Re: [Xen-devel] [linux-linus test] 25808: regressions - FAIL

[Konrad Rzeszutek Wilk <konrad.wilk@xxxxxxxxxx>]

On Apr 9, 2014 4:45 AM, Ian Campbell <Ian.Campbell@xxxxxxxxxx> wrote:
>
> On Tue, 2014-04-08 at 21:32 +0100, xen.org wrote: 
> > flight 25808 linux-linus real [real] 
> > http://www.chiark.greenend.org.uk/~xensrcts/logs/25808/ 
> > 
> > Regressions :-( 
> > 
> > Tests which did not succeed and are blocking, 
> > including tests which could not be run: 
> > test-amd64-i386-rhel6hvm-amd 5 xen-boot fail REGR. vs. 
> >12557 
> > test-amd64-i386-xl-multivcpu 5 xen-boot fail REGR. vs. 
> >12557 
> > test-amd64-i386-qemuu-rhel6hvm-amd 5 xen-boot fail REGR. vs. 
> >12557 
> >Â test-amd64-i386-xlÂÂÂÂÂÂÂÂÂÂÂ 5 xen-bootÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ fail REGR. vs. 
> >12557 
> >Â test-amd64-i386-xl-credit2ÂÂÂ 5 xen-bootÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ fail REGR. vs. 
> >12557 
> >Â test-amd64-i386-xl-winxpsp3-vcpus1Â 5 xen-bootÂÂÂÂÂÂÂÂÂÂÂ fail REGR. vs. 
> >12557 
> >Â test-amd64-i386-pairÂÂÂÂÂÂÂÂÂ 8 xen-boot/dst_hostÂÂÂÂÂÂÂÂ fail REGR. vs. 
> >12557 
> >Â test-amd64-i386-pairÂÂÂÂÂÂÂÂÂ 7 xen-boot/src_hostÂÂÂÂÂÂÂÂ fail REGR. vs. 
> >12557 
>
> Looks like we have some regressions here, different on arm and x86 
> AFAICT. 
>
> >version targeted for testing: 
> >Â linuxÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ a7963eb7f4c4b5df84d5dd5083734278ad75bafb 
> > baseline version: 
> >Â linuxÂÂÂÂÂÂÂÂÂÂÂÂÂÂÂ c16fa4f2ad19908a47c63d8fa436a1178438c7e7 
>
> For x86 I can see lots of 
> Apr 8 14:40:28.430437 [ 3.677587] Initializing CPU#1 
> Apr 8 14:40:28.430465 [ 3.677718] ------------[ cut here ]------------ 
> Apr 8 14:40:28.438410 [ 3.677729] WARNING: CPU: 1 PID: 0 at 
> kernel/time/tick-sched.c:789 tick_nohz_idle_enter+0x82/0x90() 
> Apr 8 14:40:28.450407 [ 3.677734] Modules linked in: 
> Apr 8 14:40:28.450436 [ 3.677738] CPU: 1 PID: 0 Comm: swapper/1 Not 
> tainted 3.14.0+ #1 
> Apr 8 14:40:28.458408 [ 3.677742] Hardware name: Supermicro 
> X9SCL/X9SCM/X9SCL/X9SCM, BIOS 2.0b 09/17/2012 
> Apr 8 14:40:28.458458 [ 3.677746] 00000000 00000000 db4b9f2c c169f2b7 
> 00000000 db4b9f5c c109919f c17f8038 
> Apr 8 14:40:28.470427 [ 3.677753] 00000001 00000000 c18002a6 00000315 
> c10f4a12 c10f4a12 c18b1fec 00000000 
> Apr 8 14:40:28.478421 [ 3.677760] 00000000 db4b9f6c c10991dd 00000009 
> 00000000 db4b9f74 c10f4a12 db4b9fa8 
> Apr 8 14:40:28.490412 [ 3.677768] Call Trace: 
> Apr 8 14:40:28.490440 [ 3.677773] [<c169f2b7>] dump_stack+0x41/0x52 
> Apr 8 14:40:28.490477 [ 3.677777] [<c109919f>] 
> warn_slowpath_common+0x7f/0xa0 
> Apr 8 14:40:28.498416 [ 3.677780] installing Xen timer for CPU 2 
> Apr 8 14:40:28.498449 [ 3.677785] [<c10f4a12>] ? 
> tick_nohz_idle_enter+0x82/0x90 
> Apr 8 14:40:28.510411 [ 3.677786] cpu 2 spinlock event irq 55 
> Apr 8 14:40:28.510443 [ 3.677791] [<c10f4a12>] ? 
> tick_nohz_idle_enter+0x82/0x90 
> Apr 8 14:40:28.518406 [ 3.677796] [<c10991dd>] 
> warn_slowpath_null+0x1d/0x20 
> Apr 8 14:40:28.518441 [ 3.677800] [<c10f4a12>] 
> tick_nohz_idle_enter+0x82/0x90 
> Apr 8 14:40:28.530415 [ 3.677805] [<c10d28e7>] 
> cpu_startup_entry+0x27/0x3c0 
> Apr 8 14:40:28.530451 [ 3.677810] [<c104e8b0>] ? check_events+0x8/0xc 
> Apr 8 14:40:28.538415 [ 3.677816] [<c1053c02>] 
> cpu_bringup_and_idle+0x12/0x20 
>

I believe Boris is looking at a fix for that. The same commit that triggers 
this causes the xen_iret to die.

CCing Boris here.
> (see e.g. 
> http://www.chiark.greenend.org.uk/~xensrcts/logs/25808/test-amd64-i386-xl/serial-grain-weevil.log
>  but I saw it in all of the handful of logs which I looked at. 
>
> On arm 
> http://www.chiark.greenend.org.uk/~xensrcts/logs/25808/test-armhf-armhf-xl/serial-marilith-n5.txt
>  shows: 
> [Tue Apr 8 12:31:21 2014][ 3.219710] Internal error: Oops - undefined 
> instruction: 0 [#1] SMP ARM 
> [Tue Apr 8 12:31:42 2014][ 3.226318] Unable to handle kernel NULL pointer 
> dereference at virtual address 0000017c 
> [Tue Apr 8 12:31:42 2014][ 3.226320] pgd = c0204000 
> [Tue Apr 8 12:31:42 2014][ 3.226324] [0000017c] *pgd=00000000 
>
> Then a gap and: 
> [Tue Apr 8 12:31:42 2014][ 24.230490] INFO: rcu_sched detected stalls on 
> CPUs/tasks: { 1} (detected by 2, t=2102 jiffies, g=4294967023, c=4294967022, 
> q=127) 
>
> which is just the Oops leading to a stall. Not a lot to go on, not sure 
> why there isn't even a trace, or even a pc value :-( 
>
> Ian. 
>
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel