ - memcpy(psp->fw_pri_buf, psp->kdb_start_addr, psp->kdb_bin_size); + memcpy(psp->fw_pri_buf, bin_desc->start_addr, bin_desc->size_bytes); /* Provide the PSP KDB to bootloader */ WREG32_SOC15(MP0, 0, regMP0_SMN_C2PMSG_36, (uint32_t)(psp->fw_pri_mc_addr >> 20)); - psp_gfxdrv_command_reg = PSP_BL__LOAD_KEY_DATABASE; + psp_gfxdrv_command_reg = bl_cmd; WREG32_SOC15(MP0, 0, regMP0_SMN_C2PMSG_35, psp_gfxdrv_command_reg); @@ -144,40 +152,29 @@ static int psp_v13_0_bootloader_load_kdb(struct psp_context *psp) return ret; } -static int psp_v13_0_bootloader_load_sysdrv(struct psp_context *psp) +static int psp_v13_0_bootloader_load_kdb(struct psp_context *psp) { - int ret; - uint32_t psp_gfxdrv_command_reg = 0; - struct amdgpu_device *adev = psp->adev; - - /* Check sOS sign of life register to confirm sys driver and sOS - * are already been loaded. - */ - if (psp_v13_0_is_sos_alive(psp)) - return 0; - - ret = psp_v13_0_wait_for_bootloader(psp); - if (ret) - return ret; - - memset(psp->fw_pri_buf, 0, PSP_1_MEG); - - /* Copy PSP System Driver binary to memory */ - memcpy(psp->fw_pri_buf, psp->sys_start_addr, psp->sys_bin_size); + return psp_v13_0_bootloader_load_component(psp, &psp->kdb, PSP_BL__LOAD_KEY_DATABASE); +} - /* Provide the sys driver to bootloader */ - WREG32_SOC15(MP0, 0, regMP0_SMN_C2PMSG_36, - (uint32_t)(psp->fw_pri_mc_addr >> 20)); - psp_gfxdrv_command_reg = PSP_BL__LOAD_SYSDRV; - WREG32_SOC15(MP0, 0, regMP0_SMN_C2PMSG_35, - psp_gfxdrv_command_reg); +static int psp_v13_0_bootloader_load_sysdrv(struct psp_context *psp) +{ + return psp_v13_0_bootloader_load_component(psp, &psp->sys, PSP_BL__LOAD_SYSDRV); +} - /* there might be handshake issue with hardware which needs delay */ - mdelay(20); +static int psp_v13_0_bootloader_load_soc_drv(struct psp_context *psp) +{ + return psp_v13_0_bootloader_load_component(psp, &psp->soc_drv, PSP_BL__LOAD_SOCDRV); +} - ret = psp_v13_0_wait_for_bootloader(psp); +static int psp_v13_0_bootloader_load_intf_drv(struct psp_context *psp) +{ + return psp_v13_0_bootloader_load_component(psp, &psp->intf_drv, PSP_BL__LOAD_INTFDRV); +} - return ret; +static int psp_v13_0_bootloader_load_dbg_drv(struct psp_context *psp) +{ + return psp_v13_0_bootloader_load_component(psp, &psp->dbg_drv, PSP_BL__LOAD_DBGDRV); } static int psp_v13_0_bootloader_load_sos(struct psp_context *psp) @@ -199,7 +196,7 @@ static int psp_v13_0_bootloader_load_sos(struct psp_context *psp) memset(psp->fw_pri_buf, 0, PSP_1_MEG); /* Copy Secure OS binary to PSP memory */ - memcpy(psp->fw_pri_buf, psp->sos_start_addr, psp->sos_bin_size); + memcpy(psp->fw_pri_buf, psp->sos.start_addr, psp->sos.size_bytes); /* Provide the PSP secure OS to bootloader */ WREG32_SOC15(MP0, 0, regMP0_SMN_C2PMSG_36, @@ -382,10 +379,71 @@ static void psp_v13_0_ring_set_wptr(struct psp_context *psp, uint32_t value) WREG32_SOC15(MP0, 0, regMP0_SMN_C2PMSG_67, value); } +static int psp_v13_0_load_usbc_pd_fw(struct psp_context *psp, uint64_t fw_pri_mc_addr) +{ + struct amdgpu_device *adev = psp->adev; + uint32_t reg_status; + int ret, i = 0; + + /* + * LFB address which is aligned to 1MB address and has to be + * right-shifted by 20 so that LFB address can be passed on a 32-bit C2P + * register + */ + WREG32_SOC15(MP0, 0, regMP0_SMN_C2PMSG_36, (fw_pri_mc_addr >> 20)); + + ret = psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, regMP0_SMN_C2PMSG_35), + 0x80000000, 0x80000000, false); + if (ret) + return ret; + + /* Fireup interrupt so PSP can pick up the address */ + WREG32_SOC15(MP0, 0, regMP0_SMN_C2PMSG_35, (GFX_CMD_USB_PD_USE_LFB << 16)); + + /* FW load takes very long time */ + do { + msleep(1000); + reg_status = RREG32_SOC15(MP0, 0, regMP0_SMN_C2PMSG_35); + + if (reg_status & 0x80000000) + goto done; + + } while (++i < USBC_PD_POLLING_LIMIT_S); + + return -ETIME; +done: + + if ((reg_status & 0xFFFF) != 0) { + DRM_ERROR("Address load failed - MP0_SMN_C2PMSG_35.Bits [15:0] = %04x\n", + reg_status & 0xFFFF); + return -EIO; + } + + return 0; +} + +static int psp_v13_0_read_usbc_pd_fw(struct psp_context *psp, uint32_t *fw_ver) +{ + struct amdgpu_device *adev = psp->adev; + int ret; + + WREG32_SOC15(MP0, 0, regMP0_SMN_C2PMSG_35, C2PMSG_CMD_GFX_USB_PD_FW_VER); + + ret = psp_wait_for(psp, SOC15_REG_OFFSET(MP0, 0, regMP0_SMN_C2PMSG_35), + 0x80000000, 0x80000000, false); + if (!ret) + *fw_ver = RREG32_SOC15(MP0, 0, regMP0_SMN_C2PMSG_36); + + return ret; +} + static const struct psp_funcs psp_v13_0_funcs = { .init_microcode = psp_v13_0_init_microcode, .bootloader_load_kdb = psp_v13_0_bootloader_load_kdb, .bootloader_load_sysdrv = psp_v13_0_bootloader_load_sysdrv, + .bootloader_load_soc_drv = psp_v13_0_bootloader_load_soc_drv, + .bootloader_load_intf_drv = psp_v13_0_bootloader_load_intf_drv, + .bootloader_load_dbg_drv = psp_v13_0_bootloader_load_dbg_drv, .bootloader_load_sos = psp_v13_0_bootloader_load_sos, .ring_init = psp_v13_0_ring_init, .ring_create = psp_v13_0_ring_create, @@ -393,6 +451,8 @@ static const struct psp_funcs psp_v13_0_funcs = { .ring_destroy = psp_v13_0_ring_destroy, .ring_get_wptr = psp_v13_0_ring_get_wptr, .ring_set_wptr = psp_v13_0_ring_set_wptr, + .load_usbc_pd_fw = psp_v13_0_load_usbc_pd_fw, + .read_usbc_pd_fw = psp_v13_0_read_usbc_pd_fw }; void psp_v13_0_set_psp_funcs(struct psp_context *psp) diff --git a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c index b86dcbabb635..1ed357cb0f49 100644 --- a/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c +++ b/drivers/gpu/drm/amd/amdgpu/psp_v3_1.c @@ -103,7 +103,7 @@ static int psp_v3_1_bootloader_load_sysdrv(struct psp_context *psp) return ret; /* Copy PSP System Driver binary to memory */ - psp_copy_fw(psp, psp->sys_start_addr, psp->sys_bin_size); + psp_copy_fw(psp, psp->sys.start_addr, psp->sys.size_bytes); /* Provide the sys driver to bootloader */ WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_36, @@ -142,7 +142,7 @@ static int psp_v3_1_bootloader_load_sos(struct psp_context *psp) return ret; /* Copy Secure OS binary to PSP memory */ - psp_copy_fw(psp, psp->sos_start_addr, psp->sos_bin_size); + psp_copy_fw(psp, psp->sos.start_addr, psp->sos.size_bytes); /* Provide the PSP secure OS to bootloader */ WREG32_SOC15(MP0, 0, mmMP0_SMN_C2PMSG_36, diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c index 6117ba8a4c3f..50bf3b71bc93 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c @@ -51,6 +51,12 @@ MODULE_FIRMWARE("amdgpu/navi14_sdma1.bin"); MODULE_FIRMWARE("amdgpu/navi12_sdma.bin"); MODULE_FIRMWARE("amdgpu/navi12_sdma1.bin"); +MODULE_FIRMWARE("amdgpu/cyan_skillfish_sdma.bin"); +MODULE_FIRMWARE("amdgpu/cyan_skillfish_sdma1.bin"); + +MODULE_FIRMWARE("amdgpu/cyan_skillfish2_sdma.bin"); +MODULE_FIRMWARE("amdgpu/cyan_skillfish2_sdma1.bin"); + #define SDMA1_REG_OFFSET 0x600 #define SDMA0_HYP_DEC_REG_START 0x5880 #define SDMA0_HYP_DEC_REG_END 0x5893 @@ -130,6 +136,37 @@ static const struct soc15_reg_golden golden_settings_sdma_nv12[] = { SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA1_RLC3_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000), }; +static const struct soc15_reg_golden golden_settings_sdma_cyan_skillfish[] = { + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA0_CHICKEN_BITS, 0xffbf1f0f, 0x03ab0107), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA0_GB_ADDR_CONFIG, 0x001877ff, 0x00000044), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA0_GB_ADDR_CONFIG_READ, 0x001877ff, 0x00000044), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA0_GFX_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA0_PAGE_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA0_RLC0_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA0_RLC1_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA0_RLC2_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA0_RLC3_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA0_RLC4_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA0_RLC5_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA0_RLC6_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA0_RLC7_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA0_UTCL1_PAGE, 0x007fffff, 0x004c5c00), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA1_CHICKEN_BITS, 0xffbf1f0f, 0x03ab0107), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA1_GB_ADDR_CONFIG, 0x001877ff, 0x00000044), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA1_GB_ADDR_CONFIG_READ, 0x001877ff, 0x00000044), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA1_GFX_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA1_PAGE_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA1_RLC0_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA1_RLC1_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA1_RLC2_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA1_RLC3_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA1_RLC4_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA1_RLC5_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA1_RLC6_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA1_RLC7_RB_WPTR_POLL_CNTL, 0xfffffff7, 0x00403000), + SOC15_REG_GOLDEN_VALUE(GC, 0, mmSDMA1_UTCL1_PAGE, 0x007fffff, 0x004c5c00) +}; + static u32 sdma_v5_0_get_reg_offset(struct amdgpu_device *adev, u32 instance, u32 internal_offset) { u32 base; @@ -180,6 +217,11 @@ static void sdma_v5_0_init_golden_registers(struct amdgpu_device *adev) golden_settings_sdma_nv12, (const u32)ARRAY_SIZE(golden_settings_sdma_nv12)); break; + case CHIP_CYAN_SKILLFISH: + soc15_program_register_sequence(adev, + golden_settings_sdma_cyan_skillfish, + (const u32)ARRAY_SIZE(golden_settings_sdma_cyan_skillfish)); + break; default: break; } @@ -200,7 +242,7 @@ static void sdma_v5_0_init_golden_registers(struct amdgpu_device *adev) static int sdma_v5_0_init_microcode(struct amdgpu_device *adev) { const char *chip_name; - char fw_name[30]; + char fw_name[40]; int err = 0, i; struct amdgpu_firmware_info *info = NULL; const struct common_firmware_header *header = NULL; @@ -221,6 +263,12 @@ static int sdma_v5_0_init_microcode(struct amdgpu_device *adev) case CHIP_NAVI12: chip_name = "navi12"; break; + case CHIP_CYAN_SKILLFISH: + if (adev->apu_flags & AMD_APU_IS_CYAN_SKILLFISH2) + chip_name = "cyan_skillfish2"; + else + chip_name = "cyan_skillfish"; + break; default: BUG(); } diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c b/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c index 7486e5306786..779f5c911e11 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v5_2.c @@ -87,21 +87,6 @@ static u32 sdma_v5_2_get_reg_offset(struct amdgpu_device *adev, u32 instance, u3 return base + internal_offset; } -static void sdma_v5_2_init_golden_registers(struct amdgpu_device *adev) -{ - switch (adev->asic_type) { - case CHIP_SIENNA_CICHLID: - case CHIP_NAVY_FLOUNDER: - case CHIP_VANGOGH: - case CHIP_DIMGREY_CAVEFISH: - case CHIP_BEIGE_GOBY: - case CHIP_YELLOW_CARP: - break; - default: - break; - } -} - static int sdma_v5_2_init_inst_ctx(struct amdgpu_sdma_instance *sdma_inst) { int err = 0; @@ -1345,8 +1330,6 @@ static int sdma_v5_2_hw_init(void *handle) int r; struct amdgpu_device *adev = (struct amdgpu_device *)handle; - sdma_v5_2_init_golden_registers(adev); - r = sdma_v5_2_start(adev); return r; diff --git a/drivers/gpu/drm/amd/amdgpu/smu_v11_0_i2c.c b/drivers/gpu/drm/amd/amdgpu/smu_v11_0_i2c.c index 5c7d769aee3f..73ffa8fde3df 100644 --- a/drivers/gpu/drm/amd/amdgpu/smu_v11_0_i2c.c +++ b/drivers/gpu/drm/amd/amdgpu/smu_v11_0_i2c.c @@ -41,9 +41,7 @@ #define I2C_SW_TIMEOUT 8 #define I2C_ABORT 0x10 -/* I2C transaction flags */ -#define I2C_NO_STOP 1 -#define I2C_RESTART 2 +#define I2C_X_RESTART BIT(31) #define to_amdgpu_device(x) (container_of(x, struct amdgpu_device, pm.smu_i2c)) @@ -56,12 +54,48 @@ static void smu_v11_0_i2c_set_clock_gating(struct i2c_adapter *control, bool en) WREG32_SOC15(SMUIO, 0, mmSMUIO_PWRMGT, reg); } +/* The T_I2C_POLL_US is defined as follows: + * + * "Define a timer interval (t_i2c_poll) equal to 10 times the + * signalling period for the highest I2C transfer speed used in the + * system and supported by DW_apb_i2c. For instance, if the highest + * I2C data transfer mode is 400 kb/s, then t_i2c_poll is 25 us." -- + * DesignWare DW_apb_i2c Databook, Version 1.21a, section 3.8.3.1, + * page 56, with grammar and syntax corrections. + * + * Vcc for our device is at 1.8V which puts it at 400 kHz, + * see Atmel AT24CM02 datasheet, section 8.3 DC Characteristics table, page 14. + * + * The procedure to disable the IP block is described in section + * 3.8.3 Disabling DW_apb_i2c on page 56. + */ +#define I2C_SPEED_MODE_FAST 2 +#define T_I2C_POLL_US 25 +#define I2C_MAX_T_POLL_COUNT 1000 -static void smu_v11_0_i2c_enable(struct i2c_adapter *control, bool enable) +static int smu_v11_0_i2c_enable(struct i2c_adapter *control, bool enable) { struct amdgpu_device *adev = to_amdgpu_device(control); WREG32_SOC15(SMUIO, 0, mmCKSVII2C_IC_ENABLE, enable ? 1 : 0); + + if (!enable) { + int ii; + + for (ii = I2C_MAX_T_POLL_COUNT; ii > 0; ii--) { + u32 en_stat = RREG32_SOC15(SMUIO, + 0, + mmCKSVII2C_IC_ENABLE_STATUS); + if (REG_GET_FIELD(en_stat, CKSVII2C_IC_ENABLE_STATUS, IC_EN)) + udelay(T_I2C_POLL_US); + else + return I2C_OK; + } + + return I2C_ABORT; + } + + return I2C_OK; } static void smu_v11_0_i2c_clear_status(struct i2c_adapter *control) @@ -83,8 +117,13 @@ static void smu_v11_0_i2c_configure(struct i2c_adapter *control) reg = REG_SET_FIELD(reg, CKSVII2C_IC_CON, IC_RESTART_EN, 1); reg = REG_SET_FIELD(reg, CKSVII2C_IC_CON, IC_10BITADDR_MASTER, 0); reg = REG_SET_FIELD(reg, CKSVII2C_IC_CON, IC_10BITADDR_SLAVE, 0); - /* Standard mode */ - reg = REG_SET_FIELD(reg, CKSVII2C_IC_CON, IC_MAX_SPEED_MODE, 2); + /* The values of IC_MAX_SPEED_MODE are, + * 1: standard mode, 0 - 100 Kb/s, + * 2: fast mode, <= 400 Kb/s, or fast mode plus, <= 1000 Kb/s, + * 3: high speed mode, <= 3.4 Mb/s. + */ + reg = REG_SET_FIELD(reg, CKSVII2C_IC_CON, IC_MAX_SPEED_MODE, + I2C_SPEED_MODE_FAST); reg = REG_SET_FIELD(reg, CKSVII2C_IC_CON, IC_MASTER_MODE, 1); WREG32_SOC15(SMUIO, 0, mmCKSVII2C_IC_CON, reg); @@ -113,13 +152,15 @@ static void smu_v11_0_i2c_set_clock(struct i2c_adapter *control) WREG32_SOC15(SMUIO, 0, mmCKSVII2C_IC_SDA_HOLD, 20); } -static void smu_v11_0_i2c_set_address(struct i2c_adapter *control, uint8_t address) +static void smu_v11_0_i2c_set_address(struct i2c_adapter *control, u16 address) { struct amdgpu_device *adev = to_amdgpu_device(control); - /* Convert fromr 8-bit to 7-bit address */ - address >>= 1; - WREG32_SOC15(SMUIO, 0, mmCKSVII2C_IC_TAR, (address & 0xFF)); + /* The IC_TAR::IC_TAR field is 10-bits wide. + * It takes a 7-bit or 10-bit addresses as an address, + * i.e. no read/write bit--no wire format, just the address. + */ + WREG32_SOC15(SMUIO, 0, mmCKSVII2C_IC_TAR, address & 0x3FF); } static uint32_t smu_v11_0_i2c_poll_tx_status(struct i2c_adapter *control) @@ -206,9 +247,6 @@ static uint32_t smu_v11_0_i2c_poll_rx_status(struct i2c_adapter *control) return ret; } - - - /** * smu_v11_0_i2c_transmit - Send a block of data over the I2C bus to a slave device. * @@ -221,17 +259,17 @@ static uint32_t smu_v11_0_i2c_poll_rx_status(struct i2c_adapter *control) * Returns 0 on success or error. */ static uint32_t smu_v11_0_i2c_transmit(struct i2c_adapter *control, - uint8_t address, uint8_t *data, - uint32_t numbytes, uint32_t i2c_flag) + u16 address, u8 *data, + u32 numbytes, u32 i2c_flag) { struct amdgpu_device *adev = to_amdgpu_device(control); - uint32_t bytes_sent, reg, ret = 0; + u32 bytes_sent, reg, ret = I2C_OK; unsigned long timeout_counter; bytes_sent = 0; DRM_DEBUG_DRIVER("I2C_Transmit(), address = %x, bytes = %d , data: ", - (uint16_t)address, numbytes); + address, numbytes); if (drm_debug_enabled(DRM_UT_DRIVER)) { print_hex_dump(KERN_INFO, "data: ", DUMP_PREFIX_NONE, @@ -246,53 +284,49 @@ static uint32_t smu_v11_0_i2c_transmit(struct i2c_adapter *control, /* Clear status bits */ smu_v11_0_i2c_clear_status(control); - timeout_counter = jiffies + msecs_to_jiffies(20); while (numbytes > 0) { reg = RREG32_SOC15(SMUIO, 0, mmCKSVII2C_IC_STATUS); - if (REG_GET_FIELD(reg, CKSVII2C_IC_STATUS, TFNF)) { - do { - reg = 0; - /* - * Prepare transaction, no need to set RESTART. I2C engine will send - * START as soon as it sees data in TXFIFO - */ - if (bytes_sent == 0) - reg = REG_SET_FIELD(reg, CKSVII2C_IC_DATA_CMD, RESTART, - (i2c_flag & I2C_RESTART) ? 1 : 0); - reg = REG_SET_FIELD(reg, CKSVII2C_IC_DATA_CMD, DAT, data[bytes_sent]); - - /* determine if we need to send STOP bit or not */ - if (numbytes == 1) - /* Final transaction, so send stop unless I2C_NO_STOP */ - reg = REG_SET_FIELD(reg, CKSVII2C_IC_DATA_CMD, STOP, - (i2c_flag & I2C_NO_STOP) ? 0 : 1); - /* Write */ - reg = REG_SET_FIELD(reg, CKSVII2C_IC_DATA_CMD, CMD, 0); - WREG32_SOC15(SMUIO, 0, mmCKSVII2C_IC_DATA_CMD, reg); - - /* Record that the bytes were transmitted */ - bytes_sent++; - numbytes--; - - reg = RREG32_SOC15(SMUIO, 0, mmCKSVII2C_IC_STATUS); - - } while (numbytes && REG_GET_FIELD(reg, CKSVII2C_IC_STATUS, TFNF)); - } - - /* - * We waited too long for the transmission FIFO to become not-full. - * Exit the loop with error. - */ - if (time_after(jiffies, timeout_counter)) { - ret |= I2C_SW_TIMEOUT; - goto Err; + if (!REG_GET_FIELD(reg, CKSVII2C_IC_STATUS, TFNF)) { + /* + * We waited for too long for the transmission + * FIFO to become not-full. Exit the loop + * with error. + */ + if (time_after(jiffies, timeout_counter)) { + ret |= I2C_SW_TIMEOUT; + goto Err; + } + } else { + reg = REG_SET_FIELD(reg, CKSVII2C_IC_DATA_CMD, DAT, + data[bytes_sent]); + + /* Final message, final byte, must generate a + * STOP to release the bus, i.e. don't hold + * SCL low. + */ + if (numbytes == 1 && i2c_flag & I2C_M_STOP) + reg = REG_SET_FIELD(reg, + CKSVII2C_IC_DATA_CMD, + STOP, 1); + + if (bytes_sent == 0 && i2c_flag & I2C_X_RESTART) + reg = REG_SET_FIELD(reg, + CKSVII2C_IC_DATA_CMD, + RESTART, 1); + + /* Write */ + reg = REG_SET_FIELD(reg, CKSVII2C_IC_DATA_CMD, CMD, 0); + WREG32_SOC15(SMUIO, 0, mmCKSVII2C_IC_DATA_CMD, reg); + + /* Record that the bytes were transmitted */ + bytes_sent++; + numbytes--; } } ret = smu_v11_0_i2c_poll_tx_status(control); - Err: /* Any error, no point in proceeding */ if (ret != I2C_OK) { @@ -323,8 +357,8 @@ Err: * Returns 0 on success or error. */ static uint32_t smu_v11_0_i2c_receive(struct i2c_adapter *control, - uint8_t address, uint8_t *data, - uint32_t numbytes, uint8_t i2c_flag) + u16 address, u8 *data, + u32 numbytes, u32 i2c_flag) { struct amdgpu_device *adev = to_amdgpu_device(control); uint32_t bytes_received, ret = I2C_OK; @@ -342,23 +376,21 @@ static uint32_t smu_v11_0_i2c_receive(struct i2c_adapter *control, smu_v11_0_i2c_clear_status(control); - /* Prepare transaction */ - - /* Each time we disable I2C, so this is not a restart */ - if (bytes_received == 0) - reg = REG_SET_FIELD(reg, CKSVII2C_IC_DATA_CMD, RESTART, - (i2c_flag & I2C_RESTART) ? 1 : 0); - reg = REG_SET_FIELD(reg, CKSVII2C_IC_DATA_CMD, DAT, 0); /* Read */ reg = REG_SET_FIELD(reg, CKSVII2C_IC_DATA_CMD, CMD, 1); - /* Transmitting last byte */ - if (numbytes == 1) - /* Final transaction, so send stop if requested */ - reg = REG_SET_FIELD(reg, CKSVII2C_IC_DATA_CMD, STOP, - (i2c_flag & I2C_NO_STOP) ? 0 : 1); + /* Final message, final byte, must generate a STOP + * to release the bus, i.e. don't hold SCL low. + */ + if (numbytes == 1 && i2c_flag & I2C_M_STOP) + reg = REG_SET_FIELD(reg, CKSVII2C_IC_DATA_CMD, + STOP, 1); + + if (bytes_received == 0 && i2c_flag & I2C_X_RESTART) + reg = REG_SET_FIELD(reg, CKSVII2C_IC_DATA_CMD, + RESTART, 1); WREG32_SOC15(SMUIO, 0, mmCKSVII2C_IC_DATA_CMD, reg); @@ -413,7 +445,6 @@ static void smu_v11_0_i2c_abort(struct i2c_adapter *control) DRM_DEBUG_DRIVER("I2C_Abort() Done."); } - static bool smu_v11_0_i2c_activity_done(struct i2c_adapter *control) { struct amdgpu_device *adev = to_amdgpu_device(control); @@ -425,7 +456,6 @@ static bool smu_v11_0_i2c_activity_done(struct i2c_adapter *control) reg_ic_enable_status = RREG32_SOC15(SMUIO, 0, mmCKSVII2C_IC_ENABLE_STATUS); reg_ic_enable = RREG32_SOC15(SMUIO, 0, mmCKSVII2C_IC_ENABLE); - if ((REG_GET_FIELD(reg_ic_enable, CKSVII2C_IC_ENABLE, ENABLE) == 0) && (REG_GET_FIELD(reg_ic_enable_status, CKSVII2C_IC_ENABLE_STATUS, IC_EN) == 1)) { /* @@ -455,6 +485,8 @@ static bool smu_v11_0_i2c_activity_done(struct i2c_adapter *control) static void smu_v11_0_i2c_init(struct i2c_adapter *control) { + int res; + /* Disable clock gating */ smu_v11_0_i2c_set_clock_gating(control, false); @@ -462,7 +494,9 @@ static void smu_v11_0_i2c_init(struct i2c_adapter *control) DRM_WARN("I2C busy !"); /* Disable I2C */ - smu_v11_0_i2c_enable(control, false); + res = smu_v11_0_i2c_enable(control, false); + if (res != I2C_OK) + smu_v11_0_i2c_abort(control); /* Configure I2C to operate as master and in standard mode */ smu_v11_0_i2c_configure(control); @@ -475,21 +509,22 @@ static void smu_v11_0_i2c_init(struct i2c_adapter *control) static void smu_v11_0_i2c_fini(struct i2c_adapter *control) { struct amdgpu_device *adev = to_amdgpu_device(control); - uint32_t reg_ic_enable_status, reg_ic_enable; + u32 status, enable, en_stat; + int res; - smu_v11_0_i2c_enable(control, false); + res = smu_v11_0_i2c_enable(control, false); + if (res != I2C_OK) { + status = RREG32_SOC15(SMUIO, 0, mmCKSVII2C_IC_STATUS); + enable = RREG32_SOC15(SMUIO, 0, mmCKSVII2C_IC_ENABLE); + en_stat = RREG32_SOC15(SMUIO, 0, mmCKSVII2C_IC_ENABLE_STATUS); - /* Double check if disabled, else force abort */ - reg_ic_enable_status = RREG32_SOC15(SMUIO, 0, mmCKSVII2C_IC_ENABLE_STATUS); - reg_ic_enable = RREG32_SOC15(SMUIO, 0, mmCKSVII2C_IC_ENABLE); - - if ((REG_GET_FIELD(reg_ic_enable, CKSVII2C_IC_ENABLE, ENABLE) == 0) && - (REG_GET_FIELD(reg_ic_enable_status, - CKSVII2C_IC_ENABLE_STATUS, IC_EN) == 1)) { - /* - * Nobody is using I2C engine, but engine remains active because - * someone missed to send STOP + /* Nobody is using the I2C engine, yet it remains + * active, possibly because someone missed to send + * STOP. */ + DRM_DEBUG_DRIVER("Aborting from fini: status:0x%08x " + "enable:0x%08x enable_stat:0x%08x", + status, enable, en_stat); smu_v11_0_i2c_abort(control); } @@ -531,22 +566,12 @@ static bool smu_v11_0_i2c_bus_unlock(struct i2c_adapter *control) /***************************** I2C GLUE ****************************/ static uint32_t smu_v11_0_i2c_read_data(struct i2c_adapter *control, - uint8_t address, - uint8_t *data, - uint32_t numbytes) + struct i2c_msg *msg, uint32_t i2c_flag) { - uint32_t ret = 0; - - /* First 2 bytes are dummy write to set EEPROM address */ - ret = smu_v11_0_i2c_transmit(control, address, data, 2, I2C_NO_STOP); - if (ret != I2C_OK) - goto Fail; + uint32_t ret; - /* Now read data starting with that address */ - ret = smu_v11_0_i2c_receive(control, address, data + 2, numbytes - 2, - I2C_RESTART); + ret = smu_v11_0_i2c_receive(control, msg->addr, msg->buf, msg->len, i2c_flag); -Fail: if (ret != I2C_OK) DRM_ERROR("ReadData() - I2C error occurred :%x", ret); @@ -554,28 +579,15 @@ Fail: } static uint32_t smu_v11_0_i2c_write_data(struct i2c_adapter *control, - uint8_t address, - uint8_t *data, - uint32_t numbytes) + struct i2c_msg *msg, uint32_t i2c_flag) { uint32_t ret; - ret = smu_v11_0_i2c_transmit(control, address, data, numbytes, 0); + ret = smu_v11_0_i2c_transmit(control, msg->addr, msg->buf, msg->len, i2c_flag); if (ret != I2C_OK) DRM_ERROR("WriteI2CData() - I2C error occurred :%x", ret); - else - /* - * According to EEPROM spec there is a MAX of 10 ms required for - * EEPROM to flush internal RX buffer after STOP was issued at the - * end of write transaction. During this time the EEPROM will not be - * responsive to any more commands - so wait a bit more. - * - * TODO Improve to wait for first ACK for slave address after - * internal write cycle done. - */ - msleep(10); - + return ret; } @@ -584,12 +596,11 @@ static void lock_bus(struct i2c_adapter *i2c, unsigned int flags) { struct amdgpu_device *adev = to_amdgpu_device(i2c); - if (!smu_v11_0_i2c_bus_lock(i2c)) { + mutex_lock(&adev->pm.smu_i2c_mutex); + if (!smu_v11_0_i2c_bus_lock(i2c)) DRM_ERROR("Failed to lock the bus from SMU"); - return; - } - - adev->pm.bus_locked = true; + else + adev->pm.bus_locked = true; } static int trylock_bus(struct i2c_adapter *i2c, unsigned int flags) @@ -602,12 +613,11 @@ static void unlock_bus(struct i2c_adapter *i2c, unsigned int flags) { struct amdgpu_device *adev = to_amdgpu_device(i2c); - if (!smu_v11_0_i2c_bus_unlock(i2c)) { + if (!smu_v11_0_i2c_bus_unlock(i2c)) DRM_ERROR("Failed to unlock the bus from SMU"); - return; - } - - adev->pm.bus_locked = false; + else + adev->pm.bus_locked = false; + mutex_unlock(&adev->pm.smu_i2c_mutex); } static const struct i2c_lock_operations smu_v11_0_i2c_i2c_lock_ops = { @@ -617,27 +627,60 @@ static const struct i2c_lock_operations smu_v11_0_i2c_i2c_lock_ops = { }; static int smu_v11_0_i2c_xfer(struct i2c_adapter *i2c_adap, - struct i2c_msg *msgs, int num) + struct i2c_msg *msg, int num) { int i, ret; - struct amdgpu_device *adev = to_amdgpu_device(i2c_adap); - - if (!adev->pm.bus_locked) { - DRM_ERROR("I2C bus unlocked, stopping transaction!"); - return -EIO; - } + u16 addr, dir; smu_v11_0_i2c_init(i2c_adap); + /* From the client's point of view, this sequence of + * messages-- the array i2c_msg *msg, is a single transaction + * on the bus, starting with START and ending with STOP. + * + * The client is welcome to send any sequence of messages in + * this array, as processing under this function here is + * striving to be agnostic. + * + * Record the first address and direction we see. If either + * changes for a subsequent message, generate ReSTART. The + * DW_apb_i2c databook, v1.21a, specifies that ReSTART is + * generated when the direction changes, with the default IP + * block parameter settings, but it doesn't specify if ReSTART + * is generated when the address changes (possibly...). We + * don't rely on the default IP block parameter settings as + * the block is shared and they may change. + */ + if (num > 0) { + addr = msg[0].addr; + dir = msg[0].flags & I2C_M_RD; + } + for (i = 0; i < num; i++) { - if (msgs[i].flags & I2C_M_RD) + u32 i2c_flag = 0; + + if (msg[i].addr != addr || (msg[i].flags ^ dir) & I2C_M_RD) { + addr = msg[i].addr; + dir = msg[i].flags & I2C_M_RD; + i2c_flag |= I2C_X_RESTART; + } + + if (i == num - 1) { + /* Set the STOP bit on the last message, so + * that the IP block generates a STOP after + * the last byte of the message. + */ + i2c_flag |= I2C_M_STOP; + } + + if (msg[i].flags & I2C_M_RD) ret = smu_v11_0_i2c_read_data(i2c_adap, - (uint8_t)msgs[i].addr, - msgs[i].buf, msgs[i].len); + msg + i, + i2c_flag); else ret = smu_v11_0_i2c_write_data(i2c_adap, - (uint8_t)msgs[i].addr, - msgs[i].buf, msgs[i].len); + msg + i, + i2c_flag); if (ret != I2C_OK) { num = -EIO; @@ -654,23 +697,28 @@ static u32 smu_v11_0_i2c_func(struct i2c_adapter *adap) return I2C_FUNC_I2C | I2C_FUNC_SMBUS_EMUL; } - static const struct i2c_algorithm smu_v11_0_i2c_algo = { .master_xfer = smu_v11_0_i2c_xfer, .functionality = smu_v11_0_i2c_func, }; +static const struct i2c_adapter_quirks smu_v11_0_i2c_control_quirks = { + .flags = I2C_AQ_NO_ZERO_LEN, +}; + int smu_v11_0_i2c_control_init(struct i2c_adapter *control) { struct amdgpu_device *adev = to_amdgpu_device(control); int res; + mutex_init(&adev->pm.smu_i2c_mutex); control->owner = THIS_MODULE; - control->class = I2C_CLASS_SPD; + control->class = I2C_CLASS_HWMON; control->dev.parent = &adev->pdev->dev; control->algo = &smu_v11_0_i2c_algo; snprintf(control->name, sizeof(control->name), "AMDGPU SMU"); control->lock_ops = &smu_v11_0_i2c_i2c_lock_ops; + control->quirks = &smu_v11_0_i2c_control_quirks; res = i2c_add_adapter(control); if (res) diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c index b7d350be8050..a5e085e570f7 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc15.c +++ b/drivers/gpu/drm/amd/amdgpu/soc15.c @@ -575,7 +575,7 @@ soc15_asic_reset_method(struct amdgpu_device *adev) baco_reset = amdgpu_dpm_is_baco_supported(adev); break; case CHIP_VEGA20: - if (adev->psp.sos_fw_version >= 0x80067) + if (adev->psp.sos.fw_version >= 0x80067) baco_reset = amdgpu_dpm_is_baco_supported(adev); /* @@ -635,7 +635,7 @@ static bool soc15_supports_baco(struct amdgpu_device *adev) case CHIP_ARCTURUS: return amdgpu_dpm_is_baco_supported(adev); case CHIP_VEGA20: - if (adev->psp.sos_fw_version >= 0x80067) + if (adev->psp.sos.fw_version >= 0x80067) return amdgpu_dpm_is_baco_supported(adev); return false; default: diff --git a/drivers/gpu/drm/amd/amdgpu/soc15_common.h b/drivers/gpu/drm/amd/amdgpu/soc15_common.h index 0eeb5e073be8..8a9ca87d8663 100644 --- a/drivers/gpu/drm/amd/amdgpu/soc15_common.h +++ b/drivers/gpu/drm/amd/amdgpu/soc15_common.h @@ -28,13 +28,13 @@ #define SOC15_REG_OFFSET(ip, inst, reg) (adev->reg_offset[ip##_HWIP][inst][reg##_BASE_IDX] + reg) #define __WREG32_SOC15_RLC__(reg, value, flag, hwip) \ - ((amdgpu_sriov_vf(adev) && adev->gfx.rlc.funcs && adev->gfx.rlc.funcs->rlcg_wreg) ? \ - adev->gfx.rlc.funcs->rlcg_wreg(adev, reg, value, flag, hwip) : \ + ((amdgpu_sriov_vf(adev) && adev->gfx.rlc.funcs && adev->gfx.rlc.funcs->sriov_wreg) ? \ + adev->gfx.rlc.funcs->sriov_wreg(adev, reg, value, flag, hwip) : \ WREG32(reg, value)) #define __RREG32_SOC15_RLC__(reg, flag, hwip) \ - ((amdgpu_sriov_vf(adev) && adev->gfx.rlc.funcs && adev->gfx.rlc.funcs->rlcg_rreg) ? \ - adev->gfx.rlc.funcs->rlcg_rreg(adev, reg, flag, hwip) : \ + ((amdgpu_sriov_vf(adev) && adev->gfx.rlc.funcs && adev->gfx.rlc.funcs->sriov_rreg) ? \ + adev->gfx.rlc.funcs->sriov_rreg(adev, reg, flag, hwip) : \ RREG32(reg)) #define WREG32_FIELD15(ip, idx, reg, field, val) \ diff --git a/drivers/gpu/drm/amd/amdgpu/ta_xgmi_if.h b/drivers/gpu/drm/amd/amdgpu/ta_xgmi_if.h index ac2c27b7630c..cce7127afeaa 100644 --- a/drivers/gpu/drm/amd/amdgpu/ta_xgmi_if.h +++ b/drivers/gpu/drm/amd/amdgpu/ta_xgmi_if.h @@ -33,7 +33,8 @@ enum ta_command_xgmi { TA_COMMAND_XGMI__GET_NODE_ID = 0x01, TA_COMMAND_XGMI__GET_HIVE_ID = 0x02, TA_COMMAND_XGMI__GET_GET_TOPOLOGY_INFO = 0x03, - TA_COMMAND_XGMI__SET_TOPOLOGY_INFO = 0x04 + TA_COMMAND_XGMI__SET_TOPOLOGY_INFO = 0x04, + TA_COMMAND_XGMI__GET_PEER_LINKS = 0x0B }; /* XGMI related enumerations */ @@ -75,6 +76,11 @@ struct ta_xgmi_node_info { enum ta_xgmi_assigned_sdma_engine sdma_engine; }; +struct ta_xgmi_peer_link_info { + uint64_t node_id; + uint8_t num_links; +}; + struct ta_xgmi_cmd_initialize_output { uint32_t status; }; @@ -97,6 +103,11 @@ struct ta_xgmi_cmd_get_topology_info_output { struct ta_xgmi_node_info nodes[TA_XGMI__MAX_CONNECTED_NODES]; }; +struct ta_xgmi_cmd_get_peer_link_info_output { + uint32_t num_nodes; + struct ta_xgmi_peer_link_info nodes[TA_XGMI__MAX_CONNECTED_NODES]; +}; + struct ta_xgmi_cmd_set_topology_info_input { uint32_t num_nodes; struct ta_xgmi_node_info nodes[TA_XGMI__MAX_CONNECTED_NODES]; @@ -115,6 +126,7 @@ union ta_xgmi_cmd_output { struct ta_xgmi_cmd_get_node_id_output get_node_id; struct ta_xgmi_cmd_get_hive_id_output get_hive_id; struct ta_xgmi_cmd_get_topology_info_output get_topology_info; + struct ta_xgmi_cmd_get_peer_link_info_output get_link_info; }; /**********************************************************/ diff --git a/drivers/gpu/drm/amd/amdgpu/umc_v6_7.c b/drivers/gpu/drm/amd/amdgpu/umc_v6_7.c index 3a8f787374c0..7da12110425c 100644 --- a/drivers/gpu/drm/amd/amdgpu/umc_v6_7.c +++ b/drivers/gpu/drm/amd/amdgpu/umc_v6_7.c @@ -28,6 +28,21 @@ #include "umc/umc_6_7_0_offset.h" #include "umc/umc_6_7_0_sh_mask.h" +const uint32_t + umc_v6_7_channel_idx_tbl_second[UMC_V6_7_UMC_INSTANCE_NUM][UMC_V6_7_CHANNEL_INSTANCE_NUM] = { + {28, 12, 6, 22}, {19, 3, 9, 25}, + {20, 4, 30, 14}, {11, 27, 1, 17}, + {24, 8, 2, 18}, {15, 31, 5, 21}, + {16, 0, 26, 10}, {7, 23, 29, 13} +}; +const uint32_t + umc_v6_7_channel_idx_tbl_first[UMC_V6_7_UMC_INSTANCE_NUM][UMC_V6_7_CHANNEL_INSTANCE_NUM] = { + {19, 3, 9, 25}, {28, 12, 6, 22}, + {11, 27, 1, 17}, {20, 4, 30, 14}, + {15, 31, 5, 21}, {24, 8, 2, 18}, + {7, 23, 29, 13}, {16, 0, 26, 10} +}; + static inline uint32_t get_umc_v6_7_reg_offset(struct amdgpu_device *adev, uint32_t umc_inst, uint32_t ch_inst) diff --git a/drivers/gpu/drm/amd/amdgpu/umc_v6_7.h b/drivers/gpu/drm/amd/amdgpu/umc_v6_7.h index 4eb85f247e96..81b8f1844091 100644 --- a/drivers/gpu/drm/amd/amdgpu/umc_v6_7.h +++ b/drivers/gpu/drm/amd/amdgpu/umc_v6_7.h @@ -23,6 +23,9 @@ #ifndef __UMC_V6_7_H__ #define __UMC_V6_7_H__ +#include "soc15_common.h" +#include "amdgpu.h" + /* EccErrCnt max value */ #define UMC_V6_7_CE_CNT_MAX 0xffff /* umc ce interrupt threshold */ @@ -32,6 +35,18 @@ #define UMC_V6_7_INST_DIST 0x40000 +/* number of umc channel instance with memory map register access */ +#define UMC_V6_7_CHANNEL_INSTANCE_NUM 4 +/* number of umc instance with memory map register access */ +#define UMC_V6_7_UMC_INSTANCE_NUM 8 +/* total channel instances in one umc block */ +#define UMC_V6_7_TOTAL_CHANNEL_NUM (UMC_V6_7_CHANNEL_INSTANCE_NUM * UMC_V6_7_UMC_INSTANCE_NUM) +/* UMC regiser per channel offset */ +#define UMC_V6_7_PER_CHANNEL_OFFSET 0x400 extern const struct amdgpu_umc_ras_funcs umc_v6_7_ras_funcs; +extern const uint32_t + umc_v6_7_channel_idx_tbl_second[UMC_V6_7_UMC_INSTANCE_NUM][UMC_V6_7_CHANNEL_INSTANCE_NUM]; +extern const uint32_t + umc_v6_7_channel_idx_tbl_first[UMC_V6_7_UMC_INSTANCE_NUM][UMC_V6_7_CHANNEL_INSTANCE_NUM]; #endif diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c index 284bb42d6c86..121ee9f2b8d1 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v1_0.c @@ -119,7 +119,7 @@ static int vcn_v1_0_sw_init(void *handle) adev->firmware.ucode[AMDGPU_UCODE_ID_VCN].fw = adev->vcn.fw; adev->firmware.fw_size += ALIGN(le32_to_cpu(hdr->ucode_size_bytes), PAGE_SIZE); - DRM_INFO("PSP loading VCN firmware\n"); + dev_info(adev->dev, "Will use PSP to load VCN firmware\n"); } r = amdgpu_vcn_resume(adev); diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c index 8af567c546db..f4686e918e0d 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c @@ -122,7 +122,7 @@ static int vcn_v2_0_sw_init(void *handle) adev->firmware.ucode[AMDGPU_UCODE_ID_VCN].fw = adev->vcn.fw; adev->firmware.fw_size += ALIGN(le32_to_cpu(hdr->ucode_size_bytes), PAGE_SIZE); - DRM_INFO("PSP loading VCN firmware\n"); + dev_info(adev->dev, "Will use PSP to load VCN firmware\n"); } r = amdgpu_vcn_resume(adev); diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c b/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c index 888b17d84691..e0c0c3734432 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v2_5.c @@ -152,7 +152,7 @@ static int vcn_v2_5_sw_init(void *handle) adev->firmware.fw_size += ALIGN(le32_to_cpu(hdr->ucode_size_bytes), PAGE_SIZE); } - DRM_INFO("PSP loading VCN firmware\n"); + dev_info(adev->dev, "Will use PSP to load VCN firmware\n"); } r = amdgpu_vcn_resume(adev); diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c index 47d4f04cbd69..2f017560948e 100644 --- a/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c +++ b/drivers/gpu/drm/amd/amdgpu/vcn_v3_0.c @@ -160,7 +160,7 @@ static int vcn_v3_0_sw_init(void *handle) adev->firmware.fw_size += ALIGN(le32_to_cpu(hdr->ucode_size_bytes), PAGE_SIZE); } - DRM_INFO("PSP loading VCN firmware\n"); + dev_info(adev->dev, "Will use PSP to load VCN firmware\n"); } r = amdgpu_vcn_resume(adev); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c index c6b02aee4993..cfedfb1e8596 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c @@ -1404,6 +1404,7 @@ static int kfd_fill_gpu_cache_info(struct kfd_dev *kdev, break; case CHIP_NAVI10: case CHIP_NAVI12: + case CHIP_CYAN_SKILLFISH: pcache_info = navi10_cache_info; num_of_cache_types = ARRAY_SIZE(navi10_cache_info); break; @@ -1989,8 +1990,19 @@ static int kfd_fill_gpu_direct_io_link_to_cpu(int *avail_size, sub_type_hdr->flags |= CRAT_IOLINK_FLAGS_BI_DIRECTIONAL; sub_type_hdr->io_interface_type = CRAT_IOLINK_TYPE_XGMI; sub_type_hdr->num_hops_xgmi = 1; + if (adev->asic_type == CHIP_ALDEBARAN) { + sub_type_hdr->minimum_bandwidth_mbs = + amdgpu_amdkfd_get_xgmi_bandwidth_mbytes( + kdev->kgd, NULL, true); + sub_type_hdr->maximum_bandwidth_mbs = + sub_type_hdr->minimum_bandwidth_mbs; + } } else { sub_type_hdr->io_interface_type = CRAT_IOLINK_TYPE_PCIEXPRESS; + sub_type_hdr->minimum_bandwidth_mbs = + amdgpu_amdkfd_get_pcie_bandwidth_mbytes(kdev->kgd, true); + sub_type_hdr->maximum_bandwidth_mbs = + amdgpu_amdkfd_get_pcie_bandwidth_mbytes(kdev->kgd, false); } sub_type_hdr->proximity_domain_from = proximity_domain; @@ -2033,6 +2045,11 @@ static int kfd_fill_gpu_xgmi_link_to_gpu(int *avail_size, sub_type_hdr->proximity_domain_to = proximity_domain_to; sub_type_hdr->num_hops_xgmi = amdgpu_amdkfd_get_xgmi_hops_count(kdev->kgd, peer_kdev->kgd); + sub_type_hdr->maximum_bandwidth_mbs = + amdgpu_amdkfd_get_xgmi_bandwidth_mbytes(kdev->kgd, peer_kdev->kgd, false); + sub_type_hdr->minimum_bandwidth_mbs = sub_type_hdr->maximum_bandwidth_mbs ? + amdgpu_amdkfd_get_xgmi_bandwidth_mbytes(kdev->kgd, NULL, true) : 0; + return 0; } diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c index 6b57dfd2cd2a..b551dd675085 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c @@ -84,6 +84,7 @@ static const struct kfd2kgd_calls *kfd2kgd_funcs[] = { [CHIP_DIMGREY_CAVEFISH] = &gfx_v10_3_kfd2kgd, [CHIP_BEIGE_GOBY] = &gfx_v10_3_kfd2kgd, [CHIP_YELLOW_CARP] = &gfx_v10_3_kfd2kgd, + [CHIP_CYAN_SKILLFISH] = &gfx_v10_kfd2kgd, }; #ifdef KFD_SUPPORT_IOMMU_V2 @@ -596,6 +597,24 @@ static const struct kfd_device_info yellow_carp_device_info = { .num_sdma_queues_per_engine = 2, }; +static const struct kfd_device_info cyan_skillfish_device_info = { + .asic_family = CHIP_CYAN_SKILLFISH, + .asic_name = "cyan_skillfish", + .max_pasid_bits = 16, + .max_no_of_hqd = 24, + .doorbell_size = 8, + .ih_ring_entry_size = 8 * sizeof(uint32_t), + .event_interrupt_class = &event_interrupt_class_v9, + .num_of_watch_points = 4, + .mqd_size_aligned = MQD_SIZE_ALIGNED, + .needs_iommu_device = false, + .supports_cwsr = true, + .needs_pci_atomics = true, + .num_sdma_engines = 2, + .num_xgmi_sdma_engines = 0, + .num_sdma_queues_per_engine = 8, +}; + /* For each entry, [0] is regular and [1] is virtualisation device. */ static const struct kfd_device_info *kfd_supported_devices[][2] = { #ifdef KFD_SUPPORT_IOMMU_V2 @@ -625,6 +644,7 @@ static const struct kfd_device_info *kfd_supported_devices[][2] = { [CHIP_DIMGREY_CAVEFISH] = {&dimgrey_cavefish_device_info, &dimgrey_cavefish_device_info}, [CHIP_BEIGE_GOBY] = {&beige_goby_device_info, &beige_goby_device_info}, [CHIP_YELLOW_CARP] = {&yellow_carp_device_info, NULL}, + [CHIP_CYAN_SKILLFISH] = {&cyan_skillfish_device_info, NULL}, }; static int kfd_gtt_sa_init(struct kfd_dev *kfd, unsigned int buf_size, @@ -1369,7 +1389,7 @@ void kfd_dec_compute_active(struct kfd_dev *kfd) WARN_ONCE(count < 0, "Compute profile ref. count error"); } -void kgd2kfd_smi_event_throttle(struct kfd_dev *kfd, uint32_t throttle_bitmask) +void kgd2kfd_smi_event_throttle(struct kfd_dev *kfd, uint64_t throttle_bitmask) { if (kfd && kfd->init_complete) kfd_smi_event_update_thermal_throttling(kfd, throttle_bitmask); @@ -1382,18 +1402,12 @@ void kgd2kfd_smi_event_throttle(struct kfd_dev *kfd, uint32_t throttle_bitmask) */ int kfd_debugfs_hang_hws(struct kfd_dev *dev) { - int r = 0; - if (dev->dqm->sched_policy != KFD_SCHED_POLICY_HWS) { pr_err("HWS is not enabled"); return -EINVAL; } - r = pm_debugfs_hang_hws(&dev->dqm->packets); - if (!r) - r = dqm_debugfs_execute_queues(dev->dqm); - - return r; + return dqm_debugfs_hang_hws(dev->dqm); } #endif diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c index 16a1713808c2..a972ef5eae68 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c @@ -260,7 +260,7 @@ static int allocate_vmid(struct device_queue_manager *dqm, static int flush_texture_cache_nocpsch(struct kfd_dev *kdev, struct qcm_process_device *qpd) { - const struct packet_manager_funcs *pmf = qpd->dqm->packets.pmf; + const struct packet_manager_funcs *pmf = qpd->dqm->packet_mgr.pmf; int ret; if (!qpd->ib_kaddr) @@ -1000,7 +1000,7 @@ static int start_nocpsch(struct device_queue_manager *dqm) init_interrupts(dqm); if (dqm->dev->device_info->asic_family == CHIP_HAWAII) - return pm_init(&dqm->packets, dqm); + return pm_init(&dqm->packet_mgr, dqm); dqm->sched_running = true; return 0; @@ -1009,7 +1009,7 @@ static int start_nocpsch(struct device_queue_manager *dqm) static int stop_nocpsch(struct device_queue_manager *dqm) { if (dqm->dev->device_info->asic_family == CHIP_HAWAII) - pm_uninit(&dqm->packets, false); + pm_uninit(&dqm->packet_mgr, false); dqm->sched_running = false; return 0; @@ -1124,7 +1124,7 @@ static int set_sched_resources(struct device_queue_manager *dqm) "queue mask: 0x%8llX\n", res.vmid_mask, res.queue_mask); - return pm_send_set_resources(&dqm->packets, &res); + return pm_send_set_resources(&dqm->packet_mgr, &res); } static int initialize_cpsch(struct device_queue_manager *dqm) @@ -1164,7 +1164,8 @@ static int start_cpsch(struct device_queue_manager *dqm) retval = 0; - retval = pm_init(&dqm->packets, dqm); + dqm_lock(dqm); + retval = pm_init(&dqm->packet_mgr, dqm); if (retval) goto fail_packet_manager_init; @@ -1186,7 +1187,6 @@ static int start_cpsch(struct device_queue_manager *dqm) init_interrupts(dqm); - dqm_lock(dqm); /* clear hang status when driver try to start the hw scheduler */ dqm->is_hws_hang = false; dqm->is_resetting = false; @@ -1197,8 +1197,9 @@ static int start_cpsch(struct device_queue_manager *dqm) return 0; fail_allocate_vidmem: fail_set_sched_resources: - pm_uninit(&dqm->packets, false); + pm_uninit(&dqm->packet_mgr, false); fail_packet_manager_init: + dqm_unlock(dqm); return retval; } @@ -1211,12 +1212,12 @@ static int stop_cpsch(struct device_queue_manager *dqm) unmap_queues_cpsch(dqm, KFD_UNMAP_QUEUES_FILTER_ALL_QUEUES, 0); hanging = dqm->is_hws_hang || dqm->is_resetting; dqm->sched_running = false; - dqm_unlock(dqm); - pm_release_ib(&dqm->packets); + pm_release_ib(&dqm->packet_mgr); kfd_gtt_sa_free(dqm->dev, dqm->fence_mem); - pm_uninit(&dqm->packets, hanging); + pm_uninit(&dqm->packet_mgr, hanging); + dqm_unlock(dqm); return 0; } @@ -1390,7 +1391,7 @@ static int map_queues_cpsch(struct device_queue_manager *dqm) if (dqm->active_runlist) return 0; - retval = pm_send_runlist(&dqm->packets, &dqm->queues); + retval = pm_send_runlist(&dqm->packet_mgr, &dqm->queues); pr_debug("%s sent runlist\n", __func__); if (retval) { pr_err("failed to execute runlist\n"); @@ -1416,13 +1417,13 @@ static int unmap_queues_cpsch(struct device_queue_manager *dqm, if (!dqm->active_runlist) return retval; - retval = pm_send_unmap_queue(&dqm->packets, KFD_QUEUE_TYPE_COMPUTE, + retval = pm_send_unmap_queue(&dqm->packet_mgr, KFD_QUEUE_TYPE_COMPUTE, filter, filter_param, false, 0); if (retval) return retval; *dqm->fence_addr = KFD_FENCE_INIT; - pm_send_query_status(&dqm->packets, dqm->fence_gpu_addr, + pm_send_query_status(&dqm->packet_mgr, dqm->fence_gpu_addr, KFD_FENCE_COMPLETED); /* should be timed out */ retval = amdkfd_fence_wait_timeout(dqm->fence_addr, KFD_FENCE_COMPLETED, @@ -1448,14 +1449,14 @@ static int unmap_queues_cpsch(struct device_queue_manager *dqm, * check those fields */ mqd_mgr = dqm->mqd_mgrs[KFD_MQD_TYPE_HIQ]; - if (mqd_mgr->read_doorbell_id(dqm->packets.priv_queue->queue->mqd)) { + if (mqd_mgr->read_doorbell_id(dqm->packet_mgr.priv_queue->queue->mqd)) { pr_err("HIQ MQD's queue_doorbell_id0 is not 0, Queue preemption time out\n"); while (halt_if_hws_hang) schedule(); return -ETIME; } - pm_release_ib(&dqm->packets); + pm_release_ib(&dqm->packet_mgr); dqm->active_runlist = false; return retval; @@ -1946,6 +1947,7 @@ struct device_queue_manager *device_queue_manager_init(struct kfd_dev *dev) case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: case CHIP_YELLOW_CARP: + case CHIP_CYAN_SKILLFISH: device_queue_manager_init_v10_navi10(&dqm->asic_ops); break; default: @@ -2099,11 +2101,16 @@ int dqm_debugfs_hqds(struct seq_file *m, void *data) return r; } -int dqm_debugfs_execute_queues(struct device_queue_manager *dqm) +int dqm_debugfs_hang_hws(struct device_queue_manager *dqm) { int r = 0; dqm_lock(dqm); + r = pm_debugfs_hang_hws(&dqm->packet_mgr); + if (r) { + dqm_unlock(dqm); + return r; + } dqm->active_runlist = true; r = execute_queues_cpsch(dqm, KFD_UNMAP_QUEUES_FILTER_ALL_QUEUES, 0); dqm_unlock(dqm); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h index 71e2fde56b2b..c8719682c4da 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h @@ -169,7 +169,7 @@ struct device_queue_manager { struct device_queue_manager_asic_ops asic_ops; struct mqd_manager *mqd_mgrs[KFD_MQD_TYPE_MAX]; - struct packet_manager packets; + struct packet_manager packet_mgr; struct kfd_dev *dev; struct mutex lock_hidden; /* use dqm_lock/unlock(dqm) */ struct list_head queues; diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c b/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c index a9b329f0f862..2e86692def19 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_flat_memory.c @@ -422,6 +422,7 @@ int kfd_init_apertures(struct kfd_process *process) case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: case CHIP_YELLOW_CARP: + case CHIP_CYAN_SKILLFISH: kfd_init_apertures_v9(pdd, id); break; default: diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c index d8e940f03102..e547f1f8c49f 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_packet_manager.c @@ -251,6 +251,7 @@ int pm_init(struct packet_manager *pm, struct device_queue_manager *dqm) case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: case CHIP_YELLOW_CARP: + case CHIP_CYAN_SKILLFISH: pm->pmf = &kfd_v9_pm_funcs; break; case CHIP_ALDEBARAN: @@ -278,6 +279,7 @@ void pm_uninit(struct packet_manager *pm, bool hanging) { mutex_destroy(&pm->lock); kernel_queue_uninit(pm->priv_queue, hanging); + pm->priv_queue = NULL; } int pm_send_set_resources(struct packet_manager *pm, @@ -447,6 +449,9 @@ int pm_debugfs_hang_hws(struct packet_manager *pm) uint32_t *buffer, size; int r = 0; + if (!pm->priv_queue) + return -EAGAIN; + size = pm->pmf->query_status_size; mutex_lock(&pm->lock); kq_acquire_packet_buffer(pm->priv_queue, diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h index 3426743ed228..8a5dfda224bf 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h @@ -1194,7 +1194,7 @@ int pm_debugfs_runlist(struct seq_file *m, void *data); int kfd_debugfs_hang_hws(struct kfd_dev *dev); int pm_debugfs_hang_hws(struct packet_manager *pm); -int dqm_debugfs_execute_queues(struct device_queue_manager *dqm); +int dqm_debugfs_hang_hws(struct device_queue_manager *dqm); #else diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.c b/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.c index 246522423559..ed4bc5f844ce 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.c @@ -205,23 +205,23 @@ void kfd_smi_event_update_gpu_reset(struct kfd_dev *dev, bool post_reset) } void kfd_smi_event_update_thermal_throttling(struct kfd_dev *dev, - uint32_t throttle_bitmask) + uint64_t throttle_bitmask) { struct amdgpu_device *adev = (struct amdgpu_device *)dev->kgd; /* * ThermalThrottle msg = throttle_bitmask(8): * thermal_interrupt_count(16): - * 1 byte event + 1 byte space + 8 byte throttle_bitmask + + * 1 byte event + 1 byte space + 16 byte throttle_bitmask + * 1 byte : + 16 byte thermal_interupt_counter + 1 byte \n + - * 1 byte \0 = 29 + * 1 byte \0 = 37 */ - char fifo_in[29]; + char fifo_in[37]; int len; if (list_empty(&dev->smi_clients)) return; - len = snprintf(fifo_in, sizeof(fifo_in), "%x %x:%llx\n", + len = snprintf(fifo_in, sizeof(fifo_in), "%x %llx:%llx\n", KFD_SMI_EVENT_THERMAL_THROTTLE, throttle_bitmask, atomic64_read(&adev->smu.throttle_int_counter)); diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.h b/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.h index b9b0438202e2..bffd0c32b060 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_smi_events.h @@ -26,7 +26,7 @@ int kfd_smi_event_open(struct kfd_dev *dev, uint32_t *fd); void kfd_smi_event_update_vmfault(struct kfd_dev *dev, uint16_t pasid); void kfd_smi_event_update_thermal_throttling(struct kfd_dev *dev, - uint32_t throttle_bitmask); + uint64_t throttle_bitmask); void kfd_smi_event_update_gpu_reset(struct kfd_dev *dev, bool post_reset); #endif diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c index b1ce072aa20b..ef992dd2da3a 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c @@ -1424,6 +1424,7 @@ int kfd_topology_add_device(struct kfd_dev *gpu) case CHIP_DIMGREY_CAVEFISH: case CHIP_BEIGE_GOBY: case CHIP_YELLOW_CARP: + case CHIP_CYAN_SKILLFISH: dev->node_props.capability |= ((HSA_CAP_DOORBELL_TYPE_2_0 << HSA_CAP_DOORBELL_TYPE_TOTALBITS_SHIFT) & HSA_CAP_DOORBELL_TYPE_TOTALBITS_MASK); @@ -1630,7 +1631,7 @@ int kfd_debugfs_rls_by_device(struct seq_file *m, void *data) } seq_printf(m, "Node %u, gpu_id %x:\n", i++, dev->gpu->id); - r = pm_debugfs_runlist(m, &dev->gpu->dqm->packets); + r = pm_debugfs_runlist(m, &dev->gpu->dqm->packet_mgr); if (r) break; } diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c index d3a2a5ff57e9..986c9d29d686 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -618,6 +618,7 @@ static void dm_dcn_vertical_interrupt0_high_irq(void *interrupt_params) } #endif +#define DMUB_TRACE_MAX_READ 64 /** * dm_dmub_outbox1_low_irq() - Handles Outbox interrupt * @interrupt_params: used for determining the Outbox instance @@ -625,7 +626,6 @@ static void dm_dcn_vertical_interrupt0_high_irq(void *interrupt_params) * Handles the Outbox Interrupt * event handler. */ -#define DMUB_TRACE_MAX_READ 64 static void dm_dmub_outbox1_low_irq(void *interrupt_params) { struct dmub_notification notify; @@ -2412,6 +2412,7 @@ static void update_connector_ext_caps(struct amdgpu_dm_connector *aconnector) static const u8 pre_computed_values[] = { 50, 51, 52, 53, 55, 56, 57, 58, 59, 61, 62, 63, 65, 66, 68, 69, 71, 72, 74, 75, 77, 79, 81, 82, 84, 86, 88, 90, 92, 94, 96, 98}; + int i; if (!aconnector || !aconnector->dc_link) return; @@ -2423,15 +2424,21 @@ static void update_connector_ext_caps(struct amdgpu_dm_connector *aconnector) conn_base = &aconnector->base; adev = drm_to_adev(conn_base->dev); dm = &adev->dm; - caps = &dm->backlight_caps; + for (i = 0; i < dm->num_of_edps; i++) { + if (link == dm->backlight_link[i]) + break; + } + if (i >= dm->num_of_edps) + return; + caps = &dm->backlight_caps[i]; caps->ext_caps = &aconnector->dc_link->dpcd_sink_ext_caps; caps->aux_support = false; max_cll = conn_base->hdr_sink_metadata.hdmi_type1.max_cll; min_cll = conn_base->hdr_sink_metadata.hdmi_type1.min_cll; - if (caps->ext_caps->bits.oled == 1 || + if (caps->ext_caps->bits.oled == 1 /*|| caps->ext_caps->bits.sdr_aux_backlight_control == 1 || - caps->ext_caps->bits.hdr_aux_backlight_control == 1) + caps->ext_caps->bits.hdr_aux_backlight_control == 1*/) caps->aux_support = true; if (amdgpu_backlight == 0) @@ -3423,35 +3430,36 @@ static int amdgpu_dm_mode_config_init(struct amdgpu_device *adev) #if defined(CONFIG_BACKLIGHT_CLASS_DEVICE) ||\ defined(CONFIG_BACKLIGHT_CLASS_DEVICE_MODULE) -static void amdgpu_dm_update_backlight_caps(struct amdgpu_display_manager *dm) +static void amdgpu_dm_update_backlight_caps(struct amdgpu_display_manager *dm, + int bl_idx) { #if defined(CONFIG_ACPI) struct amdgpu_dm_backlight_caps caps; memset(&caps, 0, sizeof(caps)); - if (dm->backlight_caps.caps_valid) + if (dm->backlight_caps[bl_idx].caps_valid) return; amdgpu_acpi_get_backlight_caps(&caps); if (caps.caps_valid) { - dm->backlight_caps.caps_valid = true; + dm->backlight_caps[bl_idx].caps_valid = true; if (caps.aux_support) return; - dm->backlight_caps.min_input_signal = caps.min_input_signal; - dm->backlight_caps.max_input_signal = caps.max_input_signal; + dm->backlight_caps[bl_idx].min_input_signal = caps.min_input_signal; + dm->backlight_caps[bl_idx].max_input_signal = caps.max_input_signal; } else { - dm->backlight_caps.min_input_signal = + dm->backlight_caps[bl_idx].min_input_signal = AMDGPU_DM_DEFAULT_MIN_BACKLIGHT; - dm->backlight_caps.max_input_signal = + dm->backlight_caps[bl_idx].max_input_signal = AMDGPU_DM_DEFAULT_MAX_BACKLIGHT; } #else - if (dm->backlight_caps.aux_support) + if (dm->backlight_caps[bl_idx].aux_support) return; - dm->backlight_caps.min_input_signal = AMDGPU_DM_DEFAULT_MIN_BACKLIGHT; - dm->backlight_caps.max_input_signal = AMDGPU_DM_DEFAULT_MAX_BACKLIGHT; + dm->backlight_caps[bl_idx].min_input_signal = AMDGPU_DM_DEFAULT_MIN_BACKLIGHT; + dm->backlight_caps[bl_idx].max_input_signal = AMDGPU_DM_DEFAULT_MAX_BACKLIGHT; #endif } @@ -3502,41 +3510,31 @@ static u32 convert_brightness_to_user(const struct amdgpu_dm_backlight_caps *cap } static int amdgpu_dm_backlight_set_level(struct amdgpu_display_manager *dm, + int bl_idx, u32 user_brightness) { struct amdgpu_dm_backlight_caps caps; - struct dc_link *link[AMDGPU_DM_MAX_NUM_EDP]; - u32 brightness[AMDGPU_DM_MAX_NUM_EDP]; + struct dc_link *link; + u32 brightness; bool rc; - int i; - amdgpu_dm_update_backlight_caps(dm); - caps = dm->backlight_caps; + amdgpu_dm_update_backlight_caps(dm, bl_idx); + caps = dm->backlight_caps[bl_idx]; - for (i = 0; i < dm->num_of_edps; i++) { - dm->brightness[i] = user_brightness; - brightness[i] = convert_brightness_from_user(&caps, dm->brightness[i]); - link[i] = (struct dc_link *)dm->backlight_link[i]; - } + dm->brightness[bl_idx] = user_brightness; + brightness = convert_brightness_from_user(&caps, dm->brightness[bl_idx]); + link = (struct dc_link *)dm->backlight_link[bl_idx]; /* Change brightness based on AUX property */ if (caps.aux_support) { - for (i = 0; i < dm->num_of_edps; i++) { - rc = dc_link_set_backlight_level_nits(link[i], true, brightness[i], - AUX_BL_DEFAULT_TRANSITION_TIME_MS); - if (!rc) { - DRM_DEBUG("DM: Failed to update backlight via AUX on eDP[%d]\n", i); - break; - } - } + rc = dc_link_set_backlight_level_nits(link, true, brightness, + AUX_BL_DEFAULT_TRANSITION_TIME_MS); + if (!rc) + DRM_DEBUG("DM: Failed to update backlight via AUX on eDP[%d]\n", bl_idx); } else { - for (i = 0; i < dm->num_of_edps; i++) { - rc = dc_link_set_backlight_level(dm->backlight_link[i], brightness[i], 0); - if (!rc) { - DRM_DEBUG("DM: Failed to update backlight on eDP[%d]\n", i); - break; - } - } + rc = dc_link_set_backlight_level(link, brightness, 0); + if (!rc) + DRM_DEBUG("DM: Failed to update backlight on eDP[%d]\n", bl_idx); } return rc ? 0 : 1; @@ -3545,33 +3543,41 @@ static int amdgpu_dm_backlight_set_level(struct amdgpu_display_manager *dm, static int amdgpu_dm_backlight_update_status(struct backlight_device *bd) { struct amdgpu_display_manager *dm = bl_get_data(bd); + int i; - amdgpu_dm_backlight_set_level(dm, bd->props.brightness); + for (i = 0; i < dm->num_of_edps; i++) { + if (bd == dm->backlight_dev[i]) + break; + } + if (i >= AMDGPU_DM_MAX_NUM_EDP) + i = 0; + amdgpu_dm_backlight_set_level(dm, i, bd->props.brightness); return 0; } -static u32 amdgpu_dm_backlight_get_level(struct amdgpu_display_manager *dm) +static u32 amdgpu_dm_backlight_get_level(struct amdgpu_display_manager *dm, + int bl_idx) { struct amdgpu_dm_backlight_caps caps; + struct dc_link *link = (struct dc_link *)dm->backlight_link[bl_idx]; - amdgpu_dm_update_backlight_caps(dm); - caps = dm->backlight_caps; + amdgpu_dm_update_backlight_caps(dm, bl_idx); + caps = dm->backlight_caps[bl_idx]; if (caps.aux_support) { - struct dc_link *link = (struct dc_link *)dm->backlight_link[0]; u32 avg, peak; bool rc; rc = dc_link_get_backlight_level_nits(link, &avg, &peak); if (!rc) - return dm->brightness[0]; + return dm->brightness[bl_idx]; return convert_brightness_to_user(&caps, avg); } else { - int ret = dc_link_get_backlight_level(dm->backlight_link[0]); + int ret = dc_link_get_backlight_level(link); if (ret == DC_ERROR_UNEXPECTED) - return dm->brightness[0]; + return dm->brightness[bl_idx]; return convert_brightness_to_user(&caps, ret); } } @@ -3579,8 +3585,15 @@ static u32 amdgpu_dm_backlight_get_level(struct amdgpu_display_manager *dm) static int amdgpu_dm_backlight_get_brightness(struct backlight_device *bd) { struct amdgpu_display_manager *dm = bl_get_data(bd); + int i; - return amdgpu_dm_backlight_get_level(dm); + for (i = 0; i < dm->num_of_edps; i++) { + if (bd == dm->backlight_dev[i]) + break; + } + if (i >= AMDGPU_DM_MAX_NUM_EDP) + i = 0; + return amdgpu_dm_backlight_get_level(dm, i); } static const struct backlight_ops amdgpu_dm_backlight_ops = { @@ -3594,31 +3607,28 @@ amdgpu_dm_register_backlight_device(struct amdgpu_display_manager *dm) { char bl_name[16]; struct backlight_properties props = { 0 }; - int i; - amdgpu_dm_update_backlight_caps(dm); - for (i = 0; i < dm->num_of_edps; i++) - dm->brightness[i] = AMDGPU_MAX_BL_LEVEL; + amdgpu_dm_update_backlight_caps(dm, dm->num_of_edps); + dm->brightness[dm->num_of_edps] = AMDGPU_MAX_BL_LEVEL; props.max_brightness = AMDGPU_MAX_BL_LEVEL; props.brightness = AMDGPU_MAX_BL_LEVEL; props.type = BACKLIGHT_RAW; snprintf(bl_name, sizeof(bl_name), "amdgpu_bl%d", - adev_to_drm(dm->adev)->primary->index); + adev_to_drm(dm->adev)->primary->index + dm->num_of_edps); - dm->backlight_dev = backlight_device_register(bl_name, - adev_to_drm(dm->adev)->dev, - dm, - &amdgpu_dm_backlight_ops, - &props); + dm->backlight_dev[dm->num_of_edps] = backlight_device_register(bl_name, + adev_to_drm(dm->adev)->dev, + dm, + &amdgpu_dm_backlight_ops, + &props); - if (IS_ERR(dm->backlight_dev)) + if (IS_ERR(dm->backlight_dev[dm->num_of_edps])) DRM_ERROR("DM: Backlight registration failed!\n"); else DRM_DEBUG_DRIVER("DM: Registered Backlight device: %s\n", bl_name); } - #endif static int initialize_plane(struct amdgpu_display_manager *dm, @@ -3675,10 +3685,10 @@ static void register_backlight_device(struct amdgpu_display_manager *dm, * DM initialization because not having a backlight control * is better then a black screen. */ - if (!dm->backlight_dev) + if (!dm->backlight_dev[dm->num_of_edps]) amdgpu_dm_register_backlight_device(dm); - if (dm->backlight_dev) { + if (dm->backlight_dev[dm->num_of_edps]) { dm->backlight_link[dm->num_of_edps] = link; dm->num_of_edps++; } @@ -4747,7 +4757,7 @@ fill_gfx9_plane_attributes_from_modifiers(struct amdgpu_device *adev, const bool force_disable_dcc) { const uint64_t modifier = afb->base.modifier; - int ret; + int ret = 0; fill_gfx9_tiling_info_from_modifier(adev, tiling_info, modifier); tiling_info->gfx9.swizzle = modifier_gfx9_swizzle_mode(modifier); @@ -4765,9 +4775,9 @@ fill_gfx9_plane_attributes_from_modifiers(struct amdgpu_device *adev, ret = validate_dcc(adev, format, rotation, tiling_info, dcc, address, plane_size); if (ret) - return ret; + drm_dbg_kms(adev_to_drm(adev), "validate_dcc: returned error: %d\n", ret); - return 0; + return ret; } static int @@ -6198,6 +6208,7 @@ static void amdgpu_dm_connector_destroy(struct drm_connector *connector) const struct dc_link *link = aconnector->dc_link; struct amdgpu_device *adev = drm_to_adev(connector->dev); struct amdgpu_display_manager *dm = &adev->dm; + int i; /* * Call only if mst_mgr was iniitalized before since it's not done @@ -6208,12 +6219,11 @@ static void amdgpu_dm_connector_destroy(struct drm_connector *connector) #if defined(CONFIG_BACKLIGHT_CLASS_DEVICE) ||\ defined(CONFIG_BACKLIGHT_CLASS_DEVICE_MODULE) - - if ((link->connector_signal & (SIGNAL_TYPE_EDP | SIGNAL_TYPE_LVDS)) && - link->type != dc_connection_none && - dm->backlight_dev) { - backlight_device_unregister(dm->backlight_dev); - dm->backlight_dev = NULL; + for (i = 0; i < dm->num_of_edps; i++) { + if ((link == dm->backlight_link[i]) && dm->backlight_dev[i]) { + backlight_device_unregister(dm->backlight_dev[i]); + dm->backlight_dev[i] = NULL; + } } #endif @@ -7570,8 +7580,10 @@ static uint add_fs_modes(struct amdgpu_dm_connector *aconnector) * 60 - Commonly used * 48,72,96 - Multiples of 24 */ - const uint32_t common_rates[] = { 23976, 24000, 25000, 29970, 30000, - 48000, 50000, 60000, 72000, 96000 }; + static const uint32_t common_rates[] = { + 23976, 24000, 25000, 29970, 30000, + 48000, 50000, 60000, 72000, 96000 + }; /* * Find mode with highest refresh rate with the same resolution @@ -9191,8 +9203,11 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state) #if defined(CONFIG_BACKLIGHT_CLASS_DEVICE) || \ defined(CONFIG_BACKLIGHT_CLASS_DEVICE_MODULE) /* restore the backlight level */ - if (dm->backlight_dev && (amdgpu_dm_backlight_get_level(dm) != dm->brightness[0])) - amdgpu_dm_backlight_set_level(dm, dm->brightness[0]); + for (i = 0; i < dm->num_of_edps; i++) { + if (dm->backlight_dev[i] && + (amdgpu_dm_backlight_get_level(dm, i) != dm->brightness[i])) + amdgpu_dm_backlight_set_level(dm, i, dm->brightness[i]); + } #endif /* * send vblank event on all events not handled in flip and @@ -10549,13 +10564,68 @@ static bool is_dp_capable_without_timing_msa(struct dc *dc, return capable; } -static bool parse_edid_cea(struct amdgpu_dm_connector *aconnector, +static bool dm_edid_parser_send_cea(struct amdgpu_display_manager *dm, + unsigned int offset, + unsigned int total_length, + uint8_t *data, + unsigned int length, + struct amdgpu_hdmi_vsdb_info *vsdb) +{ + bool res; + union dmub_rb_cmd cmd; + struct dmub_cmd_send_edid_cea *input; + struct dmub_cmd_edid_cea_output *output; + + if (length > DMUB_EDID_CEA_DATA_CHUNK_BYTES) + return false; + + memset(&cmd, 0, sizeof(cmd)); + + input = &cmd.edid_cea.data.input; + + cmd.edid_cea.header.type = DMUB_CMD__EDID_CEA; + cmd.edid_cea.header.sub_type = 0; + cmd.edid_cea.header.payload_bytes = + sizeof(cmd.edid_cea) - sizeof(cmd.edid_cea.header); + input->offset = offset; + input->length = length; + input->total_length = total_length; + memcpy(input->payload, data, length); + + res = dc_dmub_srv_cmd_with_reply_data(dm->dc->ctx->dmub_srv, &cmd); + if (!res) { + DRM_ERROR("EDID CEA parser failed\n"); + return false; + } + + output = &cmd.edid_cea.data.output; + + if (output->type == DMUB_CMD__EDID_CEA_ACK) { + if (!output->ack.success) { + DRM_ERROR("EDID CEA ack failed at offset %d\n", + output->ack.offset); + } + } else if (output->type == DMUB_CMD__EDID_CEA_AMD_VSDB) { + if (!output->amd_vsdb.vsdb_found) + return false; + + vsdb->freesync_supported = output->amd_vsdb.freesync_supported; + vsdb->amd_vsdb_version = output->amd_vsdb.amd_vsdb_version; + vsdb->min_refresh_rate_hz = output->amd_vsdb.min_frame_rate; + vsdb->max_refresh_rate_hz = output->amd_vsdb.max_frame_rate; + } else { + DRM_WARN("Unknown EDID CEA parser results\n"); + return false; + } + + return true; +} + +static bool parse_edid_cea_dmcu(struct amdgpu_display_manager *dm, uint8_t *edid_ext, int len, struct amdgpu_hdmi_vsdb_info *vsdb_info) { int i; - struct amdgpu_device *adev = drm_to_adev(aconnector->base.dev); - struct dc *dc = adev->dm.dc; /* send extension block to DMCU for parsing */ for (i = 0; i < len; i += 8) { @@ -10563,14 +10633,14 @@ static bool parse_edid_cea(struct amdgpu_dm_connector *aconnector, int offset; /* send 8 bytes a time */ - if (!dc_edid_parser_send_cea(dc, i, len, &edid_ext[i], 8)) + if (!dc_edid_parser_send_cea(dm->dc, i, len, &edid_ext[i], 8)) return false; if (i+8 == len) { /* EDID block sent completed, expect result */ int version, min_rate, max_rate; - res = dc_edid_parser_recv_amd_vsdb(dc, &version, &min_rate, &max_rate); + res = dc_edid_parser_recv_amd_vsdb(dm->dc, &version, &min_rate, &max_rate); if (res) { /* amd vsdb found */ vsdb_info->freesync_supported = 1; @@ -10584,7 +10654,7 @@ static bool parse_edid_cea(struct amdgpu_dm_connector *aconnector, } /* check for ack*/ - res = dc_edid_parser_recv_cea_ack(dc, &offset); + res = dc_edid_parser_recv_cea_ack(dm->dc, &offset); if (!res) return false; } @@ -10592,6 +10662,34 @@ static bool parse_edid_cea(struct amdgpu_dm_connector *aconnector, return false; } +static bool parse_edid_cea_dmub(struct amdgpu_display_manager *dm, + uint8_t *edid_ext, int len, + struct amdgpu_hdmi_vsdb_info *vsdb_info) +{ + int i; + + /* send extension block to DMCU for parsing */ + for (i = 0; i < len; i += 8) { + /* send 8 bytes a time */ + if (!dm_edid_parser_send_cea(dm, i, len, &edid_ext[i], 8, vsdb_info)) + return false; + } + + return vsdb_info->freesync_supported; +} + +static bool parse_edid_cea(struct amdgpu_dm_connector *aconnector, + uint8_t *edid_ext, int len, + struct amdgpu_hdmi_vsdb_info *vsdb_info) +{ + struct amdgpu_device *adev = drm_to_adev(aconnector->base.dev); + + if (adev->dm.dmub_srv) + return parse_edid_cea_dmub(&adev->dm, edid_ext, len, vsdb_info); + else + return parse_edid_cea_dmcu(&adev->dm, edid_ext, len, vsdb_info); +} + static int parse_hdmi_amd_vsdb(struct amdgpu_dm_connector *aconnector, struct edid *edid, struct amdgpu_hdmi_vsdb_info *vsdb_info) { diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h index 9522d4ca299e..ab1670b16b02 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h @@ -365,13 +365,13 @@ struct amdgpu_display_manager { spinlock_t irq_handler_list_table_lock; - struct backlight_device *backlight_dev; + struct backlight_device *backlight_dev[AMDGPU_DM_MAX_NUM_EDP]; const struct dc_link *backlight_link[AMDGPU_DM_MAX_NUM_EDP]; uint8_t num_of_edps; - struct amdgpu_dm_backlight_caps backlight_caps; + struct amdgpu_dm_backlight_caps backlight_caps[AMDGPU_DM_MAX_NUM_EDP]; struct mod_freesync *freesync_module; #ifdef CONFIG_DRM_AMD_DC_HDCP diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c index f1145086a468..87daa78a32b8 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c @@ -197,29 +197,29 @@ static ssize_t dp_link_settings_read(struct file *f, char __user *buf, rd_buf_ptr = rd_buf; - str_len = strlen("Current: %d %d %d "); - snprintf(rd_buf_ptr, str_len, "Current: %d %d %d ", + str_len = strlen("Current: %d 0x%x %d "); + snprintf(rd_buf_ptr, str_len, "Current: %d 0x%x %d ", link->cur_link_settings.lane_count, link->cur_link_settings.link_rate, link->cur_link_settings.link_spread); rd_buf_ptr += str_len; - str_len = strlen("Verified: %d %d %d "); - snprintf(rd_buf_ptr, str_len, "Verified: %d %d %d ", + str_len = strlen("Verified: %d 0x%x %d "); + snprintf(rd_buf_ptr, str_len, "Verified: %d 0x%x %d ", link->verified_link_cap.lane_count, link->verified_link_cap.link_rate, link->verified_link_cap.link_spread); rd_buf_ptr += str_len; - str_len = strlen("Reported: %d %d %d "); - snprintf(rd_buf_ptr, str_len, "Reported: %d %d %d ", + str_len = strlen("Reported: %d 0x%x %d "); + snprintf(rd_buf_ptr, str_len, "Reported: %d 0x%x %d ", link->reported_link_cap.lane_count, link->reported_link_cap.link_rate, link->reported_link_cap.link_spread); rd_buf_ptr += str_len; - str_len = strlen("Preferred: %d %d %d "); - snprintf(rd_buf_ptr, str_len, "Preferred: %d %d %d\n", + str_len = strlen("Preferred: %d 0x%x %d "); + snprintf(rd_buf_ptr, str_len, "Preferred: %d 0x%x %d\n", link->preferred_link_setting.lane_count, link->preferred_link_setting.link_rate, link->preferred_link_setting.link_spread); @@ -377,7 +377,7 @@ static ssize_t dp_phy_settings_read(struct file *f, char __user *buf, if (!rd_buf) return -EINVAL; - snprintf(rd_buf, rd_buf_size, " %d %d %d ", + snprintf(rd_buf, rd_buf_size, " %d %d %d\n", link->cur_lane_setting.VOLTAGE_SWING, link->cur_lane_setting.PRE_EMPHASIS, link->cur_lane_setting.POST_CURSOR2); diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c index e63c6885c757..8e39e9245d06 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c @@ -655,10 +655,8 @@ struct hdcp_workqueue *hdcp_create_workqueue(struct amdgpu_device *adev, struct INIT_DELAYED_WORK(&hdcp_work[i].property_validate_dwork, event_property_validate); hdcp_work[i].hdcp.config.psp.handle = &adev->psp; - if (dc->ctx->dce_version == DCN_VERSION_3_1) { + if (dc->ctx->dce_version == DCN_VERSION_3_1) hdcp_work[i].hdcp.config.psp.caps.dtm_v3_supported = 1; - hdcp_work[i].hdcp.config.psp.caps.opm_state_query_supported = false; - } hdcp_work[i].hdcp.config.ddc.handle = dc_get_link_at_index(dc, i); hdcp_work[i].hdcp.config.ddc.funcs.write_i2c = lp_write_i2c; hdcp_work[i].hdcp.config.ddc.funcs.read_i2c = lp_read_i2c; diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr.c index e133edc587d3..76ec8ec92efd 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr.c @@ -264,9 +264,9 @@ static void rv1_update_clocks(struct clk_mgr *clk_mgr_base, if (pp_smu->set_hard_min_fclk_by_freq && pp_smu->set_hard_min_dcfclk_by_freq && pp_smu->set_min_deep_sleep_dcfclk) { - pp_smu->set_hard_min_fclk_by_freq(&pp_smu->pp_smu, new_clocks->fclk_khz / 1000); - pp_smu->set_hard_min_dcfclk_by_freq(&pp_smu->pp_smu, new_clocks->dcfclk_khz / 1000); - pp_smu->set_min_deep_sleep_dcfclk(&pp_smu->pp_smu, (new_clocks->dcfclk_deep_sleep_khz + 999) / 1000); + pp_smu->set_hard_min_fclk_by_freq(&pp_smu->pp_smu, khz_to_mhz_ceil(new_clocks->fclk_khz)); + pp_smu->set_hard_min_dcfclk_by_freq(&pp_smu->pp_smu, khz_to_mhz_ceil(new_clocks->dcfclk_khz)); + pp_smu->set_min_deep_sleep_dcfclk(&pp_smu->pp_smu, khz_to_mhz_ceil(new_clocks->dcfclk_deep_sleep_khz)); } } @@ -284,9 +284,9 @@ static void rv1_update_clocks(struct clk_mgr *clk_mgr_base, if (pp_smu->set_hard_min_fclk_by_freq && pp_smu->set_hard_min_dcfclk_by_freq && pp_smu->set_min_deep_sleep_dcfclk) { - pp_smu->set_hard_min_fclk_by_freq(&pp_smu->pp_smu, new_clocks->fclk_khz / 1000); - pp_smu->set_hard_min_dcfclk_by_freq(&pp_smu->pp_smu, new_clocks->dcfclk_khz / 1000); - pp_smu->set_min_deep_sleep_dcfclk(&pp_smu->pp_smu, (new_clocks->dcfclk_deep_sleep_khz + 999) / 1000); + pp_smu->set_hard_min_fclk_by_freq(&pp_smu->pp_smu, khz_to_mhz_ceil(new_clocks->fclk_khz)); + pp_smu->set_hard_min_dcfclk_by_freq(&pp_smu->pp_smu, khz_to_mhz_ceil(new_clocks->dcfclk_khz)); + pp_smu->set_min_deep_sleep_dcfclk(&pp_smu->pp_smu, khz_to_mhz_ceil(new_clocks->dcfclk_deep_sleep_khz)); } } } diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr_vbios_smu.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr_vbios_smu.c index dbc7cde00433..fe18bb9e19aa 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr_vbios_smu.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr_vbios_smu.c @@ -130,7 +130,7 @@ int rv1_vbios_smu_set_dispclk(struct clk_mgr_internal *clk_mgr, int requested_di actual_dispclk_set_mhz = rv1_vbios_smu_send_msg_with_param( clk_mgr, VBIOSSMC_MSG_SetDispclkFreq, - requested_dispclk_khz / 1000); + khz_to_mhz_ceil(requested_dispclk_khz)); if (!IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) { if (dmcu && dmcu->funcs->is_dmcu_initialized(dmcu)) { @@ -150,7 +150,7 @@ int rv1_vbios_smu_set_dprefclk(struct clk_mgr_internal *clk_mgr) actual_dprefclk_set_mhz = rv1_vbios_smu_send_msg_with_param( clk_mgr, VBIOSSMC_MSG_SetDprefclkFreq, - clk_mgr->base.dprefclk_khz / 1000); + khz_to_mhz_ceil(clk_mgr->base.dprefclk_khz)); /* TODO: add code for programing DP DTO, currently this is down by command table */ diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn20/dcn20_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn20/dcn20_clk_mgr.c index 6e0c5c664fdc..0d01aa9f15a6 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn20/dcn20_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn20/dcn20_clk_mgr.c @@ -197,7 +197,7 @@ void dcn20_update_clocks_update_dentist(struct clk_mgr_internal *clk_mgr, struct REG_UPDATE(DENTIST_DISPCLK_CNTL, DENTIST_DISPCLK_WDIVIDER, dispclk_wdivider); -// REG_WAIT(DENTIST_DISPCLK_CNTL, DENTIST_DISPCLK_CHG_DONE, 1, 5, 100); + REG_WAIT(DENTIST_DISPCLK_CNTL, DENTIST_DISPCLK_CHG_DONE, 1, 50, 1000); REG_UPDATE(DENTIST_DISPCLK_CNTL, DENTIST_DPPCLK_WDIVIDER, dppclk_wdivider); REG_WAIT(DENTIST_DISPCLK_CNTL, DENTIST_DPPCLK_CHG_DONE, 1, 5, 100); @@ -253,20 +253,20 @@ void dcn2_update_clocks(struct clk_mgr *clk_mgr_base, if (should_set_clock(safe_to_lower, new_clocks->dcfclk_khz, clk_mgr_base->clks.dcfclk_khz)) { clk_mgr_base->clks.dcfclk_khz = new_clocks->dcfclk_khz; if (pp_smu && pp_smu->set_hard_min_dcfclk_by_freq) - pp_smu->set_hard_min_dcfclk_by_freq(&pp_smu->pp_smu, clk_mgr_base->clks.dcfclk_khz / 1000); + pp_smu->set_hard_min_dcfclk_by_freq(&pp_smu->pp_smu, khz_to_mhz_ceil(clk_mgr_base->clks.dcfclk_khz)); } if (should_set_clock(safe_to_lower, new_clocks->dcfclk_deep_sleep_khz, clk_mgr_base->clks.dcfclk_deep_sleep_khz)) { clk_mgr_base->clks.dcfclk_deep_sleep_khz = new_clocks->dcfclk_deep_sleep_khz; if (pp_smu && pp_smu->set_min_deep_sleep_dcfclk) - pp_smu->set_min_deep_sleep_dcfclk(&pp_smu->pp_smu, clk_mgr_base->clks.dcfclk_deep_sleep_khz / 1000); + pp_smu->set_min_deep_sleep_dcfclk(&pp_smu->pp_smu, khz_to_mhz_ceil(clk_mgr_base->clks.dcfclk_deep_sleep_khz)); } if (should_set_clock(safe_to_lower, new_clocks->socclk_khz, clk_mgr_base->clks.socclk_khz)) { clk_mgr_base->clks.socclk_khz = new_clocks->socclk_khz; if (pp_smu && pp_smu->set_hard_min_socclk_by_freq) - pp_smu->set_hard_min_socclk_by_freq(&pp_smu->pp_smu, clk_mgr_base->clks.socclk_khz / 1000); + pp_smu->set_hard_min_socclk_by_freq(&pp_smu->pp_smu, khz_to_mhz_ceil(clk_mgr_base->clks.socclk_khz)); } total_plane_count = clk_mgr_helper_get_active_plane_cnt(dc, context); @@ -281,7 +281,7 @@ void dcn2_update_clocks(struct clk_mgr *clk_mgr_base, if (should_set_clock(safe_to_lower, new_clocks->dramclk_khz, clk_mgr_base->clks.dramclk_khz)) { clk_mgr_base->clks.dramclk_khz = new_clocks->dramclk_khz; if (pp_smu && pp_smu->set_hard_min_uclk_by_freq) - pp_smu->set_hard_min_uclk_by_freq(&pp_smu->pp_smu, clk_mgr_base->clks.dramclk_khz / 1000); + pp_smu->set_hard_min_uclk_by_freq(&pp_smu->pp_smu, khz_to_mhz_ceil(clk_mgr_base->clks.dramclk_khz)); } if (should_set_clock(safe_to_lower, new_clocks->dppclk_khz, clk_mgr->base.clks.dppclk_khz)) { @@ -306,7 +306,7 @@ void dcn2_update_clocks(struct clk_mgr *clk_mgr_base, clk_mgr_base->clks.disp_dpp_voltage_level_khz = new_clocks->disp_dpp_voltage_level_khz; if (pp_smu && pp_smu->set_voltage_by_freq) - pp_smu->set_voltage_by_freq(&pp_smu->pp_smu, PP_SMU_NV_DISPCLK, clk_mgr_base->clks.disp_dpp_voltage_level_khz / 1000); + pp_smu->set_voltage_by_freq(&pp_smu->pp_smu, PP_SMU_NV_DISPCLK, khz_to_mhz_ceil(clk_mgr_base->clks.disp_dpp_voltage_level_khz)); } if (dc->config.forced_clocks == false || (force_reset && safe_to_lower)) { @@ -502,7 +502,7 @@ static void dcn2_notify_link_rate_change(struct clk_mgr *clk_mgr_base, struct dc if (max_phyclk_req != clk_mgr_base->clks.phyclk_khz) { clk_mgr_base->clks.phyclk_khz = max_phyclk_req; - pp_smu->set_voltage_by_freq(&pp_smu->pp_smu, PP_SMU_NV_PHYCLK, clk_mgr_base->clks.phyclk_khz / 1000); + pp_smu->set_voltage_by_freq(&pp_smu->pp_smu, PP_SMU_NV_PHYCLK, khz_to_mhz_ceil(clk_mgr_base->clks.phyclk_khz)); } } diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.c index 7deeec9d1c7c..9f7eed6688c4 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.c @@ -126,7 +126,7 @@ int rn_vbios_smu_set_dispclk(struct clk_mgr_internal *clk_mgr, int requested_dis actual_dispclk_set_mhz = rn_vbios_smu_send_msg_with_param( clk_mgr, VBIOSSMC_MSG_SetDispclkFreq, - requested_dispclk_khz / 1000); + khz_to_mhz_ceil(requested_dispclk_khz)); if (!IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment)) { if (dmcu && dmcu->funcs->is_dmcu_initialized(dmcu)) { @@ -138,7 +138,7 @@ int rn_vbios_smu_set_dispclk(struct clk_mgr_internal *clk_mgr, int requested_dis // pmfw always set clock more than or equal requested clock if (!IS_DIAG_DC(dc->ctx->dce_environment)) - ASSERT(actual_dispclk_set_mhz >= requested_dispclk_khz / 1000); + ASSERT(actual_dispclk_set_mhz >= khz_to_mhz_ceil(requested_dispclk_khz)); return actual_dispclk_set_mhz * 1000; } @@ -150,7 +150,7 @@ int rn_vbios_smu_set_dprefclk(struct clk_mgr_internal *clk_mgr) actual_dprefclk_set_mhz = rn_vbios_smu_send_msg_with_param( clk_mgr, VBIOSSMC_MSG_SetDprefclkFreq, - clk_mgr->base.dprefclk_khz / 1000); + khz_to_mhz_ceil(clk_mgr->base.dprefclk_khz)); /* TODO: add code for programing DP DTO, currently this is down by command table */ @@ -167,7 +167,7 @@ int rn_vbios_smu_set_hard_min_dcfclk(struct clk_mgr_internal *clk_mgr, int reque actual_dcfclk_set_mhz = rn_vbios_smu_send_msg_with_param( clk_mgr, VBIOSSMC_MSG_SetHardMinDcfclkByFreq, - requested_dcfclk_khz / 1000); + khz_to_mhz_ceil(requested_dcfclk_khz)); return actual_dcfclk_set_mhz * 1000; } @@ -182,7 +182,7 @@ int rn_vbios_smu_set_min_deep_sleep_dcfclk(struct clk_mgr_internal *clk_mgr, int actual_min_ds_dcfclk_mhz = rn_vbios_smu_send_msg_with_param( clk_mgr, VBIOSSMC_MSG_SetMinDeepSleepDcfclk, - requested_min_ds_dcfclk_khz / 1000); + khz_to_mhz_ceil(requested_min_ds_dcfclk_khz)); return actual_min_ds_dcfclk_mhz * 1000; } @@ -192,7 +192,7 @@ void rn_vbios_smu_set_phyclk(struct clk_mgr_internal *clk_mgr, int requested_phy rn_vbios_smu_send_msg_with_param( clk_mgr, VBIOSSMC_MSG_SetPhyclkVoltageByFreq, - requested_phyclk_khz / 1000); + khz_to_mhz_ceil(requested_phyclk_khz)); } int rn_vbios_smu_set_dppclk(struct clk_mgr_internal *clk_mgr, int requested_dpp_khz) @@ -203,10 +203,10 @@ int rn_vbios_smu_set_dppclk(struct clk_mgr_internal *clk_mgr, int requested_dpp_ actual_dppclk_set_mhz = rn_vbios_smu_send_msg_with_param( clk_mgr, VBIOSSMC_MSG_SetDppclkFreq, - requested_dpp_khz / 1000); + khz_to_mhz_ceil(requested_dpp_khz)); if (!IS_DIAG_DC(dc->ctx->dce_environment)) - ASSERT(actual_dppclk_set_mhz >= requested_dpp_khz / 1000); + ASSERT(actual_dppclk_set_mhz >= khz_to_mhz_ceil(requested_dpp_khz)); return actual_dppclk_set_mhz * 1000; } diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr.c index af7004b770ae..1861a147a7fa 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr.c @@ -284,12 +284,12 @@ static void dcn3_update_clocks(struct clk_mgr *clk_mgr_base, if (should_set_clock(safe_to_lower, new_clocks->dcfclk_khz, clk_mgr_base->clks.dcfclk_khz)) { clk_mgr_base->clks.dcfclk_khz = new_clocks->dcfclk_khz; - dcn30_smu_set_hard_min_by_freq(clk_mgr, PPCLK_DCEFCLK, clk_mgr_base->clks.dcfclk_khz / 1000); + dcn30_smu_set_hard_min_by_freq(clk_mgr, PPCLK_DCEFCLK, khz_to_mhz_ceil(clk_mgr_base->clks.dcfclk_khz)); } if (should_set_clock(safe_to_lower, new_clocks->dcfclk_deep_sleep_khz, clk_mgr_base->clks.dcfclk_deep_sleep_khz)) { clk_mgr_base->clks.dcfclk_deep_sleep_khz = new_clocks->dcfclk_deep_sleep_khz; - dcn30_smu_set_min_deep_sleep_dcef_clk(clk_mgr, clk_mgr_base->clks.dcfclk_deep_sleep_khz / 1000); + dcn30_smu_set_min_deep_sleep_dcef_clk(clk_mgr, khz_to_mhz_ceil(clk_mgr_base->clks.dcfclk_deep_sleep_khz)); } if (should_set_clock(safe_to_lower, new_clocks->socclk_khz, clk_mgr_base->clks.socclk_khz)) @@ -317,20 +317,20 @@ static void dcn3_update_clocks(struct clk_mgr *clk_mgr_base, /* set UCLK to requested value if P-State switching is supported, or to re-enable P-State switching */ if (clk_mgr_base->clks.p_state_change_support && (update_uclk || !clk_mgr_base->clks.prev_p_state_change_support)) - dcn30_smu_set_hard_min_by_freq(clk_mgr, PPCLK_UCLK, clk_mgr_base->clks.dramclk_khz / 1000); + dcn30_smu_set_hard_min_by_freq(clk_mgr, PPCLK_UCLK, khz_to_mhz_ceil(clk_mgr_base->clks.dramclk_khz)); if (should_set_clock(safe_to_lower, new_clocks->dppclk_khz, clk_mgr_base->clks.dppclk_khz)) { if (clk_mgr_base->clks.dppclk_khz > new_clocks->dppclk_khz) dpp_clock_lowered = true; clk_mgr_base->clks.dppclk_khz = new_clocks->dppclk_khz; - dcn30_smu_set_hard_min_by_freq(clk_mgr, PPCLK_PIXCLK, clk_mgr_base->clks.dppclk_khz / 1000); + dcn30_smu_set_hard_min_by_freq(clk_mgr, PPCLK_PIXCLK, khz_to_mhz_ceil(clk_mgr_base->clks.dppclk_khz)); update_dppclk = true; } if (should_set_clock(safe_to_lower, new_clocks->dispclk_khz, clk_mgr_base->clks.dispclk_khz)) { clk_mgr_base->clks.dispclk_khz = new_clocks->dispclk_khz; - dcn30_smu_set_hard_min_by_freq(clk_mgr, PPCLK_DISPCLK, clk_mgr_base->clks.dispclk_khz / 1000); + dcn30_smu_set_hard_min_by_freq(clk_mgr, PPCLK_DISPCLK, khz_to_mhz_ceil(clk_mgr_base->clks.dispclk_khz)); update_dispclk = true; } @@ -396,12 +396,17 @@ static void dcn3_set_hard_min_memclk(struct clk_mgr *clk_mgr_base, bool current_ if (!clk_mgr->smu_present) return; - if (current_mode) - dcn30_smu_set_hard_min_by_freq(clk_mgr, PPCLK_UCLK, - clk_mgr_base->clks.dramclk_khz / 1000); - else + if (current_mode) { + if (clk_mgr_base->clks.p_state_change_support) + dcn30_smu_set_hard_min_by_freq(clk_mgr, PPCLK_UCLK, + khz_to_mhz_ceil(clk_mgr_base->clks.dramclk_khz)); + else + dcn30_smu_set_hard_min_by_freq(clk_mgr, PPCLK_UCLK, + clk_mgr_base->bw_params->clk_table.entries[clk_mgr_base->bw_params->clk_table.num_entries - 1].memclk_mhz); + } else { dcn30_smu_set_hard_min_by_freq(clk_mgr, PPCLK_UCLK, clk_mgr_base->bw_params->clk_table.entries[0].memclk_mhz); + } } /* Set max memclk to highest DPM value */ @@ -489,7 +494,7 @@ static void dcn30_notify_link_rate_change(struct clk_mgr *clk_mgr_base, struct d if (max_phyclk_req != clk_mgr_base->clks.phyclk_khz) { clk_mgr_base->clks.phyclk_khz = max_phyclk_req; - dcn30_smu_set_hard_min_by_freq(clk_mgr, PPCLK_PHYCLK, clk_mgr_base->clks.phyclk_khz / 1000); + dcn30_smu_set_hard_min_by_freq(clk_mgr, PPCLK_PHYCLK, khz_to_mhz_ceil(clk_mgr_base->clks.phyclk_khz)); } } diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/dcn301_smu.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/dcn301_smu.c index 07774fa2c2cf..6ea642615854 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/dcn301_smu.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/dcn301_smu.c @@ -133,7 +133,7 @@ int dcn301_smu_set_dispclk(struct clk_mgr_internal *clk_mgr, int requested_dispc actual_dispclk_set_mhz = dcn301_smu_send_msg_with_param( clk_mgr, VBIOSSMC_MSG_SetDispclkFreq, - requested_dispclk_khz / 1000); + khz_to_mhz_ceil(requested_dispclk_khz)); return actual_dispclk_set_mhz * 1000; } @@ -147,7 +147,7 @@ int dcn301_smu_set_dprefclk(struct clk_mgr_internal *clk_mgr) actual_dprefclk_set_mhz = dcn301_smu_send_msg_with_param( clk_mgr, VBIOSSMC_MSG_SetDprefclkFreq, - clk_mgr->base.dprefclk_khz / 1000); + khz_to_mhz_ceil(clk_mgr->base.dprefclk_khz)); /* TODO: add code for programing DP DTO, currently this is down by command table */ @@ -163,7 +163,7 @@ int dcn301_smu_set_hard_min_dcfclk(struct clk_mgr_internal *clk_mgr, int request actual_dcfclk_set_mhz = dcn301_smu_send_msg_with_param( clk_mgr, VBIOSSMC_MSG_SetHardMinDcfclkByFreq, - requested_dcfclk_khz / 1000); + khz_to_mhz_ceil(requested_dcfclk_khz)); return actual_dcfclk_set_mhz * 1000; } @@ -177,7 +177,7 @@ int dcn301_smu_set_min_deep_sleep_dcfclk(struct clk_mgr_internal *clk_mgr, int r actual_min_ds_dcfclk_mhz = dcn301_smu_send_msg_with_param( clk_mgr, VBIOSSMC_MSG_SetMinDeepSleepDcfclk, - requested_min_ds_dcfclk_khz / 1000); + khz_to_mhz_ceil(requested_min_ds_dcfclk_khz)); return actual_min_ds_dcfclk_mhz * 1000; } @@ -191,7 +191,7 @@ int dcn301_smu_set_dppclk(struct clk_mgr_internal *clk_mgr, int requested_dpp_kh actual_dppclk_set_mhz = dcn301_smu_send_msg_with_param( clk_mgr, VBIOSSMC_MSG_SetDppclkFreq, - requested_dpp_khz / 1000); + khz_to_mhz_ceil(requested_dpp_khz)); return actual_dppclk_set_mhz * 1000; } diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/vg_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/vg_clk_mgr.c index c636b589d69d..7046da14bb2a 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/vg_clk_mgr.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn301/vg_clk_mgr.c @@ -114,7 +114,7 @@ void vg_update_clocks(struct clk_mgr *clk_mgr_base, display_count = vg_get_active_display_cnt_wa(dc, context); /* if we can go lower, go lower */ - if (display_count == 0) { + if (display_count == 0 && !IS_DIAG_DC(dc->ctx->dce_environment)) { union display_idle_optimization_u idle_info = { 0 }; idle_info.idle_info.df_request_disabled = 1; diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_smu.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_smu.c index dad4a4c18bcf..8c2b77eb9459 100644 --- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_smu.c +++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn31/dcn31_smu.c @@ -147,7 +147,7 @@ int dcn31_smu_set_dispclk(struct clk_mgr_internal *clk_mgr, int requested_dispcl actual_dispclk_set_mhz = dcn31_smu_send_msg_with_param( clk_mgr, VBIOSSMC_MSG_SetDispclkFreq, - (requested_dispclk_khz + 999) / 1000); + khz_to_mhz_ceil(requested_dispclk_khz)); return actual_dispclk_set_mhz * 1000; } @@ -162,7 +162,7 @@ int dcn31_smu_set_dprefclk(struct clk_mgr_internal *clk_mgr) actual_dprefclk_set_mhz = dcn31_smu_send_msg_with_param( clk_mgr, VBIOSSMC_MSG_SetDprefclkFreq, - (clk_mgr->base.dprefclk_khz + 999) / 1000); + khz_to_mhz_ceil(clk_mgr->base.dprefclk_khz)); /* TODO: add code for programing DP DTO, currently this is down by command table */ @@ -182,7 +182,7 @@ int dcn31_smu_set_hard_min_dcfclk(struct clk_mgr_internal *clk_mgr, int requeste actual_dcfclk_set_mhz = dcn31_smu_send_msg_with_param( clk_mgr, VBIOSSMC_MSG_SetHardMinDcfclkByFreq, - (requested_dcfclk_khz + 999) / 1000); + khz_to_mhz_ceil(requested_dcfclk_khz)); return actual_dcfclk_set_mhz * 1000; } @@ -200,7 +200,7 @@ int dcn31_smu_set_min_deep_sleep_dcfclk(struct clk_mgr_internal *clk_mgr, int re actual_min_ds_dcfclk_mhz = dcn31_smu_send_msg_with_param( clk_mgr, VBIOSSMC_MSG_SetMinDeepSleepDcfclk, - (requested_min_ds_dcfclk_khz + 999) / 1000); + khz_to_mhz_ceil(requested_min_ds_dcfclk_khz)); return actual_min_ds_dcfclk_mhz * 1000; } @@ -215,7 +215,7 @@ int dcn31_smu_set_dppclk(struct clk_mgr_internal *clk_mgr, int requested_dpp_khz actual_dppclk_set_mhz = dcn31_smu_send_msg_with_param( clk_mgr, VBIOSSMC_MSG_SetDppclkFreq, - (requested_dpp_khz + 999) / 1000); + khz_to_mhz_ceil(requested_dpp_khz)); return actual_dppclk_set_mhz * 1000; } diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c index 6132b645bfd1..8bd7f42a8053 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c @@ -541,6 +541,7 @@ static void link_disconnect_sink(struct dc_link *link) } link->dpcd_sink_count = 0; + //link->dpcd_caps.dpcd_rev.raw = 0; } static void link_disconnect_remap(struct dc_sink *prev_sink, struct dc_link *link) @@ -742,6 +743,7 @@ static bool detect_dp(struct dc_link *link, sink_caps, audio_support); link->dpcd_caps.dongle_type = sink_caps->dongle_type; + link->dpcd_caps.dpcd_rev.raw = 0; } return true; @@ -1663,6 +1665,12 @@ struct dc_link *link_create(const struct link_init_data *init_params) if (false == dc_link_construct(link, init_params)) goto construct_fail; + /* + * Must use preferred_link_setting, not reported_link_cap or verified_link_cap, + * since struct preferred_link_setting won't be reset after S3. + */ + link->preferred_link_setting.dpcd_source_device_specific_field_support = true; + return link; construct_fail: @@ -3509,61 +3517,6 @@ void dc_link_enable_hpd_filter(struct dc_link *link, bool enable) } } -uint32_t dc_bandwidth_in_kbps_from_timing( - const struct dc_crtc_timing *timing) -{ - uint32_t bits_per_channel = 0; - uint32_t kbps; - -#if defined(CONFIG_DRM_AMD_DC_DCN) - if (timing->flags.DSC) - return dc_dsc_stream_bandwidth_in_kbps(timing, - timing->dsc_cfg.bits_per_pixel, - timing->dsc_cfg.num_slices_h, - timing->dsc_cfg.is_dp); -#endif - - switch (timing->display_color_depth) { - case COLOR_DEPTH_666: - bits_per_channel = 6; - break; - case COLOR_DEPTH_888: - bits_per_channel = 8; - break; - case COLOR_DEPTH_101010: - bits_per_channel = 10; - break; - case COLOR_DEPTH_121212: - bits_per_channel = 12; - break; - case COLOR_DEPTH_141414: - bits_per_channel = 14; - break; - case COLOR_DEPTH_161616: - bits_per_channel = 16; - break; - default: - ASSERT(bits_per_channel != 0); - bits_per_channel = 8; - break; - } - - kbps = timing->pix_clk_100hz / 10; - kbps *= bits_per_channel; - - if (timing->flags.Y_ONLY != 1) { - /*Only YOnly make reduce bandwidth by 1/3 compares to RGB*/ - kbps *= 3; - if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR420) - kbps /= 2; - else if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR422) - kbps = kbps * 2 / 3; - } - - return kbps; - -} - void dc_link_set_drive_settings(struct dc *dc, struct link_training_settings *lt_settings, const struct dc_link *link) @@ -3769,3 +3722,58 @@ bool dc_link_should_enable_fec(const struct dc_link *link) return ret; } + +uint32_t dc_bandwidth_in_kbps_from_timing( + const struct dc_crtc_timing *timing) +{ + uint32_t bits_per_channel = 0; + uint32_t kbps; + +#if defined(CONFIG_DRM_AMD_DC_DCN) + if (timing->flags.DSC) + return dc_dsc_stream_bandwidth_in_kbps(timing, + timing->dsc_cfg.bits_per_pixel, + timing->dsc_cfg.num_slices_h, + timing->dsc_cfg.is_dp); +#endif + + switch (timing->display_color_depth) { + case COLOR_DEPTH_666: + bits_per_channel = 6; + break; + case COLOR_DEPTH_888: + bits_per_channel = 8; + break; + case COLOR_DEPTH_101010: + bits_per_channel = 10; + break; + case COLOR_DEPTH_121212: + bits_per_channel = 12; + break; + case COLOR_DEPTH_141414: + bits_per_channel = 14; + break; + case COLOR_DEPTH_161616: + bits_per_channel = 16; + break; + default: + ASSERT(bits_per_channel != 0); + bits_per_channel = 8; + break; + } + + kbps = timing->pix_clk_100hz / 10; + kbps *= bits_per_channel; + + if (timing->flags.Y_ONLY != 1) { + /*Only YOnly make reduce bandwidth by 1/3 compares to RGB*/ + kbps *= 3; + if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR420) + kbps /= 2; + else if (timing->pixel_encoding == PIXEL_ENCODING_YCBCR422) + kbps = kbps * 2 / 3; + } + + return kbps; + +} diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c index 9fb8c46dc606..f56e061d35bc 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c @@ -1241,29 +1241,15 @@ enum link_training_result dp_check_link_loss_status( static inline void decide_8b_10b_training_settings( struct dc_link *link, const struct dc_link_settings *link_setting, - const struct dc_link_training_overrides *overrides, struct link_training_settings *lt_settings) { - uint32_t lane; - memset(lt_settings, '\0', sizeof(struct link_training_settings)); /* Initialize link settings */ lt_settings->link_settings.use_link_rate_set = link_setting->use_link_rate_set; lt_settings->link_settings.link_rate_set = link_setting->link_rate_set; - - if (link->preferred_link_setting.link_rate != LINK_RATE_UNKNOWN) - lt_settings->link_settings.link_rate = link->preferred_link_setting.link_rate; - else - lt_settings->link_settings.link_rate = link_setting->link_rate; - - if (link->preferred_link_setting.lane_count != LANE_COUNT_UNKNOWN) - lt_settings->link_settings.lane_count = link->preferred_link_setting.lane_count; - else - lt_settings->link_settings.lane_count = link_setting->lane_count; - - /*@todo[vdevulap] move SS to LS, should not be handled by displaypath*/ - + lt_settings->link_settings.link_rate = link_setting->link_rate; + lt_settings->link_settings.lane_count = link_setting->lane_count; /* TODO hard coded to SS for now * lt_settings.link_settings.link_spread = * dal_display_path_is_ss_supported( @@ -1271,30 +1257,52 @@ static inline void decide_8b_10b_training_settings( * LINK_SPREAD_05_DOWNSPREAD_30KHZ : * LINK_SPREAD_DISABLED; */ - /* Initialize link spread */ - if (link->dp_ss_off) - lt_settings->link_settings.link_spread = LINK_SPREAD_DISABLED; - else if (overrides->downspread != NULL) - lt_settings->link_settings.link_spread - = *overrides->downspread - ? LINK_SPREAD_05_DOWNSPREAD_30KHZ - : LINK_SPREAD_DISABLED; - else - lt_settings->link_settings.link_spread = LINK_SPREAD_05_DOWNSPREAD_30KHZ; - + lt_settings->link_settings.link_spread = link->dp_ss_off ? + LINK_SPREAD_DISABLED : LINK_SPREAD_05_DOWNSPREAD_30KHZ; lt_settings->lttpr_mode = link->lttpr_mode; + lt_settings->cr_pattern_time = get_cr_training_aux_rd_interval(link, link_setting); + lt_settings->eq_pattern_time = get_eq_training_aux_rd_interval(link, link_setting); + lt_settings->pattern_for_cr = decide_cr_training_pattern(link_setting); + lt_settings->pattern_for_eq = decide_eq_training_pattern(link, link_setting); + lt_settings->enhanced_framing = 1; + lt_settings->should_set_fec_ready = true; +} + +void dp_decide_training_settings( + struct dc_link *link, + const struct dc_link_settings *link_settings, + struct link_training_settings *lt_settings) +{ + if (dp_get_link_encoding_format(link_settings) == DP_8b_10b_ENCODING) + decide_8b_10b_training_settings(link, link_settings, lt_settings); +} + +static void override_training_settings( + struct dc_link *link, + const struct dc_link_training_overrides *overrides, + struct link_training_settings *lt_settings) +{ + uint32_t lane; + + /* Override link settings */ + if (link->preferred_link_setting.link_rate != LINK_RATE_UNKNOWN) + lt_settings->link_settings.link_rate = link->preferred_link_setting.link_rate; + if (link->preferred_link_setting.lane_count != LANE_COUNT_UNKNOWN) + lt_settings->link_settings.lane_count = link->preferred_link_setting.lane_count; + + /* Override link spread */ + if (!link->dp_ss_off && overrides->downspread != NULL) + lt_settings->link_settings.link_spread = *overrides->downspread ? + LINK_SPREAD_05_DOWNSPREAD_30KHZ + : LINK_SPREAD_DISABLED; - /* Initialize lane settings overrides */ + /* Override lane settings */ if (overrides->voltage_swing != NULL) lt_settings->voltage_swing = overrides->voltage_swing; - if (overrides->pre_emphasis != NULL) lt_settings->pre_emphasis = overrides->pre_emphasis; - if (overrides->post_cursor2 != NULL) lt_settings->post_cursor2 = overrides->post_cursor2; - - /* Initialize lane settings (VS/PE/PC2) */ for (lane = 0; lane < LANE_COUNT_DP_MAX; lane++) { lt_settings->lane_settings[lane].VOLTAGE_SWING = lt_settings->voltage_swing != NULL ? @@ -1313,45 +1321,22 @@ static inline void decide_8b_10b_training_settings( /* Initialize training timings */ if (overrides->cr_pattern_time != NULL) lt_settings->cr_pattern_time = *overrides->cr_pattern_time; - else - lt_settings->cr_pattern_time = get_cr_training_aux_rd_interval(link, link_setting); if (overrides->eq_pattern_time != NULL) lt_settings->eq_pattern_time = *overrides->eq_pattern_time; - else - lt_settings->eq_pattern_time = get_eq_training_aux_rd_interval(link, link_setting); if (overrides->pattern_for_cr != NULL) lt_settings->pattern_for_cr = *overrides->pattern_for_cr; - else - lt_settings->pattern_for_cr = decide_cr_training_pattern(link_setting); if (overrides->pattern_for_eq != NULL) lt_settings->pattern_for_eq = *overrides->pattern_for_eq; - else - lt_settings->pattern_for_eq = decide_eq_training_pattern(link, link_setting); if (overrides->enhanced_framing != NULL) lt_settings->enhanced_framing = *overrides->enhanced_framing; - else - lt_settings->enhanced_framing = 1; if (link->preferred_training_settings.fec_enable != NULL) lt_settings->should_set_fec_ready = *link->preferred_training_settings.fec_enable; - else - lt_settings->should_set_fec_ready = true; -} - -void dp_decide_training_settings( - struct dc_link *link, - const struct dc_link_settings *link_settings, - const struct dc_link_training_overrides *overrides, - struct link_training_settings *lt_settings) -{ - if (dp_get_link_encoding_format(link_settings) == DP_8b_10b_ENCODING) - decide_8b_10b_training_settings(link, link_settings, overrides, lt_settings); } - uint8_t dp_convert_to_count(uint8_t lttpr_repeater_count) { switch (lttpr_repeater_count) { @@ -1581,6 +1566,9 @@ bool dc_link_dp_perform_link_training_skip_aux( dp_decide_training_settings( link, link_setting, + <_settings); + override_training_settings( + link, &link->preferred_training_settings, <_settings); @@ -1727,6 +1715,9 @@ enum link_training_result dc_link_dp_perform_link_training( dp_decide_training_settings( link, link_settings, + <_settings); + override_training_settings( + link, &link->preferred_training_settings, <_settings); @@ -1939,11 +1930,13 @@ enum link_training_result dc_link_dp_sync_lt_attempt( bool fec_enable = false; dp_decide_training_settings( - link, - link_settings, - lt_overrides, - <_settings); - + link, + link_settings, + <_settings); + override_training_settings( + link, + lt_overrides, + <_settings); /* Setup MST Mode */ if (lt_overrides->mst_enable) set_dp_mst_mode(link, *lt_overrides->mst_enable); @@ -4793,10 +4786,18 @@ void dpcd_set_source_specific_data(struct dc_link *link) uint8_t hblank_size = (uint8_t)link->dc->caps.min_horizontal_blanking_period; - result_write_min_hblank = core_link_write_dpcd(link, - DP_SOURCE_MINIMUM_HBLANK_SUPPORTED, (uint8_t *)(&hblank_size), - sizeof(hblank_size)); + if (link->preferred_link_setting.dpcd_source_device_specific_field_support) { + result_write_min_hblank = core_link_write_dpcd(link, + DP_SOURCE_MINIMUM_HBLANK_SUPPORTED, (uint8_t *)(&hblank_size), + sizeof(hblank_size)); + + if (result_write_min_hblank == DC_ERROR_UNEXPECTED) + link->preferred_link_setting.dpcd_source_device_specific_field_support = false; + } else { + DC_LOG_DC("Sink device does not support 00340h DPCD write. Skipping on purpose.\n"); + } } + DC_TRACE_LEVEL_MESSAGE(DAL_TRACE_LEVEL_INFORMATION, WPP_BIT_FLAG_DC_DETECTION_DP_CAPS, "result=%u link_index=%u enum dce_version=%d DPCD=0x%04X min_hblank=%u branch_dev_id=0x%x branch_dev_name='%c%c%c%c%c%c'", diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dpcd.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dpcd.c index fe234760a0f5..72970e49800a 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dpcd.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dpcd.c @@ -1,3 +1,28 @@ +/* + * Copyright 2021 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + * + */ + #include <inc/core_status.h> #include <dc_link.h> #include <inc/link_hwss.h> diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_stat.c b/drivers/gpu/drm/amd/display/dc/core/dc_stat.c index 31761f3595a6..28ef9760fa34 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_stat.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_stat.c @@ -62,3 +62,27 @@ void dc_stat_get_dmub_notification(const struct dc *dc, struct dmub_notification status = dmub_srv_stat_get_notification(dmub, notify); ASSERT(status == DMUB_STATUS_OK); } + +/** + ***************************************************************************** + * Function: dc_stat_get_dmub_dataout + * + * @brief + * Calls dmub layer to retrieve dmub gpint dataout + * + * @param + * [in] dc: dc structure + * [in] dataout: dmub gpint dataout + * + * @return + * None + ***************************************************************************** + */ +void dc_stat_get_dmub_dataout(const struct dc *dc, uint32_t *dataout) +{ + struct dmub_srv *dmub = dc->ctx->dmub_srv->dmub; + enum dmub_status status; + + status = dmub_srv_get_gpint_dataout(dmub, dataout); + ASSERT(status == DMUB_STATUS_OK); +} diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c index 45931ee14a6e..327fd1909c51 100644 --- a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c +++ b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c @@ -221,6 +221,9 @@ struct dc_stream_status *dc_stream_get_status_from_state( { uint8_t i; + if (state == NULL) + return NULL; + for (i = 0; i < state->stream_count; i++) { if (stream == state->streams[i]) return &state->stream_status[i]; diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h index 8dcea8ff5c5a..a948f4f48935 100644 --- a/drivers/gpu/drm/amd/display/dc/dc.h +++ b/drivers/gpu/drm/amd/display/dc/dc.h @@ -45,7 +45,7 @@ /* forward declaration */ struct aux_payload; -#define DC_VER "3.2.141" +#define DC_VER "3.2.146" #define MAX_SURFACES 3 #define MAX_PLANES 6 @@ -458,7 +458,65 @@ union mem_low_power_enable_options { uint32_t u32All; }; +struct dc_debug_data { + uint32_t ltFailCount; + uint32_t i2cErrorCount; + uint32_t auxErrorCount; +}; + +struct dc_phy_addr_space_config { + struct { + uint64_t start_addr; + uint64_t end_addr; + uint64_t fb_top; + uint64_t fb_offset; + uint64_t fb_base; + uint64_t agp_top; + uint64_t agp_bot; + uint64_t agp_base; + } system_aperture; + + struct { + uint64_t page_table_start_addr; + uint64_t page_table_end_addr; + uint64_t page_table_base_addr; + bool base_addr_is_mc_addr; + } gart_config; + + bool valid; + bool is_hvm_enabled; + uint64_t page_table_default_page_addr; +}; + +struct dc_virtual_addr_space_config { + uint64_t page_table_base_addr; + uint64_t page_table_start_addr; + uint64_t page_table_end_addr; + uint32_t page_table_block_size_in_bytes; + uint8_t page_table_depth; // 1 = 1 level, 2 = 2 level, etc. 0 = invalid +}; + +struct dc_bounding_box_overrides { + int sr_exit_time_ns; + int sr_enter_plus_exit_time_ns; + int urgent_latency_ns; + int percent_of_ideal_drambw; + int dram_clock_change_latency_ns; + int dummy_clock_change_latency_ns; + /* This forces a hard min on the DCFCLK we use + * for DML. Unlike the debug option for forcing + * DCFCLK, this override affects watermark calculations + */ + int min_dcfclk_mhz; +}; + +struct dc_state; +struct resource_pool; +struct dce_hwseq; + struct dc_debug_options { + bool native422_support; + bool disable_dsc; enum visual_confirm visual_confirm; bool sanity_checks; bool max_disp_clk; @@ -484,7 +542,6 @@ struct dc_debug_options { bool disable_dsc_power_gate; int dsc_min_slice_height_override; int dsc_bpp_increment_div; - bool native422_support; bool disable_pplib_wm_range; enum wm_report_mode pplib_wm_report_mode; unsigned int min_disp_clk_khz; @@ -554,7 +611,6 @@ struct dc_debug_options { bool validate_dml_output; bool enable_dmcub_surface_flip; bool usbc_combo_phy_reset_wa; - bool disable_dsc; bool enable_dram_clock_change_one_display_vactive; union mem_low_power_enable_options enable_mem_low_power; bool force_vblank_alignment; @@ -572,69 +628,13 @@ struct dc_debug_options { #endif }; -struct dc_debug_data { - uint32_t ltFailCount; - uint32_t i2cErrorCount; - uint32_t auxErrorCount; -}; - -struct dc_phy_addr_space_config { - struct { - uint64_t start_addr; - uint64_t end_addr; - uint64_t fb_top; - uint64_t fb_offset; - uint64_t fb_base; - uint64_t agp_top; - uint64_t agp_bot; - uint64_t agp_base; - } system_aperture; - - struct { - uint64_t page_table_start_addr; - uint64_t page_table_end_addr; - uint64_t page_table_base_addr; -#if defined(CONFIG_DRM_AMD_DC_DCN) - bool base_addr_is_mc_addr; -#endif - } gart_config; - - bool valid; - bool is_hvm_enabled; - uint64_t page_table_default_page_addr; -}; - -struct dc_virtual_addr_space_config { - uint64_t page_table_base_addr; - uint64_t page_table_start_addr; - uint64_t page_table_end_addr; - uint32_t page_table_block_size_in_bytes; - uint8_t page_table_depth; // 1 = 1 level, 2 = 2 level, etc. 0 = invalid -}; - -struct dc_bounding_box_overrides { - int sr_exit_time_ns; - int sr_enter_plus_exit_time_ns; - int urgent_latency_ns; - int percent_of_ideal_drambw; - int dram_clock_change_latency_ns; - int dummy_clock_change_latency_ns; - /* This forces a hard min on the DCFCLK we use - * for DML. Unlike the debug option for forcing - * DCFCLK, this override affects watermark calculations - */ - int min_dcfclk_mhz; -}; - -struct resource_pool; -struct dce_hwseq; struct gpu_info_soc_bounding_box_v1_0; struct dc { + struct dc_debug_options debug; struct dc_versions versions; struct dc_caps caps; struct dc_cap_funcs cap_funcs; struct dc_config config; - struct dc_debug_options debug; struct dc_bounding_box_overrides bb_overrides; struct dc_bug_wa work_arounds; struct dc_context *ctx; diff --git a/drivers/gpu/drm/amd/display/dc/dc_dp_types.h b/drivers/gpu/drm/amd/display/dc/dc_dp_types.h index 1948cd9427d7..4f54bde1bb1c 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_dp_types.h +++ b/drivers/gpu/drm/amd/display/dc/dc_dp_types.h @@ -109,6 +109,7 @@ struct dc_link_settings { enum dc_link_spread link_spread; bool use_link_rate_set; uint8_t link_rate_set; + bool dpcd_source_device_specific_field_support; }; struct dc_lane_settings { diff --git a/drivers/gpu/drm/amd/display/dc/dc_stat.h b/drivers/gpu/drm/amd/display/dc/dc_stat.h index 2a000ba54ddb..aacbfd786c6c 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_stat.h +++ b/drivers/gpu/drm/amd/display/dc/dc_stat.h @@ -38,5 +38,6 @@ #include "dmub/dmub_srv.h" void dc_stat_get_dmub_notification(const struct dc *dc, struct dmub_notification *notify); +void dc_stat_get_dmub_dataout(const struct dc *dc, uint32_t *dataout); #endif /* _DC_STAT_H_ */ diff --git a/drivers/gpu/drm/amd/display/dc/dc_types.h b/drivers/gpu/drm/amd/display/dc/dc_types.h index 8016e22114ce..c1532930169b 100644 --- a/drivers/gpu/drm/amd/display/dc/dc_types.h +++ b/drivers/gpu/drm/amd/display/dc/dc_types.h @@ -75,18 +75,6 @@ enum dce_environment { #define IS_DIAG_DC(dce_environment) \ (IS_FPGA_MAXIMUS_DC(dce_environment) || (dce_environment == DCE_ENV_DIAG)) -struct hw_asic_id { - uint32_t chip_id; - uint32_t chip_family; - uint32_t pci_revision_id; - uint32_t hw_internal_rev; - uint32_t vram_type; - uint32_t vram_width; - uint32_t feature_flags; - uint32_t fake_paths_num; - void *atombios_base_address; -}; - struct dc_perf_trace { unsigned long read_count; unsigned long write_count; @@ -94,36 +82,7 @@ struct dc_perf_trace { unsigned long last_entry_write; }; -struct dc_context { - struct dc *dc; - - void *driver_context; /* e.g. amdgpu_device */ - struct dc_perf_trace *perf_trace; - void *cgs_device; - - enum dce_environment dce_environment; - struct hw_asic_id asic_id; - - /* todo: below should probably move to dc. to facilitate removal - * of AS we will store these here - */ - enum dce_version dce_version; - struct dc_bios *dc_bios; - bool created_bios; - struct gpio_service *gpio_service; - uint32_t dc_sink_id_count; - uint32_t dc_stream_id_count; - uint32_t dc_edp_id_count; - uint64_t fbc_gpu_addr; - struct dc_dmub_srv *dmub_srv; - -#ifdef CONFIG_DRM_AMD_DC_HDCP - struct cp_psp cp_psp; -#endif -}; - - -#define DC_MAX_EDID_BUFFER_SIZE 1280 +#define DC_MAX_EDID_BUFFER_SIZE 2048 #define DC_EDID_BLOCK_SIZE 128 #define MAX_SURFACE_NUM 4 #define NUM_PIXEL_FORMATS 10 @@ -836,6 +795,46 @@ struct dc_clock_config { uint32_t current_clock_khz;/*current clock in use*/ }; +struct hw_asic_id { + uint32_t chip_id; + uint32_t chip_family; + uint32_t pci_revision_id; + uint32_t hw_internal_rev; + uint32_t vram_type; + uint32_t vram_width; + uint32_t feature_flags; + uint32_t fake_paths_num; + void *atombios_base_address; +}; + +struct dc_context { + struct dc *dc; + + void *driver_context; /* e.g. amdgpu_device */ + struct dc_perf_trace *perf_trace; + void *cgs_device; + + enum dce_environment dce_environment; + struct hw_asic_id asic_id; + + /* todo: below should probably move to dc. to facilitate removal + * of AS we will store these here + */ + enum dce_version dce_version; + struct dc_bios *dc_bios; + bool created_bios; + struct gpio_service *gpio_service; + uint32_t dc_sink_id_count; + uint32_t dc_stream_id_count; + uint32_t dc_edp_id_count; + uint64_t fbc_gpu_addr; + struct dc_dmub_srv *dmub_srv; +#ifdef CONFIG_DRM_AMD_DC_HDCP + struct cp_psp cp_psp; +#endif + +}; + /* DSC DPCD capabilities */ union dsc_slice_caps1 { struct { diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c b/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c index 2fb88e54a4bf..058a9356a39a 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dce_aux.c @@ -71,6 +71,8 @@ enum { #define DEFAULT_AUX_ENGINE_MULT 0 #define DEFAULT_AUX_ENGINE_LENGTH 69 +#define DC_TRACE_LEVEL_MESSAGE(...) /* do nothing */ + static void release_engine( struct dce_aux *engine) { @@ -743,5 +745,11 @@ bool dce_aux_transfer_with_retries(struct ddc_service *ddc, fail: if (!payload_reply) payload->reply = NULL; + + DC_TRACE_LEVEL_MESSAGE(DAL_TRACE_LEVEL_ERROR, + WPP_BIT_FLAG_DC_ERROR, + "AUX transaction failed. Result: %d", + operation_result); + return false; } diff --git a/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c b/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c index 10d42ae0cffe..1ca8b1d94bc2 100644 --- a/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c +++ b/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c @@ -29,6 +29,8 @@ #include "dmub/dmub_srv.h" #include "core_types.h" +#define DC_TRACE_LEVEL_MESSAGE(...) /* do nothing */ + #define MAX_PIPES 6 /* @@ -96,10 +98,19 @@ static void dmub_psr_get_state(struct dmub_psr *dmub, enum dc_psr_state *state, // Return invalid state when GPINT times out *state = PSR_STATE_INVALID; - // Assert if max retry hit - if (retry_count >= 1000) - ASSERT(0); } while (++retry_count <= 1000 && *state == PSR_STATE_INVALID); + + // Assert if max retry hit + if (retry_count >= 1000 && *state == PSR_STATE_INVALID) { + ASSERT(0); + DC_TRACE_LEVEL_MESSAGE(DAL_TRACE_LEVEL_ERROR, + WPP_BIT_FLAG_Firmware_PsrState, + "Unable to get PSR state from FW."); + } else + DC_TRACE_LEVEL_MESSAGE(DAL_TRACE_LEVEL_VERBOSE, + WPP_BIT_FLAG_Firmware_PsrState, + "Got PSR state from FW. PSR state: %d, Retry count: %d", + *state, retry_count); } /* @@ -207,7 +218,7 @@ static void dmub_psr_set_level(struct dmub_psr *dmub, uint16_t psr_level, uint8_ cmd.psr_set_level.header.sub_type = DMUB_CMD__PSR_SET_LEVEL; cmd.psr_set_level.header.payload_bytes = sizeof(struct dmub_cmd_psr_set_level_data); cmd.psr_set_level.psr_set_level_data.psr_level = psr_level; - cmd.psr_set_level.psr_set_level_data.cmd_version = PSR_VERSION_1; + cmd.psr_set_level.psr_set_level_data.cmd_version = DMUB_CMD_PSR_CONTROL_VERSION_1; cmd.psr_set_level.psr_set_level_data.panel_inst = panel_inst; dc_dmub_srv_cmd_queue(dc->dmub_srv, &cmd); dc_dmub_srv_cmd_execute(dc->dmub_srv); @@ -293,7 +304,7 @@ static bool dmub_psr_copy_settings(struct dmub_psr *dmub, copy_settings_data->debug.bitfields.use_hw_lock_mgr = 1; copy_settings_data->fec_enable_status = (link->fec_state == dc_link_fec_enabled); copy_settings_data->fec_enable_delay_in100us = link->dc->debug.fec_enable_delay_in100us; - copy_settings_data->cmd_version = PSR_VERSION_1; + copy_settings_data->cmd_version = DMUB_CMD_PSR_CONTROL_VERSION_1; copy_settings_data->panel_inst = panel_inst; dc_dmub_srv_cmd_queue(dc->dmub_srv, &cmd); diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h index 8d7e92d5d3e4..39485bdeb90e 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubbub.h @@ -121,6 +121,10 @@ struct dcn_hubbub_registers { uint32_t DCN_VM_AGP_BASE; uint32_t DCN_VM_PROTECTION_FAULT_DEFAULT_ADDR_MSB; uint32_t DCN_VM_PROTECTION_FAULT_DEFAULT_ADDR_LSB; + uint32_t DCN_VM_FAULT_ADDR_MSB; + uint32_t DCN_VM_FAULT_ADDR_LSB; + uint32_t DCN_VM_FAULT_CNTL; + uint32_t DCN_VM_FAULT_STATUS; uint32_t DCHUBBUB_ARB_FRAC_URG_BW_NOM_A; uint32_t DCHUBBUB_ARB_FRAC_URG_BW_NOM_B; uint32_t DCHUBBUB_ARB_FRAC_URG_BW_NOM_C; @@ -233,7 +237,19 @@ struct dcn_hubbub_registers { type DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_C;\ type DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_D;\ type DCN_VM_PROTECTION_FAULT_DEFAULT_ADDR_MSB;\ - type DCN_VM_PROTECTION_FAULT_DEFAULT_ADDR_LSB + type DCN_VM_PROTECTION_FAULT_DEFAULT_ADDR_LSB;\ + type DCN_VM_FAULT_ADDR_MSB;\ + type DCN_VM_FAULT_ADDR_LSB;\ + type DCN_VM_ERROR_STATUS_CLEAR;\ + type DCN_VM_ERROR_STATUS_MODE;\ + type DCN_VM_ERROR_INTERRUPT_ENABLE;\ + type DCN_VM_RANGE_FAULT_DISABLE;\ + type DCN_VM_PRQ_FAULT_DISABLE;\ + type DCN_VM_ERROR_STATUS;\ + type DCN_VM_ERROR_VMID;\ + type DCN_VM_ERROR_TABLE_LEVEL;\ + type DCN_VM_ERROR_PIPE;\ + type DCN_VM_ERROR_INTERRUPT_STATUS #define HUBBUB_STUTTER_REG_FIELD_LIST(type) \ type DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_A;\ @@ -303,6 +319,7 @@ struct dcn_hubbub_registers { type DET3_SIZE_CURRENT;\ type COMPBUF_SIZE;\ type COMPBUF_SIZE_CURRENT;\ + type CONFIG_ERROR;\ type COMPBUF_RESERVED_SPACE_64B;\ type COMPBUF_RESERVED_SPACE_ZS;\ type DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_Z8_A;\ diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c index 04303fe9c659..ea185c877323 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.c @@ -871,6 +871,8 @@ void hubp1_read_state_common(struct hubp *hubp) struct _vcs_dpi_display_dlg_regs_st *dlg_attr = &s->dlg_attr; struct _vcs_dpi_display_ttu_regs_st *ttu_attr = &s->ttu_attr; struct _vcs_dpi_display_rq_regs_st *rq_regs = &s->rq_regs; + uint32_t aperture_low_msb, aperture_low_lsb; + uint32_t aperture_high_msb, aperture_high_lsb; /* Requester */ REG_GET(HUBPRET_CONTROL, @@ -881,6 +883,22 @@ void hubp1_read_state_common(struct hubp *hubp) MRQ_EXPANSION_MODE, &rq_regs->mrq_expansion_mode, CRQ_EXPANSION_MODE, &rq_regs->crq_expansion_mode); + REG_GET(DCN_VM_SYSTEM_APERTURE_LOW_ADDR_MSB, + MC_VM_SYSTEM_APERTURE_LOW_ADDR_MSB, &aperture_low_msb); + + REG_GET(DCN_VM_SYSTEM_APERTURE_LOW_ADDR_LSB, + MC_VM_SYSTEM_APERTURE_LOW_ADDR_LSB, &aperture_low_lsb); + + REG_GET(DCN_VM_SYSTEM_APERTURE_HIGH_ADDR_MSB, + MC_VM_SYSTEM_APERTURE_HIGH_ADDR_MSB, &aperture_high_msb); + + REG_GET(DCN_VM_SYSTEM_APERTURE_HIGH_ADDR_LSB, + MC_VM_SYSTEM_APERTURE_HIGH_ADDR_LSB, &aperture_high_lsb); + + // On DCN1, aperture is broken down into MSB and LSB; only keep bits [47:18] to match later DCN format + rq_regs->aperture_low_addr = (aperture_low_msb << 26) | (aperture_low_lsb >> 6); + rq_regs->aperture_high_addr = (aperture_high_msb << 26) | (aperture_high_lsb >> 6); + /* DLG - Per hubp */ REG_GET_2(BLANK_OFFSET_0, REFCYC_H_BLANK_END, &dlg_attr->refcyc_h_blank_end, @@ -1037,6 +1055,17 @@ void hubp1_read_state_common(struct hubp *hubp) QoS_LEVEL_LOW_WM, &s->qos_level_low_wm, QoS_LEVEL_HIGH_WM, &s->qos_level_high_wm); + REG_GET(DCSURF_PRIMARY_SURFACE_ADDRESS, + PRIMARY_SURFACE_ADDRESS, &s->primary_surface_addr_lo); + + REG_GET(DCSURF_PRIMARY_SURFACE_ADDRESS_HIGH, + PRIMARY_SURFACE_ADDRESS, &s->primary_surface_addr_hi); + + REG_GET(DCSURF_PRIMARY_META_SURFACE_ADDRESS, + PRIMARY_META_SURFACE_ADDRESS, &s->primary_meta_addr_lo); + + REG_GET(DCSURF_PRIMARY_META_SURFACE_ADDRESS_HIGH, + PRIMARY_META_SURFACE_ADDRESS, &s->primary_meta_addr_hi); } void hubp1_read_state(struct hubp *hubp) diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h index e2f2f6995935..9cb8c383d673 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hubp.h @@ -682,6 +682,10 @@ struct dcn_hubp_state { uint32_t min_ttu_vblank; uint32_t qos_level_low_wm; uint32_t qos_level_high_wm; + uint32_t primary_surface_addr_lo; + uint32_t primary_surface_addr_hi; + uint32_t primary_meta_addr_lo; + uint32_t primary_meta_addr_hi; }; struct dcn10_hubp { diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c index c545eddabdcc..35af0401f256 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c @@ -1502,25 +1502,22 @@ void dcn10_init_hw(struct dc *dc) void dcn10_power_down_on_boot(struct dc *dc) { struct dc_link *edp_links[MAX_NUM_EDP]; - struct dc_link *edp_link; + struct dc_link *edp_link = NULL; int edp_num; int i = 0; get_edp_links(dc, edp_links, &edp_num); - - if (edp_num) { - for (i = 0; i < edp_num; i++) { - edp_link = edp_links[i]; - if (edp_link->link_enc->funcs->is_dig_enabled && - edp_link->link_enc->funcs->is_dig_enabled(edp_link->link_enc) && - dc->hwseq->funcs.edp_backlight_control && - dc->hwss.power_down && - dc->hwss.edp_power_control) { - dc->hwseq->funcs.edp_backlight_control(edp_link, false); - dc->hwss.power_down(dc); - dc->hwss.edp_power_control(edp_link, false); - } - } + if (edp_num) + edp_link = edp_links[0]; + + if (edp_link && edp_link->link_enc->funcs->is_dig_enabled && + edp_link->link_enc->funcs->is_dig_enabled(edp_link->link_enc) && + dc->hwseq->funcs.edp_backlight_control && + dc->hwss.power_down && + dc->hwss.edp_power_control) { + dc->hwseq->funcs.edp_backlight_control(edp_link, false); + dc->hwss.power_down(dc); + dc->hwss.edp_power_control(edp_link, false); } else { for (i = 0; i < dc->link_count; i++) { struct dc_link *link = dc->links[i]; @@ -3180,8 +3177,12 @@ void dcn10_update_dchub(struct dce_hwseq *hws, struct dchub_init_data *dh_data) static bool dcn10_can_pipe_disable_cursor(struct pipe_ctx *pipe_ctx) { struct pipe_ctx *test_pipe; - const struct rect *r1 = &pipe_ctx->plane_res.scl_data.recout, *r2; + const struct scaler_data *scl_data = &pipe_ctx->plane_res.scl_data; + const struct rect *r1 = &scl_data->recout, *r2; int r1_r = r1->x + r1->width, r1_b = r1->y + r1->height, r2_r, r2_b; + int cur_layer = pipe_ctx->plane_state->layer_index; + bool upper_pipe_exists = false; + struct fixed31_32 one = dc_fixpt_from_int(1); /** * Disable the cursor if there's another pipe above this with a @@ -3199,8 +3200,17 @@ static bool dcn10_can_pipe_disable_cursor(struct pipe_ctx *pipe_ctx) if (r1->x >= r2->x && r1->y >= r2->y && r1_r <= r2_r && r1_b <= r2_b) return true; + + if (test_pipe->plane_state->layer_index < cur_layer) + upper_pipe_exists = true; } + // if plane scaled, assume an upper plane can handle cursor if it exists. + if (upper_pipe_exists && + (scl_data->ratios.horz.value != one.value || + scl_data->ratios.vert.value != one.value)) + return true; + return false; } diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c index 3696faf12d86..37848f4577b1 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c @@ -1388,6 +1388,12 @@ void optc1_read_otg_state(struct optc *optc1, REG_GET(OPTC_INPUT_GLOBAL_CONTROL, OPTC_UNDERFLOW_OCCURRED_STATUS, &s->underflow_occurred_status); + + REG_GET(OTG_VERTICAL_INTERRUPT2_CONTROL, + OTG_VERTICAL_INTERRUPT2_INT_ENABLE, &s->vertical_interrupt2_en); + + REG_GET(OTG_VERTICAL_INTERRUPT2_POSITION, + OTG_VERTICAL_INTERRUPT2_LINE_START, &s->vertical_interrupt2_line); } bool optc1_get_otg_active_size(struct timing_generator *optc, diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h index 29d6fbe0093a..c50c29984d51 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h +++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h @@ -578,6 +578,8 @@ struct dcn_otg_state { uint32_t underflow_occurred_status; uint32_t otg_enabled; uint32_t blank_enabled; + uint32_t vertical_interrupt2_en; + uint32_t vertical_interrupt2_line; }; void optc1_read_otg_state(struct optc *optc1, diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubbub.c index 91a9305d42e8..aacb1fb5c73e 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubbub.c +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubbub.c @@ -605,6 +605,26 @@ static bool hubbub2_program_watermarks( return wm_pending; } +void hubbub2_read_state(struct hubbub *hubbub, struct dcn_hubbub_state *hubbub_state) +{ + struct dcn20_hubbub *hubbub1 = TO_DCN20_HUBBUB(hubbub); + + if (REG(DCN_VM_FAULT_ADDR_MSB)) + hubbub_state->vm_fault_addr_msb = REG_READ(DCN_VM_FAULT_ADDR_MSB); + + if (REG(DCN_VM_FAULT_ADDR_LSB)) + hubbub_state->vm_fault_addr_msb = REG_READ(DCN_VM_FAULT_ADDR_LSB); + + if (REG(DCN_VM_FAULT_CNTL)) + REG_GET(DCN_VM_FAULT_CNTL, DCN_VM_ERROR_STATUS_MODE, &hubbub_state->vm_error_mode); + + if (REG(DCN_VM_FAULT_STATUS)) { + REG_GET(DCN_VM_FAULT_STATUS, DCN_VM_ERROR_STATUS, &hubbub_state->vm_error_status); + REG_GET(DCN_VM_FAULT_STATUS, DCN_VM_ERROR_VMID, &hubbub_state->vm_error_vmid); + REG_GET(DCN_VM_FAULT_STATUS, DCN_VM_ERROR_PIPE, &hubbub_state->vm_error_pipe); + } +} + static const struct hubbub_funcs hubbub2_funcs = { .update_dchub = hubbub2_update_dchub, .init_dchub_sys_ctx = hubbub2_init_dchub_sys_ctx, @@ -617,6 +637,7 @@ static const struct hubbub_funcs hubbub2_funcs = { .program_watermarks = hubbub2_program_watermarks, .is_allow_self_refresh_enabled = hubbub1_is_allow_self_refresh_enabled, .allow_self_refresh_control = hubbub1_allow_self_refresh_control, + .hubbub_read_state = hubbub2_read_state, }; void hubbub2_construct(struct dcn20_hubbub *hubbub, diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubbub.h b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubbub.h index 10af257d90ef..2f6146bf1d32 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubbub.h +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubbub.h @@ -29,16 +29,6 @@ #include "dcn10/dcn10_hubbub.h" #include "dcn20_vmid.h" -#define HUBBUB_REG_LIST_DCN20_COMMON()\ - HUBBUB_REG_LIST_DCN_COMMON(), \ - SR(DCHUBBUB_CRC_CTRL), \ - SR(DCN_VM_FB_LOCATION_BASE),\ - SR(DCN_VM_FB_LOCATION_TOP),\ - SR(DCN_VM_FB_OFFSET),\ - SR(DCN_VM_AGP_BOT),\ - SR(DCN_VM_AGP_TOP),\ - SR(DCN_VM_AGP_BASE) - #define TO_DCN20_HUBBUB(hubbub)\ container_of(hubbub, struct dcn20_hubbub, base) @@ -50,7 +40,11 @@ SR(DCN_VM_FB_OFFSET),\ SR(DCN_VM_AGP_BOT),\ SR(DCN_VM_AGP_TOP),\ - SR(DCN_VM_AGP_BASE) + SR(DCN_VM_AGP_BASE),\ + SR(DCN_VM_FAULT_ADDR_MSB), \ + SR(DCN_VM_FAULT_ADDR_LSB), \ + SR(DCN_VM_FAULT_CNTL), \ + SR(DCN_VM_FAULT_STATUS) #define HUBBUB_REG_LIST_DCN20(id)\ HUBBUB_REG_LIST_DCN20_COMMON(), \ @@ -71,7 +65,19 @@ HUBBUB_SF(DCN_VM_AGP_TOP, AGP_TOP, mask_sh), \ HUBBUB_SF(DCN_VM_AGP_BASE, AGP_BASE, mask_sh), \ HUBBUB_SF(DCN_VM_PROTECTION_FAULT_DEFAULT_ADDR_MSB, DCN_VM_PROTECTION_FAULT_DEFAULT_ADDR_MSB, mask_sh), \ - HUBBUB_SF(DCN_VM_PROTECTION_FAULT_DEFAULT_ADDR_LSB, DCN_VM_PROTECTION_FAULT_DEFAULT_ADDR_LSB, mask_sh) + HUBBUB_SF(DCN_VM_PROTECTION_FAULT_DEFAULT_ADDR_LSB, DCN_VM_PROTECTION_FAULT_DEFAULT_ADDR_LSB, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_ADDR_MSB, DCN_VM_FAULT_ADDR_MSB, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_ADDR_LSB, DCN_VM_FAULT_ADDR_LSB, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_CNTL, DCN_VM_ERROR_STATUS_CLEAR, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_CNTL, DCN_VM_ERROR_STATUS_MODE, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_CNTL, DCN_VM_ERROR_INTERRUPT_ENABLE, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_CNTL, DCN_VM_RANGE_FAULT_DISABLE, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_CNTL, DCN_VM_PRQ_FAULT_DISABLE, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_STATUS, DCN_VM_ERROR_STATUS, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_STATUS, DCN_VM_ERROR_VMID, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_STATUS, DCN_VM_ERROR_TABLE_LEVEL, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_STATUS, DCN_VM_ERROR_PIPE, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_STATUS, DCN_VM_ERROR_INTERRUPT_STATUS, mask_sh) struct dcn20_hubbub { struct hubbub base; @@ -131,4 +137,7 @@ void hubbub2_get_dchub_ref_freq(struct hubbub *hubbub, void hubbub2_wm_read_state(struct hubbub *hubbub, struct dcn_hubbub_wm *wm); +void hubbub2_read_state(struct hubbub *hubbub, + struct dcn_hubbub_state *hubbub_state); + #endif diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubp.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubp.c index 7e54058715aa..5adf42a7cc27 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubp.c +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hubp.c @@ -1080,6 +1080,12 @@ void hubp2_read_state_common(struct hubp *hubp) MRQ_EXPANSION_MODE, &rq_regs->mrq_expansion_mode, CRQ_EXPANSION_MODE, &rq_regs->crq_expansion_mode); + REG_GET(DCN_VM_SYSTEM_APERTURE_HIGH_ADDR, + MC_VM_SYSTEM_APERTURE_HIGH_ADDR, &rq_regs->aperture_high_addr); + + REG_GET(DCN_VM_SYSTEM_APERTURE_LOW_ADDR, + MC_VM_SYSTEM_APERTURE_LOW_ADDR, &rq_regs->aperture_low_addr); + /* DLG - Per hubp */ REG_GET_2(BLANK_OFFSET_0, REFCYC_H_BLANK_END, &dlg_attr->refcyc_h_blank_end, @@ -1236,6 +1242,17 @@ void hubp2_read_state_common(struct hubp *hubp) QoS_LEVEL_LOW_WM, &s->qos_level_low_wm, QoS_LEVEL_HIGH_WM, &s->qos_level_high_wm); + REG_GET(DCSURF_PRIMARY_SURFACE_ADDRESS, + PRIMARY_SURFACE_ADDRESS, &s->primary_surface_addr_lo); + + REG_GET(DCSURF_PRIMARY_SURFACE_ADDRESS_HIGH, + PRIMARY_SURFACE_ADDRESS, &s->primary_surface_addr_hi); + + REG_GET(DCSURF_PRIMARY_META_SURFACE_ADDRESS, + PRIMARY_META_SURFACE_ADDRESS, &s->primary_meta_addr_lo); + + REG_GET(DCSURF_PRIMARY_META_SURFACE_ADDRESS_HIGH, + PRIMARY_META_SURFACE_ADDRESS, &s->primary_meta_addr_hi); } void hubp2_read_state(struct hubp *hubp) diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c index b173fa3653b5..0b1cd1dbed8b 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c @@ -2462,7 +2462,7 @@ void dcn20_set_mcif_arb_params( wb_arb_params->cli_watermark[k] = get_wm_writeback_urgent(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; wb_arb_params->pstate_watermark[k] = get_wm_writeback_dram_clock_change(&context->bw_ctx.dml, pipes, pipe_cnt) * 1000; } - wb_arb_params->time_per_pixel = 16.0 / context->res_ctx.pipe_ctx[i].stream->phy_pix_clk; /* 4 bit fraction, ms */ + wb_arb_params->time_per_pixel = 16.0 * 1000 / (context->res_ctx.pipe_ctx[i].stream->phy_pix_clk / 1000); /* 4 bit fraction, ms */ wb_arb_params->slice_lines = 32; wb_arb_params->arbitration_slice = 2; wb_arb_params->max_scaled_time = dcn20_calc_max_scaled_time(wb_arb_params->time_per_pixel, @@ -2531,16 +2531,16 @@ struct pipe_ctx *dcn20_find_secondary_pipe(struct dc *dc, * pick that pipe as secondary * Same logic applies for ODM pipes */ - if (dc->current_state->res_ctx.pipe_ctx[primary_pipe->pipe_idx].bottom_pipe) { - preferred_pipe_idx = dc->current_state->res_ctx.pipe_ctx[primary_pipe->pipe_idx].bottom_pipe->pipe_idx; + if (dc->current_state->res_ctx.pipe_ctx[primary_pipe->pipe_idx].next_odm_pipe) { + preferred_pipe_idx = dc->current_state->res_ctx.pipe_ctx[primary_pipe->pipe_idx].next_odm_pipe->pipe_idx; if (res_ctx->pipe_ctx[preferred_pipe_idx].stream == NULL) { secondary_pipe = &res_ctx->pipe_ctx[preferred_pipe_idx]; secondary_pipe->pipe_idx = preferred_pipe_idx; } } if (secondary_pipe == NULL && - dc->current_state->res_ctx.pipe_ctx[primary_pipe->pipe_idx].next_odm_pipe) { - preferred_pipe_idx = dc->current_state->res_ctx.pipe_ctx[primary_pipe->pipe_idx].next_odm_pipe->pipe_idx; + dc->current_state->res_ctx.pipe_ctx[primary_pipe->pipe_idx].bottom_pipe) { + preferred_pipe_idx = dc->current_state->res_ctx.pipe_ctx[primary_pipe->pipe_idx].bottom_pipe->pipe_idx; if (res_ctx->pipe_ctx[preferred_pipe_idx].stream == NULL) { secondary_pipe = &res_ctx->pipe_ctx[preferred_pipe_idx]; secondary_pipe->pipe_idx = preferred_pipe_idx; diff --git a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.c index 42fbb5e6d505..36044cb8ec83 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.c +++ b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.c @@ -701,6 +701,7 @@ static const struct hubbub_funcs hubbub21_funcs = { .program_watermarks = hubbub21_program_watermarks, .allow_self_refresh_control = hubbub1_allow_self_refresh_control, .apply_DEDCN21_147_wa = hubbub21_apply_DEDCN21_147_wa, + .hubbub_read_state = hubbub2_read_state, }; void hubbub21_construct(struct dcn20_hubbub *hubbub, diff --git a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.h b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.h index ef3ef28509ed..d8eb2bb7282c 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.h +++ b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubbub.h @@ -108,7 +108,19 @@ HUBBUB_SF(DCN_VM_FB_OFFSET, FB_OFFSET, mask_sh), \ HUBBUB_SF(DCN_VM_AGP_BOT, AGP_BOT, mask_sh), \ HUBBUB_SF(DCN_VM_AGP_TOP, AGP_TOP, mask_sh), \ - HUBBUB_SF(DCN_VM_AGP_BASE, AGP_BASE, mask_sh) + HUBBUB_SF(DCN_VM_AGP_BASE, AGP_BASE, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_ADDR_MSB, DCN_VM_FAULT_ADDR_MSB, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_ADDR_LSB, DCN_VM_FAULT_ADDR_LSB, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_CNTL, DCN_VM_ERROR_STATUS_CLEAR, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_CNTL, DCN_VM_ERROR_STATUS_MODE, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_CNTL, DCN_VM_ERROR_INTERRUPT_ENABLE, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_CNTL, DCN_VM_RANGE_FAULT_DISABLE, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_CNTL, DCN_VM_PRQ_FAULT_DISABLE, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_STATUS, DCN_VM_ERROR_STATUS, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_STATUS, DCN_VM_ERROR_VMID, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_STATUS, DCN_VM_ERROR_TABLE_LEVEL, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_STATUS, DCN_VM_ERROR_PIPE, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_STATUS, DCN_VM_ERROR_INTERRUPT_STATUS, mask_sh) void dcn21_dchvm_init(struct hubbub *hubbub); int hubbub21_init_dchub(struct hubbub *hubbub, diff --git a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubp.c b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubp.c index b0c9180b808f..3de1bcf9b3d8 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubp.c +++ b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_hubp.c @@ -833,7 +833,7 @@ static struct hubp_funcs dcn21_hubp_funcs = { .dmdata_set_attributes = hubp2_dmdata_set_attributes, .dmdata_load = hubp2_dmdata_load, .dmdata_status_done = hubp2_dmdata_status_done, - .hubp_read_state = hubp1_read_state, + .hubp_read_state = hubp2_read_state, .hubp_clear_underflow = hubp1_clear_underflow, .hubp_set_flip_control_surface_gsl = hubp2_set_flip_control_surface_gsl, .hubp_init = hubp21_init, diff --git a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c index f3d98e3ba624..f27fc2acac57 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c +++ b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c @@ -109,6 +109,7 @@ struct _vcs_dpi_ip_params_st dcn2_1_ip = { .max_page_table_levels = 4, .pte_chunk_size_kbytes = 2, .meta_chunk_size_kbytes = 2, + .min_meta_chunk_size_bytes = 256, .writeback_chunk_size_kbytes = 2, .line_buffer_size_bits = 789504, .is_line_buffer_bpp_fixed = 0, @@ -883,7 +884,8 @@ static const struct dc_debug_options debug_defaults_drv = { .disable_48mhz_pwrdwn = false, .usbc_combo_phy_reset_wa = true, .dmub_command_table = true, - .use_max_lb = true + .use_max_lb = true, + .optimize_edp_link_rate = true }; static const struct dc_debug_options debug_defaults_diags = { diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hubbub.c index c0980da6dc49..f4414de96acc 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hubbub.c +++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hubbub.c @@ -451,6 +451,7 @@ static const struct hubbub_funcs hubbub30_funcs = { .force_wm_propagate_to_pipes = hubbub3_force_wm_propagate_to_pipes, .force_pstate_change_control = hubbub3_force_pstate_change_control, .init_watermarks = hubbub3_init_watermarks, + .hubbub_read_state = hubbub2_read_state, }; void hubbub3_construct(struct dcn20_hubbub *hubbub3, diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hubbub.h b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hubbub.h index c0bd0fb09455..7b597908b937 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hubbub.h +++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hubbub.h @@ -87,7 +87,19 @@ HUBBUB_SF(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_A, DCHUBBUB_ARB_VM_ROW_ALLOW_DRAM_CLK_CHANGE_WATERMARK_A, mask_sh), \ HUBBUB_SF(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_B, DCHUBBUB_ARB_VM_ROW_ALLOW_DRAM_CLK_CHANGE_WATERMARK_B, mask_sh), \ HUBBUB_SF(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_C, DCHUBBUB_ARB_VM_ROW_ALLOW_DRAM_CLK_CHANGE_WATERMARK_C, mask_sh), \ - HUBBUB_SF(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_D, DCHUBBUB_ARB_VM_ROW_ALLOW_DRAM_CLK_CHANGE_WATERMARK_D, mask_sh) + HUBBUB_SF(DCHUBBUB_ARB_ALLOW_DRAM_CLK_CHANGE_WATERMARK_D, DCHUBBUB_ARB_VM_ROW_ALLOW_DRAM_CLK_CHANGE_WATERMARK_D, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_ADDR_MSB, DCN_VM_FAULT_ADDR_MSB, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_ADDR_LSB, DCN_VM_FAULT_ADDR_LSB, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_CNTL, DCN_VM_ERROR_STATUS_CLEAR, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_CNTL, DCN_VM_ERROR_STATUS_MODE, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_CNTL, DCN_VM_ERROR_INTERRUPT_ENABLE, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_CNTL, DCN_VM_RANGE_FAULT_DISABLE, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_CNTL, DCN_VM_PRQ_FAULT_DISABLE, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_STATUS, DCN_VM_ERROR_STATUS, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_STATUS, DCN_VM_ERROR_VMID, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_STATUS, DCN_VM_ERROR_TABLE_LEVEL, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_STATUS, DCN_VM_ERROR_PIPE, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_STATUS, DCN_VM_ERROR_INTERRUPT_STATUS, mask_sh) void hubbub3_construct(struct dcn20_hubbub *hubbub3, struct dc_context *ctx, diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c index c68e3a708a33..2e8ab9775fa3 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_hwseq.c @@ -580,22 +580,19 @@ void dcn30_init_hw(struct dc *dc) */ if (dc->config.power_down_display_on_boot) { struct dc_link *edp_links[MAX_NUM_EDP]; - struct dc_link *edp_link; + struct dc_link *edp_link = NULL; get_edp_links(dc, edp_links, &edp_num); - if (edp_num) { - for (i = 0; i < edp_num; i++) { - edp_link = edp_links[i]; - if (edp_link->link_enc->funcs->is_dig_enabled && - edp_link->link_enc->funcs->is_dig_enabled(edp_link->link_enc) && - dc->hwss.edp_backlight_control && - dc->hwss.power_down && - dc->hwss.edp_power_control) { - dc->hwss.edp_backlight_control(edp_link, false); - dc->hwss.power_down(dc); - dc->hwss.edp_power_control(edp_link, false); - } - } + if (edp_num) + edp_link = edp_links[0]; + if (edp_link && edp_link->link_enc->funcs->is_dig_enabled && + edp_link->link_enc->funcs->is_dig_enabled(edp_link->link_enc) && + dc->hwss.edp_backlight_control && + dc->hwss.power_down && + dc->hwss.edp_power_control) { + dc->hwss.edp_backlight_control(edp_link, false); + dc->hwss.power_down(dc); + dc->hwss.edp_power_control(edp_link, false); } else { for (i = 0; i < dc->link_count; i++) { struct dc_link *link = dc->links[i]; diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_optc.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_optc.c index f37e8254df21..089be7347591 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_optc.c +++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_optc.c @@ -109,11 +109,9 @@ void optc3_lock(struct timing_generator *optc) REG_SET(OTG_MASTER_UPDATE_LOCK, 0, OTG_MASTER_UPDATE_LOCK, 1); - /* Should be fast, status does not update on maximus */ - if (optc->ctx->dce_environment != DCE_ENV_FPGA_MAXIMUS) - REG_WAIT(OTG_MASTER_UPDATE_LOCK, - UPDATE_LOCK_STATUS, 1, - 1, 10); + REG_WAIT(OTG_MASTER_UPDATE_LOCK, + UPDATE_LOCK_STATUS, 1, + 1, 10); } void optc3_set_out_mux(struct timing_generator *optc, enum otg_out_mux_dest dest) diff --git a/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_hubbub.c index a0b96b3c083f..1e3bd2e9cdcc 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_hubbub.c +++ b/drivers/gpu/drm/amd/display/dc/dcn301/dcn301_hubbub.c @@ -62,6 +62,7 @@ static const struct hubbub_funcs hubbub301_funcs = { .is_allow_self_refresh_enabled = hubbub1_is_allow_self_refresh_enabled, .force_wm_propagate_to_pipes = hubbub3_force_wm_propagate_to_pipes, .force_pstate_change_control = hubbub3_force_pstate_change_control, + .hubbub_read_state = hubbub2_read_state, }; void hubbub301_construct(struct dcn20_hubbub *hubbub3, diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hubbub.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hubbub.c index bb9648488900..90c73a1cb986 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hubbub.c +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hubbub.c @@ -124,8 +124,8 @@ static void dcn31_program_compbuf_size(struct hubbub *hubbub, unsigned int compb ASSERT(hubbub2->det0_size + hubbub2->det1_size + hubbub2->det2_size + hubbub2->det3_size + compbuf_size_segments <= hubbub2->crb_size_segs); REG_UPDATE(DCHUBBUB_COMPBUF_CTRL, COMPBUF_SIZE, compbuf_size_segments); - REG_WAIT(DCHUBBUB_COMPBUF_CTRL, COMPBUF_SIZE_CURRENT, compbuf_size_segments, 1, 100); hubbub2->compbuf_size_segments = compbuf_size_segments; + ASSERT(REG_GET(DCHUBBUB_COMPBUF_CTRL, CONFIG_ERROR, &compbuf_size_segments) && !compbuf_size_segments); } } @@ -876,7 +876,33 @@ static bool hubbub31_get_dcc_compression_cap(struct hubbub *hubbub, static int hubbub31_init_dchub_sys_ctx(struct hubbub *hubbub, struct dcn_hubbub_phys_addr_config *pa_config) { - hubbub3_init_dchub_sys_ctx(hubbub, pa_config); + struct dcn20_hubbub *hubbub2 = TO_DCN20_HUBBUB(hubbub); + struct dcn_vmid_page_table_config phys_config; + + REG_SET(DCN_VM_FB_LOCATION_BASE, 0, + FB_BASE, pa_config->system_aperture.fb_base >> 24); + REG_SET(DCN_VM_FB_LOCATION_TOP, 0, + FB_TOP, pa_config->system_aperture.fb_top >> 24); + REG_SET(DCN_VM_FB_OFFSET, 0, + FB_OFFSET, pa_config->system_aperture.fb_offset >> 24); + REG_SET(DCN_VM_AGP_BOT, 0, + AGP_BOT, pa_config->system_aperture.agp_bot >> 24); + REG_SET(DCN_VM_AGP_TOP, 0, + AGP_TOP, pa_config->system_aperture.agp_top >> 24); + REG_SET(DCN_VM_AGP_BASE, 0, + AGP_BASE, pa_config->system_aperture.agp_base >> 24); + + if (pa_config->gart_config.page_table_start_addr != pa_config->gart_config.page_table_end_addr) { + phys_config.page_table_start_addr = pa_config->gart_config.page_table_start_addr >> 12; + phys_config.page_table_end_addr = pa_config->gart_config.page_table_end_addr >> 12; + phys_config.page_table_base_addr = pa_config->gart_config.page_table_base_addr; + phys_config.depth = 0; + phys_config.block_size = 0; + // Init VMID 0 based on PA config + dcn20_vmid_setup(&hubbub2->vmid[0], &phys_config); + + dcn20_vmid_setup(&hubbub2->vmid[15], &phys_config); + } dcn21_dchvm_init(hubbub); @@ -934,7 +960,8 @@ static const struct hubbub_funcs hubbub31_funcs = { .is_allow_self_refresh_enabled = hubbub1_is_allow_self_refresh_enabled, .program_det_size = dcn31_program_det_size, .program_compbuf_size = dcn31_program_compbuf_size, - .init_crb = dcn31_init_crb + .init_crb = dcn31_init_crb, + .hubbub_read_state = hubbub2_read_state, }; void hubbub31_construct(struct dcn20_hubbub *hubbub31, diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hubbub.h b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hubbub.h index 8ec98cbcbd47..e3a654bf04e8 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hubbub.h +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hubbub.h @@ -98,6 +98,7 @@ HUBBUB_SF(DCHUBBUB_DET3_CTRL, DET3_SIZE_CURRENT, mask_sh),\ HUBBUB_SF(DCHUBBUB_COMPBUF_CTRL, COMPBUF_SIZE, mask_sh),\ HUBBUB_SF(DCHUBBUB_COMPBUF_CTRL, COMPBUF_SIZE_CURRENT, mask_sh),\ + HUBBUB_SF(DCHUBBUB_COMPBUF_CTRL, CONFIG_ERROR, mask_sh),\ HUBBUB_SF(COMPBUF_RESERVED_SPACE, COMPBUF_RESERVED_SPACE_64B, mask_sh),\ HUBBUB_SF(COMPBUF_RESERVED_SPACE, COMPBUF_RESERVED_SPACE_ZS, mask_sh),\ HUBBUB_SF(DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_Z8_A, DCHUBBUB_ARB_ALLOW_SR_ENTER_WATERMARK_Z8_A, mask_sh), \ @@ -107,7 +108,19 @@ HUBBUB_SF(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_Z8_A, DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_Z8_A, mask_sh), \ HUBBUB_SF(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_Z8_B, DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_Z8_B, mask_sh), \ HUBBUB_SF(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_Z8_C, DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_Z8_C, mask_sh), \ - HUBBUB_SF(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_Z8_D, DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_Z8_D, mask_sh) + HUBBUB_SF(DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_Z8_D, DCHUBBUB_ARB_ALLOW_SR_EXIT_WATERMARK_Z8_D, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_ADDR_MSB, DCN_VM_FAULT_ADDR_MSB, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_ADDR_LSB, DCN_VM_FAULT_ADDR_LSB, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_CNTL, DCN_VM_ERROR_STATUS_CLEAR, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_CNTL, DCN_VM_ERROR_STATUS_MODE, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_CNTL, DCN_VM_ERROR_INTERRUPT_ENABLE, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_CNTL, DCN_VM_RANGE_FAULT_DISABLE, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_CNTL, DCN_VM_PRQ_FAULT_DISABLE, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_STATUS, DCN_VM_ERROR_STATUS, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_STATUS, DCN_VM_ERROR_VMID, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_STATUS, DCN_VM_ERROR_TABLE_LEVEL, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_STATUS, DCN_VM_ERROR_PIPE, mask_sh), \ + HUBBUB_SF(DCN_VM_FAULT_STATUS, DCN_VM_ERROR_INTERRUPT_STATUS, mask_sh) void hubbub31_construct(struct dcn20_hubbub *hubbub3, diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c index 6ac6faf0c533..83f7904630e6 100644 --- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c +++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_hwseq.c @@ -226,6 +226,7 @@ void dcn31_init_hw(struct dc *dc) if (dc->config.power_down_display_on_boot) { struct dc_link *edp_links[MAX_NUM_EDP]; struct dc_link *edp_link; + bool power_down = false; get_edp_links(dc, edp_links, &edp_num); if (edp_num) { @@ -239,9 +240,11 @@ void dcn31_init_hw(struct dc *dc) dc->hwss.edp_backlight_control(edp_link, false); dc->hwss.power_down(dc); dc->hwss.edp_power_control(edp_link, false); + power_down = true; } } - } else { + } + if (!power_down) { for (i = 0; i < dc->link_count; i++) { struct dc_link *link = dc->links[i]; diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_mode_vba_21.c b/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_mode_vba_21.c index d25a7d38d21f..4136eb8256cb 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_mode_vba_21.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn21/display_mode_vba_21.c @@ -841,6 +841,9 @@ static bool CalculatePrefetchSchedule( else *DestinationLinesForPrefetch = dst_y_prefetch_equ; + // Limit to prevent overflow in DST_Y_PREFETCH register + *DestinationLinesForPrefetch = dml_min(*DestinationLinesForPrefetch, 63.75); + dml_print("DML: VStartup: %d\n", VStartup); dml_print("DML: TCalc: %f\n", TCalc); dml_print("DML: TWait: %f\n", TWait); @@ -2267,7 +2270,7 @@ static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerforman &locals->UrgentBurstFactorLumaPre[k], &locals->UrgentBurstFactorChroma[k], &locals->UrgentBurstFactorChromaPre[k], - &locals->NotEnoughUrgentLatencyHiding, + &locals->NotEnoughUrgentLatencyHiding[0][0], &locals->NotEnoughUrgentLatencyHidingPre); if (mode_lib->vba.UseUrgentBurstBandwidth == false) { @@ -2300,7 +2303,8 @@ static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerforman } mode_lib->vba.FractionOfUrgentBandwidth = MaxTotalRDBandwidthNoUrgentBurst / mode_lib->vba.ReturnBW; - if (MaxTotalRDBandwidth <= mode_lib->vba.ReturnBW && locals->NotEnoughUrgentLatencyHiding == 0 && locals->NotEnoughUrgentLatencyHidingPre == 0 && !VRatioPrefetchMoreThan4 + if (MaxTotalRDBandwidth <= mode_lib->vba.ReturnBW && locals->NotEnoughUrgentLatencyHiding[0][0] == 0 && + locals->NotEnoughUrgentLatencyHidingPre == 0 && !VRatioPrefetchMoreThan4 && !DestinationLineTimesForPrefetchLessThan2) mode_lib->vba.PrefetchModeSupported = true; else { @@ -4821,7 +4825,7 @@ void dml21_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l &locals->UrgentBurstFactorLumaPre[k], &locals->UrgentBurstFactorChroma[k], &locals->UrgentBurstFactorChromaPre[k], - &locals->NotEnoughUrgentLatencyHiding, + &locals->NotEnoughUrgentLatencyHiding[0][0], &locals->NotEnoughUrgentLatencyHidingPre); if (mode_lib->vba.UseUrgentBurstBandwidth == false) { @@ -4848,13 +4852,13 @@ void dml21_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l } locals->BandwidthWithoutPrefetchSupported[i][0] = true; if (mode_lib->vba.MaximumReadBandwidthWithoutPrefetch > locals->ReturnBWPerState[i][0] - || locals->NotEnoughUrgentLatencyHiding == 1) { + || locals->NotEnoughUrgentLatencyHiding[0][0] == 1) { locals->BandwidthWithoutPrefetchSupported[i][0] = false; } locals->PrefetchSupported[i][j] = true; if (mode_lib->vba.MaximumReadBandwidthWithPrefetch > locals->ReturnBWPerState[i][0] - || locals->NotEnoughUrgentLatencyHiding == 1 + || locals->NotEnoughUrgentLatencyHiding[0][0] == 1 || locals->NotEnoughUrgentLatencyHidingPre == 1) { locals->PrefetchSupported[i][j] = false; } diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_mode_vba_30.c b/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_mode_vba_30.c index 9d2016d8fafe..e1a961a62add 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_mode_vba_30.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_mode_vba_30.c @@ -2596,7 +2596,7 @@ static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerforman } } - v->NotEnoughUrgentLatencyHiding = false; + v->NotEnoughUrgentLatencyHiding[0][0] = false; v->NotEnoughUrgentLatencyHidingPre = false; for (k = 0; k < v->NumberOfActivePlanes; ++k) { @@ -2681,7 +2681,7 @@ static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerforman if (v->VRatioPrefetchY[k] > 4 || v->VRatioPrefetchC[k] > 4) VRatioPrefetchMoreThan4 = true; if (v->NoUrgentLatencyHiding[k] == true) - v->NotEnoughUrgentLatencyHiding = true; + v->NotEnoughUrgentLatencyHiding[0][0] = true; if (v->NoUrgentLatencyHidingPre[k] == true) v->NotEnoughUrgentLatencyHidingPre = true; @@ -2689,7 +2689,8 @@ static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerforman v->FractionOfUrgentBandwidth = MaxTotalRDBandwidthNoUrgentBurst / v->ReturnBW; - if (MaxTotalRDBandwidth <= v->ReturnBW && v->NotEnoughUrgentLatencyHiding == 0 && v->NotEnoughUrgentLatencyHidingPre == 0 && !VRatioPrefetchMoreThan4 + if (MaxTotalRDBandwidth <= v->ReturnBW && v->NotEnoughUrgentLatencyHiding[0][0] == 0 + && v->NotEnoughUrgentLatencyHidingPre == 0 && !VRatioPrefetchMoreThan4 && !DestinationLineTimesForPrefetchLessThan2) v->PrefetchModeSupported = true; else { @@ -2794,8 +2795,9 @@ static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerforman } v->VStartupLines = v->VStartupLines + 1; - v->PrefetchAndImmediateFlipSupported = (v->PrefetchModeSupported == true && ((!v->ImmediateFlipSupport && !v->HostVMEnable && v->ImmediateFlipRequirement != dm_immediate_flip_required) || v->ImmediateFlipSupported)) ? true : false; - + v->PrefetchModeSupported = (v->PrefetchModeSupported == true && ((!v->ImmediateFlipSupport && + !v->HostVMEnable && v->ImmediateFlipRequirement[0] != dm_immediate_flip_required) || + v->ImmediateFlipSupported)) ? true : false; } while (!v->PrefetchModeSupported && v->VStartupLines <= v->MaximumMaxVStartupLines); ASSERT(v->PrefetchModeSupported); @@ -4753,7 +4755,7 @@ void dml30_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l v->HostVMMinPageSize, v->HostVMMaxNonCachedPageTableLevels, v->DynamicMetadataVMEnabled, - v->ImmediateFlipRequirement, + v->ImmediateFlipRequirement[0], v->ProgressiveToInterlaceUnitInOPP, v->MaxAveragePercentOfIdealSDPPortBWDisplayCanUseInNormalSystemOperation, v->PercentOfIdealDRAMFabricAndSDPPortBWReceivedAfterUrgLatencyPixelMixedWithVMData, @@ -5164,7 +5166,7 @@ void dml30_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l v->NextMaxVStartup = v->NextMaxVStartup - 1; } } while (!((v->PrefetchSupported[i][j] == true && v->DynamicMetadataSupported[i][j] == true && v->VRatioInPrefetchSupported[i][j] == true - && ((v->HostVMEnable == false && v->ImmediateFlipRequirement != dm_immediate_flip_required) + && ((v->HostVMEnable == false && v->ImmediateFlipRequirement[0] != dm_immediate_flip_required) || v->ImmediateFlipSupportedForState[i][j] == true)) || (v->NextMaxVStartup == v->MaxMaxVStartup[i][j] && NextPrefetchModeState > MaxPrefetchMode))); @@ -5305,7 +5307,7 @@ void dml30_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l && ViewportExceedsSurface == 0 && v->PrefetchSupported[i][j] == 1 && v->DynamicMetadataSupported[i][j] == 1 && v->TotalVerticalActiveBandwidthSupport[i][j] == 1 && v->VRatioInPrefetchSupported[i][j] == 1 && v->PTEBufferSizeNotExceeded[i][j] == 1 && v->NonsupportedDSCInputBPC == 0 - && ((v->HostVMEnable == 0 && v->ImmediateFlipRequirement != dm_immediate_flip_required) + && ((v->HostVMEnable == 0 && v->ImmediateFlipRequirement[0] != dm_immediate_flip_required) || v->ImmediateFlipSupportedForState[i][j] == true)) { v->ModeSupport[i][j] = true; } else { diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c b/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c index a9667068c690..ce55c9caf9a2 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c +++ b/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c @@ -3036,10 +3036,9 @@ static void DISPCLKDPPCLKDCFCLKDeepSleepPrefetchParametersWatermarksAndPerforman } v->PrefetchAndImmediateFlipSupported = - (v->PrefetchModeSupported == true - && ((!v->ImmediateFlipSupport && !v->HostVMEnable - && v->ImmediateFlipRequirement != dm_immediate_flip_required) || v->ImmediateFlipSupported)) ? - true : false; + (v->PrefetchModeSupported == true && ((!v->ImmediateFlipSupport && !v->HostVMEnable + && v->ImmediateFlipRequirement[0] != dm_immediate_flip_required) || + v->ImmediateFlipSupported)) ? true : false; #ifdef __DML_VBA_DEBUG__ dml_print("DML::%s: PrefetchModeSupported %d\n", __func__, v->PrefetchModeSupported); dml_print("DML::%s: ImmediateFlipRequirement %d\n", __func__, v->ImmediateFlipRequirement == dm_immediate_flip_required); @@ -5103,7 +5102,7 @@ void dml31_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l v->HostVMMinPageSize, v->HostVMMaxNonCachedPageTableLevels, v->DynamicMetadataVMEnabled, - v->ImmediateFlipRequirement, + v->ImmediateFlipRequirement[0], v->ProgressiveToInterlaceUnitInOPP, v->MaxAveragePercentOfIdealFabricAndSDPPortBWDisplayCanUseInNormalSystemOperation, v->PercentOfIdealFabricAndSDPPortBWReceivedAfterUrgLatency, @@ -5542,7 +5541,8 @@ void dml31_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l } v->NextPrefetchMode = v->NextPrefetchMode + 1; } while (!((v->PrefetchSupported[i][j] == true && v->DynamicMetadataSupported[i][j] == true && v->VRatioInPrefetchSupported[i][j] == true - && ((v->HostVMEnable == false && v->ImmediateFlipRequirement != dm_immediate_flip_required) + && ((v->HostVMEnable == false && + v->ImmediateFlipRequirement[0] != dm_immediate_flip_required) || v->ImmediateFlipSupportedForState[i][j] == true)) || (v->NextMaxVStartup == v->MaxMaxVStartup[i][j] && NextPrefetchModeState > MaxPrefetchMode))); @@ -5702,7 +5702,8 @@ void dml31_ModeSupportAndSystemConfigurationFull(struct display_mode_lib *mode_l && v->PrefetchSupported[i][j] == true && v->DynamicMetadataSupported[i][j] == true && v->TotalVerticalActiveBandwidthSupport[i][j] == true && v->VRatioInPrefetchSupported[i][j] == true && v->PTEBufferSizeNotExceeded[i][j] == true && v->NonsupportedDSCInputBPC == false - && ((v->HostVMEnable == false && v->ImmediateFlipRequirement != dm_immediate_flip_required) + && ((v->HostVMEnable == false + && v->ImmediateFlipRequirement[0] != dm_immediate_flip_required) || v->ImmediateFlipSupportedForState[i][j] == true) && FMTBufferExceeded == false) { v->ModeSupport[i][j] = true; diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_enums.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_enums.h index 64f9c735f74d..1051ca1a23b8 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_enums.h +++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_enums.h @@ -109,7 +109,9 @@ enum clock_change_support { }; enum output_standard { - dm_std_uninitialized = 0, dm_std_cvtr2, dm_std_cvt + dm_std_uninitialized = 0, + dm_std_cvtr2, + dm_std_cvt }; enum mpc_combine_affinity { diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h index 64daa0507393..d46a2733024c 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h +++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_structs.h @@ -536,6 +536,8 @@ struct _vcs_dpi_display_rq_regs_st { unsigned int mrq_expansion_mode; unsigned int crq_expansion_mode; unsigned int plane1_base_address; + unsigned int aperture_low_addr; // bits [47:18] + unsigned int aperture_high_addr; // bits [47:18] }; struct _vcs_dpi_display_dlg_sys_params_st { diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.c b/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.c index d3b1b6d4ce2f..73f5be26abc4 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.c +++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.c @@ -396,7 +396,6 @@ static void fetch_pipe_params(struct display_mode_lib *mode_lib) mode_lib->vba.NumberOfActivePlanes = 0; mode_lib->vba.ImmediateFlipSupport = false; - mode_lib->vba.ImmediateFlipRequirement = dm_immediate_flip_not_required; for (j = 0; j < mode_lib->vba.cache_num_pipes; ++j) { display_pipe_source_params_st *src = &pipes[j].pipe.src; display_pipe_dest_params_st *dst = &pipes[j].pipe.dest; @@ -409,6 +408,7 @@ static void fetch_pipe_params(struct display_mode_lib *mode_lib) continue; visited[j] = true; + mode_lib->vba.ImmediateFlipRequirement[j] = dm_immediate_flip_not_required; mode_lib->vba.pipe_plane[j] = mode_lib->vba.NumberOfActivePlanes; mode_lib->vba.DPPPerPlane[mode_lib->vba.NumberOfActivePlanes] = 1; mode_lib->vba.SourceScan[mode_lib->vba.NumberOfActivePlanes] = @@ -667,9 +667,9 @@ static void fetch_pipe_params(struct display_mode_lib *mode_lib) mode_lib->vba.ViewportHeightChroma[mode_lib->vba.NumberOfActivePlanes] = src->viewport_height_max / vdiv_c; } - if (pipes[k].pipe.src.immediate_flip) { + if (pipes[j].pipe.src.immediate_flip) { mode_lib->vba.ImmediateFlipSupport = true; - mode_lib->vba.ImmediateFlipRequirement = dm_immediate_flip_required; + mode_lib->vba.ImmediateFlipRequirement[j] = dm_immediate_flip_required; } mode_lib->vba.NumberOfActivePlanes++; @@ -845,9 +845,10 @@ void PixelClockAdjustmentForProgressiveToInterlaceUnit(struct display_mode_lib * //Progressive To Interlace Unit Effect for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { + mode_lib->vba.PixelClockBackEnd[k] = mode_lib->vba.PixelClock[k]; if (mode_lib->vba.Interlace[k] == 1 && mode_lib->vba.ProgressiveToInterlaceUnitInOPP == true) { - mode_lib->vba.PixelClock[k] = 2 * mode_lib->vba.PixelClockBackEnd[k]; + mode_lib->vba.PixelClock[k] = 2 * mode_lib->vba.PixelClock[k]; } } } @@ -890,8 +891,9 @@ void ModeSupportAndSystemConfiguration(struct display_mode_lib *mode_lib) mode_lib->vba.DISPCLK = soc->clock_limits[mode_lib->vba.VoltageLevel].dispclk_mhz; // Total Available Pipes Support Check - for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) + for (k = 0; k < mode_lib->vba.NumberOfActivePlanes; ++k) { total_pipes += mode_lib->vba.DPPPerPlane[k]; + } ASSERT(total_pipes <= DC__NUM_DPP__MAX); } diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.h index d18a021d4d32..90e87961fe3e 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.h +++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.h @@ -676,7 +676,7 @@ struct vba_vars_st { double AlignedDCCMetaPitchY[DC__NUM_DPP__MAX]; double AlignedDCCMetaPitchC[DC__NUM_DPP__MAX]; - unsigned int NotEnoughUrgentLatencyHiding; + unsigned int NotEnoughUrgentLatencyHiding[DC__VOLTAGE_STATES][2]; unsigned int NotEnoughUrgentLatencyHidingPre; int PTEBufferSizeInRequestsForLuma; int PTEBufferSizeInRequestsForChroma; @@ -877,7 +877,7 @@ struct vba_vars_st { int PercentMarginOverMinimumRequiredDCFCLK; bool DynamicMetadataSupported[DC__VOLTAGE_STATES][2]; - enum immediate_flip_requirement ImmediateFlipRequirement; + enum immediate_flip_requirement ImmediateFlipRequirement[DC__NUM_DPP__MAX]; unsigned int DETBufferSizeYThisState[DC__NUM_DPP__MAX]; unsigned int DETBufferSizeCThisState[DC__NUM_DPP__MAX]; bool NoUrgentLatencyHiding[DC__NUM_DPP__MAX]; diff --git a/drivers/gpu/drm/amd/display/dc/dsc/dc_dsc.c b/drivers/gpu/drm/amd/display/dc/dsc/dc_dsc.c index f403d8e84a8c..f5b7da0e64c0 100644 --- a/drivers/gpu/drm/amd/display/dc/dsc/dc_dsc.c +++ b/drivers/gpu/drm/amd/display/dc/dsc/dc_dsc.c @@ -28,6 +28,7 @@ #include <drm/drm_dp_helper.h> #include "dc.h" #include "rc_calc.h" +#include "fixed31_32.h" /* This module's internal functions */ @@ -39,6 +40,47 @@ static bool dsc_policy_enable_dsc_when_not_needed; static bool dsc_policy_disable_dsc_stream_overhead; +/* Forward Declerations */ +static void get_dsc_bandwidth_range( + const uint32_t min_bpp_x16, + const uint32_t max_bpp_x16, + const uint32_t num_slices_h, + const struct dsc_enc_caps *dsc_caps, + const struct dc_crtc_timing *timing, + struct dc_dsc_bw_range *range); + +static uint32_t compute_bpp_x16_from_target_bandwidth( + const uint32_t bandwidth_in_kbps, + const struct dc_crtc_timing *timing, + const uint32_t num_slices_h, + const uint32_t bpp_increment_div, + const bool is_dp); + +static void get_dsc_enc_caps( + const struct display_stream_compressor *dsc, + struct dsc_enc_caps *dsc_enc_caps, + int pixel_clock_100Hz); + +static bool intersect_dsc_caps( + const struct dsc_dec_dpcd_caps *dsc_sink_caps, + const struct dsc_enc_caps *dsc_enc_caps, + enum dc_pixel_encoding pixel_encoding, + struct dsc_enc_caps *dsc_common_caps); + +static bool setup_dsc_config( + const struct dsc_dec_dpcd_caps *dsc_sink_caps, + const struct dsc_enc_caps *dsc_enc_caps, + int target_bandwidth_kbps, + const struct dc_crtc_timing *timing, + int min_slice_height_override, + int max_dsc_target_bpp_limit_override_x16, + struct dc_dsc_config *dsc_cfg); + +static struct fixed31_32 compute_dsc_max_bandwidth_overhead( + const struct dc_crtc_timing *timing, + const int num_slices_h, + const bool is_dp); + static bool dsc_buff_block_size_from_dpcd(int dpcd_buff_block_size, int *buff_block_size) { @@ -171,10 +213,164 @@ static bool dsc_bpp_increment_div_from_dpcd(uint8_t bpp_increment_dpcd, uint32_t return true; } + + +bool dc_dsc_parse_dsc_dpcd(const struct dc *dc, + const uint8_t *dpcd_dsc_basic_data, + const uint8_t *dpcd_dsc_branch_decoder_caps, + struct dsc_dec_dpcd_caps *dsc_sink_caps) +{ + if (!dpcd_dsc_basic_data) + return false; + + dsc_sink_caps->is_dsc_supported = + (dpcd_dsc_basic_data[DP_DSC_SUPPORT - DP_DSC_SUPPORT] & DP_DSC_DECOMPRESSION_IS_SUPPORTED) != 0; + if (!dsc_sink_caps->is_dsc_supported) + return false; + + dsc_sink_caps->dsc_version = dpcd_dsc_basic_data[DP_DSC_REV - DP_DSC_SUPPORT]; + + { + int buff_block_size; + int buff_size; + + if (!dsc_buff_block_size_from_dpcd(dpcd_dsc_basic_data[DP_DSC_RC_BUF_BLK_SIZE - DP_DSC_SUPPORT], + &buff_block_size)) + return false; + + buff_size = dpcd_dsc_basic_data[DP_DSC_RC_BUF_SIZE - DP_DSC_SUPPORT] + 1; + dsc_sink_caps->rc_buffer_size = buff_size * buff_block_size; + } + + dsc_sink_caps->slice_caps1.raw = dpcd_dsc_basic_data[DP_DSC_SLICE_CAP_1 - DP_DSC_SUPPORT]; + if (!dsc_line_buff_depth_from_dpcd(dpcd_dsc_basic_data[DP_DSC_LINE_BUF_BIT_DEPTH - DP_DSC_SUPPORT], + &dsc_sink_caps->lb_bit_depth)) + return false; + + dsc_sink_caps->is_block_pred_supported = + (dpcd_dsc_basic_data[DP_DSC_BLK_PREDICTION_SUPPORT - DP_DSC_SUPPORT] & + DP_DSC_BLK_PREDICTION_IS_SUPPORTED) != 0; + + dsc_sink_caps->edp_max_bits_per_pixel = + dpcd_dsc_basic_data[DP_DSC_MAX_BITS_PER_PIXEL_LOW - DP_DSC_SUPPORT] | + dpcd_dsc_basic_data[DP_DSC_MAX_BITS_PER_PIXEL_HI - DP_DSC_SUPPORT] << 8; + + dsc_sink_caps->color_formats.raw = dpcd_dsc_basic_data[DP_DSC_DEC_COLOR_FORMAT_CAP - DP_DSC_SUPPORT]; + dsc_sink_caps->color_depth.raw = dpcd_dsc_basic_data[DP_DSC_DEC_COLOR_DEPTH_CAP - DP_DSC_SUPPORT]; + + { + int dpcd_throughput = dpcd_dsc_basic_data[DP_DSC_PEAK_THROUGHPUT - DP_DSC_SUPPORT]; + + if (!dsc_throughput_from_dpcd(dpcd_throughput & DP_DSC_THROUGHPUT_MODE_0_MASK, + &dsc_sink_caps->throughput_mode_0_mps)) + return false; + + dpcd_throughput = (dpcd_throughput & DP_DSC_THROUGHPUT_MODE_1_MASK) >> DP_DSC_THROUGHPUT_MODE_1_SHIFT; + if (!dsc_throughput_from_dpcd(dpcd_throughput, &dsc_sink_caps->throughput_mode_1_mps)) + return false; + } + + dsc_sink_caps->max_slice_width = dpcd_dsc_basic_data[DP_DSC_MAX_SLICE_WIDTH - DP_DSC_SUPPORT] * 320; + dsc_sink_caps->slice_caps2.raw = dpcd_dsc_basic_data[DP_DSC_SLICE_CAP_2 - DP_DSC_SUPPORT]; + + if (!dsc_bpp_increment_div_from_dpcd(dpcd_dsc_basic_data[DP_DSC_BITS_PER_PIXEL_INC - DP_DSC_SUPPORT], + &dsc_sink_caps->bpp_increment_div)) + return false; + + if (dc->debug.dsc_bpp_increment_div) { + /* dsc_bpp_increment_div should onl be 1, 2, 4, 8 or 16, but rather than rejecting invalid values, + * we'll accept all and get it into range. This also makes the above check against 0 redundant, + * but that one stresses out the override will be only used if it's not 0. + */ + if (dc->debug.dsc_bpp_increment_div >= 1) + dsc_sink_caps->bpp_increment_div = 1; + if (dc->debug.dsc_bpp_increment_div >= 2) + dsc_sink_caps->bpp_increment_div = 2; + if (dc->debug.dsc_bpp_increment_div >= 4) + dsc_sink_caps->bpp_increment_div = 4; + if (dc->debug.dsc_bpp_increment_div >= 8) + dsc_sink_caps->bpp_increment_div = 8; + if (dc->debug.dsc_bpp_increment_div >= 16) + dsc_sink_caps->bpp_increment_div = 16; + } + + /* Extended caps */ + if (dpcd_dsc_branch_decoder_caps == NULL) { // branch decoder DPCD DSC data can be null for non branch device + dsc_sink_caps->branch_overall_throughput_0_mps = 0; + dsc_sink_caps->branch_overall_throughput_1_mps = 0; + dsc_sink_caps->branch_max_line_width = 0; + return true; + } + + dsc_sink_caps->branch_overall_throughput_0_mps = + dpcd_dsc_branch_decoder_caps[DP_DSC_BRANCH_OVERALL_THROUGHPUT_0 - DP_DSC_BRANCH_OVERALL_THROUGHPUT_0]; + if (dsc_sink_caps->branch_overall_throughput_0_mps == 0) + dsc_sink_caps->branch_overall_throughput_0_mps = 0; + else if (dsc_sink_caps->branch_overall_throughput_0_mps == 1) + dsc_sink_caps->branch_overall_throughput_0_mps = 680; + else { + dsc_sink_caps->branch_overall_throughput_0_mps *= 50; + dsc_sink_caps->branch_overall_throughput_0_mps += 600; + } + + dsc_sink_caps->branch_overall_throughput_1_mps = + dpcd_dsc_branch_decoder_caps[DP_DSC_BRANCH_OVERALL_THROUGHPUT_1 - DP_DSC_BRANCH_OVERALL_THROUGHPUT_0]; + if (dsc_sink_caps->branch_overall_throughput_1_mps == 0) + dsc_sink_caps->branch_overall_throughput_1_mps = 0; + else if (dsc_sink_caps->branch_overall_throughput_1_mps == 1) + dsc_sink_caps->branch_overall_throughput_1_mps = 680; + else { + dsc_sink_caps->branch_overall_throughput_1_mps *= 50; + dsc_sink_caps->branch_overall_throughput_1_mps += 600; + } + + dsc_sink_caps->branch_max_line_width = + dpcd_dsc_branch_decoder_caps[DP_DSC_BRANCH_MAX_LINE_WIDTH - DP_DSC_BRANCH_OVERALL_THROUGHPUT_0] * 320; + ASSERT(dsc_sink_caps->branch_max_line_width == 0 || dsc_sink_caps->branch_max_line_width >= 5120); + + dsc_sink_caps->is_dp = true; + return true; +} + + +/* If DSC is possbile, get DSC bandwidth range based on [min_bpp, max_bpp] target bitrate range and + * timing's pixel clock and uncompressed bandwidth. + * If DSC is not possible, leave '*range' untouched. + */ +bool dc_dsc_compute_bandwidth_range( + const struct display_stream_compressor *dsc, + uint32_t dsc_min_slice_height_override, + uint32_t min_bpp_x16, + uint32_t max_bpp_x16, + const struct dsc_dec_dpcd_caps *dsc_sink_caps, + const struct dc_crtc_timing *timing, + struct dc_dsc_bw_range *range) +{ + bool is_dsc_possible = false; + struct dsc_enc_caps dsc_enc_caps; + struct dsc_enc_caps dsc_common_caps; + struct dc_dsc_config config; + + get_dsc_enc_caps(dsc, &dsc_enc_caps, timing->pix_clk_100hz); + + is_dsc_possible = intersect_dsc_caps(dsc_sink_caps, &dsc_enc_caps, + timing->pixel_encoding, &dsc_common_caps); + + if (is_dsc_possible) + is_dsc_possible = setup_dsc_config(dsc_sink_caps, &dsc_enc_caps, 0, timing, + dsc_min_slice_height_override, max_bpp_x16, &config); + + if (is_dsc_possible) + get_dsc_bandwidth_range(min_bpp_x16, max_bpp_x16, + config.num_slices_h, &dsc_common_caps, timing, range); + + return is_dsc_possible; +} + static void get_dsc_enc_caps( - const struct display_stream_compressor *dsc, - struct dsc_enc_caps *dsc_enc_caps, - int pixel_clock_100Hz) + const struct display_stream_compressor *dsc, + struct dsc_enc_caps *dsc_enc_caps, + int pixel_clock_100Hz) { // This is a static HW query, so we can use any DSC @@ -187,14 +383,14 @@ static void get_dsc_enc_caps( } } -/* Returns 'false' if no intersection was found for at least one capablity. +/* Returns 'false' if no intersection was found for at least one capability. * It also implicitly validates some sink caps against invalid value of zero. */ static bool intersect_dsc_caps( - const struct dsc_dec_dpcd_caps *dsc_sink_caps, - const struct dsc_enc_caps *dsc_enc_caps, - enum dc_pixel_encoding pixel_encoding, - struct dsc_enc_caps *dsc_common_caps) + const struct dsc_dec_dpcd_caps *dsc_sink_caps, + const struct dsc_enc_caps *dsc_enc_caps, + enum dc_pixel_encoding pixel_encoding, + struct dsc_enc_caps *dsc_common_caps) { int32_t max_slices; int32_t total_sink_throughput; @@ -205,10 +401,14 @@ static bool intersect_dsc_caps( if (!dsc_common_caps->dsc_version) return false; - dsc_common_caps->slice_caps.bits.NUM_SLICES_1 = dsc_sink_caps->slice_caps1.bits.NUM_SLICES_1 && dsc_enc_caps->slice_caps.bits.NUM_SLICES_1; - dsc_common_caps->slice_caps.bits.NUM_SLICES_2 = dsc_sink_caps->slice_caps1.bits.NUM_SLICES_2 && dsc_enc_caps->slice_caps.bits.NUM_SLICES_2; - dsc_common_caps->slice_caps.bits.NUM_SLICES_4 = dsc_sink_caps->slice_caps1.bits.NUM_SLICES_4 && dsc_enc_caps->slice_caps.bits.NUM_SLICES_4; - dsc_common_caps->slice_caps.bits.NUM_SLICES_8 = dsc_sink_caps->slice_caps1.bits.NUM_SLICES_8 && dsc_enc_caps->slice_caps.bits.NUM_SLICES_8; + dsc_common_caps->slice_caps.bits.NUM_SLICES_1 = + dsc_sink_caps->slice_caps1.bits.NUM_SLICES_1 && dsc_enc_caps->slice_caps.bits.NUM_SLICES_1; + dsc_common_caps->slice_caps.bits.NUM_SLICES_2 = + dsc_sink_caps->slice_caps1.bits.NUM_SLICES_2 && dsc_enc_caps->slice_caps.bits.NUM_SLICES_2; + dsc_common_caps->slice_caps.bits.NUM_SLICES_4 = + dsc_sink_caps->slice_caps1.bits.NUM_SLICES_4 && dsc_enc_caps->slice_caps.bits.NUM_SLICES_4; + dsc_common_caps->slice_caps.bits.NUM_SLICES_8 = + dsc_sink_caps->slice_caps1.bits.NUM_SLICES_8 && dsc_enc_caps->slice_caps.bits.NUM_SLICES_8; if (!dsc_common_caps->slice_caps.raw) return false; @@ -216,7 +416,8 @@ static bool intersect_dsc_caps( if (!dsc_common_caps->lb_bit_depth) return false; - dsc_common_caps->is_block_pred_supported = dsc_sink_caps->is_block_pred_supported && dsc_enc_caps->is_block_pred_supported; + dsc_common_caps->is_block_pred_supported = + dsc_sink_caps->is_block_pred_supported && dsc_enc_caps->is_block_pred_supported; dsc_common_caps->color_formats.raw = dsc_sink_caps->color_formats.raw & dsc_enc_caps->color_formats.raw; if (!dsc_common_caps->color_formats.raw) @@ -288,11 +489,11 @@ static struct fixed31_32 compute_dsc_max_bandwidth_overhead( } static uint32_t compute_bpp_x16_from_target_bandwidth( - const uint32_t bandwidth_in_kbps, - const struct dc_crtc_timing *timing, - const uint32_t num_slices_h, - const uint32_t bpp_increment_div, - const bool is_dp) + const uint32_t bandwidth_in_kbps, + const struct dc_crtc_timing *timing, + const uint32_t num_slices_h, + const uint32_t bpp_increment_div, + const bool is_dp) { struct fixed31_32 overhead_in_kbps; struct fixed31_32 effective_bandwidth_in_kbps; @@ -769,146 +970,6 @@ done: return is_dsc_possible; } -bool dc_dsc_parse_dsc_dpcd(const struct dc *dc, const uint8_t *dpcd_dsc_basic_data, const uint8_t *dpcd_dsc_branch_decoder_caps, struct dsc_dec_dpcd_caps *dsc_sink_caps) -{ - if (!dpcd_dsc_basic_data) - return false; - - dsc_sink_caps->is_dsc_supported = (dpcd_dsc_basic_data[DP_DSC_SUPPORT - DP_DSC_SUPPORT] & DP_DSC_DECOMPRESSION_IS_SUPPORTED) != 0; - if (!dsc_sink_caps->is_dsc_supported) - return false; - - dsc_sink_caps->dsc_version = dpcd_dsc_basic_data[DP_DSC_REV - DP_DSC_SUPPORT]; - - { - int buff_block_size; - int buff_size; - - if (!dsc_buff_block_size_from_dpcd(dpcd_dsc_basic_data[DP_DSC_RC_BUF_BLK_SIZE - DP_DSC_SUPPORT], &buff_block_size)) - return false; - - buff_size = dpcd_dsc_basic_data[DP_DSC_RC_BUF_SIZE - DP_DSC_SUPPORT] + 1; - dsc_sink_caps->rc_buffer_size = buff_size * buff_block_size; - } - - dsc_sink_caps->slice_caps1.raw = dpcd_dsc_basic_data[DP_DSC_SLICE_CAP_1 - DP_DSC_SUPPORT]; - if (!dsc_line_buff_depth_from_dpcd(dpcd_dsc_basic_data[DP_DSC_LINE_BUF_BIT_DEPTH - DP_DSC_SUPPORT], &dsc_sink_caps->lb_bit_depth)) - return false; - - dsc_sink_caps->is_block_pred_supported = - (dpcd_dsc_basic_data[DP_DSC_BLK_PREDICTION_SUPPORT - DP_DSC_SUPPORT] & DP_DSC_BLK_PREDICTION_IS_SUPPORTED) != 0; - - dsc_sink_caps->edp_max_bits_per_pixel = - dpcd_dsc_basic_data[DP_DSC_MAX_BITS_PER_PIXEL_LOW - DP_DSC_SUPPORT] | - dpcd_dsc_basic_data[DP_DSC_MAX_BITS_PER_PIXEL_HI - DP_DSC_SUPPORT] << 8; - - dsc_sink_caps->color_formats.raw = dpcd_dsc_basic_data[DP_DSC_DEC_COLOR_FORMAT_CAP - DP_DSC_SUPPORT]; - dsc_sink_caps->color_depth.raw = dpcd_dsc_basic_data[DP_DSC_DEC_COLOR_DEPTH_CAP - DP_DSC_SUPPORT]; - - { - int dpcd_throughput = dpcd_dsc_basic_data[DP_DSC_PEAK_THROUGHPUT - DP_DSC_SUPPORT]; - - if (!dsc_throughput_from_dpcd(dpcd_throughput & DP_DSC_THROUGHPUT_MODE_0_MASK, &dsc_sink_caps->throughput_mode_0_mps)) - return false; - - dpcd_throughput = (dpcd_throughput & DP_DSC_THROUGHPUT_MODE_1_MASK) >> DP_DSC_THROUGHPUT_MODE_1_SHIFT; - if (!dsc_throughput_from_dpcd(dpcd_throughput, &dsc_sink_caps->throughput_mode_1_mps)) - return false; - } - - dsc_sink_caps->max_slice_width = dpcd_dsc_basic_data[DP_DSC_MAX_SLICE_WIDTH - DP_DSC_SUPPORT] * 320; - dsc_sink_caps->slice_caps2.raw = dpcd_dsc_basic_data[DP_DSC_SLICE_CAP_2 - DP_DSC_SUPPORT]; - - if (!dsc_bpp_increment_div_from_dpcd(dpcd_dsc_basic_data[DP_DSC_BITS_PER_PIXEL_INC - DP_DSC_SUPPORT], &dsc_sink_caps->bpp_increment_div)) - return false; - - if (dc->debug.dsc_bpp_increment_div) { - /* dsc_bpp_increment_div should onl be 1, 2, 4, 8 or 16, but rather than rejecting invalid values, - * we'll accept all and get it into range. This also makes the above check against 0 redundant, - * but that one stresses out the override will be only used if it's not 0. - */ - if (dc->debug.dsc_bpp_increment_div >= 1) - dsc_sink_caps->bpp_increment_div = 1; - if (dc->debug.dsc_bpp_increment_div >= 2) - dsc_sink_caps->bpp_increment_div = 2; - if (dc->debug.dsc_bpp_increment_div >= 4) - dsc_sink_caps->bpp_increment_div = 4; - if (dc->debug.dsc_bpp_increment_div >= 8) - dsc_sink_caps->bpp_increment_div = 8; - if (dc->debug.dsc_bpp_increment_div >= 16) - dsc_sink_caps->bpp_increment_div = 16; - } - - /* Extended caps */ - if (dpcd_dsc_branch_decoder_caps == NULL) { // branch decoder DPCD DSC data can be null for non branch device - dsc_sink_caps->branch_overall_throughput_0_mps = 0; - dsc_sink_caps->branch_overall_throughput_1_mps = 0; - dsc_sink_caps->branch_max_line_width = 0; - return true; - } - - dsc_sink_caps->branch_overall_throughput_0_mps = dpcd_dsc_branch_decoder_caps[DP_DSC_BRANCH_OVERALL_THROUGHPUT_0 - DP_DSC_BRANCH_OVERALL_THROUGHPUT_0]; - if (dsc_sink_caps->branch_overall_throughput_0_mps == 0) - dsc_sink_caps->branch_overall_throughput_0_mps = 0; - else if (dsc_sink_caps->branch_overall_throughput_0_mps == 1) - dsc_sink_caps->branch_overall_throughput_0_mps = 680; - else { - dsc_sink_caps->branch_overall_throughput_0_mps *= 50; - dsc_sink_caps->branch_overall_throughput_0_mps += 600; - } - - dsc_sink_caps->branch_overall_throughput_1_mps = dpcd_dsc_branch_decoder_caps[DP_DSC_BRANCH_OVERALL_THROUGHPUT_1 - DP_DSC_BRANCH_OVERALL_THROUGHPUT_0]; - if (dsc_sink_caps->branch_overall_throughput_1_mps == 0) - dsc_sink_caps->branch_overall_throughput_1_mps = 0; - else if (dsc_sink_caps->branch_overall_throughput_1_mps == 1) - dsc_sink_caps->branch_overall_throughput_1_mps = 680; - else { - dsc_sink_caps->branch_overall_throughput_1_mps *= 50; - dsc_sink_caps->branch_overall_throughput_1_mps += 600; - } - - dsc_sink_caps->branch_max_line_width = dpcd_dsc_branch_decoder_caps[DP_DSC_BRANCH_MAX_LINE_WIDTH - DP_DSC_BRANCH_OVERALL_THROUGHPUT_0] * 320; - ASSERT(dsc_sink_caps->branch_max_line_width == 0 || dsc_sink_caps->branch_max_line_width >= 5120); - - dsc_sink_caps->is_dp = true; - return true; -} - - -/* If DSC is possbile, get DSC bandwidth range based on [min_bpp, max_bpp] target bitrate range and - * timing's pixel clock and uncompressed bandwidth. - * If DSC is not possible, leave '*range' untouched. - */ -bool dc_dsc_compute_bandwidth_range( - const struct display_stream_compressor *dsc, - uint32_t dsc_min_slice_height_override, - uint32_t min_bpp_x16, - uint32_t max_bpp_x16, - const struct dsc_dec_dpcd_caps *dsc_sink_caps, - const struct dc_crtc_timing *timing, - struct dc_dsc_bw_range *range) -{ - bool is_dsc_possible = false; - struct dsc_enc_caps dsc_enc_caps; - struct dsc_enc_caps dsc_common_caps; - struct dc_dsc_config config; - - get_dsc_enc_caps(dsc, &dsc_enc_caps, timing->pix_clk_100hz); - - is_dsc_possible = intersect_dsc_caps(dsc_sink_caps, &dsc_enc_caps, - timing->pixel_encoding, &dsc_common_caps); - - if (is_dsc_possible) - is_dsc_possible = setup_dsc_config(dsc_sink_caps, &dsc_enc_caps, 0, timing, - dsc_min_slice_height_override, max_bpp_x16, &config); - - if (is_dsc_possible) - get_dsc_bandwidth_range(min_bpp_x16, max_bpp_x16, - config.num_slices_h, &dsc_common_caps, timing, range); - - return is_dsc_possible; -} - bool dc_dsc_compute_config( const struct display_stream_compressor *dsc, const struct dsc_dec_dpcd_caps *dsc_sink_caps, @@ -923,22 +984,22 @@ bool dc_dsc_compute_config( get_dsc_enc_caps(dsc, &dsc_enc_caps, timing->pix_clk_100hz); is_dsc_possible = setup_dsc_config(dsc_sink_caps, - &dsc_enc_caps, - target_bandwidth_kbps, - timing, dsc_min_slice_height_override, - max_target_bpp_limit_override * 16, dsc_cfg); + &dsc_enc_caps, + target_bandwidth_kbps, + timing, dsc_min_slice_height_override, + max_target_bpp_limit_override * 16, dsc_cfg); return is_dsc_possible; } uint32_t dc_dsc_stream_bandwidth_in_kbps(const struct dc_crtc_timing *timing, - uint32_t bpp_x16, uint32_t num_slices_h, bool is_dp) + uint32_t bpp_x16, uint32_t num_slices_h, bool is_dp) { struct fixed31_32 overhead_in_kbps; struct fixed31_32 bpp; struct fixed31_32 actual_bandwidth_in_kbps; overhead_in_kbps = compute_dsc_max_bandwidth_overhead( - timing, num_slices_h, is_dp); + timing, num_slices_h, is_dp); bpp = dc_fixpt_from_fraction(bpp_x16, 16); actual_bandwidth_in_kbps = dc_fixpt_from_fraction(timing->pix_clk_100hz, 10); actual_bandwidth_in_kbps = dc_fixpt_mul(actual_bandwidth_in_kbps, bpp); @@ -946,7 +1007,9 @@ uint32_t dc_dsc_stream_bandwidth_in_kbps(const struct dc_crtc_timing *timing, return dc_fixpt_ceil(actual_bandwidth_in_kbps); } -void dc_dsc_get_policy_for_timing(const struct dc_crtc_timing *timing, uint32_t max_target_bpp_limit_override_x16, struct dc_dsc_policy *policy) +void dc_dsc_get_policy_for_timing(const struct dc_crtc_timing *timing, + uint32_t max_target_bpp_limit_override_x16, + struct dc_dsc_policy *policy) { uint32_t bpc = 0; diff --git a/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h b/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h index e2b58ec9912d..01c3a31be191 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h +++ b/drivers/gpu/drm/amd/display/dc/inc/dc_link_dp.h @@ -174,7 +174,6 @@ bool dp_set_dsc_on_rx(struct pipe_ctx *pipe_ctx, bool enable); void dp_decide_training_settings( struct dc_link *link, const struct dc_link_settings *link_setting, - const struct dc_link_training_overrides *overrides, struct link_training_settings *lt_settings); /* Convert PHY repeater count read from DPCD uint8_t. */ diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr_internal.h b/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr_internal.h index 316301fc1e30..a262f3278c21 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr_internal.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr_internal.h @@ -318,6 +318,11 @@ static inline bool should_update_pstate_support(bool safe_to_lower, bool calc_su return false; } +static inline int khz_to_mhz_ceil(int khz) +{ + return (khz + 999) / 1000; +} + int clk_mgr_helper_get_active_display_cnt( struct dc *dc, struct dc_state *context); diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h index 0638b337f143..713f5558f5e1 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h +++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dchubbub.h @@ -102,6 +102,15 @@ struct hubbub_addr_config { } default_addrs; }; +struct dcn_hubbub_state { + uint32_t vm_fault_addr_msb; + uint32_t vm_fault_addr_lsb; + uint32_t vm_error_status; + uint32_t vm_error_vmid; + uint32_t vm_error_pipe; + uint32_t vm_error_mode; +}; + struct hubbub_funcs { void (*update_dchub)( struct hubbub *hubbub, @@ -149,6 +158,8 @@ struct hubbub_funcs { void (*force_wm_propagate_to_pipes)(struct hubbub *hubbub); + void (*hubbub_read_state)(struct hubbub *hubbub, struct dcn_hubbub_state *hubbub_state); + void (*force_pstate_change_control)(struct hubbub *hubbub, bool force, bool allow); void (*init_watermarks)(struct hubbub *hubbub); diff --git a/drivers/gpu/drm/amd/display/dc/inc/link_dpcd.h b/drivers/gpu/drm/amd/display/dc/inc/link_dpcd.h index d4d52ef1b165..3f12b1600d2a 100644 --- a/drivers/gpu/drm/amd/display/dc/inc/link_dpcd.h +++ b/drivers/gpu/drm/amd/display/dc/inc/link_dpcd.h @@ -1,3 +1,28 @@ +/* + * Copyright 2021 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + * Authors: AMD + * + */ + #ifndef __LINK_DPCD_H__ #define __LINK_DPCD_H__ #include <inc/core_status.h> diff --git a/drivers/gpu/drm/amd/display/dc/irq_types.h b/drivers/gpu/drm/amd/display/dc/irq_types.h index 1139b9eb9f6f..530c2578db40 100644 --- a/drivers/gpu/drm/amd/display/dc/irq_types.h +++ b/drivers/gpu/drm/amd/display/dc/irq_types.h @@ -152,7 +152,7 @@ enum dc_irq_source { DC_IRQ_SOURCE_DC6_VLINE1, DC_IRQ_SOURCE_DMCUB_OUTBOX, DC_IRQ_SOURCE_DMCUB_OUTBOX0, - + DC_IRQ_SOURCE_DMCUB_GENERAL_DATAOUT, DAL_IRQ_SOURCES_NUMBER }; diff --git a/drivers/gpu/drm/amd/display/dmub/dmub_srv.h b/drivers/gpu/drm/amd/display/dmub/dmub_srv.h index abbf7ae584c9..caf961bb633f 100644 --- a/drivers/gpu/drm/amd/display/dmub/dmub_srv.h +++ b/drivers/gpu/drm/amd/display/dmub/dmub_srv.h @@ -352,6 +352,8 @@ struct dmub_srv_hw_funcs { uint32_t (*get_gpint_response)(struct dmub_srv *dmub); + uint32_t (*get_gpint_dataout)(struct dmub_srv *dmub); + void (*send_inbox0_cmd)(struct dmub_srv *dmub, union dmub_inbox0_data_register data); uint32_t (*get_current_time)(struct dmub_srv *dmub); @@ -677,6 +679,22 @@ enum dmub_status dmub_srv_get_gpint_response(struct dmub_srv *dmub, uint32_t *response); /** + * dmub_srv_get_gpint_dataout() - Queries the GPINT DATAOUT. + * @dmub: the dmub service + * @dataout: the data for the GPINT DATAOUT + * + * Returns the response code for the last GPINT DATAOUT interrupt. + * + * Can be called after software initialization. + * + * Return: + * DMUB_STATUS_OK - success + * DMUB_STATUS_INVALID - unspecified error + */ +enum dmub_status dmub_srv_get_gpint_dataout(struct dmub_srv *dmub, + uint32_t *dataout); + +/** * dmub_flush_buffer_mem() - Read back entire frame buffer region. * This ensures that the write from x86 has been flushed and will not * hang the DMCUB. diff --git a/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h index 7c4734f905d9..02921ad22310 100644 --- a/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h +++ b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h @@ -23,8 +23,8 @@ * */ -#ifndef _DMUB_CMD_H_ -#define _DMUB_CMD_H_ +#ifndef DMUB_CMD_H +#define DMUB_CMD_H #if defined(_TEST_HARNESS) || defined(FPGA_USB4) #include "dmub_fw_types.h" @@ -47,10 +47,10 @@ /* Firmware versioning. */ #ifdef DMUB_EXPOSE_VERSION -#define DMUB_FW_VERSION_GIT_HASH 0xf3da2b656 +#define DMUB_FW_VERSION_GIT_HASH 0xe599e0896 #define DMUB_FW_VERSION_MAJOR 0 #define DMUB_FW_VERSION_MINOR 0 -#define DMUB_FW_VERSION_REVISION 71 +#define DMUB_FW_VERSION_REVISION 76 #define DMUB_FW_VERSION_TEST 0 #define DMUB_FW_VERSION_VBIOS 0 #define DMUB_FW_VERSION_HOTFIX 0 @@ -334,6 +334,7 @@ enum dmub_fw_boot_status_bit { DMUB_FW_BOOT_STATUS_BIT_MAILBOX_READY = (1 << 1), /**< 1 if mailbox ready */ DMUB_FW_BOOT_STATUS_BIT_OPTIMIZED_INIT_DONE = (1 << 2), /**< 1 if init done */ DMUB_FW_BOOT_STATUS_BIT_RESTORE_REQUIRED = (1 << 3), /**< 1 if driver should call restore */ + DMUB_FW_BOOT_STATUS_BIT_DEFERRED_LOADED = (1 << 4), /**< 1 if VBIOS data is deferred programmed */ }; /* Register bit definition for SCRATCH5 */ @@ -352,7 +353,7 @@ enum dmub_lvtma_status_bit { }; /** - * union dmub_fw_boot_options - Boot option definitions for SCRATCH15 + * union dmub_fw_boot_options - Boot option definitions for SCRATCH14 */ union dmub_fw_boot_options { struct { @@ -363,7 +364,10 @@ union dmub_fw_boot_options { uint32_t disable_clk_gate: 1; /**< 1 if clock gating should be disabled */ uint32_t skip_phy_init_panel_sequence: 1; /**< 1 to skip panel init seq */ uint32_t z10_disable: 1; /**< 1 to disable z10 */ - uint32_t reserved : 25; /**< reserved */ + uint32_t reserved2: 1; /**< reserved for an unreleased feature */ + uint32_t reserved_unreleased1: 1; /**< reserved for an unreleased feature */ + uint32_t invalid_vbios_data: 1; /**< 1 if VBIOS data table is invalid */ + uint32_t reserved : 23; /**< reserved */ } bits; /**< boot bits */ uint32_t all; /**< 32-bit access to bits */ }; @@ -1406,6 +1410,10 @@ struct dmub_cmd_psr_copy_settings_data { * Currently the support is only for 0 or 1 */ uint8_t panel_inst; + /** + * Explicit padding to 4 byte boundary. + */ + uint8_t pad3[4]; }; /** @@ -2462,16 +2470,14 @@ static inline bool dmub_rb_full(struct dmub_rb *rb) static inline bool dmub_rb_push_front(struct dmub_rb *rb, const union dmub_rb_cmd *cmd) { - uint64_t volatile *dst = (uint64_t volatile *)(rb->base_address) + rb->wrpt / sizeof(uint64_t); - const uint64_t *src = (const uint64_t *)cmd; - uint8_t i; + uint8_t *dst = (uint8_t *)(rb->base_address) + rb->wrpt; + const uint8_t *src = (const uint8_t *)cmd; if (dmub_rb_full(rb)) return false; // copying data - for (i = 0; i < DMUB_RB_CMD_SIZE / sizeof(uint64_t); i++) - *dst++ = *src++; + dmub_memcpy(dst, src, DMUB_RB_CMD_SIZE); rb->wrpt += DMUB_RB_CMD_SIZE; @@ -2493,7 +2499,7 @@ static inline bool dmub_rb_out_push_front(struct dmub_rb *rb, const union dmub_rb_out_cmd *cmd) { uint8_t *dst = (uint8_t *)(rb->base_address) + rb->wrpt; - const uint8_t *src = (uint8_t *)cmd; + const uint8_t *src = (const uint8_t *)cmd; if (dmub_rb_full(rb)) return false; @@ -2578,18 +2584,16 @@ static inline bool dmub_rb_peek_offset(struct dmub_rb *rb, * @return false otherwise */ static inline bool dmub_rb_out_front(struct dmub_rb *rb, - union dmub_rb_out_cmd *cmd) + union dmub_rb_out_cmd *cmd) { - const uint64_t volatile *src = (const uint64_t volatile *)(rb->base_address) + rb->rptr / sizeof(uint64_t); - uint64_t *dst = (uint64_t *)cmd; - uint8_t i; + const uint8_t *src = (const uint8_t *)(rb->base_address) + rb->rptr; + uint8_t *dst = (uint8_t *)cmd; if (dmub_rb_empty(rb)) return false; // copying data - for (i = 0; i < DMUB_RB_CMD_SIZE / sizeof(uint64_t); i++) - *dst++ = *src++; + dmub_memcpy(dst, src, DMUB_RB_CMD_SIZE); return true; } @@ -2624,15 +2628,14 @@ static inline bool dmub_rb_pop_front(struct dmub_rb *rb) */ static inline void dmub_rb_flush_pending(const struct dmub_rb *rb) { + uint8_t buf[DMUB_RB_CMD_SIZE]; uint32_t rptr = rb->rptr; uint32_t wptr = rb->wrpt; while (rptr != wptr) { - uint64_t volatile *data = (uint64_t volatile *)rb->base_address + rptr / sizeof(uint64_t); - uint8_t i; + const uint8_t *data = (const uint8_t *)rb->base_address + rptr; - for (i = 0; i < DMUB_RB_CMD_SIZE / sizeof(uint64_t); i++) - *data++; + dmub_memcpy(buf, data, DMUB_RB_CMD_SIZE); rptr += DMUB_RB_CMD_SIZE; if (rptr >= rb->capacity) diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.c index 973de346410d..61446170056e 100644 --- a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.c +++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.c @@ -38,7 +38,10 @@ const struct dmub_srv_dcn31_regs dmub_srv_dcn31_regs = { #define DMUB_SR(reg) REG_OFFSET_EXP(reg), - { DMUB_DCN31_REGS() }, + { + DMUB_DCN31_REGS() + DMCUB_INTERNAL_REGS() + }, #undef DMUB_SR #define DMUB_SF(reg, field) FD_MASK(reg, field), @@ -305,6 +308,21 @@ uint32_t dmub_dcn31_get_gpint_response(struct dmub_srv *dmub) return REG_READ(DMCUB_SCRATCH7); } +uint32_t dmub_dcn31_get_gpint_dataout(struct dmub_srv *dmub) +{ + uint32_t dataout = REG_READ(DMCUB_GPINT_DATAOUT); + + REG_UPDATE(DMCUB_INTERRUPT_ENABLE, DMCUB_GPINT_IH_INT_EN, 0); + + REG_WRITE(DMCUB_GPINT_DATAOUT, 0); + REG_UPDATE(DMCUB_INTERRUPT_ACK, DMCUB_GPINT_IH_INT_ACK, 1); + REG_UPDATE(DMCUB_INTERRUPT_ACK, DMCUB_GPINT_IH_INT_ACK, 0); + + REG_UPDATE(DMCUB_INTERRUPT_ENABLE, DMCUB_GPINT_IH_INT_EN, 1); + + return dataout; +} + union dmub_fw_boot_status dmub_dcn31_get_fw_boot_status(struct dmub_srv *dmub) { union dmub_fw_boot_status status; diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.h b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.h index 9456a6a2d518..bb62605d2ac8 100644 --- a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.h +++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.h @@ -114,7 +114,9 @@ struct dmub_srv; DMUB_SR(DMCUB_TIMER_CURRENT) \ DMUB_SR(DMCUB_INST_FETCH_FAULT_ADDR) \ DMUB_SR(DMCUB_UNDEFINED_ADDRESS_FAULT_ADDR) \ - DMUB_SR(DMCUB_DATA_WRITE_FAULT_ADDR) + DMUB_SR(DMCUB_DATA_WRITE_FAULT_ADDR) \ + DMUB_SR(DMCUB_INTERRUPT_ENABLE) \ + DMUB_SR(DMCUB_INTERRUPT_ACK) #define DMUB_DCN31_FIELDS() \ DMUB_SF(DMCUB_CNTL, DMCUB_ENABLE) \ @@ -147,7 +149,9 @@ struct dmub_srv; DMUB_SF(MMHUBBUB_SOFT_RESET, DMUIF_SOFT_RESET) \ DMUB_SF(DCN_VM_FB_LOCATION_BASE, FB_BASE) \ DMUB_SF(DCN_VM_FB_OFFSET, FB_OFFSET) \ - DMUB_SF(DMCUB_INBOX0_WPTR, DMCUB_INBOX0_WPTR) + DMUB_SF(DMCUB_INBOX0_WPTR, DMCUB_INBOX0_WPTR) \ + DMUB_SF(DMCUB_INTERRUPT_ENABLE, DMCUB_GPINT_IH_INT_EN) \ + DMUB_SF(DMCUB_INTERRUPT_ACK, DMCUB_GPINT_IH_INT_ACK) struct dmub_srv_dcn31_reg_offset { #define DMUB_SR(reg) uint32_t reg; @@ -222,6 +226,8 @@ bool dmub_dcn31_is_gpint_acked(struct dmub_srv *dmub, uint32_t dmub_dcn31_get_gpint_response(struct dmub_srv *dmub); +uint32_t dmub_dcn31_get_gpint_dataout(struct dmub_srv *dmub); + void dmub_dcn31_enable_dmub_boot_options(struct dmub_srv *dmub, const struct dmub_srv_hw_params *params); void dmub_dcn31_skip_dmub_panel_power_sequence(struct dmub_srv *dmub, bool skip); diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c index 2bdbd7406f56..75a91cfaf036 100644 --- a/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c +++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c @@ -224,6 +224,7 @@ static bool dmub_srv_hw_setup(struct dmub_srv *dmub, enum dmub_asic asic) funcs->set_gpint = dmub_dcn31_set_gpint; funcs->is_gpint_acked = dmub_dcn31_is_gpint_acked; funcs->get_gpint_response = dmub_dcn31_get_gpint_response; + funcs->get_gpint_dataout = dmub_dcn31_get_gpint_dataout; funcs->get_fw_status = dmub_dcn31_get_fw_boot_status; funcs->enable_dmub_boot_options = dmub_dcn31_enable_dmub_boot_options; funcs->skip_dmub_panel_power_sequence = dmub_dcn31_skip_dmub_panel_power_sequence; @@ -719,6 +720,22 @@ enum dmub_status dmub_srv_get_gpint_response(struct dmub_srv *dmub, return DMUB_STATUS_OK; } +enum dmub_status dmub_srv_get_gpint_dataout(struct dmub_srv *dmub, + uint32_t *dataout) +{ + *dataout = 0; + + if (!dmub->sw_init) + return DMUB_STATUS_INVALID; + + if (!dmub->hw_funcs.get_gpint_dataout) + return DMUB_STATUS_INVALID; + + *dataout = dmub->hw_funcs.get_gpint_dataout(dmub); + + return DMUB_STATUS_OK; +} + enum dmub_status dmub_srv_get_fw_boot_status(struct dmub_srv *dmub, union dmub_fw_boot_status *status) { diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.c b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.c index b963226e8af4..06d60f031a06 100644 --- a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.c +++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.c @@ -39,8 +39,12 @@ static void push_error_status(struct mod_hdcp *hdcp, if (is_hdcp1(hdcp)) { hdcp->connection.hdcp1_retry_count++; + if (hdcp->connection.hdcp1_retry_count == MAX_NUM_OF_ATTEMPTS) + hdcp->connection.link.adjust.hdcp1.disable = 1; } else if (is_hdcp2(hdcp)) { hdcp->connection.hdcp2_retry_count++; + if (hdcp->connection.hdcp2_retry_count == MAX_NUM_OF_ATTEMPTS) + hdcp->connection.link.adjust.hdcp2.disable = 1; } } @@ -59,8 +63,7 @@ static uint8_t is_cp_desired_hdcp1(struct mod_hdcp *hdcp) } } - return (hdcp->connection.hdcp1_retry_count < MAX_NUM_OF_ATTEMPTS) && - is_auth_needed && + return is_auth_needed && !hdcp->connection.link.adjust.hdcp1.disable && !hdcp->connection.is_hdcp1_revoked; } @@ -80,8 +83,7 @@ static uint8_t is_cp_desired_hdcp2(struct mod_hdcp *hdcp) } } - return (hdcp->connection.hdcp2_retry_count < MAX_NUM_OF_ATTEMPTS) && - is_auth_needed && + return is_auth_needed && !hdcp->connection.link.adjust.hdcp2.disable && !hdcp->connection.is_hdcp2_revoked; } @@ -313,9 +315,6 @@ enum mod_hdcp_status mod_hdcp_add_display(struct mod_hdcp *hdcp, goto out; } - /* save current encryption states to restore after next authentication */ - mod_hdcp_save_current_encryption_states(hdcp); - /* reset existing authentication status */ status = reset_authentication(hdcp, output); if (status != MOD_HDCP_STATUS_SUCCESS) @@ -362,9 +361,6 @@ enum mod_hdcp_status mod_hdcp_remove_display(struct mod_hdcp *hdcp, goto out; } - /* save current encryption states to restore after next authentication */ - mod_hdcp_save_current_encryption_states(hdcp); - /* stop current authentication */ status = reset_authentication(hdcp, output); if (status != MOD_HDCP_STATUS_SUCCESS) @@ -392,6 +388,60 @@ out: return status; } +enum mod_hdcp_status mod_hdcp_update_authentication(struct mod_hdcp *hdcp, + uint8_t index, + struct mod_hdcp_link_adjustment *link_adjust, + struct mod_hdcp_display_adjustment *display_adjust, + struct mod_hdcp_output *output) +{ + enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS; + struct mod_hdcp_display *display = NULL; + + HDCP_TOP_INTERFACE_TRACE_WITH_INDEX(hdcp, index); + memset(output, 0, sizeof(struct mod_hdcp_output)); + + /* find display in connection */ + display = get_active_display_at_index(hdcp, index); + if (!display) { + status = MOD_HDCP_STATUS_DISPLAY_NOT_FOUND; + goto out; + } + + /* skip if no changes */ + if (memcmp(link_adjust, &hdcp->connection.link.adjust, + sizeof(struct mod_hdcp_link_adjustment)) == 0 && + memcmp(display_adjust, &display->adjust, + sizeof(struct mod_hdcp_display_adjustment)) == 0) { + status = MOD_HDCP_STATUS_SUCCESS; + goto out; + } + + /* stop current authentication */ + status = reset_authentication(hdcp, output); + if (status != MOD_HDCP_STATUS_SUCCESS) + goto out; + + /* clear retry counters */ + reset_retry_counts(hdcp); + + /* reset error trace */ + memset(&hdcp->connection.trace, 0, sizeof(hdcp->connection.trace)); + + /* set new adjustment */ + hdcp->connection.link.adjust = *link_adjust; + display->adjust = *display_adjust; + + /* request authentication when connection is not reset */ + if (current_state(hdcp) != HDCP_UNINITIALIZED) + /* wait 100ms to debounce simultaneous updates for different indices */ + callback_in_ms(100, output); + +out: + if (status != MOD_HDCP_STATUS_SUCCESS) + push_error_status(hdcp, status); + return status; +} + enum mod_hdcp_status mod_hdcp_query_display(struct mod_hdcp *hdcp, uint8_t index, struct mod_hdcp_display_query *query) { diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h index 3ce91db560d1..7123f0915706 100644 --- a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h +++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h @@ -331,8 +331,6 @@ enum mod_hdcp_status mod_hdcp_add_display_to_topology( struct mod_hdcp *hdcp, struct mod_hdcp_display *display); enum mod_hdcp_status mod_hdcp_remove_display_from_topology( struct mod_hdcp *hdcp, uint8_t index); -bool mod_hdcp_is_link_encryption_enabled(struct mod_hdcp *hdcp); -void mod_hdcp_save_current_encryption_states(struct mod_hdcp *hdcp); enum mod_hdcp_status mod_hdcp_hdcp1_create_session(struct mod_hdcp *hdcp); enum mod_hdcp_status mod_hdcp_hdcp1_destroy_session(struct mod_hdcp *hdcp); enum mod_hdcp_status mod_hdcp_hdcp1_validate_rx(struct mod_hdcp *hdcp); diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp1_execution.c b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp1_execution.c index de872e7958b0..6ec918af3bff 100644 --- a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp1_execution.c +++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp1_execution.c @@ -266,9 +266,6 @@ static enum mod_hdcp_status authenticated(struct mod_hdcp *hdcp, mod_hdcp_execute_and_set(mod_hdcp_hdcp1_link_maintenance, &input->link_maintenance, &status, hdcp, "link_maintenance"); - - if (status != MOD_HDCP_STATUS_SUCCESS) - mod_hdcp_save_current_encryption_states(hdcp); out: return status; } @@ -447,9 +444,6 @@ static enum mod_hdcp_status authenticated_dp(struct mod_hdcp *hdcp, mod_hdcp_execute_and_set(check_no_reauthentication_request_dp, &input->reauth_request_check, &status, hdcp, "reauth_request_check"); - - if (status != MOD_HDCP_STATUS_SUCCESS) - mod_hdcp_save_current_encryption_states(hdcp); out: return status; } diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_execution.c b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_execution.c index 117c6b45f718..91c22b96ebde 100644 --- a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_execution.c +++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_execution.c @@ -571,9 +571,6 @@ static enum mod_hdcp_status authenticated(struct mod_hdcp *hdcp, } process_rxstatus(hdcp, event_ctx, input, &status); - - if (status != MOD_HDCP_STATUS_SUCCESS) - mod_hdcp_save_current_encryption_states(hdcp); out: return status; } diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_psp.c b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_psp.c index 1b02056bc8bd..ade86a042398 100644 --- a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_psp.c +++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_psp.c @@ -45,7 +45,7 @@ static void hdcp2_message_init(struct mod_hdcp *hdcp, in->process.msg3_desc.msg_size = 0; } -static enum mod_hdcp_status mod_hdcp_remove_display_from_topology_v2( +static enum mod_hdcp_status remove_display_from_topology_v2( struct mod_hdcp *hdcp, uint8_t index) { struct psp_context *psp = hdcp->config.psp.handle; @@ -81,7 +81,7 @@ static enum mod_hdcp_status mod_hdcp_remove_display_from_topology_v2( return status; } -static enum mod_hdcp_status mod_hdcp_remove_display_from_topology_v3( +static enum mod_hdcp_status remove_display_from_topology_v3( struct mod_hdcp *hdcp, uint8_t index) { struct psp_context *psp = hdcp->config.psp.handle; @@ -107,7 +107,7 @@ static enum mod_hdcp_status mod_hdcp_remove_display_from_topology_v3( psp_dtm_invoke(psp, dtm_cmd->cmd_id); if (dtm_cmd->dtm_status != TA_DTM_STATUS__SUCCESS) { - status = mod_hdcp_remove_display_from_topology_v2(hdcp, index); + status = remove_display_from_topology_v2(hdcp, index); if (status != MOD_HDCP_STATUS_SUCCESS) display->state = MOD_HDCP_DISPLAY_INACTIVE; } else { @@ -120,20 +120,7 @@ static enum mod_hdcp_status mod_hdcp_remove_display_from_topology_v3( return status; } -enum mod_hdcp_status mod_hdcp_remove_display_from_topology( - struct mod_hdcp *hdcp, uint8_t index) -{ - enum mod_hdcp_status status = MOD_HDCP_STATUS_UPDATE_TOPOLOGY_FAILURE; - - if (hdcp->config.psp.caps.dtm_v3_supported) - status = mod_hdcp_remove_display_from_topology_v3(hdcp, index); - else - status = mod_hdcp_remove_display_from_topology_v2(hdcp, index); - - return status; -} - -static enum mod_hdcp_status mod_hdcp_add_display_to_topology_v2( +static enum mod_hdcp_status add_display_to_topology_v2( struct mod_hdcp *hdcp, struct mod_hdcp_display *display) { struct psp_context *psp = hdcp->config.psp.handle; @@ -180,7 +167,7 @@ static enum mod_hdcp_status mod_hdcp_add_display_to_topology_v2( return status; } -static enum mod_hdcp_status mod_hdcp_add_display_to_topology_v3( +static enum mod_hdcp_status add_display_to_topology_v3( struct mod_hdcp *hdcp, struct mod_hdcp_display *display) { struct psp_context *psp = hdcp->config.psp.handle; @@ -220,7 +207,7 @@ static enum mod_hdcp_status mod_hdcp_add_display_to_topology_v3( psp_dtm_invoke(psp, dtm_cmd->cmd_id); if (dtm_cmd->dtm_status != TA_DTM_STATUS__SUCCESS) { - status = mod_hdcp_add_display_to_topology_v2(hdcp, display); + status = add_display_to_topology_v2(hdcp, display); if (status != MOD_HDCP_STATUS_SUCCESS) display->state = MOD_HDCP_DISPLAY_INACTIVE; } else { @@ -232,15 +219,28 @@ static enum mod_hdcp_status mod_hdcp_add_display_to_topology_v3( return status; } +enum mod_hdcp_status mod_hdcp_remove_display_from_topology( + struct mod_hdcp *hdcp, uint8_t index) +{ + enum mod_hdcp_status status = MOD_HDCP_STATUS_UPDATE_TOPOLOGY_FAILURE; + + if (hdcp->config.psp.caps.dtm_v3_supported) + status = remove_display_from_topology_v3(hdcp, index); + else + status = remove_display_from_topology_v2(hdcp, index); + + return status; +} + enum mod_hdcp_status mod_hdcp_add_display_to_topology(struct mod_hdcp *hdcp, struct mod_hdcp_display *display) { enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS; if (hdcp->config.psp.caps.dtm_v3_supported) - status = mod_hdcp_add_display_to_topology_v3(hdcp, display); + status = add_display_to_topology_v3(hdcp, display); else - status = mod_hdcp_add_display_to_topology_v2(hdcp, display); + status = add_display_to_topology_v2(hdcp, display); return status; } @@ -1021,14 +1021,3 @@ enum mod_hdcp_status mod_hdcp_hdcp2_validate_stream_ready(struct mod_hdcp *hdcp) mutex_unlock(&psp->hdcp_context.mutex); return status; } - -bool mod_hdcp_is_link_encryption_enabled(struct mod_hdcp *hdcp) -{ - /* unsupported */ - return true; -} - -void mod_hdcp_save_current_encryption_states(struct mod_hdcp *hdcp) -{ - /* unsupported */ -} diff --git a/drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h b/drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h index c590493fd293..a6eb86de8d5c 100644 --- a/drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h +++ b/drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h @@ -123,7 +123,6 @@ enum mod_hdcp_display_state { struct mod_hdcp_psp_caps { uint8_t dtm_v3_supported; - uint8_t opm_state_query_supported; }; enum mod_hdcp_display_disable_option { @@ -282,15 +281,22 @@ enum mod_hdcp_status mod_hdcp_setup(struct mod_hdcp *hdcp, /* called per link on link destroy */ enum mod_hdcp_status mod_hdcp_teardown(struct mod_hdcp *hdcp); -/* called per display on cp_desired set to true */ +/* called per display after stream is enabled */ enum mod_hdcp_status mod_hdcp_add_display(struct mod_hdcp *hdcp, struct mod_hdcp_link *link, struct mod_hdcp_display *display, struct mod_hdcp_output *output); -/* called per display on cp_desired set to false */ +/* called per display before stream is disabled */ enum mod_hdcp_status mod_hdcp_remove_display(struct mod_hdcp *hdcp, uint8_t index, struct mod_hdcp_output *output); +/* called per display to apply new authentication adjustment */ +enum mod_hdcp_status mod_hdcp_update_authentication(struct mod_hdcp *hdcp, + uint8_t index, + struct mod_hdcp_link_adjustment *link_adjust, + struct mod_hdcp_display_adjustment *display_adjust, + struct mod_hdcp_output *output); + /* called to query hdcp information on a specific index */ enum mod_hdcp_status mod_hdcp_query_display(struct mod_hdcp *hdcp, uint8_t index, struct mod_hdcp_display_query *query); diff --git a/drivers/gpu/drm/amd/include/amd_shared.h b/drivers/gpu/drm/amd/include/amd_shared.h index ff1d3d4a6488..257f280d3d53 100644 --- a/drivers/gpu/drm/amd/include/amd_shared.h +++ b/drivers/gpu/drm/amd/include/amd_shared.h @@ -47,6 +47,7 @@ enum amd_apu_flags { AMD_APU_IS_RENOIR = 0x00000008UL, AMD_APU_IS_GREEN_SARDINE = 0x00000010UL, AMD_APU_IS_VANGOGH = 0x00000020UL, + AMD_APU_IS_CYAN_SKILLFISH2 = 0x00000040UL, }; /** diff --git a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_1_2_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_1_2_sh_mask.h index e5fd0121ceff..a9d553ef26c0 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_1_2_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/dcn/dcn_3_1_2_sh_mask.h @@ -5971,6 +5971,7 @@ #define DMCUB_INTERRUPT_ENABLE__DMCUB_GPINT1_INT_EN__SHIFT 0xb #define DMCUB_INTERRUPT_ENABLE__DMCUB_GPINT2_INT_EN__SHIFT 0xc #define DMCUB_INTERRUPT_ENABLE__DMCUB_UNDEFINED_ADDRESS_FAULT_INT_EN__SHIFT 0xd +#define DMCUB_INTERRUPT_ENABLE__DMCUB_GPINT_IH_INT_EN__SHIFT 0x11 #define DMCUB_INTERRUPT_ENABLE__DMCUB_TIMER0_INT_EN_MASK 0x00000001L #define DMCUB_INTERRUPT_ENABLE__DMCUB_TIMER1_INT_EN_MASK 0x00000002L #define DMCUB_INTERRUPT_ENABLE__DMCUB_INBOX0_READY_INT_EN_MASK 0x00000004L @@ -5985,6 +5986,7 @@ #define DMCUB_INTERRUPT_ENABLE__DMCUB_GPINT1_INT_EN_MASK 0x00000800L #define DMCUB_INTERRUPT_ENABLE__DMCUB_GPINT2_INT_EN_MASK 0x00001000L #define DMCUB_INTERRUPT_ENABLE__DMCUB_UNDEFINED_ADDRESS_FAULT_INT_EN_MASK 0x00002000L +#define DMCUB_INTERRUPT_ENABLE__DMCUB_GPINT_IH_INT_EN_MASK 0x00020000L //DMCUB_INTERRUPT_ACK #define DMCUB_INTERRUPT_ACK__DMCUB_TIMER0_INT_ACK__SHIFT 0x0 #define DMCUB_INTERRUPT_ACK__DMCUB_TIMER1_INT_ACK__SHIFT 0x1 @@ -6000,6 +6002,7 @@ #define DMCUB_INTERRUPT_ACK__DMCUB_GPINT1_INT_ACK__SHIFT 0xb #define DMCUB_INTERRUPT_ACK__DMCUB_GPINT2_INT_ACK__SHIFT 0xc #define DMCUB_INTERRUPT_ACK__DMCUB_UNDEFINED_ADDRESS_FAULT_ACK__SHIFT 0xd +#define DMCUB_INTERRUPT_ACK__DMCUB_GPINT_IH_INT_ACK__SHIFT 0x11 #define DMCUB_INTERRUPT_ACK__DMCUB_TIMER0_INT_ACK_MASK 0x00000001L #define DMCUB_INTERRUPT_ACK__DMCUB_TIMER1_INT_ACK_MASK 0x00000002L #define DMCUB_INTERRUPT_ACK__DMCUB_INBOX0_READY_INT_ACK_MASK 0x00000004L @@ -6014,6 +6017,7 @@ #define DMCUB_INTERRUPT_ACK__DMCUB_GPINT1_INT_ACK_MASK 0x00000800L #define DMCUB_INTERRUPT_ACK__DMCUB_GPINT2_INT_ACK_MASK 0x00001000L #define DMCUB_INTERRUPT_ACK__DMCUB_UNDEFINED_ADDRESS_FAULT_ACK_MASK 0x00002000L +#define DMCUB_INTERRUPT_ACK__DMCUB_GPINT_IH_INT_ACK_MASK 0x00020000L //DMCUB_INTERRUPT_STATUS #define DMCUB_INTERRUPT_STATUS__DMCUB_TIMER0_INT_STAT__SHIFT 0x0 #define DMCUB_INTERRUPT_STATUS__DMCUB_TIMER1_INT_STAT__SHIFT 0x1 diff --git a/drivers/gpu/drm/amd/include/asic_reg/mp/mp_11_0_8_offset.h b/drivers/gpu/drm/amd/include/asic_reg/mp/mp_11_0_8_offset.h new file mode 100644 index 000000000000..19293ccaec23 --- /dev/null +++ b/drivers/gpu/drm/amd/include/asic_reg/mp/mp_11_0_8_offset.h @@ -0,0 +1,352 @@ +/* + * Copyright (C) 2021 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included + * in all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS + * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN + * AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. + */ +#ifndef _mp_11_0_8_OFFSET_HEADER +#define _mp_11_0_8_OFFSET_HEADER + + + +// addressBlock: mp_SmuMp0_SmnDec +// base address: 0x0 +#define mmMP0_SMN_C2PMSG_32 0x0060 +#define mmMP0_SMN_C2PMSG_32_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_33 0x0061 +#define mmMP0_SMN_C2PMSG_33_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_34 0x0062 +#define mmMP0_SMN_C2PMSG_34_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_35 0x0063 +#define mmMP0_SMN_C2PMSG_35_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_36 0x0064 +#define mmMP0_SMN_C2PMSG_36_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_37 0x0065 +#define mmMP0_SMN_C2PMSG_37_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_38 0x0066 +#define mmMP0_SMN_C2PMSG_38_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_39 0x0067 +#define mmMP0_SMN_C2PMSG_39_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_40 0x0068 +#define mmMP0_SMN_C2PMSG_40_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_41 0x0069 +#define mmMP0_SMN_C2PMSG_41_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_42 0x006a +#define mmMP0_SMN_C2PMSG_42_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_43 0x006b +#define mmMP0_SMN_C2PMSG_43_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_44 0x006c +#define mmMP0_SMN_C2PMSG_44_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_45 0x006d +#define mmMP0_SMN_C2PMSG_45_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_46 0x006e +#define mmMP0_SMN_C2PMSG_46_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_47 0x006f +#define mmMP0_SMN_C2PMSG_47_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_48 0x0070 +#define mmMP0_SMN_C2PMSG_48_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_49 0x0071 +#define mmMP0_SMN_C2PMSG_49_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_50 0x0072 +#define mmMP0_SMN_C2PMSG_50_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_51 0x0073 +#define mmMP0_SMN_C2PMSG_51_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_52 0x0074 +#define mmMP0_SMN_C2PMSG_52_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_53 0x0075 +#define mmMP0_SMN_C2PMSG_53_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_54 0x0076 +#define mmMP0_SMN_C2PMSG_54_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_55 0x0077 +#define mmMP0_SMN_C2PMSG_55_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_56 0x0078 +#define mmMP0_SMN_C2PMSG_56_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_57 0x0079 +#define mmMP0_SMN_C2PMSG_57_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_58 0x007a +#define mmMP0_SMN_C2PMSG_58_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_59 0x007b +#define mmMP0_SMN_C2PMSG_59_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_60 0x007c +#define mmMP0_SMN_C2PMSG_60_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_61 0x007d +#define mmMP0_SMN_C2PMSG_61_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_62 0x007e +#define mmMP0_SMN_C2PMSG_62_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_63 0x007f +#define mmMP0_SMN_C2PMSG_63_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_64 0x0080 +#define mmMP0_SMN_C2PMSG_64_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_65 0x0081 +#define mmMP0_SMN_C2PMSG_65_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_66 0x0082 +#define mmMP0_SMN_C2PMSG_66_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_67 0x0083 +#define mmMP0_SMN_C2PMSG_67_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_68 0x0084 +#define mmMP0_SMN_C2PMSG_68_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_69 0x0085 +#define mmMP0_SMN_C2PMSG_69_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_70 0x0086 +#define mmMP0_SMN_C2PMSG_70_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_71 0x0087 +#define mmMP0_SMN_C2PMSG_71_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_72 0x0088 +#define mmMP0_SMN_C2PMSG_72_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_73 0x0089 +#define mmMP0_SMN_C2PMSG_73_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_74 0x008a +#define mmMP0_SMN_C2PMSG_74_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_75 0x008b +#define mmMP0_SMN_C2PMSG_75_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_76 0x008c +#define mmMP0_SMN_C2PMSG_76_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_77 0x008d +#define mmMP0_SMN_C2PMSG_77_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_78 0x008e +#define mmMP0_SMN_C2PMSG_78_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_79 0x008f +#define mmMP0_SMN_C2PMSG_79_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_80 0x0090 +#define mmMP0_SMN_C2PMSG_80_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_81 0x0091 +#define mmMP0_SMN_C2PMSG_81_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_82 0x0092 +#define mmMP0_SMN_C2PMSG_82_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_83 0x0093 +#define mmMP0_SMN_C2PMSG_83_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_84 0x0094 +#define mmMP0_SMN_C2PMSG_84_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_85 0x0095 +#define mmMP0_SMN_C2PMSG_85_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_86 0x0096 +#define mmMP0_SMN_C2PMSG_86_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_87 0x0097 +#define mmMP0_SMN_C2PMSG_87_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_88 0x0098 +#define mmMP0_SMN_C2PMSG_88_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_89 0x0099 +#define mmMP0_SMN_C2PMSG_89_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_90 0x009a +#define mmMP0_SMN_C2PMSG_90_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_91 0x009b +#define mmMP0_SMN_C2PMSG_91_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_92 0x009c +#define mmMP0_SMN_C2PMSG_92_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_93 0x009d +#define mmMP0_SMN_C2PMSG_93_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_94 0x009e +#define mmMP0_SMN_C2PMSG_94_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_95 0x009f +#define mmMP0_SMN_C2PMSG_95_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_96 0x00a0 +#define mmMP0_SMN_C2PMSG_96_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_97 0x00a1 +#define mmMP0_SMN_C2PMSG_97_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_98 0x00a2 +#define mmMP0_SMN_C2PMSG_98_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_99 0x00a3 +#define mmMP0_SMN_C2PMSG_99_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_100 0x00a4 +#define mmMP0_SMN_C2PMSG_100_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_101 0x00a5 +#define mmMP0_SMN_C2PMSG_101_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_102 0x00a6 +#define mmMP0_SMN_C2PMSG_102_BASE_IDX 0 +#define mmMP0_SMN_C2PMSG_103 0x00a7 +#define mmMP0_SMN_C2PMSG_103_BASE_IDX 0 +#define mmMP0_SMN_IH_CREDIT 0x00c1 +#define mmMP0_SMN_IH_CREDIT_BASE_IDX 0 +#define mmMP0_SMN_IH_SW_INT 0x00c2 +#define mmMP0_SMN_IH_SW_INT_BASE_IDX 0 +#define mmMP0_SMN_IH_SW_INT_CTRL 0x00c3 +#define mmMP0_SMN_IH_SW_INT_CTRL_BASE_IDX 0 + + +// addressBlock: mp_SmuMp1_SmnDec +// base address: 0x0 +#define mmMP1_SMN_C2PMSG_32 0x0260 +#define mmMP1_SMN_C2PMSG_32_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_33 0x0261 +#define mmMP1_SMN_C2PMSG_33_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_34 0x0262 +#define mmMP1_SMN_C2PMSG_34_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_35 0x0263 +#define mmMP1_SMN_C2PMSG_35_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_36 0x0264 +#define mmMP1_SMN_C2PMSG_36_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_37 0x0265 +#define mmMP1_SMN_C2PMSG_37_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_38 0x0266 +#define mmMP1_SMN_C2PMSG_38_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_39 0x0267 +#define mmMP1_SMN_C2PMSG_39_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_40 0x0268 +#define mmMP1_SMN_C2PMSG_40_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_41 0x0269 +#define mmMP1_SMN_C2PMSG_41_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_42 0x026a +#define mmMP1_SMN_C2PMSG_42_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_43 0x026b +#define mmMP1_SMN_C2PMSG_43_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_44 0x026c +#define mmMP1_SMN_C2PMSG_44_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_45 0x026d +#define mmMP1_SMN_C2PMSG_45_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_46 0x026e +#define mmMP1_SMN_C2PMSG_46_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_47 0x026f +#define mmMP1_SMN_C2PMSG_47_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_48 0x0270 +#define mmMP1_SMN_C2PMSG_48_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_49 0x0271 +#define mmMP1_SMN_C2PMSG_49_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_50 0x0272 +#define mmMP1_SMN_C2PMSG_50_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_51 0x0273 +#define mmMP1_SMN_C2PMSG_51_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_52 0x0274 +#define mmMP1_SMN_C2PMSG_52_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_53 0x0275 +#define mmMP1_SMN_C2PMSG_53_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_54 0x0276 +#define mmMP1_SMN_C2PMSG_54_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_55 0x0277 +#define mmMP1_SMN_C2PMSG_55_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_56 0x0278 +#define mmMP1_SMN_C2PMSG_56_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_57 0x0279 +#define mmMP1_SMN_C2PMSG_57_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_58 0x027a +#define mmMP1_SMN_C2PMSG_58_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_59 0x027b +#define mmMP1_SMN_C2PMSG_59_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_60 0x027c +#define mmMP1_SMN_C2PMSG_60_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_61 0x027d +#define mmMP1_SMN_C2PMSG_61_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_62 0x027e +#define mmMP1_SMN_C2PMSG_62_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_63 0x027f +#define mmMP1_SMN_C2PMSG_63_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_64 0x0280 +#define mmMP1_SMN_C2PMSG_64_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_65 0x0281 +#define mmMP1_SMN_C2PMSG_65_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_66 0x0282 +#define mmMP1_SMN_C2PMSG_66_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_67 0x0283 +#define mmMP1_SMN_C2PMSG_67_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_68 0x0284 +#define mmMP1_SMN_C2PMSG_68_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_69 0x0285 +#define mmMP1_SMN_C2PMSG_69_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_70 0x0286 +#define mmMP1_SMN_C2PMSG_70_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_71 0x0287 +#define mmMP1_SMN_C2PMSG_71_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_72 0x0288 +#define mmMP1_SMN_C2PMSG_72_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_73 0x0289 +#define mmMP1_SMN_C2PMSG_73_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_74 0x028a +#define mmMP1_SMN_C2PMSG_74_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_75 0x028b +#define mmMP1_SMN_C2PMSG_75_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_76 0x028c +#define mmMP1_SMN_C2PMSG_76_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_77 0x028d +#define mmMP1_SMN_C2PMSG_77_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_78 0x028e +#define mmMP1_SMN_C2PMSG_78_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_79 0x028f +#define mmMP1_SMN_C2PMSG_79_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_80 0x0290 +#define mmMP1_SMN_C2PMSG_80_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_81 0x0291 +#define mmMP1_SMN_C2PMSG_81_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_82 0x0292 +#define mmMP1_SMN_C2PMSG_82_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_83 0x0293 +#define mmMP1_SMN_C2PMSG_83_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_84 0x0294 +#define mmMP1_SMN_C2PMSG_84_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_85 0x0295 +#define mmMP1_SMN_C2PMSG_85_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_86 0x0296 +#define mmMP1_SMN_C2PMSG_86_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_87 0x0297 +#define mmMP1_SMN_C2PMSG_87_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_88 0x0298 +#define mmMP1_SMN_C2PMSG_88_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_89 0x0299 +#define mmMP1_SMN_C2PMSG_89_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_90 0x029a +#define mmMP1_SMN_C2PMSG_90_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_91 0x029b +#define mmMP1_SMN_C2PMSG_91_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_92 0x029c +#define mmMP1_SMN_C2PMSG_92_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_93 0x029d +#define mmMP1_SMN_C2PMSG_93_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_94 0x029e +#define mmMP1_SMN_C2PMSG_94_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_95 0x029f +#define mmMP1_SMN_C2PMSG_95_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_96 0x02a0 +#define mmMP1_SMN_C2PMSG_96_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_97 0x02a1 +#define mmMP1_SMN_C2PMSG_97_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_98 0x02a2 +#define mmMP1_SMN_C2PMSG_98_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_99 0x02a3 +#define mmMP1_SMN_C2PMSG_99_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_100 0x02a4 +#define mmMP1_SMN_C2PMSG_100_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_101 0x02a5 +#define mmMP1_SMN_C2PMSG_101_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_102 0x02a6 +#define mmMP1_SMN_C2PMSG_102_BASE_IDX 0 +#define mmMP1_SMN_C2PMSG_103 0x02a7 +#define mmMP1_SMN_C2PMSG_103_BASE_IDX 0 +#define mmMP1_SMN_IH_CREDIT 0x02c1 +#define mmMP1_SMN_IH_CREDIT_BASE_IDX 0 +#define mmMP1_SMN_IH_SW_INT 0x02c2 +#define mmMP1_SMN_IH_SW_INT_BASE_IDX 0 +#define mmMP1_SMN_IH_SW_INT_CTRL 0x02c3 +#define mmMP1_SMN_IH_SW_INT_CTRL_BASE_IDX 0 +#define mmMP1_SMN_FPS_CNT 0x02c4 +#define mmMP1_SMN_FPS_CNT_BASE_IDX 0 +#define mmMP1_SMN_EXT_SCRATCH0 0x03c0 +#define mmMP1_SMN_EXT_SCRATCH0_BASE_IDX 0 +#define mmMP1_SMN_EXT_SCRATCH1 0x03c1 +#define mmMP1_SMN_EXT_SCRATCH1_BASE_IDX 0 +#define mmMP1_SMN_EXT_SCRATCH2 0x03c2 +#define mmMP1_SMN_EXT_SCRATCH2_BASE_IDX 0 +#define mmMP1_SMN_EXT_SCRATCH3 0x03c3 +#define mmMP1_SMN_EXT_SCRATCH3_BASE_IDX 0 +#define mmMP1_SMN_EXT_SCRATCH4 0x03c4 +#define mmMP1_SMN_EXT_SCRATCH4_BASE_IDX 0 +#define mmMP1_SMN_EXT_SCRATCH5 0x03c5 +#define mmMP1_SMN_EXT_SCRATCH5_BASE_IDX 0 +#define mmMP1_SMN_EXT_SCRATCH6 0x03c6 +#define mmMP1_SMN_EXT_SCRATCH6_BASE_IDX 0 +#define mmMP1_SMN_EXT_SCRATCH7 0x03c7 +#define mmMP1_SMN_EXT_SCRATCH7_BASE_IDX 0 + + +#endif diff --git a/drivers/gpu/drm/amd/include/asic_reg/umc/umc_8_7_0_sh_mask.h b/drivers/gpu/drm/amd/include/asic_reg/umc/umc_8_7_0_sh_mask.h index 4c5097fa0c09..20329da53b0a 100644 --- a/drivers/gpu/drm/amd/include/asic_reg/umc/umc_8_7_0_sh_mask.h +++ b/drivers/gpu/drm/amd/include/asic_reg/umc/umc_8_7_0_sh_mask.h @@ -1,79 +1,100 @@ -#ifndef _umc_8_7_0_SH_MASK_HEADER
-#define _umc_8_7_0_SH_MASK_HEADER
-
-//UMCCH0_0_GeccErrCntSel
-#define UMCCH0_0_GeccErrCntSel__GeccErrCntCsSel__SHIFT 0x0
-#define UMCCH0_0_GeccErrCntSel__GeccErrInt__SHIFT 0xc
-#define UMCCH0_0_GeccErrCntSel__GeccErrCntEn__SHIFT 0xf
-#define UMCCH0_0_GeccErrCntSel__PoisonCntEn__SHIFT 0x10
-#define UMCCH0_0_GeccErrCntSel__GeccErrCntCsSel_MASK 0x0000000FL
-#define UMCCH0_0_GeccErrCntSel__GeccErrInt_MASK 0x00003000L
-#define UMCCH0_0_GeccErrCntSel__GeccErrCntEn_MASK 0x00008000L
-#define UMCCH0_0_GeccErrCntSel__PoisonCntEn_MASK 0x00030000L
-//UMCCH0_0_GeccErrCnt
-#define UMCCH0_0_GeccErrCnt__GeccErrCnt__SHIFT 0x0
-#define UMCCH0_0_GeccErrCnt__GeccUnCorrErrCnt__SHIFT 0x10
-#define UMCCH0_0_GeccErrCnt__GeccErrCnt_MASK 0x0000FFFFL
-#define UMCCH0_0_GeccErrCnt__GeccUnCorrErrCnt_MASK 0xFFFF0000L
-//MCA_UMC_UMC0_MCUMC_STATUST0
-#define MCA_UMC_UMC0_MCUMC_STATUST0__ErrorCode__SHIFT 0x0
-#define MCA_UMC_UMC0_MCUMC_STATUST0__ErrorCodeExt__SHIFT 0x10
-#define MCA_UMC_UMC0_MCUMC_STATUST0__RESERV22__SHIFT 0x16
-#define MCA_UMC_UMC0_MCUMC_STATUST0__AddrLsb__SHIFT 0x18
-#define MCA_UMC_UMC0_MCUMC_STATUST0__RESERV30__SHIFT 0x1e
-#define MCA_UMC_UMC0_MCUMC_STATUST0__ErrCoreId__SHIFT 0x20
-#define MCA_UMC_UMC0_MCUMC_STATUST0__RESERV38__SHIFT 0x26
-#define MCA_UMC_UMC0_MCUMC_STATUST0__Scrub__SHIFT 0x28
-#define MCA_UMC_UMC0_MCUMC_STATUST0__RESERV41__SHIFT 0x29
-#define MCA_UMC_UMC0_MCUMC_STATUST0__Poison__SHIFT 0x2b
-#define MCA_UMC_UMC0_MCUMC_STATUST0__Deferred__SHIFT 0x2c
-#define MCA_UMC_UMC0_MCUMC_STATUST0__UECC__SHIFT 0x2d
-#define MCA_UMC_UMC0_MCUMC_STATUST0__CECC__SHIFT 0x2e
-#define MCA_UMC_UMC0_MCUMC_STATUST0__RESERV47__SHIFT 0x2f
-#define MCA_UMC_UMC0_MCUMC_STATUST0__Transparent__SHIFT 0x34
-#define MCA_UMC_UMC0_MCUMC_STATUST0__SyndV__SHIFT 0x35
-#define MCA_UMC_UMC0_MCUMC_STATUST0__RESERV54__SHIFT 0x36
-#define MCA_UMC_UMC0_MCUMC_STATUST0__TCC__SHIFT 0x37
-#define MCA_UMC_UMC0_MCUMC_STATUST0__ErrCoreIdVal__SHIFT 0x38
-#define MCA_UMC_UMC0_MCUMC_STATUST0__PCC__SHIFT 0x39
-#define MCA_UMC_UMC0_MCUMC_STATUST0__AddrV__SHIFT 0x3a
-#define MCA_UMC_UMC0_MCUMC_STATUST0__MiscV__SHIFT 0x3b
-#define MCA_UMC_UMC0_MCUMC_STATUST0__En__SHIFT 0x3c
-#define MCA_UMC_UMC0_MCUMC_STATUST0__UC__SHIFT 0x3d
-#define MCA_UMC_UMC0_MCUMC_STATUST0__Overflow__SHIFT 0x3e
-#define MCA_UMC_UMC0_MCUMC_STATUST0__Val__SHIFT 0x3f
-#define MCA_UMC_UMC0_MCUMC_STATUST0__ErrorCode_MASK 0x000000000000FFFFL
-#define MCA_UMC_UMC0_MCUMC_STATUST0__ErrorCodeExt_MASK 0x00000000003F0000L
-#define MCA_UMC_UMC0_MCUMC_STATUST0__RESERV22_MASK 0x0000000000C00000L
-#define MCA_UMC_UMC0_MCUMC_STATUST0__AddrLsb_MASK 0x000000003F000000L
-#define MCA_UMC_UMC0_MCUMC_STATUST0__RESERV30_MASK 0x00000000C0000000L
-#define MCA_UMC_UMC0_MCUMC_STATUST0__ErrCoreId_MASK 0x0000003F00000000L
-#define MCA_UMC_UMC0_MCUMC_STATUST0__RESERV38_MASK 0x000000C000000000L
-#define MCA_UMC_UMC0_MCUMC_STATUST0__Scrub_MASK 0x0000010000000000L
-#define MCA_UMC_UMC0_MCUMC_STATUST0__RESERV41_MASK 0x0000060000000000L
-#define MCA_UMC_UMC0_MCUMC_STATUST0__Poison_MASK 0x0000080000000000L
-#define MCA_UMC_UMC0_MCUMC_STATUST0__Deferred_MASK 0x0000100000000000L
-#define MCA_UMC_UMC0_MCUMC_STATUST0__UECC_MASK 0x0000200000000000L
-#define MCA_UMC_UMC0_MCUMC_STATUST0__CECC_MASK 0x0000400000000000L
-#define MCA_UMC_UMC0_MCUMC_STATUST0__RESERV47_MASK 0x000F800000000000L
-#define MCA_UMC_UMC0_MCUMC_STATUST0__Transparent_MASK 0x0010000000000000L
-#define MCA_UMC_UMC0_MCUMC_STATUST0__SyndV_MASK 0x0020000000000000L
-#define MCA_UMC_UMC0_MCUMC_STATUST0__RESERV54_MASK 0x0040000000000000L
-#define MCA_UMC_UMC0_MCUMC_STATUST0__TCC_MASK 0x0080000000000000L
-#define MCA_UMC_UMC0_MCUMC_STATUST0__ErrCoreIdVal_MASK 0x0100000000000000L
-#define MCA_UMC_UMC0_MCUMC_STATUST0__PCC_MASK 0x0200000000000000L
-#define MCA_UMC_UMC0_MCUMC_STATUST0__AddrV_MASK 0x0400000000000000L
-#define MCA_UMC_UMC0_MCUMC_STATUST0__MiscV_MASK 0x0800000000000000L
-#define MCA_UMC_UMC0_MCUMC_STATUST0__En_MASK 0x1000000000000000L
-#define MCA_UMC_UMC0_MCUMC_STATUST0__UC_MASK 0x2000000000000000L
-#define MCA_UMC_UMC0_MCUMC_STATUST0__Overflow_MASK 0x4000000000000000L
-#define MCA_UMC_UMC0_MCUMC_STATUST0__Val_MASK 0x8000000000000000L
-//MCA_UMC_UMC0_MCUMC_ADDRT0
-#define MCA_UMC_UMC0_MCUMC_ADDRT0__ErrorAddr__SHIFT 0x0
-#define MCA_UMC_UMC0_MCUMC_ADDRT0__LSB__SHIFT 0x38
-#define MCA_UMC_UMC0_MCUMC_ADDRT0__Reserved__SHIFT 0x3e
-#define MCA_UMC_UMC0_MCUMC_ADDRT0__ErrorAddr_MASK 0x00FFFFFFFFFFFFFFL
-#define MCA_UMC_UMC0_MCUMC_ADDRT0__LSB_MASK 0x3F00000000000000L
-#define MCA_UMC_UMC0_MCUMC_ADDRT0__Reserved_MASK 0xC000000000000000L
-
-#endif
+/* + * Copyright (C) 2020 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included + * in all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS + * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN + * AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. + */ + +#ifndef _umc_8_7_0_SH_MASK_HEADER +#define _umc_8_7_0_SH_MASK_HEADER + +//UMCCH0_0_GeccErrCntSel +#define UMCCH0_0_GeccErrCntSel__GeccErrCntCsSel__SHIFT 0x0 +#define UMCCH0_0_GeccErrCntSel__GeccErrInt__SHIFT 0xc +#define UMCCH0_0_GeccErrCntSel__GeccErrCntEn__SHIFT 0xf +#define UMCCH0_0_GeccErrCntSel__PoisonCntEn__SHIFT 0x10 +#define UMCCH0_0_GeccErrCntSel__GeccErrCntCsSel_MASK 0x0000000FL +#define UMCCH0_0_GeccErrCntSel__GeccErrInt_MASK 0x00003000L +#define UMCCH0_0_GeccErrCntSel__GeccErrCntEn_MASK 0x00008000L +#define UMCCH0_0_GeccErrCntSel__PoisonCntEn_MASK 0x00030000L +//UMCCH0_0_GeccErrCnt +#define UMCCH0_0_GeccErrCnt__GeccErrCnt__SHIFT 0x0 +#define UMCCH0_0_GeccErrCnt__GeccUnCorrErrCnt__SHIFT 0x10 +#define UMCCH0_0_GeccErrCnt__GeccErrCnt_MASK 0x0000FFFFL +#define UMCCH0_0_GeccErrCnt__GeccUnCorrErrCnt_MASK 0xFFFF0000L +//MCA_UMC_UMC0_MCUMC_STATUST0 +#define MCA_UMC_UMC0_MCUMC_STATUST0__ErrorCode__SHIFT 0x0 +#define MCA_UMC_UMC0_MCUMC_STATUST0__ErrorCodeExt__SHIFT 0x10 +#define MCA_UMC_UMC0_MCUMC_STATUST0__RESERV22__SHIFT 0x16 +#define MCA_UMC_UMC0_MCUMC_STATUST0__AddrLsb__SHIFT 0x18 +#define MCA_UMC_UMC0_MCUMC_STATUST0__RESERV30__SHIFT 0x1e +#define MCA_UMC_UMC0_MCUMC_STATUST0__ErrCoreId__SHIFT 0x20 +#define MCA_UMC_UMC0_MCUMC_STATUST0__RESERV38__SHIFT 0x26 +#define MCA_UMC_UMC0_MCUMC_STATUST0__Scrub__SHIFT 0x28 +#define MCA_UMC_UMC0_MCUMC_STATUST0__RESERV41__SHIFT 0x29 +#define MCA_UMC_UMC0_MCUMC_STATUST0__Poison__SHIFT 0x2b +#define MCA_UMC_UMC0_MCUMC_STATUST0__Deferred__SHIFT 0x2c +#define MCA_UMC_UMC0_MCUMC_STATUST0__UECC__SHIFT 0x2d +#define MCA_UMC_UMC0_MCUMC_STATUST0__CECC__SHIFT 0x2e +#define MCA_UMC_UMC0_MCUMC_STATUST0__RESERV47__SHIFT 0x2f +#define MCA_UMC_UMC0_MCUMC_STATUST0__Transparent__SHIFT 0x34 +#define MCA_UMC_UMC0_MCUMC_STATUST0__SyndV__SHIFT 0x35 +#define MCA_UMC_UMC0_MCUMC_STATUST0__RESERV54__SHIFT 0x36 +#define MCA_UMC_UMC0_MCUMC_STATUST0__TCC__SHIFT 0x37 +#define MCA_UMC_UMC0_MCUMC_STATUST0__ErrCoreIdVal__SHIFT 0x38 +#define MCA_UMC_UMC0_MCUMC_STATUST0__PCC__SHIFT 0x39 +#define MCA_UMC_UMC0_MCUMC_STATUST0__AddrV__SHIFT 0x3a +#define MCA_UMC_UMC0_MCUMC_STATUST0__MiscV__SHIFT 0x3b +#define MCA_UMC_UMC0_MCUMC_STATUST0__En__SHIFT 0x3c +#define MCA_UMC_UMC0_MCUMC_STATUST0__UC__SHIFT 0x3d +#define MCA_UMC_UMC0_MCUMC_STATUST0__Overflow__SHIFT 0x3e +#define MCA_UMC_UMC0_MCUMC_STATUST0__Val__SHIFT 0x3f +#define MCA_UMC_UMC0_MCUMC_STATUST0__ErrorCode_MASK 0x000000000000FFFFL +#define MCA_UMC_UMC0_MCUMC_STATUST0__ErrorCodeExt_MASK 0x00000000003F0000L +#define MCA_UMC_UMC0_MCUMC_STATUST0__RESERV22_MASK 0x0000000000C00000L +#define MCA_UMC_UMC0_MCUMC_STATUST0__AddrLsb_MASK 0x000000003F000000L +#define MCA_UMC_UMC0_MCUMC_STATUST0__RESERV30_MASK 0x00000000C0000000L +#define MCA_UMC_UMC0_MCUMC_STATUST0__ErrCoreId_MASK 0x0000003F00000000L +#define MCA_UMC_UMC0_MCUMC_STATUST0__RESERV38_MASK 0x000000C000000000L +#define MCA_UMC_UMC0_MCUMC_STATUST0__Scrub_MASK 0x0000010000000000L +#define MCA_UMC_UMC0_MCUMC_STATUST0__RESERV41_MASK 0x0000060000000000L +#define MCA_UMC_UMC0_MCUMC_STATUST0__Poison_MASK 0x0000080000000000L +#define MCA_UMC_UMC0_MCUMC_STATUST0__Deferred_MASK 0x0000100000000000L +#define MCA_UMC_UMC0_MCUMC_STATUST0__UECC_MASK 0x0000200000000000L +#define MCA_UMC_UMC0_MCUMC_STATUST0__CECC_MASK 0x0000400000000000L +#define MCA_UMC_UMC0_MCUMC_STATUST0__RESERV47_MASK 0x000F800000000000L +#define MCA_UMC_UMC0_MCUMC_STATUST0__Transparent_MASK 0x0010000000000000L +#define MCA_UMC_UMC0_MCUMC_STATUST0__SyndV_MASK 0x0020000000000000L +#define MCA_UMC_UMC0_MCUMC_STATUST0__RESERV54_MASK 0x0040000000000000L +#define MCA_UMC_UMC0_MCUMC_STATUST0__TCC_MASK 0x0080000000000000L +#define MCA_UMC_UMC0_MCUMC_STATUST0__ErrCoreIdVal_MASK 0x0100000000000000L +#define MCA_UMC_UMC0_MCUMC_STATUST0__PCC_MASK 0x0200000000000000L +#define MCA_UMC_UMC0_MCUMC_STATUST0__AddrV_MASK 0x0400000000000000L +#define MCA_UMC_UMC0_MCUMC_STATUST0__MiscV_MASK 0x0800000000000000L +#define MCA_UMC_UMC0_MCUMC_STATUST0__En_MASK 0x1000000000000000L +#define MCA_UMC_UMC0_MCUMC_STATUST0__UC_MASK 0x2000000000000000L +#define MCA_UMC_UMC0_MCUMC_STATUST0__Overflow_MASK 0x4000000000000000L +#define MCA_UMC_UMC0_MCUMC_STATUST0__Val_MASK 0x8000000000000000L +//MCA_UMC_UMC0_MCUMC_ADDRT0 +#define MCA_UMC_UMC0_MCUMC_ADDRT0__ErrorAddr__SHIFT 0x0 +#define MCA_UMC_UMC0_MCUMC_ADDRT0__LSB__SHIFT 0x38 +#define MCA_UMC_UMC0_MCUMC_ADDRT0__Reserved__SHIFT 0x3e +#define MCA_UMC_UMC0_MCUMC_ADDRT0__ErrorAddr_MASK 0x00FFFFFFFFFFFFFFL +#define MCA_UMC_UMC0_MCUMC_ADDRT0__LSB_MASK 0x3F00000000000000L +#define MCA_UMC_UMC0_MCUMC_ADDRT0__Reserved_MASK 0xC000000000000000L + +#endif diff --git a/drivers/gpu/drm/amd/include/cyan_skillfish_ip_offset.h b/drivers/gpu/drm/amd/include/cyan_skillfish_ip_offset.h new file mode 100644 index 000000000000..9cb5f3631c60 --- /dev/null +++ b/drivers/gpu/drm/amd/include/cyan_skillfish_ip_offset.h @@ -0,0 +1,714 @@ +/* + * Copyright (C) 2018 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included + * in all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS + * OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN + * AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. + */ +#ifndef _cyan_skillfish_ip_offset_HEADER +#define _cyan_skillfish_ip_offset_HEADER + +#define MAX_INSTANCE 6 +#define MAX_SEGMENT 5 + + +struct IP_BASE_INSTANCE +{ + unsigned int segment[MAX_SEGMENT]; +}; + +struct IP_BASE +{ + struct IP_BASE_INSTANCE instance[MAX_INSTANCE]; +}; + + +static const struct IP_BASE ATHUB_BASE ={ { { { 0x00000C00, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } } } }; +static const struct IP_BASE CLK_BASE ={ { { { 0x00016C00, 0, 0, 0, 0 } }, + { { 0x00016E00, 0, 0, 0, 0 } }, + { { 0x00017000, 0, 0, 0, 0 } }, + { { 0x00017200, 0, 0, 0, 0 } }, + { { 0x00017E00, 0, 0, 0, 0 } }, + { { 0x0001B000, 0, 0, 0, 0 } } } }; +static const struct IP_BASE DF_BASE ={ { { { 0x00007000, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } } } }; +static const struct IP_BASE DMU_BASE ={ { { { 0x00000012, 0x000000C0, 0x000034C0, 0x00009000, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } } } }; +static const struct IP_BASE FUSE_BASE ={ { { { 0x00017400, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } } } }; +static const struct IP_BASE GC_BASE ={ { { { 0x00001260, 0x0000A000, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } } } }; +static const struct IP_BASE HDP_BASE ={ { { { 0x00000F20, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } } } }; +static const struct IP_BASE MMHUB_BASE ={ { { { 0x0001A000, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } } } }; +static const struct IP_BASE MP0_BASE ={ { { { 0x00016000, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } } } }; +static const struct IP_BASE MP1_BASE ={ { { { 0x00016000, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } } } }; +static const struct IP_BASE NBIO_BASE ={ { { { 0x00000000, 0x00000014, 0x00000D20, 0x00010400, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } } } }; +static const struct IP_BASE OSSSYS_BASE ={ { { { 0x000010A0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } } } }; +static const struct IP_BASE SMUIO_BASE ={ { { { 0x00016800, 0x00016A00, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } } } }; +static const struct IP_BASE THM_BASE ={ { { { 0x00016600, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } } } }; +static const struct IP_BASE UMC0_BASE ={ { { { 0x00014000, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } } } }; +static const struct IP_BASE UVD0_BASE ={ { { { 0x00007800, 0x00007E00, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } }, + { { 0, 0, 0, 0, 0 } } } }; + + +#define ATHUB_BASE__INST0_SEG0 0x00000C00 +#define ATHUB_BASE__INST0_SEG1 0 +#define ATHUB_BASE__INST0_SEG2 0 +#define ATHUB_BASE__INST0_SEG3 0 +#define ATHUB_BASE__INST0_SEG4 0 + +#define ATHUB_BASE__INST1_SEG0 0 +#define ATHUB_BASE__INST1_SEG1 0 +#define ATHUB_BASE__INST1_SEG2 0 +#define ATHUB_BASE__INST1_SEG3 0 +#define ATHUB_BASE__INST1_SEG4 0 + +#define ATHUB_BASE__INST2_SEG0 0 +#define ATHUB_BASE__INST2_SEG1 0 +#define ATHUB_BASE__INST2_SEG2 0 +#define ATHUB_BASE__INST2_SEG3 0 +#define ATHUB_BASE__INST2_SEG4 0 + +#define ATHUB_BASE__INST3_SEG0 0 +#define ATHUB_BASE__INST3_SEG1 0 +#define ATHUB_BASE__INST3_SEG2 0 +#define ATHUB_BASE__INST3_SEG3 0 +#define ATHUB_BASE__INST3_SEG4 0 + +#define ATHUB_BASE__INST4_SEG0 0 +#define ATHUB_BASE__INST4_SEG1 0 +#define ATHUB_BASE__INST4_SEG2 0 +#define ATHUB_BASE__INST4_SEG3 0 +#define ATHUB_BASE__INST4_SEG4 0 + +#define ATHUB_BASE__INST5_SEG0 0 +#define ATHUB_BASE__INST5_SEG1 0 +#define ATHUB_BASE__INST5_SEG2 0 +#define ATHUB_BASE__INST5_SEG3 0 +#define ATHUB_BASE__INST5_SEG4 0 + +#define CLK_BASE__INST0_SEG0 0x00016C00 +#define CLK_BASE__INST0_SEG1 0 +#define CLK_BASE__INST0_SEG2 0 +#define CLK_BASE__INST0_SEG3 0 +#define CLK_BASE__INST0_SEG4 0 + +#define CLK_BASE__INST1_SEG0 0x00016E00 +#define CLK_BASE__INST1_SEG1 0 +#define CLK_BASE__INST1_SEG2 0 +#define CLK_BASE__INST1_SEG3 0 +#define CLK_BASE__INST1_SEG4 0 + +#define CLK_BASE__INST2_SEG0 0x00017000 +#define CLK_BASE__INST2_SEG1 0 +#define CLK_BASE__INST2_SEG2 0 +#define CLK_BASE__INST2_SEG3 0 +#define CLK_BASE__INST2_SEG4 0 + +#define CLK_BASE__INST3_SEG0 0x00017200 +#define CLK_BASE__INST3_SEG1 0 +#define CLK_BASE__INST3_SEG2 0 +#define CLK_BASE__INST3_SEG3 0 +#define CLK_BASE__INST3_SEG4 0 + +#define CLK_BASE__INST4_SEG0 0x00017E00 +#define CLK_BASE__INST4_SEG1 0 +#define CLK_BASE__INST4_SEG2 0 +#define CLK_BASE__INST4_SEG3 0 +#define CLK_BASE__INST4_SEG4 0 + +#define CLK_BASE__INST5_SEG0 0x0001B000 +#define CLK_BASE__INST5_SEG1 0 +#define CLK_BASE__INST5_SEG2 0 +#define CLK_BASE__INST5_SEG3 0 +#define CLK_BASE__INST5_SEG4 0 + +#define DF_BASE__INST0_SEG0 0x00007000 +#define DF_BASE__INST0_SEG1 0 +#define DF_BASE__INST0_SEG2 0 +#define DF_BASE__INST0_SEG3 0 +#define DF_BASE__INST0_SEG4 0 + +#define DF_BASE__INST1_SEG0 0 +#define DF_BASE__INST1_SEG1 0 +#define DF_BASE__INST1_SEG2 0 +#define DF_BASE__INST1_SEG3 0 +#define DF_BASE__INST1_SEG4 0 + +#define DF_BASE__INST2_SEG0 0 +#define DF_BASE__INST2_SEG1 0 +#define DF_BASE__INST2_SEG2 0 +#define DF_BASE__INST2_SEG3 0 +#define DF_BASE__INST2_SEG4 0 + +#define DF_BASE__INST3_SEG0 0 +#define DF_BASE__INST3_SEG1 0 +#define DF_BASE__INST3_SEG2 0 +#define DF_BASE__INST3_SEG3 0 +#define DF_BASE__INST3_SEG4 0 + +#define DF_BASE__INST4_SEG0 0 +#define DF_BASE__INST4_SEG1 0 +#define DF_BASE__INST4_SEG2 0 +#define DF_BASE__INST4_SEG3 0 +#define DF_BASE__INST4_SEG4 0 + +#define DF_BASE__INST5_SEG0 0 +#define DF_BASE__INST5_SEG1 0 +#define DF_BASE__INST5_SEG2 0 +#define DF_BASE__INST5_SEG3 0 +#define DF_BASE__INST5_SEG4 0 + +#define DMU_BASE__INST0_SEG0 0x00000012 +#define DMU_BASE__INST0_SEG1 0x000000C0 +#define DMU_BASE__INST0_SEG2 0x000034C0 +#define DMU_BASE__INST0_SEG3 0x00009000 +#define DMU_BASE__INST0_SEG4 0 + +#define DMU_BASE__INST1_SEG0 0 +#define DMU_BASE__INST1_SEG1 0 +#define DMU_BASE__INST1_SEG2 0 +#define DMU_BASE__INST1_SEG3 0 +#define DMU_BASE__INST1_SEG4 0 + +#define DMU_BASE__INST2_SEG0 0 +#define DMU_BASE__INST2_SEG1 0 +#define DMU_BASE__INST2_SEG2 0 +#define DMU_BASE__INST2_SEG3 0 +#define DMU_BASE__INST2_SEG4 0 + +#define DMU_BASE__INST3_SEG0 0 +#define DMU_BASE__INST3_SEG1 0 +#define DMU_BASE__INST3_SEG2 0 +#define DMU_BASE__INST3_SEG3 0 +#define DMU_BASE__INST3_SEG4 0 + +#define DMU_BASE__INST4_SEG0 0 +#define DMU_BASE__INST4_SEG1 0 +#define DMU_BASE__INST4_SEG2 0 +#define DMU_BASE__INST4_SEG3 0 +#define DMU_BASE__INST4_SEG4 0 + +#define DMU_BASE__INST5_SEG0 0 +#define DMU_BASE__INST5_SEG1 0 +#define DMU_BASE__INST5_SEG2 0 +#define DMU_BASE__INST5_SEG3 0 +#define DMU_BASE__INST5_SEG4 0 + +#define FUSE_BASE__INST0_SEG0 0x00017400 +#define FUSE_BASE__INST0_SEG1 0 +#define FUSE_BASE__INST0_SEG2 0 +#define FUSE_BASE__INST0_SEG3 0 +#define FUSE_BASE__INST0_SEG4 0 + +#define FUSE_BASE__INST1_SEG0 0 +#define FUSE_BASE__INST1_SEG1 0 +#define FUSE_BASE__INST1_SEG2 0 +#define FUSE_BASE__INST1_SEG3 0 +#define FUSE_BASE__INST1_SEG4 0 + +#define FUSE_BASE__INST2_SEG0 0 +#define FUSE_BASE__INST2_SEG1 0 +#define FUSE_BASE__INST2_SEG2 0 +#define FUSE_BASE__INST2_SEG3 0 +#define FUSE_BASE__INST2_SEG4 0 + +#define FUSE_BASE__INST3_SEG0 0 +#define FUSE_BASE__INST3_SEG1 0 +#define FUSE_BASE__INST3_SEG2 0 +#define FUSE_BASE__INST3_SEG3 0 +#define FUSE_BASE__INST3_SEG4 0 + +#define FUSE_BASE__INST4_SEG0 0 +#define FUSE_BASE__INST4_SEG1 0 +#define FUSE_BASE__INST4_SEG2 0 +#define FUSE_BASE__INST4_SEG3 0 +#define FUSE_BASE__INST4_SEG4 0 + +#define FUSE_BASE__INST5_SEG0 0 +#define FUSE_BASE__INST5_SEG1 0 +#define FUSE_BASE__INST5_SEG2 0 +#define FUSE_BASE__INST5_SEG3 0 +#define FUSE_BASE__INST5_SEG4 0 + +#define GC_BASE__INST0_SEG0 0x00001260 +#define GC_BASE__INST0_SEG1 0x0000A000 +#define GC_BASE__INST0_SEG2 0 +#define GC_BASE__INST0_SEG3 0 +#define GC_BASE__INST0_SEG4 0 + +#define GC_BASE__INST1_SEG0 0 +#define GC_BASE__INST1_SEG1 0 +#define GC_BASE__INST1_SEG2 0 +#define GC_BASE__INST1_SEG3 0 +#define GC_BASE__INST1_SEG4 0 + +#define GC_BASE__INST2_SEG0 0 +#define GC_BASE__INST2_SEG1 0 +#define GC_BASE__INST2_SEG2 0 +#define GC_BASE__INST2_SEG3 0 +#define GC_BASE__INST2_SEG4 0 + +#define GC_BASE__INST3_SEG0 0 +#define GC_BASE__INST3_SEG1 0 +#define GC_BASE__INST3_SEG2 0 +#define GC_BASE__INST3_SEG3 0 +#define GC_BASE__INST3_SEG4 0 + +#define GC_BASE__INST4_SEG0 0 +#define GC_BASE__INST4_SEG1 0 +#define GC_BASE__INST4_SEG2 0 +#define GC_BASE__INST4_SEG3 0 +#define GC_BASE__INST4_SEG4 0 + +#define GC_BASE__INST5_SEG0 0 +#define GC_BASE__INST5_SEG1 0 +#define GC_BASE__INST5_SEG2 0 +#define GC_BASE__INST5_SEG3 0 +#define GC_BASE__INST5_SEG4 0 + +#define HDP_BASE__INST0_SEG0 0x00000F20 +#define HDP_BASE__INST0_SEG1 0 +#define HDP_BASE__INST0_SEG2 0 +#define HDP_BASE__INST0_SEG3 0 +#define HDP_BASE__INST0_SEG4 0 + +#define HDP_BASE__INST1_SEG0 0 +#define HDP_BASE__INST1_SEG1 0 +#define HDP_BASE__INST1_SEG2 0 +#define HDP_BASE__INST1_SEG3 0 +#define HDP_BASE__INST1_SEG4 0 + +#define HDP_BASE__INST2_SEG0 0 +#define HDP_BASE__INST2_SEG1 0 +#define HDP_BASE__INST2_SEG2 0 +#define HDP_BASE__INST2_SEG3 0 +#define HDP_BASE__INST2_SEG4 0 + +#define HDP_BASE__INST3_SEG0 0 +#define HDP_BASE__INST3_SEG1 0 +#define HDP_BASE__INST3_SEG2 0 +#define HDP_BASE__INST3_SEG3 0 +#define HDP_BASE__INST3_SEG4 0 + +#define HDP_BASE__INST4_SEG0 0 +#define HDP_BASE__INST4_SEG1 0 +#define HDP_BASE__INST4_SEG2 0 +#define HDP_BASE__INST4_SEG3 0 +#define HDP_BASE__INST4_SEG4 0 + +#define HDP_BASE__INST5_SEG0 0 +#define HDP_BASE__INST5_SEG1 0 +#define HDP_BASE__INST5_SEG2 0 +#define HDP_BASE__INST5_SEG3 0 +#define HDP_BASE__INST5_SEG4 0 + +#define MMHUB_BASE__INST0_SEG0 0x0001A000 +#define MMHUB_BASE__INST0_SEG1 0 +#define MMHUB_BASE__INST0_SEG2 0 +#define MMHUB_BASE__INST0_SEG3 0 +#define MMHUB_BASE__INST0_SEG4 0 + +#define MMHUB_BASE__INST1_SEG0 0 +#define MMHUB_BASE__INST1_SEG1 0 +#define MMHUB_BASE__INST1_SEG2 0 +#define MMHUB_BASE__INST1_SEG3 0 +#define MMHUB_BASE__INST1_SEG4 0 + +#define MMHUB_BASE__INST2_SEG0 0 +#define MMHUB_BASE__INST2_SEG1 0 +#define MMHUB_BASE__INST2_SEG2 0 +#define MMHUB_BASE__INST2_SEG3 0 +#define MMHUB_BASE__INST2_SEG4 0 + +#define MMHUB_BASE__INST3_SEG0 0 +#define MMHUB_BASE__INST3_SEG1 0 +#define MMHUB_BASE__INST3_SEG2 0 +#define MMHUB_BASE__INST3_SEG3 0 +#define MMHUB_BASE__INST3_SEG4 0 + +#define MMHUB_BASE__INST4_SEG0 0 +#define MMHUB_BASE__INST4_SEG1 0 +#define MMHUB_BASE__INST4_SEG2 0 +#define MMHUB_BASE__INST4_SEG3 0 +#define MMHUB_BASE__INST4_SEG4 0 + +#define MMHUB_BASE__INST5_SEG0 0 +#define MMHUB_BASE__INST5_SEG1 0 +#define MMHUB_BASE__INST5_SEG2 0 +#define MMHUB_BASE__INST5_SEG3 0 +#define MMHUB_BASE__INST5_SEG4 0 + +#define MP0_BASE__INST0_SEG0 0x00016000 +#define MP0_BASE__INST0_SEG1 0 +#define MP0_BASE__INST0_SEG2 0 +#define MP0_BASE__INST0_SEG3 0 +#define MP0_BASE__INST0_SEG4 0 + +#define MP0_BASE__INST1_SEG0 0 +#define MP0_BASE__INST1_SEG1 0 +#define MP0_BASE__INST1_SEG2 0 +#define MP0_BASE__INST1_SEG3 0 +#define MP0_BASE__INST1_SEG4 0 + +#define MP0_BASE__INST2_SEG0 0 +#define MP0_BASE__INST2_SEG1 0 +#define MP0_BASE__INST2_SEG2 0 +#define MP0_BASE__INST2_SEG3 0 +#define MP0_BASE__INST2_SEG4 0 + +#define MP0_BASE__INST3_SEG0 0 +#define MP0_BASE__INST3_SEG1 0 +#define MP0_BASE__INST3_SEG2 0 +#define MP0_BASE__INST3_SEG3 0 +#define MP0_BASE__INST3_SEG4 0 + +#define MP0_BASE__INST4_SEG0 0 +#define MP0_BASE__INST4_SEG1 0 +#define MP0_BASE__INST4_SEG2 0 +#define MP0_BASE__INST4_SEG3 0 +#define MP0_BASE__INST4_SEG4 0 + +#define MP0_BASE__INST5_SEG0 0 +#define MP0_BASE__INST5_SEG1 0 +#define MP0_BASE__INST5_SEG2 0 +#define MP0_BASE__INST5_SEG3 0 +#define MP0_BASE__INST5_SEG4 0 + +#define MP1_BASE__INST0_SEG0 0x00016000 +#define MP1_BASE__INST0_SEG1 0 +#define MP1_BASE__INST0_SEG2 0 +#define MP1_BASE__INST0_SEG3 0 +#define MP1_BASE__INST0_SEG4 0 + +#define MP1_BASE__INST1_SEG0 0 +#define MP1_BASE__INST1_SEG1 0 +#define MP1_BASE__INST1_SEG2 0 +#define MP1_BASE__INST1_SEG3 0 +#define MP1_BASE__INST1_SEG4 0 + +#define MP1_BASE__INST2_SEG0 0 +#define MP1_BASE__INST2_SEG1 0 +#define MP1_BASE__INST2_SEG2 0 +#define MP1_BASE__INST2_SEG3 0 +#define MP1_BASE__INST2_SEG4 0 + +#define MP1_BASE__INST3_SEG0 0 +#define MP1_BASE__INST3_SEG1 0 +#define MP1_BASE__INST3_SEG2 0 +#define MP1_BASE__INST3_SEG3 0 +#define MP1_BASE__INST3_SEG4 0 + +#define MP1_BASE__INST4_SEG0 0 +#define MP1_BASE__INST4_SEG1 0 +#define MP1_BASE__INST4_SEG2 0 +#define MP1_BASE__INST4_SEG3 0 +#define MP1_BASE__INST4_SEG4 0 + +#define MP1_BASE__INST5_SEG0 0 +#define MP1_BASE__INST5_SEG1 0 +#define MP1_BASE__INST5_SEG2 0 +#define MP1_BASE__INST5_SEG3 0 +#define MP1_BASE__INST5_SEG4 0 + +#define NBIO_BASE__INST0_SEG0 0x00000000 +#define NBIO_BASE__INST0_SEG1 0x00000014 +#define NBIO_BASE__INST0_SEG2 0x00000D20 +#define NBIO_BASE__INST0_SEG3 0x00010400 +#define NBIO_BASE__INST0_SEG4 0 + +#define NBIO_BASE__INST1_SEG0 0 +#define NBIO_BASE__INST1_SEG1 0 +#define NBIO_BASE__INST1_SEG2 0 +#define NBIO_BASE__INST1_SEG3 0 +#define NBIO_BASE__INST1_SEG4 0 + +#define NBIO_BASE__INST2_SEG0 0 +#define NBIO_BASE__INST2_SEG1 0 +#define NBIO_BASE__INST2_SEG2 0 +#define NBIO_BASE__INST2_SEG3 0 +#define NBIO_BASE__INST2_SEG4 0 + +#define NBIO_BASE__INST3_SEG0 0 +#define NBIO_BASE__INST3_SEG1 0 +#define NBIO_BASE__INST3_SEG2 0 +#define NBIO_BASE__INST3_SEG3 0 +#define NBIO_BASE__INST3_SEG4 0 + +#define NBIO_BASE__INST4_SEG0 0 +#define NBIO_BASE__INST4_SEG1 0 +#define NBIO_BASE__INST4_SEG2 0 +#define NBIO_BASE__INST4_SEG3 0 +#define NBIO_BASE__INST4_SEG4 0 + +#define NBIO_BASE__INST5_SEG0 0 +#define NBIO_BASE__INST5_SEG1 0 +#define NBIO_BASE__INST5_SEG2 0 +#define NBIO_BASE__INST5_SEG3 0 +#define NBIO_BASE__INST5_SEG4 0 + +#define OSSSYS_BASE__INST0_SEG0 0x000010A0 +#define OSSSYS_BASE__INST0_SEG1 0 +#define OSSSYS_BASE__INST0_SEG2 0 +#define OSSSYS_BASE__INST0_SEG3 0 +#define OSSSYS_BASE__INST0_SEG4 0 + +#define OSSSYS_BASE__INST1_SEG0 0 +#define OSSSYS_BASE__INST1_SEG1 0 +#define OSSSYS_BASE__INST1_SEG2 0 +#define OSSSYS_BASE__INST1_SEG3 0 +#define OSSSYS_BASE__INST1_SEG4 0 + +#define OSSSYS_BASE__INST2_SEG0 0 +#define OSSSYS_BASE__INST2_SEG1 0 +#define OSSSYS_BASE__INST2_SEG2 0 +#define OSSSYS_BASE__INST2_SEG3 0 +#define OSSSYS_BASE__INST2_SEG4 0 + +#define OSSSYS_BASE__INST3_SEG0 0 +#define OSSSYS_BASE__INST3_SEG1 0 +#define OSSSYS_BASE__INST3_SEG2 0 +#define OSSSYS_BASE__INST3_SEG3 0 +#define OSSSYS_BASE__INST3_SEG4 0 + +#define OSSSYS_BASE__INST4_SEG0 0 +#define OSSSYS_BASE__INST4_SEG1 0 +#define OSSSYS_BASE__INST4_SEG2 0 +#define OSSSYS_BASE__INST4_SEG3 0 +#define OSSSYS_BASE__INST4_SEG4 0 + +#define OSSSYS_BASE__INST5_SEG0 0 +#define OSSSYS_BASE__INST5_SEG1 0 +#define OSSSYS_BASE__INST5_SEG2 0 +#define OSSSYS_BASE__INST5_SEG3 0 +#define OSSSYS_BASE__INST5_SEG4 0 + +#define SMUIO_BASE__INST0_SEG0 0x00016800 +#define SMUIO_BASE__INST0_SEG1 0x00016A00 +#define SMUIO_BASE__INST0_SEG2 0 +#define SMUIO_BASE__INST0_SEG3 0 +#define SMUIO_BASE__INST0_SEG4 0 + +#define SMUIO_BASE__INST1_SEG0 0 +#define SMUIO_BASE__INST1_SEG1 0 +#define SMUIO_BASE__INST1_SEG2 0 +#define SMUIO_BASE__INST1_SEG3 0 +#define SMUIO_BASE__INST1_SEG4 0 + +#define SMUIO_BASE__INST2_SEG0 0 +#define SMUIO_BASE__INST2_SEG1 0 +#define SMUIO_BASE__INST2_SEG2 0 +#define SMUIO_BASE__INST2_SEG3 0 +#define SMUIO_BASE__INST2_SEG4 0 + +#define SMUIO_BASE__INST3_SEG0 0 +#define SMUIO_BASE__INST3_SEG1 0 +#define SMUIO_BASE__INST3_SEG2 0 +#define SMUIO_BASE__INST3_SEG3 0 +#define SMUIO_BASE__INST3_SEG4 0 + +#define SMUIO_BASE__INST4_SEG0 0 +#define SMUIO_BASE__INST4_SEG1 0 +#define SMUIO_BASE__INST4_SEG2 0 +#define SMUIO_BASE__INST4_SEG3 0 +#define SMUIO_BASE__INST4_SEG4 0 + +#define SMUIO_BASE__INST5_SEG0 0 +#define SMUIO_BASE__INST5_SEG1 0 +#define SMUIO_BASE__INST5_SEG2 0 +#define SMUIO_BASE__INST5_SEG3 0 +#define SMUIO_BASE__INST5_SEG4 0 + +#define THM_BASE__INST0_SEG0 0x00016600 +#define THM_BASE__INST0_SEG1 0 +#define THM_BASE__INST0_SEG2 0 +#define THM_BASE__INST0_SEG3 0 +#define THM_BASE__INST0_SEG4 0 + +#define THM_BASE__INST1_SEG0 0 +#define THM_BASE__INST1_SEG1 0 +#define THM_BASE__INST1_SEG2 0 +#define THM_BASE__INST1_SEG3 0 +#define THM_BASE__INST1_SEG4 0 + +#define THM_BASE__INST2_SEG0 0 +#define THM_BASE__INST2_SEG1 0 +#define THM_BASE__INST2_SEG2 0 +#define THM_BASE__INST2_SEG3 0 +#define THM_BASE__INST2_SEG4 0 + +#define THM_BASE__INST3_SEG0 0 +#define THM_BASE__INST3_SEG1 0 +#define THM_BASE__INST3_SEG2 0 +#define THM_BASE__INST3_SEG3 0 +#define THM_BASE__INST3_SEG4 0 + +#define THM_BASE__INST4_SEG0 0 +#define THM_BASE__INST4_SEG1 0 +#define THM_BASE__INST4_SEG2 0 +#define THM_BASE__INST4_SEG3 0 +#define THM_BASE__INST4_SEG4 0 + +#define THM_BASE__INST5_SEG0 0 +#define THM_BASE__INST5_SEG1 0 +#define THM_BASE__INST5_SEG2 0 +#define THM_BASE__INST5_SEG3 0 +#define THM_BASE__INST5_SEG4 0 + +#define UMC0_BASE__INST0_SEG0 0x00014000 +#define UMC0_BASE__INST0_SEG1 0 +#define UMC0_BASE__INST0_SEG2 0 +#define UMC0_BASE__INST0_SEG3 0 +#define UMC0_BASE__INST0_SEG4 0 + +#define UMC0_BASE__INST1_SEG0 0 +#define UMC0_BASE__INST1_SEG1 0 +#define UMC0_BASE__INST1_SEG2 0 +#define UMC0_BASE__INST1_SEG3 0 +#define UMC0_BASE__INST1_SEG4 0 + +#define UMC0_BASE__INST2_SEG0 0 +#define UMC0_BASE__INST2_SEG1 0 +#define UMC0_BASE__INST2_SEG2 0 +#define UMC0_BASE__INST2_SEG3 0 +#define UMC0_BASE__INST2_SEG4 0 + +#define UMC0_BASE__INST3_SEG0 0 +#define UMC0_BASE__INST3_SEG1 0 +#define UMC0_BASE__INST3_SEG2 0 +#define UMC0_BASE__INST3_SEG3 0 +#define UMC0_BASE__INST3_SEG4 0 + +#define UMC0_BASE__INST4_SEG0 0 +#define UMC0_BASE__INST4_SEG1 0 +#define UMC0_BASE__INST4_SEG2 0 +#define UMC0_BASE__INST4_SEG3 0 +#define UMC0_BASE__INST4_SEG4 0 + +#define UMC0_BASE__INST5_SEG0 0 +#define UMC0_BASE__INST5_SEG1 0 +#define UMC0_BASE__INST5_SEG2 0 +#define UMC0_BASE__INST5_SEG3 0 +#define UMC0_BASE__INST5_SEG4 0 + +#define UVD0_BASE__INST0_SEG0 0x00007800 +#define UVD0_BASE__INST0_SEG1 0x00007E00 +#define UVD0_BASE__INST0_SEG2 0 +#define UVD0_BASE__INST0_SEG3 0 +#define UVD0_BASE__INST0_SEG4 0 + +#define UVD0_BASE__INST1_SEG0 0 +#define UVD0_BASE__INST1_SEG1 0 +#define UVD0_BASE__INST1_SEG2 0 +#define UVD0_BASE__INST1_SEG3 0 +#define UVD0_BASE__INST1_SEG4 0 + +#define UVD0_BASE__INST2_SEG0 0 +#define UVD0_BASE__INST2_SEG1 0 +#define UVD0_BASE__INST2_SEG2 0 +#define UVD0_BASE__INST2_SEG3 0 +#define UVD0_BASE__INST2_SEG4 0 + +#define UVD0_BASE__INST3_SEG0 0 +#define UVD0_BASE__INST3_SEG1 0 +#define UVD0_BASE__INST3_SEG2 0 +#define UVD0_BASE__INST3_SEG3 0 +#define UVD0_BASE__INST3_SEG4 0 + +#define UVD0_BASE__INST4_SEG0 0 +#define UVD0_BASE__INST4_SEG1 0 +#define UVD0_BASE__INST4_SEG2 0 +#define UVD0_BASE__INST4_SEG3 0 +#define UVD0_BASE__INST4_SEG4 0 + +#define UVD0_BASE__INST5_SEG0 0 +#define UVD0_BASE__INST5_SEG1 0 +#define UVD0_BASE__INST5_SEG2 0 +#define UVD0_BASE__INST5_SEG3 0 +#define UVD0_BASE__INST5_SEG4 0 + +#endif + diff --git a/drivers/gpu/drm/amd/pm/inc/amdgpu_dpm.h b/drivers/gpu/drm/amd/pm/inc/amdgpu_dpm.h index f6e0e7d8a007..d03e6fa2bf1a 100644 --- a/drivers/gpu/drm/amd/pm/inc/amdgpu_dpm.h +++ b/drivers/gpu/drm/amd/pm/inc/amdgpu_dpm.h @@ -450,6 +450,7 @@ struct amdgpu_pm { /* Used for I2C access to various EEPROMs on relevant ASICs */ struct i2c_adapter smu_i2c; + struct mutex smu_i2c_mutex; struct list_head pm_attr_list; }; diff --git a/drivers/gpu/drm/amd/pm/inc/amdgpu_smu.h b/drivers/gpu/drm/amd/pm/inc/amdgpu_smu.h index 3e89852e4820..c2c201b8e3cf 100644 --- a/drivers/gpu/drm/amd/pm/inc/amdgpu_smu.h +++ b/drivers/gpu/drm/amd/pm/inc/amdgpu_smu.h @@ -231,6 +231,7 @@ struct smu_user_dpm_profile { uint32_t power_limit; uint32_t fan_speed_percent; uint32_t flags; + uint32_t user_od; /* user clock state information */ uint32_t clk_mask[SMU_CLK_COUNT]; @@ -352,6 +353,7 @@ struct smu_table_context void *overdrive_table; void *boot_overdrive_table; + void *user_overdrive_table; uint32_t gpu_metrics_table_size; void *gpu_metrics_table; @@ -624,6 +626,12 @@ struct pptable_funcs { long *input, uint32_t size); /** + * @restore_user_od_settings: Restore the user customized + * OD settings on S3/S4/Runpm resume. + */ + int (*restore_user_od_settings)(struct smu_context *smu); + + /** * @get_clock_by_type_with_latency: Get the speed and latency of a clock * domain. */ diff --git a/drivers/gpu/drm/amd/pm/inc/smu11_driver_if_cyan_skillfish.h b/drivers/gpu/drm/amd/pm/inc/smu11_driver_if_cyan_skillfish.h new file mode 100644 index 000000000000..8a08ecc34c69 --- /dev/null +++ b/drivers/gpu/drm/amd/pm/inc/smu11_driver_if_cyan_skillfish.h @@ -0,0 +1,95 @@ +/* + * Copyright 2021 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ +#ifndef __SMU11_DRIVER_IF_CYAN_SKILLFISH_H__ +#define __SMU11_DRIVER_IF_CYAN_SKILLFISH_H__ + +// *** IMPORTANT *** +// Always increment the interface version if +// any structure is changed in this file +#define MP1_DRIVER_IF_VERSION 0x8 + +#define TABLE_BIOS_IF 0 // Called by BIOS +#define TABLE_WATERMARKS 1 // Called by Driver; defined here, but not used, for backward compatible +#define TABLE_PMSTATUSLOG 3 // Called by Tools for Agm logging +#define TABLE_DPMCLOCKS 4 // Called by Driver; defined here, but not used, for backward compatible +#define TABLE_MOMENTARY_PM 5 // Called by Tools; defined here, but not used, for backward compatible +#define TABLE_COUNT 6 + +#define NUM_DSPCLK_LEVELS 8 +#define NUM_SOCCLK_DPM_LEVELS 8 +#define NUM_DCEFCLK_DPM_LEVELS 4 +#define NUM_FCLK_DPM_LEVELS 4 +#define NUM_MEMCLK_DPM_LEVELS 4 + +#define NUMBER_OF_PSTATES 8 +#define NUMBER_OF_CORES 8 + +typedef enum { + S3_TYPE_ENTRY, + S5_TYPE_ENTRY, +} Sleep_Type_e; + +typedef enum { + GFX_OFF = 0, + GFX_ON = 1, +} GFX_Mode_e; + +typedef enum { + CPU_P0 = 0, + CPU_P1, + CPU_P2, + CPU_P3, + CPU_P4, + CPU_P5, + CPU_P6, + CPU_P7 +} CPU_PState_e; + +typedef enum { + CPU_CORE0 = 0, + CPU_CORE1, + CPU_CORE2, + CPU_CORE3, + CPU_CORE4, + CPU_CORE5, + CPU_CORE6, + CPU_CORE7 +} CORE_ID_e; + +typedef enum { + DF_DPM0 = 0, + DF_DPM1, + DF_DPM2, + DF_DPM3, + DF_PState_Count +} DF_PState_e; + +typedef enum { + GFX_DPM0 = 0, + GFX_DPM1, + GFX_DPM2, + GFX_DPM3, + GFX_PState_Count +} GFX_PState_e; + +#endif diff --git a/drivers/gpu/drm/amd/pm/inc/smu11_driver_if_sienna_cichlid.h b/drivers/gpu/drm/amd/pm/inc/smu11_driver_if_sienna_cichlid.h index 61c87c39be80..63b8701fd466 100644 --- a/drivers/gpu/drm/amd/pm/inc/smu11_driver_if_sienna_cichlid.h +++ b/drivers/gpu/drm/amd/pm/inc/smu11_driver_if_sienna_cichlid.h @@ -131,7 +131,7 @@ #define FEATURE_GFX_EDC_BIT 49 #define FEATURE_GFX_PER_PART_VMIN_BIT 50 #define FEATURE_SMART_SHIFT_BIT 51 -#define FEATURE_SPARE_52_BIT 52 +#define FEATURE_APT_BIT 52 #define FEATURE_SPARE_53_BIT 53 #define FEATURE_SPARE_54_BIT 54 #define FEATURE_SPARE_55_BIT 55 @@ -211,6 +211,7 @@ typedef enum { #define THROTTLER_FIT_BIT 17 #define THROTTLER_PPM_BIT 18 #define THROTTLER_APCC_BIT 19 +#define THROTTLER_COUNT 20 // FW DState Features Control Bits // FW DState Features Control Bits @@ -1406,7 +1407,67 @@ typedef struct { } SmuMetrics_t; typedef struct { - SmuMetrics_t SmuMetrics; + uint32_t CurrClock[PPCLK_COUNT]; + + uint16_t AverageGfxclkFrequencyPreDs; + uint16_t AverageGfxclkFrequencyPostDs; + uint16_t AverageFclkFrequencyPreDs; + uint16_t AverageFclkFrequencyPostDs; + uint16_t AverageUclkFrequencyPreDs ; + uint16_t AverageUclkFrequencyPostDs ; + + + uint16_t AverageGfxActivity ; + uint16_t AverageUclkActivity ; + uint8_t CurrSocVoltageOffset ; + uint8_t CurrGfxVoltageOffset ; + uint8_t CurrMemVidOffset ; + uint8_t Padding8 ; + uint16_t AverageSocketPower ; + uint16_t TemperatureEdge ; + uint16_t TemperatureHotspot ; + uint16_t TemperatureMem ; + uint16_t TemperatureVrGfx ; + uint16_t TemperatureVrMem0 ; + uint16_t TemperatureVrMem1 ; + uint16_t TemperatureVrSoc ; + uint16_t TemperatureLiquid0 ; + uint16_t TemperatureLiquid1 ; + uint16_t TemperaturePlx ; + uint16_t Padding16 ; + uint32_t AccCnt ; + uint8_t ThrottlingPercentage[THROTTLER_COUNT]; + + + uint8_t LinkDpmLevel; + uint8_t CurrFanPwm; + uint16_t CurrFanSpeed; + + //BACO metrics, PMFW-1721 + //metrics for D3hot entry/exit and driver ARM msgs + uint8_t D3HotEntryCountPerMode[D3HOT_SEQUENCE_COUNT]; + uint8_t D3HotExitCountPerMode[D3HOT_SEQUENCE_COUNT]; + uint8_t ArmMsgReceivedCountPerMode[D3HOT_SEQUENCE_COUNT]; + + //PMFW-4362 + uint32_t EnergyAccumulator; + uint16_t AverageVclk0Frequency ; + uint16_t AverageDclk0Frequency ; + uint16_t AverageVclk1Frequency ; + uint16_t AverageDclk1Frequency ; + uint16_t VcnActivityPercentage ; //place holder, David N. to provide full sequence + uint8_t PcieRate ; + uint8_t PcieWidth ; + uint16_t AverageGfxclkFrequencyTarget; + uint16_t Padding16_2; + +} SmuMetrics_V2_t; + +typedef struct { + union { + SmuMetrics_t SmuMetrics; + SmuMetrics_V2_t SmuMetrics_V2; + }; uint32_t Spare[1]; // Padding - ignore diff --git a/drivers/gpu/drm/amd/pm/inc/smu_v11_0.h b/drivers/gpu/drm/amd/pm/inc/smu_v11_0.h index f61b5c914a3d..403bc1bf8a77 100644 --- a/drivers/gpu/drm/amd/pm/inc/smu_v11_0.h +++ b/drivers/gpu/drm/amd/pm/inc/smu_v11_0.h @@ -30,11 +30,12 @@ #define SMU11_DRIVER_IF_VERSION_NV10 0x37 #define SMU11_DRIVER_IF_VERSION_NV12 0x38 #define SMU11_DRIVER_IF_VERSION_NV14 0x38 -#define SMU11_DRIVER_IF_VERSION_Sienna_Cichlid 0x3D +#define SMU11_DRIVER_IF_VERSION_Sienna_Cichlid 0x40 #define SMU11_DRIVER_IF_VERSION_Navy_Flounder 0xE #define SMU11_DRIVER_IF_VERSION_VANGOGH 0x03 #define SMU11_DRIVER_IF_VERSION_Dimgrey_Cavefish 0xF #define SMU11_DRIVER_IF_VERSION_Beige_Goby 0xD +#define SMU11_DRIVER_IF_VERSION_Cyan_Skillfish 0x8 /* MP Apertures */ #define MP0_Public 0x03800000 @@ -302,5 +303,7 @@ void smu_v11_0_interrupt_work(struct smu_context *smu); int smu_v11_0_set_light_sbr(struct smu_context *smu, bool enable); +int smu_v11_0_restore_user_od_settings(struct smu_context *smu); + #endif #endif diff --git a/drivers/gpu/drm/amd/pm/inc/smu_v11_8_pmfw.h b/drivers/gpu/drm/amd/pm/inc/smu_v11_8_pmfw.h new file mode 100644 index 000000000000..bd4fcb6b9610 --- /dev/null +++ b/drivers/gpu/drm/amd/pm/inc/smu_v11_8_pmfw.h @@ -0,0 +1,152 @@ +/* + * Copyright 2021 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#ifndef __SMU_V11_8_0_PMFW_H__ +#define __SMU_V11_8_0_PMFW_H__ + +#pragma pack(push, 1) + +#define ENABLE_DEBUG_FEATURES + +// Feature Control Defines +#define FEATURE_CCLK_CONTROLLER_BIT 0 +#define FEATURE_GFXCLK_EFFT_FREQ_BIT 1 +#define FEATURE_DATA_CALCULATION_BIT 2 +#define FEATURE_THERMAL_BIT 3 +#define FEATURE_PLL_POWER_DOWN_BIT 4 +#define FEATURE_FCLK_DPM_BIT 5 +#define FEATURE_GFX_DPM_BIT 6 +#define FEATURE_DS_GFXCLK_BIT 7 +#define FEATURE_DS_SOCCLK_BIT 8 +#define FEATURE_DS_LCLK_BIT 9 +#define FEATURE_CORE_CSTATES_BIT 10 +#define FEATURE_G6_SSC_BIT 11 //G6 memory UCLK and UCLK_DIV SS +#define FEATURE_RM_BIT 12 +#define FEATURE_SOC_DPM_BIT 13 +#define FEATURE_DS_SMNCLK_BIT 14 +#define FEATURE_DS_MP1CLK_BIT 15 +#define FEATURE_DS_MP0CLK_BIT 16 +#define FEATURE_MGCG_BIT 17 +#define FEATURE_DS_FUSE_SRAM_BIT 18 +#define FEATURE_GFX_CKS_BIT 19 +#define FEATURE_FP_THROTTLING_BIT 20 +#define FEATURE_PROCHOT_BIT 21 +#define FEATURE_CPUOFF_BIT 22 +#define FEATURE_UMC_THROTTLE_BIT 23 +#define FEATURE_DF_THROTTLE_BIT 24 +#define FEATURE_DS_MP3CLK_BIT 25 +#define FEATURE_DS_SHUBCLK_BIT 26 +#define FEATURE_TDC_BIT 27 //Legacy APM_BIT +#define FEATURE_UMC_CAL_SHARING_BIT 28 +#define FEATURE_DFLL_BTC_CALIBRATION_BIT 29 +#define FEATURE_EDC_BIT 30 +#define FEATURE_DLDO_BIT 31 +#define FEATURE_MEAS_DRAM_BLACKOUT_BIT 32 +#define FEATURE_CC1_BIT 33 +#define FEATURE_PPT_BIT 34 +#define FEATURE_STAPM_BIT 35 +#define FEATURE_CSTATE_BOOST_BIT 36 +#define FEATURE_SPARE_37_BIT 37 +#define FEATURE_SPARE_38_BIT 38 +#define FEATURE_SPARE_39_BIT 39 +#define FEATURE_SPARE_40_BIT 40 +#define FEATURE_SPARE_41_BIT 41 +#define FEATURE_SPARE_42_BIT 42 +#define FEATURE_SPARE_43_BIT 43 +#define FEATURE_SPARE_44_BIT 44 +#define FEATURE_SPARE_45_BIT 45 +#define FEATURE_SPARE_46_BIT 46 +#define FEATURE_SPARE_47_BIT 47 +#define FEATURE_SPARE_48_BIT 48 +#define FEATURE_SPARE_49_BIT 49 +#define FEATURE_SPARE_50_BIT 50 +#define FEATURE_SPARE_51_BIT 51 +#define FEATURE_SPARE_52_BIT 52 +#define FEATURE_SPARE_53_BIT 53 +#define FEATURE_SPARE_54_BIT 54 +#define FEATURE_SPARE_55_BIT 55 +#define FEATURE_SPARE_56_BIT 56 +#define FEATURE_SPARE_57_BIT 57 +#define FEATURE_SPARE_58_BIT 58 +#define FEATURE_SPARE_59_BIT 59 +#define FEATURE_SPARE_60_BIT 60 +#define FEATURE_SPARE_61_BIT 61 +#define FEATURE_SPARE_62_BIT 62 +#define FEATURE_SPARE_63_BIT 63 + +#define NUM_FEATURES 64 + +#define FEATURE_CCLK_CONTROLLER_MASK (1 << FEATURE_CCLK_CONTROLLER_BIT) +#define FEATURE_DATA_CALCULATION_MASK (1 << FEATURE_DATA_CALCULATION_BIT) +#define FEATURE_THERMAL_MASK (1 << FEATURE_THERMAL_BIT) +#define FEATURE_PLL_POWER_DOWN_MASK (1 << FEATURE_PLL_POWER_DOWN_BIT) +#define FEATURE_FCLK_DPM_MASK (1 << FEATURE_FCLK_DPM_BIT) +#define FEATURE_GFX_DPM_MASK (1 << FEATURE_GFX_DPM_BIT) +#define FEATURE_DS_GFXCLK_MASK (1 << FEATURE_DS_GFXCLK_BIT) +#define FEATURE_DS_SOCCLK_MASK (1 << FEATURE_DS_SOCCLK_BIT) +#define FEATURE_DS_LCLK_MASK (1 << FEATURE_DS_LCLK_BIT) +#define FEATURE_RM_MASK (1 << FEATURE_RM_BIT) +#define FEATURE_DS_SMNCLK_MASK (1 << FEATURE_DS_SMNCLK_BIT) +#define FEATURE_DS_MP1CLK_MASK (1 << FEATURE_DS_MP1CLK_BIT) +#define FEATURE_DS_MP0CLK_MASK (1 << FEATURE_DS_MP0CLK_BIT) +#define FEATURE_MGCG_MASK (1 << FEATURE_MGCG_BIT) +#define FEATURE_DS_FUSE_SRAM_MASK (1 << FEATURE_DS_FUSE_SRAM_BIT) +#define FEATURE_PROCHOT_MASK (1 << FEATURE_PROCHOT_BIT) +#define FEATURE_CPUOFF_MASK (1 << FEATURE_CPUOFF_BIT) +#define FEATURE_GFX_CKS_MASK (1 << FEATURE_GFX_CKS_BIT) +#define FEATURE_UMC_THROTTLE_MASK (1 << FEATURE_UMC_THROTTLE_BIT) +#define FEATURE_DF_THROTTLE_MASK (1 << FEATURE_DF_THROTTLE_BIT) +#define FEATURE_SOC_DPM_MASK (1 << FEATURE_SOC_DPM_BIT) + +typedef struct { + // MP1_EXT_SCRATCH0 + uint32_t SPARE1 : 4; + uint32_t SPARE2 : 4; + uint32_t SPARE3 : 4; + uint32_t CurrLevel_LCLK : 4; + uint32_t CurrLevel_MP0CLK : 4; + uint32_t CurrLevel_FCLK : 4; + uint32_t CurrLevel_SOCCLK : 4; + uint32_t CurrLevel_DCEFCLK : 4; + // MP1_EXT_SCRATCH1 + uint32_t SPARE4 : 4; + uint32_t SPARE5 : 4; + uint32_t SPARE6 : 4; + uint32_t TargLevel_LCLK : 4; + uint32_t TargLevel_MP0CLK : 4; + uint32_t TargLevel_FCLK : 4; + uint32_t TargLevel_SOCCLK : 4; + uint32_t TargLevel_DCEFCLK : 4; + // MP1_EXT_SCRATCH2 + uint32_t CurrLevel_SHUBCLK : 4; + uint32_t TargLevel_SHUBCLK : 4; + uint32_t Reserved : 24; + // MP1_EXT_SCRATCH3-4 + uint32_t Reserved2[2]; + // MP1_EXT_SCRATCH5 + uint32_t FeatureStatus[NUM_FEATURES / 32]; +} FwStatus_t; + +#pragma pack(pop) + +#endif diff --git a/drivers/gpu/drm/amd/pm/inc/smu_v11_8_ppsmc.h b/drivers/gpu/drm/amd/pm/inc/smu_v11_8_ppsmc.h new file mode 100644 index 000000000000..6e6088760b18 --- /dev/null +++ b/drivers/gpu/drm/amd/pm/inc/smu_v11_8_ppsmc.h @@ -0,0 +1,70 @@ +/* + * Copyright 2021 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#ifndef SMU_11_8_0_PPSMC_H +#define SMU_11_8_0_PPSMC_H + +// SMU Response Codes: +#define PPSMC_Result_OK 0x1 +#define PPSMC_Result_Failed 0xFF +#define PPSMC_Result_UnknownCmd 0xFE +#define PPSMC_Result_CmdRejectedPrereq 0xFD +#define PPSMC_Result_CmdRejectedBusy 0xFC + +// Message Definitions: +#define PPSMC_MSG_TestMessage 0x1 +#define PPSMC_MSG_GetSmuVersion 0x2 +#define PPSMC_MSG_GetDriverIfVersion 0x3 +#define PPSMC_MSG_SetDriverTableDramAddrHigh 0x4 +#define PPSMC_MSG_SetDriverTableDramAddrLow 0x5 +#define PPSMC_MSG_TransferTableSmu2Dram 0x6 +#define PPSMC_MSG_TransferTableDram2Smu 0x7 +#define PPSMC_MSG_Rsvd1 0xA +#define PPSMC_MSG_RequestCorePstate 0xB +#define PPSMC_MSG_QueryCorePstate 0xC +#define PPSMC_MSG_Rsvd2 0xD +#define PPSMC_MSG_RequestGfxclk 0xE +#define PPSMC_MSG_QueryGfxclk 0xF +#define PPSMC_MSG_QueryVddcrSocClock 0x11 +#define PPSMC_MSG_QueryDfPstate 0x13 +#define PPSMC_MSG_Rsvd3 0x14 +#define PPSMC_MSG_ConfigureS3PwrOffRegisterAddressHigh 0x16 +#define PPSMC_MSG_ConfigureS3PwrOffRegisterAddressLow 0x17 +#define PPSMC_MSG_RequestActiveWgp 0x18 +#define PPSMC_MSG_SetMinDeepSleepGfxclkFreq 0x19 +#define PPSMC_MSG_SetMaxDeepSleepDfllGfxDiv 0x1A +#define PPSMC_MSG_StartTelemetryReporting 0x1B +#define PPSMC_MSG_StopTelemetryReporting 0x1C +#define PPSMC_MSG_ClearTelemetryMax 0x1D +#define PPSMC_MSG_QueryActiveWgp 0x1E +#define PPSMC_MSG_SetCoreEnableMask 0x2C +#define PPSMC_MSG_InitiateGcRsmuSoftReset 0x2E +#define PPSMC_MSG_GfxCacWeightOperation 0x2F +#define PPSMC_MSG_L3CacWeightOperation 0x30 +#define PPSMC_MSG_PackCoreCacWeight 0x31 +#define PPSMC_MSG_SetDriverTableVMID 0x34 +#define PPSMC_MSG_SetSoftMinCclk 0x35 +#define PPSMC_MSG_SetSoftMaxCclk 0x36 +#define PPSMC_Message_Count 0x37 + +#endif diff --git a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c index ebe672142808..71afc2d20b12 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c +++ b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c @@ -36,6 +36,7 @@ #include "vangogh_ppt.h" #include "aldebaran_ppt.h" #include "yellow_carp_ppt.h" +#include "cyan_skillfish_ppt.h" #include "amd_pcie.h" /* @@ -416,6 +417,15 @@ static void smu_restore_dpm_user_profile(struct smu_context *smu) } } + /* Restore user customized OD settings */ + if (smu->user_dpm_profile.user_od) { + if (smu->ppt_funcs->restore_user_od_settings) { + ret = smu->ppt_funcs->restore_user_od_settings(smu); + if (ret) + dev_err(smu->adev->dev, "Failed to upload customized OD settings\n"); + } + } + /* Disable restore flag */ smu->user_dpm_profile.flags &= ~SMU_DPM_USER_PROFILE_RESTORE; } @@ -588,6 +598,9 @@ static int smu_set_funcs(struct amdgpu_device *adev) case CHIP_YELLOW_CARP: yellow_carp_set_ppt_funcs(smu); break; + case CHIP_CYAN_SKILLFISH: + cyan_skillfish_set_ppt_funcs(smu); + break; default: return -EINVAL; } diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/Makefile b/drivers/gpu/drm/amd/pm/swsmu/smu11/Makefile index 0138c982dfd3..f9b2e16f6431 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/Makefile +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/Makefile @@ -27,6 +27,7 @@ SMU11_MGR = arcturus_ppt.o \ navi10_ppt.o \ sienna_cichlid_ppt.o \ vangogh_ppt.o \ + cyan_skillfish_ppt.o \ smu_v11_0.o AMD_SWSMU_SMU11MGR = $(addprefix $(AMD_SWSMU_PATH)/smu11/,$(SMU11_MGR)) diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c index 094df6f87cfc..6ec8492f71f5 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c @@ -1936,197 +1936,77 @@ static int arcturus_dpm_set_vcn_enable(struct smu_context *smu, bool enable) return ret; } -static void arcturus_fill_i2c_req(SwI2cRequest_t *req, bool write, - uint8_t address, uint32_t numbytes, - uint8_t *data) -{ - int i; - - req->I2CcontrollerPort = 0; - req->I2CSpeed = 2; - req->SlaveAddress = address; - req->NumCmds = numbytes; - - for (i = 0; i < numbytes; i++) { - SwI2cCmd_t *cmd = &req->SwI2cCmds[i]; - - /* First 2 bytes are always write for lower 2b EEPROM address */ - if (i < 2) - cmd->Cmd = 1; - else - cmd->Cmd = write; - - - /* Add RESTART for read after address filled */ - cmd->CmdConfig |= (i == 2 && !write) ? CMDCONFIG_RESTART_MASK : 0; - - /* Add STOP in the end */ - cmd->CmdConfig |= (i == (numbytes - 1)) ? CMDCONFIG_STOP_MASK : 0; - - /* Fill with data regardless if read or write to simplify code */ - cmd->RegisterAddr = data[i]; - } -} - -static int arcturus_i2c_read_data(struct i2c_adapter *control, - uint8_t address, - uint8_t *data, - uint32_t numbytes) +static int arcturus_i2c_xfer(struct i2c_adapter *i2c_adap, + struct i2c_msg *msg, int num_msgs) { - uint32_t i, ret = 0; - SwI2cRequest_t req; - struct amdgpu_device *adev = to_amdgpu_device(control); + struct amdgpu_device *adev = to_amdgpu_device(i2c_adap); struct smu_table_context *smu_table = &adev->smu.smu_table; struct smu_table *table = &smu_table->driver_table; + SwI2cRequest_t *req, *res = (SwI2cRequest_t *)table->cpu_addr; + int i, j, r, c; + u16 dir; - if (numbytes > MAX_SW_I2C_COMMANDS) { - dev_err(adev->dev, "numbytes requested %d is over max allowed %d\n", - numbytes, MAX_SW_I2C_COMMANDS); - return -EINVAL; - } - - memset(&req, 0, sizeof(req)); - arcturus_fill_i2c_req(&req, false, address, numbytes, data); - - mutex_lock(&adev->smu.mutex); - /* Now read data starting with that address */ - ret = smu_cmn_update_table(&adev->smu, SMU_TABLE_I2C_COMMANDS, 0, &req, - true); - mutex_unlock(&adev->smu.mutex); - - if (!ret) { - SwI2cRequest_t *res = (SwI2cRequest_t *)table->cpu_addr; - - /* Assume SMU fills res.SwI2cCmds[i].Data with read bytes */ - for (i = 0; i < numbytes; i++) - data[i] = res->SwI2cCmds[i].Data; - - dev_dbg(adev->dev, "arcturus_i2c_read_data, address = %x, bytes = %d, data :", - (uint16_t)address, numbytes); - - print_hex_dump(KERN_DEBUG, "data: ", DUMP_PREFIX_NONE, - 8, 1, data, numbytes, false); - } else - dev_err(adev->dev, "arcturus_i2c_read_data - error occurred :%x", ret); + req = kzalloc(sizeof(*req), GFP_KERNEL); + if (!req) + return -ENOMEM; - return ret; -} + req->I2CcontrollerPort = 0; + req->I2CSpeed = I2C_SPEED_FAST_400K; + req->SlaveAddress = msg[0].addr << 1; /* wants an 8-bit address */ + dir = msg[0].flags & I2C_M_RD; + + for (c = i = 0; i < num_msgs; i++) { + for (j = 0; j < msg[i].len; j++, c++) { + SwI2cCmd_t *cmd = &req->SwI2cCmds[c]; + + if (!(msg[i].flags & I2C_M_RD)) { + /* write */ + cmd->Cmd = I2C_CMD_WRITE; + cmd->RegisterAddr = msg[i].buf[j]; + } -static int arcturus_i2c_write_data(struct i2c_adapter *control, - uint8_t address, - uint8_t *data, - uint32_t numbytes) -{ - uint32_t ret; - SwI2cRequest_t req; - struct amdgpu_device *adev = to_amdgpu_device(control); + if ((dir ^ msg[i].flags) & I2C_M_RD) { + /* The direction changes. + */ + dir = msg[i].flags & I2C_M_RD; + cmd->CmdConfig |= CMDCONFIG_RESTART_MASK; + } - if (numbytes > MAX_SW_I2C_COMMANDS) { - dev_err(adev->dev, "numbytes requested %d is over max allowed %d\n", - numbytes, MAX_SW_I2C_COMMANDS); - return -EINVAL; + req->NumCmds++; + + /* + * Insert STOP if we are at the last byte of either last + * message for the transaction or the client explicitly + * requires a STOP at this particular message. + */ + if ((j == msg[i].len - 1) && + ((i == num_msgs - 1) || (msg[i].flags & I2C_M_STOP))) { + cmd->CmdConfig &= ~CMDCONFIG_RESTART_MASK; + cmd->CmdConfig |= CMDCONFIG_STOP_MASK; + } + } } - - memset(&req, 0, sizeof(req)); - arcturus_fill_i2c_req(&req, true, address, numbytes, data); - mutex_lock(&adev->smu.mutex); - ret = smu_cmn_update_table(&adev->smu, SMU_TABLE_I2C_COMMANDS, 0, &req, true); + r = smu_cmn_update_table(&adev->smu, SMU_TABLE_I2C_COMMANDS, 0, req, true); mutex_unlock(&adev->smu.mutex); + if (r) + goto fail; - if (!ret) { - dev_dbg(adev->dev, "arcturus_i2c_write(), address = %x, bytes = %d , data: ", - (uint16_t)address, numbytes); - - print_hex_dump(KERN_DEBUG, "data: ", DUMP_PREFIX_NONE, - 8, 1, data, numbytes, false); - /* - * According to EEPROM spec there is a MAX of 10 ms required for - * EEPROM to flush internal RX buffer after STOP was issued at the - * end of write transaction. During this time the EEPROM will not be - * responsive to any more commands - so wait a bit more. - */ - msleep(10); - - } else - dev_err(adev->dev, "arcturus_i2c_write- error occurred :%x", ret); - - return ret; -} - -static int arcturus_i2c_xfer(struct i2c_adapter *i2c_adap, - struct i2c_msg *msgs, int num) -{ - uint32_t i, j, ret, data_size, data_chunk_size, next_eeprom_addr = 0; - uint8_t *data_ptr, data_chunk[MAX_SW_I2C_COMMANDS] = { 0 }; - - for (i = 0; i < num; i++) { - /* - * SMU interface allows at most MAX_SW_I2C_COMMANDS bytes of data at - * once and hence the data needs to be spliced into chunks and sent each - * chunk separately - */ - data_size = msgs[i].len - 2; - data_chunk_size = MAX_SW_I2C_COMMANDS - 2; - next_eeprom_addr = (msgs[i].buf[0] << 8 & 0xff00) | (msgs[i].buf[1] & 0xff); - data_ptr = msgs[i].buf + 2; - - for (j = 0; j < data_size / data_chunk_size; j++) { - /* Insert the EEPROM dest addess, bits 0-15 */ - data_chunk[0] = ((next_eeprom_addr >> 8) & 0xff); - data_chunk[1] = (next_eeprom_addr & 0xff); - - if (msgs[i].flags & I2C_M_RD) { - ret = arcturus_i2c_read_data(i2c_adap, - (uint8_t)msgs[i].addr, - data_chunk, MAX_SW_I2C_COMMANDS); - - memcpy(data_ptr, data_chunk + 2, data_chunk_size); - } else { - - memcpy(data_chunk + 2, data_ptr, data_chunk_size); - - ret = arcturus_i2c_write_data(i2c_adap, - (uint8_t)msgs[i].addr, - data_chunk, MAX_SW_I2C_COMMANDS); - } - - if (ret) { - num = -EIO; - goto fail; - } - - next_eeprom_addr += data_chunk_size; - data_ptr += data_chunk_size; + for (c = i = 0; i < num_msgs; i++) { + if (!(msg[i].flags & I2C_M_RD)) { + c += msg[i].len; + continue; } + for (j = 0; j < msg[i].len; j++, c++) { + SwI2cCmd_t *cmd = &res->SwI2cCmds[c]; - if (data_size % data_chunk_size) { - data_chunk[0] = ((next_eeprom_addr >> 8) & 0xff); - data_chunk[1] = (next_eeprom_addr & 0xff); - - if (msgs[i].flags & I2C_M_RD) { - ret = arcturus_i2c_read_data(i2c_adap, - (uint8_t)msgs[i].addr, - data_chunk, (data_size % data_chunk_size) + 2); - - memcpy(data_ptr, data_chunk + 2, data_size % data_chunk_size); - } else { - memcpy(data_chunk + 2, data_ptr, data_size % data_chunk_size); - - ret = arcturus_i2c_write_data(i2c_adap, - (uint8_t)msgs[i].addr, - data_chunk, (data_size % data_chunk_size) + 2); - } - - if (ret) { - num = -EIO; - goto fail; - } + msg[i].buf[j] = cmd->Data; } } - + r = num_msgs; fail: - return num; + kfree(req); + return r; } static u32 arcturus_i2c_func(struct i2c_adapter *adap) @@ -2140,15 +2020,25 @@ static const struct i2c_algorithm arcturus_i2c_algo = { .functionality = arcturus_i2c_func, }; + +static const struct i2c_adapter_quirks arcturus_i2c_control_quirks = { + .flags = I2C_AQ_COMB | I2C_AQ_COMB_SAME_ADDR | I2C_AQ_NO_ZERO_LEN, + .max_read_len = MAX_SW_I2C_COMMANDS, + .max_write_len = MAX_SW_I2C_COMMANDS, + .max_comb_1st_msg_len = 2, + .max_comb_2nd_msg_len = MAX_SW_I2C_COMMANDS - 2, +}; + static int arcturus_i2c_control_init(struct smu_context *smu, struct i2c_adapter *control) { struct amdgpu_device *adev = to_amdgpu_device(control); int res; control->owner = THIS_MODULE; - control->class = I2C_CLASS_SPD; + control->class = I2C_CLASS_HWMON; control->dev.parent = &adev->pdev->dev; control->algo = &arcturus_i2c_algo; + control->quirks = &arcturus_i2c_control_quirks; snprintf(control->name, sizeof(control->name), "AMDGPU SMU"); res = i2c_add_adapter(control); @@ -2288,7 +2178,9 @@ static void arcturus_log_thermal_throttling_event(struct smu_context *smu) dev_warn(adev->dev, "WARN: GPU thermal throttling temperature reached, expect performance decrease. %s.\n", log_buf); - kgd2kfd_smi_event_throttle(smu->adev->kfd.dev, throttler_status); + kgd2kfd_smi_event_throttle(smu->adev->kfd.dev, + smu_cmn_get_indep_throttler_status(throttler_status, + arcturus_throttler_map)); } static uint16_t arcturus_get_current_pcie_link_speed(struct smu_context *smu) diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/cyan_skillfish_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/cyan_skillfish_ppt.c new file mode 100644 index 000000000000..b05f9541accc --- /dev/null +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/cyan_skillfish_ppt.c @@ -0,0 +1,76 @@ +/* + * Copyright 2021 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#define SWSMU_CODE_LAYER_L2 + +#include "amdgpu.h" +#include "amdgpu_smu.h" +#include "smu_v11_0.h" +#include "smu11_driver_if_cyan_skillfish.h" +#include "cyan_skillfish_ppt.h" +#include "smu_v11_8_ppsmc.h" +#include "smu_v11_8_pmfw.h" +#include "smu_cmn.h" +#include "soc15_common.h" + +/* + * DO NOT use these for err/warn/info/debug messages. + * Use dev_err, dev_warn, dev_info and dev_dbg instead. + * They are more MGPU friendly. + */ + +#undef pr_err +#undef pr_warn +#undef pr_info +#undef pr_debug + +static struct cmn2asic_msg_mapping cyan_skillfish_message_map[SMU_MSG_MAX_COUNT] = { + MSG_MAP(TestMessage, PPSMC_MSG_TestMessage, 0), + MSG_MAP(GetSmuVersion, PPSMC_MSG_GetSmuVersion, 0), + MSG_MAP(GetDriverIfVersion, PPSMC_MSG_GetDriverIfVersion, 0), + MSG_MAP(SetDriverDramAddrHigh, PPSMC_MSG_SetDriverTableDramAddrHigh, 0), + MSG_MAP(SetDriverDramAddrLow, PPSMC_MSG_SetDriverTableDramAddrLow, 0), + MSG_MAP(TransferTableSmu2Dram, PPSMC_MSG_TransferTableSmu2Dram, 0), + MSG_MAP(TransferTableDram2Smu, PPSMC_MSG_TransferTableDram2Smu, 0), +}; + +static const struct pptable_funcs cyan_skillfish_ppt_funcs = { + + .check_fw_status = smu_v11_0_check_fw_status, + .check_fw_version = smu_v11_0_check_fw_version, + .init_power = smu_v11_0_init_power, + .fini_power = smu_v11_0_fini_power, + .register_irq_handler = smu_v11_0_register_irq_handler, + .notify_memory_pool_location = smu_v11_0_notify_memory_pool_location, + .send_smc_msg_with_param = smu_cmn_send_smc_msg_with_param, + .send_smc_msg = smu_cmn_send_smc_msg, + .set_driver_table_location = smu_v11_0_set_driver_table_location, + .interrupt_work = smu_v11_0_interrupt_work, +}; + +void cyan_skillfish_set_ppt_funcs(struct smu_context *smu) +{ + smu->ppt_funcs = &cyan_skillfish_ppt_funcs; + smu->message_map = cyan_skillfish_message_map; + smu->is_apu = true; +} diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/cyan_skillfish_ppt.h b/drivers/gpu/drm/amd/pm/swsmu/smu11/cyan_skillfish_ppt.h new file mode 100644 index 000000000000..76cd7229e383 --- /dev/null +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/cyan_skillfish_ppt.h @@ -0,0 +1,29 @@ +/* + * Copyright 2021 Advanced Micro Devices, Inc. + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice shall be included in + * all copies or substantial portions of the Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR + * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, + * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR + * OTHER DEALINGS IN THE SOFTWARE. + * + */ + +#ifndef __CYAN_SKILLFISH_PPT_H__ +#define __CYAN_SKILLFISH_PPT_H__ + +extern void cyan_skillfish_set_ppt_funcs(struct smu_context *smu); + +#endif diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c index 1ba42b69ce74..d7722c229ddd 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c @@ -2294,41 +2294,52 @@ static int navi10_set_default_od_settings(struct smu_context *smu) (OverDriveTable_t *)smu->smu_table.overdrive_table; OverDriveTable_t *boot_od_table = (OverDriveTable_t *)smu->smu_table.boot_overdrive_table; + OverDriveTable_t *user_od_table = + (OverDriveTable_t *)smu->smu_table.user_overdrive_table; int ret = 0; - ret = smu_cmn_update_table(smu, SMU_TABLE_OVERDRIVE, 0, (void *)od_table, false); + /* + * For S3/S4/Runpm resume, no need to setup those overdrive tables again as + * - either they already have the default OD settings got during cold bootup + * - or they have some user customized OD settings which cannot be overwritten + */ + if (smu->adev->in_suspend) + return 0; + + ret = smu_cmn_update_table(smu, SMU_TABLE_OVERDRIVE, 0, (void *)boot_od_table, false); if (ret) { dev_err(smu->adev->dev, "Failed to get overdrive table!\n"); return ret; } - if (!od_table->GfxclkVolt1) { + if (!boot_od_table->GfxclkVolt1) { ret = navi10_overdrive_get_gfx_clk_base_voltage(smu, - &od_table->GfxclkVolt1, - od_table->GfxclkFreq1); + &boot_od_table->GfxclkVolt1, + boot_od_table->GfxclkFreq1); if (ret) return ret; } - if (!od_table->GfxclkVolt2) { + if (!boot_od_table->GfxclkVolt2) { ret = navi10_overdrive_get_gfx_clk_base_voltage(smu, - &od_table->GfxclkVolt2, - od_table->GfxclkFreq2); + &boot_od_table->GfxclkVolt2, + boot_od_table->GfxclkFreq2); if (ret) return ret; } - if (!od_table->GfxclkVolt3) { + if (!boot_od_table->GfxclkVolt3) { ret = navi10_overdrive_get_gfx_clk_base_voltage(smu, - &od_table->GfxclkVolt3, - od_table->GfxclkFreq3); + &boot_od_table->GfxclkVolt3, + boot_od_table->GfxclkFreq3); if (ret) return ret; } - memcpy(boot_od_table, od_table, sizeof(OverDriveTable_t)); + navi10_dump_od_table(smu, boot_od_table); - navi10_dump_od_table(smu, od_table); + memcpy(od_table, boot_od_table, sizeof(OverDriveTable_t)); + memcpy(user_od_table, boot_od_table, sizeof(OverDriveTable_t)); return 0; } @@ -2429,11 +2440,20 @@ static int navi10_od_edit_dpm_table(struct smu_context *smu, enum PP_OD_DPM_TABL memcpy(table_context->overdrive_table, table_context->boot_overdrive_table, sizeof(OverDriveTable_t)); break; case PP_OD_COMMIT_DPM_TABLE: - navi10_dump_od_table(smu, od_table); - ret = smu_cmn_update_table(smu, SMU_TABLE_OVERDRIVE, 0, (void *)od_table, true); - if (ret) { - dev_err(smu->adev->dev, "Failed to import overdrive table!\n"); - return ret; + if (memcmp(od_table, table_context->user_overdrive_table, sizeof(OverDriveTable_t))) { + navi10_dump_od_table(smu, od_table); + ret = smu_cmn_update_table(smu, SMU_TABLE_OVERDRIVE, 0, (void *)od_table, true); + if (ret) { + dev_err(smu->adev->dev, "Failed to import overdrive table!\n"); + return ret; + } + memcpy(table_context->user_overdrive_table, od_table, sizeof(OverDriveTable_t)); + smu->user_dpm_profile.user_od = true; + + if (!memcmp(table_context->user_overdrive_table, + table_context->boot_overdrive_table, + sizeof(OverDriveTable_t))) + smu->user_dpm_profile.user_od = false; } break; case PP_OD_EDIT_VDDC_CURVE: @@ -2735,6 +2755,122 @@ static ssize_t navi10_get_legacy_gpu_metrics(struct smu_context *smu, return sizeof(struct gpu_metrics_v1_3); } +static int navi10_i2c_xfer(struct i2c_adapter *i2c_adap, + struct i2c_msg *msg, int num_msgs) +{ + struct amdgpu_device *adev = to_amdgpu_device(i2c_adap); + struct smu_table_context *smu_table = &adev->smu.smu_table; + struct smu_table *table = &smu_table->driver_table; + SwI2cRequest_t *req, *res = (SwI2cRequest_t *)table->cpu_addr; + int i, j, r, c; + u16 dir; + + req = kzalloc(sizeof(*req), GFP_KERNEL); + if (!req) + return -ENOMEM; + + req->I2CcontrollerPort = 0; + req->I2CSpeed = I2C_SPEED_FAST_400K; + req->SlaveAddress = msg[0].addr << 1; /* wants an 8-bit address */ + dir = msg[0].flags & I2C_M_RD; + + for (c = i = 0; i < num_msgs; i++) { + for (j = 0; j < msg[i].len; j++, c++) { + SwI2cCmd_t *cmd = &req->SwI2cCmds[c]; + + if (!(msg[i].flags & I2C_M_RD)) { + /* write */ + cmd->Cmd = I2C_CMD_WRITE; + cmd->RegisterAddr = msg[i].buf[j]; + } + + if ((dir ^ msg[i].flags) & I2C_M_RD) { + /* The direction changes. + */ + dir = msg[i].flags & I2C_M_RD; + cmd->CmdConfig |= CMDCONFIG_RESTART_MASK; + } + + req->NumCmds++; + + /* + * Insert STOP if we are at the last byte of either last + * message for the transaction or the client explicitly + * requires a STOP at this particular message. + */ + if ((j == msg[i].len - 1) && + ((i == num_msgs - 1) || (msg[i].flags & I2C_M_STOP))) { + cmd->CmdConfig &= ~CMDCONFIG_RESTART_MASK; + cmd->CmdConfig |= CMDCONFIG_STOP_MASK; + } + } + } + mutex_lock(&adev->smu.mutex); + r = smu_cmn_update_table(&adev->smu, SMU_TABLE_I2C_COMMANDS, 0, req, true); + mutex_unlock(&adev->smu.mutex); + if (r) + goto fail; + + for (c = i = 0; i < num_msgs; i++) { + if (!(msg[i].flags & I2C_M_RD)) { + c += msg[i].len; + continue; + } + for (j = 0; j < msg[i].len; j++, c++) { + SwI2cCmd_t *cmd = &res->SwI2cCmds[c]; + + msg[i].buf[j] = cmd->Data; + } + } + r = num_msgs; +fail: + kfree(req); + return r; +} + +static u32 navi10_i2c_func(struct i2c_adapter *adap) +{ + return I2C_FUNC_I2C | I2C_FUNC_SMBUS_EMUL; +} + + +static const struct i2c_algorithm navi10_i2c_algo = { + .master_xfer = navi10_i2c_xfer, + .functionality = navi10_i2c_func, +}; + +static const struct i2c_adapter_quirks navi10_i2c_control_quirks = { + .flags = I2C_AQ_COMB | I2C_AQ_COMB_SAME_ADDR | I2C_AQ_NO_ZERO_LEN, + .max_read_len = MAX_SW_I2C_COMMANDS, + .max_write_len = MAX_SW_I2C_COMMANDS, + .max_comb_1st_msg_len = 2, + .max_comb_2nd_msg_len = MAX_SW_I2C_COMMANDS - 2, +}; + +static int navi10_i2c_control_init(struct smu_context *smu, struct i2c_adapter *control) +{ + struct amdgpu_device *adev = to_amdgpu_device(control); + int res; + + control->owner = THIS_MODULE; + control->class = I2C_CLASS_HWMON; + control->dev.parent = &adev->pdev->dev; + control->algo = &navi10_i2c_algo; + snprintf(control->name, sizeof(control->name), "AMDGPU SMU"); + control->quirks = &navi10_i2c_control_quirks; + + res = i2c_add_adapter(control); + if (res) + DRM_ERROR("Failed to register hw i2c, err: %d\n", res); + + return res; +} + +static void navi10_i2c_control_fini(struct smu_context *smu, struct i2c_adapter *control) +{ + i2c_del_adapter(control); +} + static ssize_t navi10_get_gpu_metrics(struct smu_context *smu, void **table) { @@ -3078,6 +3214,8 @@ static const struct pptable_funcs navi10_ppt_funcs = { .set_default_dpm_table = navi10_set_default_dpm_table, .dpm_set_vcn_enable = navi10_dpm_set_vcn_enable, .dpm_set_jpeg_enable = navi10_dpm_set_jpeg_enable, + .i2c_init = navi10_i2c_control_init, + .i2c_fini = navi10_i2c_control_fini, .print_clk_levels = navi10_print_clk_levels, .force_clk_levels = navi10_force_clk_levels, .populate_umd_state_clk = navi10_populate_umd_state_clk, @@ -3144,6 +3282,7 @@ static const struct pptable_funcs navi10_ppt_funcs = { .set_soft_freq_limited_range = smu_v11_0_set_soft_freq_limited_range, .set_default_od_settings = navi10_set_default_od_settings, .od_edit_dpm_table = navi10_od_edit_dpm_table, + .restore_user_od_settings = smu_v11_0_restore_user_od_settings, .run_btc = navi10_run_btc, .set_power_source = smu_v11_0_set_power_source, .get_pp_feature_mask = smu_cmn_get_pp_feature_mask, diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c index c751f717a0da..90e40aacb8f6 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c @@ -489,6 +489,26 @@ err0_out: return -ENOMEM; } +static uint32_t sienna_cichlid_get_throttler_status_locked(struct smu_context *smu) +{ + struct smu_table_context *smu_table= &smu->smu_table; + SmuMetricsExternal_t *metrics_ext = + (SmuMetricsExternal_t *)(smu_table->metrics_table); + uint32_t throttler_status = 0; + int i; + + if ((smu->adev->asic_type == CHIP_SIENNA_CICHLID) && + (smu->smc_fw_version >= 0x3A4300)) { + for (i = 0; i < THROTTLER_COUNT; i++) + throttler_status |= + (metrics_ext->SmuMetrics_V2.ThrottlingPercentage[i] ? 1U << i : 0); + } else { + throttler_status = metrics_ext->SmuMetrics.ThrottlerStatus; + } + + return throttler_status; +} + static int sienna_cichlid_get_smu_metrics_data(struct smu_context *smu, MetricsMember_t member, uint32_t *value) @@ -496,6 +516,11 @@ static int sienna_cichlid_get_smu_metrics_data(struct smu_context *smu, struct smu_table_context *smu_table= &smu->smu_table; SmuMetrics_t *metrics = &(((SmuMetricsExternal_t *)(smu_table->metrics_table))->SmuMetrics); + SmuMetrics_V2_t *metrics_v2 = + &(((SmuMetricsExternal_t *)(smu_table->metrics_table))->SmuMetrics_V2); + bool use_metrics_v2 = ((smu->adev->asic_type == CHIP_SIENNA_CICHLID) && + (smu->smc_fw_version >= 0x3A4300)) ? true : false; + uint16_t average_gfx_activity; int ret = 0; mutex_lock(&smu->metrics_lock); @@ -510,78 +535,96 @@ static int sienna_cichlid_get_smu_metrics_data(struct smu_context *smu, switch (member) { case METRICS_CURR_GFXCLK: - *value = metrics->CurrClock[PPCLK_GFXCLK]; + *value = use_metrics_v2 ? metrics_v2->CurrClock[PPCLK_GFXCLK] : + metrics->CurrClock[PPCLK_GFXCLK]; break; case METRICS_CURR_SOCCLK: - *value = metrics->CurrClock[PPCLK_SOCCLK]; + *value = use_metrics_v2 ? metrics_v2->CurrClock[PPCLK_SOCCLK] : + metrics->CurrClock[PPCLK_SOCCLK]; break; case METRICS_CURR_UCLK: - *value = metrics->CurrClock[PPCLK_UCLK]; + *value = use_metrics_v2 ? metrics_v2->CurrClock[PPCLK_UCLK] : + metrics->CurrClock[PPCLK_UCLK]; break; case METRICS_CURR_VCLK: - *value = metrics->CurrClock[PPCLK_VCLK_0]; + *value = use_metrics_v2 ? metrics_v2->CurrClock[PPCLK_VCLK_0] : + metrics->CurrClock[PPCLK_VCLK_0]; break; case METRICS_CURR_VCLK1: - *value = metrics->CurrClock[PPCLK_VCLK_1]; + *value = use_metrics_v2 ? metrics_v2->CurrClock[PPCLK_VCLK_1] : + metrics->CurrClock[PPCLK_VCLK_1]; break; case METRICS_CURR_DCLK: - *value = metrics->CurrClock[PPCLK_DCLK_0]; + *value = use_metrics_v2 ? metrics_v2->CurrClock[PPCLK_DCLK_0] : + metrics->CurrClock[PPCLK_DCLK_0]; break; case METRICS_CURR_DCLK1: - *value = metrics->CurrClock[PPCLK_DCLK_1]; + *value = use_metrics_v2 ? metrics_v2->CurrClock[PPCLK_DCLK_1] : + metrics->CurrClock[PPCLK_DCLK_1]; break; case METRICS_CURR_DCEFCLK: - *value = metrics->CurrClock[PPCLK_DCEFCLK]; + *value = use_metrics_v2 ? metrics_v2->CurrClock[PPCLK_DCEFCLK] : + metrics->CurrClock[PPCLK_DCEFCLK]; break; case METRICS_CURR_FCLK: - *value = metrics->CurrClock[PPCLK_FCLK]; + *value = use_metrics_v2 ? metrics_v2->CurrClock[PPCLK_FCLK] : + metrics->CurrClock[PPCLK_FCLK]; break; case METRICS_AVERAGE_GFXCLK: - if (metrics->AverageGfxActivity <= SMU_11_0_7_GFX_BUSY_THRESHOLD) - *value = metrics->AverageGfxclkFrequencyPostDs; + average_gfx_activity = use_metrics_v2 ? metrics_v2->AverageGfxActivity : + metrics->AverageGfxActivity; + if (average_gfx_activity <= SMU_11_0_7_GFX_BUSY_THRESHOLD) + *value = use_metrics_v2 ? metrics_v2->AverageGfxclkFrequencyPostDs : + metrics->AverageGfxclkFrequencyPostDs; else - *value = metrics->AverageGfxclkFrequencyPreDs; + *value = use_metrics_v2 ? metrics_v2->AverageGfxclkFrequencyPreDs : + metrics->AverageGfxclkFrequencyPreDs; break; case METRICS_AVERAGE_FCLK: - *value = metrics->AverageFclkFrequencyPostDs; + *value = use_metrics_v2 ? metrics_v2->AverageFclkFrequencyPostDs : + metrics->AverageFclkFrequencyPostDs; break; case METRICS_AVERAGE_UCLK: - *value = metrics->AverageUclkFrequencyPostDs; + *value = use_metrics_v2 ? metrics_v2->AverageUclkFrequencyPostDs : + metrics->AverageUclkFrequencyPostDs; break; case METRICS_AVERAGE_GFXACTIVITY: - *value = metrics->AverageGfxActivity; + *value = use_metrics_v2 ? metrics_v2->AverageGfxActivity : + metrics->AverageGfxActivity; break; case METRICS_AVERAGE_MEMACTIVITY: - *value = metrics->AverageUclkActivity; + *value = use_metrics_v2 ? metrics_v2->AverageUclkActivity : + metrics->AverageUclkActivity; break; case METRICS_AVERAGE_SOCKETPOWER: - *value = metrics->AverageSocketPower << 8; + *value = use_metrics_v2 ? metrics_v2->AverageSocketPower << 8 : + metrics->AverageSocketPower << 8; break; case METRICS_TEMPERATURE_EDGE: - *value = metrics->TemperatureEdge * + *value = (use_metrics_v2 ? metrics_v2->TemperatureEdge : metrics->TemperatureEdge) * SMU_TEMPERATURE_UNITS_PER_CENTIGRADES; break; case METRICS_TEMPERATURE_HOTSPOT: - *value = metrics->TemperatureHotspot * + *value = (use_metrics_v2 ? metrics_v2->TemperatureHotspot : metrics->TemperatureHotspot) * SMU_TEMPERATURE_UNITS_PER_CENTIGRADES; break; case METRICS_TEMPERATURE_MEM: - *value = metrics->TemperatureMem * + *value = (use_metrics_v2 ? metrics_v2->TemperatureMem : metrics->TemperatureMem) * SMU_TEMPERATURE_UNITS_PER_CENTIGRADES; break; case METRICS_TEMPERATURE_VRGFX: - *value = metrics->TemperatureVrGfx * + *value = (use_metrics_v2 ? metrics_v2->TemperatureVrGfx : metrics->TemperatureVrGfx) * SMU_TEMPERATURE_UNITS_PER_CENTIGRADES; break; case METRICS_TEMPERATURE_VRSOC: - *value = metrics->TemperatureVrSoc * + *value = (use_metrics_v2 ? metrics_v2->TemperatureVrSoc : metrics->TemperatureVrSoc) * SMU_TEMPERATURE_UNITS_PER_CENTIGRADES; break; case METRICS_THROTTLER_STATUS: - *value = metrics->ThrottlerStatus; + *value = sienna_cichlid_get_throttler_status_locked(smu); break; case METRICS_CURR_FANSPEED: - *value = metrics->CurrFanSpeed; + *value = use_metrics_v2 ? metrics_v2->CurrFanSpeed : metrics->CurrFanSpeed; break; default: *value = UINT_MAX; @@ -1911,18 +1954,29 @@ static int sienna_cichlid_set_default_od_settings(struct smu_context *smu) (OverDriveTable_t *)smu->smu_table.overdrive_table; OverDriveTable_t *boot_od_table = (OverDriveTable_t *)smu->smu_table.boot_overdrive_table; + OverDriveTable_t *user_od_table = + (OverDriveTable_t *)smu->smu_table.user_overdrive_table; int ret = 0; + /* + * For S3/S4/Runpm resume, no need to setup those overdrive tables again as + * - either they already have the default OD settings got during cold bootup + * - or they have some user customized OD settings which cannot be overwritten + */ + if (smu->adev->in_suspend) + return 0; + ret = smu_cmn_update_table(smu, SMU_TABLE_OVERDRIVE, - 0, (void *)od_table, false); + 0, (void *)boot_od_table, false); if (ret) { dev_err(smu->adev->dev, "Failed to get overdrive table!\n"); return ret; } - memcpy(boot_od_table, od_table, sizeof(OverDriveTable_t)); + sienna_cichlid_dump_od_table(smu, boot_od_table); - sienna_cichlid_dump_od_table(smu, od_table); + memcpy(od_table, boot_od_table, sizeof(OverDriveTable_t)); + memcpy(user_od_table, boot_od_table, sizeof(OverDriveTable_t)); return 0; } @@ -2085,13 +2139,20 @@ static int sienna_cichlid_od_edit_dpm_table(struct smu_context *smu, fallthrough; case PP_OD_COMMIT_DPM_TABLE: - sienna_cichlid_dump_od_table(smu, od_table); + if (memcmp(od_table, table_context->user_overdrive_table, sizeof(OverDriveTable_t))) { + sienna_cichlid_dump_od_table(smu, od_table); + ret = smu_cmn_update_table(smu, SMU_TABLE_OVERDRIVE, 0, (void *)od_table, true); + if (ret) { + dev_err(smu->adev->dev, "Failed to import overdrive table!\n"); + return ret; + } + memcpy(table_context->user_overdrive_table, od_table, sizeof(OverDriveTable_t)); + smu->user_dpm_profile.user_od = true; - ret = smu_cmn_update_table(smu, SMU_TABLE_OVERDRIVE, - 0, (void *)od_table, true); - if (ret) { - dev_err(smu->adev->dev, "Failed to import overdrive table!\n"); - return ret; + if (!memcmp(table_context->user_overdrive_table, + table_context->boot_overdrive_table, + sizeof(OverDriveTable_t))) + smu->user_dpm_profile.user_od = false; } break; @@ -3442,197 +3503,77 @@ static void sienna_cichlid_dump_pptable(struct smu_context *smu) dev_info(smu->adev->dev, "MmHubPadding[7] = 0x%x\n", pptable->MmHubPadding[7]); } -static void sienna_cichlid_fill_i2c_req(SwI2cRequest_t *req, bool write, - uint8_t address, uint32_t numbytes, - uint8_t *data) -{ - int i; - - req->I2CcontrollerPort = 1; - req->I2CSpeed = 2; - req->SlaveAddress = address; - req->NumCmds = numbytes; - - for (i = 0; i < numbytes; i++) { - SwI2cCmd_t *cmd = &req->SwI2cCmds[i]; - - /* First 2 bytes are always write for lower 2b EEPROM address */ - if (i < 2) - cmd->CmdConfig = CMDCONFIG_READWRITE_MASK; - else - cmd->CmdConfig = write ? CMDCONFIG_READWRITE_MASK : 0; - - - /* Add RESTART for read after address filled */ - cmd->CmdConfig |= (i == 2 && !write) ? CMDCONFIG_RESTART_MASK : 0; - - /* Add STOP in the end */ - cmd->CmdConfig |= (i == (numbytes - 1)) ? CMDCONFIG_STOP_MASK : 0; - - /* Fill with data regardless if read or write to simplify code */ - cmd->ReadWriteData = data[i]; - } -} - -static int sienna_cichlid_i2c_read_data(struct i2c_adapter *control, - uint8_t address, - uint8_t *data, - uint32_t numbytes) +static int sienna_cichlid_i2c_xfer(struct i2c_adapter *i2c_adap, + struct i2c_msg *msg, int num_msgs) { - uint32_t i, ret = 0; - SwI2cRequest_t req; - struct amdgpu_device *adev = to_amdgpu_device(control); + struct amdgpu_device *adev = to_amdgpu_device(i2c_adap); struct smu_table_context *smu_table = &adev->smu.smu_table; struct smu_table *table = &smu_table->driver_table; + SwI2cRequest_t *req, *res = (SwI2cRequest_t *)table->cpu_addr; + int i, j, r, c; + u16 dir; - if (numbytes > MAX_SW_I2C_COMMANDS) { - dev_err(adev->dev, "numbytes requested %d is over max allowed %d\n", - numbytes, MAX_SW_I2C_COMMANDS); - return -EINVAL; - } - - memset(&req, 0, sizeof(req)); - sienna_cichlid_fill_i2c_req(&req, false, address, numbytes, data); - - mutex_lock(&adev->smu.mutex); - /* Now read data starting with that address */ - ret = smu_cmn_update_table(&adev->smu, SMU_TABLE_I2C_COMMANDS, 0, &req, - true); - mutex_unlock(&adev->smu.mutex); - - if (!ret) { - SwI2cRequest_t *res = (SwI2cRequest_t *)table->cpu_addr; - - /* Assume SMU fills res.SwI2cCmds[i].Data with read bytes */ - for (i = 0; i < numbytes; i++) - data[i] = res->SwI2cCmds[i].ReadWriteData; - - dev_dbg(adev->dev, "sienna_cichlid_i2c_read_data, address = %x, bytes = %d, data :", - (uint16_t)address, numbytes); - - print_hex_dump(KERN_DEBUG, "data: ", DUMP_PREFIX_NONE, - 8, 1, data, numbytes, false); - } else - dev_err(adev->dev, "sienna_cichlid_i2c_read_data - error occurred :%x", ret); + req = kzalloc(sizeof(*req), GFP_KERNEL); + if (!req) + return -ENOMEM; - return ret; -} + req->I2CcontrollerPort = 1; + req->I2CSpeed = I2C_SPEED_FAST_400K; + req->SlaveAddress = msg[0].addr << 1; /* wants an 8-bit address */ + dir = msg[0].flags & I2C_M_RD; + + for (c = i = 0; i < num_msgs; i++) { + for (j = 0; j < msg[i].len; j++, c++) { + SwI2cCmd_t *cmd = &req->SwI2cCmds[c]; + + if (!(msg[i].flags & I2C_M_RD)) { + /* write */ + cmd->CmdConfig |= CMDCONFIG_READWRITE_MASK; + cmd->ReadWriteData = msg[i].buf[j]; + } -static int sienna_cichlid_i2c_write_data(struct i2c_adapter *control, - uint8_t address, - uint8_t *data, - uint32_t numbytes) -{ - uint32_t ret; - SwI2cRequest_t req; - struct amdgpu_device *adev = to_amdgpu_device(control); + if ((dir ^ msg[i].flags) & I2C_M_RD) { + /* The direction changes. + */ + dir = msg[i].flags & I2C_M_RD; + cmd->CmdConfig |= CMDCONFIG_RESTART_MASK; + } - if (numbytes > MAX_SW_I2C_COMMANDS) { - dev_err(adev->dev, "numbytes requested %d is over max allowed %d\n", - numbytes, MAX_SW_I2C_COMMANDS); - return -EINVAL; + req->NumCmds++; + + /* + * Insert STOP if we are at the last byte of either last + * message for the transaction or the client explicitly + * requires a STOP at this particular message. + */ + if ((j == msg[i].len - 1) && + ((i == num_msgs - 1) || (msg[i].flags & I2C_M_STOP))) { + cmd->CmdConfig &= ~CMDCONFIG_RESTART_MASK; + cmd->CmdConfig |= CMDCONFIG_STOP_MASK; + } + } } - - memset(&req, 0, sizeof(req)); - sienna_cichlid_fill_i2c_req(&req, true, address, numbytes, data); - mutex_lock(&adev->smu.mutex); - ret = smu_cmn_update_table(&adev->smu, SMU_TABLE_I2C_COMMANDS, 0, &req, true); + r = smu_cmn_update_table(&adev->smu, SMU_TABLE_I2C_COMMANDS, 0, req, true); mutex_unlock(&adev->smu.mutex); + if (r) + goto fail; - if (!ret) { - dev_dbg(adev->dev, "sienna_cichlid_i2c_write(), address = %x, bytes = %d , data: ", - (uint16_t)address, numbytes); - - print_hex_dump(KERN_DEBUG, "data: ", DUMP_PREFIX_NONE, - 8, 1, data, numbytes, false); - /* - * According to EEPROM spec there is a MAX of 10 ms required for - * EEPROM to flush internal RX buffer after STOP was issued at the - * end of write transaction. During this time the EEPROM will not be - * responsive to any more commands - so wait a bit more. - */ - msleep(10); - - } else - dev_err(adev->dev, "sienna_cichlid_i2c_write- error occurred :%x", ret); - - return ret; -} - -static int sienna_cichlid_i2c_xfer(struct i2c_adapter *i2c_adap, - struct i2c_msg *msgs, int num) -{ - uint32_t i, j, ret, data_size, data_chunk_size, next_eeprom_addr = 0; - uint8_t *data_ptr, data_chunk[MAX_SW_I2C_COMMANDS] = { 0 }; - - for (i = 0; i < num; i++) { - /* - * SMU interface allows at most MAX_SW_I2C_COMMANDS bytes of data at - * once and hence the data needs to be spliced into chunks and sent each - * chunk separately - */ - data_size = msgs[i].len - 2; - data_chunk_size = MAX_SW_I2C_COMMANDS - 2; - next_eeprom_addr = (msgs[i].buf[0] << 8 & 0xff00) | (msgs[i].buf[1] & 0xff); - data_ptr = msgs[i].buf + 2; - - for (j = 0; j < data_size / data_chunk_size; j++) { - /* Insert the EEPROM dest addess, bits 0-15 */ - data_chunk[0] = ((next_eeprom_addr >> 8) & 0xff); - data_chunk[1] = (next_eeprom_addr & 0xff); - - if (msgs[i].flags & I2C_M_RD) { - ret = sienna_cichlid_i2c_read_data(i2c_adap, - (uint8_t)msgs[i].addr, - data_chunk, MAX_SW_I2C_COMMANDS); - - memcpy(data_ptr, data_chunk + 2, data_chunk_size); - } else { - - memcpy(data_chunk + 2, data_ptr, data_chunk_size); - - ret = sienna_cichlid_i2c_write_data(i2c_adap, - (uint8_t)msgs[i].addr, - data_chunk, MAX_SW_I2C_COMMANDS); - } - - if (ret) { - num = -EIO; - goto fail; - } - - next_eeprom_addr += data_chunk_size; - data_ptr += data_chunk_size; + for (c = i = 0; i < num_msgs; i++) { + if (!(msg[i].flags & I2C_M_RD)) { + c += msg[i].len; + continue; } + for (j = 0; j < msg[i].len; j++, c++) { + SwI2cCmd_t *cmd = &res->SwI2cCmds[c]; - if (data_size % data_chunk_size) { - data_chunk[0] = ((next_eeprom_addr >> 8) & 0xff); - data_chunk[1] = (next_eeprom_addr & 0xff); - - if (msgs[i].flags & I2C_M_RD) { - ret = sienna_cichlid_i2c_read_data(i2c_adap, - (uint8_t)msgs[i].addr, - data_chunk, (data_size % data_chunk_size) + 2); - - memcpy(data_ptr, data_chunk + 2, data_size % data_chunk_size); - } else { - memcpy(data_chunk + 2, data_ptr, data_size % data_chunk_size); - - ret = sienna_cichlid_i2c_write_data(i2c_adap, - (uint8_t)msgs[i].addr, - data_chunk, (data_size % data_chunk_size) + 2); - } - - if (ret) { - num = -EIO; - goto fail; - } + msg[i].buf[j] = cmd->ReadWriteData; } } - + r = num_msgs; fail: - return num; + kfree(req); + return r; } static u32 sienna_cichlid_i2c_func(struct i2c_adapter *adap) @@ -3646,16 +3587,25 @@ static const struct i2c_algorithm sienna_cichlid_i2c_algo = { .functionality = sienna_cichlid_i2c_func, }; +static const struct i2c_adapter_quirks sienna_cichlid_i2c_control_quirks = { + .flags = I2C_AQ_COMB | I2C_AQ_COMB_SAME_ADDR | I2C_AQ_NO_ZERO_LEN, + .max_read_len = MAX_SW_I2C_COMMANDS, + .max_write_len = MAX_SW_I2C_COMMANDS, + .max_comb_1st_msg_len = 2, + .max_comb_2nd_msg_len = MAX_SW_I2C_COMMANDS - 2, +}; + static int sienna_cichlid_i2c_control_init(struct smu_context *smu, struct i2c_adapter *control) { struct amdgpu_device *adev = to_amdgpu_device(control); int res; control->owner = THIS_MODULE; - control->class = I2C_CLASS_SPD; + control->class = I2C_CLASS_HWMON; control->dev.parent = &adev->pdev->dev; control->algo = &sienna_cichlid_i2c_algo; snprintf(control->name, sizeof(control->name), "AMDGPU SMU"); + control->quirks = &sienna_cichlid_i2c_control_quirks; res = i2c_add_adapter(control); if (res) @@ -3678,65 +3628,94 @@ static ssize_t sienna_cichlid_get_gpu_metrics(struct smu_context *smu, SmuMetricsExternal_t metrics_external; SmuMetrics_t *metrics = &(metrics_external.SmuMetrics); + SmuMetrics_V2_t *metrics_v2 = + &(metrics_external.SmuMetrics_V2); struct amdgpu_device *adev = smu->adev; - uint32_t smu_version; + bool use_metrics_v2 = ((adev->asic_type == CHIP_SIENNA_CICHLID) && + (smu->smc_fw_version >= 0x3A4300)) ? true : false; + uint16_t average_gfx_activity; int ret = 0; - ret = smu_cmn_get_metrics_table(smu, - &metrics_external, - true); - if (ret) + mutex_lock(&smu->metrics_lock); + ret = smu_cmn_get_metrics_table_locked(smu, + &metrics_external, + true); + if (ret) { + mutex_unlock(&smu->metrics_lock); return ret; + } smu_cmn_init_soft_gpu_metrics(gpu_metrics, 1, 3); - gpu_metrics->temperature_edge = metrics->TemperatureEdge; - gpu_metrics->temperature_hotspot = metrics->TemperatureHotspot; - gpu_metrics->temperature_mem = metrics->TemperatureMem; - gpu_metrics->temperature_vrgfx = metrics->TemperatureVrGfx; - gpu_metrics->temperature_vrsoc = metrics->TemperatureVrSoc; - gpu_metrics->temperature_vrmem = metrics->TemperatureVrMem0; - - gpu_metrics->average_gfx_activity = metrics->AverageGfxActivity; - gpu_metrics->average_umc_activity = metrics->AverageUclkActivity; - gpu_metrics->average_mm_activity = metrics->VcnActivityPercentage; - - gpu_metrics->average_socket_power = metrics->AverageSocketPower; - gpu_metrics->energy_accumulator = metrics->EnergyAccumulator; - - if (metrics->AverageGfxActivity <= SMU_11_0_7_GFX_BUSY_THRESHOLD) - gpu_metrics->average_gfxclk_frequency = metrics->AverageGfxclkFrequencyPostDs; + gpu_metrics->temperature_edge = + use_metrics_v2 ? metrics_v2->TemperatureEdge : metrics->TemperatureEdge; + gpu_metrics->temperature_hotspot = + use_metrics_v2 ? metrics_v2->TemperatureHotspot : metrics->TemperatureHotspot; + gpu_metrics->temperature_mem = + use_metrics_v2 ? metrics_v2->TemperatureMem : metrics->TemperatureMem; + gpu_metrics->temperature_vrgfx = + use_metrics_v2 ? metrics_v2->TemperatureVrGfx : metrics->TemperatureVrGfx; + gpu_metrics->temperature_vrsoc = + use_metrics_v2 ? metrics_v2->TemperatureVrSoc : metrics->TemperatureVrSoc; + gpu_metrics->temperature_vrmem = + use_metrics_v2 ? metrics_v2->TemperatureVrMem0 : metrics->TemperatureVrMem0; + + gpu_metrics->average_gfx_activity = + use_metrics_v2 ? metrics_v2->AverageGfxActivity : metrics->AverageGfxActivity; + gpu_metrics->average_umc_activity = + use_metrics_v2 ? metrics_v2->AverageUclkActivity : metrics->AverageUclkActivity; + gpu_metrics->average_mm_activity = + use_metrics_v2 ? metrics_v2->VcnActivityPercentage : metrics->VcnActivityPercentage; + + gpu_metrics->average_socket_power = + use_metrics_v2 ? metrics_v2->AverageSocketPower : metrics->AverageSocketPower; + gpu_metrics->energy_accumulator = + use_metrics_v2 ? metrics_v2->EnergyAccumulator : metrics->EnergyAccumulator; + + average_gfx_activity = use_metrics_v2 ? metrics_v2->AverageGfxActivity : metrics->AverageGfxActivity; + if (average_gfx_activity <= SMU_11_0_7_GFX_BUSY_THRESHOLD) + gpu_metrics->average_gfxclk_frequency = + use_metrics_v2 ? metrics_v2->AverageGfxclkFrequencyPostDs : metrics->AverageGfxclkFrequencyPostDs; else - gpu_metrics->average_gfxclk_frequency = metrics->AverageGfxclkFrequencyPreDs; - gpu_metrics->average_uclk_frequency = metrics->AverageUclkFrequencyPostDs; - gpu_metrics->average_vclk0_frequency = metrics->AverageVclk0Frequency; - gpu_metrics->average_dclk0_frequency = metrics->AverageDclk0Frequency; - gpu_metrics->average_vclk1_frequency = metrics->AverageVclk1Frequency; - gpu_metrics->average_dclk1_frequency = metrics->AverageDclk1Frequency; - - gpu_metrics->current_gfxclk = metrics->CurrClock[PPCLK_GFXCLK]; - gpu_metrics->current_socclk = metrics->CurrClock[PPCLK_SOCCLK]; - gpu_metrics->current_uclk = metrics->CurrClock[PPCLK_UCLK]; - gpu_metrics->current_vclk0 = metrics->CurrClock[PPCLK_VCLK_0]; - gpu_metrics->current_dclk0 = metrics->CurrClock[PPCLK_DCLK_0]; - gpu_metrics->current_vclk1 = metrics->CurrClock[PPCLK_VCLK_1]; - gpu_metrics->current_dclk1 = metrics->CurrClock[PPCLK_DCLK_1]; - - gpu_metrics->throttle_status = metrics->ThrottlerStatus; + gpu_metrics->average_gfxclk_frequency = + use_metrics_v2 ? metrics_v2->AverageGfxclkFrequencyPreDs : metrics->AverageGfxclkFrequencyPreDs; + gpu_metrics->average_uclk_frequency = + use_metrics_v2 ? metrics_v2->AverageUclkFrequencyPostDs : metrics->AverageUclkFrequencyPostDs; + gpu_metrics->average_vclk0_frequency = + use_metrics_v2 ? metrics_v2->AverageVclk0Frequency : metrics->AverageVclk0Frequency; + gpu_metrics->average_dclk0_frequency = + use_metrics_v2 ? metrics_v2->AverageDclk0Frequency : metrics->AverageDclk0Frequency; + gpu_metrics->average_vclk1_frequency = + use_metrics_v2 ? metrics_v2->AverageVclk1Frequency : metrics->AverageVclk1Frequency; + gpu_metrics->average_dclk1_frequency = + use_metrics_v2 ? metrics_v2->AverageDclk1Frequency : metrics->AverageDclk1Frequency; + + gpu_metrics->current_gfxclk = + use_metrics_v2 ? metrics_v2->CurrClock[PPCLK_GFXCLK] : metrics->CurrClock[PPCLK_GFXCLK]; + gpu_metrics->current_socclk = + use_metrics_v2 ? metrics_v2->CurrClock[PPCLK_SOCCLK] : metrics->CurrClock[PPCLK_SOCCLK]; + gpu_metrics->current_uclk = + use_metrics_v2 ? metrics_v2->CurrClock[PPCLK_UCLK] : metrics->CurrClock[PPCLK_UCLK]; + gpu_metrics->current_vclk0 = + use_metrics_v2 ? metrics_v2->CurrClock[PPCLK_VCLK_0] : metrics->CurrClock[PPCLK_VCLK_0]; + gpu_metrics->current_dclk0 = + use_metrics_v2 ? metrics_v2->CurrClock[PPCLK_DCLK_0] : metrics->CurrClock[PPCLK_DCLK_0]; + gpu_metrics->current_vclk1 = + use_metrics_v2 ? metrics_v2->CurrClock[PPCLK_VCLK_1] : metrics->CurrClock[PPCLK_VCLK_1]; + gpu_metrics->current_dclk1 = + use_metrics_v2 ? metrics_v2->CurrClock[PPCLK_DCLK_1] : metrics->CurrClock[PPCLK_DCLK_1]; + + gpu_metrics->throttle_status = sienna_cichlid_get_throttler_status_locked(smu); gpu_metrics->indep_throttle_status = - smu_cmn_get_indep_throttler_status(metrics->ThrottlerStatus, + smu_cmn_get_indep_throttler_status(gpu_metrics->throttle_status, sienna_cichlid_throttler_map); - gpu_metrics->current_fan_speed = metrics->CurrFanSpeed; + gpu_metrics->current_fan_speed = use_metrics_v2 ? metrics_v2->CurrFanSpeed : metrics->CurrFanSpeed; - ret = smu_cmn_get_smc_version(smu, NULL, &smu_version); - if (ret) - return ret; - - if (((adev->asic_type == CHIP_SIENNA_CICHLID) && smu_version > 0x003A1E00) || - ((adev->asic_type == CHIP_NAVY_FLOUNDER) && smu_version > 0x00410400)) { - gpu_metrics->pcie_link_width = metrics->PcieWidth; - gpu_metrics->pcie_link_speed = link_speed[metrics->PcieRate]; + if (((adev->asic_type == CHIP_SIENNA_CICHLID) && smu->smc_fw_version > 0x003A1E00) || + ((adev->asic_type == CHIP_NAVY_FLOUNDER) && smu->smc_fw_version > 0x00410400)) { + gpu_metrics->pcie_link_width = use_metrics_v2 ? metrics_v2->PcieWidth : metrics->PcieWidth; + gpu_metrics->pcie_link_speed = link_speed[use_metrics_v2 ? metrics_v2->PcieRate : metrics->PcieRate]; } else { gpu_metrics->pcie_link_width = smu_v11_0_get_current_pcie_link_width(smu); @@ -3744,6 +3723,8 @@ static ssize_t sienna_cichlid_get_gpu_metrics(struct smu_context *smu, smu_v11_0_get_current_pcie_link_speed(smu); } + mutex_unlock(&smu->metrics_lock); + gpu_metrics->system_clock_counter = ktime_get_boottime_ns(); *table = (void *)gpu_metrics; @@ -3939,6 +3920,7 @@ static const struct pptable_funcs sienna_cichlid_ppt_funcs = { .set_soft_freq_limited_range = smu_v11_0_set_soft_freq_limited_range, .set_default_od_settings = sienna_cichlid_set_default_od_settings, .od_edit_dpm_table = sienna_cichlid_od_edit_dpm_table, + .restore_user_od_settings = smu_v11_0_restore_user_od_settings, .run_btc = sienna_cichlid_run_btc, .set_power_source = smu_v11_0_set_power_source, .get_pp_feature_mask = smu_cmn_get_pp_feature_mask, diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/smu_v11_0.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/smu_v11_0.c index 0a5d46ac9ccd..b5419e8eba89 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/smu_v11_0.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/smu_v11_0.c @@ -266,6 +266,9 @@ int smu_v11_0_check_fw_version(struct smu_context *smu) case CHIP_BEIGE_GOBY: smu->smc_driver_if_version = SMU11_DRIVER_IF_VERSION_Beige_Goby; break; + case CHIP_CYAN_SKILLFISH: + smu->smc_driver_if_version = SMU11_DRIVER_IF_VERSION_Cyan_Skillfish; + break; default: dev_err(smu->adev->dev, "smu unsupported asic type:%d.\n", smu->adev->asic_type); smu->smc_driver_if_version = SMU11_DRIVER_IF_VERSION_INV; @@ -422,10 +425,20 @@ int smu_v11_0_init_smc_tables(struct smu_context *smu) ret = -ENOMEM; goto err3_out; } + + smu_table->user_overdrive_table = + kzalloc(tables[SMU_TABLE_OVERDRIVE].size, GFP_KERNEL); + if (!smu_table->user_overdrive_table) { + ret = -ENOMEM; + goto err4_out; + } + } return 0; +err4_out: + kfree(smu_table->boot_overdrive_table); err3_out: kfree(smu_table->overdrive_table); err2_out: @@ -442,12 +455,14 @@ int smu_v11_0_fini_smc_tables(struct smu_context *smu) struct smu_dpm_context *smu_dpm = &smu->smu_dpm; kfree(smu_table->gpu_metrics_table); + kfree(smu_table->user_overdrive_table); kfree(smu_table->boot_overdrive_table); kfree(smu_table->overdrive_table); kfree(smu_table->max_sustainable_clocks); kfree(smu_table->driver_pptable); kfree(smu_table->clocks_table); smu_table->gpu_metrics_table = NULL; + smu_table->user_overdrive_table = NULL; smu_table->boot_overdrive_table = NULL; smu_table->overdrive_table = NULL; smu_table->max_sustainable_clocks = NULL; @@ -2101,3 +2116,16 @@ int smu_v11_0_deep_sleep_control(struct smu_context *smu, return ret; } + +int smu_v11_0_restore_user_od_settings(struct smu_context *smu) +{ + struct smu_table_context *table_context = &smu->smu_table; + void *user_od_table = table_context->user_overdrive_table; + int ret = 0; + + ret = smu_cmn_update_table(smu, SMU_TABLE_OVERDRIVE, 0, (void *)user_od_table, true); + if (ret) + dev_err(smu->adev->dev, "Failed to import overdrive table!\n"); + + return ret; +} diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c index cb5485cf243f..856eeaf293b8 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/aldebaran_ppt.c @@ -1451,197 +1451,77 @@ static bool aldebaran_is_dpm_running(struct smu_context *smu) return !!(feature_enabled & SMC_DPM_FEATURE); } -static void aldebaran_fill_i2c_req(SwI2cRequest_t *req, bool write, - uint8_t address, uint32_t numbytes, - uint8_t *data) -{ - int i; - - req->I2CcontrollerPort = 0; - req->I2CSpeed = 2; - req->SlaveAddress = address; - req->NumCmds = numbytes; - - for (i = 0; i < numbytes; i++) { - SwI2cCmd_t *cmd = &req->SwI2cCmds[i]; - - /* First 2 bytes are always write for lower 2b EEPROM address */ - if (i < 2) - cmd->CmdConfig = CMDCONFIG_READWRITE_MASK; - else - cmd->CmdConfig = write ? CMDCONFIG_READWRITE_MASK : 0; - - - /* Add RESTART for read after address filled */ - cmd->CmdConfig |= (i == 2 && !write) ? CMDCONFIG_RESTART_MASK : 0; - - /* Add STOP in the end */ - cmd->CmdConfig |= (i == (numbytes - 1)) ? CMDCONFIG_STOP_MASK : 0; - - /* Fill with data regardless if read or write to simplify code */ - cmd->ReadWriteData = data[i]; - } -} - -static int aldebaran_i2c_read_data(struct i2c_adapter *control, - uint8_t address, - uint8_t *data, - uint32_t numbytes) +static int aldebaran_i2c_xfer(struct i2c_adapter *i2c_adap, + struct i2c_msg *msg, int num_msgs) { - uint32_t i, ret = 0; - SwI2cRequest_t req; - struct amdgpu_device *adev = to_amdgpu_device(control); + struct amdgpu_device *adev = to_amdgpu_device(i2c_adap); struct smu_table_context *smu_table = &adev->smu.smu_table; struct smu_table *table = &smu_table->driver_table; + SwI2cRequest_t *req, *res = (SwI2cRequest_t *)table->cpu_addr; + int i, j, r, c; + u16 dir; - if (numbytes > MAX_SW_I2C_COMMANDS) { - dev_err(adev->dev, "numbytes requested %d is over max allowed %d\n", - numbytes, MAX_SW_I2C_COMMANDS); - return -EINVAL; - } - - memset(&req, 0, sizeof(req)); - aldebaran_fill_i2c_req(&req, false, address, numbytes, data); - - mutex_lock(&adev->smu.mutex); - /* Now read data starting with that address */ - ret = smu_cmn_update_table(&adev->smu, SMU_TABLE_I2C_COMMANDS, 0, &req, - true); - mutex_unlock(&adev->smu.mutex); - - if (!ret) { - SwI2cRequest_t *res = (SwI2cRequest_t *)table->cpu_addr; - - /* Assume SMU fills res.SwI2cCmds[i].Data with read bytes */ - for (i = 0; i < numbytes; i++) - data[i] = res->SwI2cCmds[i].ReadWriteData; - - dev_dbg(adev->dev, "aldebaran_i2c_read_data, address = %x, bytes = %d, data :", - (uint16_t)address, numbytes); - - print_hex_dump(KERN_DEBUG, "data: ", DUMP_PREFIX_NONE, - 8, 1, data, numbytes, false); - } else - dev_err(adev->dev, "aldebaran_i2c_read_data - error occurred :%x", ret); + req = kzalloc(sizeof(*req), GFP_KERNEL); + if (!req) + return -ENOMEM; - return ret; -} + req->I2CcontrollerPort = 0; + req->I2CSpeed = I2C_SPEED_FAST_400K; + req->SlaveAddress = msg[0].addr << 1; /* wants an 8-bit address */ + dir = msg[0].flags & I2C_M_RD; + + for (c = i = 0; i < num_msgs; i++) { + for (j = 0; j < msg[i].len; j++, c++) { + SwI2cCmd_t *cmd = &req->SwI2cCmds[c]; + + if (!(msg[i].flags & I2C_M_RD)) { + /* write */ + cmd->CmdConfig |= CMDCONFIG_READWRITE_MASK; + cmd->ReadWriteData = msg[i].buf[j]; + } -static int aldebaran_i2c_write_data(struct i2c_adapter *control, - uint8_t address, - uint8_t *data, - uint32_t numbytes) -{ - uint32_t ret; - SwI2cRequest_t req; - struct amdgpu_device *adev = to_amdgpu_device(control); + if ((dir ^ msg[i].flags) & I2C_M_RD) { + /* The direction changes. + */ + dir = msg[i].flags & I2C_M_RD; + cmd->CmdConfig |= CMDCONFIG_RESTART_MASK; + } - if (numbytes > MAX_SW_I2C_COMMANDS) { - dev_err(adev->dev, "numbytes requested %d is over max allowed %d\n", - numbytes, MAX_SW_I2C_COMMANDS); - return -EINVAL; + req->NumCmds++; + + /* + * Insert STOP if we are at the last byte of either last + * message for the transaction or the client explicitly + * requires a STOP at this particular message. + */ + if ((j == msg[i].len - 1) && + ((i == num_msgs - 1) || (msg[i].flags & I2C_M_STOP))) { + cmd->CmdConfig &= ~CMDCONFIG_RESTART_MASK; + cmd->CmdConfig |= CMDCONFIG_STOP_MASK; + } + } } - - memset(&req, 0, sizeof(req)); - aldebaran_fill_i2c_req(&req, true, address, numbytes, data); - mutex_lock(&adev->smu.mutex); - ret = smu_cmn_update_table(&adev->smu, SMU_TABLE_I2C_COMMANDS, 0, &req, true); + r = smu_cmn_update_table(&adev->smu, SMU_TABLE_I2C_COMMANDS, 0, req, true); mutex_unlock(&adev->smu.mutex); + if (r) + goto fail; - if (!ret) { - dev_dbg(adev->dev, "aldebaran_i2c_write(), address = %x, bytes = %d , data: ", - (uint16_t)address, numbytes); - - print_hex_dump(KERN_DEBUG, "data: ", DUMP_PREFIX_NONE, - 8, 1, data, numbytes, false); - /* - * According to EEPROM spec there is a MAX of 10 ms required for - * EEPROM to flush internal RX buffer after STOP was issued at the - * end of write transaction. During this time the EEPROM will not be - * responsive to any more commands - so wait a bit more. - */ - msleep(10); - - } else - dev_err(adev->dev, "aldebaran_i2c_write- error occurred :%x", ret); - - return ret; -} - -static int aldebaran_i2c_xfer(struct i2c_adapter *i2c_adap, - struct i2c_msg *msgs, int num) -{ - uint32_t i, j, ret, data_size, data_chunk_size, next_eeprom_addr = 0; - uint8_t *data_ptr, data_chunk[MAX_SW_I2C_COMMANDS] = { 0 }; - - for (i = 0; i < num; i++) { - /* - * SMU interface allows at most MAX_SW_I2C_COMMANDS bytes of data at - * once and hence the data needs to be spliced into chunks and sent each - * chunk separately - */ - data_size = msgs[i].len - 2; - data_chunk_size = MAX_SW_I2C_COMMANDS - 2; - next_eeprom_addr = (msgs[i].buf[0] << 8 & 0xff00) | (msgs[i].buf[1] & 0xff); - data_ptr = msgs[i].buf + 2; - - for (j = 0; j < data_size / data_chunk_size; j++) { - /* Insert the EEPROM dest addess, bits 0-15 */ - data_chunk[0] = ((next_eeprom_addr >> 8) & 0xff); - data_chunk[1] = (next_eeprom_addr & 0xff); - - if (msgs[i].flags & I2C_M_RD) { - ret = aldebaran_i2c_read_data(i2c_adap, - (uint8_t)msgs[i].addr, - data_chunk, MAX_SW_I2C_COMMANDS); - - memcpy(data_ptr, data_chunk + 2, data_chunk_size); - } else { - - memcpy(data_chunk + 2, data_ptr, data_chunk_size); - - ret = aldebaran_i2c_write_data(i2c_adap, - (uint8_t)msgs[i].addr, - data_chunk, MAX_SW_I2C_COMMANDS); - } - - if (ret) { - num = -EIO; - goto fail; - } - - next_eeprom_addr += data_chunk_size; - data_ptr += data_chunk_size; + for (c = i = 0; i < num_msgs; i++) { + if (!(msg[i].flags & I2C_M_RD)) { + c += msg[i].len; + continue; } + for (j = 0; j < msg[i].len; j++, c++) { + SwI2cCmd_t *cmd = &res->SwI2cCmds[c]; - if (data_size % data_chunk_size) { - data_chunk[0] = ((next_eeprom_addr >> 8) & 0xff); - data_chunk[1] = (next_eeprom_addr & 0xff); - - if (msgs[i].flags & I2C_M_RD) { - ret = aldebaran_i2c_read_data(i2c_adap, - (uint8_t)msgs[i].addr, - data_chunk, (data_size % data_chunk_size) + 2); - - memcpy(data_ptr, data_chunk + 2, data_size % data_chunk_size); - } else { - memcpy(data_chunk + 2, data_ptr, data_size % data_chunk_size); - - ret = aldebaran_i2c_write_data(i2c_adap, - (uint8_t)msgs[i].addr, - data_chunk, (data_size % data_chunk_size) + 2); - } - - if (ret) { - num = -EIO; - goto fail; - } + msg[i].buf[j] = cmd->ReadWriteData; } } - + r = num_msgs; fail: - return num; + kfree(req); + return r; } static u32 aldebaran_i2c_func(struct i2c_adapter *adap) @@ -1655,6 +1535,14 @@ static const struct i2c_algorithm aldebaran_i2c_algo = { .functionality = aldebaran_i2c_func, }; +static const struct i2c_adapter_quirks aldebaran_i2c_control_quirks = { + .flags = I2C_AQ_COMB | I2C_AQ_COMB_SAME_ADDR | I2C_AQ_NO_ZERO_LEN, + .max_read_len = MAX_SW_I2C_COMMANDS, + .max_write_len = MAX_SW_I2C_COMMANDS, + .max_comb_1st_msg_len = 2, + .max_comb_2nd_msg_len = MAX_SW_I2C_COMMANDS - 2, +}; + static int aldebaran_i2c_control_init(struct smu_context *smu, struct i2c_adapter *control) { struct amdgpu_device *adev = to_amdgpu_device(control); @@ -1665,6 +1553,7 @@ static int aldebaran_i2c_control_init(struct smu_context *smu, struct i2c_adapte control->dev.parent = &adev->pdev->dev; control->algo = &aldebaran_i2c_algo; snprintf(control->name, sizeof(control->name), "AMDGPU SMU"); + control->quirks = &aldebaran_i2c_control_quirks; res = i2c_add_adapter(control); if (res) @@ -1764,7 +1653,9 @@ static void aldebaran_log_thermal_throttling_event(struct smu_context *smu) dev_warn(adev->dev, "WARN: GPU thermal throttling temperature reached, expect performance decrease. %s.\n", log_buf); - kgd2kfd_smi_event_throttle(smu->adev->kfd.dev, throttler_status); + kgd2kfd_smi_event_throttle(smu->adev->kfd.dev, + smu_cmn_get_indep_throttler_status(throttler_status, + aldebaran_throttler_map)); } static int aldebaran_get_current_pcie_link_speed(struct smu_context *smu) diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c b/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c index e802f9a95f08..415be74df28c 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c @@ -55,7 +55,7 @@ #undef __SMU_DUMMY_MAP #define __SMU_DUMMY_MAP(type) #type -static const char* __smu_message_names[] = { +static const char * const __smu_message_names[] = { SMU_MESSAGE_TYPES }; @@ -76,55 +76,256 @@ static void smu_cmn_read_arg(struct smu_context *smu, *arg = RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_82); } -int smu_cmn_wait_for_response(struct smu_context *smu) +/* Redefine the SMU error codes here. + * + * Note that these definitions are redundant and should be removed + * when the SMU has exported a unified header file containing these + * macros, which header file we can just include and use the SMU's + * macros. At the moment, these error codes are defined by the SMU + * per-ASIC unfortunately, yet we're a one driver for all ASICs. + */ +#define SMU_RESP_NONE 0 +#define SMU_RESP_OK 1 +#define SMU_RESP_CMD_FAIL 0xFF +#define SMU_RESP_CMD_UNKNOWN 0xFE +#define SMU_RESP_CMD_BAD_PREREQ 0xFD +#define SMU_RESP_BUSY_OTHER 0xFC +#define SMU_RESP_DEBUG_END 0xFB + +/** + * __smu_cmn_poll_stat -- poll for a status from the SMU + * smu: a pointer to SMU context + * + * Returns the status of the SMU, which could be, + * 0, the SMU is busy with your previous command; + * 1, execution status: success, execution result: success; + * 0xFF, execution status: success, execution result: failure; + * 0xFE, unknown command; + * 0xFD, valid command, but bad (command) prerequisites; + * 0xFC, the command was rejected as the SMU is busy; + * 0xFB, "SMC_Result_DebugDataDumpEnd". + * + * The values here are not defined by macros, because I'd rather we + * include a single header file which defines them, which is + * maintained by the SMU FW team, so that we're impervious to firmware + * changes. At the moment those values are defined in various header + * files, one for each ASIC, yet here we're a single ASIC-agnostic + * interface. Such a change can be followed-up by a subsequent patch. + */ +static u32 __smu_cmn_poll_stat(struct smu_context *smu) { struct amdgpu_device *adev = smu->adev; - uint32_t cur_value, i, timeout = adev->usec_timeout * 20; + int timeout = adev->usec_timeout * 20; + u32 reg; - for (i = 0; i < timeout; i++) { - cur_value = RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90); - if ((cur_value & MP1_C2PMSG_90__CONTENT_MASK) != 0) - return cur_value; + for ( ; timeout > 0; timeout--) { + reg = RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90); + if ((reg & MP1_C2PMSG_90__CONTENT_MASK) != 0) + break; udelay(1); } - /* timeout means wrong logic */ - if (i == timeout) - return -ETIME; - - return RREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90); + return reg; } -int smu_cmn_send_msg_without_waiting(struct smu_context *smu, - uint16_t msg, uint32_t param) +static void __smu_cmn_reg_print_error(struct smu_context *smu, + u32 reg_c2pmsg_90, + int msg_index, + u32 param, + enum smu_message_type msg) { struct amdgpu_device *adev = smu->adev; - int ret; + const char *message = smu_get_message_name(smu, msg); - ret = smu_cmn_wait_for_response(smu); - if (ret != 0x1) { - dev_err(adev->dev, "Msg issuing pre-check failed(0x%x) and " - "SMU may be not in the right state!\n", ret); - if (ret != -ETIME) - ret = -EIO; - return ret; + switch (reg_c2pmsg_90) { + case SMU_RESP_NONE: + dev_err_ratelimited(adev->dev, + "SMU: I'm not done with your previous command!"); + break; + case SMU_RESP_OK: + /* The SMU executed the command. It completed with a + * successful result. + */ + break; + case SMU_RESP_CMD_FAIL: + /* The SMU executed the command. It completed with an + * unsuccessful result. + */ + break; + case SMU_RESP_CMD_UNKNOWN: + dev_err_ratelimited(adev->dev, + "SMU: unknown command: index:%d param:0x%08X message:%s", + msg_index, param, message); + break; + case SMU_RESP_CMD_BAD_PREREQ: + dev_err_ratelimited(adev->dev, + "SMU: valid command, bad prerequisites: index:%d param:0x%08X message:%s", + msg_index, param, message); + break; + case SMU_RESP_BUSY_OTHER: + dev_err_ratelimited(adev->dev, + "SMU: I'm very busy for your command: index:%d param:0x%08X message:%s", + msg_index, param, message); + break; + case SMU_RESP_DEBUG_END: + dev_err_ratelimited(adev->dev, + "SMU: I'm debugging!"); + break; + default: + dev_err_ratelimited(adev->dev, + "SMU: response:0x%08X for index:%d param:0x%08X message:%s?", + reg_c2pmsg_90, msg_index, param, message); + break; } +} + +static int __smu_cmn_reg2errno(struct smu_context *smu, u32 reg_c2pmsg_90) +{ + int res; + + switch (reg_c2pmsg_90) { + case SMU_RESP_NONE: + /* The SMU is busy--still executing your command. + */ + res = -ETIME; + break; + case SMU_RESP_OK: + res = 0; + break; + case SMU_RESP_CMD_FAIL: + /* Command completed successfully, but the command + * status was failure. + */ + res = -EIO; + break; + case SMU_RESP_CMD_UNKNOWN: + /* Unknown command--ignored by the SMU. + */ + res = -EOPNOTSUPP; + break; + case SMU_RESP_CMD_BAD_PREREQ: + /* Valid command--bad prerequisites. + */ + res = -EINVAL; + break; + case SMU_RESP_BUSY_OTHER: + /* The SMU is busy with other commands. The client + * should retry in 10 us. + */ + res = -EBUSY; + break; + default: + /* Unknown or debug response from the SMU. + */ + res = -EREMOTEIO; + break; + } + + return res; +} + +static void __smu_cmn_send_msg(struct smu_context *smu, + u16 msg, + u32 param) +{ + struct amdgpu_device *adev = smu->adev; WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_90, 0); WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_82, param); WREG32_SOC15(MP1, 0, mmMP1_SMN_C2PMSG_66, msg); +} - return 0; +/** + * smu_cmn_send_msg_without_waiting -- send the message; don't wait for status + * @smu: pointer to an SMU context + * @msg_index: message index + * @param: message parameter to send to the SMU + * + * Send a message to the SMU with the parameter passed. Do not wait + * for status/result of the message, thus the "without_waiting". + * + * Return 0 on success, -errno on error if we weren't able to _send_ + * the message for some reason. See __smu_cmn_reg2errno() for details + * of the -errno. + */ +int smu_cmn_send_msg_without_waiting(struct smu_context *smu, + uint16_t msg_index, + uint32_t param) +{ + u32 reg; + int res; + + if (smu->adev->no_hw_access) + return 0; + + reg = __smu_cmn_poll_stat(smu); + res = __smu_cmn_reg2errno(smu, reg); + if (reg == SMU_RESP_NONE || + reg == SMU_RESP_BUSY_OTHER || + res == -EREMOTEIO) + goto Out; + __smu_cmn_send_msg(smu, msg_index, param); + res = 0; +Out: + return res; +} + +/** + * smu_cmn_wait_for_response -- wait for response from the SMU + * @smu: pointer to an SMU context + * + * Wait for status from the SMU. + * + * Return 0 on success, -errno on error, indicating the execution + * status and result of the message being waited for. See + * __smu_cmn_reg2errno() for details of the -errno. + */ +int smu_cmn_wait_for_response(struct smu_context *smu) +{ + u32 reg; + + reg = __smu_cmn_poll_stat(smu); + return __smu_cmn_reg2errno(smu, reg); } +/** + * smu_cmn_send_smc_msg_with_param -- send a message with parameter + * @smu: pointer to an SMU context + * @msg: message to send + * @param: parameter to send to the SMU + * @read_arg: pointer to u32 to return a value from the SMU back + * to the caller + * + * Send the message @msg with parameter @param to the SMU, wait for + * completion of the command, and return back a value from the SMU in + * @read_arg pointer. + * + * Return 0 on success, -errno on error, if we weren't able to send + * the message or if the message completed with some kind of + * error. See __smu_cmn_reg2errno() for details of the -errno. + * + * If we weren't able to send the message to the SMU, we also print + * the error to the standard log. + * + * Command completion status is printed only if the -errno is + * -EREMOTEIO, indicating that the SMU returned back an + * undefined/unknown/unspecified result. All other cases are + * well-defined, not printed, but instead given back to the client to + * decide what further to do. + * + * The return value, @read_arg is read back regardless, to give back + * more information to the client, which on error would most likely be + * @param, but we can't assume that. This also eliminates more + * conditionals. + */ int smu_cmn_send_smc_msg_with_param(struct smu_context *smu, enum smu_message_type msg, uint32_t param, uint32_t *read_arg) { - struct amdgpu_device *adev = smu->adev; - int ret = 0, index = 0; + int res, index; + u32 reg; if (smu->adev->no_hw_access) return 0; @@ -136,31 +337,24 @@ int smu_cmn_send_smc_msg_with_param(struct smu_context *smu, return index == -EACCES ? 0 : index; mutex_lock(&smu->message_lock); - ret = smu_cmn_send_msg_without_waiting(smu, (uint16_t)index, param); - if (ret) - goto out; - - ret = smu_cmn_wait_for_response(smu); - if (ret != 0x1) { - if (ret == -ETIME) { - dev_err(adev->dev, "message: %15s (%d) \tparam: 0x%08x is timeout (no response)\n", - smu_get_message_name(smu, msg), index, param); - } else { - dev_err(adev->dev, "failed send message: %15s (%d) \tparam: 0x%08x response %#x\n", - smu_get_message_name(smu, msg), index, param, - ret); - ret = -EIO; - } - goto out; + reg = __smu_cmn_poll_stat(smu); + res = __smu_cmn_reg2errno(smu, reg); + if (reg == SMU_RESP_NONE || + reg == SMU_RESP_BUSY_OTHER || + res == -EREMOTEIO) { + __smu_cmn_reg_print_error(smu, reg, index, param, msg); + goto Out; } - + __smu_cmn_send_msg(smu, (uint16_t) index, param); + reg = __smu_cmn_poll_stat(smu); + res = __smu_cmn_reg2errno(smu, reg); + if (res == -EREMOTEIO) + __smu_cmn_reg_print_error(smu, reg, index, param, msg); if (read_arg) smu_cmn_read_arg(smu, read_arg); - - ret = 0; /* 0 as driver return value */ -out: +Out: mutex_unlock(&smu->message_lock); - return ret; + return res; } int smu_cmn_send_smc_msg(struct smu_context *smu, diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.h b/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.h index 9add5f16ff56..16993daa2ae0 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.h +++ b/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.h @@ -27,7 +27,8 @@ #if defined(SWSMU_CODE_LAYER_L2) || defined(SWSMU_CODE_LAYER_L3) || defined(SWSMU_CODE_LAYER_L4) int smu_cmn_send_msg_without_waiting(struct smu_context *smu, - uint16_t msg, uint32_t param); + uint16_t msg_index, + uint32_t param); int smu_cmn_send_smc_msg_with_param(struct smu_context *smu, enum smu_message_type msg, uint32_t param, |