mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Suresh Gumpula <suresh.gumpula@nutanix.com>
To: Paolo Bonzini <pbonzini@redhat.com>,
	"kvm@vger.kernel.org" <kvm@vger.kernel.org>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: Re: ata driver loading hang on qemu/kvm intel
Date: Mon, 6 Apr 2020 15:30:36 +0000	[thread overview]
Message-ID: <D7D964C2-DD4B-4F17-BA3D-C45C992A4B15@nutanix.com> (raw)
In-Reply-To: <56486177-b629-081e-2785-b6e2ca626e88@redhat.com>

Hi Paolo,

Thanks much for the quick response.
We are not running nested guest here , just enabled in the host, so the host is capable of running nested, but we are not running nested guests.

The guest kernel(not a nested guest) boot iso. i.e its regular VM on a host is hanging with following errors.
Its consistently reproducible with some load on the host.

Guest serial log:
[    1.414035] Write protecting the kernel read-only data: 6144k
[    1.418006] Freeing unused kernel memory: 1080K
[    1.423033] Freeing unused kernel memory: 1004K
[    1.466783] scsi host0: ata_piix
[    1.469539] scsi host1: ata_piix
[    1.472039] ata1: PATA max MWDMA2 cmd 0x1f0 ctl 0x3f6 bmdma 0xc300 irq 14
[    1.475740] ata2: PATA max MWDMA2 cmd 0x170 ctl 0x376 bmdma 0xc308 irq 15
 
Guest vcpu registers:
[root@NTNX-MADMAX01-A ~]# virsh qemu-monitor-command --domain 3014 --hmp --cmd info registers -a |egrep -i ‘eip|rip’
RIP=ffffffff842f9887 RFL=00000246 [---Z-P-] CPL=0 II=0 A20=1 SMM=0 HLT=1
EIP=000fd0f5 EFL=00000012 [----A--] CPL=0 II=0 A20=1 SMM=0 HLT=1
EIP=000fd0f5 EFL=00000012 [----A--] CPL=0 II=0 A20=1 SMM=0 HLT=1
EIP=000fd0f5 EFL=00000012 [----A--] CPL=0 II=0 A20=1 SMM=0 HLT=1
EIP=000fd0f5 EFL=00000012 [----A--] CPL=0 II=0 A20=1 SMM=0 HLT=1
EIP=000fd0f5 EFL=00000012 [----A--] CPL=0 II=0 A20=1 SMM=0 HLT=1
EIP=000fd0f5 EFL=00000012 [----A--] CPL=0 II=0 A20=1 SMM=0 HLT=1
EIP=000fd0f5 EFL=00000012 [----A--] CPL=0 II=0 A20=1 SMM=0 HLT=1
EIP=000fd0f5 EFL=00000012 [----A--] CPL=0 II=0 A20=1 SMM=0 HLT=1
EIP=000fd0f5 EFL=00000012 [----A--] CPL=0 II=0 A20=1 SMM=0 HLT=1
EIP=000fd0f5 EFL=00000012 [----A--] CPL=0 II=0 A20=1 SMM=0 HLT=1
EIP=000fd0f5 EFL=00000012 [----A--] CPL=0 II=0 A20=1 SMM=0 HLT=1
  

Qemu trace logs:
cd_read_sector 0.866 pid=68853 lba=0x60f
cd_read_sector_cb 89.730 pid=68853 lba=0x60f ret=0x0
ide_atapi_cmd_reply_end 1.084 pid=68853 s=0x55dd90cde0a8 tx_size=0x800 elem_tx_size=0x0 index=0x0
ide_atapi_cmd_reply_end_bcl 0.342 pid=68853 s=0x55dd90cde0a8 bcl=0x800
ide_atapi_cmd_reply_end_new 0.485 pid=68853 s=0x55dd90cde0a8 status=0x58
ide_atapi_cmd_reply_end 298.203 pid=68853 s=0x55dd90cde0a8 tx_size=0x0 elem_tx_size=0x0 index=0x800
ide_atapi_cmd_reply_end_eot 0.554 pid=68853 s=0x55dd90cde0a8 status=0x50
bmdma_read 1550089.160 pid=68853 addr=0x2 val=0x0
bmdma_read 9.003 pid=68853 addr=0x2 val=0x0
bmdma_read 512.186 pid=68853 addr=0x2 val=0x0
bmdma_write 7.776 pid=68853 addr=0x2 val=0x0
bmdma_read 26.277 pid=68853 addr=0x2 val=0x0
bmdma_write 7.294 pid=68853 addr=0x2 val=0x0
bmdma_read 13917.904 pid=68853 addr=0x2 val=0x0
bmdma_write 7.843 pid=68853 addr=0x2 val=0x0
bmdma_read 382.849 pid=68853 addr=0x2 val=0x0
bmdma_write 7.280 pid=68853 addr=0x2 val=0x0
 

All the guest vcpus are in halt state(hlt=1) and only one vcpu is in 64 bit and rest all in 32.
I am thinking probably guest missed an interrupt and waiting for ever.
Not sure if this is ide emulation issue in the qemu or a bug in the kvm module.


Thanks,
Suresh



On 4/6/20, 1:13 AM, "Paolo Bonzini" <pbonzini@redhat.com> wrote:

    On 05/04/20 01:36, Suresh Gumpula wrote:
    > Hi,
    > 
    > I am seeing this old problem with qemu 2.12/kernel 4.19.84(intel kvm
    > with nesting enabled) and guest kernel 4.10
    > 
    > https://urldefense.proofpoint.com/v2/url?u=https-3A__bugzilla.redhat.com_show-5Fbug.cgi-3Fid-3D1345964&d=DwIFaQ&c=s883GpUCOChKOHiocYtGcg&r=F-ygQ993NOjzeEnzE4m1YQM2dkkoF0uqOdC0cuSJSQc&m=37pz6pqZuAoO_2NHYfODlXoxXpQBcgBrh6axdah6aqc&s=T9ZcI9Rz-PSG6M0tDOf11DnKk_OTKXIAoxWPsKuzRVE&e= 
    > 
    > Could you please advise me on this?  I am seeing the exact same problem
    > where the ata driver load was stuck.
    > 
    > Really appreciate if you could give some pointers?
    > 
    > Is it a known issue with nesting? Or is this bug fixed in latest kvm module?
    
    Hi, that bug was specific to running under Ravello, not on bare metal.
    I have not received any other reports.
    
    To clarify you have:
    
    - QEMU 2.12 and kernel 4.19.x on the host
    
    - kernel 4.10.x on the guest
    
    What QEMU version is running in the guest and what kernel version in the
    nested guest?
    
    Paolo
    
    


  reply	other threads:[~2020-04-06 15:30 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <7C92AFF4-D479-4F80-8BED-6E9B226DFB72@nutanix.com>
2020-04-06  8:13 ` Paolo Bonzini
2020-04-06 15:30   ` Suresh Gumpula [this message]
2020-04-06 17:01     ` John Snow
2020-04-07  4:43       ` Suresh Gumpula

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=D7D964C2-DD4B-4F17-BA3D-C45C992A4B15@nutanix.com \
    --to=suresh.gumpula@nutanix.com \
    --cc=kvm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=pbonzini@redhat.com \
    /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

all inboxes | Powered by JetHome®