[llvm-branch-commits] [llvm] [AMDGPU] Use different name scope for MIMGEncoding and MIMGDim (PR #171166)

2025-12-08 Thread Mirko Brkušanin via llvm-branch-commits
https://github.com/mbrkusanin created https://github.com/llvm/llvm-project/pull/171166 Use new scoped enums with type set to uint8_t. From 46da7be139d82c2271688faaa5cb8364be91f930 Mon Sep 17 00:00:00 2001 From: Mirko Brkusanin Date: Mon, 8 Dec 2025 18:05:45 +0100 Subject: [PATCH] [AMDGPU] Use

[llvm-branch-commits] [llvm] [AMDGPU] Use different name scope for MIMGBaseOpcode (PR #170904)

2025-12-08 Thread Mirko Brkušanin via llvm-branch-commits
https://github.com/mbrkusanin updated https://github.com/llvm/llvm-project/pull/170904 From 882c6ec5c217e4a29963ba735a7068b28d92bf96 Mon Sep 17 00:00:00 2001 From: Mirko Brkusanin Date: Fri, 5 Dec 2025 19:21:27 +0100 Subject: [PATCH 1/2] [AMDGPU] Use different name scope for MIMGBaseOpcode Def

[llvm-branch-commits] [llvm] [AMDGPU] Use different name scope for MIMGBaseOpcode (PR #170904)

2025-12-05 Thread Mirko Brkušanin via llvm-branch-commits
https://github.com/mbrkusanin created https://github.com/llvm/llvm-project/pull/170904 Define MIMGBaseOpcode values with `enum class` instead of regular `enum` so they will be in a separate name scope from regular opcodes. These two groups of opcodes should not be mixed and keeping them in diffe

[llvm-branch-commits] [llvm] AMDGPU/GlobalISel: RegBankLegalize rules for G_FABS and G_FNEG (PR #168411)

2025-11-28 Thread Mirko Brkušanin via llvm-branch-commits
https://github.com/mbrkusanin approved this pull request. https://github.com/llvm/llvm-project/pull/168411 ___ llvm-branch-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-branch-commits

[llvm-branch-commits] [llvm] [AMDGPU] Support image_bvh8_intersect_ray instruction and intrinsic. (PR #130041)

2025-04-05 Thread Mirko Brkušanin via llvm-branch-commits
@@ -1509,18 +1509,18 @@ multiclass MIMG_Gather : MIMG_Gather; -class MIMG_IntersectRay_Helper { - int num_addrs = !if(Is64, !if(IsA16, 9, 12), !if(IsA16, 8, 11)); +class MIMG_IntersectRay_Helper { + int num_addrs = !if(isBVH8, 11, !if(Is64, !if(IsA16, 9, 12), !if(IsA16,

[llvm-branch-commits] [llvm] [AMDGPU] Dynamic VGPR support for llvm.amdgcn.cs.chain (PR #130094)

2025-03-14 Thread Mirko Brkušanin via llvm-branch-commits
@@ -1200,34 +1225,79 @@ bool AMDGPUCallLowering::lowerTailCall( if (!IsSibCall) CallSeqStart = MIRBuilder.buildInstr(AMDGPU::ADJCALLSTACKUP); - unsigned Opc = - getCallOpcode(MF, Info.Callee.isReg(), true, ST.isWave32(), CalleeCC); + bool IsChainCall = AMDGPU::isC

[llvm-branch-commits] [llvm] [AMDGPU] Dynamic VGPR support for llvm.amdgcn.cs.chain (PR #130094)

2025-03-14 Thread Mirko Brkušanin via llvm-branch-commits
@@ -1200,34 +1225,79 @@ bool AMDGPUCallLowering::lowerTailCall( if (!IsSibCall) CallSeqStart = MIRBuilder.buildInstr(AMDGPU::ADJCALLSTACKUP); - unsigned Opc = - getCallOpcode(MF, Info.Callee.isReg(), true, ST.isWave32(), CalleeCC); + bool IsChainCall = AMDGPU::isC