summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* anv: Disable snooping for allocator pools againKristian Høgsberg Kristensen2016-02-121-2/+0
| | | | | | The race we were seeing on cherryview was caused by the multi-submit problem with fences. We can now turn snooping off again an rely on clflush and we intended.
* anv: Submit fence bo only after all command buffersKristian Høgsberg Kristensen2016-02-121-9/+9
| | | | | | We were submitting the fence bo after each command buffer in a multi command buffer submit, causing us to occasionally complete the fence too early.
* anv: Implement VkPipelineCacheKristian Høgsberg Kristensen2016-02-125-198/+589
| | | | | | | We hash the input SPIR-V, specialization constants, entrypoint and the shader key using SHA1 to determine a unique identifier for the combination. A VkPipelineCache is then a hash table mapping these identifiers to the corresponding prog_data and kernel data.
* anv/meta_blit: Remove references to clearingChad Versace2016-02-121-13/+10
| | | | | | | | Long ago, the blit code used to handle clearing and blitting. - Fix any comments that refer to clearing. - Rename shader var 'attr' to 'tex_pos'. The name 'attr' is an artifact of the time when the shader was used for blitting as well as clearing.
* anv/meta_blit: Coalesce glsl_vec4_type varsChad Versace2016-02-121-12/+9
| | | | | | | Just a refactor. No behavior change. Several expressions have the same value: they point to glsl_vec4_type(). Coalesce them into a single variable.
* anv/device: clflush simple batches if !LLCJason Ekstrand2016-02-121-0/+2
|
* anv: Add a clfush_range helper functionJason Ekstrand2016-02-121-7/+10
|
* nir/spirv/glsl: Clean up the row-skipping swizzle logic a bitJason Ekstrand2016-02-121-10/+4
|
* anv/meta: Move blit code to anv_meta_blit.cChad Versace2016-02-124-1428/+1453
| | | | | | | | The clear code lived in anv_meta_clear.c. The resolve code in anv_meta_resolve.c. Only the blit code lived in anv_meta.c, alongside the shareed meta code. This is just a copy-paste patch. No change in behavior.
* anv/meta: Hardcode smooth texcoord interpolation in blit shadersChad Versace2016-02-121-4/+3
| | | | | | | Trivial cleanup. No change in behavior. Function argument 'attr_flat', in anv_meta.c:build_nir_vertex_shader(), was always false.
* anv/device: Use a normal BO in submit_simple_batchJason Ekstrand2016-02-111-10/+12
|
* anv: Add a vk_icdGetInstanceProcAddr entrypointJason Ekstrand2016-02-111-0/+14
| | | | | | Aparently there are some issues in symbol resolution if an application packages its own loader and you have a system-installed one. I don't really understand the details, but it's not onorous to add.
* anv/event: Use a 64-bit valueJason Ekstrand2016-02-112-2/+2
| | | | | | | The immediate write from PIPE_CONTROL is 64-bits at least on BDW. This used to work on 64-bit archs because the compiler would align the following anv_state struct up for us. However, in 32-bit builds, they overlap and it causes problems.
* gen8/pipeline: Properly set bits in PS_EXTRA for W, depth, and samaple maskJason Ekstrand2016-02-112-1/+10
|
* nir/spirv: Allow the clip distance capability.Jason Ekstrand2016-02-111-0/+1
|
* gen8/pipeline: Pull gs_vertex_count from prog_dataJason Ekstrand2016-02-113-3/+1
|
* Merge remote-tracking branch 'mesa-public/master' into vulkanJason Ekstrand2016-02-1160-1636/+976
|\
| * i965/gs: Pass VerticesIn though prog_dataJason Ekstrand2016-02-113-1/+5
| | | | | | | | Reviewed-by: Kenneth Graunke <[email protected]>
| * i965/fs: Pass usage of depth, W, and sample mask through prog_dataJason Ekstrand2016-02-116-20/+30
| | | | | | | | | | | | | | | | | | | | We really need to stop pulling information directly out of shaders for state setup. For one thing, if we want any sort of an on-disk shader cache, having all of this metadata in one place is going to be crucial. Also, passing it all through prog_data cleans up the compiler <-> state setup API substantially. Reviewed-by: Kenneth Graunke <[email protected]>
| * i965/fs: Refactor setup_payload_gen6 to assume FSJason Ekstrand2016-02-113-15/+15
| | | | | | | | | | | | | | | | | | It's extremely FS specific so the fact that we have a stage check in the middle of it is rather bogus. While were here, we rename setup_payload_gen4 and setup_payload_gen6 to make it obvious that they are both FS specific. Reviewed-by: Kenneth Graunke <[email protected]>
| * nv50,nvc0: remove unused parameter in nvXX_state_validate()Samuel Pitoiset2016-02-118-14/+10
| | | | | | | | | | | | | | | | This 'words' parameter is there since 2011 but it has never been used. While we are at it, get rid of the extern declaration. Signed-off-by: Samuel Pitoiset <[email protected]> Reviewed-by: Ilia Mirkin <[email protected]>
| * glsl: don't validate interface blocks twiceTimothy Arceri2016-02-121-23/+21
| | | | | | | | | | | | | | We already check for opaque types so don't recheck for atomics and images. Reviewed-by: Iago Toral Quiroga <[email protected]>
| * glsl: remove duplicate embedded struct validationTimothy Arceri2016-02-123-51/+17
| | | | | | | | | | | | | | | | | | | | | | Commit c98deb18d5836f in 2010 disallowed embedded struct definitions in ES. Then in 2013 d9bb8b7b56ce65b disallowed it for everything but GLSL 1.10. Commit c98deb18d5836f seemed the cleanest way to do the check so its been extended to cover GL and the other version has been removed. Reviewed-by: Iago Toral Quiroga <[email protected]>
| * include,gallium: Remove pre-MSVC 2013 compatibility.Jose Fonseca2016-02-117-649/+10
| | | | | | | | | | Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Roland Scheidegger <[email protected]>
| * scons: Eliminate MSVC2008 compatibility.Jose Fonseca2016-02-117-22/+6
| | | | | | | | | | Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Roland Scheidegger <[email protected]>
| * configure: Eliminate MSVC2008 compatibility.Jose Fonseca2016-02-116-14/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We no longer need to build any part of Mesa with Windows SDK 7.0.7600 or MSVC 2008. MSVC 2013 will be the oldest we support. In practice this means people are now free to declare variables in the middle of blocks, on the whole Mesa tree. Care should still be taken with variable length arrays and void pointer arithmetic. Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Roland Scheidegger <[email protected]> Hella-acked-by: Ian Romanick <[email protected]>
| * i965: ir: dump floats as %-g rather than %f, so we can see denormalsChris Forbes2016-02-111-1/+1
| | | | | | | | | | Signed-off-by: Chris Forbes <[email protected]> Reviewed-by: Ben Widawsky <[email protected]>
| * i965/gen7: Require kernel cmd_parser 5 for ARB_compute_shaderJordan Justen2016-02-111-1/+2
| | | | | | | | | | | | | | | | The indirect dispatch registers were whitelisted in command parser version 5. (Version 5 is available as of Linux 4.4) Signed-off-by: Jordan Justen <[email protected]> Reviewed-by: Kenneth Graunke <[email protected]>
| * st/mesa: release GLSL IR in LinkShader after it's not neededMarek Olšák2016-02-111-0/+4
| | | | | | | | Reviewed-by: Brian Paul <[email protected]>
| * mesa: call build_program_resource_list inside Driver.LinkShaderMarek Olšák2016-02-114-3/+11
| | | | | | | | | | | | to allow LinkShader to free the GLSL IR. Reviewed-by: Ian Romanick <[email protected]>
| * st/mesa: use correct pipe functions to create tess shadersMarek Olšák2016-02-113-7/+27
| | | | | | | | | | | | | | | | | | Broken by one of my cleanups. Spotted by luck. Radeonsi doesn't care, because all shader create callbacks go to the same function. Reviewed-by: Brian Paul <[email protected]>
| * gallium/radeon: drop support for LLVM 3.5Marek Olšák2016-02-119-186/+11
| | | | | | | | | | | | Reviewed-by: Nicolai Hähnle <[email protected]> v2: adjust the comment in the amdgpu winsys
| * radeonsi: obtain commonly used LLVM types only onceMarek Olšák2016-02-111-215/+194
| | | | | | | | Reviewed-by: Nicolai Hähnle <[email protected]>
| * radeonsi: cleanup shader codegenMarek Olšák2016-02-111-425/+425
| | | | | | | | | | | | | | | | si_shader_ctx -> ctx type * ptr -> type *ptr si_shader_context *shader -> si_shader_context *ctx Reviewed-by: Nicolai Hähnle <[email protected]>
| * radeonsi: fix a crash when binding a sampler bufferMarek Olšák2016-02-111-1/+2
| | | | | | | | | | | | | | | | | | Buffers don't contain r600_texture. Broken by 7aedbbacae6d3ec3d06735fff2eb66: "radeonsi: put image, fmask, and sampler descriptors into one array" Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=94091
| * docs: add news item and link release notes for 11.1.2Emil Velikov2016-02-112-0/+7
| | | | | | | | Signed-off-by: Emil Velikov <[email protected]>
| * docs: add sha256 checksums for 11.1.2Emil Velikov2016-02-111-1/+2
| | | | | | | | | | Signed-off-by: Emil Velikov <[email protected]> (cherry picked from commit e49dd21bcbabdb330620d48f5915828cfd5eb983)
| * docs: add release notes for 11.1.2Emil Velikov2016-02-111-0/+181
| | | | | | | | | | Signed-off-by: Emil Velikov <[email protected]> (cherry picked from commit 7bcd827806b0816d61122ba3d37dd40178d96d98)
* | anv/pack: Handle case where a struct field covers multiple dwordsKristian Høgsberg Kristensen2016-02-101-4/+7
| | | | | | | | | | We also didn't add start to field.end to get the absolute field end position.
* | Merge remote-tracking branch 'mesa-public/master' into vulkanJason Ekstrand2016-02-1035-383/+437
|\| | | | | | | | | This also reverts commit 1d65abfa582a371558113f699ffbf16d60b64c90 because now NIR handles texture offsets in a much more sane way.
| * nir: Remove the const_offset from nir_tex_instrJason Ekstrand2016-02-108-72/+27
| | | | | | | | | | | | | | | | | | | | | | When NIR was originally drafted, there was no easy way to determine if something was constant or not. The result was that we had lots of special-casing for constant values such as this. Now that load_const instructions are SSA-only, it's really easy to find constants and this isn't really needed anymore. Reviewed-by: Connor Abbott <[email protected]> Reviewed-by: Rob Clark <[email protected]>
| * nir/lower_vec_to_movs: Better report channels handled by insert_movJason Ekstrand2016-02-101-1/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | This fixes two issues. First, we had a use-after-free in the case where the instruction got deleted and we tried to return mov->dest.write_mask. Second, in the case where we are doing a self-mov of a register, we delete those channels that are moved to themselves from the write-mask. This means that those channels aren't reported as being handled even though they are. We now stash off the write-mask before remove unneeded channels so that they still get reported as handled. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=94073 Reviewed-by: Matt Turner <[email protected]> Cc: "11.0 11.1" <[email protected]>
| * radeonsi: don't emit unnecessary NULL exports for unbound targets (v3)Marek Olšák2016-02-101-26/+68
| | | | | | | | | | | | | | | | v2: remove semantic index == 0 checks add the else statement to remove shadowing of args v3: fix fbo-alphatest-nocolor regression Reviewed-by: Nicolai Hähnle <[email protected]> (v2)
| * i965: Make sure we blit a full compressed blockBen Widawsky2016-02-101-0/+14
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This fixes an assertion failure in [at least] one of the Unreal Engine Linux demo/games that uses DXT1 compression. Specifically, the "Vehicle Game". At some point, the game ends up trying to blit mip level whose size is 2x2, which is smaller than a DXT1 block. As a result, the assertion in the blit path is triggered. It should be safe to simply make sure we align the width and height, which is sadly an example of compression being less efficient. NOTE: The demo seems to work fine without the assert, and therefore release builds of mesa wouldn't stumble over this. Perhaps there is some unnoticeable corruption, but I had trouble spotting it. Thanks to Jason for looking at my backtrace and figuring out what was going on. v2: Use NPOT alignment to make sure ASTC is handled properly (Ilia) Remove comment about how this doesn't fix other bugs, because it does. Cc: "11.0 11.1" <[email protected] Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=93358 Signed-off-by: Ben Widawsky <[email protected]> Tested-by: Matt Turner <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
| * radeon/uvd: silence a warningMarek Olšák2016-02-101-0/+3
| |
| * r300g: silence warningsMarek Olšák2016-02-101-2/+0
| |
| * meta/decompress: Don't pollute the renderbuffer namespaceIan Romanick2016-02-101-6/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | tl;dr: For many types of GL object, we can *NEVER* use the Gen function. In OpenGL ES (all versions!) and OpenGL compatibility profile, applications don't have to call Gen functions. The GL spec is very clear about how you can mix-and-match generated names and non-generated names: you can use any name you want for a particular object type until you call the Gen function for that object type. Here's the problem scenario: - Application calls a meta function that generates a name. The first Gen will probably return 1. - Application decides to use the same name for an object of the same type without calling Gen. Many demo programs use names 1, 2, 3, etc. without calling Gen. - Application calls the meta function again, and the meta function replaces the data. The application's data is lost, and the app fails. Have fun debugging that. Fixes piglit 'object-namespace-pollution glGetTexImage-compressed renderbuffer' test. Signed-off-by: Ian Romanick <[email protected]> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=92363 Reviewed-by: Anuj Phogat <[email protected]>
| * meta: Use internal functions for renderbuffer accessIan Romanick2016-02-102-14/+8
| | | | | | | | | | Signed-off-by: Ian Romanick <[email protected]> Reviewed-by: Anuj Phogat <[email protected]>
| * meta/decompress: Track renderbuffer using gl_renderbuffer instead of GL API ↵Ian Romanick2016-02-102-5/+11
| | | | | | | | | | | | | | object handle Signed-off-by: Ian Romanick <[email protected]> Reviewed-by: Anuj Phogat <[email protected]>
| * i965/meta: Don't pollute the renderbuffer namespaceIan Romanick2016-02-103-16/+11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | tl;dr: For many types of GL object, we can *NEVER* use the Gen function. In OpenGL ES (all versions!) and OpenGL compatibility profile, applications don't have to call Gen functions. The GL spec is very clear about how you can mix-and-match generated names and non-generated names: you can use any name you want for a particular object type until you call the Gen function for that object type. Here's the problem scenario: - Application calls a meta function that generates a name. The first Gen will probably return 1. - Application decides to use the same name for an object of the same type without calling Gen. Many demo programs use names 1, 2, 3, etc. without calling Gen. - Application calls the meta function again, and the meta function replaces the data. The application's data is lost, and the app fails. Have fun debugging that. Signed-off-by: Ian Romanick <[email protected]> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=92363 Reviewed-by: Anuj Phogat <[email protected]>
generated by cgit v1.2.3 (git 2.39.1) at 2024-11-30 00:39:08 +0000