File: 2003-08-03-CallArgLiveRanges.ll

package info (click to toggle)
llvm-2.9 2.9%2Bdfsg-7
  • links: PTS, VCS
  • area: main
  • in suites: wheezy
  • size: 69,236 kB
  • sloc: cpp: 439,289; ansic: 16,211; asm: 13,295; sh: 12,912; python: 4,459; ml: 4,274; perl: 2,093; makefile: 1,808; pascal: 1,481; exp: 427; objc: 334; xml: 283; ada: 235; lisp: 187; csh: 117; f90: 36
file content (18 lines) | stat: -rw-r--r-- 598 bytes parent folder | download | duplicates (4)
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: llc < %s -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
}