Floating-point warning fixes; fix round-to-overflow
commitfab3a6c9de9c56f1db7770c7c5e0271e33455ea2
authorH. Peter Anvin <hpa@zytor.com>
Tue, 16 Oct 2007 18:48:07 +0000 (16 11:48 -0700)
committerH. Peter Anvin <hpa@zytor.com>
Tue, 16 Oct 2007 18:48:07 +0000 (16 11:48 -0700)
treec46a56d27737c008cdd2b8c30edcacf433847379
parent125c878e96b5aeb50ae0c7f68847200262a9340f
Floating-point warning fixes; fix round-to-overflow

Actually generate the appropriate floating-point warnings, and only
one per assembly, pretty please.

Correct the round-to-overflow condition; as written all numbers with a
positive exponent were considered overflows!
float.c
nasm.c
nasmlib.h