Merge tag 'v6.9-rc1' into sched/core, to pick up fixes and to refresh the branch

Signed-off-by: Ingo Molnar <mingo@kernel.org>

1 files changed, 40 insertions, 9 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index 94bdb5fa6ebc..5dc24c971b41 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -96,6 +96,9 @@ MODULE_FIRMWARE("amdgpu/navi12_gpu_info.bin");
 #define AMDGPU_RESUME_MS		2000
 #define AMDGPU_MAX_RETRY_LIMIT		2
 #define AMDGPU_RETRY_SRIOV_RESET(r) ((r) == -EBUSY || (r) == -ETIMEDOUT || (r) == -EINVAL)
+#define AMDGPU_PCIE_INDEX_FALLBACK (0x38 >> 2)
+#define AMDGPU_PCIE_INDEX_HI_FALLBACK (0x44 >> 2)
+#define AMDGPU_PCIE_DATA_FALLBACK (0x3C >> 2)
 
 static const struct drm_driver amdgpu_kms_driver;
 
@@ -781,12 +784,22 @@ u32 amdgpu_device_indirect_rreg_ext(struct amdgpu_device *adev,
 	void __iomem *pcie_index_hi_offset;
 	void __iomem *pcie_data_offset;
 
-	pcie_index = adev->nbio.funcs->get_pcie_index_offset(adev);
-	pcie_data = adev->nbio.funcs->get_pcie_data_offset(adev);
-	if ((reg_addr >> 32) && (adev->nbio.funcs->get_pcie_index_hi_offset))
-		pcie_index_hi = adev->nbio.funcs->get_pcie_index_hi_offset(adev);
-	else
+	if (unlikely(!adev->nbio.funcs)) {
+		pcie_index = AMDGPU_PCIE_INDEX_FALLBACK;
+		pcie_data = AMDGPU_PCIE_DATA_FALLBACK;
+	} else {
+		pcie_index = adev->nbio.funcs->get_pcie_index_offset(adev);
+		pcie_data = adev->nbio.funcs->get_pcie_data_offset(adev);
+	}
+
+	if (reg_addr >> 32) {
+		if (unlikely(!adev->nbio.funcs))
+			pcie_index_hi = AMDGPU_PCIE_INDEX_HI_FALLBACK;
+		else
+			pcie_index_hi = adev->nbio.funcs->get_pcie_index_hi_offset(adev);
+	} else {
 		pcie_index_hi = 0;
+	}
 
 	spin_lock_irqsave(&adev->pcie_idx_lock, flags);
 	pcie_index_offset = (void __iomem *)adev->rmmio + pcie_index * 4;
@@ -1218,8 +1231,6 @@ static int amdgpu_device_asic_init(struct amdgpu_device *adev)
 	    amdgpu_ip_version(adev, GC_HWIP, 0) >= IP_VERSION(11, 0, 0)) {
 		amdgpu_psp_wait_for_bootloader(adev);
 		ret = amdgpu_atomfirmware_asic_init(adev, true);
-		/* TODO: check the return val and stop device initialization if boot fails */
-		amdgpu_psp_query_boot_status(adev);
 		return ret;
 	} else {
 		return amdgpu_atom_asic_init(adev->mode_info.atom_context);
@@ -1442,6 +1453,10 @@ int amdgpu_device_resize_fb_bar(struct amdgpu_device *adev)
 	if (amdgpu_sriov_vf(adev))
 		return 0;
 
+	/* PCI_EXT_CAP_ID_VNDR extended capability is located at 0x100 */
+	if (!pci_find_ext_capability(adev->pdev, PCI_EXT_CAP_ID_VNDR))
+		DRM_WARN("System can't access extended configuration space,please check!!\n");
+
 	/* skip if the bios has already enabled large BAR */
 	if (adev->gmc.real_vram_size &&
 	    (pci_resource_len(adev->pdev, 0) >= adev->gmc.real_vram_size))
@@ -4039,13 +4054,13 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 		return r;
 	}
 
+	amdgpu_device_set_mcbp(adev);
+
 	/* early init functions */
 	r = amdgpu_device_ip_early_init(adev);
 	if (r)
 		return r;
 
-	amdgpu_device_set_mcbp(adev);
-
 	/* Get rid of things like offb */
 	r = drm_aperture_remove_conflicting_pci_framebuffers(adev->pdev, &amdgpu_kms_driver);
 	if (r)
@@ -5306,6 +5321,7 @@ int amdgpu_do_asic_reset(struct list_head *device_list_handle,
 	list_for_each_entry(tmp_adev, device_list_handle, reset_list) {
 		if (need_full_reset) {
 			/* post card */
+			amdgpu_ras_set_fed(tmp_adev, false);
 			r = amdgpu_device_asic_init(tmp_adev);
 			if (r) {
 				dev_warn(tmp_adev->dev, "asic atom init failed!");
@@ -5686,6 +5702,7 @@ retry:	/* Rest of adevs pre asic reset from XGMI hive. */
 		/* Aldebaran and gfx_11_0_3 support ras in SRIOV, so need resume ras during reset */
 		if (amdgpu_ip_version(adev, GC_HWIP, 0) ==
 			    IP_VERSION(9, 4, 2) ||
+		    amdgpu_ip_version(adev, GC_HWIP, 0) == IP_VERSION(9, 4, 3) ||
 		    amdgpu_ip_version(adev, GC_HWIP, 0) == IP_VERSION(11, 0, 3))
 			amdgpu_ras_resume(adev);
 	} else {
@@ -6107,6 +6124,20 @@ pci_ers_result_t amdgpu_pci_slot_reset(struct pci_dev *pdev)
 	struct amdgpu_reset_context reset_context;
 	u32 memsize;
 	struct list_head device_list;
+	struct amdgpu_hive_info *hive;
+	int hive_ras_recovery = 0;
+	struct amdgpu_ras *ras;
+
+	/* PCI error slot reset should be skipped During RAS recovery */
+	hive = amdgpu_get_xgmi_hive(adev);
+	if (hive) {
+		hive_ras_recovery = atomic_read(&hive->ras_recovery);
+		amdgpu_put_xgmi_hive(hive);
+	}
+	ras = amdgpu_ras_get_context(adev);
+	if ((amdgpu_ip_version(adev, GC_HWIP, 0) == IP_VERSION(9, 4, 3)) &&
+		 ras && (atomic_read(&ras->in_recovery) || hive_ras_recovery))
+		return PCI_ERS_RESULT_RECOVERED;
 
 	DRM_INFO("PCI error: slot reset callback!!\n");