qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Avihai Horon <avihaih@nvidia.com>
To: "Cédric Le Goater" <clg@redhat.com>, qemu-devel@nongnu.org
Cc: Alex Williamson <alex.williamson@redhat.com>,
	Paolo Bonzini <pbonzini@redhat.com>,
	Joao Martins <joao.m.martins@oracle.com>,
	Yishai Hadas <yishaih@nvidia.com>,
	Jason Gunthorpe <jgg@nvidia.com>,
	Maor Gottlieb <maorg@nvidia.com>,
	Kirti Wankhede <kwankhede@nvidia.com>,
	Tarun Gupta <targupta@nvidia.com>
Subject: Re: [PATCH for-8.2 1/6] vfio/migration: Move from STOP_COPY to STOP in vfio_save_cleanup()
Date: Mon, 31 Jul 2023 09:32:00 +0300	[thread overview]
Message-ID: <26daa2db-13fd-3705-fbbf-22b8643dc9ee@nvidia.com> (raw)
In-Reply-To: <dc2365df-4e71-e2d8-988d-f74431800123@redhat.com>


On 30/07/2023 19:25, Cédric Le Goater wrote:
> External email: Use caution opening links or attachments
>
>
> On 7/16/23 10:15, Avihai Horon wrote:
>> Changing the device state from STOP_COPY to STOP can take time as the
>> device may need to free resources and do other operations as part of the
>> transition. Currently, this is done in vfio_save_complete_precopy() and
>> therefore it is counted in the migration downtime.
>>
>> To avoid this, change the device state from STOP_COPY to STOP in
>> vfio_save_cleanup(), which is called after migration has completed and
>> thus is not part of migration downtime.
>>
>> Signed-off-by: Avihai Horon <avihaih@nvidia.com>
>
> Have you tried this series with vGPUs ? If so, are there any improvement
> to report ?
>
Not with a vGPU.
But I tried it with a ConnectX-7 VF and I could see up to 6% downtime 
improvement.
I mentioned this in the cover letter.
Do you want to mention it also in the commit message?

Thanks.

>
>> ---
>>   hw/vfio/migration.c | 19 +++++++++++++------
>>   1 file changed, 13 insertions(+), 6 deletions(-)
>>
>> diff --git a/hw/vfio/migration.c b/hw/vfio/migration.c
>> index 2674f4bc47..8acd182a8b 100644
>> --- a/hw/vfio/migration.c
>> +++ b/hw/vfio/migration.c
>> @@ -383,6 +383,19 @@ static void vfio_save_cleanup(void *opaque)
>>       VFIODevice *vbasedev = opaque;
>>       VFIOMigration *migration = vbasedev->migration;
>>
>> +    /*
>> +     * Changing device state from STOP_COPY to STOP can take time. 
>> Do it here,
>> +     * after migration has completed, so it won't increase downtime.
>> +     */
>> +    if (migration->device_state == VFIO_DEVICE_STATE_STOP_COPY) {
>> +        /*
>> +         * If setting the device in STOP state fails, the device 
>> should be
>> +         * reset. To do so, use ERROR state as a recover state.
>> +         */
>> +        vfio_migration_set_state(vbasedev, VFIO_DEVICE_STATE_STOP,
>> +                                 VFIO_DEVICE_STATE_ERROR);
>> +    }
>> +
>>       g_free(migration->data_buffer);
>>       migration->data_buffer = NULL;
>>       migration->precopy_init_size = 0;
>> @@ -508,12 +521,6 @@ static int vfio_save_complete_precopy(QEMUFile 
>> *f, void *opaque)
>>           return ret;
>>       }
>>
>> -    /*
>> -     * If setting the device in STOP state fails, the device should 
>> be reset.
>> -     * To do so, use ERROR state as a recover state.
>> -     */
>> -    ret = vfio_migration_set_state(vbasedev, VFIO_DEVICE_STATE_STOP,
>> -                                   VFIO_DEVICE_STATE_ERROR);
>>       trace_vfio_save_complete_precopy(vbasedev->name, ret);
>>
>>       return ret;
>


  reply	other threads:[~2023-07-31  6:37 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-07-16  8:15 [PATCH for-8.2 0/6] vfio/migration: Add P2P support for VFIO migration Avihai Horon
2023-07-16  8:15 ` [PATCH for-8.2 1/6] vfio/migration: Move from STOP_COPY to STOP in vfio_save_cleanup() Avihai Horon
2023-07-30 16:25   ` Cédric Le Goater
2023-07-31  6:32     ` Avihai Horon [this message]
2023-07-31  7:05       ` Cédric Le Goater
2023-07-16  8:15 ` [PATCH for-8.2 2/6] sysemu: Add pre VM state change callback Avihai Horon
2023-07-27 16:23   ` Cédric Le Goater
2023-07-30  6:31     ` Avihai Horon
2023-07-30 16:22       ` Cédric Le Goater
2023-07-16  8:15 ` [PATCH for-8.2 3/6] qdev: Add qdev_add_vm_change_state_handler_full() Avihai Horon
2023-07-16  8:15 ` [PATCH for-8.2 4/6] vfio/migration: Refactor PRE_COPY and RUNNING state checks Avihai Horon
2023-07-21 11:33   ` Cédric Le Goater
2023-07-16  8:15 ` [PATCH for-8.2 5/6] vfio/migration: Add P2P support for VFIO migration Avihai Horon
2023-07-21 11:48   ` Cédric Le Goater
2023-07-23 10:37     ` Avihai Horon
2023-07-16  8:15 ` [PATCH for-8.2 6/6] vfio/migration: Allow migration of multiple P2P supporting devices Avihai Horon
2023-07-21 12:09   ` Cédric Le Goater
2023-07-23 10:42     ` Avihai Horon
2023-07-18 15:46 ` [PATCH for-8.2 0/6] vfio/migration: Add P2P support for VFIO migration Jason Gunthorpe
2023-07-30  6:50   ` Avihai Horon

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=26daa2db-13fd-3705-fbbf-22b8643dc9ee@nvidia.com \
    --to=avihaih@nvidia.com \
    --cc=alex.williamson@redhat.com \
    --cc=clg@redhat.com \
    --cc=jgg@nvidia.com \
    --cc=joao.m.martins@oracle.com \
    --cc=kwankhede@nvidia.com \
    --cc=maorg@nvidia.com \
    --cc=pbonzini@redhat.com \
    --cc=qemu-devel@nongnu.org \
    --cc=targupta@nvidia.com \
    --cc=yishaih@nvidia.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;
as well as URLs for NNTP newsgroup(s).