summaryrefslogtreecommitdiffstats
path: root/drivers/gpu
diff options
context:
space:
mode:
authorAlex Deucher <alexdeucher@gmail.com>2009-09-10 02:53:50 -0400
committerDave Airlie <airlied@redhat.com>2009-09-10 18:44:16 +1000
commit119e20dc149581db3064661b2e659f308f97b663 (patch)
tree44971a17736e127ae503c608dd76874472103b10 /drivers/gpu
parenta513c184d99fe10e7b20771ef86f5f807769318f (diff)
downloadblackbird-op-linux-119e20dc149581db3064661b2e659f308f97b663.tar.gz
blackbird-op-linux-119e20dc149581db3064661b2e659f308f97b663.zip
drm/radeon/kms/r600: fix blit support
select the correct max number of bytes per blit based on whether the size is multiple of 4 bytes. This determines whether we can use 8 or 32 bit pixels for the blit. airlied: also merged the IB padding patch + correcting the VS offset for context Signed-off-by: Alex Deucher <alexdeucher@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
Diffstat (limited to 'drivers/gpu')
-rw-r--r--drivers/gpu/drm/radeon/r600_blit_kms.c39
1 files changed, 31 insertions, 8 deletions
diff --git a/drivers/gpu/drm/radeon/r600_blit_kms.c b/drivers/gpu/drm/radeon/r600_blit_kms.c
index 5755647e688a..bbb0d615ac1c 100644
--- a/drivers/gpu/drm/radeon/r600_blit_kms.c
+++ b/drivers/gpu/drm/radeon/r600_blit_kms.c
@@ -129,6 +129,7 @@ set_shaders(struct radeon_device *rdev)
radeon_ring_write(rdev, (SQ_PGM_CF_OFFSET_PS - PACKET3_SET_CONTEXT_REG_OFFSET) >> 2);
radeon_ring_write(rdev, 0);
+ gpu_addr = rdev->r600_blit.shader_gpu_addr + rdev->r600_blit.vs_offset;
cp_set_surface_sync(rdev, PACKET3_SH_ACTION_ENA, 512, gpu_addr);
}
@@ -248,6 +249,7 @@ set_default_state(struct radeon_device *rdev)
int num_ps_threads, num_vs_threads, num_gs_threads, num_es_threads;
int num_ps_stack_entries, num_vs_stack_entries, num_gs_stack_entries, num_es_stack_entries;
u64 gpu_addr;
+ int dwords;
switch (rdev->family) {
case CHIP_R600:
@@ -394,11 +396,12 @@ set_default_state(struct radeon_device *rdev)
NUM_ES_STACK_ENTRIES(num_es_stack_entries));
/* emit an IB pointing at default state */
+ dwords = (rdev->r600_blit.state_len + 0xf) & ~0xf;
gpu_addr = rdev->r600_blit.shader_gpu_addr + rdev->r600_blit.state_offset;
radeon_ring_write(rdev, PACKET3(PACKET3_INDIRECT_BUFFER, 2));
radeon_ring_write(rdev, gpu_addr & 0xFFFFFFFC);
radeon_ring_write(rdev, upper_32_bits(gpu_addr) & 0xFF);
- radeon_ring_write(rdev, (rdev->r600_blit.state_len / 4));
+ radeon_ring_write(rdev, dwords);
radeon_ring_write(rdev, PACKET3(PACKET3_EVENT_WRITE, 0));
radeon_ring_write(rdev, CACHE_FLUSH_AND_INV_EVENT);
@@ -441,17 +444,25 @@ static inline uint32_t i2f(uint32_t input)
int r600_blit_init(struct radeon_device *rdev)
{
u32 obj_size;
- int r;
+ int r, dwords;
void *ptr;
+ u32 packet2s[16];
+ int num_packet2s = 0;
rdev->r600_blit.state_offset = 0;
if (rdev->family >= CHIP_RV770)
- rdev->r600_blit.state_len = r7xx_default_size * 4;
+ rdev->r600_blit.state_len = r7xx_default_size;
else
- rdev->r600_blit.state_len = r6xx_default_size * 4;
+ rdev->r600_blit.state_len = r6xx_default_size;
+
+ dwords = rdev->r600_blit.state_len;
+ while (dwords & 0xf) {
+ packet2s[num_packet2s++] = PACKET2(0);
+ dwords++;
+ }
- obj_size = rdev->r600_blit.state_len;
+ obj_size = dwords * 4;
obj_size = ALIGN(obj_size, 256);
rdev->r600_blit.vs_offset = obj_size;
@@ -488,9 +499,15 @@ int r600_blit_init(struct radeon_device *rdev)
}
if (rdev->family >= CHIP_RV770)
- memcpy_toio(ptr + rdev->r600_blit.state_offset, r7xx_default_state, rdev->r600_blit.state_len);
+ memcpy_toio(ptr + rdev->r600_blit.state_offset,
+ r7xx_default_state, rdev->r600_blit.state_len * 4);
else
- memcpy_toio(ptr + rdev->r600_blit.state_offset, r6xx_default_state, rdev->r600_blit.state_len);
+ memcpy_toio(ptr + rdev->r600_blit.state_offset,
+ r6xx_default_state, rdev->r600_blit.state_len * 4);
+ if (num_packet2s)
+ memcpy_toio(ptr + rdev->r600_blit.state_offset + (rdev->r600_blit.state_len * 4),
+ packet2s, num_packet2s * 4);
+
memcpy(ptr + rdev->r600_blit.vs_offset, r6xx_vs, r6xx_vs_size * 4);
memcpy(ptr + rdev->r600_blit.ps_offset, r6xx_ps, r6xx_ps_size * 4);
@@ -532,7 +549,13 @@ int r600_blit_prepare_copy(struct radeon_device *rdev, int size_bytes)
{
int r;
int ring_size;
- const int max_size = 8192*8192;
+ int max_size;
+
+ /* 8 bpp vs 32 bpp for xfer unit */
+ if (size_bytes & 3)
+ max_size = 8192*8192;
+ else
+ max_size = 8192*8192*4;
r = r600_vb_ib_get(rdev);
WARN_ON(r);
OpenPOWER on IntegriCloud