diff --git a/llvm/test/CodeGen/AMDGPU/match-perm-extract-vector-elt-bug.ll b/llvm/test/CodeGen/AMDGPU/match-perm-extract-vector-elt-bug.ll index 78757b00cea44d..c7a831185b83c6 100644 --- a/llvm/test/CodeGen/AMDGPU/match-perm-extract-vector-elt-bug.ll +++ b/llvm/test/CodeGen/AMDGPU/match-perm-extract-vector-elt-bug.ll @@ -9,7 +9,6 @@ define amdgpu_kernel void @test(ptr addrspace(1) %src, ptr addrspace(1) %dst) { ; GFX9-NEXT: s_load_dword s7, s[4:5], 0x1c ; GFX9-NEXT: s_load_dword s8, s[4:5], 0x38 ; GFX9-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x0 -; GFX9-NEXT: v_mov_b32_e32 v6, -1 ; GFX9-NEXT: s_waitcnt lgkmcnt(0) ; GFX9-NEXT: s_and_b32 s4, s7, 0xffff ; GFX9-NEXT: s_mul_i32 s6, s6, s4 @@ -21,22 +20,14 @@ define amdgpu_kernel void @test(ptr addrspace(1) %src, ptr addrspace(1) %dst) { ; GFX9-NEXT: v_add_co_u32_e32 v0, vcc, s0, v4 ; GFX9-NEXT: v_addc_co_u32_e32 v1, vcc, v1, v5, vcc ; GFX9-NEXT: global_load_dwordx4 v[0:3], v[0:1], off -; GFX9-NEXT: v_mov_b32_e32 v7, s3 +; GFX9-NEXT: v_mov_b32_e32 v6, s3 ; GFX9-NEXT: v_add_co_u32_e32 v4, vcc, s2, v4 -; GFX9-NEXT: v_addc_co_u32_e32 v5, vcc, v7, v5, vcc +; GFX9-NEXT: v_addc_co_u32_e32 v5, vcc, v6, v5, vcc ; GFX9-NEXT: s_waitcnt vmcnt(0) -; GFX9-NEXT: v_xor_b32_e32 v8, -1, v3 -; GFX9-NEXT: v_xor_b32_sdwa v3, v3, v6 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:DWORD -; GFX9-NEXT: v_xor_b32_e32 v9, -1, v2 -; GFX9-NEXT: v_xor_b32_sdwa v2, v2, v6 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:DWORD -; GFX9-NEXT: v_xor_b32_e32 v10, -1, v1 -; GFX9-NEXT: v_xor_b32_sdwa v1, v1, v6 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:DWORD -; GFX9-NEXT: v_xor_b32_e32 v11, -1, v0 -; GFX9-NEXT: v_xor_b32_sdwa v0, v0, v6 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:DWORD -; GFX9-NEXT: v_or_b32_sdwa v3, v8, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:DWORD -; GFX9-NEXT: v_or_b32_sdwa v2, v9, v2 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:DWORD -; GFX9-NEXT: v_or_b32_sdwa v1, v10, v1 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:DWORD -; GFX9-NEXT: v_or_b32_sdwa v0, v11, v0 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:DWORD +; GFX9-NEXT: v_not_b32_e32 v3, v3 +; GFX9-NEXT: v_not_b32_e32 v2, v2 +; GFX9-NEXT: v_not_b32_e32 v1, v1 +; GFX9-NEXT: v_not_b32_e32 v0, v0 ; GFX9-NEXT: global_store_dwordx4 v[4:5], v[0:3], off ; GFX9-NEXT: s_endpgm ; @@ -46,7 +37,6 @@ define amdgpu_kernel void @test(ptr addrspace(1) %src, ptr addrspace(1) %dst) { ; GFX10-NEXT: s_load_dword s7, s[4:5], 0x1c ; GFX10-NEXT: s_load_dword s8, s[4:5], 0x38 ; GFX10-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x0 -; GFX10-NEXT: v_mov_b32_e32 v6, -1 ; GFX10-NEXT: s_waitcnt lgkmcnt(0) ; GFX10-NEXT: s_and_b32 s4, s7, 0xffff ; GFX10-NEXT: s_mul_i32 s6, s6, s4 @@ -59,18 +49,10 @@ define amdgpu_kernel void @test(ptr addrspace(1) %src, ptr addrspace(1) %dst) { ; GFX10-NEXT: v_add_co_ci_u32_e32 v5, vcc_lo, s3, v5, vcc_lo ; GFX10-NEXT: global_load_dwordx4 v[0:3], v[0:1], off ; GFX10-NEXT: s_waitcnt vmcnt(0) -; GFX10-NEXT: v_xor_b32_e32 v7, -1, v3 -; GFX10-NEXT: v_xor_b32_sdwa v3, v3, v6 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:DWORD -; GFX10-NEXT: v_xor_b32_e32 v8, -1, v2 -; GFX10-NEXT: v_xor_b32_sdwa v2, v2, v6 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:DWORD -; GFX10-NEXT: v_xor_b32_e32 v9, -1, v1 -; GFX10-NEXT: v_xor_b32_sdwa v1, v1, v6 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:DWORD -; GFX10-NEXT: v_xor_b32_e32 v10, -1, v0 -; GFX10-NEXT: v_xor_b32_sdwa v0, v0, v6 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:DWORD -; GFX10-NEXT: v_or_b32_sdwa v3, v7, v3 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:DWORD -; GFX10-NEXT: v_or_b32_sdwa v2, v8, v2 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:DWORD -; GFX10-NEXT: v_or_b32_sdwa v1, v9, v1 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:DWORD -; GFX10-NEXT: v_or_b32_sdwa v0, v10, v0 dst_sel:DWORD dst_unused:UNUSED_PAD src0_sel:WORD_0 src1_sel:DWORD +; GFX10-NEXT: v_not_b32_e32 v3, v3 +; GFX10-NEXT: v_not_b32_e32 v2, v2 +; GFX10-NEXT: v_not_b32_e32 v1, v1 +; GFX10-NEXT: v_not_b32_e32 v0, v0 ; GFX10-NEXT: global_store_dwordx4 v[4:5], v[0:3], off ; GFX10-NEXT: s_endpgm ; @@ -95,32 +77,10 @@ define amdgpu_kernel void @test(ptr addrspace(1) %src, ptr addrspace(1) %dst) { ; GFX11-NEXT: v_add_co_ci_u32_e32 v5, vcc_lo, s3, v5, vcc_lo ; GFX11-NEXT: global_load_b128 v[0:3], v[0:1], off ; GFX11-NEXT: s_waitcnt vmcnt(0) -; GFX11-NEXT: v_lshrrev_b32_e32 v6, 16, v0 -; GFX11-NEXT: v_lshrrev_b32_e32 v7, 16, v1 -; GFX11-NEXT: v_lshrrev_b32_e32 v8, 16, v2 -; GFX11-NEXT: v_xor_b32_e32 v9, -1, v3 -; GFX11-NEXT: v_lshrrev_b32_e32 v3, 16, v3 -; GFX11-NEXT: v_xor_b32_e32 v2, -1, v2 -; GFX11-NEXT: v_xor_b32_e32 v1, -1, v1 -; GFX11-NEXT: v_xor_b32_e32 v0, -1, v0 -; GFX11-NEXT: v_xor_b32_e32 v8, -1, v8 -; GFX11-NEXT: v_xor_b32_e32 v3, -1, v3 -; GFX11-NEXT: v_xor_b32_e32 v7, -1, v7 -; GFX11-NEXT: v_xor_b32_e32 v6, -1, v6 -; GFX11-NEXT: v_and_b32_e32 v9, 0xffff, v9 -; GFX11-NEXT: v_and_b32_e32 v2, 0xffff, v2 -; GFX11-NEXT: v_and_b32_e32 v1, 0xffff, v1 -; GFX11-NEXT: v_and_b32_e32 v0, 0xffff, v0 -; GFX11-NEXT: v_lshlrev_b32_e32 v3, 16, v3 -; GFX11-NEXT: v_lshlrev_b32_e32 v8, 16, v8 -; GFX11-NEXT: v_lshlrev_b32_e32 v7, 16, v7 -; GFX11-NEXT: v_lshlrev_b32_e32 v6, 16, v6 -; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_4) | instskip(NEXT) | instid1(VALU_DEP_4) -; GFX11-NEXT: v_or_b32_e32 v3, v9, v3 -; GFX11-NEXT: v_or_b32_e32 v2, v2, v8 -; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_4) | instskip(NEXT) | instid1(VALU_DEP_4) -; GFX11-NEXT: v_or_b32_e32 v1, v1, v7 -; GFX11-NEXT: v_or_b32_e32 v0, v0, v6 +; GFX11-NEXT: v_not_b32_e32 v3, v3 +; GFX11-NEXT: v_not_b32_e32 v2, v2 +; GFX11-NEXT: v_not_b32_e32 v1, v1 +; GFX11-NEXT: v_not_b32_e32 v0, v0 ; GFX11-NEXT: global_store_b128 v[4:5], v[0:3], off ; GFX11-NEXT: s_nop 0 ; GFX11-NEXT: s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) diff --git a/revert_patches.txt b/revert_patches.txt index 92cf35fc4145f7..88bc092e8140c2 100644 --- a/revert_patches.txt +++ b/revert_patches.txt @@ -31,6 +31,9 @@ Revert: 8009bbe [2024-05-01T03:18:40.779Z] /jenkins/workspace/compiler-psdb-amd-staging_2/Libs/rocThrust/thrust/../thrust/optional.h:2756:11: error: no member named 'construct' in 'optional' [2024-05-01T03:18:40.779Z] 2756 | this->construct(std::forward(args)...); + fd87d765c045 [Clang][Sema] Don't build CXXDependentScopeMemberExprs for potent + acf5ad2a4ed9 [Clang][Sema] Diagnose current instantiation used as an incomplete + Contacts: Sam, Konstantin --- @@ -45,8 +48,3 @@ Revert: 41f9c78 Konstantin --- -Reverts: breaks oclConformance - 41466a177a95 [SelectionDAG] Correct the implementation of m_AllOnes. (#90776) - SWDEV-460097 - merges: revert : 41466a177a95 [SelectionDAG] Correct the implementation of m_AllOnes. (#90776) -