linux/drivers/gpu/drm/radeon/radeon_uvd.c
Christian König 55b51c88c5 drm/radeon: raise UVD clocks only on demand
That not only saves some power, but also solves problems with
older chips where an idle UVD block on higher clocks can
cause problems.

Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2013-04-23 17:45:44 -04:00

695 lines
16 KiB
C

/*
* Copyright 2011 Advanced Micro Devices, Inc.
* All Rights Reserved.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and associated documentation files (the
* "Software"), to deal in the Software without restriction, including
* without limitation the rights to use, copy, modify, merge, publish,
* distribute, sub license, and/or sell copies of the Software, and to
* permit persons to whom the Software is furnished to do so, subject to
* the following conditions:
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
* THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
* DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
* OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
* USE OR OTHER DEALINGS IN THE SOFTWARE.
*
* The above copyright notice and this permission notice (including the
* next paragraph) shall be included in all copies or substantial portions
* of the Software.
*
*/
/*
* Authors:
* Christian König <deathsimple@vodafone.de>
*/
#include <linux/firmware.h>
#include <linux/module.h>
#include <drm/drmP.h>
#include <drm/drm.h>
#include "radeon.h"
#include "r600d.h"
/* 1 second timeout */
#define UVD_IDLE_TIMEOUT_MS 1000
/* Firmware Names */
#define FIRMWARE_RV710 "radeon/RV710_uvd.bin"
#define FIRMWARE_CYPRESS "radeon/CYPRESS_uvd.bin"
#define FIRMWARE_SUMO "radeon/SUMO_uvd.bin"
#define FIRMWARE_TAHITI "radeon/TAHITI_uvd.bin"
MODULE_FIRMWARE(FIRMWARE_RV710);
MODULE_FIRMWARE(FIRMWARE_CYPRESS);
MODULE_FIRMWARE(FIRMWARE_SUMO);
MODULE_FIRMWARE(FIRMWARE_TAHITI);
static void radeon_uvd_idle_work_handler(struct work_struct *work);
int radeon_uvd_init(struct radeon_device *rdev)
{
struct platform_device *pdev;
unsigned long bo_size;
const char *fw_name;
int i, r;
INIT_DELAYED_WORK(&rdev->uvd.idle_work, radeon_uvd_idle_work_handler);
pdev = platform_device_register_simple("radeon_uvd", 0, NULL, 0);
r = IS_ERR(pdev);
if (r) {
dev_err(rdev->dev, "radeon_uvd: Failed to register firmware\n");
return -EINVAL;
}
switch (rdev->family) {
case CHIP_RV710:
case CHIP_RV730:
case CHIP_RV740:
fw_name = FIRMWARE_RV710;
break;
case CHIP_CYPRESS:
case CHIP_HEMLOCK:
case CHIP_JUNIPER:
case CHIP_REDWOOD:
case CHIP_CEDAR:
fw_name = FIRMWARE_CYPRESS;
break;
case CHIP_SUMO:
case CHIP_SUMO2:
case CHIP_PALM:
case CHIP_CAYMAN:
case CHIP_BARTS:
case CHIP_TURKS:
case CHIP_CAICOS:
fw_name = FIRMWARE_SUMO;
break;
case CHIP_TAHITI:
case CHIP_VERDE:
case CHIP_PITCAIRN:
case CHIP_ARUBA:
fw_name = FIRMWARE_TAHITI;
break;
default:
return -EINVAL;
}
r = request_firmware(&rdev->uvd_fw, fw_name, &pdev->dev);
if (r) {
dev_err(rdev->dev, "radeon_uvd: Can't load firmware \"%s\"\n",
fw_name);
platform_device_unregister(pdev);
return r;
}
platform_device_unregister(pdev);
bo_size = RADEON_GPU_PAGE_ALIGN(rdev->uvd_fw->size + 8) +
RADEON_UVD_STACK_SIZE + RADEON_UVD_HEAP_SIZE;
r = radeon_bo_create(rdev, bo_size, PAGE_SIZE, true,
RADEON_GEM_DOMAIN_VRAM, NULL, &rdev->uvd.vcpu_bo);
if (r) {
dev_err(rdev->dev, "(%d) failed to allocate UVD bo\n", r);
return r;
}
r = radeon_uvd_resume(rdev);
if (r)
return r;
memset(rdev->uvd.cpu_addr, 0, bo_size);
memcpy(rdev->uvd.cpu_addr, rdev->uvd_fw->data, rdev->uvd_fw->size);
r = radeon_uvd_suspend(rdev);
if (r)
return r;
for (i = 0; i < RADEON_MAX_UVD_HANDLES; ++i) {
atomic_set(&rdev->uvd.handles[i], 0);
rdev->uvd.filp[i] = NULL;
}
return 0;
}
void radeon_uvd_fini(struct radeon_device *rdev)
{
radeon_uvd_suspend(rdev);
radeon_bo_unref(&rdev->uvd.vcpu_bo);
}
int radeon_uvd_suspend(struct radeon_device *rdev)
{
int r;
if (rdev->uvd.vcpu_bo == NULL)
return 0;
r = radeon_bo_reserve(rdev->uvd.vcpu_bo, false);
if (!r) {
radeon_bo_kunmap(rdev->uvd.vcpu_bo);
radeon_bo_unpin(rdev->uvd.vcpu_bo);
radeon_bo_unreserve(rdev->uvd.vcpu_bo);
}
return r;
}
int radeon_uvd_resume(struct radeon_device *rdev)
{
int r;
if (rdev->uvd.vcpu_bo == NULL)
return -EINVAL;
r = radeon_bo_reserve(rdev->uvd.vcpu_bo, false);
if (r) {
radeon_bo_unref(&rdev->uvd.vcpu_bo);
dev_err(rdev->dev, "(%d) failed to reserve UVD bo\n", r);
return r;
}
r = radeon_bo_pin(rdev->uvd.vcpu_bo, RADEON_GEM_DOMAIN_VRAM,
&rdev->uvd.gpu_addr);
if (r) {
radeon_bo_unreserve(rdev->uvd.vcpu_bo);
radeon_bo_unref(&rdev->uvd.vcpu_bo);
dev_err(rdev->dev, "(%d) UVD bo pin failed\n", r);
return r;
}
r = radeon_bo_kmap(rdev->uvd.vcpu_bo, &rdev->uvd.cpu_addr);
if (r) {
dev_err(rdev->dev, "(%d) UVD map failed\n", r);
return r;
}
radeon_bo_unreserve(rdev->uvd.vcpu_bo);
return 0;
}
void radeon_uvd_force_into_uvd_segment(struct radeon_bo *rbo)
{
rbo->placement.fpfn = 0 >> PAGE_SHIFT;
rbo->placement.lpfn = (256 * 1024 * 1024) >> PAGE_SHIFT;
}
void radeon_uvd_free_handles(struct radeon_device *rdev, struct drm_file *filp)
{
int i, r;
for (i = 0; i < RADEON_MAX_UVD_HANDLES; ++i) {
if (rdev->uvd.filp[i] == filp) {
uint32_t handle = atomic_read(&rdev->uvd.handles[i]);
struct radeon_fence *fence;
r = radeon_uvd_get_destroy_msg(rdev,
R600_RING_TYPE_UVD_INDEX, handle, &fence);
if (r) {
DRM_ERROR("Error destroying UVD (%d)!\n", r);
continue;
}
radeon_fence_wait(fence, false);
radeon_fence_unref(&fence);
rdev->uvd.filp[i] = NULL;
atomic_set(&rdev->uvd.handles[i], 0);
}
}
}
static int radeon_uvd_cs_msg_decode(uint32_t *msg, unsigned 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 width_in_mb = width / 16;
unsigned height_in_mb = ALIGN(height / 16, 2);
unsigned image_size, tmp, min_dpb_size;
image_size = width * height;
image_size += image_size / 2;
image_size = ALIGN(image_size, 1024);
switch (stream_type) {
case 0: /* H264 */
/* reference picture buffer */
min_dpb_size = image_size * 17;
/* macroblock context buffer */
min_dpb_size += width_in_mb * height_in_mb * 17 * 192;
/* IT surface buffer */
min_dpb_size += width_in_mb * height_in_mb * 32;
break;
case 1: /* VC1 */
/* reference picture buffer */
min_dpb_size = image_size * 3;
/* CONTEXT_BUFFER */
min_dpb_size += width_in_mb * height_in_mb * 128;
/* IT surface buffer */
min_dpb_size += width_in_mb * 64;
/* DB surface buffer */
min_dpb_size += width_in_mb * 128;
/* BP */
tmp = max(width_in_mb, height_in_mb);
min_dpb_size += ALIGN(tmp * 7 * 16, 64);
break;
case 3: /* MPEG2 */
/* reference picture buffer */
min_dpb_size = image_size * 3;
break;
case 4: /* MPEG4 */
/* reference picture buffer */
min_dpb_size = image_size * 3;
/* CM */
min_dpb_size += width_in_mb * height_in_mb * 64;
/* IT surface buffer */
min_dpb_size += ALIGN(width_in_mb * height_in_mb * 32, 64);
break;
default:
DRM_ERROR("UVD codec not handled %d!\n", stream_type);
return -EINVAL;
}
if (width > pitch) {
DRM_ERROR("Invalid UVD decoding target pitch!\n");
return -EINVAL;
}
if (dpb_size < min_dpb_size) {
DRM_ERROR("Invalid dpb_size in UVD message (%d / %d)!\n",
dpb_size, min_dpb_size);
return -EINVAL;
}
buf_sizes[0x1] = dpb_size;
buf_sizes[0x2] = image_size;
return 0;
}
static int radeon_uvd_cs_msg(struct radeon_cs_parser *p, struct radeon_bo *bo,
unsigned offset, unsigned buf_sizes[])
{
int32_t *msg, msg_type, handle;
void *ptr;
int i, r;
if (offset & 0x3F) {
DRM_ERROR("UVD messages must be 64 byte aligned!\n");
return -EINVAL;
}
r = radeon_bo_kmap(bo, &ptr);
if (r)
return r;
msg = ptr + offset;
msg_type = msg[1];
handle = msg[2];
if (handle == 0) {
DRM_ERROR("Invalid UVD handle!\n");
return -EINVAL;
}
if (msg_type == 1) {
/* it's a decode msg, calc buffer sizes */
r = radeon_uvd_cs_msg_decode(msg, buf_sizes);
radeon_bo_kunmap(bo);
if (r)
return r;
} else if (msg_type == 2) {
/* it's a destroy msg, free the handle */
for (i = 0; i < RADEON_MAX_UVD_HANDLES; ++i)
atomic_cmpxchg(&p->rdev->uvd.handles[i], handle, 0);
radeon_bo_kunmap(bo);
return 0;
} else {
/* it's a create msg, no special handling needed */
radeon_bo_kunmap(bo);
}
/* create or decode, validate the handle */
for (i = 0; i < RADEON_MAX_UVD_HANDLES; ++i) {
if (atomic_read(&p->rdev->uvd.handles[i]) == handle)
return 0;
}
/* handle not found try to alloc a new one */
for (i = 0; i < RADEON_MAX_UVD_HANDLES; ++i) {
if (!atomic_cmpxchg(&p->rdev->uvd.handles[i], 0, handle)) {
p->rdev->uvd.filp[i] = p->filp;
return 0;
}
}
DRM_ERROR("No more free UVD handles!\n");
return -EINVAL;
}
static int radeon_uvd_cs_reloc(struct radeon_cs_parser *p,
int data0, int data1,
unsigned buf_sizes[])
{
struct radeon_cs_chunk *relocs_chunk;
struct radeon_cs_reloc *reloc;
unsigned idx, cmd, offset;
uint64_t start, end;
int r;
relocs_chunk = &p->chunks[p->chunk_relocs_idx];
offset = radeon_get_ib_value(p, data0);
idx = radeon_get_ib_value(p, data1);
if (idx >= relocs_chunk->length_dw) {
DRM_ERROR("Relocs at %d after relocations chunk end %d !\n",
idx, relocs_chunk->length_dw);
return -EINVAL;
}
reloc = p->relocs_ptr[(idx / 4)];
start = reloc->lobj.gpu_offset;
end = start + radeon_bo_size(reloc->robj);
start += offset;
p->ib.ptr[data0] = start & 0xFFFFFFFF;
p->ib.ptr[data1] = start >> 32;
cmd = radeon_get_ib_value(p, p->idx) >> 1;
if (cmd < 0x4) {
if ((end - start) < buf_sizes[cmd]) {
DRM_ERROR("buffer to small (%d / %d)!\n",
(unsigned)(end - start), buf_sizes[cmd]);
return -EINVAL;
}
} else if (cmd != 0x100) {
DRM_ERROR("invalid UVD command %X!\n", cmd);
return -EINVAL;
}
if ((start >> 28) != (end >> 28)) {
DRM_ERROR("reloc %LX-%LX crossing 256MB boundary!\n",
start, end);
return -EINVAL;
}
/* TODO: is this still necessary on NI+ ? */
if ((cmd == 0 || cmd == 0x3) &&
(start >> 28) != (p->rdev->uvd.gpu_addr >> 28)) {
DRM_ERROR("msg/fb buffer %LX-%LX out of 256MB segment!\n",
start, end);
return -EINVAL;
}
if (cmd == 0) {
r = radeon_uvd_cs_msg(p, reloc->robj, offset, buf_sizes);
if (r)
return r;
}
return 0;
}
static int radeon_uvd_cs_reg(struct radeon_cs_parser *p,
struct radeon_cs_packet *pkt,
int *data0, int *data1,
unsigned buf_sizes[])
{
int i, r;
p->idx++;
for (i = 0; i <= pkt->count; ++i) {
switch (pkt->reg + i*4) {
case UVD_GPCOM_VCPU_DATA0:
*data0 = p->idx;
break;
case UVD_GPCOM_VCPU_DATA1:
*data1 = p->idx;
break;
case UVD_GPCOM_VCPU_CMD:
r = radeon_uvd_cs_reloc(p, *data0, *data1, buf_sizes);
if (r)
return r;
break;
case UVD_ENGINE_CNTL:
break;
default:
DRM_ERROR("Invalid reg 0x%X!\n",
pkt->reg + i*4);
return -EINVAL;
}
p->idx++;
}
return 0;
}
int radeon_uvd_cs_parse(struct radeon_cs_parser *p)
{
struct radeon_cs_packet pkt;
int r, data0 = 0, data1 = 0;
/* minimum buffer sizes */
unsigned buf_sizes[] = {
[0x00000000] = 2048,
[0x00000001] = 32 * 1024 * 1024,
[0x00000002] = 2048 * 1152 * 3,
[0x00000003] = 2048,
};
if (p->chunks[p->chunk_ib_idx].length_dw % 16) {
DRM_ERROR("UVD IB length (%d) not 16 dwords aligned!\n",
p->chunks[p->chunk_ib_idx].length_dw);
return -EINVAL;
}
if (p->chunk_relocs_idx == -1) {
DRM_ERROR("No relocation chunk !\n");
return -EINVAL;
}
do {
r = radeon_cs_packet_parse(p, &pkt, p->idx);
if (r)
return r;
switch (pkt.type) {
case RADEON_PACKET_TYPE0:
r = radeon_uvd_cs_reg(p, &pkt, &data0,
&data1, buf_sizes);
if (r)
return r;
break;
case RADEON_PACKET_TYPE2:
p->idx += pkt.count + 2;
break;
default:
DRM_ERROR("Unknown packet type %d !\n", pkt.type);
return -EINVAL;
}
} while (p->idx < p->chunks[p->chunk_ib_idx].length_dw);
return 0;
}
static int radeon_uvd_send_msg(struct radeon_device *rdev,
int ring, struct radeon_bo *bo,
struct radeon_fence **fence)
{
struct ttm_validate_buffer tv;
struct list_head head;
struct radeon_ib ib;
uint64_t addr;
int i, r;
memset(&tv, 0, sizeof(tv));
tv.bo = &bo->tbo;
INIT_LIST_HEAD(&head);
list_add(&tv.head, &head);
r = ttm_eu_reserve_buffers(&head);
if (r)
return r;
radeon_ttm_placement_from_domain(bo, RADEON_GEM_DOMAIN_VRAM);
radeon_uvd_force_into_uvd_segment(bo);
r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false);
if (r) {
ttm_eu_backoff_reservation(&head);
return r;
}
r = radeon_ib_get(rdev, ring, &ib, NULL, 16);
if (r) {
ttm_eu_backoff_reservation(&head);
return r;
}
addr = radeon_bo_gpu_offset(bo);
ib.ptr[0] = PACKET0(UVD_GPCOM_VCPU_DATA0, 0);
ib.ptr[1] = addr;
ib.ptr[2] = PACKET0(UVD_GPCOM_VCPU_DATA1, 0);
ib.ptr[3] = addr >> 32;
ib.ptr[4] = PACKET0(UVD_GPCOM_VCPU_CMD, 0);
ib.ptr[5] = 0;
for (i = 6; i < 16; ++i)
ib.ptr[i] = PACKET2(0);
ib.length_dw = 16;
r = radeon_ib_schedule(rdev, &ib, NULL);
if (r) {
ttm_eu_backoff_reservation(&head);
return r;
}
ttm_eu_fence_buffer_objects(&head, ib.fence);
if (fence)
*fence = radeon_fence_ref(ib.fence);
radeon_ib_free(rdev, &ib);
radeon_bo_unref(&bo);
return 0;
}
/* 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 */
int radeon_uvd_get_create_msg(struct radeon_device *rdev, int ring,
uint32_t handle, struct radeon_fence **fence)
{
struct radeon_bo *bo;
uint32_t *msg;
int r, i;
r = radeon_bo_create(rdev, 1024, PAGE_SIZE, true,
RADEON_GEM_DOMAIN_VRAM, NULL, &bo);
if (r)
return r;
r = radeon_bo_reserve(bo, false);
if (r) {
radeon_bo_unref(&bo);
return r;
}
r = radeon_bo_kmap(bo, (void **)&msg);
if (r) {
radeon_bo_unreserve(bo);
radeon_bo_unref(&bo);
return r;
}
/* stitch together an UVD create msg */
msg[0] = 0x00000de4;
msg[1] = 0x00000000;
msg[2] = handle;
msg[3] = 0x00000000;
msg[4] = 0x00000000;
msg[5] = 0x00000000;
msg[6] = 0x00000000;
msg[7] = 0x00000780;
msg[8] = 0x00000440;
msg[9] = 0x00000000;
msg[10] = 0x01b37000;
for (i = 11; i < 1024; ++i)
msg[i] = 0x0;
radeon_bo_kunmap(bo);
radeon_bo_unreserve(bo);
return radeon_uvd_send_msg(rdev, ring, bo, fence);
}
int radeon_uvd_get_destroy_msg(struct radeon_device *rdev, int ring,
uint32_t handle, struct radeon_fence **fence)
{
struct radeon_bo *bo;
uint32_t *msg;
int r, i;
r = radeon_bo_create(rdev, 1024, PAGE_SIZE, true,
RADEON_GEM_DOMAIN_VRAM, NULL, &bo);
if (r)
return r;
r = radeon_bo_reserve(bo, false);
if (r) {
radeon_bo_unref(&bo);
return r;
}
r = radeon_bo_kmap(bo, (void **)&msg);
if (r) {
radeon_bo_unreserve(bo);
radeon_bo_unref(&bo);
return r;
}
/* stitch together an UVD destroy msg */
msg[0] = 0x00000de4;
msg[1] = 0x00000002;
msg[2] = handle;
msg[3] = 0x00000000;
for (i = 4; i < 1024; ++i)
msg[i] = 0x0;
radeon_bo_kunmap(bo);
radeon_bo_unreserve(bo);
return radeon_uvd_send_msg(rdev, ring, bo, fence);
}
static void radeon_uvd_idle_work_handler(struct work_struct *work)
{
struct radeon_device *rdev =
container_of(work, struct radeon_device, uvd.idle_work.work);
if (radeon_fence_count_emitted(rdev, R600_RING_TYPE_UVD_INDEX) == 0)
radeon_set_uvd_clocks(rdev, 0, 0);
else
schedule_delayed_work(&rdev->uvd.idle_work,
msecs_to_jiffies(UVD_IDLE_TIMEOUT_MS));
}
void radeon_uvd_note_usage(struct radeon_device *rdev)
{
bool set_clocks = !cancel_delayed_work_sync(&rdev->uvd.idle_work);
set_clocks &= schedule_delayed_work(&rdev->uvd.idle_work,
msecs_to_jiffies(UVD_IDLE_TIMEOUT_MS));
if (set_clocks)
radeon_set_uvd_clocks(rdev, 53300, 40000);
}