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.c158
1 files changed, 78 insertions, 80 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
index 632a6ded5735..5c38f0d30c87 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_uvd.c
@@ -96,16 +96,16 @@
*/
struct amdgpu_uvd_cs_ctx {
struct amdgpu_cs_parser *parser;
- unsigned reg, count;
- unsigned data0, data1;
- unsigned idx;
+ unsigned int reg, count;
+ unsigned int data0, data1;
+ unsigned int idx;
struct amdgpu_ib *ib;
/* does the IB has a msg command */
bool has_msg_cmd;
/* minimum buffer sizes */
- unsigned *buf_sizes;
+ unsigned int *buf_sizes;
};
#ifdef CONFIG_DRM_AMDGPU_SI
@@ -186,7 +186,7 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
unsigned long bo_size;
const char *fw_name;
const struct common_firmware_header *hdr;
- unsigned family_id;
+ unsigned int family_id;
int i, j, r;
INIT_DELAYED_WORK(&adev->uvd.idle_work, amdgpu_uvd_idle_work_handler);
@@ -260,7 +260,7 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
return -EINVAL;
}
- r = amdgpu_ucode_request(adev, &adev->uvd.fw, fw_name);
+ r = amdgpu_ucode_request(adev, &adev->uvd.fw, AMDGPU_UCODE_REQUIRED, "%s", fw_name);
if (r) {
dev_err(adev->dev, "amdgpu_uvd: Can't validate firmware \"%s\"\n",
fw_name);
@@ -275,7 +275,7 @@ int amdgpu_uvd_sw_init(struct amdgpu_device *adev)
family_id = le32_to_cpu(hdr->ucode_version) & 0xff;
if (adev->asic_type < CHIP_VEGA20) {
- unsigned version_major, version_minor;
+ unsigned int version_major, version_minor;
version_major = (le32_to_cpu(hdr->ucode_version) >> 24) & 0xff;
version_minor = (le32_to_cpu(hdr->ucode_version) >> 8) & 0xff;
@@ -398,32 +398,32 @@ int amdgpu_uvd_sw_fini(struct amdgpu_device *adev)
* amdgpu_uvd_entity_init - init entity
*
* @adev: amdgpu_device pointer
+ * @ring: amdgpu_ring pointer to check
*
+ * Initialize the entity used for handle management in the kernel driver.
*/
-int amdgpu_uvd_entity_init(struct amdgpu_device *adev)
+int amdgpu_uvd_entity_init(struct amdgpu_device *adev, struct amdgpu_ring *ring)
{
- struct amdgpu_ring *ring;
- struct drm_gpu_scheduler *sched;
- int r;
+ if (ring == &adev->uvd.inst[0].ring) {
+ struct drm_gpu_scheduler *sched = &ring->sched;
+ int r;
- ring = &adev->uvd.inst[0].ring;
- sched = &ring->sched;
- r = drm_sched_entity_init(&adev->uvd.entity, DRM_SCHED_PRIORITY_NORMAL,
- &sched, 1, NULL);
- if (r) {
- DRM_ERROR("Failed setting up UVD kernel entity.\n");
- return r;
+ r = drm_sched_entity_init(&adev->uvd.entity, DRM_SCHED_PRIORITY_NORMAL,
+ &sched, 1, NULL);
+ if (r) {
+ DRM_ERROR("Failed setting up UVD kernel entity.\n");
+ return r;
+ }
}
return 0;
}
-int amdgpu_uvd_suspend(struct amdgpu_device *adev)
+int amdgpu_uvd_prepare_suspend(struct amdgpu_device *adev)
{
- unsigned size;
+ unsigned int size;
void *ptr;
int i, j, idx;
- bool in_ras_intr = amdgpu_ras_intr_triggered();
cancel_delayed_work_sync(&adev->uvd.idle_work);
@@ -452,7 +452,7 @@ int amdgpu_uvd_suspend(struct amdgpu_device *adev)
if (drm_dev_enter(adev_to_drm(adev), &idx)) {
/* re-write 0 since err_event_athub will corrupt VCPU buffer */
- if (in_ras_intr)
+ if (amdgpu_ras_intr_triggered())
memset(adev->uvd.inst[j].saved_bo, 0, size);
else
memcpy_fromio(adev->uvd.inst[j].saved_bo, ptr, size);
@@ -461,7 +461,12 @@ int amdgpu_uvd_suspend(struct amdgpu_device *adev)
}
}
- if (in_ras_intr)
+ return 0;
+}
+
+int amdgpu_uvd_suspend(struct amdgpu_device *adev)
+{
+ if (amdgpu_ras_intr_triggered())
DRM_WARN("UVD VCPU state may lost due to RAS ERREVENT_ATHUB_INTERRUPT\n");
return 0;
@@ -469,7 +474,7 @@ int amdgpu_uvd_suspend(struct amdgpu_device *adev)
int amdgpu_uvd_resume(struct amdgpu_device *adev)
{
- unsigned size;
+ unsigned int size;
void *ptr;
int i, idx;
@@ -491,7 +496,7 @@ int amdgpu_uvd_resume(struct amdgpu_device *adev)
adev->uvd.inst[i].saved_bo = NULL;
} else {
const struct common_firmware_header *hdr;
- unsigned offset;
+ unsigned int offset;
hdr = (const struct common_firmware_header *)adev->uvd.fw->data;
if (adev->firmware.load_type != AMDGPU_FW_LOAD_PSP) {
@@ -542,9 +547,12 @@ void amdgpu_uvd_free_handles(struct amdgpu_device *adev, struct drm_file *filp)
static void amdgpu_uvd_force_into_uvd_segment(struct amdgpu_bo *abo)
{
int i;
+
for (i = 0; i < abo->placement.num_placement; ++i) {
abo->placements[i].fpfn = 0 >> PAGE_SHIFT;
abo->placements[i].lpfn = (256 * 1024 * 1024) >> PAGE_SHIFT;
+ if (abo->placements[i].mem_type == TTM_PL_VRAM)
+ abo->placements[i].flags |= TTM_PL_FLAG_CONTIGUOUS;
}
}
@@ -579,7 +587,7 @@ static int amdgpu_uvd_cs_pass1(struct amdgpu_uvd_cs_ctx *ctx)
r = amdgpu_cs_find_mapping(ctx->parser, addr, &bo, &mapping);
if (r) {
- DRM_ERROR("Can't find BO for addr 0x%08Lx\n", addr);
+ DRM_ERROR("Can't find BO for addr 0x%08llx\n", addr);
return r;
}
@@ -589,6 +597,7 @@ static int amdgpu_uvd_cs_pass1(struct amdgpu_uvd_cs_ctx *ctx)
if (cmd == 0x0 || cmd == 0x3) {
/* yes, force it into VRAM */
uint32_t domain = AMDGPU_GEM_DOMAIN_VRAM;
+
amdgpu_bo_placement_from_domain(bo, domain);
}
amdgpu_uvd_force_into_uvd_segment(bo);
@@ -609,21 +618,21 @@ static int amdgpu_uvd_cs_pass1(struct amdgpu_uvd_cs_ctx *ctx)
* Peek into the decode message and calculate the necessary buffer sizes.
*/
static int amdgpu_uvd_cs_msg_decode(struct amdgpu_device *adev, uint32_t *msg,
- unsigned buf_sizes[])
+ unsigned int buf_sizes[])
{
- unsigned stream_type = msg[4];
- unsigned width = msg[6];
- unsigned height = msg[7];
- unsigned dpb_size = msg[9];
- unsigned pitch = msg[28];
- unsigned level = msg[57];
+ unsigned int stream_type = msg[4];
+ unsigned int width = msg[6];
+ unsigned int height = msg[7];
+ unsigned int dpb_size = msg[9];
+ unsigned int pitch = msg[28];
+ unsigned int level = msg[57];
- unsigned width_in_mb = width / 16;
- unsigned height_in_mb = ALIGN(height / 16, 2);
- unsigned fs_in_mb = width_in_mb * height_in_mb;
+ unsigned int width_in_mb = width / 16;
+ unsigned int height_in_mb = ALIGN(height / 16, 2);
+ unsigned int fs_in_mb = width_in_mb * height_in_mb;
- unsigned image_size, tmp, min_dpb_size, num_dpb_buffer;
- unsigned min_ctx_size = ~0;
+ unsigned int image_size, tmp, min_dpb_size, num_dpb_buffer;
+ unsigned int min_ctx_size = ~0;
image_size = width * height;
image_size += image_size / 2;
@@ -631,7 +640,7 @@ static int amdgpu_uvd_cs_msg_decode(struct amdgpu_device *adev, uint32_t *msg,
switch (stream_type) {
case 0: /* H264 */
- switch(level) {
+ switch (level) {
case 30:
num_dpb_buffer = 8100 / fs_in_mb;
break;
@@ -709,7 +718,7 @@ static int amdgpu_uvd_cs_msg_decode(struct amdgpu_device *adev, uint32_t *msg,
break;
case 7: /* H264 Perf */
- switch(level) {
+ switch (level) {
case 30:
num_dpb_buffer = 8100 / fs_in_mb;
break;
@@ -742,7 +751,7 @@ static int amdgpu_uvd_cs_msg_decode(struct amdgpu_device *adev, uint32_t *msg,
/* reference picture buffer */
min_dpb_size = image_size * num_dpb_buffer;
- if (!adev->uvd.use_ctx_buf){
+ if (!adev->uvd.use_ctx_buf) {
/* macroblock context buffer */
min_dpb_size +=
width_in_mb * height_in_mb * num_dpb_buffer * 192;
@@ -805,7 +814,7 @@ static int amdgpu_uvd_cs_msg_decode(struct amdgpu_device *adev, uint32_t *msg,
* Make sure that we don't open up to many sessions.
*/
static int amdgpu_uvd_cs_msg(struct amdgpu_uvd_cs_ctx *ctx,
- struct amdgpu_bo *bo, unsigned offset)
+ struct amdgpu_bo *bo, unsigned int offset)
{
struct amdgpu_device *adev = ctx->parser->adev;
int32_t *msg, msg_type, handle;
@@ -911,7 +920,7 @@ static int amdgpu_uvd_cs_pass2(struct amdgpu_uvd_cs_ctx *ctx)
r = amdgpu_cs_find_mapping(ctx->parser, addr, &bo, &mapping);
if (r) {
- DRM_ERROR("Can't find BO for addr 0x%08Lx\n", addr);
+ DRM_ERROR("Can't find BO for addr 0x%08llx\n", addr);
return r;
}
@@ -930,7 +939,7 @@ static int amdgpu_uvd_cs_pass2(struct amdgpu_uvd_cs_ctx *ctx)
if (cmd < 0x4) {
if ((end - start) < ctx->buf_sizes[cmd]) {
DRM_ERROR("buffer (%d) to small (%d / %d)!\n", cmd,
- (unsigned)(end - start),
+ (unsigned int)(end - start),
ctx->buf_sizes[cmd]);
return -EINVAL;
}
@@ -938,7 +947,7 @@ static int amdgpu_uvd_cs_pass2(struct amdgpu_uvd_cs_ctx *ctx)
} else if (cmd == 0x206) {
if ((end - start) < ctx->buf_sizes[4]) {
DRM_ERROR("buffer (%d) to small (%d / %d)!\n", cmd,
- (unsigned)(end - start),
+ (unsigned int)(end - start),
ctx->buf_sizes[4]);
return -EINVAL;
}
@@ -949,14 +958,14 @@ static int amdgpu_uvd_cs_pass2(struct amdgpu_uvd_cs_ctx *ctx)
if (!ctx->parser->adev->uvd.address_64_bit) {
if ((start >> 28) != ((end - 1) >> 28)) {
- DRM_ERROR("reloc %LX-%LX crossing 256MB boundary!\n",
+ DRM_ERROR("reloc %llx-%llx crossing 256MB boundary!\n",
start, end);
return -EINVAL;
}
if ((cmd == 0 || cmd == 0x3) &&
(start >> 28) != (ctx->parser->adev->uvd.inst->gpu_addr >> 28)) {
- DRM_ERROR("msg/fb buffer %LX-%LX out of 256MB segment!\n",
+ DRM_ERROR("msg/fb buffer %llx-%llx out of 256MB segment!\n",
start, end);
return -EINVAL;
}
@@ -990,7 +999,7 @@ static int amdgpu_uvd_cs_reg(struct amdgpu_uvd_cs_ctx *ctx,
ctx->idx++;
for (i = 0; i <= ctx->count; ++i) {
- unsigned reg = ctx->reg + i;
+ unsigned int reg = ctx->reg + i;
if (ctx->idx >= ctx->ib->length_dw) {
DRM_ERROR("Register command after end of CS!\n");
@@ -1036,7 +1045,8 @@ static int amdgpu_uvd_cs_packets(struct amdgpu_uvd_cs_ctx *ctx,
for (ctx->idx = 0 ; ctx->idx < ctx->ib->length_dw; ) {
uint32_t cmd = amdgpu_ib_get_value(ctx->ib, ctx->idx);
- unsigned type = CP_PACKET_GET_TYPE(cmd);
+ unsigned int type = CP_PACKET_GET_TYPE(cmd);
+
switch (type) {
case PACKET_TYPE0:
ctx->reg = CP_PACKET0_GET_REG(cmd);
@@ -1070,7 +1080,7 @@ int amdgpu_uvd_ring_parse_cs(struct amdgpu_cs_parser *parser,
struct amdgpu_ib *ib)
{
struct amdgpu_uvd_cs_ctx ctx = {};
- unsigned buf_sizes[] = {
+ unsigned int buf_sizes[] = {
[0x00000000] = 2048,
[0x00000001] = 0xFFFFFFFF,
[0x00000002] = 0xFFFFFFFF,
@@ -1080,7 +1090,6 @@ int amdgpu_uvd_ring_parse_cs(struct amdgpu_cs_parser *parser,
int r;
job->vm = NULL;
- ib->gpu_addr = amdgpu_sa_bo_gpu_addr(ib->sa_bo);
if (ib->length_dw % 16) {
DRM_ERROR("UVD IB length (%d) not 16 dwords aligned!\n",
@@ -1118,32 +1127,29 @@ static int amdgpu_uvd_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo,
{
struct amdgpu_device *adev = ring->adev;
struct dma_fence *f = NULL;
+ uint32_t offset, data[4];
struct amdgpu_job *job;
struct amdgpu_ib *ib;
- uint32_t data[4];
uint64_t addr;
- long r;
- int i;
- unsigned offset_idx = 0;
- unsigned offset[3] = { UVD_BASE_SI, 0, 0 };
+ int i, r;
r = amdgpu_job_alloc_with_ib(ring->adev, &adev->uvd.entity,
AMDGPU_FENCE_OWNER_UNDEFINED,
64, direct ? AMDGPU_IB_POOL_DIRECT :
- AMDGPU_IB_POOL_DELAYED, &job);
+ AMDGPU_IB_POOL_DELAYED, &job,
+ AMDGPU_KERNEL_JOB_ID_VCN_RING_TEST);
if (r)
return r;
- if (adev->asic_type >= CHIP_VEGA10) {
- offset_idx = 1 + ring->me;
- offset[1] = adev->reg_offset[UVD_HWIP][0][1];
- offset[2] = adev->reg_offset[UVD_HWIP][1][1];
- }
+ if (adev->asic_type >= CHIP_VEGA10)
+ offset = adev->reg_offset[UVD_HWIP][ring->me][1];
+ else
+ offset = UVD_BASE_SI;
- data[0] = PACKET0(offset[offset_idx] + UVD_GPCOM_VCPU_DATA0, 0);
- data[1] = PACKET0(offset[offset_idx] + UVD_GPCOM_VCPU_DATA1, 0);
- data[2] = PACKET0(offset[offset_idx] + UVD_GPCOM_VCPU_CMD, 0);
- data[3] = PACKET0(offset[offset_idx] + UVD_NO_OP, 0);
+ data[0] = PACKET0(offset + UVD_GPCOM_VCPU_DATA0, 0);
+ data[1] = PACKET0(offset + UVD_GPCOM_VCPU_DATA1, 0);
+ data[2] = PACKET0(offset + UVD_GPCOM_VCPU_CMD, 0);
+ data[3] = PACKET0(offset + UVD_NO_OP, 0);
ib = &job->ibs[0];
addr = amdgpu_bo_gpu_offset(bo);
@@ -1160,14 +1166,6 @@ static int amdgpu_uvd_send_msg(struct amdgpu_ring *ring, struct amdgpu_bo *bo,
ib->length_dw = 16;
if (direct) {
- r = dma_resv_wait_timeout(bo->tbo.base.resv,
- DMA_RESV_USAGE_KERNEL, false,
- msecs_to_jiffies(10));
- if (r == 0)
- r = -ETIMEDOUT;
- if (r < 0)
- goto err_free;
-
r = amdgpu_job_submit_direct(job, ring, &f);
if (r)
goto err_free;
@@ -1197,8 +1195,9 @@ err_free:
}
/* multiple fence commands without any stream commands in between can
- crash the vcpu so just try to emmit a dummy create/destroy msg to
- avoid this */
+ * crash the vcpu so just try to emmit a dummy create/destroy msg to
+ * avoid this
+ */
int amdgpu_uvd_get_create_msg(struct amdgpu_ring *ring, uint32_t handle,
struct dma_fence **fence)
{
@@ -1264,15 +1263,14 @@ static void amdgpu_uvd_idle_work_handler(struct work_struct *work)
{
struct amdgpu_device *adev =
container_of(work, struct amdgpu_device, uvd.idle_work.work);
- unsigned fences = 0, i, j;
+ unsigned int fences = 0, i, j;
for (i = 0; i < adev->uvd.num_uvd_inst; ++i) {
if (adev->uvd.harvest_config & (1 << i))
continue;
fences += amdgpu_fence_count_emitted(&adev->uvd.inst[i].ring);
- for (j = 0; j < adev->uvd.num_enc_rings; ++j) {
+ for (j = 0; j < adev->uvd.num_enc_rings; ++j)
fences += amdgpu_fence_count_emitted(&adev->uvd.inst[i].ring_enc[j]);
- }
}
if (fences == 0) {
@@ -1368,7 +1366,7 @@ error:
*/
uint32_t amdgpu_uvd_used_handles(struct amdgpu_device *adev)
{
- unsigned i;
+ unsigned int i;
uint32_t used_handles = 0;
for (i = 0; i < adev->uvd.max_handles; ++i) {
generated by cgit (git 2.34.1) at 2025-12-21 18:31:57 +0000