This is the mail archive of the
mailing list for the GCC project.
The -m4-nofpu fix didn't work...
- To: gcc-patches at gcc dot gnu dot org
- Subject: The -m4-nofpu fix didn't work...
- From: Bernd Schmidt <bernds at redhat dot co dot uk>
- Date: Fri, 15 Sep 2000 18:52:35 +0100 (BST)
- cc: Elena Zannoni <ezannoni at cygnus dot com>
I checked in a patch yesterday to fix a problem on the sh with the -m4-nofpu
option. That patch would have worked, except for
MULTILIB_MATCHES = m2=m3 m2=m4-nofpu
in t-sh which means that we don't actually _build_ a multilib for -m4-nofpu.
I'm not sure what the best way to fix this would be - either remove the
"m2=m4-nofpu" (slows down builds), or make sure we don't call the sh4-specific
functions in lib1calls.asm (possibly slows down executables). Any comments?
I'm favouring solution #1 (at least it doesn't make the _tests_ slower).