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/fma4-fneg-combine.ll

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

112 lines
5.0 KiB
LLVM
Raw Normal View History

[X86] Add separate intrinsics for scalar FMA4 instructions. Summary: These instructions zero the non-scalar part of the lower 128-bits which makes them different than the FMA3 instructions which pass through the non-scalar part of the lower 128-bits. I've only added fmadd because we should be able to derive all other variants using operand negation in the intrinsic header like we do for AVX512. I think there are still some missed negate folding opportunities with the FMA4 instructions in light of this behavior difference that I hadn't noticed before. I've split the tests so that we can use different intrinsics for scalar testing between the two. I just copied the tests split the RUN lines and changed out the scalar intrinsics. fma4-fneg-combine.ll is a new test to make sure we negate the fma4 intrinsics correctly though there are a couple TODOs in it. Reviewers: RKSimon, spatel Reviewed By: RKSimon Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D39851 llvm-svn: 318984
2017-11-26 02:32:43 +08:00
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc < %s -mtriple=x86_64-unknown-linux-gnu -mattr=+fma4,-fma | FileCheck %s
; RUN: llc < %s -mtriple=x86_64-unknown-linux-gnu -mattr=+fma4,+fma | FileCheck %s
declare <4 x float> @llvm.x86.fma4.vfmadd.ss(<4 x float> %a, <4 x float> %b, <4 x float> %c)
declare <2 x double> @llvm.x86.fma4.vfmadd.sd(<2 x double> %a, <2 x double> %b, <2 x double> %c)
; TODO this can be negated
define <4 x float> @test1(<4 x float> %a, <4 x float> %b, <4 x float> %c) {
; CHECK-LABEL: test1:
[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] Add separate intrinsics for scalar FMA4 instructions. Summary: These instructions zero the non-scalar part of the lower 128-bits which makes them different than the FMA3 instructions which pass through the non-scalar part of the lower 128-bits. I've only added fmadd because we should be able to derive all other variants using operand negation in the intrinsic header like we do for AVX512. I think there are still some missed negate folding opportunities with the FMA4 instructions in light of this behavior difference that I hadn't noticed before. I've split the tests so that we can use different intrinsics for scalar testing between the two. I just copied the tests split the RUN lines and changed out the scalar intrinsics. fma4-fneg-combine.ll is a new test to make sure we negate the fma4 intrinsics correctly though there are a couple TODOs in it. Reviewers: RKSimon, spatel Reviewed By: RKSimon Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D39851 llvm-svn: 318984
2017-11-26 02:32:43 +08:00
; CHECK-NEXT: vfmaddss %xmm2, %xmm1, %xmm0, %xmm0
; CHECK-NEXT: vxorps {{.*}}(%rip), %xmm0, %xmm0
; CHECK-NEXT: retq
%res = tail call <4 x float> @llvm.x86.fma4.vfmadd.ss(<4 x float> %a, <4 x float> %b, <4 x float> %c)
%sub.i = fsub <4 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %res
ret <4 x float> %sub.i
}
define <4 x float> @test2(<4 x float> %a, <4 x float> %b, <4 x float> %c) {
; CHECK-LABEL: test2:
[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] Enhance combineFMA to look for FNEG behind an EXTRACT_VECTOR_ELT. llvm-svn: 336514
2018-07-09 02:04:00 +08:00
; CHECK-NEXT: vfmsubss %xmm2, %xmm1, %xmm0, %xmm0
[X86] Add separate intrinsics for scalar FMA4 instructions. Summary: These instructions zero the non-scalar part of the lower 128-bits which makes them different than the FMA3 instructions which pass through the non-scalar part of the lower 128-bits. I've only added fmadd because we should be able to derive all other variants using operand negation in the intrinsic header like we do for AVX512. I think there are still some missed negate folding opportunities with the FMA4 instructions in light of this behavior difference that I hadn't noticed before. I've split the tests so that we can use different intrinsics for scalar testing between the two. I just copied the tests split the RUN lines and changed out the scalar intrinsics. fma4-fneg-combine.ll is a new test to make sure we negate the fma4 intrinsics correctly though there are a couple TODOs in it. Reviewers: RKSimon, spatel Reviewed By: RKSimon Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D39851 llvm-svn: 318984
2017-11-26 02:32:43 +08:00
; CHECK-NEXT: retq
%sub.i = fsub <4 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %c
%res = tail call <4 x float> @llvm.x86.fma4.vfmadd.ss(<4 x float> %a, <4 x float> %b, <4 x float> %sub.i)
ret <4 x float> %res
}
define <4 x float> @test3(<4 x float> %a, <4 x float> %b, <4 x float> %c) {
; CHECK-LABEL: test3:
[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] Enhance combineFMA to look for FNEG behind an EXTRACT_VECTOR_ELT. llvm-svn: 336514
2018-07-09 02:04:00 +08:00
; CHECK-NEXT: vfnmaddss %xmm2, %xmm1, %xmm0, %xmm0
[X86] Add separate intrinsics for scalar FMA4 instructions. Summary: These instructions zero the non-scalar part of the lower 128-bits which makes them different than the FMA3 instructions which pass through the non-scalar part of the lower 128-bits. I've only added fmadd because we should be able to derive all other variants using operand negation in the intrinsic header like we do for AVX512. I think there are still some missed negate folding opportunities with the FMA4 instructions in light of this behavior difference that I hadn't noticed before. I've split the tests so that we can use different intrinsics for scalar testing between the two. I just copied the tests split the RUN lines and changed out the scalar intrinsics. fma4-fneg-combine.ll is a new test to make sure we negate the fma4 intrinsics correctly though there are a couple TODOs in it. Reviewers: RKSimon, spatel Reviewed By: RKSimon Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D39851 llvm-svn: 318984
2017-11-26 02:32:43 +08:00
; CHECK-NEXT: retq
%sub.i = fsub <4 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %b
%res = tail call <4 x float> @llvm.x86.fma4.vfmadd.ss(<4 x float> %a, <4 x float> %sub.i, <4 x float> %c)
ret <4 x float> %res
}
define <4 x float> @test4(<4 x float> %a, <4 x float> %b, <4 x float> %c) {
; CHECK-LABEL: test4:
[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] Enhance combineFMA to look for FNEG behind an EXTRACT_VECTOR_ELT. llvm-svn: 336514
2018-07-09 02:04:00 +08:00
; CHECK-NEXT: vfnmaddss %xmm2, %xmm1, %xmm0, %xmm0
[X86] Add separate intrinsics for scalar FMA4 instructions. Summary: These instructions zero the non-scalar part of the lower 128-bits which makes them different than the FMA3 instructions which pass through the non-scalar part of the lower 128-bits. I've only added fmadd because we should be able to derive all other variants using operand negation in the intrinsic header like we do for AVX512. I think there are still some missed negate folding opportunities with the FMA4 instructions in light of this behavior difference that I hadn't noticed before. I've split the tests so that we can use different intrinsics for scalar testing between the two. I just copied the tests split the RUN lines and changed out the scalar intrinsics. fma4-fneg-combine.ll is a new test to make sure we negate the fma4 intrinsics correctly though there are a couple TODOs in it. Reviewers: RKSimon, spatel Reviewed By: RKSimon Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D39851 llvm-svn: 318984
2017-11-26 02:32:43 +08:00
; CHECK-NEXT: retq
%sub.i = fsub <4 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %a
%res = tail call <4 x float> @llvm.x86.fma4.vfmadd.ss(<4 x float> %sub.i, <4 x float> %b, <4 x float> %c)
ret <4 x float> %res
}
define <4 x float> @test5(<4 x float> %a, <4 x float> %b, <4 x float> %c) {
; CHECK-LABEL: test5:
[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] Enhance combineFMA to look for FNEG behind an EXTRACT_VECTOR_ELT. llvm-svn: 336514
2018-07-09 02:04:00 +08:00
; CHECK-NEXT: vfnmsubss %xmm2, %xmm1, %xmm0, %xmm0
[X86] Add separate intrinsics for scalar FMA4 instructions. Summary: These instructions zero the non-scalar part of the lower 128-bits which makes them different than the FMA3 instructions which pass through the non-scalar part of the lower 128-bits. I've only added fmadd because we should be able to derive all other variants using operand negation in the intrinsic header like we do for AVX512. I think there are still some missed negate folding opportunities with the FMA4 instructions in light of this behavior difference that I hadn't noticed before. I've split the tests so that we can use different intrinsics for scalar testing between the two. I just copied the tests split the RUN lines and changed out the scalar intrinsics. fma4-fneg-combine.ll is a new test to make sure we negate the fma4 intrinsics correctly though there are a couple TODOs in it. Reviewers: RKSimon, spatel Reviewed By: RKSimon Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D39851 llvm-svn: 318984
2017-11-26 02:32:43 +08:00
; CHECK-NEXT: retq
%sub.i = fsub <4 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %a
%sub.i.2 = fsub <4 x float> <float -0.000000e+00, float -0.000000e+00, float -0.000000e+00, float -0.000000e+00>, %c
%res = tail call <4 x float> @llvm.x86.fma4.vfmadd.ss(<4 x float> %sub.i, <4 x float> %b, <4 x float> %sub.i.2)
ret <4 x float> %res
}
define <2 x double> @test6(<2 x double> %a, <2 x double> %b, <2 x double> %c) {
; CHECK-LABEL: test6:
[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] Add separate intrinsics for scalar FMA4 instructions. Summary: These instructions zero the non-scalar part of the lower 128-bits which makes them different than the FMA3 instructions which pass through the non-scalar part of the lower 128-bits. I've only added fmadd because we should be able to derive all other variants using operand negation in the intrinsic header like we do for AVX512. I think there are still some missed negate folding opportunities with the FMA4 instructions in light of this behavior difference that I hadn't noticed before. I've split the tests so that we can use different intrinsics for scalar testing between the two. I just copied the tests split the RUN lines and changed out the scalar intrinsics. fma4-fneg-combine.ll is a new test to make sure we negate the fma4 intrinsics correctly though there are a couple TODOs in it. Reviewers: RKSimon, spatel Reviewed By: RKSimon Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D39851 llvm-svn: 318984
2017-11-26 02:32:43 +08:00
; CHECK-NEXT: vfmaddsd %xmm2, %xmm1, %xmm0, %xmm0
; CHECK-NEXT: vxorpd {{.*}}(%rip), %xmm0, %xmm0
; CHECK-NEXT: retq
%res = tail call <2 x double> @llvm.x86.fma4.vfmadd.sd(<2 x double> %a, <2 x double> %b, <2 x double> %c)
%sub.i = fsub <2 x double> <double -0.000000e+00, double -0.000000e+00>, %res
ret <2 x double> %sub.i
}
define <2 x double> @test7(<2 x double> %a, <2 x double> %b, <2 x double> %c) {
; CHECK-LABEL: test7:
[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] Enhance combineFMA to look for FNEG behind an EXTRACT_VECTOR_ELT. llvm-svn: 336514
2018-07-09 02:04:00 +08:00
; CHECK-NEXT: vfmsubsd %xmm2, %xmm1, %xmm0, %xmm0
[X86] Add separate intrinsics for scalar FMA4 instructions. Summary: These instructions zero the non-scalar part of the lower 128-bits which makes them different than the FMA3 instructions which pass through the non-scalar part of the lower 128-bits. I've only added fmadd because we should be able to derive all other variants using operand negation in the intrinsic header like we do for AVX512. I think there are still some missed negate folding opportunities with the FMA4 instructions in light of this behavior difference that I hadn't noticed before. I've split the tests so that we can use different intrinsics for scalar testing between the two. I just copied the tests split the RUN lines and changed out the scalar intrinsics. fma4-fneg-combine.ll is a new test to make sure we negate the fma4 intrinsics correctly though there are a couple TODOs in it. Reviewers: RKSimon, spatel Reviewed By: RKSimon Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D39851 llvm-svn: 318984
2017-11-26 02:32:43 +08:00
; CHECK-NEXT: retq
%sub.i = fsub <2 x double> <double -0.000000e+00, double -0.000000e+00>, %c
%res = tail call <2 x double> @llvm.x86.fma4.vfmadd.sd(<2 x double> %a, <2 x double> %b, <2 x double> %sub.i)
ret <2 x double> %res
}
define <2 x double> @test8(<2 x double> %a, <2 x double> %b, <2 x double> %c) {
; CHECK-LABEL: test8:
[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] Enhance combineFMA to look for FNEG behind an EXTRACT_VECTOR_ELT. llvm-svn: 336514
2018-07-09 02:04:00 +08:00
; CHECK-NEXT: vfnmaddsd %xmm2, %xmm1, %xmm0, %xmm0
[X86] Add separate intrinsics for scalar FMA4 instructions. Summary: These instructions zero the non-scalar part of the lower 128-bits which makes them different than the FMA3 instructions which pass through the non-scalar part of the lower 128-bits. I've only added fmadd because we should be able to derive all other variants using operand negation in the intrinsic header like we do for AVX512. I think there are still some missed negate folding opportunities with the FMA4 instructions in light of this behavior difference that I hadn't noticed before. I've split the tests so that we can use different intrinsics for scalar testing between the two. I just copied the tests split the RUN lines and changed out the scalar intrinsics. fma4-fneg-combine.ll is a new test to make sure we negate the fma4 intrinsics correctly though there are a couple TODOs in it. Reviewers: RKSimon, spatel Reviewed By: RKSimon Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D39851 llvm-svn: 318984
2017-11-26 02:32:43 +08:00
; CHECK-NEXT: retq
%sub.i = fsub <2 x double> <double -0.000000e+00, double -0.000000e+00>, %b
%res = tail call <2 x double> @llvm.x86.fma4.vfmadd.sd(<2 x double> %a, <2 x double> %sub.i, <2 x double> %c)
ret <2 x double> %res
}
define <2 x double> @test9(<2 x double> %a, <2 x double> %b, <2 x double> %c) {
; CHECK-LABEL: test9:
[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] Enhance combineFMA to look for FNEG behind an EXTRACT_VECTOR_ELT. llvm-svn: 336514
2018-07-09 02:04:00 +08:00
; CHECK-NEXT: vfnmaddsd %xmm2, %xmm1, %xmm0, %xmm0
[X86] Add separate intrinsics for scalar FMA4 instructions. Summary: These instructions zero the non-scalar part of the lower 128-bits which makes them different than the FMA3 instructions which pass through the non-scalar part of the lower 128-bits. I've only added fmadd because we should be able to derive all other variants using operand negation in the intrinsic header like we do for AVX512. I think there are still some missed negate folding opportunities with the FMA4 instructions in light of this behavior difference that I hadn't noticed before. I've split the tests so that we can use different intrinsics for scalar testing between the two. I just copied the tests split the RUN lines and changed out the scalar intrinsics. fma4-fneg-combine.ll is a new test to make sure we negate the fma4 intrinsics correctly though there are a couple TODOs in it. Reviewers: RKSimon, spatel Reviewed By: RKSimon Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D39851 llvm-svn: 318984
2017-11-26 02:32:43 +08:00
; CHECK-NEXT: retq
%sub.i = fsub <2 x double> <double -0.000000e+00, double -0.000000e+00>, %a
%res = tail call <2 x double> @llvm.x86.fma4.vfmadd.sd(<2 x double> %sub.i, <2 x double> %b, <2 x double> %c)
ret <2 x double> %res
}
define <2 x double> @test10(<2 x double> %a, <2 x double> %b, <2 x double> %c) {
; CHECK-LABEL: test10:
[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] Enhance combineFMA to look for FNEG behind an EXTRACT_VECTOR_ELT. llvm-svn: 336514
2018-07-09 02:04:00 +08:00
; CHECK-NEXT: vfnmsubsd %xmm2, %xmm1, %xmm0, %xmm0
[X86] Add separate intrinsics for scalar FMA4 instructions. Summary: These instructions zero the non-scalar part of the lower 128-bits which makes them different than the FMA3 instructions which pass through the non-scalar part of the lower 128-bits. I've only added fmadd because we should be able to derive all other variants using operand negation in the intrinsic header like we do for AVX512. I think there are still some missed negate folding opportunities with the FMA4 instructions in light of this behavior difference that I hadn't noticed before. I've split the tests so that we can use different intrinsics for scalar testing between the two. I just copied the tests split the RUN lines and changed out the scalar intrinsics. fma4-fneg-combine.ll is a new test to make sure we negate the fma4 intrinsics correctly though there are a couple TODOs in it. Reviewers: RKSimon, spatel Reviewed By: RKSimon Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D39851 llvm-svn: 318984
2017-11-26 02:32:43 +08:00
; CHECK-NEXT: retq
%sub.i = fsub <2 x double> <double -0.000000e+00, double -0.000000e+00>, %a
%sub.i.2 = fsub <2 x double> <double -0.000000e+00, double -0.000000e+00>, %c
%res = tail call <2 x double> @llvm.x86.fma4.vfmadd.sd(<2 x double> %sub.i, <2 x double> %b, <2 x double> %sub.i.2)
ret <2 x double> %res
}