[Bug 943107] New: BUG: scheduling while atomic: swapper/1/0/0x00000145 (i586)
http://bugzilla.suse.com/show_bug.cgi?id=943107 Bug ID: 943107 Summary: BUG: scheduling while atomic: swapper/1/0/0x00000145 (i586) Classification: openSUSE Product: openSUSE Factory Version: 201505* Hardware: i586 OS: Other Status: NEW Severity: Normal Priority: P5 - None Component: Kernel Assignee: kernel-maintainers@forge.provo.novell.com Reporter: chrubis@suse.com QA Contact: qa-bugs@suse.de Found By: --- Blocker: --- I've installed Tumbleweed from openSUSE-Tumleweed-NET-i586-Snapshot20150821-Media.iso into KVM with 4CPUs and 2G RAM. The machine fails to boot after the installation and the log is filled with: BUG: scheduling while atomic: swapper/1/0/0x00000145 with the last number increasing by one in each call. sometimes it's interleaved with: rcu_preempt self-deteceted stall on CPU { 1} rcu_preempt kthread starved for 60002 jiffies! Kernel version is 4.1.6-1-desktop and it boots fine when I reduce the number of CPUs to 1. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=943107 http://bugzilla.suse.com/show_bug.cgi?id=943107#c1 Takashi Iwai <tiwai@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |chrubis@suse.com, | |tiwai@suse.com Flags| |needinfo?(chrubis@suse.com) --- Comment #1 from Takashi Iwai <tiwai@suse.com> --- Can't reproduced here. I freshly installed the given Tumbleweed image (minimal text installation) on KVM with -smp 4 -m 2024, but it works as expected. What's the exact reproducer? -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=943107 Martin Pluskal <mpluskal@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |mpluskal@suse.com -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=943107 http://bugzilla.suse.com/show_bug.cgi?id=943107#c2 Cyril Hrubis <chrubis@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Flags|needinfo?(chrubis@suse.com) | --- Comment #2 from Cyril Hrubis <chrubis@suse.com> --- The whole command line for qemu-system-x86_64 was "-enable-kvm -smp 4 -m 2G" + path to hda image and the cdrom image. The qemu version is 2.3.0-r5. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=943107 http://bugzilla.suse.com/show_bug.cgi?id=943107#c3 --- Comment #3 from Takashi Iwai <tiwai@suse.com> --- So there is nothing special there. I'm using virtio-scsi-pci, and -soundhw hdaudio -vga cirrus -net nic -net user, but these shouldn't matter... Does this happen on different machines you all tested? -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=943107 http://bugzilla.suse.com/show_bug.cgi?id=943107#c4 Cyril Hrubis <chrubis@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Flags|needinfo?(chrubis@suse.com) | --- Comment #4 from Cyril Hrubis <chrubis@suse.com> --- Hmm, looks like I can boot the image just fine on different machine (has intel cpu not AMD, different distribution, etc.). So I may have been just unlucky. I will try to debug this further. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=943107 http://bugzilla.suse.com/show_bug.cgi?id=943107#c7 Takashi Iwai <tiwai@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Flags| |needinfo?(chrubis@suse.com) --- Comment #7 from Takashi Iwai <tiwai@suse.com> --- Could you check whether the issue persists with the latest TW? -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=943107 http://bugzilla.suse.com/show_bug.cgi?id=943107#c8 Cyril Hrubis <chrubis@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Flags|needinfo?(chrubis@suse.com) | --- Comment #8 from Cyril Hrubis <chrubis@suse.com> --- I've tried with Snapshot20151128 and everything works fine now. Greping for BUG in dmesg showed nothing. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=943107 http://bugzilla.suse.com/show_bug.cgi?id=943107#c9 --- Comment #9 from Cyril Hrubis <chrubis@suse.com> --- (In reply to Cyril Hrubis from comment #8)
I've tried with Snapshot20151128 and everything works fine now. Greping for BUG in dmesg showed nothing.
Ah, forget that, that was x86_64 installation. Will try x86 now. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=943107 http://bugzilla.suse.com/show_bug.cgi?id=943107#c10 --- Comment #10 from Cyril Hrubis <chrubis@suse.com> --- The Snapshot20151128 i586 seems to be broken the same way. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=943107 http://bugzilla.suse.com/show_bug.cgi?id=943107#c12 Cyril Hrubis <chrubis@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Flags|needinfo?(chrubis@suse.com) | --- Comment #12 from Cyril Hrubis <chrubis@suse.com> --- Created attachment 658031 --> http://bugzilla.suse.com/attachment.cgi?id=658031&action=edit Boot console log Boot log attached. The interesting part seems to be: [ 0.008000] ------------[ cut here ]------------^M [ 0.008000] WARNING: CPU: 1 PID: 0 at ../arch/x86/kernel/cpu/amd.c:265 init_amd+0x912/0xa80() [ 0.008000] WARNING: This combination of AMD processors is not suitable for SMP. [ 0.008000] Modules linked in: [ 0.008000] CPU: 1 PID: 0 Comm: swapper/1 Not tainted 4.3.0-2-pae #1 [ 0.008000] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.8.2-0-g33fbe13 by qemu-project.org 04/01/2014 [ 0.008000] 00000000 00000000 f4d3be94 c0528adf f4d3bed4 f4d3bec4 c0257f47 c09ad15c [ 0.008000] f4d3bef0 00000000 c09b6e70 00000109 c02160d2 c02160d2 f5155040 f51550b9 [ 0.008000] f4d3bf10 f4d3bedc c0257fb3 00000009 f4d3bed4 c09ad15c f4d3bef0 f4d3bf3c [ 0.008000] Call Trace: [ 0.008000] [<c0206432>] try_stack_unwind+0x132/0x140 [ 0.008000] [<c0205495>] dump_trace+0x65/0x230 [ 0.008000] [<c0206486>] show_trace_log_lvl+0x46/0x50 [ 0.008000] [<c020572b>] show_stack_log_lvl+0xcb/0x110 [ 0.008000] [<c0206516>] show_stack+0x26/0x50 [ 0.008000] [<c0528adf>] dump_stack+0x48/0x69 [ 0.008000] [<c0257f47>] warn_slowpath_common+0x87/0xc0 [ 0.008000] [<c0257fb3>] warn_slowpath_fmt+0x33/0x40 [ 0.008000] [<c02160d2>] init_amd+0x912/0xa80 [ 0.008000] [<c02135ed>] identify_cpu+0x11d/0x4c0 [ 0.008000] [<c0213a2f>] identify_secondary_cpu+0xf/0x30 [ 0.008000] [<c0238302>] smp_store_cpu_info+0x32/0x40 [ 0.008000] [<c02387ac>] start_secondary+0x4c/0x1a0 [ 0.008000] ---[ end trace 09a0667241288c54 ]--- -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=943107 http://bugzilla.suse.com/show_bug.cgi?id=943107#c14 Cyril Hrubis <chrubis@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Flags|needinfo?(chrubis@suse.com) | --- Comment #14 from Cyril Hrubis <chrubis@suse.com> --- Created attachment 658082 --> http://bugzilla.suse.com/attachment.cgi?id=658082&action=edit Boot console log for lockdep kernel Boot log from the lockdep kernel. The virtual machine ends up frozen on "Loading initial ramdisk ..." -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=943107 http://bugzilla.suse.com/show_bug.cgi?id=943107#c17 --- Comment #17 from Cyril Hrubis <chrubis@suse.com> --- (In reply to Takashi Iwai from comment #16)
I just wonder whether it's an issue rather in QEMU/KVM side. Changing the CPU, e.g. passing "-cpu host" option makes any difference?
Hmm, with -cpu host ("AMD Phenom(tm) 9750 Quad-Core Processor (fam: 10, model: 02, stepping: 03)") the older image (Snapshot20150821 kernel 4.1.6-desktop) works fine. The problem occurs there only with the "AMD QEMU Virtual CPU version 2.4.0 (fam: 06, model: 06, stepping: 03)" And the newer image (Snapshot20151128) with the lockdep kernel 4.3.0-30 fails just the same way even with -cpu host. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=943107 http://bugzilla.suse.com/show_bug.cgi?id=943107#c23 --- Comment #23 from Cyril Hrubis <chrubis@suse.com> --- Sorry for the delay, I'm back from vacation now. Looks like the virtual machine I've installed previously does not boot at all with the debug kernel I've installed there. I will try to do a new Tumbleweed installation. Have the fix made it to the Tumbleweed snapshot meanwhile? -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=943107 http://bugzilla.suse.com/show_bug.cgi?id=943107#c24 Cyril Hrubis <chrubis@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Flags|needinfo?(chrubis@suse.com) | --- Comment #24 from Cyril Hrubis <chrubis@suse.com> --- And I've just did a network Tumbleweed installation which ended up with kernel 4.3.3-3-pae which works like a charm. -- You are receiving this mail because: You are on the CC list for the bug.
participants (1)
-
bugzilla_noreply＠novell.com