File: uniformshift.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 (25 lines) | stat: -rw-r--r-- 828 bytes parent folder | download | duplicates (6) 
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
 
; RUN: opt -mtriple=x86_64-apple-darwin -mattr=+sse2 -passes=loop-vectorize -debug-only=loop-vectorize -S < %s 2>&1 | FileCheck %s
; REQUIRES: asserts

; CHECK: 'foo'
; CHECK: LV: Found an estimated cost of 1 for VF 1 For instruction:   %shift = ashr i32 %val, %k
; CHECK: Cost of 2 for VF 2: WIDEN ir<%shift> = ashr ir<%val>, ir<%k>
; CHECK: Cost of 2 for VF 4: WIDEN ir<%shift> = ashr ir<%val>, ir<%k>
define void @foo(ptr nocapture %p, i32 %k) local_unnamed_addr #0 {
entry:
  br label %body

body:
  %i = phi i64 [ 0, %entry ], [ %next, %body ]
  %ptr = getelementptr inbounds i32, ptr %p, i64 %i
  %val = load i32, ptr %ptr, align 4
  %shift = ashr i32 %val, %k
  store i32 %shift, ptr %ptr, align 4
  %next = add nuw nsw i64 %i, 1
  %cmp = icmp eq i64 %next, 16
  br i1 %cmp, label %exit, label %body

exit:
  ret void

}