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: [patch][IRA] Really record loop exits


On Fri, Oct 12, 2012 at 3:31 PM, Vladimir Makarov wrote:
> Ops.  Sorry, Steven.  I did a wrong conclusion because I thought I would
> have found such code generation problem if it had an affect.

Oh, the patch shouldn't (and doesn't) change the generated code, that
is not how cfgloops works: record_loop_exits is just a cache.

If you use record_loop_exits, cfgloops sets up a cache of the loop
exits so that if you look them up frequently it can return the VEC of
edges faster. But if you don't use record_loop_exits, it still returns
the same set of edges (and in the same order, too). It just will be
slower because cfgloop will search for the loop exit edges by loading
the loop body and scanning the successor edges of the loop body basic
blocks.


> So your patch is ok.  Thanks for working on this.

I'll commit the patch later this weekend. Thanks for the quick review!

Ciao!
Steven


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