summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* radv: handle layer export from vs->fs properlyDave Airlie2017-01-313-2/+23
| | | | | | | | Fixes: dEQP-VK.geometry.layered.1d_array.fragment_layer Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv: emit esgs itemsize register.Dave Airlie2017-01-311-0/+2
| | | | | Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv: handle prim id inputs to fragment shader.Dave Airlie2017-01-311-1/+15
| | | | | Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv: emit geometry shaders to hardwareDave Airlie2017-01-311-2/+96
| | | | | | | This emits the compiled geometry shader and other state registers. Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv: emit geometry ring size and pointers via preamble (v2)Dave Airlie2017-01-313-11/+230
| | | | | | | | | | | | This uses the scratch infrastructure to handle the esgs and gsvs rings. (this replaces the old code that did this with patching). v2: fix correct ring sizes, reset sizes (Bas) Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv: add gs ring size calculations to pipeline.Dave Airlie2017-01-312-0/+34
| | | | | Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv: add pipeline creation support for geometry shaders (v2.1)Dave Airlie2017-01-313-8/+124
| | | | | | | | | | | This adds gs copy shader support to the pipeline cache, and few geometry related changes. v2: rebase for spill changes. v2.1: fix incorrect pipeline destruction. Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv/ac: handle primitive idDave Airlie2017-01-312-1/+11
| | | | | Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv/ac: handle emitting vertex outputs to esgs ring.Dave Airlie2017-01-312-1/+38
| | | | | Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv/ac: handle gs inputsDave Airlie2017-01-311-0/+56
| | | | | | | | This handles geometry shader inputs written by the vertex (es) shader to the esgs ring. Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv/ac: add geom input support to get deref offset.Dave Airlie2017-01-311-8/+14
| | | | | | | This just adds the API and fixes up the callers. Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv/ac: handle invocation and primitive id intrinsicsDave Airlie2017-01-311-0/+9
| | | | | Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv/ac: handle geometry emit vertex and end prim intrinsics.Dave Airlie2017-01-311-0/+126
| | | | | | | | This handles emitting things to the gsvs ring, and sending the correct GS msgs. Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv/ac: handle emitting gs epilogueDave Airlie2017-01-311-0/+14
| | | | | Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv/ac: add copy shader creationDave Airlie2017-01-312-0/+88
| | | | | | | This create the gs copy shader and compiles it. Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv/ac: setup function parameters for vs as es and copy shader.Dave Airlie2017-01-311-17/+32
| | | | | Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv: pass some necessary gs info back to state handling.Dave Airlie2017-01-312-0/+7
| | | | | | | We need this info to program some registers. Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv: emit vertex shader to correct hw block.Dave Airlie2017-01-311-1/+20
| | | | | | | This emits the shader to the ES block in the correct case. Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv/ac: propogate as_es flag into shader info from key.Dave Airlie2017-01-312-0/+4
| | | | | | | | This just places the flag into the shader info so we can use it from the driver after we create the shader. Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv: extend shader stage code to cover geometry shaders.Dave Airlie2017-01-311-8/+22
| | | | | | | This enables the paths for setting up user ptrs to vs/es and gs. Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv/ac: start setting up the geom shader rings (v2)Dave Airlie2017-01-312-3/+63
| | | | | | | | | This sets up the rings and adds the variables needed to make them work. v2: rework for sharing ring and scratch Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv/ac: handle geom shader sgpr/vgpr inputsDave Airlie2017-01-312-0/+41
| | | | | | | This just sets up the gpr inputs. Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv/ac: add geom shader sendmsg defines.Dave Airlie2017-01-311-0/+8
| | | | | | | This just adds some defines needed for geom shaders. Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv/ac: add some geom shader info from nir->ac shader.Dave Airlie2017-01-312-0/+12
| | | | | Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv: move hw vertex shader emit to separate functionDave Airlie2017-01-311-29/+40
| | | | | | | This is to later allow ES shaders to be emitted. Review-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv: fixup ia multi vgt param code to handle geom shaders.Dave Airlie2017-01-311-11/+9
| | | | | | | This fixes up a few of the commented out blocks. Reviewed-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv: add code to set gs_table_depth.Dave Airlie2017-01-312-0/+35
| | | | | Review-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv: add small helper to denote when a geom shader is in the pipeline.Dave Airlie2017-01-311-0/+5
| | | | | Review-by: Bas Nieuwenhuizen <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv: Prevent Coverity warningRobert Foss2017-01-301-0/+2
| | | | | | | | | Prevent Coverity seeing potential errors when src is no initialized in the switch case. Coverity-Id: 1396397 Signed-off-by: Robert Foss <[email protected]> Reviewed-by: Bas Nieuwenhuizen <[email protected]>
* mesa: add new MESA_GLSL flag for printing shader cache debug infoTimothy Arceri2017-01-313-0/+4
| | | | Reviewed-by: Eric Anholt <[email protected]>
* glsl: add cache to ctx and add sha1 string fieldsCarl Worth2017-01-312-0/+15
| | | | | | | | | We also add a flag for detecting shaders written to shader cache. V2: dont leak cache Signed-off-by: Timothy Arceri <[email protected]> Reviewed-by: Eric Anholt <[email protected]>
* glsl: add new uniform fields to be used to restore state from cacheCarl Worth2017-01-312-0/+8
| | | | | Signed-off-by: Timothy Arceri <[email protected]> Reviewed-by: Eric Anholt <[email protected]>
* glsl: Switch to disable-by-default for the GLSL shader cacheCarl Worth2017-01-312-0/+12
| | | | | | | | | | | | | | | | | The shader cache is expected to be developed incrementally over a fairly long series of commits. For that period of instability, we require users to opt into the shader cache by setting: MESA_GLSL_CACHE_ENABLE=1 In the future, when the shader cache is complete, we can revert this commit so that the cache will be on by default. The user can always disable the cache with MESA_GLSL_CACHE_DISABLE=1. That functionality is not affected by this commit, (nor will it be affected by the future revert). Reviewed-by: Eric Anholt <[email protected]>
* radv/ac: implement txs for buffer textures.Dave Airlie2017-01-311-0/+5
| | | | | | | | | | This fixes a bunch of buffer related: dEQP-VK.memory.pipeline_barrier.* tests, that were crashing in LLVM due to this being missing. Reviewed-by: Andres Rodriguez<[email protected]> Cc: "17.0" <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv/ac: handle nir irem opcode.Dave Airlie2017-01-311-0/+3
| | | | | | | | | This fixes: dEQP-VK.spirv_assembly.instruction.compute.opsrem.* Reviewed-by: Andres Rodriguez <[email protected]> Cc: "17.0" <[email protected]" Signed-off-by: Dave Airlie <[email protected]>
* radv/ac: fix multisample subpass image.Dave Airlie2017-01-311-5/+6
| | | | | | | | We weren't adding the fragment position properly. Reviewed-by: Bas Nieuwenhuizen <[email protected]> Cc: "17.0" <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* radv: handle transfer_write as a dst flag.Dave Airlie2017-01-311-0/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | It appears we can get image barriers like: srcStageMask: VkPipelineStageFlags = 4096 (VK_PIPELINE_STAGE_TRANSFER_BIT) dstStageMask: VkPipelineStageFlags = 4096 (VK_PIPELINE_STAGE_TRANSFER_BIT) dependencyFlags: VkDependencyFlags = 0 memoryBarrierCount: uint32_t = 0 pMemoryBarriers: const VkMemoryBarrier* = NULL bufferMemoryBarrierCount: uint32_t = 0 pBufferMemoryBarriers: const VkBufferMemoryBarrier* = NULL imageMemoryBarrierCount: uint32_t = 1 pImageMemoryBarriers: const VkImageMemoryBarrier* = 0x7ffc882367b0 pImageMemoryBarriers[0]: const VkImageMemoryBarrier = 0x7ffc882367b0: sType: VkStructureType = VK_STRUCTURE_TYPE_IMAGE_MEMORY_BARRIER (45) pNext: const void* = NULL srcAccessMask: VkAccessFlags = 4096 (VK_ACCESS_TRANSFER_WRITE_BIT) dstAccessMask: VkAccessFlags = 4096 (VK_ACCESS_TRANSFER_WRITE_BIT) oldLayout: VkImageLayout = VK_IMAGE_LAYOUT_TRANSFER_DST_OPTIMAL (7) newLayout: VkImageLayout = VK_IMAGE_LAYOUT_GENERAL (1) srcQueueFamilyIndex: uint32_t = 4294967295 dstQueueFamilyIndex: uint32_t = 4294967295 image: VkImage = 0x2df55e0 subresourceRange: VkImageSubresourceRange = 0x7ffc882367e0: aspectMask: VkImageAspectFlags = 1 (VK_IMAGE_ASPECT_COLOR_BIT) baseMipLevel: uint32_t = 0 levelCount: uint32_t = 1 baseArrayLayer: uint32_t = 0 layerCount: uint32_t = 1 This fixes all the CTS dEQP-VK.memory.pipeline_barrier.transfer_dst tests here, not sure if this is a too large hammer. Reviewed-by: Bas Nieuwenhuizen <[email protected]> Cc: "17.0" <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* r600: fix a compilation warning in r600_screen_create()Samuel Pitoiset2017-01-301-1/+1
| | | | | | | | Should be r600_common_screen instead of r600_screen. Fixes: 80157a2c20 ("gallium/radeon: clean up r600_query_init_backend_mask") Signed-off-by: Samuel Pitoiset <[email protected]> Reviewed-by: Nicolai Hähnle <[email protected]>
* gallium/radeon: merge dirty_fb_counter and dirty_tex_descriptor_counterMarek Olšák2017-01-304-41/+22
| | | | | | to simplify things in draw_vbo a little Reviewed-by: Nicolai Hähnle <[email protected]>
* winsys/radeon: clamp vram_vis_size to 256MBMarek Olšák2017-01-301-1/+1
| | | | | | the value from the kernel is wrong Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: handle count_from_stream_output in a few IA_MULTI_VGT_PARAM casesMarek Olšák2017-01-301-2/+4
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: don't invoke DCC decompression in update_all_texture_descriptorsMarek Olšák2017-01-301-5/+6
| | | | | | | | | | | | | This fixes a bug uncovered by the 17-part patch series, specifically: "gallium/radeon: merge dirty_fb_counter and dirty_tex_descriptor_counter" If dirty_tex_counter has been updated and set_shader_image invokes DCC decompression, the DCC decompression itself checks the counter and updates descriptors, which in turn invokes the same DCC decompression. The blitter can't handle the recursion and the driver eventually crashes. Cc: 17.0 <[email protected]> Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: fold info->indirect conditionals into the last one in draw_vboMarek Olšák2017-01-301-12/+13
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: atomize the scratch buffer stateMarek Olšák2017-01-306-29/+32
| | | | | | | | | The update frequency is very low. Difference: Only account for the size when allocating a new one and when starting a new IB, and check for NULL. (v3) Reviewed-by: Nicolai Hähnle <[email protected]>
* r600: Fix stack overflowBartosz Tomczyk2017-01-301-1/+1
| | | | | | | | | Commit 7b5878ee0491e7a93914389a8369cd6752b9757d increased number of outputs to 64, but left output array intact. This caused stack overflow when number of outputs is bigger then 32. Found by ASAN. Cc: "12.0 13.0 17.0" <[email protected]> Reviewed-by: Nicolai Hähnle <[email protected]>
* gallium/radeon: add new HUD queries for monitoring the CPSamuel Pitoiset2017-01-304-3/+80
| | | | | | | | | | There are even more counters in the CP_STAT register but I think these ones are enough for now. v2: only read (and expose) CP_STAT on VI+ Signed-off-by: Samuel Pitoiset <[email protected]> Reviewed-by: Marek Olšák <[email protected]>
* gallium/radeon: add new GPU-sdma-busy HUD querySamuel Pitoiset2017-01-304-1/+21
| | | | | | | | | For simplicity, GPU-sdma-busy will return 0 on previous gens. v2: only read SRBM_STATUS2 on Evergreen+ Signed-off-by: Samuel Pitoiset <[email protected]> Reviewed-by: Marek Olšák <[email protected]>
* gallium/radeon: rename grbm to mmio in the gpu load pathSamuel Pitoiset2017-01-302-32/+33
| | | | | | | | We also want to monitor other MMIO counters like SRBM_STATUS2 in order to know if SDMA is busy. Signed-off-by: Samuel Pitoiset <[email protected]> Reviewed-by: Marek Olšák <[email protected]>
* winsys/amdgpu: add a fast exit path into amdgpu_cs_add_bufferMarek Olšák2017-01-302-0/+21
| | | | | | The time spent in the function dropped by 37% for torcs. Reviewed-by: Nicolai Hähnle <[email protected]>
* winsys/amdgpu: do not iterate twice when adding fence dependenciesSamuel Pitoiset2017-01-301-31/+32
| | | | | | | | The perf difference is very small, 3.25->2.84% in amdgpu_cs_flush() in the DXMD benchmark. Signed-off-by: Samuel Pitoiset <[email protected]> Reviewed-by: Marek Olšák <[email protected]>
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-28 15:55:25 +0000