File: 2002-04-07-HexFloatConstants.ll

package info (click to toggle)
llvm-toolchain-21 1%3A21.1.0-1
  • links: PTS, VCS
  • area: main
  • in suites: sid
  • size: 2,235,796 kB
  • sloc: cpp: 7,617,614; ansic: 1,433,901; asm: 1,058,726; python: 252,096; f90: 94,671; objc: 70,753; lisp: 42,813; pascal: 18,401; sh: 10,032; ml: 5,111; perl: 4,720; awk: 3,523; makefile: 3,401; javascript: 2,272; xml: 892; fortran: 770
file content (17 lines) | stat: -rw-r--r-- 730 bytes parent folder | download | duplicates (13)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
; This testcase checks to make sure that the assembler can handle floating 
; point constants in IEEE hex format. This also checks that the disassembler,
; when presented with a FP constant that cannot be represented exactly in 
; exponential form, outputs it correctly in hex format.  This is a distillation
; of the bug that was causing the Olden Health benchmark to output incorrect
; results!
;
; RUN: opt -passes=instsimplify -S > %t.1 < %s
; RUN: llvm-as < %s | llvm-dis | llvm-as | opt -passes=instsimplify | \
; RUN: llvm-dis > %t.2
; RUN: diff %t.1 %t.2
; RUN: verify-uselistorder %s

define double @test() {
        %tmp = fmul double 7.200000e+101, 0x427F4000             ; <double> [#uses=1]
        ret double %tmp
}