[LLVMbugs] [Bug 22698] New: Infinite loop in DAGCombine trying to combine fmuls

bugzilla-daemon at llvm.org bugzilla-daemon at llvm.org
Wed Feb 25 07:27:37 PST 2015


http://llvm.org/bugs/show_bug.cgi?id=22698

            Bug ID: 22698
           Summary: Infinite loop in DAGCombine trying to combine fmuls
           Product: libraries
           Version: 3.6
          Hardware: PC
                OS: Windows NT
            Status: NEW
          Severity: normal
          Priority: P
         Component: Common Code Generator Code
          Assignee: unassignedbugs at nondot.org
          Reporter: michael.m.kuperstein at intel.com
                CC: hans at chromium.org, llvmbugs at cs.uiuc.edu
    Classification: Unclassified

Happens both in 3.6 and trunk.

To reproduce, run through llc:

target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
target triple = "x86_64-pc-windows-msvc"

; Function Attrs: nounwind readnone uwtable
define <4 x float> @foo(<4 x float> %a) #0 {
entry:
  %add.i = fadd fast <4 x float> %a, %a
  %add.i13 = fadd fast <4 x float> %add.i, %add.i
  %add.i12 = fadd fast <4 x float> %add.i13, %add.i13
  %add.i11 = fadd fast <4 x float> %add.i12, %add.i12
  ret <4 x float> %add.i11
}

attributes #0 = { nounwind readnone uwtable "less-precise-fpmad"="false"
"no-frame-pointer-elim"="false" "no-infs-fp-math"="true"
"no-nans-fp-math"="true" "stack-protector-buffer-size"="8"
"unsafe-fp-math"="true" "use-soft-float"="false" }

!llvm.module.flags = !{!0}
!llvm.ident = !{!1}

!0 = !{i32 1, !"PIC Level", i32 2}
!1 = !{!"clang version 3.7.0 (trunk 229307) (llvm/trunk 229306)"}

I believe this is a regression from 3.5. 
At the very least, this is a regression on the Clang level (don't have 3.5
built right now, but tried running the C source that produced it through
gcc.godbolt, 3.5 succeeds, 3.6 rc2 and 3.7 fail).

-- 
You are receiving this mail because:
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-bugs/attachments/20150225/35a8f4cd/attachment.html>


More information about the llvm-bugs mailing list