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: Speedup CSE by 5%


Richard Guenther <richard.guenther@gmail.com> writes:
>
> Why is this not optimized by gcc itself?  Does marking approx_reg_cost_1
> inline help?

I don't think gcc can join multiple loops. Although it may be sometimes
useful, agreed, but I suspect it would need at least a new pass.

In the case of pointer chasing being involved like in this loop it
would probably need far more advanced alias analysis than gcc has
today.

-Andi


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