testcase question (portability)

Aldy Hernandez aldyh@redhat.com
Wed Feb 25 12:42:00 GMT 2004


> How in the world can SPE get this wrong?

Uhhhh... When the specs are wrong?  Kumar just sent me an email saying 
it should be efsctsiz/efsctuiz which round to zero (not nearest).  
Interestingly enough, the specs say efsct*iz round to nearest.

Yay.  Documentation bug.

Patch later today when I'm done with my SLOW testing cycle.

Cheers.
Aldy



More information about the Gcc mailing list