PATCH: enable odd floating-point registers for MIPS32 and MIPS64

[Sandra Loosemore]

This patch enables the use of both halves of paired floating-point registers to 
hold 32-bit values on MIPS32 and MIPS64.  (Currently, GCC treats each pair as a 
unit, as older MIPS architectures did.)  I did my original testing against gcc 
version 4.2.  OK for mainline once sanity check re-testing is complete?

-Sandra

-------------- next part --------------
A non-text attachment was scrubbed...
Name: 04-mips32-use-odd-floats.log
Type: text/x-log
Size: 539 bytes
Desc: not available
URL: <http://gcc.gnu.org/pipermail/gcc-patches/attachments/20070507/546aa4ac/attachment.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 04-mips32-use-odd-floats.patch
Type: text/x-patch
Size: 6803 bytes
Desc: not available
URL: <http://gcc.gnu.org/pipermail/gcc-patches/attachments/20070507/546aa4ac/attachment-0001.bin>