This is the mail archive of the gcc@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]

Re: Code enlargement


<<      - avoid this trick with '-Os', may be also with '-O2', because speed
        increase is very little, zero or below zero. Brnach prediction
        seems to work very well.
>>

Branch prediction is not much use for forward jumps not in a loop, so I don't
understand your statement here. Certainly the rearrangement of jumps that
GCC 3 does is in some cases a big win, so this needs careful analysis.


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