int vce_v1_0_init(struct radeon_device *rdev)
{
struct radeon_ring *ring;
int r;
r = vce_v1_0_start(rdev);
if (r)
return r;
ring = &rdev->ring[TN_RING_TYPE_VCE1_INDEX];
ring->ready = true;
r = radeon_ring_test(rdev, TN_RING_TYPE_VCE1_INDEX, ring);
if (r) {
ring->ready = false;
return r;
}
ring = &rdev->ring[TN_RING_TYPE_VCE2_INDEX];
ring->ready = true;
r = radeon_ring_test(rdev, TN_RING_TYPE_VCE2_INDEX, ring);
if (r) {
ring->ready = false;
return r;
}
DRM_INFO("VCE initialized successfully.\n");
return 0;
}
/**
* uvd_v1_0_init - start and test UVD block
*
* @rdev: radeon_device pointer
*
* Initialize the hardware, boot up the VCPU and do some testing
*/
int uvd_v1_0_init(struct radeon_device *rdev)
{
struct radeon_ring *ring = &rdev->ring[R600_RING_TYPE_UVD_INDEX];
uint32_t tmp;
int r;
/* raise clocks while booting up the VCPU */
if (rdev->family < CHIP_RV740)
radeon_set_uvd_clocks(rdev, 10000, 10000);
else
radeon_set_uvd_clocks(rdev, 53300, 40000);
r = uvd_v1_0_start(rdev);
if (r)
goto done;
ring->ready = true;
r = radeon_ring_test(rdev, R600_RING_TYPE_UVD_INDEX, ring);
if (r) {
ring->ready = false;
goto done;
}
r = radeon_ring_lock(rdev, ring, 10);
if (r) {
DRM_ERROR("radeon: ring failed to lock UVD ring (%d).\n", r);
goto done;
}
tmp = PACKET0(UVD_SEMA_WAIT_FAULT_TIMEOUT_CNTL, 0);
radeon_ring_write(ring, tmp);
radeon_ring_write(ring, 0xFFFFF);
tmp = PACKET0(UVD_SEMA_WAIT_INCOMPLETE_TIMEOUT_CNTL, 0);
radeon_ring_write(ring, tmp);
radeon_ring_write(ring, 0xFFFFF);
tmp = PACKET0(UVD_SEMA_SIGNAL_INCOMPLETE_TIMEOUT_CNTL, 0);
radeon_ring_write(ring, tmp);
radeon_ring_write(ring, 0xFFFFF);
/* Clear timeout status bits */
radeon_ring_write(ring, PACKET0(UVD_SEMA_TIMEOUT_STATUS, 0));
radeon_ring_write(ring, 0x8);
radeon_ring_write(ring, PACKET0(UVD_SEMA_CNTL, 0));
radeon_ring_write(ring, 3);
radeon_ring_unlock_commit(rdev, ring);
done:
/* lower clocks again */
radeon_set_uvd_clocks(rdev, 0, 0);
if (!r)
DRM_INFO("UVD initialized successfully.\n");
return r;
}
/**
* cik_sdma_gfx_resume - setup and start the async dma engines
*
* @rdev: radeon_device pointer
*
* Set up the gfx DMA ring buffers and enable them (CIK).
* Returns 0 for success, error for failure.
*/
static int cik_sdma_gfx_resume(struct radeon_device *rdev)
{
struct radeon_ring *ring;
u32 rb_cntl, ib_cntl;
u32 rb_bufsz;
u32 reg_offset, wb_offset;
int i, r;
for (i = 0; i < 2; i++) {
if (i == 0) {
ring = &rdev->ring[R600_RING_TYPE_DMA_INDEX];
reg_offset = SDMA0_REGISTER_OFFSET;
wb_offset = R600_WB_DMA_RPTR_OFFSET;
} else {
ring = &rdev->ring[CAYMAN_RING_TYPE_DMA1_INDEX];
reg_offset = SDMA1_REGISTER_OFFSET;
wb_offset = CAYMAN_WB_DMA1_RPTR_OFFSET;
}
WREG32(SDMA0_SEM_INCOMPLETE_TIMER_CNTL + reg_offset, 0);
WREG32(SDMA0_SEM_WAIT_FAIL_TIMER_CNTL + reg_offset, 0);
/* Set ring buffer size in dwords */
rb_bufsz = order_base_2(ring->ring_size / 4);
rb_cntl = rb_bufsz << 1;
#ifdef __BIG_ENDIAN
rb_cntl |= SDMA_RB_SWAP_ENABLE | SDMA_RPTR_WRITEBACK_SWAP_ENABLE;
#endif
WREG32(SDMA0_GFX_RB_CNTL + reg_offset, rb_cntl);
/* Initialize the ring buffer's read and write pointers */
WREG32(SDMA0_GFX_RB_RPTR + reg_offset, 0);
WREG32(SDMA0_GFX_RB_WPTR + reg_offset, 0);
/* set the wb address whether it's enabled or not */
WREG32(SDMA0_GFX_RB_RPTR_ADDR_HI + reg_offset,
upper_32_bits(rdev->wb.gpu_addr + wb_offset) & 0xFFFFFFFF);
WREG32(SDMA0_GFX_RB_RPTR_ADDR_LO + reg_offset,
((rdev->wb.gpu_addr + wb_offset) & 0xFFFFFFFC));
if (rdev->wb.enabled)
rb_cntl |= SDMA_RPTR_WRITEBACK_ENABLE;
WREG32(SDMA0_GFX_RB_BASE + reg_offset, ring->gpu_addr >> 8);
WREG32(SDMA0_GFX_RB_BASE_HI + reg_offset, ring->gpu_addr >> 40);
ring->wptr = 0;
WREG32(SDMA0_GFX_RB_WPTR + reg_offset, ring->wptr << 2);
/* enable DMA RB */
WREG32(SDMA0_GFX_RB_CNTL + reg_offset, rb_cntl | SDMA_RB_ENABLE);
ib_cntl = SDMA_IB_ENABLE;
#ifdef __BIG_ENDIAN
ib_cntl |= SDMA_IB_SWAP_ENABLE;
#endif
/* enable DMA IBs */
WREG32(SDMA0_GFX_IB_CNTL + reg_offset, ib_cntl);
ring->ready = true;
r = radeon_ring_test(rdev, ring->idx, ring);
if (r) {
ring->ready = false;
return r;
}
}
if ((rdev->asic->copy.copy_ring_index == R600_RING_TYPE_DMA_INDEX) ||
(rdev->asic->copy.copy_ring_index == CAYMAN_RING_TYPE_DMA1_INDEX))
radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size);
return 0;
}
/**
* r600_dma_resume - setup and start the async dma engine
*
* @rdev: radeon_device pointer
*
* Set up the DMA ring buffer and enable it. (r6xx-evergreen).
* Returns 0 for success, error for failure.
*/
int r600_dma_resume(struct radeon_device *rdev)
{
struct radeon_ring *ring = &rdev->ring[R600_RING_TYPE_DMA_INDEX];
u32 rb_cntl, dma_cntl, ib_cntl;
u32 rb_bufsz;
int r;
/* Reset dma */
if (rdev->family >= CHIP_RV770)
WREG32(SRBM_SOFT_RESET, RV770_SOFT_RESET_DMA);
else
WREG32(SRBM_SOFT_RESET, SOFT_RESET_DMA);
RREG32(SRBM_SOFT_RESET);
udelay(50);
WREG32(SRBM_SOFT_RESET, 0);
WREG32(DMA_SEM_INCOMPLETE_TIMER_CNTL, 0);
WREG32(DMA_SEM_WAIT_FAIL_TIMER_CNTL, 0);
/* Set ring buffer size in dwords */
rb_bufsz = order_base_2(ring->ring_size / 4);
rb_cntl = rb_bufsz << 1;
#ifdef __BIG_ENDIAN
rb_cntl |= DMA_RB_SWAP_ENABLE | DMA_RPTR_WRITEBACK_SWAP_ENABLE;
#endif
WREG32(DMA_RB_CNTL, rb_cntl);
/* Initialize the ring buffer's read and write pointers */
WREG32(DMA_RB_RPTR, 0);
WREG32(DMA_RB_WPTR, 0);
/* set the wb address whether it's enabled or not */
WREG32(DMA_RB_RPTR_ADDR_HI,
upper_32_bits(rdev->wb.gpu_addr + R600_WB_DMA_RPTR_OFFSET) & 0xFF);
WREG32(DMA_RB_RPTR_ADDR_LO,
((rdev->wb.gpu_addr + R600_WB_DMA_RPTR_OFFSET) & 0xFFFFFFFC));
if (rdev->wb.enabled)
rb_cntl |= DMA_RPTR_WRITEBACK_ENABLE;
WREG32(DMA_RB_BASE, ring->gpu_addr >> 8);
/* enable DMA IBs */
ib_cntl = DMA_IB_ENABLE;
#ifdef __BIG_ENDIAN
ib_cntl |= DMA_IB_SWAP_ENABLE;
#endif
WREG32(DMA_IB_CNTL, ib_cntl);
dma_cntl = RREG32(DMA_CNTL);
dma_cntl &= ~CTXEMPTY_INT_ENABLE;
WREG32(DMA_CNTL, dma_cntl);
if (rdev->family >= CHIP_RV770)
WREG32(DMA_MODE, 1);
ring->wptr = 0;
WREG32(DMA_RB_WPTR, ring->wptr << 2);
ring->rptr = RREG32(DMA_RB_RPTR) >> 2;
WREG32(DMA_RB_CNTL, rb_cntl | DMA_RB_ENABLE);
ring->ready = true;
r = radeon_ring_test(rdev, R600_RING_TYPE_DMA_INDEX, ring);
if (r) {
ring->ready = false;
return r;
}
if (rdev->asic->copy.copy_ring_index == R600_RING_TYPE_DMA_INDEX)
radeon_ttm_set_active_vram_size(rdev, rdev->mc.real_vram_size);
return 0;
}
int cayman_cp_resume(struct radeon_device *rdev)
{
struct radeon_ring *ring;
u32 tmp;
u32 rb_bufsz;
int r;
/* Reset cp; if cp is reset, then PA, SH, VGT also need to be reset */
WREG32(GRBM_SOFT_RESET, (SOFT_RESET_CP |
SOFT_RESET_PA |
SOFT_RESET_SH |
SOFT_RESET_VGT |
SOFT_RESET_SPI |
SOFT_RESET_SX));
RREG32(GRBM_SOFT_RESET);
mdelay(15);
WREG32(GRBM_SOFT_RESET, 0);
RREG32(GRBM_SOFT_RESET);
WREG32(CP_SEM_WAIT_TIMER, 0x0);
WREG32(CP_SEM_INCOMPLETE_TIMER_CNTL, 0x0);
/* Set the write pointer delay */
WREG32(CP_RB_WPTR_DELAY, 0);
WREG32(CP_DEBUG, (1 << 27));
/* ring 0 - compute and gfx */
/* Set ring buffer size */
ring = &rdev->ring[RADEON_RING_TYPE_GFX_INDEX];
rb_bufsz = drm_order(ring->ring_size / 8);
tmp = (drm_order(RADEON_GPU_PAGE_SIZE/8) << 8) | rb_bufsz;
#ifdef __BIG_ENDIAN
tmp |= BUF_SWAP_32BIT;
#endif
WREG32(CP_RB0_CNTL, tmp);
/* Initialize the ring buffer's read and write pointers */
WREG32(CP_RB0_CNTL, tmp | RB_RPTR_WR_ENA);
ring->wptr = 0;
WREG32(CP_RB0_WPTR, ring->wptr);
/* set the wb address wether it's enabled or not */
WREG32(CP_RB0_RPTR_ADDR, (rdev->wb.gpu_addr + RADEON_WB_CP_RPTR_OFFSET) & 0xFFFFFFFC);
WREG32(CP_RB0_RPTR_ADDR_HI, upper_32_bits(rdev->wb.gpu_addr + RADEON_WB_CP_RPTR_OFFSET) & 0xFF);
WREG32(SCRATCH_ADDR, ((rdev->wb.gpu_addr + RADEON_WB_SCRATCH_OFFSET) >> 8) & 0xFFFFFFFF);
if (rdev->wb.enabled)
WREG32(SCRATCH_UMSK, 0xff);
else {
tmp |= RB_NO_UPDATE;
WREG32(SCRATCH_UMSK, 0);
}
mdelay(1);
WREG32(CP_RB0_CNTL, tmp);
WREG32(CP_RB0_BASE, ring->gpu_addr >> 8);
ring->rptr = RREG32(CP_RB0_RPTR);
/* ring1 - compute only */
/* Set ring buffer size */
ring = &rdev->ring[CAYMAN_RING_TYPE_CP1_INDEX];
rb_bufsz = drm_order(ring->ring_size / 8);
tmp = (drm_order(RADEON_GPU_PAGE_SIZE/8) << 8) | rb_bufsz;
#ifdef __BIG_ENDIAN
tmp |= BUF_SWAP_32BIT;
#endif
WREG32(CP_RB1_CNTL, tmp);
/* Initialize the ring buffer's read and write pointers */
WREG32(CP_RB1_CNTL, tmp | RB_RPTR_WR_ENA);
ring->wptr = 0;
WREG32(CP_RB1_WPTR, ring->wptr);
/* set the wb address wether it's enabled or not */
WREG32(CP_RB1_RPTR_ADDR, (rdev->wb.gpu_addr + RADEON_WB_CP1_RPTR_OFFSET) & 0xFFFFFFFC);
WREG32(CP_RB1_RPTR_ADDR_HI, upper_32_bits(rdev->wb.gpu_addr + RADEON_WB_CP1_RPTR_OFFSET) & 0xFF);
mdelay(1);
WREG32(CP_RB1_CNTL, tmp);
WREG32(CP_RB1_BASE, ring->gpu_addr >> 8);
ring->rptr = RREG32(CP_RB1_RPTR);
/* ring2 - compute only */
/* Set ring buffer size */
ring = &rdev->ring[CAYMAN_RING_TYPE_CP2_INDEX];
rb_bufsz = drm_order(ring->ring_size / 8);
tmp = (drm_order(RADEON_GPU_PAGE_SIZE/8) << 8) | rb_bufsz;
#ifdef __BIG_ENDIAN
tmp |= BUF_SWAP_32BIT;
#endif
WREG32(CP_RB2_CNTL, tmp);
/* Initialize the ring buffer's read and write pointers */
WREG32(CP_RB2_CNTL, tmp | RB_RPTR_WR_ENA);
ring->wptr = 0;
WREG32(CP_RB2_WPTR, ring->wptr);
/* set the wb address wether it's enabled or not */
WREG32(CP_RB2_RPTR_ADDR, (rdev->wb.gpu_addr + RADEON_WB_CP2_RPTR_OFFSET) & 0xFFFFFFFC);
WREG32(CP_RB2_RPTR_ADDR_HI, upper_32_bits(rdev->wb.gpu_addr + RADEON_WB_CP2_RPTR_OFFSET) & 0xFF);
mdelay(1);
WREG32(CP_RB2_CNTL, tmp);
WREG32(CP_RB2_BASE, ring->gpu_addr >> 8);
ring->rptr = RREG32(CP_RB2_RPTR);
/* start the rings */
cayman_cp_start(rdev);
rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ready = true;
rdev->ring[CAYMAN_RING_TYPE_CP1_INDEX].ready = false;
rdev->ring[CAYMAN_RING_TYPE_CP2_INDEX].ready = false;
/* this only test cp0 */
r = radeon_ring_test(rdev, RADEON_RING_TYPE_GFX_INDEX, &rdev->ring[RADEON_RING_TYPE_GFX_INDEX]);
if (r) {
rdev->ring[RADEON_RING_TYPE_GFX_INDEX].ready = false;
rdev->ring[CAYMAN_RING_TYPE_CP1_INDEX].ready = false;
rdev->ring[CAYMAN_RING_TYPE_CP2_INDEX].ready = false;
return r;
}
return 0;
}
