File: fma-mutate.ll

package info (click to toggle)
llvm-toolchain-20 1%3A20.1.6-1~exp1
  • links: PTS, VCS
  • area: main
  • in suites: experimental
  • size: 2,111,304 kB
  • sloc: cpp: 7,438,677; ansic: 1,393,822; asm: 1,012,926; python: 241,650; f90: 86,635; objc: 75,479; lisp: 42,144; pascal: 17,286; sh: 10,027; ml: 5,082; perl: 4,730; awk: 3,523; makefile: 3,349; javascript: 2,251; xml: 892; fortran: 672
file content (45 lines) | stat: -rw-r--r-- 1,457 bytes parent folder | download | duplicates (10)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -verify-machineinstrs < %s -mtriple=powerpc64-unknown-linux-gnu -mcpu=pwr7 -mattr=+vsx -disable-ppc-vsx-fma-mutation=false | FileCheck %s

declare double @llvm.sqrt.f64(double)

; Test several VSX FMA mutation opportunities.

; This is reasonable transformation since it eliminates extra register copy.
define double @foo3_fmf(double %a) nounwind {
; CHECK-LABEL: foo3_fmf:
; CHECK:       # %bb.0:
; CHECK-NEXT:    xstsqrtdp 0, 1
; CHECK-NEXT:    bc 12, 2, .LBB0_2
; CHECK-NEXT:  # %bb.1:
; CHECK-NEXT:    vspltisw 2, -3
; CHECK-NEXT:    xsrsqrtedp 0, 1
; CHECK-NEXT:    addis 3, 2, .LCPI0_0@toc@ha
; CHECK-NEXT:    xvcvsxwdp 3, 34
; CHECK-NEXT:    xsmuldp 2, 1, 0
; CHECK-NEXT:    fmr 4, 3
; CHECK-NEXT:    xsmaddadp 4, 2, 0
; CHECK-NEXT:    lfs 2, .LCPI0_0@toc@l(3)
; CHECK-NEXT:    xsmuldp 0, 0, 2
; CHECK-NEXT:    xsmuldp 0, 0, 4
; CHECK-NEXT:    xsmuldp 1, 1, 0
; CHECK-NEXT:    xsmaddadp 3, 1, 0
; CHECK-NEXT:    xsmuldp 0, 1, 2
; CHECK-NEXT:    xsmuldp 1, 0, 3
; CHECK-NEXT:    blr
; CHECK-NEXT:  .LBB0_2:
; CHECK-NEXT:    xssqrtdp 1, 1
; CHECK-NEXT:    blr
  %r = call contract reassoc afn ninf double @llvm.sqrt.f64(double %a)
  ret double %r
}

define double @foo3_safe(double %a) nounwind {
; CHECK-LABEL: foo3_safe:
; CHECK:       # %bb.0:
; CHECK-NEXT:    xssqrtdp 1, 1
; CHECK-NEXT:    blr
  %r = call double @llvm.sqrt.f64(double %a)
  ret double %r
}