real.c (encode_ibm_extended): Don't bother rounding low double.
authorAlan Modra <amodra@bigpond.net.au>
Thu, 4 Mar 2004 10:23:23 +0000 (10:23 +0000)
committerAlan Modra <amodra@gcc.gnu.org>
Thu, 4 Mar 2004 10:23:23 +0000 (20:53 +1030)
commitddc6856430bcef3e68563b08cc513f794a5cb57e
tree2a5b34507a773fc68de86a6f3a19dec1a8df7b3e
parentca87076c9bfac0090fcc26c06dd1c24dc5e72bc9
real.c (encode_ibm_extended): Don't bother rounding low double.

* real.c (encode_ibm_extended): Don't bother rounding low double.
* c-cppbuiltin.c (builtin_define_float_constants): Tweak MAX
when fmt->pnan < fmt->p.

From-SVN: r78900
gcc/ChangeLog
gcc/c-cppbuiltin.c
gcc/real.c