File: gv-offset-folding.ll

package info (click to toggle)
llvm-toolchain-15 1%3A15.0.6-4
  • links: PTS, VCS
  • area: main
  • in suites: bookworm
  • size: 1,554,644 kB
  • sloc: cpp: 5,922,452; ansic: 1,012,136; asm: 674,362; python: 191,568; objc: 73,855; f90: 42,327; lisp: 31,913; pascal: 11,973; javascript: 10,144; sh: 9,421; perl: 7,447; ml: 5,527; awk: 3,523; makefile: 2,520; xml: 885; cs: 573; fortran: 567
file content (21 lines) | stat: -rw-r--r-- 866 bytes parent folder | download | duplicates (15) 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
 
; RUN: llc -mtriple=amdgcn--amdhsa -mcpu=fiji -mattr=-flat-for-global -relocation-model=static < %s | FileCheck %s

@lds = external addrspace(3) global [4 x i32]

; Function Attrs: nounwind

; Offset folding is an optimization done for global variables with relocations,
; which allows you to store the offset in the r_addend of the relocation entry.
; The offset is apllied to the variables address at link time, which eliminates
; the need to emit shader instructions to do this calculation.
; We don't use relocations for local memory, so we should never fold offsets
; for local memory globals.

; CHECK-LABEL: lds_no_offset:
; CHECK: ds_write_b32 v{{[0-9]+}}, v{{[0-9]+}} offset:4
define amdgpu_kernel void @lds_no_offset() {
entry:
  %ptr = getelementptr [4 x i32], [4 x i32] addrspace(3)* @lds, i32 0, i32 1
  store i32 0, i32 addrspace(3)* %ptr
  ret void
}