On Tue, Jul 14, 2015 at 11:46 AM, Siarhei Siamashka <[email protected]> wrote: > On Thu, 2 Jul 2015 13:04:16 +0300 > Oded Gabbay <[email protected]> wrote: > >> no changes were observed when running cairo trimmed benchmarks. >> >> Signed-off-by: Oded Gabbay <[email protected]> > > This seems to cause performance regression in "lowlevel-blt-bench > src_0565_8888" test on my Playstation3: > > == before == > > src_0565_8888 = L1: 106.96 L2: 106.88 M: 85.11 ( 97.18%) > HT: 50.47 VT: 34.76 R: 34.82 RT: 13.19 ( 115Kops/s) > > == after == > > src_0565_8888 = L1: 59.04 L2: 57.74 M: 52.01 ( 59.39%) > HT: 35.05 VT: 26.37 R: 26.78 RT: 11.73 ( 106Kops/s) > > -- > Best regards, > Siarhei Siamashka
ok, I'll remove it for now, and I will check it offline on my POWER7 machine Oded _______________________________________________ Pixman mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/pixman
