http://bugzilla.novell.com/show_bug.cgi?id=568647 http://bugzilla.novell.com/show_bug.cgi?id=568647#c3 Stefan Seyfried <seife@novell.slipkontur.de> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO Info Provider| |robermann@gmail.com --- Comment #3 from Stefan Seyfried <seife@novell.slipkontur.de> 2010-01-07 11:35:10 CET --- BTW: I can confirm that "-kernel-kqemu" is slower here on my machine processor : 0 vendor_id : GenuineIntel cpu family : 6 model : 13 model name : Intel(R) Pentium(R) M processor 1.70GHz stepping : 6 cpu MHz : 1700.000 cache size : 2048 KB for a specific testcase: start the memtest86 from 11.2 CD, and the memory test speed with "-kernel-kqemu" is ~128MB/second, with "-enable-kqemu" (and without kqemu at all) it is ~440MB/second. But for other test cases it seems to perform mostly as expected: Booting an 11.2 rescue system without kqemu took 4:55 minutes, with "-enable-kqemu" it took 2:13 minutes, with "-kernel-kqemu" it did not find the CD drive, so there are other problems with "-kernel-kqemu". Then I measured how long it took until "starting udev... ok" and "loading basic drivers... ok" appeared during rescue system boot: starting udev | loading basic drivers kernel-kqemu: 28 seconds 0:51 minutes enable-kqemu: 26 seconds 1:24 minutes no kqemu: 37 seconds 2:49 minutes Those measurements were taken with a stopwatch and run only once, so they are not too exact, but generally it seems to work as expected. kernel-kqemu seems to be negatively impact some of the workloads (starting udev) and hugely improve some of the others (loading modules), but that's probably also to be expected. So unless you can show me more facts, I'm inclined to close this as "worksforme". Note: all tests were run on FACTORY, but I dont think too much has changed wrt. kqemu since 11.2. -- Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.