summaryrefslogtreecommitdiff
path: root/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c')
-rw-r--r--drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c119
1 files changed, 82 insertions, 37 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
index d451c359606a..6f8de11a17f1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
@@ -134,6 +134,51 @@ MODULE_FIRMWARE(FIRMWARE_VEGA12);
MODULE_FIRMWARE(FIRMWARE_VEGA20);
static void amdgpu_uvd_idle_work_handler(struct work_struct *work);
+static void amdgpu_uvd_force_into_uvd_segment(struct amdgpu_bo *abo);
+
+static int amdgpu_uvd_create_msg_bo_helper(struct amdgpu_device *adev,
+ uint32_t size,
+ struct amdgpu_bo **bo_ptr)
+{
+ struct ttm_operation_ctx ctx = { true, false };
+ struct amdgpu_bo *bo = NULL;
+ void *addr;
+ int r;
+
+ r = amdgpu_bo_create_reserved(adev, size, PAGE_SIZE,
+ AMDGPU_GEM_DOMAIN_GTT,
+ &bo, NULL, &addr);
+ if (r)
+ return r;
+
+ if (adev->uvd.address_64_bit)
+ goto succ;
+
+ amdgpu_bo_kunmap(bo);
+ amdgpu_bo_unpin(bo);
+ amdgpu_bo_placement_from_domain(bo, AMDGPU_GEM_DOMAIN_VRAM);
+ amdgpu_uvd_force_into_uvd_segment(bo);
+ r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
+ if (r)
+ goto err;
+ r = amdgpu_bo_pin(bo, AMDGPU_GEM_DOMAIN_VRAM);
+ if (r)
+ goto err_pin;
+ r = amdgpu_bo_kmap(bo, &addr);
+ if (r)
+ goto err_kmap;
+succ:
+ amdgpu_bo_unreserve(bo);
+ *bo_ptr = bo;
+ return 0;
+err_kmap:
+ amdgpu_bo_unpin(bo);
+err_pin:
+err:
+ amdgpu_bo_unreserve(bo);
+ amdgpu_bo_unref(&bo);
+ return r;
+}
int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
{
@@ -302,6 +347,10 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
if (!amdgpu_device_ip_block_version_cmp(adev, AMD_IP_BLOCK_TYPE_UVD, 5, 0))
adev->uvd.address_64_bit = true;
+ r = amdgpu_uvd_create_msg_bo_helper(adev, 128 << 10, &adev->uvd.ib_bo);
+ if (r)
+ return r;
+
switch (adev->asic_type) {
case CHIP_TONGA:
adev->uvd.use_ctx_buf = adev->uvd.fw_version >= FW_1_65_10;
@@ -324,6 +373,7 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
int amdgpu_uvd_sw_fini(struct amdgpu_device *adev)
{
+ void *addr = amdgpu_bo_kptr(adev->uvd.ib_bo);
int i, j;
drm_sched_entity_destroy(&adev->uvd.entity);
@@ -342,6 +392,7 @@ int amdgpu_uvd_sw_fini(struct amdgpu_device *adev)
for (i = 0; i < AMDGPU_MAX_UVD_ENC_RINGS; ++i)
amdgpu_ring_fini(&adev->uvd.inst[j].ring_enc[i]);
}
+ amdgpu_bo_free_kernel(&adev->uvd.ib_bo, NULL, &addr);
release_firmware(adev->uvd.fw);
return 0;
@@ -403,7 +454,7 @@ int amdgpu_uvd_suspend(struct amdgpu_device *adev)
if (!adev->uvd.inst[j].saved_bo)
return -ENOMEM;
- if (drm_dev_enter(&adev->ddev, &idx)) {
+ if (drm_dev_enter(adev_to_drm(adev), &idx)) {
/* re-write 0 since err_event_athub will corrupt VCPU buffer */
if (in_ras_intr)
memset(adev->uvd.inst[j].saved_bo, 0, size);
@@ -436,7 +487,7 @@ int amdgpu_uvd_resume(struct amdgpu_device *adev)
ptr = adev->uvd.inst[i].cpu_addr;
if (adev->uvd.inst[i].saved_bo != NULL) {
- if (drm_dev_enter(&adev->ddev, &idx)) {
+ if (drm_dev_enter(adev_to_drm(adev), &idx)) {
memcpy_toio(ptr, adev->uvd.inst[i].saved_bo, size);
drm_dev_exit(idx);
}
@@ -449,7 +500,7 @@ int amdgpu_uvd_resume(struct amdgpu_device *adev)
hdr = (const struct common_firmware_header *)adev->uvd.fw->data;
if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) {
offset = le32_to_cpu(hdr->ucode_array_offset_bytes);
- if (drm_dev_enter(&adev->ddev, &idx)) {
+ if (drm_dev_enter(adev_to_drm(adev), &idx)) {
memcpy_toio(adev->uvd.inst[i].cpu_addr, adev->uvd.fw->data + offset,
le32_to_cpu(hdr->ucode_size_bytes));
drm_dev_exit(idx);
@@ -1080,23 +1131,10 @@ static int amdgpu_uvd_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo,
unsigned offset_idx = 0;
unsigned offset[3] = { UVD_BASE_SI, 0, 0 };
- amdgpu_bo_kunmap(bo);
- amdgpu_bo_unpin(bo);
-
- if (!ring->adev->uvd.address_64_bit) {
- struct ttm_operation_ctx ctx = { true, false };
-
- amdgpu_bo_placement_from_domain(bo, AMDGPU_GEM_DOMAIN_VRAM);
- amdgpu_uvd_force_into_uvd_segment(bo);
- r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
- if (r)
- goto err;
- }
-
r = amdgpu_job_alloc_with_ib(adev, 64, direct ? AMDGPU_IB_POOL_DIRECT :
AMDGPU_IB_POOL_DELAYED, &job);
if (r)
- goto err;
+ return r;
if (adev->asic_type >= CHIP_VEGA10) {
offset_idx = 1 + ring->me;
@@ -1147,9 +1185,9 @@ static int amdgpu_uvd_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo,
goto err_free;
}
+ amdgpu_bo_reserve(bo, true);
amdgpu_bo_fence(bo, f, false);
amdgpu_bo_unreserve(bo);
- amdgpu_bo_unref(&bo);
if (fence)
*fence = dma_fence_get(f);
@@ -1159,10 +1197,6 @@ static int amdgpu_uvd_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo,
err_free:
amdgpu_job_free(job);
-
-err:
- amdgpu_bo_unreserve(bo);
- amdgpu_bo_unref(&bo);
return r;
}
@@ -1173,16 +1207,11 @@ int amdgpu_uvd_get_create_msg(struct amdgpu_ring *ring, uint32_t handle,
struct dma_fence **fence)
{
struct amdgpu_device *adev = ring->adev;
- struct amdgpu_bo *bo = NULL;
+ struct amdgpu_bo *bo = adev->uvd.ib_bo;
uint32_t *msg;
- int r, i;
-
- r = amdgpu_bo_create_reserved(adev, 1024, PAGE_SIZE,
- AMDGPU_GEM_DOMAIN_GTT,
- &bo, NULL, (void **)&msg);
- if (r)
- return r;
+ int i;
+ msg = amdgpu_bo_kptr(bo);
/* stitch together an UVD create msg */
msg[0] = cpu_to_le32(0x00000de4);
msg[1] = cpu_to_le32(0x00000000);
@@ -1199,6 +1228,7 @@ int amdgpu_uvd_get_create_msg(struct amdgpu_ring *ring, uint32_t handle,
msg[i] = cpu_to_le32(0x0);
return amdgpu_uvd_send_msg(ring, bo, true, fence);
+
}
int amdgpu_uvd_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle,
@@ -1209,12 +1239,15 @@ int amdgpu_uvd_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle,
uint32_t *msg;
int r, i;
- r = amdgpu_bo_create_reserved(adev, 1024, PAGE_SIZE,
- AMDGPU_GEM_DOMAIN_GTT,
- &bo, NULL, (void **)&msg);
- if (r)
- return r;
+ if (direct) {
+ bo = adev->uvd.ib_bo;
+ } else {
+ r = amdgpu_uvd_create_msg_bo_helper(adev, 4096, &bo);
+ if (r)
+ return r;
+ }
+ msg = amdgpu_bo_kptr(bo);
/* stitch together an UVD destroy msg */
msg[0] = cpu_to_le32(0x00000de4);
msg[1] = cpu_to_le32(0x00000002);
@@ -1223,7 +1256,12 @@ int amdgpu_uvd_get_destroy_msg(struct amdgpu_ring *ring, uint32_t handle,
for (i = 4; i < 1024; ++i)
msg[i] = cpu_to_le32(0x0);
- return amdgpu_uvd_send_msg(ring, bo, direct, fence);
+ r = amdgpu_uvd_send_msg(ring, bo, direct, fence);
+
+ if (!direct)
+ amdgpu_bo_free_kernel(&bo, NULL, (void **)&msg);
+
+ return r;
}
static void amdgpu_uvd_idle_work_handler(struct work_struct *work)
@@ -1298,10 +1336,17 @@ int amdgpu_uvd_ring_test_ib(struct amdgpu_ring *ring, long timeout)
struct dma_fence *fence;
long r;
- r = amdgpu_uvd_get_create_msg(ring, 1, NULL);
+ r = amdgpu_uvd_get_create_msg(ring, 1, &fence);
if (r)
goto error;
+ r = dma_fence_wait_timeout(fence, false, timeout);
+ dma_fence_put(fence);
+ if (r == 0)
+ r = -ETIMEDOUT;
+ if (r < 0)
+ goto error;
+
r = amdgpu_uvd_get_destroy_msg(ring, 1, true, &fence);
if (r)
goto error;
generated by cgit v1.2.3 (git 2.39.0) at 2025-01-09 18:24:19 +0300