xen-devel.lists.xenproject.org archive mirror
 help / color / mirror / Atom feed
From: Wei Liu <Wei.Liu2@citrix.com>
To: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: wei.liu2@citrix.com, "xen-devel@lists.xen.org" <xen-devel@lists.xen.org>
Subject: Re: HVM bug: system crashes after offline online a vcpu
Date: Wed, 19 Dec 2012 16:18:22 +0000	[thread overview]
Message-ID: <1355933902.10526.27.camel@iceland> (raw)
In-Reply-To: <20121219160455.GA12077@phenom.dumpdata.com>

On Wed, 2012-12-19 at 16:04 +0000, Konrad Rzeszutek Wilk wrote:
> On Thu, Dec 13, 2012 at 03:12:17PM +0000, Wei Liu wrote:
> > Hi Konrad
> > 
> > I encountered a bug when trying to bring offline a cpu then online it
> > again in HVM. As I'm not very familiar with HVM stuffs I cannot come up
> > with a quick fix.
> 
> I took your two patches that you posted and they are in v3.8 now.
> 
> It seems that there are bugs in the offline/online code thought.
> 
> I did this:
> # echo 0 > /sys/devices/system/cpu/cpu3/online
> # echo 1 > /sys/devices/system/cpu/cpu3/online
> 
> With a PV guest and it blows up (with or without your patches).
> 
> Have you seen something similar to this:
> 
> [  106.166795] BUG: scheduling while atomic: swapper/2/0/0x00000000
> [  106.167168] microcode: CPU2 sig=0x206a7, pf=0x2, revision=0x17
> [  106.167566] Modules linked in: sg sd_mod dm_multipath dm_mod xen_evtchn iscsi_boot_sysfs iscsi_tcp libiscsi_tcp libiscsi scsi_transport_iscsi scsi_mod libcrc32c crc32c radeon fbcon tileblit font bitblit softcursor ttm drm_kms_helper crc32c_intel xen_blkfront xen_netfront xen_fbfront fb_sys_fops sysimgblt sysfillrect syscopyarea xen_kbdfront xenfs xen_privcmd [last unloaded: dump_dma]
> [  106.169286] Pid: 0, comm: swapper/2 Tainted: G           O 3.5.0-rc3upstream-00139-gb1849b3-dirty #1
> [  106.170152] Call Trace:
> [  106.170598]  [<ffffffff8109bcbd>] __schedule_bug+0x4d/0x60
> [  106.171042]  [<ffffffff815be0fc>] __schedule+0x69c/0x760
> [  106.171469]  [<ffffffff815be284>] schedule+0x24/0x70
> [  106.171890]  [<ffffffff8103fbe9>] cpu_idle+0xc9/0xe0
> [  106.172309]  [<ffffffff81033e79>] ? xen_irq_enable_direct_reloc+0x4/0x4
> [  106.172726]  [<ffffffff815b1c5d>] cpu_bringup_and_idle+0xe/0x10
> [  106.174533] BUG: scheduling while atomic: swapper/2/0/0x00000000
> ?
> 

IIRC I didn't see this. I was using your xen.git kernel tree, not the
upstream one. The PV path was fixed after applying my patch and the HVM
path I didn't have much idea. I didn't play much with online/offline
after fixing what I could fix.

Let me play with upstream kernel and give you some feedback.


Wei

  reply	other threads:[~2012-12-19 16:18 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-12-13 15:12 HVM bug: system crashes after offline online a vcpu Wei Liu
2012-12-19 16:04 ` Konrad Rzeszutek Wilk
2012-12-19 16:18   ` Wei Liu [this message]
2012-12-19 17:01   ` Wei Liu
2012-12-19 17:40     ` Konrad Rzeszutek Wilk
2012-12-19 17:52       ` Wei Liu
2012-12-19 19:24         ` Konrad Rzeszutek Wilk
2012-12-20 17:48           ` Wei Liu

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1355933902.10526.27.camel@iceland \
    --to=wei.liu2@citrix.com \
    --cc=konrad.wilk@oracle.com \
    --cc=xen-devel@lists.xen.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).