Age | Commit message (Collapse) | Author | Files | Lines | |
---|---|---|---|---|---|
2012-01-31 | r600g: remove unused variable num_dest_buffers | Marek Olšák | 3 | -5/+0 | |
2012-01-31 | r600g: use the new code for streamout flush as well | Marek Olšák | 1 | -11/+6 | |
2012-01-31 | r600g: rename r600_reg::flush_flags -> sbu_flags | Marek Olšák | 3 | -4/+4 | |
There is no other use for that. | |||||
2012-01-31 | r600g: fix computation of how many dwords is needed for a flush at the end of CS | Marek Olšák | 2 | -11/+4 | |
2012-01-31 | r600g: remove unused r600_reg::flush_mask | Marek Olšák | 3 | -1179/+1176 | |
2012-01-31 | r600g: remove more dead code | Marek Olšák | 2 | -9/+0 | |
2012-01-31 | r600g: remove dead code for tracking relocations | Marek Olšák | 4 | -32/+1 | |
2012-01-31 | r600g: remove unused flush code | Marek Olšák | 2 | -40/+0 | |
2012-01-31 | r600g: rework cache flushing | Marek Olšák | 9 | -190/+193 | |
This also significantly improves the RV670 flush by using the CB1 flush *always* and also DEST_BASE_0_ENA, which appears to magically fix some tests. I am not entirely sure, but it's possible that RV670 flushing is fixed completely. v2: fix cayman by flushing texture cache instead of vertex cache Thanks to Dave Airlie for testing Cayman. | |||||
2012-01-31 | r600g: add a new simple API for state emission | Marek Olšák | 4 | -0/+54 | |
2012-01-31 | r600g: remove redundant pm4 and pm4_cdwords in r600_context | Marek Olšák | 3 | -215/+234 | |
These only mirrored radeon_winsys_cs. | |||||
2012-01-31 | r600g: consolidate some context_draw code | Marek Olšák | 5 | -56/+23 | |
2012-01-31 | r600g: remove the now-useless internal flush callback | Marek Olšák | 3 | -6/+1 | |
2012-01-31 | r600g: don't initialize the screen and winsys pointer twice | Marek Olšák | 4 | -14/+8 | |
2012-01-31 | r600g: merge r600_context with r600_pipe_context | Marek Olšák | 18 | -296/+291 | |
The split made no sense. | |||||
2012-01-31 | r600g: remove u8,u16,u32,u64 types | Marek Olšák | 10 | -61/+56 | |
2012-01-31 | r600g: get rid of r600_context_reg | Marek Olšák | 2 | -30/+0 | |
Reviewed-by: Dave Airlie <airlied@redhat.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> | |||||
2012-01-31 | r600g: don't use r600_context_reg on r6xx-r7xx | Marek Olšák | 5 | -23/+47 | |
Reviewed-by: Dave Airlie <airlied@redhat.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> | |||||
2012-01-31 | r600g: don't use r600_context_reg on evergreen | Marek Olšák | 6 | -14/+23 | |
Just getting rid of things which use the register mask. Reviewed-by: Dave Airlie <airlied@redhat.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> | |||||
2012-01-31 | r600g: get rid of the mask in r600_pipe_reg | Marek Olšák | 3 | -9/+2 | |
Reviewed-by: Dave Airlie <airlied@redhat.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> | |||||
2012-01-31 | r600g: get rid of the mask parameter in pipe_state_add_reg | Marek Olšák | 4 | -478/+478 | |
Reviewed-by: Dave Airlie <airlied@redhat.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> | |||||
2012-01-31 | r600g: don't use register mask for TA_CNTL_AUX | Marek Olšák | 1 | -12/+8 | |
Reviewed-by: Dave Airlie <airlied@redhat.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> | |||||
2012-01-31 | r600g: don't use register mask for PA_CL_CLIP_CNTL | Marek Olšák | 4 | -22/+19 | |
Reviewed-by: Dave Airlie <airlied@redhat.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> | |||||
2012-01-31 | r600g: don't use register mask for PA_CL_VS_OUT_CNTL | Marek Olšák | 4 | -30/+15 | |
Reviewed-by: Dave Airlie <airlied@redhat.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> | |||||
2012-01-31 | r600g: set full register mask for PA_CL_CLIP_CNTL | Marek Olšák | 2 | -6/+2 | |
We don't set the other bits anywhere else. Reviewed-by: Dave Airlie <airlied@redhat.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> | |||||
2012-01-31 | r600g: don't use register mask for PA_SU_SC_MODE_CNTL | Marek Olšák | 4 | -35/+40 | |
It's always emitted in draw_vbo. Reviewed-by: Dave Airlie <airlied@redhat.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> | |||||
2012-01-31 | r600g: don't add PA_SC_LINE_STIPPLE to rasterizer_state | Marek Olšák | 4 | -19/+11 | |
It's always emitted in draw_vbo. Reviewed-by: Dave Airlie <airlied@redhat.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> | |||||
2012-01-31 | r600g: don't use register mask for CB_COLOR_CONTROL on r6xx-r7xx | Marek Olšák | 3 | -7/+17 | |
Reviewed-by: Dave Airlie <airlied@redhat.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> | |||||
2012-01-31 | r600g: don't set CB_TARGET_MASK in set_framebuffer_state | Marek Olšák | 2 | -13/+2 | |
It's emitted in draw_vbo, always. Reviewed-by: Dave Airlie <airlied@redhat.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> | |||||
2012-01-31 | r600g: set full register mask for CB_COLOR_CONTROL on evergreen | Marek Olšák | 1 | -1/+1 | |
We don't set the other bits anywhere else. Reviewed-by: Dave Airlie <airlied@redhat.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> | |||||
2012-01-31 | r600g: use a more clever way to disable per-vertex point size | Marek Olšák | 4 | -14/+26 | |
This uses point size clamping to force point size to a particular value, making the vertex shader output irrelevant. Reviewed-by: Dave Airlie <airlied@redhat.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> | |||||
2012-01-31 | r600g: don't use register mask for DB_RENDER_CONTROL | Marek Olšák | 2 | -6/+2 | |
We don't set the other bits anywhere else except the other DSA states, which are mutually-exclusive with this one. Reviewed-by: Dave Airlie <airlied@redhat.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> | |||||
2012-01-31 | r600g: cleanup setting DB_SHADER_CONTROL | Marek Olšák | 2 | -32/+7 | |
Reviewed-by: Dave Airlie <airlied@redhat.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> | |||||
2012-01-31 | r600g: rework and consolidate stencilref state setting | Marek Olšák | 4 | -84/+85 | |
Stop using the register mask. Reviewed-by: Dave Airlie <airlied@redhat.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> | |||||
2012-01-31 | r600g: don't use register mask for SQ_GPR_RESOURCE_MGMT_1 | Marek Olšák | 2 | -1/+4 | |
Reviewed-by: Dave Airlie <airlied@redhat.com> Reviewed-by: Alex Deucher <alexander.deucher@amd.com> | |||||
2012-01-31 | r600g: set minimum point size to 1.0 for non-sprite non-aa points | Marek Olšák | 4 | -2/+31 | |
2012-01-30 | r600g: fix udiv/idiv/umod/imod on cayman | Dave Airlie | 1 | -91/+280 | |
Cayman needs some of the MUL instructions spread across a full slot of vectors. It also no longer has RECIP_UINT, the recommendation is to replace it with a U2F + RECIP_IEEE + MUL + F2U. Signed-off-by: Dave Airlie <airlied@redhat.com> | |||||
2012-01-30 | r600g: enable blend_bypass for uint/sint | Dave Airlie | 1 | -1/+12 | |
2012-01-30 | r600g: fix fragcoord.w test on cayman | Dave Airlie | 1 | -11/+28 | |
2012-01-30 | r600g: remove unused streamout code | Marek Olšák | 1 | -29/+0 | |
2012-01-30 | r600g: disable printing debug info for streamout | Marek Olšák | 1 | -1/+2 | |
2012-01-29 | st/mesa: add PIPE_CAP_GLSL_FEATURE_LEVEL, cleanup st_extensions.c | Marek Olšák | 1 | -0/+3 | |
v2: handle the cap in r300 and r600 as well Additional info for r600g: The env var R600_GLSL130=1 enables GLSL 1.3. Along with R600_STREAMOUT=1, it enables full GL 3. | |||||
2012-01-28 | r600g: fix linear and flat interpolation | Vadim Girlin | 4 | -17/+27 | |
Signed-off-by: Vadim Girlin <vadimgirlin@gmail.com> Signed-off-by: Dave Airlie <airlied@redhat.com> | |||||
2012-01-27 | r600g: remove unsupported evergreen CB formats | Alex Deucher | 1 | -7/+0 | |
The evergreen+ CB no longer supports the following formats compared to 6xx/7xx: - COLOR_4_4 - COLOR_3_3_2 - COLOR_6_5_5 - COLOR_8_24_FLOAT - COLOR_24_8_FLOAT - COLOR_11_11_10 - COLOR_11_11_10_FLOAT Signed-off-by: Alex Deucher <alexander.deucher@amd.com> | |||||
2012-01-27 | Revert "r600g: use S_028B20_BUFFER_0_EN macro for streamout buffer enable" | Marek Olšák | 1 | -1/+1 | |
This reverts commit d6cd514edbeca0de38561f66189748078d0dc602. It broke multiple streamout buffers, because it only enabled the first one. | |||||
2012-01-25 | r600g: handle start_component from stream_output_info when possible | Marek Olšák | 1 | -4/+4 | |
2012-01-25 | r600g: don't expose transform_feedback2 without kernel support | Marek Olšák | 1 | -1/+2 | |
2012-01-25 | r600g: nuke the fallback for vertex and fragment color clamping | Marek Olšák | 6 | -53/+1 | |
2012-01-25 | st/mesa: do vertex and fragment color clamping in shaders | Marek Olšák | 1 | -1/+3 | |
For ARB_color_buffer_float. Most hardware can't do it and st/mesa is the perfect place for a fallback. The exceptions are: - r500 (vertex clamp only) - nv50 (both) - nvc0 (both) - softpipe (both) We also have to take into account that r300 can do CLAMPED vertex colors only, while r600 can do UNCLAMPED vertex colors only. The difference can be expressed with the two new CAPs. | |||||
2012-01-24 | r600g: remove unused variable | Brian Paul | 1 | -1/+0 | |