Promotion from AMD internal branch for 2024.Q1.1 #2893
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Update gfx900-based lgc tests to gfx1010 Update hasher with some missing options Update llvm-dialects Handle constant GEPs into the RT LDS allocation Implement color-related compatibility mode built-in variables. Improve in/out matching printing info Include GNUInstallDirs for llpc_version Increase lgc::Options size to fix build error Make lgc::ShaderStage more class-like Make ShaderStage optional in some places Mask off type field of null descriptor Prepare for trans OpAccessChain with multi value map Recalculate the arraysize when the dimension exceeds 1. Refine build pipeline code Refine Vertex fetch. Support ext formats Remove unused class member m_atomicInsts Rename pipeline dump file for color export shader Repair the malfunctioning oglp OSIBuild pipeline Replace deprecated starts/endswith Replace the IA_MULTI_VGT_PARAM_PIPED Return full pipeline when enabling xfb Set noReciprocalFragCoordW Simplify how we ensure a single unified return block for outputs Support 'start-lower-after' for llpc spv lower Support atomic counter array with multiple dimensions Support disablePerCompFetch in uber fetch shader Support waveMatrix_Fill Use a stable function signature when compiling libraries Use ShaderStageMask for masks Use strided buffers for shader record buffer [Continuations] Add _AmdContStack* interface [Continuations] Add global memory payload test. [Continuations] Add header for gpurt [Continuations] Add llvm_unreachable [Continuations] Add LowerSwitchPass [Continuations] Add ShaderStageHelper class. [Continuations] Add store-to-load forwarding [Continuations] Change lgc.cps. stack lowering to work with i32 offsets. [Continuations] Cleanup uses of amd.dx prefix [Continuations] DX: Directly access cont state on the CPS stack [CONTINUATIONS] Fix Gpurt hlsl spirv compiling [CONTINUATIONS] Guard getStructName() with hasName() check [Continuations] Merge payload type MD removal into `fixupDxilMetadata`. [CONTINUATIONS] Move setLocalRootIndex at shader entry out of LowerRayTracingPipeline [Continuations] Move type helpers together [CONTINUATIONS] Pass RCR/ShaderIndex to lgc.cps.jump [Continuations] Remove CpsMutator from LowerRaytracingPipelinePass. [Continuations] Remove DialectUtils namespace. [Continuations] Remove getSystemData intrinsic [Continuations] Remove using namespace in CpsStackLowering [CONTINUATIONS] Rename dxil.payload.type [Continuations] Simplify some assertions. [Continuations] Specify namespace for llvm_dialects [Continuations] Support compilation of launch kernel [CONTINUATIONS] Switch to LgcRt HitAttribute metadata [Continuations] Unify GPURT function discovery + signature precondition checks in LowerRaytracingPipeline. [Continuations] Use Arg::getParent to get function [Continuations] Use isa with multiple args [Continuations]Add Function type Add constant load dword intrinsics Add floatOpWithRoundMode intrinsic for lgc Add helper for getWithSamePointeeType Add LibraryMode and library summaries Add type map for struct/array type with sampler Adjust the order of AlwaysInline/GlobalDCE passes Change transCooperativeMatrixKHRFromConstruct to CreateCooperativeMatrixFill Clean up translation of binary opcodes. NFC. Continuations: Implement store to load forwarding Continuations: Lower lgr.rt intrinsic in CleanupContinuations CPS lowing adjustments for dVGPR Don't fill xfb info to xfb_matadata if xfb_offset is not assigned Emit shader data SQTT tokens when required Expose KHR_SHADER_QUAD_CONTROL support in lgc Fix a regression of handling TCS per-vertex outputs Fix assert caused by OpDecorationGroup Fix bug in building shader module usage Fix bug in translate image pointer Fix inline compiles when Continufy mode is enabled Fix struct copy bug during translate spirv OpStore Fix the addrspace of all transitive users of output globals Fix typo in building shader module usage Fix unresolved GPURT dialect in build BVH shaders
Test summary for commit 143d484CTS tests (Failed: 14/212150)
Rhel 8.9, Gfx10Ubuntu navi3x, SrdcvkUbuntu navi2x, Srdcvk |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.