This is the mail archive of the
mailing list for the GCC project.
Re: Testsuite tweaks on SPARC
> Perhaps. Consider someone who has a biarch gcc installed, and shell
> scripts called powerpc-linux-gcc and powerpc64-linux-gcc, with the
> script providing -m32 or -m64. Is it reasonable to abort if I run
> powerpc64-linux-gcc -m32? Of course, you can fix this with suitable
> editing of args inside the script, but why should -m32/-m64 be treated
> any differently than say -O1 and -O2? Those are mutually exclusive too!
Yes, consistency should probably win here. But note that the current
behaviour has not generated much outcry so far.