summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* radeonsi: use the correct LLVMTargetMachineRef in si_build_shader_variantNicolai Hähnle2017-06-221-6/+22
| | | | | | | | | | | | si_build_shader_variant can actually be called directly from one of normal-priority compiler threads. In that case, the thread_index is only valid for the normal tm array. v2: - use the correct sel/shader->compiler_ctx_state Fixes: 86cc8097266c ("radeonsi: use a compiler queue with a low priority for optimized shaders") Reviewed-by: Marek Olšák <[email protected]>
* radeonsi/gfx9: keep reusing the same buffer/address for the gfx9 flush fenceMarek Olšák2017-06-223-8/+28
| | | | | | | | instead of using a monotonic suballocator v2: initialize the memory at context creation Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi/gfx9: enable the constant engineMarek Olšák2017-06-221-4/+1
| | | | | | | I think this kernel commit fixes it: drm/amdgpu:use FRAME_CNTL for new GFX ucode Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi/gfx9: indirect buffers and all CP packets use TC L2Marek Olšák2017-06-224-13/+21
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: flush CB after MSAA only when transitioning from CB to texturesMarek Olšák2017-06-222-14/+60
| | | | | | | | | | | | | The main flush before texturing is done after the FMASK decompress pass. CB after MSAA rendering is not flushed in set_framebuffer_state and also not in memory_barrier if the current color buffer is MSAA. We fully rely on the FMASK decompress pass for the flushing. Some CB decompress and resolve passes need an explicit flush before and after. Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: unify CB_RESOLVE blitter invocation codeMarek Olšák2017-06-221-17/+18
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: flush DB caches only when transitioning from DB to texturingMarek Olšák2017-06-225-25/+56
| | | | | | | | | Use the mechanism of si_decompress_textures, but instead of doing the actual decompression, just flag the DB cache flush there. This removes a lot of unnecessary DB cache flushes. Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: add separate HUD counters for CB and DB cache flushesMarek Olšák2017-06-224-10/+20
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: don't set the border color if it's unusedMarek Olšák2017-06-221-4/+18
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: don't set 16 scissors and 16 viewports if they're unusedMarek Olšák2017-06-224-6/+34
| | | | | | | Only do so if there is a shader writing gl_ViewportIndex. This removes a lot of CPU overhead for the most common case. Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: fix pipe_rasterizer_state::scissor with multiple viewportsMarek Olšák2017-06-221-1/+1
| | | | | Cc: 17.1 <[email protected]> Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: simplify st_update_viewportMarek Olšák2017-06-221-25/+10
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: remove redundant sample_mask checkingMarek Olšák2017-06-222-7/+1
| | | | | | cso does that too Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: use precomputed st_fb_orientationMarek Olšák2017-06-224-8/+8
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* mesa: don't call _mesa_update_clip_plane in the GL core profileMarek Olšák2017-06-221-1/+2
| | | | | | It uses the projection matrix to transform the clip plane. Reviewed-by: Ilia Mirkin <[email protected]>
* st/mesa: set st_context::...num_samplers to 0 when there are no samplersMarek Olšák2017-06-221-1/+3
| | | | | | This was missed during my st/mesa series. Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: unify fail paths for update_single_textureMarek Olšák2017-06-222-7/+2
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: don't call u_sampler_view_default_template for sampler viewsMarek Olšák2017-06-221-1/+2
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: always set sampler swizzle according to the texture base formatMarek Olšák2017-06-221-32/+9
| | | | | | | | | Mainly don't (indirectly) call util_format_description here. If the driver supports texture swizzling, this will always do the right thing. If the driver doesn't support it, it doesn't matter. Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: samplers only need to track whether GLSL >= 130Marek Olšák2017-06-225-22/+25
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: simplify get_texture_format_swizzleMarek Olšák2017-06-221-23/+20
| | | | | | | - Don't check GL_NONE (that was only for buffers). - Don't use util_format_is_depth_or_stencil. Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: add an entirely separate codepath for setting up buffer viewsMarek Olšák2017-06-226-100/+134
| | | | | | | | | | | | Remove handling of buffers from all texture paths. This simplifies things for both buffers and textures. get_sampler_view_format is also cleaned up not to call util_format_is_depth_and_stencil. v2: also update st_NewTextureHandle Reviewed-by: Nicolai Hähnle <[email protected]> (v1)
* st/mesa: don't return an error from update_single_textureMarek Olšák2017-06-223-11/+7
| | | | | | | It can just return a NULL sampler view, which is better than not doing anything at all. Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: clean up trivial dereferences in update_texturesMarek Olšák2017-06-221-6/+7
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: don't check MaxTextureImageUnits in update_texturesMarek Olšák2017-06-221-12/+2
| | | | | | The linker takes care of it. Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: don't call st_shader_stage_to_ptarget in update_texturesMarek Olšák2017-06-221-8/+7
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* cso: inline a few frequently-used functionsMarek Olšák2017-06-222-31/+26
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* cso: don't return errors from sampler functionsMarek Olšák2017-06-222-18/+8
| | | | | | No code checks the errors. Reviewed-by: Nicolai Hähnle <[email protected]>
* cso: don't track the number of sampler states boundMarek Olšák2017-06-221-36/+23
| | | | | | | | | | | | This removes 2 loops from hot codepaths and adds 1 loop to a rare codepath (restore_sampler_states), and makes sanitize_hash() slightly worse. Sampler states, when bound, are not unbound for draw calls that don't need them. That's OK, because bound sampler states don't add any overhead. This results in lower CPU overhead in most cases. Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: sink and simplify texBaseFormat getting for sampler statesMarek Olšák2017-06-221-9/+9
| | | | | Reviewed-by: Timothy Arceri <[email protected]> Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: don't set sampler states for TBOsMarek Olšák2017-06-222-4/+12
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: optimize sampler state translation codeMarek Olšák2017-06-221-48/+20
| | | | | Reviewed-by: Timothy Arceri <[email protected]> Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: sink code needed for apply_texture_swizzle_to_border_colorMarek Olšák2017-06-221-28/+33
| | | | | Reviewed-by: Timothy Arceri <[email protected]> Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: simplify update_shader_samplersMarek Olšák2017-06-221-16/+8
| | | | | Reviewed-by: Timothy Arceri <[email protected]> Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: when binding sampler states, don't check the max sampler limitMarek Olšák2017-06-221-8/+1
| | | | | | | The GLSL linker takes care of it. Reviewed-by: Timothy Arceri <[email protected]> Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: don't unbind sampler states if none are usedMarek Olšák2017-06-221-1/+1
| | | | | Reviewed-by: Nicolai Hähnle <[email protected]> Reviewed-by: Timothy Arceri <[email protected]>
* st/mesa: unify update_gp/tcp/tep codeMarek Olšák2017-06-221-63/+28
| | | | | Reviewed-by: Timothy Arceri <[email protected]> Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: don't search through shader variants if there is only oneMarek Olšák2017-06-221-51/+90
| | | | | Reviewed-by: Timothy Arceri <[email protected]> Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: don't track shader variants in st_contextMarek Olšák2017-06-222-23/+14
| | | | | Reviewed-by: Timothy Arceri <[email protected]> Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: move blend color into its own state atomMarek Olšák2017-06-223-6/+10
| | | | | | | This is now sensible thanks to the NewBlendColor flag. Reviewed-by: Timothy Arceri <[email protected]> Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: check correctly if multisampling is enabledMarek Olšák2017-06-222-3/+4
| | | | | Reviewed-by: Timothy Arceri <[email protected]> Reviewed-by: Nicolai Hähnle <[email protected]>
* st/mesa: don't invoke st_finalize_texture & st_convert_sampler for TBOsMarek Olšák2017-06-221-4/+6
| | | | | | This is a v2 of the previous patch (v1 didn't skip st_finalize_texture). Reviewed-by: Samuel Pitoiset <[email protected]>
* mesa: simplify _mesa_is_image_unit_valid for buffersMarek Olšák2017-06-221-12/+13
| | | | | | Reviewed-by: Nicolai Hähnle <[email protected]> Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Timothy Arceri <[email protected]>
* mesa: don't flag _NEW_PROGRAM_CONSTANTS for GLSL programs for st/mesaMarek Olšák2017-06-224-8/+30
| | | | | | | | | v2: also update _mesa_uniform_handle for bindless textures Reviewed-by: Nicolai Hähnle <[email protected]> (v1) Reviewed-by: Brian Paul <[email protected]> (v1) Reviewed-by: Timothy Arceri <[email protected]> (v1) Reviewed-by: Samuel Pitoiset <[email protected]>
* glsl: Track whether uniforms are active per stageKenneth Graunke2017-06-223-0/+9
| | | | | | | | | | for finer granularity state flagging v2: Marek - use a bitmask, add shader cache support Reviewed-by: Nicolai Hähnle <[email protected]> Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Timothy Arceri <[email protected]>
* mesa: don't flag _NEW_PROGRAM_CONSTANTS for non-GLSL programs for st/mesaMarek Olšák2017-06-224-8/+49
| | | | | | | | | This has the benefit that we get to set up constants for exactly the shader stage that needs it. Reviewed-by: Nicolai Hähnle <[email protected]> Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Timothy Arceri <[email protected]>
* mesa: flush vertices before updating ctx->_ShaderMarek Olšák2017-06-221-2/+2
| | | | | | | Cc: 17.1 <[email protected]> Reviewed-by: Nicolai Hähnle <[email protected]> Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Timothy Arceri <[email protected]>
* mesa: set driver flags for glPopAttrib(GL_ENABLE_BIT) properlyMarek Olšák2017-06-221-0/+14
| | | | | | Reviewed-by: Nicolai Hähnle <[email protected]> Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Timothy Arceri <[email protected]>
* mesa: don't flag _NEW_POLYGON_STIPPLE for st/mesaMarek Olšák2017-06-224-5/+13
| | | | | | Reviewed-by: Nicolai Hähnle <[email protected]> Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Timothy Arceri <[email protected]>
* mesa: don't flag _NEW_LINE for st/mesaMarek Olšák2017-06-224-5/+12
| | | | | | Reviewed-by: Nicolai Hähnle <[email protected]> Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Timothy Arceri <[email protected]>