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: mainline "exploding"


> On Sat, 18 Oct 2003, Diego Novillo wrote:
> >> Here is some data for testcase from PR8361:
> >>
> >> 	  	-O0 	-O1	-O2	-O3
> >> ----------------------------------------------
> >> GCC 3.2.3 	28.98	54.27	70.90	73.9
> >> GCC 3.3.2 	30.13	54.59	76.39	77.89
> >> mainline  	31.91	113.22	105.29	106.30
> >>  w/o checking   26.37   90.78   87.88   89.16
> >> tree-ssa        32.60   71.49   93.78   99.14
> >>
> >> Mainline (w/o checking) is nearly twice as slow as 3.3.2 for -O1.
> > Are the timings for tree-ssa with checking enabled?
> 
> No, disabled as well.
tree-ssa has unit-at-a-time inline heuristics disabled, so I guess 99 is
pretty good result (it produces same amount of tree nodes as 3.3.2 -O1
and compiles them faster at -O2).
It may be interetsting to try out the speed with my patch to enable
unit-at-a-time on tree-ssa.

Hmm, on the other hand the function size estimates is much larger on tree-ssa
branch because of new nodes introduced by gimplification process.  I
will write proper generic code size estimation soonish so we can see
better comparisons then.

Honza
> 
> Gerald


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