forked from OSchip/llvm-project
Tidy up this rather crufty test. Put the declarations at the top to make
my C-brain happy. Remove the unnecessary bits of pedantic IR fluff like nounwind. Remove stray uses comments. Name things semantically rather than tN so that adding a new test in the middle doesn't cause pain, and so that new tests can be grouped semantically. This exposes how little systematic testing is going on here. I noticed this by finding several bugs via inspection and wondering why this test wasn't catching any of them. =[ llvm-svn: 147248
This commit is contained in:
parent
767bbe48c1
commit
44cf07228b
|
|
@ -1,90 +1,89 @@
|
||||||
; RUN: llc < %s -march=x86 -mcpu=yonah | FileCheck %s
|
; RUN: llc < %s -march=x86 -mcpu=yonah | FileCheck %s
|
||||||
|
|
||||||
define i32 @t1(i32 %x) nounwind {
|
declare i32 @llvm.cttz.i32(i32, i1)
|
||||||
%tmp = tail call i32 @llvm.ctlz.i32( i32 %x, i1 true )
|
declare i16 @llvm.ctlz.i16(i16, i1)
|
||||||
ret i32 %tmp
|
declare i32 @llvm.ctlz.i32(i32, i1)
|
||||||
; CHECK: t1:
|
|
||||||
; CHECK: bsrl
|
|
||||||
; CHECK-NOT: cmov
|
|
||||||
; CHECK: xorl $31,
|
|
||||||
; CHECK: ret
|
|
||||||
}
|
|
||||||
|
|
||||||
declare i32 @llvm.ctlz.i32(i32, i1) nounwind readnone
|
define i32 @cttz_i32(i32 %x) {
|
||||||
|
%tmp = call i32 @llvm.cttz.i32( i32 %x, i1 true )
|
||||||
define i32 @t2(i32 %x) nounwind {
|
|
||||||
%tmp = tail call i32 @llvm.cttz.i32( i32 %x, i1 true )
|
|
||||||
ret i32 %tmp
|
ret i32 %tmp
|
||||||
; CHECK: t2:
|
; CHECK: cttz_i32:
|
||||||
; CHECK: bsfl
|
; CHECK: bsfl
|
||||||
; CHECK-NOT: cmov
|
; CHECK-NOT: cmov
|
||||||
; CHECK: ret
|
; CHECK: ret
|
||||||
}
|
}
|
||||||
|
|
||||||
declare i32 @llvm.cttz.i32(i32, i1) nounwind readnone
|
define i16 @ctlz_i16(i16 %x, i16 %y) {
|
||||||
|
|
||||||
define i16 @t3(i16 %x, i16 %y) nounwind {
|
|
||||||
entry:
|
entry:
|
||||||
%tmp1 = add i16 %x, %y
|
%tmp1 = add i16 %x, %y
|
||||||
%tmp2 = tail call i16 @llvm.ctlz.i16( i16 %tmp1, i1 true ) ; <i16> [#uses=1]
|
%tmp2 = call i16 @llvm.ctlz.i16( i16 %tmp1, i1 true )
|
||||||
ret i16 %tmp2
|
ret i16 %tmp2
|
||||||
; CHECK: t3:
|
; CHECK: ctlz_i16:
|
||||||
; CHECK: bsrw
|
; CHECK: bsrw
|
||||||
; CHECK-NOT: cmov
|
; CHECK-NOT: cmov
|
||||||
; CHECK: xorl $15,
|
; CHECK: xorl $15,
|
||||||
; CHECK: ret
|
; CHECK: ret
|
||||||
}
|
}
|
||||||
|
|
||||||
declare i16 @llvm.ctlz.i16(i16, i1) nounwind readnone
|
define i32 @ctlz_i32(i32 %x) {
|
||||||
|
%tmp = call i32 @llvm.ctlz.i32( i32 %x, i1 true )
|
||||||
|
ret i32 %tmp
|
||||||
|
; CHECK: ctlz_i32:
|
||||||
|
; CHECK: bsrl
|
||||||
|
; CHECK-NOT: cmov
|
||||||
|
; CHECK: xorl $31,
|
||||||
|
; CHECK: ret
|
||||||
|
}
|
||||||
|
|
||||||
define i32 @t4(i32 %n) nounwind {
|
define i32 @ctlz_i32_cmov(i32 %n) {
|
||||||
entry:
|
entry:
|
||||||
; Generate a cmov to handle zero inputs when necessary.
|
; Generate a cmov to handle zero inputs when necessary.
|
||||||
; CHECK: t4:
|
; CHECK: ctlz_i32_cmov:
|
||||||
; CHECK: bsrl
|
; CHECK: bsrl
|
||||||
; CHECK: cmov
|
; CHECK: cmov
|
||||||
; CHECK: xorl $31,
|
; CHECK: xorl $31,
|
||||||
; CHECK: ret
|
; CHECK: ret
|
||||||
%tmp1 = tail call i32 @llvm.ctlz.i32(i32 %n, i1 false)
|
%tmp1 = call i32 @llvm.ctlz.i32(i32 %n, i1 false)
|
||||||
ret i32 %tmp1
|
ret i32 %tmp1
|
||||||
}
|
}
|
||||||
|
|
||||||
define i32 @t5(i32 %n) nounwind {
|
define i32 @ctlz_i32_fold_cmov(i32 %n) {
|
||||||
entry:
|
entry:
|
||||||
; Don't generate the cmovne when the source is known non-zero (and bsr would
|
; Don't generate the cmovne when the source is known non-zero (and bsr would
|
||||||
; not set ZF).
|
; not set ZF).
|
||||||
; rdar://9490949
|
; rdar://9490949
|
||||||
; CHECK: t5:
|
; CHECK: ctlz_i32_fold_cmov:
|
||||||
; CHECK: bsrl
|
; CHECK: bsrl
|
||||||
; CHECK-NOT: cmov
|
; CHECK-NOT: cmov
|
||||||
; CHECK: xorl $31,
|
; CHECK: xorl $31,
|
||||||
; CHECK: ret
|
; CHECK: ret
|
||||||
%or = or i32 %n, 1
|
%or = or i32 %n, 1
|
||||||
%tmp1 = tail call i32 @llvm.ctlz.i32(i32 %or, i1 false)
|
%tmp1 = call i32 @llvm.ctlz.i32(i32 %or, i1 false)
|
||||||
ret i32 %tmp1
|
ret i32 %tmp1
|
||||||
}
|
}
|
||||||
|
|
||||||
define i32 @t6(i32 %n) nounwind {
|
define i32 @ctlz_bsr(i32 %n) {
|
||||||
entry:
|
entry:
|
||||||
; Don't generate any xors when a 'ctlz' intrinsic is actually used to compute
|
; Don't generate any xors when a 'ctlz' intrinsic is actually used to compute
|
||||||
; the most significant bit, which is what 'bsr' does natively.
|
; the most significant bit, which is what 'bsr' does natively.
|
||||||
; CHECK: t6:
|
; CHECK: ctlz_bsr:
|
||||||
; CHECK: bsrl
|
; CHECK: bsrl
|
||||||
; CHECK-NOT: xorl
|
; CHECK-NOT: xorl
|
||||||
; CHECK: ret
|
; CHECK: ret
|
||||||
%ctlz = tail call i32 @llvm.ctlz.i32(i32 %n, i1 true)
|
%ctlz = call i32 @llvm.ctlz.i32(i32 %n, i1 true)
|
||||||
%bsr = xor i32 %ctlz, 31
|
%bsr = xor i32 %ctlz, 31
|
||||||
ret i32 %bsr
|
ret i32 %bsr
|
||||||
}
|
}
|
||||||
|
|
||||||
define i32 @t7(i32 %n) nounwind {
|
define i32 @ctlz_bsr_cmov(i32 %n) {
|
||||||
entry:
|
entry:
|
||||||
; Same as t6, but ensure this happens even when there is a potential zero.
|
; Same as ctlz_bsr, but ensure this happens even when there is a potential
|
||||||
; CHECK: t7:
|
; zero.
|
||||||
|
; CHECK: ctlz_bsr_cmov:
|
||||||
; CHECK: bsrl
|
; CHECK: bsrl
|
||||||
; CHECK-NOT: xorl
|
; CHECK-NOT: xorl
|
||||||
; CHECK: ret
|
; CHECK: ret
|
||||||
%ctlz = tail call i32 @llvm.ctlz.i32(i32 %n, i1 false)
|
%ctlz = call i32 @llvm.ctlz.i32(i32 %n, i1 false)
|
||||||
%bsr = xor i32 %ctlz, 31
|
%bsr = xor i32 %ctlz, 31
|
||||||
ret i32 %bsr
|
ret i32 %bsr
|
||||||
}
|
}
|
||||||
|
|
|
||||||
Loading…
Reference in New Issue