maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/llvm/test/CodeGen/X86/avx-cmp.ll

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

227 lines
7.1 KiB
LLVM
Raw Normal View History

[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx | FileCheck %s
define <8 x i32> @cmp00(<8 x float> %a, <8 x float> %b) nounwind {
; CHECK-LABEL: cmp00:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; CHECK: # %bb.0:
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-NEXT: vcmpltps %ymm1, %ymm0, %ymm0
; CHECK-NEXT: retq
Rename and tidy up tests llvm-svn: 137103
2011-08-09 11:04:23 +08:00
%bincmp = fcmp olt <8 x float> %a, %b
%s = sext <8 x i1> %bincmp to <8 x i32>
ret <8 x i32> %s
}
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
define <4 x i64> @cmp01(<4 x double> %a, <4 x double> %b) nounwind {
; CHECK-LABEL: cmp01:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; CHECK: # %bb.0:
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-NEXT: vcmpltpd %ymm1, %ymm0, %ymm0
; CHECK-NEXT: retq
Rename and tidy up tests llvm-svn: 137103
2011-08-09 11:04:23 +08:00
%bincmp = fcmp olt <4 x double> %a, %b
%s = sext <4 x i1> %bincmp to <4 x i64>
ret <4 x i64> %s
}
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
declare void @scale() nounwind
[X86] Fix some tests using fcmp with undef arguments Prep work for PR40800 (Add UNDEF handling to SelectionDAG::FoldSetCC) llvm-svn: 357278
2019-03-30 01:20:27 +08:00
define void @render(double %a0) nounwind {
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-LABEL: render:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; CHECK: # %bb.0: # %entry
Generalize MergeBlockIntoPredecessor. Replace uses of MergeBasicBlockIntoOnlyPred. Summary: Two utils methods have essentially the same functionality. This is an attempt to merge them into one. 1. lib/Transforms/Utils/Local.cpp : MergeBasicBlockIntoOnlyPred 2. lib/Transforms/Utils/BasicBlockUtils.cpp : MergeBlockIntoPredecessor Prior to the patch: 1. MergeBasicBlockIntoOnlyPred Updates either DomTree or DeferredDominance Moves all instructions from Pred to BB, deletes Pred Asserts BB has single predecessor If address was taken, replace the block address with constant 1 (?) 2. MergeBlockIntoPredecessor Updates DomTree, LoopInfo and MemoryDependenceResults Moves all instruction from BB to Pred, deletes BB Returns if doesn't have a single predecessor Returns if BB's address was taken After the patch: Method 2. MergeBlockIntoPredecessor is attempting to become the new default: Updates DomTree or DeferredDominance, and LoopInfo and MemoryDependenceResults Moves all instruction from BB to Pred, deletes BB Returns if doesn't have a single predecessor Returns if BB's address was taken Uses of MergeBasicBlockIntoOnlyPred that need to be replaced: 1. lib/Transforms/Scalar/LoopSimplifyCFG.cpp Updated in this patch. No challenges. 2. lib/CodeGen/CodeGenPrepare.cpp Updated in this patch. i. eliminateFallThrough is straightforward, but I added using a temporary array to avoid the iterator invalidation. ii. eliminateMostlyEmptyBlock(s) methods also now use a temporary array for blocks Some interesting aspects: - Since Pred is not deleted (BB is), the entry block does not need updating. - The entry block was being updated with the deleted block in eliminateMostlyEmptyBlock. Added assert to make obvious that BB=SinglePred. - isMergingEmptyBlockProfitable assumes BB is the one to be deleted. - eliminateMostlyEmptyBlock(BB) does not delete BB on one path, it deletes its unique predecessor instead. - adding some test owner as subscribers for the interesting tests modified: test/CodeGen/X86/avx-cmp.ll test/CodeGen/AMDGPU/nested-loop-conditions.ll test/CodeGen/AMDGPU/si-annotate-cf.ll test/CodeGen/X86/hoist-spill.ll test/CodeGen/X86/2006-11-17-IllegalMove.ll 3. lib/Transforms/Scalar/JumpThreading.cpp Not covered in this patch. It is the only use case using the DeferredDominance. I would defer to Brian Rzycki to make this replacement. Reviewers: chandlerc, spatel, davide, brzycki, bkramer, javed.absar Subscribers: qcolombet, sanjoy, nemanjai, nhaehnle, jlebar, tpr, kbarton, RKSimon, wmi, arsenm, llvm-commits Differential Revision: https://reviews.llvm.org/D48202 llvm-svn: 335183
2018-06-21 06:01:04 +08:00
; CHECK-NEXT: pushq %rbp
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-NEXT: pushq %rbx
Generalize MergeBlockIntoPredecessor. Replace uses of MergeBasicBlockIntoOnlyPred. Summary: Two utils methods have essentially the same functionality. This is an attempt to merge them into one. 1. lib/Transforms/Utils/Local.cpp : MergeBasicBlockIntoOnlyPred 2. lib/Transforms/Utils/BasicBlockUtils.cpp : MergeBlockIntoPredecessor Prior to the patch: 1. MergeBasicBlockIntoOnlyPred Updates either DomTree or DeferredDominance Moves all instructions from Pred to BB, deletes Pred Asserts BB has single predecessor If address was taken, replace the block address with constant 1 (?) 2. MergeBlockIntoPredecessor Updates DomTree, LoopInfo and MemoryDependenceResults Moves all instruction from BB to Pred, deletes BB Returns if doesn't have a single predecessor Returns if BB's address was taken After the patch: Method 2. MergeBlockIntoPredecessor is attempting to become the new default: Updates DomTree or DeferredDominance, and LoopInfo and MemoryDependenceResults Moves all instruction from BB to Pred, deletes BB Returns if doesn't have a single predecessor Returns if BB's address was taken Uses of MergeBasicBlockIntoOnlyPred that need to be replaced: 1. lib/Transforms/Scalar/LoopSimplifyCFG.cpp Updated in this patch. No challenges. 2. lib/CodeGen/CodeGenPrepare.cpp Updated in this patch. i. eliminateFallThrough is straightforward, but I added using a temporary array to avoid the iterator invalidation. ii. eliminateMostlyEmptyBlock(s) methods also now use a temporary array for blocks Some interesting aspects: - Since Pred is not deleted (BB is), the entry block does not need updating. - The entry block was being updated with the deleted block in eliminateMostlyEmptyBlock. Added assert to make obvious that BB=SinglePred. - isMergingEmptyBlockProfitable assumes BB is the one to be deleted. - eliminateMostlyEmptyBlock(BB) does not delete BB on one path, it deletes its unique predecessor instead. - adding some test owner as subscribers for the interesting tests modified: test/CodeGen/X86/avx-cmp.ll test/CodeGen/AMDGPU/nested-loop-conditions.ll test/CodeGen/AMDGPU/si-annotate-cf.ll test/CodeGen/X86/hoist-spill.ll test/CodeGen/X86/2006-11-17-IllegalMove.ll 3. lib/Transforms/Scalar/JumpThreading.cpp Not covered in this patch. It is the only use case using the DeferredDominance. I would defer to Brian Rzycki to make this replacement. Reviewers: chandlerc, spatel, davide, brzycki, bkramer, javed.absar Subscribers: qcolombet, sanjoy, nemanjai, nhaehnle, jlebar, tpr, kbarton, RKSimon, wmi, arsenm, llvm-commits Differential Revision: https://reviews.llvm.org/D48202 llvm-svn: 335183
2018-06-21 06:01:04 +08:00
; CHECK-NEXT: pushq %rax
[X86] Fix some tests using fcmp with undef arguments Prep work for PR40800 (Add UNDEF handling to SelectionDAG::FoldSetCC) llvm-svn: 357278
2019-03-30 01:20:27 +08:00
; CHECK-NEXT: vmovsd %xmm0, (%rsp) # 8-byte Spill
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-NEXT: xorl %eax, %eax
; CHECK-NEXT: testb %al, %al
; CHECK-NEXT: jne .LBB2_6
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; CHECK-NEXT: # %bb.1: # %for.cond5.preheader
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-NEXT: xorl %ebx, %ebx
Generalize MergeBlockIntoPredecessor. Replace uses of MergeBasicBlockIntoOnlyPred. Summary: Two utils methods have essentially the same functionality. This is an attempt to merge them into one. 1. lib/Transforms/Utils/Local.cpp : MergeBasicBlockIntoOnlyPred 2. lib/Transforms/Utils/BasicBlockUtils.cpp : MergeBlockIntoPredecessor Prior to the patch: 1. MergeBasicBlockIntoOnlyPred Updates either DomTree or DeferredDominance Moves all instructions from Pred to BB, deletes Pred Asserts BB has single predecessor If address was taken, replace the block address with constant 1 (?) 2. MergeBlockIntoPredecessor Updates DomTree, LoopInfo and MemoryDependenceResults Moves all instruction from BB to Pred, deletes BB Returns if doesn't have a single predecessor Returns if BB's address was taken After the patch: Method 2. MergeBlockIntoPredecessor is attempting to become the new default: Updates DomTree or DeferredDominance, and LoopInfo and MemoryDependenceResults Moves all instruction from BB to Pred, deletes BB Returns if doesn't have a single predecessor Returns if BB's address was taken Uses of MergeBasicBlockIntoOnlyPred that need to be replaced: 1. lib/Transforms/Scalar/LoopSimplifyCFG.cpp Updated in this patch. No challenges. 2. lib/CodeGen/CodeGenPrepare.cpp Updated in this patch. i. eliminateFallThrough is straightforward, but I added using a temporary array to avoid the iterator invalidation. ii. eliminateMostlyEmptyBlock(s) methods also now use a temporary array for blocks Some interesting aspects: - Since Pred is not deleted (BB is), the entry block does not need updating. - The entry block was being updated with the deleted block in eliminateMostlyEmptyBlock. Added assert to make obvious that BB=SinglePred. - isMergingEmptyBlockProfitable assumes BB is the one to be deleted. - eliminateMostlyEmptyBlock(BB) does not delete BB on one path, it deletes its unique predecessor instead. - adding some test owner as subscribers for the interesting tests modified: test/CodeGen/X86/avx-cmp.ll test/CodeGen/AMDGPU/nested-loop-conditions.ll test/CodeGen/AMDGPU/si-annotate-cf.ll test/CodeGen/X86/hoist-spill.ll test/CodeGen/X86/2006-11-17-IllegalMove.ll 3. lib/Transforms/Scalar/JumpThreading.cpp Not covered in this patch. It is the only use case using the DeferredDominance. I would defer to Brian Rzycki to make this replacement. Reviewers: chandlerc, spatel, davide, brzycki, bkramer, javed.absar Subscribers: qcolombet, sanjoy, nemanjai, nhaehnle, jlebar, tpr, kbarton, RKSimon, wmi, arsenm, llvm-commits Differential Revision: https://reviews.llvm.org/D48202 llvm-svn: 335183
2018-06-21 06:01:04 +08:00
; CHECK-NEXT: movb $1, %bpl
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-NEXT: .p2align 4, 0x90
; CHECK-NEXT: .LBB2_2: # %for.cond5
; CHECK-NEXT: # =>This Inner Loop Header: Depth=1
; CHECK-NEXT: testb %bl, %bl
; CHECK-NEXT: jne .LBB2_2
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; CHECK-NEXT: # %bb.3: # %for.cond5
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-NEXT: # in Loop: Header=BB2_2 Depth=1
Generalize MergeBlockIntoPredecessor. Replace uses of MergeBasicBlockIntoOnlyPred. Summary: Two utils methods have essentially the same functionality. This is an attempt to merge them into one. 1. lib/Transforms/Utils/Local.cpp : MergeBasicBlockIntoOnlyPred 2. lib/Transforms/Utils/BasicBlockUtils.cpp : MergeBlockIntoPredecessor Prior to the patch: 1. MergeBasicBlockIntoOnlyPred Updates either DomTree or DeferredDominance Moves all instructions from Pred to BB, deletes Pred Asserts BB has single predecessor If address was taken, replace the block address with constant 1 (?) 2. MergeBlockIntoPredecessor Updates DomTree, LoopInfo and MemoryDependenceResults Moves all instruction from BB to Pred, deletes BB Returns if doesn't have a single predecessor Returns if BB's address was taken After the patch: Method 2. MergeBlockIntoPredecessor is attempting to become the new default: Updates DomTree or DeferredDominance, and LoopInfo and MemoryDependenceResults Moves all instruction from BB to Pred, deletes BB Returns if doesn't have a single predecessor Returns if BB's address was taken Uses of MergeBasicBlockIntoOnlyPred that need to be replaced: 1. lib/Transforms/Scalar/LoopSimplifyCFG.cpp Updated in this patch. No challenges. 2. lib/CodeGen/CodeGenPrepare.cpp Updated in this patch. i. eliminateFallThrough is straightforward, but I added using a temporary array to avoid the iterator invalidation. ii. eliminateMostlyEmptyBlock(s) methods also now use a temporary array for blocks Some interesting aspects: - Since Pred is not deleted (BB is), the entry block does not need updating. - The entry block was being updated with the deleted block in eliminateMostlyEmptyBlock. Added assert to make obvious that BB=SinglePred. - isMergingEmptyBlockProfitable assumes BB is the one to be deleted. - eliminateMostlyEmptyBlock(BB) does not delete BB on one path, it deletes its unique predecessor instead. - adding some test owner as subscribers for the interesting tests modified: test/CodeGen/X86/avx-cmp.ll test/CodeGen/AMDGPU/nested-loop-conditions.ll test/CodeGen/AMDGPU/si-annotate-cf.ll test/CodeGen/X86/hoist-spill.ll test/CodeGen/X86/2006-11-17-IllegalMove.ll 3. lib/Transforms/Scalar/JumpThreading.cpp Not covered in this patch. It is the only use case using the DeferredDominance. I would defer to Brian Rzycki to make this replacement. Reviewers: chandlerc, spatel, davide, brzycki, bkramer, javed.absar Subscribers: qcolombet, sanjoy, nemanjai, nhaehnle, jlebar, tpr, kbarton, RKSimon, wmi, arsenm, llvm-commits Differential Revision: https://reviews.llvm.org/D48202 llvm-svn: 335183
2018-06-21 06:01:04 +08:00
; CHECK-NEXT: testb %bpl, %bpl
; CHECK-NEXT: jne .LBB2_2
[X86] Fix some tests using fcmp with undef arguments Prep work for PR40800 (Add UNDEF handling to SelectionDAG::FoldSetCC) llvm-svn: 357278
2019-03-30 01:20:27 +08:00
; CHECK-NEXT: # %bb.4: # %for.body33.preheader
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-NEXT: # in Loop: Header=BB2_2 Depth=1
[X86] Fix some tests using fcmp with undef arguments Prep work for PR40800 (Add UNDEF handling to SelectionDAG::FoldSetCC) llvm-svn: 357278
2019-03-30 01:20:27 +08:00
; CHECK-NEXT: vmovsd (%rsp), %xmm0 # 8-byte Reload
; CHECK-NEXT: # xmm0 = mem[0],zero
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-NEXT: vucomisd {{\.LCPI.*}}, %xmm0
; CHECK-NEXT: jne .LBB2_5
[MBP] Don't move bottom block before header if it can't reduce taken branches If bottom of block BB has only one successor OldTop, in most cases it is profitable to move it before OldTop, except the following case: -->OldTop<- | . | | . | | . | ---Pred | | | BB----- Move BB before OldTop can't reduce the number of taken branches, this patch detects this case and prevent the moving. Differential Revision: https://reviews.llvm.org/D57067 llvm-svn: 352236
2019-01-26 03:45:13 +08:00
; CHECK-NEXT: jnp .LBB2_2
; CHECK-NEXT: .LBB2_5: # %if.then
[X86] Fix some tests using fcmp with undef arguments Prep work for PR40800 (Add UNDEF handling to SelectionDAG::FoldSetCC) llvm-svn: 357278
2019-03-30 01:20:27 +08:00
; CHECK-NEXT: # in Loop: Header=BB2_2 Depth=1
; CHECK-NEXT: callq scale
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-NEXT: jmp .LBB2_2
; CHECK-NEXT: .LBB2_6: # %for.end52
Generalize MergeBlockIntoPredecessor. Replace uses of MergeBasicBlockIntoOnlyPred. Summary: Two utils methods have essentially the same functionality. This is an attempt to merge them into one. 1. lib/Transforms/Utils/Local.cpp : MergeBasicBlockIntoOnlyPred 2. lib/Transforms/Utils/BasicBlockUtils.cpp : MergeBlockIntoPredecessor Prior to the patch: 1. MergeBasicBlockIntoOnlyPred Updates either DomTree or DeferredDominance Moves all instructions from Pred to BB, deletes Pred Asserts BB has single predecessor If address was taken, replace the block address with constant 1 (?) 2. MergeBlockIntoPredecessor Updates DomTree, LoopInfo and MemoryDependenceResults Moves all instruction from BB to Pred, deletes BB Returns if doesn't have a single predecessor Returns if BB's address was taken After the patch: Method 2. MergeBlockIntoPredecessor is attempting to become the new default: Updates DomTree or DeferredDominance, and LoopInfo and MemoryDependenceResults Moves all instruction from BB to Pred, deletes BB Returns if doesn't have a single predecessor Returns if BB's address was taken Uses of MergeBasicBlockIntoOnlyPred that need to be replaced: 1. lib/Transforms/Scalar/LoopSimplifyCFG.cpp Updated in this patch. No challenges. 2. lib/CodeGen/CodeGenPrepare.cpp Updated in this patch. i. eliminateFallThrough is straightforward, but I added using a temporary array to avoid the iterator invalidation. ii. eliminateMostlyEmptyBlock(s) methods also now use a temporary array for blocks Some interesting aspects: - Since Pred is not deleted (BB is), the entry block does not need updating. - The entry block was being updated with the deleted block in eliminateMostlyEmptyBlock. Added assert to make obvious that BB=SinglePred. - isMergingEmptyBlockProfitable assumes BB is the one to be deleted. - eliminateMostlyEmptyBlock(BB) does not delete BB on one path, it deletes its unique predecessor instead. - adding some test owner as subscribers for the interesting tests modified: test/CodeGen/X86/avx-cmp.ll test/CodeGen/AMDGPU/nested-loop-conditions.ll test/CodeGen/AMDGPU/si-annotate-cf.ll test/CodeGen/X86/hoist-spill.ll test/CodeGen/X86/2006-11-17-IllegalMove.ll 3. lib/Transforms/Scalar/JumpThreading.cpp Not covered in this patch. It is the only use case using the DeferredDominance. I would defer to Brian Rzycki to make this replacement. Reviewers: chandlerc, spatel, davide, brzycki, bkramer, javed.absar Subscribers: qcolombet, sanjoy, nemanjai, nhaehnle, jlebar, tpr, kbarton, RKSimon, wmi, arsenm, llvm-commits Differential Revision: https://reviews.llvm.org/D48202 llvm-svn: 335183
2018-06-21 06:01:04 +08:00
; CHECK-NEXT: addq $8, %rsp
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-NEXT: popq %rbx
Generalize MergeBlockIntoPredecessor. Replace uses of MergeBasicBlockIntoOnlyPred. Summary: Two utils methods have essentially the same functionality. This is an attempt to merge them into one. 1. lib/Transforms/Utils/Local.cpp : MergeBasicBlockIntoOnlyPred 2. lib/Transforms/Utils/BasicBlockUtils.cpp : MergeBlockIntoPredecessor Prior to the patch: 1. MergeBasicBlockIntoOnlyPred Updates either DomTree or DeferredDominance Moves all instructions from Pred to BB, deletes Pred Asserts BB has single predecessor If address was taken, replace the block address with constant 1 (?) 2. MergeBlockIntoPredecessor Updates DomTree, LoopInfo and MemoryDependenceResults Moves all instruction from BB to Pred, deletes BB Returns if doesn't have a single predecessor Returns if BB's address was taken After the patch: Method 2. MergeBlockIntoPredecessor is attempting to become the new default: Updates DomTree or DeferredDominance, and LoopInfo and MemoryDependenceResults Moves all instruction from BB to Pred, deletes BB Returns if doesn't have a single predecessor Returns if BB's address was taken Uses of MergeBasicBlockIntoOnlyPred that need to be replaced: 1. lib/Transforms/Scalar/LoopSimplifyCFG.cpp Updated in this patch. No challenges. 2. lib/CodeGen/CodeGenPrepare.cpp Updated in this patch. i. eliminateFallThrough is straightforward, but I added using a temporary array to avoid the iterator invalidation. ii. eliminateMostlyEmptyBlock(s) methods also now use a temporary array for blocks Some interesting aspects: - Since Pred is not deleted (BB is), the entry block does not need updating. - The entry block was being updated with the deleted block in eliminateMostlyEmptyBlock. Added assert to make obvious that BB=SinglePred. - isMergingEmptyBlockProfitable assumes BB is the one to be deleted. - eliminateMostlyEmptyBlock(BB) does not delete BB on one path, it deletes its unique predecessor instead. - adding some test owner as subscribers for the interesting tests modified: test/CodeGen/X86/avx-cmp.ll test/CodeGen/AMDGPU/nested-loop-conditions.ll test/CodeGen/AMDGPU/si-annotate-cf.ll test/CodeGen/X86/hoist-spill.ll test/CodeGen/X86/2006-11-17-IllegalMove.ll 3. lib/Transforms/Scalar/JumpThreading.cpp Not covered in this patch. It is the only use case using the DeferredDominance. I would defer to Brian Rzycki to make this replacement. Reviewers: chandlerc, spatel, davide, brzycki, bkramer, javed.absar Subscribers: qcolombet, sanjoy, nemanjai, nhaehnle, jlebar, tpr, kbarton, RKSimon, wmi, arsenm, llvm-commits Differential Revision: https://reviews.llvm.org/D48202 llvm-svn: 335183
2018-06-21 06:01:04 +08:00
; CHECK-NEXT: popq %rbp
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-NEXT: retq
Fix PR10422 by adding the necessary AVX UCOMISD memory versions to load folding logic llvm-svn: 135801
2011-07-23 04:53:20 +08:00
entry:
br i1 undef, label %for.cond5, label %for.end52
for.cond5:
%or.cond = and i1 undef, false
br i1 %or.cond, label %for.body33, label %for.cond5
for.cond30:
br i1 false, label %for.body33, label %for.cond5
for.body33:
[X86] Fix some tests using fcmp with undef arguments Prep work for PR40800 (Add UNDEF handling to SelectionDAG::FoldSetCC) llvm-svn: 357278
2019-03-30 01:20:27 +08:00
%tobool = fcmp une double %a0, 0.000000e+00
Fix PR10422 by adding the necessary AVX UCOMISD memory versions to load folding logic llvm-svn: 135801
2011-07-23 04:53:20 +08:00
br i1 %tobool, label %if.then, label %for.cond30
if.then:
call void @scale()
br label %for.cond30
for.end52:
ret void
}
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
define <8 x i32> @int256_cmp(<8 x i32> %i, <8 x i32> %j) nounwind {
; CHECK-LABEL: int256_cmp:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; CHECK: # %bb.0:
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-NEXT: vextractf128 $1, %ymm0, %xmm2
; CHECK-NEXT: vextractf128 $1, %ymm1, %xmm3
; CHECK-NEXT: vpcmpgtd %xmm2, %xmm3, %xmm2
; CHECK-NEXT: vpcmpgtd %xmm0, %xmm1, %xmm0
; CHECK-NEXT: vinsertf128 $1, %xmm2, %ymm0, %ymm0
; CHECK-NEXT: retq
Add support for breaking 256-bit int VETCC into two 128-bit ones, avoding scalarization of the compare. Reduces code from 59 to 6 instructions. Fix PR10712. llvm-svn: 138271
2011-08-23 04:31:04 +08:00
%bincmp = icmp slt <8 x i32> %i, %j
%x = sext <8 x i1> %bincmp to <8 x i32>
ret <8 x i32> %x
}
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
define <4 x i64> @v4i64_cmp(<4 x i64> %i, <4 x i64> %j) nounwind {
; CHECK-LABEL: v4i64_cmp:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; CHECK: # %bb.0:
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-NEXT: vextractf128 $1, %ymm0, %xmm2
; CHECK-NEXT: vextractf128 $1, %ymm1, %xmm3
; CHECK-NEXT: vpcmpgtq %xmm2, %xmm3, %xmm2
; CHECK-NEXT: vpcmpgtq %xmm0, %xmm1, %xmm0
; CHECK-NEXT: vinsertf128 $1, %xmm2, %ymm0, %ymm0
; CHECK-NEXT: retq
Add support for breaking 256-bit v16i16 and v32i8 VSETCC into two 128-bit ones, avoiding sclarization. Add vex form of pcmpeqq and pcmpgtq. Fixes more cases for PR10712. llvm-svn: 138321
2011-08-23 12:36:33 +08:00
%bincmp = icmp slt <4 x i64> %i, %j
%x = sext <4 x i1> %bincmp to <4 x i64>
ret <4 x i64> %x
}
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
define <16 x i16> @v16i16_cmp(<16 x i16> %i, <16 x i16> %j) nounwind {
; CHECK-LABEL: v16i16_cmp:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; CHECK: # %bb.0:
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-NEXT: vextractf128 $1, %ymm0, %xmm2
; CHECK-NEXT: vextractf128 $1, %ymm1, %xmm3
; CHECK-NEXT: vpcmpgtw %xmm2, %xmm3, %xmm2
; CHECK-NEXT: vpcmpgtw %xmm0, %xmm1, %xmm0
; CHECK-NEXT: vinsertf128 $1, %xmm2, %ymm0, %ymm0
; CHECK-NEXT: retq
Add support for breaking 256-bit v16i16 and v32i8 VSETCC into two 128-bit ones, avoiding sclarization. Add vex form of pcmpeqq and pcmpgtq. Fixes more cases for PR10712. llvm-svn: 138321
2011-08-23 12:36:33 +08:00
%bincmp = icmp slt <16 x i16> %i, %j
%x = sext <16 x i1> %bincmp to <16 x i16>
ret <16 x i16> %x
}
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
define <32 x i8> @v32i8_cmp(<32 x i8> %i, <32 x i8> %j) nounwind {
; CHECK-LABEL: v32i8_cmp:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; CHECK: # %bb.0:
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-NEXT: vextractf128 $1, %ymm0, %xmm2
; CHECK-NEXT: vextractf128 $1, %ymm1, %xmm3
; CHECK-NEXT: vpcmpgtb %xmm2, %xmm3, %xmm2
; CHECK-NEXT: vpcmpgtb %xmm0, %xmm1, %xmm0
; CHECK-NEXT: vinsertf128 $1, %xmm2, %ymm0, %ymm0
; CHECK-NEXT: retq
Add support for breaking 256-bit v16i16 and v32i8 VSETCC into two 128-bit ones, avoiding sclarization. Add vex form of pcmpeqq and pcmpgtq. Fixes more cases for PR10712. llvm-svn: 138321
2011-08-23 12:36:33 +08:00
%bincmp = icmp slt <32 x i8> %i, %j
%x = sext <32 x i1> %bincmp to <32 x i8>
ret <32 x i8> %x
}
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
define <8 x i32> @int256_cmpeq(<8 x i32> %i, <8 x i32> %j) nounwind {
; CHECK-LABEL: int256_cmpeq:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; CHECK: # %bb.0:
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-NEXT: vextractf128 $1, %ymm1, %xmm2
; CHECK-NEXT: vextractf128 $1, %ymm0, %xmm3
; CHECK-NEXT: vpcmpeqd %xmm2, %xmm3, %xmm2
; CHECK-NEXT: vpcmpeqd %xmm1, %xmm0, %xmm0
; CHECK-NEXT: vinsertf128 $1, %xmm2, %ymm0, %ymm0
; CHECK-NEXT: retq
Add support for breaking 256-bit v16i16 and v32i8 VSETCC into two 128-bit ones, avoiding sclarization. Add vex form of pcmpeqq and pcmpgtq. Fixes more cases for PR10712. llvm-svn: 138321
2011-08-23 12:36:33 +08:00
%bincmp = icmp eq <8 x i32> %i, %j
%x = sext <8 x i1> %bincmp to <8 x i32>
ret <8 x i32> %x
}
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
define <4 x i64> @v4i64_cmpeq(<4 x i64> %i, <4 x i64> %j) nounwind {
; CHECK-LABEL: v4i64_cmpeq:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; CHECK: # %bb.0:
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-NEXT: vextractf128 $1, %ymm1, %xmm2
; CHECK-NEXT: vextractf128 $1, %ymm0, %xmm3
; CHECK-NEXT: vpcmpeqq %xmm2, %xmm3, %xmm2
; CHECK-NEXT: vpcmpeqq %xmm1, %xmm0, %xmm0
; CHECK-NEXT: vinsertf128 $1, %xmm2, %ymm0, %ymm0
; CHECK-NEXT: retq
Add support for breaking 256-bit v16i16 and v32i8 VSETCC into two 128-bit ones, avoiding sclarization. Add vex form of pcmpeqq and pcmpgtq. Fixes more cases for PR10712. llvm-svn: 138321
2011-08-23 12:36:33 +08:00
%bincmp = icmp eq <4 x i64> %i, %j
%x = sext <4 x i1> %bincmp to <4 x i64>
ret <4 x i64> %x
}
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
define <16 x i16> @v16i16_cmpeq(<16 x i16> %i, <16 x i16> %j) nounwind {
; CHECK-LABEL: v16i16_cmpeq:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; CHECK: # %bb.0:
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-NEXT: vextractf128 $1, %ymm1, %xmm2
; CHECK-NEXT: vextractf128 $1, %ymm0, %xmm3
; CHECK-NEXT: vpcmpeqw %xmm2, %xmm3, %xmm2
; CHECK-NEXT: vpcmpeqw %xmm1, %xmm0, %xmm0
; CHECK-NEXT: vinsertf128 $1, %xmm2, %ymm0, %ymm0
; CHECK-NEXT: retq
Add support for breaking 256-bit v16i16 and v32i8 VSETCC into two 128-bit ones, avoiding sclarization. Add vex form of pcmpeqq and pcmpgtq. Fixes more cases for PR10712. llvm-svn: 138321
2011-08-23 12:36:33 +08:00
%bincmp = icmp eq <16 x i16> %i, %j
%x = sext <16 x i1> %bincmp to <16 x i16>
ret <16 x i16> %x
}
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
define <32 x i8> @v32i8_cmpeq(<32 x i8> %i, <32 x i8> %j) nounwind {
; CHECK-LABEL: v32i8_cmpeq:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; CHECK: # %bb.0:
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-NEXT: vextractf128 $1, %ymm1, %xmm2
; CHECK-NEXT: vextractf128 $1, %ymm0, %xmm3
; CHECK-NEXT: vpcmpeqb %xmm2, %xmm3, %xmm2
; CHECK-NEXT: vpcmpeqb %xmm1, %xmm0, %xmm0
; CHECK-NEXT: vinsertf128 $1, %xmm2, %ymm0, %ymm0
; CHECK-NEXT: retq
Add support for breaking 256-bit v16i16 and v32i8 VSETCC into two 128-bit ones, avoiding sclarization. Add vex form of pcmpeqq and pcmpgtq. Fixes more cases for PR10712. llvm-svn: 138321
2011-08-23 12:36:33 +08:00
%bincmp = icmp eq <32 x i8> %i, %j
%x = sext <32 x i1> %bincmp to <32 x i8>
ret <32 x i8> %x
}
Match X86ISD::FSETCCsd and X86ISD::FSETCCss while in AVX mode. This fix PR10955 and PR10948. llvm-svn: 140069
2011-09-20 05:29:24 +08:00
;; Scalar comparison
[X86] Fix some tests using fcmp with undef arguments Prep work for PR40800 (Add UNDEF handling to SelectionDAG::FoldSetCC) llvm-svn: 357278
2019-03-30 01:20:27 +08:00
define i32 @scalarcmpA(double %a0) uwtable ssp {
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-LABEL: scalarcmpA:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; CHECK: # %bb.0:
[X86] Fix some tests using fcmp with undef arguments Prep work for PR40800 (Add UNDEF handling to SelectionDAG::FoldSetCC) llvm-svn: 357278
2019-03-30 01:20:27 +08:00
; CHECK-NEXT: vxorpd %xmm1, %xmm1, %xmm1
; CHECK-NEXT: vcmpeqsd %xmm1, %xmm0, %xmm0
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-NEXT: vmovq %xmm0, %rax
; CHECK-NEXT: andl $1, %eax
Followup on Proposal to move MIR physical register namespace to '$' sigil. Discussed here: http://lists.llvm.org/pipermail/llvm-dev/2018-January/120320.html In preparation for adding support for named vregs we are changing the sigil for physical registers in MIR to '$' from '%'. This will prevent name clashes of named physical register with named vregs. llvm-svn: 323922
2018-02-01 06:04:26 +08:00
; CHECK-NEXT: # kill: def $eax killed $eax killed $rax
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-NEXT: retq
[X86] Fix some tests using fcmp with undef arguments Prep work for PR40800 (Add UNDEF handling to SelectionDAG::FoldSetCC) llvm-svn: 357278
2019-03-30 01:20:27 +08:00
%cmp29 = fcmp oeq double %a0, 0.000000e+00
Match X86ISD::FSETCCsd and X86ISD::FSETCCss while in AVX mode. This fix PR10955 and PR10948. llvm-svn: 140069
2011-09-20 05:29:24 +08:00
%res = zext i1 %cmp29 to i32
ret i32 %res
}
[X86] Fix some tests using fcmp with undef arguments Prep work for PR40800 (Add UNDEF handling to SelectionDAG::FoldSetCC) llvm-svn: 357278
2019-03-30 01:20:27 +08:00
define i32 @scalarcmpB(float %a0) uwtable ssp {
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-LABEL: scalarcmpB:
[CodeGen] Unify MBB reference format in both MIR and debug output As part of the unification of the debug format and the MIR format, print MBB references as '%bb.5'. The MIR printer prints the IR name of a MBB only for block definitions. * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)->getNumber\(\)/" << printMBBReference(*\1)/g' * find . \( -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#" << ([a-zA-Z0-9_]+)\.getNumber\(\)/" << printMBBReference(\1)/g' * find . \( -name "*.txt" -o -name "*.s" -o -name "*.mir" -o -name "*.cpp" -o -name "*.h" -o -name "*.ll" \) -type f -print0 | xargs -0 sed -i '' -E 's/BB#([0-9]+)/%bb.\1/g' * grep -nr 'BB#' and fix Differential Revision: https://reviews.llvm.org/D40422 llvm-svn: 319665
2017-12-05 01:18:51 +08:00
; CHECK: # %bb.0:
[X86] Fix some tests using fcmp with undef arguments Prep work for PR40800 (Add UNDEF handling to SelectionDAG::FoldSetCC) llvm-svn: 357278
2019-03-30 01:20:27 +08:00
; CHECK-NEXT: vxorps %xmm1, %xmm1, %xmm1
; CHECK-NEXT: vcmpeqss %xmm1, %xmm0, %xmm0
[x86] auto-generate complete checks for tests; NFC These all used 'CHECK-NOT' which isn't necessary if we have complete checks. There were also several over-specifications in the RUN params such as CPU model or OS requirement llvm-svn: 307028
2017-07-03 23:04:05 +08:00
; CHECK-NEXT: vmovd %xmm0, %eax
; CHECK-NEXT: andl $1, %eax
; CHECK-NEXT: retq
[X86] Fix some tests using fcmp with undef arguments Prep work for PR40800 (Add UNDEF handling to SelectionDAG::FoldSetCC) llvm-svn: 357278
2019-03-30 01:20:27 +08:00
%cmp29 = fcmp oeq float %a0, 0.000000e+00
Match X86ISD::FSETCCsd and X86ISD::FSETCCss while in AVX mode. This fix PR10955 and PR10948. llvm-svn: 140069
2011-09-20 05:29:24 +08:00
%res = zext i1 %cmp29 to i32
ret i32 %res
}