On Tue, Sep 16, 2014 at 11:28 AM, Jakub Jelinek <ja...@redhat.com> wrote: > On Tue, Sep 16, 2014 at 11:20:37AM +0200, Richard Biener wrote: >> > This confuses me, but, no matter. Isn’t 8hrs time data? :-) > > It is, but not time(1) data, just wall clock computed from subtracting > mtimes of my make check output log and make -j48 bootstrap log. > >> >> patch toplevel make -j48 -k check took: >> >> real 40m21.984s >> >> user 341m51.675s >> >> sys 112m46.993s >> >> and with the patch make -j48 -k check took: >> >> real 32m22.066s >> >> user 355m1.788s >> >> sys 117m5.809s >> > >> > These numbers are useful to try and ensure the overhead (scaling factor) >> > is reasonable, thanks. >> >> A nice improvement indeed. The patched result is 15 times faster >> than the serial unpatched run. So there is room for improvement > > Note, the box used was oldish AMD 16-core, no ht, box, haven't tried it on > anything
Ah, I assumed -j48 testing means you have 48 cores. I usually test with -j12 on my 6-core HT-enabled box. A factor 15 scaling for 16 CPUs is of course close to the best we can achieve. Richard. > more parallel, also it was normal hard disk, etc. No idea whether anything > from this is relevant to that though. > Some CPU time goes into the expect processes, I can retry the build tonight > and grab also time(1) info from make -k check to see the user/sys times for > serial testing. > > Jakub