lgc: Add ShaderDbg support for ElfLinker
lgc: Cooperative rowacc code clean up
lgc: finish builder-ification of PatchInOutImportExport
lgc: Fix cooperative matrix enum build error
lgc: Move cooperative matrix enum into lgcDialect
lgc: SamplePosOffset should be v2f32 instead of i32
lgc: Shrink down include list.
lgc: Support waveMatrix rowacc expand
lgc: Support waveMatrix rowacc scalar
lgc: Support waveMatrix rowacc splat (fill)
lgc: Support waveMatrix rowacc sum accumulate
lgc: Treat patch + GS w/o tess to be point mode
lgc: Undo moving code into LowerSubgroupOps
lgc: Use llvm-dialects with new name mangling
llpc: fix a corner case with LDS in task/mesh pipelines
[Continuations] Add _AmdContStackLoadLastUse builtin support.
[Continuations] Add _cont_-prefixed contants.
[Continuations] Do not change function name during linking
[Continuations] Do not remove types metadata
[Continuations] Fix continuations-units test warning.
[Continuations] Fix undef / poison test changes.
[Continuations] Fix use-after-free
[Continuations] Handle InstanceInclusionMaskOp
[Continuations] Handle mangled _AmdComplete calls.
[Continuations] Implement lgc.cps lowering in DXILContPostProcess.
[Continuations] Lower GetResumePoint
[Continuations] Pass GPURT library to LowerRaytracingPipelinePass
[Continuations] Process lgc.cps function with no await call correctly
[Continuations] Refactor DXILContIntrinsicPrepare
[Continuations] Refactor pipeline-scope payload size handling
[Continuations] Remove continuation.complete intrinsic.
[Continuations] Remove more deprecated 'startswith' calls.
[Continuations] Remove types metadata sometimes
[Continuations] Replace payload spill allocation with alloca.
[Continuations] Terminate shader after inlining _cont_SetupRayGen.
[RT] Bad code on calculating hash for RT pipeline which contains library
[RT] Fix tracedata for function return
[RT] Remove all unnecessary functions for Traversal
[RT] Replace deprecated startswith call.
[RT] Swap InstanceIndexOp and InstanceIdOp
[RT] Use more GpuRt Funcs instead of hardcoded ir.
[RT] use the getinstance index/id from gpurt directly.
Abort for -emit-llvm and color export shader
Add new field 'numComponents' to 'InOutInfo'
Add td file in cmake source files
Add the alignment for cooperativeMatrixLoad and cooperativeMatrixStore
Changes for non-intrisic debug info mode
Cleanup-gfx9: Remove check for major >= 9
Correct the condition for executing color export operation
Move shared/continuations/ to llvmraytracing/
Port the optimization for the pattern: ((b==0.0 ? 0.0 : a) * (a==0.0 ? 0.0 : b))
Preserve necessary _Amd function calls
rayquery: Add first version of lgc.rtq dialect
Refactor handleColorExportInstructions
Refine VKGC pipeline dump and vk sections in vfx
ShaderStages test check lines more flexible
Support PipelineShaderOptions::clientHash in pipeline hash calculation
Update llpc tests to allow for upstream llvm changes
Update to handle changes in PassModel parameters
Use VKI_ macros in llpc_version
Wrap PipelineContext and RaytracingContext in SPIRVReader.
Bump Gpurt Version to 45
Update llvm-dialects.
Fix Attributes for Cooperative Matrices
Fix fp and packed formats' norm emulation issue.
Fix RT demo perf-drop due to handling global access for null desc
Fix some CTS failures for dynamicDualSource on NAVI10