Merge remote-tracking branch 'torvalds/master' into perf/core

To get closer to upstream and check if we need to sync more UAPI
headers, pick up fixes for libbpf that prevent perf's container tests
from completing successfuly, etc.

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>

9 files changed, 143 insertions, 60 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
index 1d3ee9c42f7e..6a5c96e519b1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
@@ -1140,7 +1140,8 @@ int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu(
 			adev->asic_type != CHIP_FIJI &&
 			adev->asic_type != CHIP_POLARIS10 &&
 			adev->asic_type != CHIP_POLARIS11 &&
-			adev->asic_type != CHIP_POLARIS12) ?
+			adev->asic_type != CHIP_POLARIS12 &&
+			adev->asic_type != CHIP_VEGAM) ?
 			VI_BO_SIZE_ALIGN : 1;
 
 	mapping_flags = AMDGPU_VM_PAGE_READABLE;
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
index e069de8b54e6..4e4094f842e7 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
@@ -1044,29 +1044,27 @@ static int amdgpu_cs_process_fence_dep(struct amdgpu_cs_parser *p,
 			return r;
 		}
 
-		fence = amdgpu_ctx_get_fence(ctx, entity,
-					     deps[i].handle);
+		fence = amdgpu_ctx_get_fence(ctx, entity, deps[i].handle);
+		amdgpu_ctx_put(ctx);
+
+		if (IS_ERR(fence))
+			return PTR_ERR(fence);
+		else if (!fence)
+			continue;
 
 		if (chunk->chunk_id == AMDGPU_CHUNK_ID_SCHEDULED_DEPENDENCIES) {
-			struct drm_sched_fence *s_fence = to_drm_sched_fence(fence);
+			struct drm_sched_fence *s_fence;
 			struct dma_fence *old = fence;
 
+			s_fence = to_drm_sched_fence(fence);
 			fence = dma_fence_get(&s_fence->scheduled);
 			dma_fence_put(old);
 		}
 
-		if (IS_ERR(fence)) {
-			r = PTR_ERR(fence);
-			amdgpu_ctx_put(ctx);
+		r = amdgpu_sync_fence(p->adev, &p->job->sync, fence, true);
+		dma_fence_put(fence);
+		if (r)
 			return r;
-		} else if (fence) {
-			r = amdgpu_sync_fence(p->adev, &p->job->sync, fence,
-					true);
-			dma_fence_put(fence);
-			amdgpu_ctx_put(ctx);
-			if (r)
-				return r;
-		}
 	}
 	return 0;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
index 6d54decef7f8..5652cc72ed3a 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c
@@ -707,7 +707,7 @@ static ssize_t amdgpu_debugfs_gpr_read(struct file *f, char __user *buf,
 	thread = (*pos & GENMASK_ULL(59, 52)) >> 52;
 	bank = (*pos & GENMASK_ULL(61, 60)) >> 60;
 
-	data = kmalloc_array(1024, sizeof(*data), GFP_KERNEL);
+	data = kcalloc(1024, sizeof(*data), GFP_KERNEL);
 	if (!data)
 		return -ENOMEM;
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gds.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_gds.h
index df8a23554831..f6ac1e9548f2 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gds.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gds.h
@@ -32,7 +32,6 @@ struct amdgpu_gds {
 	uint32_t gws_size;
 	uint32_t oa_size;
 	uint32_t gds_compute_max_wave_id;
-	uint32_t vgt_gs_max_wave_id;
 };
 
 struct amdgpu_gds_reg_offset {
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
index 03ca8c69114f..2b546567853b 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c
@@ -159,12 +159,16 @@ static ssize_t amdgpu_get_dpm_state(struct device *dev,
 	struct amdgpu_device *adev = ddev->dev_private;
 	enum amd_pm_state_type pm;
 
-	if (is_support_sw_smu(adev) && adev->smu.ppt_funcs->get_current_power_state)
-		pm = amdgpu_smu_get_current_power_state(adev);
-	else if (adev->powerplay.pp_funcs->get_current_power_state)
+	if (is_support_sw_smu(adev)) {
+		if (adev->smu.ppt_funcs->get_current_power_state)
+			pm = amdgpu_smu_get_current_power_state(adev);
+		else
+			pm = adev->pm.dpm.user_state;
+	} else if (adev->powerplay.pp_funcs->get_current_power_state) {
 		pm = amdgpu_dpm_get_current_power_state(adev);
-	else
+	} else {
 		pm = adev->pm.dpm.user_state;
+	}
 
 	return snprintf(buf, PAGE_SIZE, "%s\n",
 			(pm == POWER_STATE_TYPE_BATTERY) ? "battery" :
@@ -191,7 +195,11 @@ static ssize_t amdgpu_set_dpm_state(struct device *dev,
 		goto fail;
 	}
 
-	if (adev->powerplay.pp_funcs->dispatch_tasks) {
+	if (is_support_sw_smu(adev)) {
+		mutex_lock(&adev->pm.mutex);
+		adev->pm.dpm.user_state = state;
+		mutex_unlock(&adev->pm.mutex);
+	} else if (adev->powerplay.pp_funcs->dispatch_tasks) {
 		amdgpu_dpm_dispatch_task(adev, AMD_PP_TASK_ENABLE_USER_STATE, &state);
 	} else {
 		mutex_lock(&adev->pm.mutex);
@@ -3067,28 +3075,44 @@ static int amdgpu_debugfs_pm_info_pp(struct seq_file *m, struct amdgpu_device *a
 	if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_ENABLED_SMC_FEATURES_MASK, (void *)&value64, &size))
 		seq_printf(m, "SMC Feature Mask: 0x%016llx\n", value64);
 
-	/* UVD clocks */
-	if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_UVD_POWER, (void *)&value, &size)) {
-		if (!value) {
-			seq_printf(m, "UVD: Disabled\n");
-		} else {
-			seq_printf(m, "UVD: Enabled\n");
-			if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_UVD_DCLK, (void *)&value, &size))
-				seq_printf(m, "\t%u MHz (DCLK)\n", value/100);
-			if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_UVD_VCLK, (void *)&value, &size))
-				seq_printf(m, "\t%u MHz (VCLK)\n", value/100);
+	if (adev->asic_type > CHIP_VEGA20) {
+		/* VCN clocks */
+		if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_VCN_POWER_STATE, (void *)&value, &size)) {
+			if (!value) {
+				seq_printf(m, "VCN: Disabled\n");
+			} else {
+				seq_printf(m, "VCN: Enabled\n");
+				if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_UVD_DCLK, (void *)&value, &size))
+					seq_printf(m, "\t%u MHz (DCLK)\n", value/100);
+				if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_UVD_VCLK, (void *)&value, &size))
+					seq_printf(m, "\t%u MHz (VCLK)\n", value/100);
+			}
 		}
-	}
-	seq_printf(m, "\n");
+		seq_printf(m, "\n");
+	} else {
+		/* UVD clocks */
+		if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_UVD_POWER, (void *)&value, &size)) {
+			if (!value) {
+				seq_printf(m, "UVD: Disabled\n");
+			} else {
+				seq_printf(m, "UVD: Enabled\n");
+				if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_UVD_DCLK, (void *)&value, &size))
+					seq_printf(m, "\t%u MHz (DCLK)\n", value/100);
+				if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_UVD_VCLK, (void *)&value, &size))
+					seq_printf(m, "\t%u MHz (VCLK)\n", value/100);
+			}
+		}
+		seq_printf(m, "\n");
 
-	/* VCE clocks */
-	if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_VCE_POWER, (void *)&value, &size)) {
-		if (!value) {
-			seq_printf(m, "VCE: Disabled\n");
-		} else {
-			seq_printf(m, "VCE: Enabled\n");
-			if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_VCE_ECCLK, (void *)&value, &size))
-				seq_printf(m, "\t%u MHz (ECCLK)\n", value/100);
+		/* VCE clocks */
+		if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_VCE_POWER, (void *)&value, &size)) {
+			if (!value) {
+				seq_printf(m, "VCE: Disabled\n");
+			} else {
+				seq_printf(m, "VCE: Enabled\n");
+				if (!amdgpu_dpm_read_sensor(adev, AMDGPU_PP_SENSOR_VCE_ECCLK, (void *)&value, &size))
+					seq_printf(m, "\t%u MHz (ECCLK)\n", value/100);
+			}
 		}
 	}
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h
index 99f14fcc1460..19661c645703 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_vcn.h
@@ -30,6 +30,7 @@
 #define AMDGPU_VCN_FIRMWARE_OFFSET	256
 #define AMDGPU_VCN_MAX_ENC_RINGS	3
 
+#define VCN_DEC_KMD_CMD 		0x80000000
 #define VCN_DEC_CMD_FENCE		0x00000000
 #define VCN_DEC_CMD_TRAP		0x00000001
 #define VCN_DEC_CMD_WRITE_REG		0x00000004
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
index 32773b7523d2..f41287f9000d 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v10_0.c
@@ -4206,15 +4206,6 @@ static void gfx_v10_0_ring_emit_ib_gfx(struct amdgpu_ring *ring,
 	unsigned vmid = AMDGPU_JOB_GET_VMID(job);
 	u32 header, control = 0;
 
-	/* Prevent a hw deadlock due to a wave ID mismatch between ME and GDS.
-	 * This resets the wave ID counters. (needed by transform feedback)
-	 * TODO: This might only be needed on a VMID switch when we change
-	 *       the GDS OA mapping, not sure.
-	 */
-	amdgpu_ring_write(ring, PACKET3(PACKET3_SET_CONFIG_REG, 1));
-	amdgpu_ring_write(ring, mmVGT_GS_MAX_WAVE_ID);
-	amdgpu_ring_write(ring, ring->adev->gds.vgt_gs_max_wave_id);
-
 	if (ib->flags & AMDGPU_IB_FLAG_CE)
 		header = PACKET3(PACKET3_INDIRECT_BUFFER_CNST, 2);
 	else
@@ -4961,7 +4952,7 @@ static const struct amdgpu_ring_funcs gfx_v10_0_ring_funcs_gfx = {
 		5 + /* HDP_INVL */
 		8 + 8 + /* FENCE x2 */
 		2, /* SWITCH_BUFFER */
-	.emit_ib_size =	7, /* gfx_v10_0_ring_emit_ib_gfx */
+	.emit_ib_size =	4, /* gfx_v10_0_ring_emit_ib_gfx */
 	.emit_ib = gfx_v10_0_ring_emit_ib_gfx,
 	.emit_fence = gfx_v10_0_ring_emit_fence,
 	.emit_pipeline_sync = gfx_v10_0_ring_emit_pipeline_sync,
@@ -5112,7 +5103,6 @@ static void gfx_v10_0_set_gds_init(struct amdgpu_device *adev)
 	default:
 		adev->gds.gds_size = 0x10000;
 		adev->gds.gds_compute_max_wave_id = 0x4ff;
-		adev->gds.vgt_gs_max_wave_id = 0x3ff;
 		break;
 	}
 
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
index 751567f78567..ee1ccdcf2d30 100644
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
@@ -1321,6 +1321,39 @@ static int gfx_v8_0_rlc_init(struct amdgpu_device *adev)
 	return 0;
 }
 
+static int gfx_v8_0_csb_vram_pin(struct amdgpu_device *adev)
+{
+	int r;
+
+	r = amdgpu_bo_reserve(adev->gfx.rlc.clear_state_obj, false);
+	if (unlikely(r != 0))
+		return r;
+
+	r = amdgpu_bo_pin(adev->gfx.rlc.clear_state_obj,
+			AMDGPU_GEM_DOMAIN_VRAM);
+	if (!r)
+		adev->gfx.rlc.clear_state_gpu_addr =
+			amdgpu_bo_gpu_offset(adev->gfx.rlc.clear_state_obj);
+
+	amdgpu_bo_unreserve(adev->gfx.rlc.clear_state_obj);
+
+	return r;
+}
+
+static void gfx_v8_0_csb_vram_unpin(struct amdgpu_device *adev)
+{
+	int r;
+
+	if (!adev->gfx.rlc.clear_state_obj)
+		return;
+
+	r = amdgpu_bo_reserve(adev->gfx.rlc.clear_state_obj, true);
+	if (likely(r == 0)) {
+		amdgpu_bo_unpin(adev->gfx.rlc.clear_state_obj);
+		amdgpu_bo_unreserve(adev->gfx.rlc.clear_state_obj);
+	}
+}
+
 static void gfx_v8_0_mec_fini(struct amdgpu_device *adev)
 {
 	amdgpu_bo_free_kernel(&adev->gfx.mec.hpd_eop_obj, NULL, NULL);
@@ -4785,6 +4818,10 @@ static int gfx_v8_0_hw_init(void *handle)
 	gfx_v8_0_init_golden_registers(adev);
 	gfx_v8_0_constants_init(adev);
 
+	r = gfx_v8_0_csb_vram_pin(adev);
+	if (r)
+		return r;
+
 	r = adev->gfx.rlc.funcs->resume(adev);
 	if (r)
 		return r;
@@ -4901,6 +4938,9 @@ static int gfx_v8_0_hw_fini(void *handle)
 	else
 		pr_err("rlc is busy, skip halt rlc\n");
 	amdgpu_gfx_rlc_exit_safe_mode(adev);
+
+	gfx_v8_0_csb_vram_unpin(adev);
+
 	return 0;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c b/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c
index 1cfc2620b2dd..dfde886cc6bd 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v2_0.c
@@ -1485,7 +1485,7 @@ static void vcn_v2_0_dec_ring_insert_start(struct amdgpu_ring *ring)
 	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_DATA0_INTERNAL_OFFSET, 0));
 	amdgpu_ring_write(ring, 0);
 	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_CMD_INTERNAL_OFFSET, 0));
-	amdgpu_ring_write(ring, VCN_DEC_CMD_PACKET_START << 1);
+	amdgpu_ring_write(ring, VCN_DEC_KMD_CMD | (VCN_DEC_CMD_PACKET_START << 1));
 }
 
 /**
@@ -1498,7 +1498,7 @@ static void vcn_v2_0_dec_ring_insert_start(struct amdgpu_ring *ring)
 static void vcn_v2_0_dec_ring_insert_end(struct amdgpu_ring *ring)
 {
 	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_CMD_INTERNAL_OFFSET, 0));
-	amdgpu_ring_write(ring, VCN_DEC_CMD_PACKET_END << 1);
+	amdgpu_ring_write(ring, VCN_DEC_KMD_CMD | (VCN_DEC_CMD_PACKET_END << 1));
 }
 
 /**
@@ -1543,7 +1543,7 @@ static void vcn_v2_0_dec_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64
 	amdgpu_ring_write(ring, upper_32_bits(addr) & 0xff);
 
 	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_CMD_INTERNAL_OFFSET, 0));
-	amdgpu_ring_write(ring, VCN_DEC_CMD_FENCE << 1);
+	amdgpu_ring_write(ring, VCN_DEC_KMD_CMD | (VCN_DEC_CMD_FENCE << 1));
 
 	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_DATA0_INTERNAL_OFFSET, 0));
 	amdgpu_ring_write(ring, 0);
@@ -1553,7 +1553,7 @@ static void vcn_v2_0_dec_ring_emit_fence(struct amdgpu_ring *ring, u64 addr, u64
 
 	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_CMD_INTERNAL_OFFSET, 0));
 
-	amdgpu_ring_write(ring, VCN_DEC_CMD_TRAP << 1);
+	amdgpu_ring_write(ring, VCN_DEC_KMD_CMD | (VCN_DEC_CMD_TRAP << 1));
 }
 
 /**
@@ -1597,7 +1597,7 @@ static void vcn_v2_0_dec_ring_emit_reg_wait(struct amdgpu_ring *ring,
 
 	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_CMD_INTERNAL_OFFSET, 0));
 
-	amdgpu_ring_write(ring, VCN_DEC_CMD_REG_READ_COND_WAIT << 1);
+	amdgpu_ring_write(ring, VCN_DEC_KMD_CMD | (VCN_DEC_CMD_REG_READ_COND_WAIT << 1));
 }
 
 static void vcn_v2_0_dec_ring_emit_vm_flush(struct amdgpu_ring *ring,
@@ -1626,7 +1626,7 @@ static void vcn_v2_0_dec_ring_emit_wreg(struct amdgpu_ring *ring,
 
 	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_CMD_INTERNAL_OFFSET, 0));
 
-	amdgpu_ring_write(ring, VCN_DEC_CMD_WRITE_REG << 1);
+	amdgpu_ring_write(ring, VCN_DEC_KMD_CMD | (VCN_DEC_CMD_WRITE_REG << 1));
 }
 
 /**
@@ -2079,6 +2079,36 @@ static int vcn_v2_0_process_interrupt(struct amdgpu_device *adev,
 	return 0;
 }
 
+static int vcn_v2_0_dec_ring_test_ring(struct amdgpu_ring *ring)
+{
+	struct amdgpu_device *adev = ring->adev;
+	uint32_t tmp = 0;
+	unsigned i;
+	int r;
+
+	WREG32(adev->vcn.external.scratch9, 0xCAFEDEAD);
+	r = amdgpu_ring_alloc(ring, 4);
+	if (r)
+		return r;
+	amdgpu_ring_write(ring, PACKET0(mmUVD_GPCOM_VCPU_CMD_INTERNAL_OFFSET, 0));
+	amdgpu_ring_write(ring, VCN_DEC_KMD_CMD | (VCN_DEC_CMD_PACKET_START << 1));
+	amdgpu_ring_write(ring, PACKET0(adev->vcn.internal.scratch9, 0));
+	amdgpu_ring_write(ring, 0xDEADBEEF);
+	amdgpu_ring_commit(ring);
+	for (i = 0; i < adev->usec_timeout; i++) {
+		tmp = RREG32(adev->vcn.external.scratch9);
+		if (tmp == 0xDEADBEEF)
+			break;
+		DRM_UDELAY(1);
+	}
+
+	if (i >= adev->usec_timeout)
+		r = -ETIMEDOUT;
+
+	return r;
+}
+
+
 static int vcn_v2_0_set_powergating_state(void *handle,
 					  enum amd_powergating_state state)
 {
@@ -2142,7 +2172,7 @@ static const struct amdgpu_ring_funcs vcn_v2_0_dec_ring_vm_funcs = {
 	.emit_ib = vcn_v2_0_dec_ring_emit_ib,
 	.emit_fence = vcn_v2_0_dec_ring_emit_fence,
 	.emit_vm_flush = vcn_v2_0_dec_ring_emit_vm_flush,
-	.test_ring = amdgpu_vcn_dec_ring_test_ring,
+	.test_ring = vcn_v2_0_dec_ring_test_ring,
 	.test_ib = amdgpu_vcn_dec_ring_test_ib,
 	.insert_nop = vcn_v2_0_dec_ring_insert_nop,
 	.insert_start = vcn_v2_0_dec_ring_insert_start,