This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: max-inline-insns



On 22-jan-04, at 7:51, Andreas Jaeger wrote:


Segher Boessenkool <segher@kernel.crashing.org> writes:

On 16-jan-04, at 22:32, Jan Hubicka wrote:
in the recent few day I made quite serious testing of the inlining.

As an aside, I noticed powerpc64-linux glibc does not compile with TOT
GCC, because the param max-inline-insns is gone. Would it make sense
for this parameter to alias to one (or more) of the max-inline-insns-XXX
parameters, for compatibility (maybe with a warning)?

This should be fixed inside glibc in a portable way, there're better ways for this (using attributes etc),

Sure, no question about that, glibc is just how I found out about this, I'm not proposing to change the compiler for just one library on one platform's sake.

Would only emitting a warning telling to use one of the new params be
accepted?

This might technically count as a regression against 3.3 ;-)


Segher



Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]