File: MachineSink-convergent.ll

package info (click to toggle)
llvm-toolchain-7 1%3A7.0.1-8~deb9u3
  • links: PTS, VCS
  • area: main
  • in suites: stretch
  • size: 733,456 kB
  • sloc: cpp: 3,776,651; ansic: 633,271; asm: 350,301; python: 142,716; objc: 107,612; sh: 22,626; lisp: 11,056; perl: 7,999; pascal: 6,742; ml: 5,537; awk: 3,536; makefile: 2,557; cs: 2,027; xml: 841; ruby: 156
file content (23 lines) | stat: -rw-r--r-- 645 bytes parent folder | download | duplicates (16)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
; RUN: llc < %s | FileCheck %s
target triple = "nvptx64-nvidia-cuda"

declare void @llvm.nvvm.barrier0()

; Load a value, then syncthreads.  Branch, and use the loaded value only on one
; side of the branch.  The load shouldn't be sunk beneath the call, because
; syncthreads is modeled as maystore.
define i32 @f(i32 %x, i32* %ptr, i1 %cond) {
Start:
  ; CHECK: ld.u32
  %ptr_val = load i32, i32* %ptr
  ; CHECK: bar.sync
  call void @llvm.nvvm.barrier0()
  br i1 %cond, label %L1, label %L2
L1:
  %ptr_val2 = add i32 %ptr_val, 100
  br label %L2
L2:
  %v4 = phi i32 [ %x, %Start ], [ %ptr_val2, %L1 ]
  %v5 = add i32 %v4, 1000
  ret i32 %v5
}