Messages in this thread | | | Date | Fri, 02 Apr 2004 22:53:54 +0200 | From | Andreas Hartmann <> | Subject | Re: Very poor performance with 2.6.4 |
| |
Andrew Morton wrote: > Andreas Hartmann <andihartmann@freenet.de> wrote: >> >> Now, I tested 2.6.5-rc3-mm4. Same procedure. >> The good news first: >> 2.6.5-rc3-mm4 is nearly as fast as 2.4.25 - it is about 2% slower than >> 2.4.25 (with preemption turned on). >> >> Now the bad news: >> The system-processor-time is unchanged abnormal high: it is 34% (!) higher >> than in 2.4.25 (and about 1% more than in 2.4.6). >> >> >> Btw: Did the other profile outputs help to find the problem? >> >> These are the profile-values for an example run (make of kernel 2.6.5rc2) >> with 2.6.5rc3mm4: > > Spending a lot of time on do_softirq() while compiling stuff is peculiar. > > What device drivers are running at the time? disk/network/usb/etc?
Module Size Used by Not tainted eepro100 19828 1 (autoclean) mii 2480 0 (autoclean) [eepro100] sis900 13036 1 (autoclean) crc32 2880 0 (autoclean) [sis900] usb-storage 26416 0 (unused) scsi_mod 87488 0 [usb-storage] uhci 25436 0 (unused) usbcore 62316 0 [usb-storage uhci] lvm-mod 44416 12 (autoclean) unix 15308 13 (autoclean)
These are all modules (drivers), which are running - in both cases (2.4.25 and 2.6.x).
Regards, Andreas Hartmann - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |