On Nov 28, 2005, at 6:12 PM, carmen wrote:
On Mon, Nov 28, 2005 at 03:35:30PM -0500, Hans-Christoph Steiner wrote:
Has anyone ever done any profiling to see whether it makes any difference if you compile TclTk with the CPU compiler flags?
switching to 8.5 did seem to help a decent amount, more so than any optimilization flags.. kind of like the sIMD additions to devel helps much more than slopping on a bunch of -O3 -fomit-frame-pointer gunk. but youre welcome to play funroll-loops and let us know what you find :)
Yes, I can imagine that SIMD stuff would help. Have you run into any bugs with 8.5?
One of the next steps is to make a global OPTIM_FLAGS variable, so that the
the last 'pd wont compile' to this list was actually caused by aggressive optimization flags coupled with redhat's screwy compilers..
Yeah, this would just be for personal gratification. The default would be conservative.
.hc
whole package can be built for a certain CPU. I wonder whether TclTk
PD-dev mailing list PD-dev@iem.at http://lists.puredata.info/listinfo/pd-dev
________________________________________________________________________ ____
"Computer science is no more related to the computer than astronomy is related to the telescope." -Edsger Dykstra