[Bug tree-optimization/35254] New: Should gcc fold DFP expression?

[hjl dot tools at gmail dot com]

On Linux/Intel64,

[EMAIL PROTECTED] stack-7]$ cat y.c
extern void abort (void);

void
sse_test (void)
{
  if (7.999999999999999999999999999999999E6144dl + 3.0E6144dl
      !=  9.999999999999999999999999999999999E+6144dl)
    abort ();
}
[EMAIL PROTECTED] stack-7]$ /usr/gcc-4.3/bin/gcc -O2 -c -std=gnu99 y.c
[EMAIL PROTECTED] stack-7]$ readelf -r y.o

Relocation section '.rela.text' at offset 0x5f0 contains 3 entries:
  Offset          Info           Type           Sym. Value    Sym. Name +
Addend
000000000049  000800000002 R_X86_64_PC32     0000000000000000 __bid_addtd3 +
fffffffffffffffc
000000000070  000900000002 R_X86_64_PC32     0000000000000000 __bid_netd2 +
fffffffffffffffc
00000000007a  000b00000002 R_X86_64_PC32     0000000000000000 abort +
fffffffffffffffc

Relocation section '.rela.eh_frame' at offset 0x638 contains 1 entries:
  Offset          Info           Type           Sym. Value    Sym. Name +
Addend
000000000020  00020000000a R_X86_64_32       0000000000000000 .text + 0
[EMAIL PROTECTED] stack-7]$


-- 
           Summary: Should gcc fold DFP expression?
           Product: gcc
           Version: 4.4.0
            Status: UNCONFIRMED
          Severity: enhancement
          Priority: P3
         Component: tree-optimization
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: hjl dot tools at gmail dot com


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=35254