[LLVMbugs] [Bug 17088] New: SIGFPE due to illegal FP optimization

bugzilla-daemon at llvm.org bugzilla-daemon at llvm.org
Tue Sep 3 11:44:59 PDT 2013


            Bug ID: 17088
           Summary: SIGFPE due to illegal FP optimization
           Product: new-bugs
           Version: trunk
          Hardware: PC
                OS: Linux
            Status: NEW
          Severity: normal
          Priority: P
         Component: new bugs
          Assignee: unassignedbugs at nondot.org
          Reporter: kcc at google.com
                CC: llvmbugs at cs.uiuc.edu
    Classification: Unclassified

<this is fresh clang trunk r189814 on linux x86_64>

% cat fpe.cc 
#include <fenv.h>
double zero = 0.0;
#define max(a, b) (a > b ? a : b)
int main() {
  double z = 3.0 / max(1.0, zero);
  return z == 3.0;
% clang  fpe.cc -lm -g && ./a.out 
% gcc  fpe.cc -lm -g && ./a.out 
% gcc -O1  fpe.cc -lm -g && ./a.out 
% clang -O1  fpe.cc -lm -g && ./a.out 
Floating point exception (core dumped)
% gdb ./a.out 
Reading symbols from /home/kcc/tmp/./a.out...done.
(gdb) r
Starting program: /home/kcc/tmp/./a.out 

Program received signal SIGFPE, Arithmetic exception.
0x000000000040057e in main () at fpe.cc:6
6      double z = 3.0 / max(1.0, zero);
(gdb) disassemble 
Dump of assembler code for function main():
   0x0000000000400560 <+0>:    push   %rax
   0x0000000000400561 <+1>:    mov    $0x4,%edi
   0x0000000000400566 <+6>:    callq  0x400460 <feenableexcept at plt>
   0x000000000040056b <+11>:    movsd  0x12d(%rip),%xmm0        # 0x4006a0
   0x0000000000400573 <+19>:    movsd  0x200ab5(%rip),%xmm1        # 0x601030
   0x000000000040057b <+27>:    movaps %xmm0,%xmm2
=> 0x000000000040057e <+30>:    divsd  %xmm1,%xmm2
   0x0000000000400582 <+34>:    movsd  0x11e(%rip),%xmm3        # 0x4006a8
   0x000000000040058a <+42>:    ucomisd %xmm1,%xmm3
   0x000000000040058e <+46>:    seta   %al
   0x0000000000400591 <+49>:    ucomisd %xmm0,%xmm2
   0x0000000000400595 <+53>:    setnp  %cl
   0x0000000000400598 <+56>:    sete   %dl
   0x000000000040059b <+59>:    and    %cl,%dl
   0x000000000040059d <+61>:    or     %al,%dl
   0x000000000040059f <+63>:    movzbl %dl,%eax
   0x00000000004005a2 <+66>:    pop    %rdx
   0x00000000004005a3 <+67>:    retq   
End of assembler dump.

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/20130903/55940a9d/attachment.html>

More information about the llvm-bugs mailing list