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/vec-copysign.ll

160 lines
5.1 KiB
LLVM
Raw Normal View History

[X86][SSE] Canonicalize OR(AND(X,C),AND(Y,~C)) -> OR(AND(X,C),ANDNP(C,Y)) For constant bit select patterns, replace one AND with a ANDNP, allowing us to reuse the constant mask. Only do this if the mask has multiple uses (to avoid losing load folding) or if we have XOP as its VPCMOV can handle most folding commutations. This also requires computeKnownBitsForTargetNode support for X86ISD::ANDNP and X86ISD::FOR to prevent regressions in fabs/fcopysign patterns. Differential Revision: https://reviews.llvm.org/D55935 llvm-svn: 351819
2019-01-22 21:44:49 +08:00
; RUN: llc < %s -mtriple=x86_64-apple-macosx10.10.0 -mattr=+sse2 | FileCheck %s --check-prefixes=CHECK,SSE2
; RUN: llc < %s -mtriple=x86_64-apple-macosx10.10.0 -mattr=+avx | FileCheck %s --check-prefixes=CHECK,AVX
[x86] add test to show unnecessary scalarization of copysign intrinsics (PR30433) llvm-svn: 283071
2016-10-03 00:31:35 +08:00
[X86] Removed reference to update_test_checks.py llvm-svn: 306959
2017-07-02 00:34:29 +08:00
; Assertions have been enhanced from utils/update_llc_test_checks.py to show the constant pool values.
[x86, SSE/AVX] allow 128/256-bit lowering for copysign vector intrinsics (PR30433) This should fix: https://llvm.org/bugs/show_bug.cgi?id=30433 There are a couple of open questions about the codegen: 1. Should we let scalar ops be scalars and avoid vector constant loads/splats? 2. Should we have a pass to combine constants such as the inverted pair that we have here? Differential Revision: https://reviews.llvm.org/D25165 llvm-svn: 283119
2016-10-04 00:38:27 +08:00
; Use a macosx triple to make sure the format of those constant strings is exact.
; CHECK: [[SIGNMASK1:L.+]]:
; CHECK-NEXT: .long 2147483648
; CHECK-NEXT: .long 2147483648
; CHECK-NEXT: .long 2147483648
; CHECK-NEXT: .long 2147483648
; CHECK: [[MAGMASK1:L.+]]:
; CHECK-NEXT: .long 2147483647
; CHECK-NEXT: .long 2147483647
; CHECK-NEXT: .long 2147483647
; CHECK-NEXT: .long 2147483647
[x86] add test to show unnecessary scalarization of copysign intrinsics (PR30433) llvm-svn: 283071
2016-10-03 00:31:35 +08:00
define <4 x float> @v4f32(<4 x float> %a, <4 x float> %b) nounwind {
; SSE2-LABEL: v4f32:
[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
; SSE2: # %bb.0:
[x86, SSE/AVX] allow 128/256-bit lowering for copysign vector intrinsics (PR30433) This should fix: https://llvm.org/bugs/show_bug.cgi?id=30433 There are a couple of open questions about the codegen: 1. Should we let scalar ops be scalars and avoid vector constant loads/splats? 2. Should we have a pass to combine constants such as the inverted pair that we have here? Differential Revision: https://reviews.llvm.org/D25165 llvm-svn: 283119
2016-10-04 00:38:27 +08:00
; SSE2-NEXT: andps [[SIGNMASK1]](%rip), %xmm1
; SSE2-NEXT: andps [[MAGMASK1]](%rip), %xmm0
[x86] add test to show unnecessary scalarization of copysign intrinsics (PR30433) llvm-svn: 283071
2016-10-03 00:31:35 +08:00
; SSE2-NEXT: orps %xmm1, %xmm0
; SSE2-NEXT: retq
;
; AVX-LABEL: v4f32:
[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
; AVX: # %bb.0:
[x86, SSE/AVX] allow 128/256-bit lowering for copysign vector intrinsics (PR30433) This should fix: https://llvm.org/bugs/show_bug.cgi?id=30433 There are a couple of open questions about the codegen: 1. Should we let scalar ops be scalars and avoid vector constant loads/splats? 2. Should we have a pass to combine constants such as the inverted pair that we have here? Differential Revision: https://reviews.llvm.org/D25165 llvm-svn: 283119
2016-10-04 00:38:27 +08:00
; AVX-NEXT: vandps [[SIGNMASK1]](%rip), %xmm1, %xmm1
; AVX-NEXT: vandps [[MAGMASK1]](%rip), %xmm0, %xmm0
[x86] add test to show unnecessary scalarization of copysign intrinsics (PR30433) llvm-svn: 283071
2016-10-03 00:31:35 +08:00
; AVX-NEXT: vorps %xmm1, %xmm0, %xmm0
; AVX-NEXT: retq
;
%tmp = tail call <4 x float> @llvm.copysign.v4f32( <4 x float> %a, <4 x float> %b )
ret <4 x float> %tmp
}
[x86, SSE/AVX] allow 128/256-bit lowering for copysign vector intrinsics (PR30433) This should fix: https://llvm.org/bugs/show_bug.cgi?id=30433 There are a couple of open questions about the codegen: 1. Should we let scalar ops be scalars and avoid vector constant loads/splats? 2. Should we have a pass to combine constants such as the inverted pair that we have here? Differential Revision: https://reviews.llvm.org/D25165 llvm-svn: 283119
2016-10-04 00:38:27 +08:00
; SSE2: [[MAGMASK2:L.+]]:
; SSE2-NEXT: .long 2147483647
; SSE2-NEXT: .long 2147483647
; SSE2-NEXT: .long 2147483647
; SSE2-NEXT: .long 2147483647
; AVX: [[SIGNMASK2:L.+]]:
; AVX-NEXT: .long 2147483648
; AVX-NEXT: .long 2147483648
; AVX-NEXT: .long 2147483648
; AVX-NEXT: .long 2147483648
; AVX-NEXT: .long 2147483648
; AVX-NEXT: .long 2147483648
; AVX-NEXT: .long 2147483648
; AVX-NEXT: .long 2147483648
; AVX: [[MAGMASK2:L.+]]:
; AVX-NEXT: .long 2147483647
; AVX-NEXT: .long 2147483647
; AVX-NEXT: .long 2147483647
; AVX-NEXT: .long 2147483647
; AVX-NEXT: .long 2147483647
; AVX-NEXT: .long 2147483647
; AVX-NEXT: .long 2147483647
; AVX-NEXT: .long 2147483647
[x86] add test to show unnecessary scalarization of copysign intrinsics (PR30433) llvm-svn: 283071
2016-10-03 00:31:35 +08:00
define <8 x float> @v8f32(<8 x float> %a, <8 x float> %b) nounwind {
; SSE2-LABEL: v8f32:
[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
; SSE2: # %bb.0:
[X86][SSE] Canonicalize OR(AND(X,C),AND(Y,~C)) -> OR(AND(X,C),ANDNP(C,Y)) For constant bit select patterns, replace one AND with a ANDNP, allowing us to reuse the constant mask. Only do this if the mask has multiple uses (to avoid losing load folding) or if we have XOP as its VPCMOV can handle most folding commutations. This also requires computeKnownBitsForTargetNode support for X86ISD::ANDNP and X86ISD::FOR to prevent regressions in fabs/fcopysign patterns. Differential Revision: https://reviews.llvm.org/D55935 llvm-svn: 351819
2019-01-22 21:44:49 +08:00
; SSE2-NEXT: movaps [[MAGMASK2]](%rip), %xmm4
; SSE2-NEXT: movaps %xmm4, %xmm5
; SSE2-NEXT: andnps %xmm2, %xmm5
; SSE2-NEXT: andps %xmm4, %xmm0
; SSE2-NEXT: orps %xmm5, %xmm0
; SSE2-NEXT: andps %xmm4, %xmm1
; SSE2-NEXT: andnps %xmm3, %xmm4
; SSE2-NEXT: orps %xmm4, %xmm1
[x86] add test to show unnecessary scalarization of copysign intrinsics (PR30433) llvm-svn: 283071
2016-10-03 00:31:35 +08:00
; SSE2-NEXT: retq
;
; AVX-LABEL: v8f32:
[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
; AVX: # %bb.0:
[x86, SSE/AVX] allow 128/256-bit lowering for copysign vector intrinsics (PR30433) This should fix: https://llvm.org/bugs/show_bug.cgi?id=30433 There are a couple of open questions about the codegen: 1. Should we let scalar ops be scalars and avoid vector constant loads/splats? 2. Should we have a pass to combine constants such as the inverted pair that we have here? Differential Revision: https://reviews.llvm.org/D25165 llvm-svn: 283119
2016-10-04 00:38:27 +08:00
; AVX-NEXT: vandps [[SIGNMASK2]](%rip), %ymm1, %ymm1
; AVX-NEXT: vandps [[MAGMASK2]](%rip), %ymm0, %ymm0
; AVX-NEXT: vorps %ymm1, %ymm0, %ymm0
[x86] add test to show unnecessary scalarization of copysign intrinsics (PR30433) llvm-svn: 283071
2016-10-03 00:31:35 +08:00
; AVX-NEXT: retq
;
%tmp = tail call <8 x float> @llvm.copysign.v8f32( <8 x float> %a, <8 x float> %b )
ret <8 x float> %tmp
}
[x86, SSE/AVX] allow 128/256-bit lowering for copysign vector intrinsics (PR30433) This should fix: https://llvm.org/bugs/show_bug.cgi?id=30433 There are a couple of open questions about the codegen: 1. Should we let scalar ops be scalars and avoid vector constant loads/splats? 2. Should we have a pass to combine constants such as the inverted pair that we have here? Differential Revision: https://reviews.llvm.org/D25165 llvm-svn: 283119
2016-10-04 00:38:27 +08:00
; CHECK: [[SIGNMASK3:L.+]]:
; CHECK-NEXT: .quad -9223372036854775808
; CHECK-NEXT: .quad -9223372036854775808
; CHECK: [[MAGMASK3:L.+]]:
; CHECK-NEXT: .quad 9223372036854775807
; CHECK-NEXT: .quad 9223372036854775807
[x86] add test to show unnecessary scalarization of copysign intrinsics (PR30433) llvm-svn: 283071
2016-10-03 00:31:35 +08:00
define <2 x double> @v2f64(<2 x double> %a, <2 x double> %b) nounwind {
; SSE2-LABEL: v2f64:
[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
; SSE2: # %bb.0:
[x86, SSE/AVX] allow 128/256-bit lowering for copysign vector intrinsics (PR30433) This should fix: https://llvm.org/bugs/show_bug.cgi?id=30433 There are a couple of open questions about the codegen: 1. Should we let scalar ops be scalars and avoid vector constant loads/splats? 2. Should we have a pass to combine constants such as the inverted pair that we have here? Differential Revision: https://reviews.llvm.org/D25165 llvm-svn: 283119
2016-10-04 00:38:27 +08:00
; SSE2-NEXT: andps [[SIGNMASK3]](%rip), %xmm1
; SSE2-NEXT: andps [[MAGMASK3]](%rip), %xmm0
[x86] add test to show unnecessary scalarization of copysign intrinsics (PR30433) llvm-svn: 283071
2016-10-03 00:31:35 +08:00
; SSE2-NEXT: orps %xmm1, %xmm0
; SSE2-NEXT: retq
;
; AVX-LABEL: v2f64:
[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
; AVX: # %bb.0:
[x86, SSE/AVX] allow 128/256-bit lowering for copysign vector intrinsics (PR30433) This should fix: https://llvm.org/bugs/show_bug.cgi?id=30433 There are a couple of open questions about the codegen: 1. Should we let scalar ops be scalars and avoid vector constant loads/splats? 2. Should we have a pass to combine constants such as the inverted pair that we have here? Differential Revision: https://reviews.llvm.org/D25165 llvm-svn: 283119
2016-10-04 00:38:27 +08:00
; AVX-NEXT: vandps [[SIGNMASK3]](%rip), %xmm1, %xmm1
; AVX-NEXT: vandps [[MAGMASK3]](%rip), %xmm0, %xmm0
; AVX-NEXT: vorps %xmm1, %xmm0, %xmm0
[x86] add test to show unnecessary scalarization of copysign intrinsics (PR30433) llvm-svn: 283071
2016-10-03 00:31:35 +08:00
; AVX-NEXT: retq
;
%tmp = tail call <2 x double> @llvm.copysign.v2f64( <2 x double> %a, <2 x double> %b )
ret <2 x double> %tmp
}
[x86, SSE/AVX] allow 128/256-bit lowering for copysign vector intrinsics (PR30433) This should fix: https://llvm.org/bugs/show_bug.cgi?id=30433 There are a couple of open questions about the codegen: 1. Should we let scalar ops be scalars and avoid vector constant loads/splats? 2. Should we have a pass to combine constants such as the inverted pair that we have here? Differential Revision: https://reviews.llvm.org/D25165 llvm-svn: 283119
2016-10-04 00:38:27 +08:00
; SSE2: [[MAGMASK4:L.+]]:
; SSE2-NEXT: .quad 9223372036854775807
; SSE2-NEXT: .quad 9223372036854775807
; AVX: [[SIGNMASK4:L.+]]:
; AVX-NEXT: .quad -9223372036854775808
; AVX-NEXT: .quad -9223372036854775808
; AVX-NEXT: .quad -9223372036854775808
; AVX-NEXT: .quad -9223372036854775808
; AVX: [[MAGMASK4:L.+]]:
; AVX-NEXT: .quad 9223372036854775807
; AVX-NEXT: .quad 9223372036854775807
; AVX-NEXT: .quad 9223372036854775807
; AVX-NEXT: .quad 9223372036854775807
[x86] add test to show unnecessary scalarization of copysign intrinsics (PR30433) llvm-svn: 283071
2016-10-03 00:31:35 +08:00
define <4 x double> @v4f64(<4 x double> %a, <4 x double> %b) nounwind {
; SSE2-LABEL: v4f64:
[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
; SSE2: # %bb.0:
[X86][SSE] Canonicalize OR(AND(X,C),AND(Y,~C)) -> OR(AND(X,C),ANDNP(C,Y)) For constant bit select patterns, replace one AND with a ANDNP, allowing us to reuse the constant mask. Only do this if the mask has multiple uses (to avoid losing load folding) or if we have XOP as its VPCMOV can handle most folding commutations. This also requires computeKnownBitsForTargetNode support for X86ISD::ANDNP and X86ISD::FOR to prevent regressions in fabs/fcopysign patterns. Differential Revision: https://reviews.llvm.org/D55935 llvm-svn: 351819
2019-01-22 21:44:49 +08:00
; SSE2-NEXT: movaps [[MAGMASK4]](%rip), %xmm4
; SSE2-NEXT: movaps %xmm4, %xmm5
; SSE2-NEXT: andnps %xmm2, %xmm5
; SSE2-NEXT: andps %xmm4, %xmm0
; SSE2-NEXT: orps %xmm5, %xmm0
; SSE2-NEXT: andps %xmm4, %xmm1
; SSE2-NEXT: andnps %xmm3, %xmm4
; SSE2-NEXT: orps %xmm4, %xmm1
[x86] add test to show unnecessary scalarization of copysign intrinsics (PR30433) llvm-svn: 283071
2016-10-03 00:31:35 +08:00
; SSE2-NEXT: retq
;
; AVX-LABEL: v4f64:
[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
; AVX: # %bb.0:
[x86, SSE/AVX] allow 128/256-bit lowering for copysign vector intrinsics (PR30433) This should fix: https://llvm.org/bugs/show_bug.cgi?id=30433 There are a couple of open questions about the codegen: 1. Should we let scalar ops be scalars and avoid vector constant loads/splats? 2. Should we have a pass to combine constants such as the inverted pair that we have here? Differential Revision: https://reviews.llvm.org/D25165 llvm-svn: 283119
2016-10-04 00:38:27 +08:00
; AVX-NEXT: vandps [[SIGNMASK4]](%rip), %ymm1, %ymm1
; AVX-NEXT: vandps [[MAGMASK4]](%rip), %ymm0, %ymm0
; AVX-NEXT: vorps %ymm1, %ymm0, %ymm0
[x86] add test to show unnecessary scalarization of copysign intrinsics (PR30433) llvm-svn: 283071
2016-10-03 00:31:35 +08:00
; AVX-NEXT: retq
;
%tmp = tail call <4 x double> @llvm.copysign.v4f64( <4 x double> %a, <4 x double> %b )
ret <4 x double> %tmp
}
declare <4 x float> @llvm.copysign.v4f32(<4 x float> %Mag, <4 x float> %Sgn)
declare <8 x float> @llvm.copysign.v8f32(<8 x float> %Mag, <8 x float> %Sgn)
declare <2 x double> @llvm.copysign.v2f64(<2 x double> %Mag, <2 x double> %Sgn)
declare <4 x double> @llvm.copysign.v4f64(<4 x double> %Mag, <4 x double> %Sgn)