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: Reduce inline-insns-auto


On Wed, Sep 16, 2009 at 4:33 AM, Richard Guenther
<richard.guenther@gmail.com> wrote:

> The units are sort of gimple statements and are target independent.
>
>>?You report that the eon benchmark is very sensitive to
>> the result and have
>> tuned the parameters for x86_64, which seems like it will push all
>> RISC architectures
>> over the limit.
>
> How do you come to this conclusion?

Thanks for the clarification.  The parameter and variables are named
"insn", not "stmt", and I did not see any documentation of the units.
I misunderstood the code and thought that the pass was "estimating"
the size through trial RTL expansion, similar to the loop optimizer.

Thanks, David


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