File: 2003-08-03-CallArgLiveRanges.ll

package info (click to toggle)
llvm 2.6-9.1
  • links: PTS
  • area: main
  • in suites: squeeze
  • size: 57,604 kB
  • ctags: 44,336
  • sloc: cpp: 344,766; sh: 12,407; ansic: 10,617; ada: 3,070; ml: 2,505; perl: 2,496; makefile: 1,426; pascal: 1,163; exp: 389; asm: 307; python: 298; objc: 260; lisp: 182; csh: 117; xml: 38; f90: 36; tcl: 20
file content (18 lines) | stat: -rw-r--r-- 608 bytes parent folder | download
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
; The old instruction selector used to load all arguments to a call up in 
; registers, then start pushing them all onto the stack.  This is bad news as
; it makes a ton of annoying overlapping live ranges.  This code should not
; cause spills!
;
; RUN: llvm-as < %s | llc -march=x86 -stats |& not grep spilled

target datalayout = "e-p:32:32"

define i32 @test(i32, i32, i32, i32, i32, i32, i32, i32, i32, i32) {
        ret i32 0
}

define i32 @main() {
        %X = call i32 @test( i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10 )            ; <i32> [#uses=1]
        ret i32 %X
}