
In https://github.com/llvm/llvm-project/issues/57452, we found that IRTranslator is translating `i1 true` into `i32 -1`. This is because IRTranslator uses SExt for indices. In this fix, we change the expected behavior of extractelement's index, moving from SExt to ZExt. This change includes both documentation, SelectionDAG and IRTranslator. We also included a test for AMDGPU, updated tests for AArch64, Mips, PowerPC, RISCV, VE, WebAssembly and X86 This patch fixes issue #57452. Differential Revision: https://reviews.llvm.org/D132978
181 lines
5.1 KiB
LLVM
181 lines
5.1 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
|
|
; RUN: llc -verify-machineinstrs -mtriple=powerpc64-ibm-aix-xcoff -vec-extabi -mcpu=pwr9 < %s | FileCheck %s -check-prefix=CHECK-64
|
|
; RUN: llc -verify-machineinstrs -mtriple=powerpc-ibm-aix-xcoff -vec-extabi -mcpu=pwr9 < %s | FileCheck %s -check-prefix=CHECK-32
|
|
|
|
define zeroext i8 @test1(<16 x i8> %a, i32 signext %index) {
|
|
; CHECK-64-LABEL: test1:
|
|
; CHECK-64: # %bb.0: # %entry
|
|
; CHECK-64-NEXT: clrldi 3, 3, 32
|
|
; CHECK-64-NEXT: vextublx 3, 3, 2
|
|
; CHECK-64-NEXT: clrldi 3, 3, 56
|
|
; CHECK-64-NEXT: blr
|
|
;
|
|
; CHECK-32-LABEL: test1:
|
|
; CHECK-32: # %bb.0: # %entry
|
|
; CHECK-32-NEXT: addi 4, 1, -16
|
|
; CHECK-32-NEXT: clrlwi 3, 3, 28
|
|
; CHECK-32-NEXT: stxv 34, -16(1)
|
|
; CHECK-32-NEXT: lbzx 3, 4, 3
|
|
; CHECK-32-NEXT: blr
|
|
entry:
|
|
%vecext = extractelement <16 x i8> %a, i32 %index
|
|
ret i8 %vecext
|
|
}
|
|
|
|
define signext i8 @test2(<16 x i8> %a, i32 signext %index) {
|
|
; CHECK-64-LABEL: test2:
|
|
; CHECK-64: # %bb.0: # %entry
|
|
; CHECK-64-NEXT: clrldi 3, 3, 32
|
|
; CHECK-64-NEXT: vextublx 3, 3, 2
|
|
; CHECK-64-NEXT: extsb 3, 3
|
|
; CHECK-64-NEXT: blr
|
|
;
|
|
; CHECK-32-LABEL: test2:
|
|
; CHECK-32: # %bb.0: # %entry
|
|
; CHECK-32-NEXT: addi 4, 1, -16
|
|
; CHECK-32-NEXT: clrlwi 3, 3, 28
|
|
; CHECK-32-NEXT: stxv 34, -16(1)
|
|
; CHECK-32-NEXT: lbzx 3, 4, 3
|
|
; CHECK-32-NEXT: extsb 3, 3
|
|
; CHECK-32-NEXT: blr
|
|
entry:
|
|
%vecext = extractelement <16 x i8> %a, i32 %index
|
|
ret i8 %vecext
|
|
}
|
|
|
|
define zeroext i16 @test3(<8 x i16> %a, i32 signext %index) {
|
|
; CHECK-64-LABEL: test3:
|
|
; CHECK-64: # %bb.0: # %entry
|
|
; CHECK-64-NEXT: clrldi 3, 3, 32
|
|
; CHECK-64-NEXT: rlwinm 3, 3, 1, 28, 30
|
|
; CHECK-64-NEXT: vextuhlx 3, 3, 2
|
|
; CHECK-64-NEXT: clrldi 3, 3, 48
|
|
; CHECK-64-NEXT: blr
|
|
;
|
|
; CHECK-32-LABEL: test3:
|
|
; CHECK-32: # %bb.0: # %entry
|
|
; CHECK-32-NEXT: addi 4, 1, -16
|
|
; CHECK-32-NEXT: rlwinm 3, 3, 1, 28, 30
|
|
; CHECK-32-NEXT: stxv 34, -16(1)
|
|
; CHECK-32-NEXT: lhzx 3, 4, 3
|
|
; CHECK-32-NEXT: blr
|
|
entry:
|
|
%vecext = extractelement <8 x i16> %a, i32 %index
|
|
ret i16 %vecext
|
|
}
|
|
|
|
define signext i16 @test4(<8 x i16> %a, i32 signext %index) {
|
|
; CHECK-64-LABEL: test4:
|
|
; CHECK-64: # %bb.0: # %entry
|
|
; CHECK-64-NEXT: clrldi 3, 3, 32
|
|
; CHECK-64-NEXT: rlwinm 3, 3, 1, 28, 30
|
|
; CHECK-64-NEXT: vextuhlx 3, 3, 2
|
|
; CHECK-64-NEXT: extsh 3, 3
|
|
; CHECK-64-NEXT: blr
|
|
;
|
|
; CHECK-32-LABEL: test4:
|
|
; CHECK-32: # %bb.0: # %entry
|
|
; CHECK-32-NEXT: addi 4, 1, -16
|
|
; CHECK-32-NEXT: rlwinm 3, 3, 1, 28, 30
|
|
; CHECK-32-NEXT: stxv 34, -16(1)
|
|
; CHECK-32-NEXT: lhax 3, 4, 3
|
|
; CHECK-32-NEXT: blr
|
|
entry:
|
|
%vecext = extractelement <8 x i16> %a, i32 %index
|
|
ret i16 %vecext
|
|
}
|
|
|
|
define zeroext i32 @test5(<4 x i32> %a, i32 signext %index) {
|
|
; CHECK-64-LABEL: test5:
|
|
; CHECK-64: # %bb.0: # %entry
|
|
; CHECK-64-NEXT: clrldi 3, 3, 32
|
|
; CHECK-64-NEXT: rlwinm 3, 3, 2, 28, 29
|
|
; CHECK-64-NEXT: vextuwlx 3, 3, 2
|
|
; CHECK-64-NEXT: blr
|
|
;
|
|
; CHECK-32-LABEL: test5:
|
|
; CHECK-32: # %bb.0: # %entry
|
|
; CHECK-32-NEXT: addi 4, 1, -16
|
|
; CHECK-32-NEXT: rlwinm 3, 3, 2, 28, 29
|
|
; CHECK-32-NEXT: stxv 34, -16(1)
|
|
; CHECK-32-NEXT: lwzx 3, 4, 3
|
|
; CHECK-32-NEXT: blr
|
|
entry:
|
|
%vecext = extractelement <4 x i32> %a, i32 %index
|
|
ret i32 %vecext
|
|
}
|
|
|
|
define signext i32 @test6(<4 x i32> %a, i32 signext %index) {
|
|
; CHECK-64-LABEL: test6:
|
|
; CHECK-64: # %bb.0: # %entry
|
|
; CHECK-64-NEXT: clrldi 3, 3, 32
|
|
; CHECK-64-NEXT: rlwinm 3, 3, 2, 28, 29
|
|
; CHECK-64-NEXT: vextuwlx 3, 3, 2
|
|
; CHECK-64-NEXT: extsw 3, 3
|
|
; CHECK-64-NEXT: blr
|
|
;
|
|
; CHECK-32-LABEL: test6:
|
|
; CHECK-32: # %bb.0: # %entry
|
|
; CHECK-32-NEXT: addi 4, 1, -16
|
|
; CHECK-32-NEXT: rlwinm 3, 3, 2, 28, 29
|
|
; CHECK-32-NEXT: stxv 34, -16(1)
|
|
; CHECK-32-NEXT: lwzx 3, 4, 3
|
|
; CHECK-32-NEXT: blr
|
|
entry:
|
|
%vecext = extractelement <4 x i32> %a, i32 %index
|
|
ret i32 %vecext
|
|
}
|
|
|
|
; Test with immediate index
|
|
define zeroext i8 @test7(<16 x i8> %a) {
|
|
; CHECK-64-LABEL: test7:
|
|
; CHECK-64: # %bb.0: # %entry
|
|
; CHECK-64-NEXT: li 3, 1
|
|
; CHECK-64-NEXT: vextublx 3, 3, 2
|
|
; CHECK-64-NEXT: clrldi 3, 3, 56
|
|
; CHECK-64-NEXT: blr
|
|
;
|
|
; CHECK-32-LABEL: test7:
|
|
; CHECK-32: # %bb.0: # %entry
|
|
; CHECK-32-NEXT: stxv 34, -16(1)
|
|
; CHECK-32-NEXT: lbz 3, -15(1)
|
|
; CHECK-32-NEXT: blr
|
|
entry:
|
|
%vecext = extractelement <16 x i8> %a, i32 1
|
|
ret i8 %vecext
|
|
}
|
|
|
|
define zeroext i16 @test8(<8 x i16> %a) {
|
|
; CHECK-64-LABEL: test8:
|
|
; CHECK-64: # %bb.0: # %entry
|
|
; CHECK-64-NEXT: li 3, 2
|
|
; CHECK-64-NEXT: vextuhlx 3, 3, 2
|
|
; CHECK-64-NEXT: clrldi 3, 3, 48
|
|
; CHECK-64-NEXT: blr
|
|
;
|
|
; CHECK-32-LABEL: test8:
|
|
; CHECK-32: # %bb.0: # %entry
|
|
; CHECK-32-NEXT: stxv 34, -16(1)
|
|
; CHECK-32-NEXT: lhz 3, -14(1)
|
|
; CHECK-32-NEXT: blr
|
|
entry:
|
|
%vecext = extractelement <8 x i16> %a, i32 1
|
|
ret i16 %vecext
|
|
}
|
|
|
|
define zeroext i32 @test9(<4 x i32> %a) {
|
|
; CHECK-64-LABEL: test9:
|
|
; CHECK-64: # %bb.0:
|
|
; CHECK-64-NEXT: li 3, 12
|
|
; CHECK-64-NEXT: vextuwlx 3, 3, 2
|
|
; CHECK-64-NEXT: blr
|
|
;
|
|
; CHECK-32-LABEL: test9:
|
|
; CHECK-32: # %bb.0:
|
|
; CHECK-32-NEXT: stxv 34, -16(1)
|
|
; CHECK-32-NEXT: lwz 3, -4(1)
|
|
; CHECK-32-NEXT: blr
|
|
%vecext = extractelement <4 x i32> %a, i32 3
|
|
ret i32 %vecext
|
|
}
|