Merge tag 'amd-drm-fixes-6.13-2024-12-18' of https://gitlab.freedesktop.org/agd5f/linux into drm-fixes

amd-drm-fixes-6.13-2024-12-18:

amdgpu:
- Disable BOCO when CONFIG_HOTPLUG_PCI_PCIE is not enabled
- scheduler job fixes
- IP version check fixes
- devcoredump fix
- GPUVM update fix
- NBIO 2.5 fix

Signed-off-by: Dave Airlie <airlied@redhat.com>

From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20241218204637.2966198-1-alexander.deucher@amd.com

Author: airlied

drivers/gpu/drm/amd/amdgpu/amdgpu_dev_coredump.c

@@ -343,11 +343,10 @@ void amdgpu_coredump(struct amdgpu_device *adev, bool skip_vram_check,
 	coredump->skip_vram_check = skip_vram_check;
 	coredump->reset_vram_lost = vram_lost;
 
-	if (job && job->vm) {
-		struct amdgpu_vm *vm = job->vm;
+	if (job && job->pasid) {
 		struct amdgpu_task_info *ti;
 
-		ti = amdgpu_vm_get_task_info_vm(vm);
+		ti = amdgpu_vm_get_task_info_pasid(adev, job->pasid);
 		if (ti) {
 			coredump->reset_task_info = *ti;
 			amdgpu_vm_put_task_info(ti);

drivers/gpu/drm/amd/amdgpu/amdgpu_device.c

@@ -417,6 +417,9 @@ bool amdgpu_device_supports_boco(struct drm_device *dev)
 {
 	struct amdgpu_device *adev = drm_to_adev(dev);
 
+	if (!IS_ENABLED(CONFIG_HOTPLUG_PCI_PCIE))
+		return false;
+
 	if (adev->has_pr3 ||
 	    ((adev->flags & AMD_IS_PX) && amdgpu_is_atpx_hybrid()))
 		return true;

drivers/gpu/drm/amd/amdgpu/amdgpu_job.c

@@ -255,7 +255,6 @@ void amdgpu_job_set_resources(struct amdgpu_job *job, struct amdgpu_bo *gds,
 
 void amdgpu_job_free_resources(struct amdgpu_job *job)
 {
-	struct amdgpu_ring *ring = to_amdgpu_ring(job->base.sched);
 	struct dma_fence *f;
 	unsigned i;
 
@@ -268,7 +267,7 @@ void amdgpu_job_free_resources(struct amdgpu_job *job)
 		f = NULL;
 
 	for (i = 0; i < job->num_ibs; ++i)
-		amdgpu_ib_free(ring->adev, &job->ibs[i], f);
+		amdgpu_ib_free(NULL, &job->ibs[i], f);
 }
 
 static void amdgpu_job_free_cb(struct drm_sched_job *s_job)

drivers/gpu/drm/amd/amdgpu/amdgpu_vm.c

@@ -1266,10 +1266,9 @@ int amdgpu_vm_bo_update(struct amdgpu_device *adev, struct amdgpu_bo_va *bo_va,
 	 * next command submission.
 	 */
 	if (amdgpu_vm_is_bo_always_valid(vm, bo)) {
-		uint32_t mem_type = bo->tbo.resource->mem_type;
-
-		if (!(bo->preferred_domains &
-		      amdgpu_mem_type_to_domain(mem_type)))
+		if (bo->tbo.resource &&
+		    !(bo->preferred_domains &
+		      amdgpu_mem_type_to_domain(bo->tbo.resource->mem_type)))
 			amdgpu_vm_bo_evicted(&bo_va->base);
 		else
 			amdgpu_vm_bo_idle(&bo_va->base);

drivers/gpu/drm/amd/amdgpu/gfx_v12_0.c

@@ -4123,7 +4123,7 @@ static int gfx_v12_0_set_clockgating_state(void *handle,
 	if (amdgpu_sriov_vf(adev))
 		return 0;
 
-	switch (adev->ip_versions[GC_HWIP][0]) {
+	switch (amdgpu_ip_version(adev, GC_HWIP, 0)) {
 	case IP_VERSION(12, 0, 0):
 	case IP_VERSION(12, 0, 1):
 		gfx_v12_0_update_gfx_clock_gating(adev,

drivers/gpu/drm/amd/amdgpu/mmhub_v4_1_0.c

@@ -108,7 +108,7 @@ mmhub_v4_1_0_print_l2_protection_fault_status(struct amdgpu_device *adev,
 	dev_err(adev->dev,
 		"MMVM_L2_PROTECTION_FAULT_STATUS_LO32:0x%08X\n",
 		status);
-	switch (adev->ip_versions[MMHUB_HWIP][0]) {
+	switch (amdgpu_ip_version(adev, MMHUB_HWIP, 0)) {
 	case IP_VERSION(4, 1, 0):
 		mmhub_cid = mmhub_client_ids_v4_1_0[cid][rw];
 		break;

drivers/gpu/drm/amd/amdgpu/nbio_v7_0.c

@@ -271,8 +271,19 @@ const struct nbio_hdp_flush_reg nbio_v7_0_hdp_flush_reg = {
 	.ref_and_mask_sdma1 = GPU_HDP_FLUSH_DONE__SDMA1_MASK,
 };
 
+#define regRCC_DEV0_EPF6_STRAP4                                                                         0xd304
+#define regRCC_DEV0_EPF6_STRAP4_BASE_IDX                                                                5
+
 static void nbio_v7_0_init_registers(struct amdgpu_device *adev)
 {
+	uint32_t data;
+
+	switch (amdgpu_ip_version(adev, NBIO_HWIP, 0)) {
+	case IP_VERSION(2, 5, 0):
+		data = RREG32_SOC15(NBIO, 0, regRCC_DEV0_EPF6_STRAP4) & ~BIT(23);
+		WREG32_SOC15(NBIO, 0, regRCC_DEV0_EPF6_STRAP4, data);
+		break;
+	}
 }
 
 #define MMIO_REG_HOLE_OFFSET (0x80000 - PAGE_SIZE)

drivers/gpu/drm/amd/amdgpu/nbio_v7_11.c

@@ -275,7 +275,7 @@ static void nbio_v7_11_init_registers(struct amdgpu_device *adev)
 	if (def != data)
 		WREG32_SOC15(NBIO, 0, regBIF_BIF256_CI256_RC3X4_USB4_PCIE_MST_CTRL_3, data);
 
-	switch (adev->ip_versions[NBIO_HWIP][0]) {
+	switch (amdgpu_ip_version(adev, NBIO_HWIP, 0)) {
 	case IP_VERSION(7, 11, 0):
 	case IP_VERSION(7, 11, 1):
 	case IP_VERSION(7, 11, 2):

drivers/gpu/drm/amd/amdgpu/nbio_v7_7.c

@@ -247,7 +247,7 @@ static void nbio_v7_7_init_registers(struct amdgpu_device *adev)
 	if (def != data)
 		WREG32_SOC15(NBIO, 0, regBIF0_PCIE_MST_CTRL_3, data);
 
-	switch (adev->ip_versions[NBIO_HWIP][0]) {
+	switch (amdgpu_ip_version(adev, NBIO_HWIP, 0)) {
 	case IP_VERSION(7, 7, 0):
 		data = RREG32_SOC15(NBIO, 0, regRCC_DEV0_EPF5_STRAP4) & ~BIT(23);
 		WREG32_SOC15(NBIO, 0, regRCC_DEV0_EPF5_STRAP4, data);

drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_2_ppt.c

@@ -2096,7 +2096,7 @@ static int smu_v14_0_2_enable_gfx_features(struct smu_context *smu)
 {
 	struct amdgpu_device *adev = smu->adev;
 
-	if (adev->ip_versions[MP1_HWIP][0] == IP_VERSION(14, 0, 2))
+	if (amdgpu_ip_version(adev, MP1_HWIP, 0) == IP_VERSION(14, 0, 2))
 		return smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_EnableAllSmuFeatures,
 										   FEATURE_PWR_GFX, NULL);
 	else