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]
Other format: [Raw text]

Re: gcc compile-time performance


Stan Shebs writes:

 > Then there's the extreme approach of having maintainers only
 > accept patches that either remove code or make the compiler run
 > faster... :-)

Hmmm, my attempts at tidying up handling of bit sets for GCC fell upon
deaf ears (apart from the Bison maintainer).  Without micro optimising
the implementations, I measured several percent improvement
bootstrapping GCC. 

Further improvements could be made I believe by better selection of
the bit set implementation depending on the number and distribution of
the bits set.  For example, my analysis found that most of the time we
are dealing with bit sets with one or fewer bits set (CSE in
particular spends a lot of time hunting the one set bit).

Michael.



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