https://bugzilla.novell.com/show_bug.cgi?id=302327#c1 Jeff Mahoney changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO Info Provider| |wns@comcast.net --- Comment #1 from Jeff Mahoney 2007-08-25 09:35:58 MST --- More information on this report can be found on the opensuse-kernel mailing list, here: http://lists.opensuse.org/opensuse-kernel/2007-08/msg00070.html The key thing from that discussion is that the vanilla 2.6.22.3 doesn't exhibit the lockups that the openSUSE kernel does. Warren, have you had any luck setting up a console mechanism? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=302327#c3 Stephan Kulow changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |coolo@novell.com Status|NEW |NEEDINFO Info Provider| |wns@comcast.net Flag| |SHIP_STOPPER- --- Comment #3 from Stephan Kulow 2007-09-14 06:55:23 MST --- is this fixed in later beta releases? Because I know several installations of x86_64 -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=302327#c5 --- Comment #5 from Warren Stockton 2007-09-20 23:02:16 MST --- I have updated to 10.3RC1 but the update kernels were also installed (so I don't know if the kernel-default-2.6.22.5-23 had the problem): # rpm -qa kernel-* kernel-source-2.6.22.5-25 kernel-vanilla-2.6.22.5-23 kernel-default-2.6.22.5-25 I still get the hang with the latest kernel. I have noticed that if I set the RTC via the BIOS setup to the correct time and then run the -default kernel until it hangs, I find than when I reboot the RTC has a wrong (corrupted) time. The RTC date appears to still be correct. I am running ntpd and wondering if this has anything to do with the kernel's sync_cmos_clock() I don't know if this ties back to issues I encounter with hwclock where it can also hang then system just reading the RTC: # hwclock --hctosys -u --test --debug hwclock from util-linux-2.13-rc2 Using /dev interface to clock. Assuming hardware clock is kept in UTC time. Waiting for clock tick... /dev/rtc does not have interrupt functions. Waiting in loop for time from /dev/rtc to change <<< hang will intermittently occur here >>> ..got clock tick Time read from Hardware Clock: 2007/10/21 00:55:51 Hw clock time : 2007/10/21 00:55:51 = 1192928151 seconds since 1969 Calling settimeofday: tv.tv_sec = 1192928151, tv.tv_usec = 0 tz.tz_minuteswest = 360 Not setting system clock because running in test mode. (intermittent hangs are encountered with both --localtime or --utc) Comparing /boot/config-2.6.22.1-16-default (which works) and /boot/config-2.6.22.5-25-default (which hangs), I notice that CONFIG_HPET_EMULATE_RTC=y is defined in 2.6.22.1-16-default This laptop (TL-56) is using hpet: # cat /sys/devices/system/clocksource/clocksource0/current_clocksource hpet The "older" laptop (TL-52) does not have hpet and uses acpi_pm as a clocksource. Booting the TL-56 with "clocksource=acpi_pm" and -default kernel does not avoid the hang. I plan to build a 2.6.22.5-25-default kernel that defines CONFIG_HPET_EMULATE_RTC=y to see if this narrows down the problem, but this will have to wait for tomorrow. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=302327#c7 Warren Stockton changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |NEW Info Provider|wns@comcast.net | --- Comment #7 from Warren Stockton 2007-10-18 15:04:54 MST --- Yes. I found that this lookup occurred with all the kernel-default-2.6.22.x kernels that have been released for 10.3 (After updating to the 10.3-GM, I would run the kernel-default from the 10.3-Beta1 media to avoid the problem.) The last 2.6.22 kernel that I ran that still encountered the problem was: ftp://ftp.suse.com/pub/projects/kernel/kotd/HEAD/x86_64/kernel-default-2.6.22.10-20071011115904.x86_64.rpm I have recently updated to 10.3.1 (on Oct 15) along with the kernel from http://download.opensuse.org/repositories/Kernel:/HEAD/openSUSE_Factory # cat /proc/version Linux version 2.6.23.1-62-default (geeko@buildhost) (gcc version 4.2.1 (SUSE Linux)) #1 SMP 2007/10/17 17:52:09 UTC As I type up this response, I have the following uptime: # uptime 3:00pm up 5:55, 2 users, load average: 0.61, 0.34, 0.17 This kernel appears to be quite stable and appears to avoid the lockup. I wish I could be more positive but I never found a test case that can reproduce the hang at will. The previous 2.6.23.1-46-default was the first 2.6.23 kernel I tried and it also appeared to avoid the lockup. I can drop back to 10.3-GM and any other other kernel if I need to try some tests. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=302327#c9 Warren Stockton changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |NEW Info Provider|wns@comcast.net | --- Comment #9 from Warren Stockton 2007-10-23 12:55:43 MST --- Sure. Andi previously built 5 kernels that I tried:

0 is the kernel with all patches for a control run, 1-4 are without ACPI, libata, ALSA, ieee1394 respectively. All of these still encountered the lockup.

I did notice that it requires runlevel 5 to expose this issue. I booted to runlevel 1 and left the system for about two hours and it was still OK, but within 15 minutes of switching to runlevel 5, it had locked up. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=302327#c11 Warren Stockton changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |NEW Info Provider|wns@comcast.net | --- Comment #11 from Warren Stockton 2007-10-23 14:18:36 MST --- Umm... That is an i586 kernel!!! At least as far as the filename is concerned. I will download and confirm. Can you provide and x86_64 kernel? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=302327#c13 Jeff Mahoney changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO Info Provider| |wns@comcast.net --- Comment #13 from Jeff Mahoney 2007-10-23 14:58:35 MST --- Ok, updated. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=302327#c15 Jeff Mahoney changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO Info Provider| |wns@comcast.net --- Comment #15 from Jeff Mahoney 2007-11-07 14:15:51 MST --- Ok, let's try a newer version then. I've updated the RPMs in the URL from comment #10. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=302327#c17 Jeff Mahoney changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO Info Provider| |wns@comcast.net --- Comment #17 from Jeff Mahoney 2007-11-08 11:28:51 MST --- Ok, that eliminates most of them. I expect that this kernel is probably the one which will end up not working for you: ftp://ftp.suse.com/pub/people/jeffm/suse/testpkgs/302327/10.3-x86_64/kernel-default-2.6.22.2-jeffm.e5e61a519e4f860c4595f966ecb897e5eb294d19.x86_64.rpm -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=302327#c19 Jeff Mahoney changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO Info Provider| |wns@comcast.net --- Comment #19 from Jeff Mahoney 2007-11-08 12:09:49 MST --- Ok, that corresponds to this change: Wed Aug 15 15:16:04 CEST 2007 - ak@suse.de - patches.arch/x86-early-quirks-unificiation: x86: Unify i386 and x86-64 early quirks. - patches.arch/x86-nvidia-timer-quirk: Replace nvidia timer override quirk with pci id list and unify quirks. - patches.arch/x86-fam10-l3cache: i386: Add L3 cache support to AMD CPUID4 emulation. - patches.arch/i386-compat-vdso: i386: allow debuggers to access the vsyscall page with compat vDSO. - patches.xen/xen3-auto-arch-i386.diff: Merge - patches.xen/xen-x86-no-lapic: Merge - patches.xen/xen3-x86-early-quirks-unificiation: x86: Unify i386 and x86-64 early quirks. - patches.xen/xen3-x86-fam10-l3cache: i386: Add L3 cache support to AMD CPUID4 emulation. Given your hardware, it's probably patches.arch/x86-nvidia-timer-quirk. Can you attach the full output of 'hwinfo' using a working kernel? -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=302327#c21 --- Comment #21 from Warren Stockton 2007-11-08 12:56:09 MST --- Created an attachment (id=182707) --> (https://bugzilla.novell.com/attachment.cgi?id=182707) Output of hwinfo (while running kernel-default-2.6.23.1-12.x86_64) -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=302327#c23 Warren Stockton changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |NEW Info Provider|wns@comcast.net | --- Comment #23 from Warren Stockton 2007-11-09 09:51:11 MST --- Kernel 2.6.22.2-jeffm.test1 ran fine with over 9 hours of uptime. I am currently running 2.6.22.2-jeffm.e5e61a519e4f860c4595f966ecb897e5eb294d19 w/ acpi_use_timer_override and have over an hour of uptime, which (as per comment #18) didn't survive 5 minutes w/o acpi_use_timer_override. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.

https://bugzilla.novell.com/show_bug.cgi?id=302327#c25 --- Comment #25 from Warren Stockton 2007-11-09 15:02:20 MST --- kernel-default-2.6.22.12-0.1.x86_64.rpm from 10.3/update has now been running for 2 hours w/ acpi_use_timer_override. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.