qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: "Cédric Le Goater" <clg@redhat.com>
To: Avihai Horon <avihaih@nvidia.com>, qemu-devel@nongnu.org
Cc: Peter Xu <peterx@redhat.com>, Fabiano Rosas <farosas@suse.de>,
	Alex Williamson <alex.williamson@redhat.com>
Subject: Re: [PATCH 08/14] vfio: Use new Error** argument in vfio_save_setup()
Date: Mon, 12 Feb 2024 18:54:10 +0100	[thread overview]
Message-ID: <01f3e1e5-75e0-474f-91fc-27d20a4c96ed@redhat.com> (raw)
In-Reply-To: <589c8a79-4266-41c9-a019-ce1d55d7fab9@nvidia.com>

On 2/12/24 10:17, Avihai Horon wrote:
> Hi Cedric,
> 
> On 07/02/2024 15:33, Cédric Le Goater wrote:
>> External email: Use caution opening links or attachments
>>
>>
>> Add an Error** argument to vfio_migration_set_state() and adjust
>> callers, including vfio_save_setup(). The error will be propagated up
>> to qemu_savevm_state_setup() where the save_setup() handler is
>> executed.
>>
>> Signed-off-by: Cédric Le Goater <clg@redhat.com>
>> ---
>>   hw/vfio/migration.c | 62 +++++++++++++++++++++++++++++----------------
>>   1 file changed, 40 insertions(+), 22 deletions(-)
>>
>> diff --git a/hw/vfio/migration.c b/hw/vfio/migration.c
>> index 2dfbe671f6f45aa530c7341177bb532d8292cecd..2e0a79967cc97f44d9be5575c3cfe18c9f349dab 100644
>> --- a/hw/vfio/migration.c
>> +++ b/hw/vfio/migration.c
>> @@ -84,7 +84,8 @@ static const char *mig_state_to_str(enum vfio_device_mig_state state)
>>
>>   static int vfio_migration_set_state(VFIODevice *vbasedev,
>>                                       enum vfio_device_mig_state new_state,
>> -                                    enum vfio_device_mig_state recover_state)
>> +                                    enum vfio_device_mig_state recover_state,
>> +                                    Error **errp)
>>   {
>>       VFIOMigration *migration = vbasedev->migration;
>>       uint64_t buf[DIV_ROUND_UP(sizeof(struct vfio_device_feature) +
>> @@ -104,15 +105,15 @@ static int vfio_migration_set_state(VFIODevice *vbasedev,
>>           ret = -errno;
>>
>>           if (recover_state == VFIO_DEVICE_STATE_ERROR) {
>> -            error_report("%s: Failed setting device state to %s, err: %s. "
>> -                         "Recover state is ERROR. Resetting device",
>> -                         vbasedev->name, mig_state_to_str(new_state),
>> -                         strerror(errno));
>> +            error_setg(errp, "%s: Failed setting device state to %s, err: %s. "
>> +                       "Recover state is ERROR. Resetting device",
>> +                       vbasedev->name, mig_state_to_str(new_state),
>> +                       strerror(errno));
>>
>>               goto reset_device;
>>           }
>>
>> -        error_report(
>> +        error_setg(errp,
>>               "%s: Failed setting device state to %s, err: %s. Setting device in recover state %s",
>>                        vbasedev->name, mig_state_to_str(new_state),
>>                        strerror(errno), mig_state_to_str(recover_state));
>> @@ -120,7 +121,7 @@ static int vfio_migration_set_state(VFIODevice *vbasedev,
>>           mig_state->device_state = recover_state;
>>           if (ioctl(vbasedev->fd, VFIO_DEVICE_FEATURE, feature)) {
>>               ret = -errno;
>> -            error_report(
>> +            error_setg(errp,
>>                   "%s: Failed setting device in recover state, err: %s. Resetting device",
>>                            vbasedev->name, strerror(errno));
> 
> I think here we will assert because errp is already set.
> 
> Adding an error_append() API would be useful here I guess.

yes.

> Otherwise, we need to move the first error_setg() below, to before we return from a successful recover state change, and construct the error message differently (e.g., provide a full error message for the recover state fail case containing also the first error).
> 
> Do you have other ideas?

Errors for :

     if (ioctl(vbasedev->fd, VFIO_DEVICE_RESET)) {

should be treated as the others with and error_append() and not
hw_error(). This needs a rework before any new changes.

I also wonder why we have twice :

     migration->device_state = recover_state;

It looks redundant. The ioctl VFIO_DEVICE_FEATURE should leave the
state unmodified.

Thanks,

C.




  reply	other threads:[~2024-02-12 17:54 UTC|newest]

Thread overview: 65+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-02-07 13:33 [PATCH 00/14] migration: Improve error reporting Cédric Le Goater
2024-02-07 13:33 ` [PATCH 01/14] migration: Add Error** argument to .save_setup() handler Cédric Le Goater
2024-02-07 20:11   ` Philippe Mathieu-Daudé
2024-02-08 13:27     ` Cédric Le Goater
2024-02-08  4:26   ` Peter Xu
2024-02-12  8:36   ` Avihai Horon
2024-02-12 14:49     ` Cédric Le Goater
2024-02-12 15:57       ` Avihai Horon
2024-02-07 13:33 ` [PATCH 02/14] migration: Add Error** argument to .load_setup() handler Cédric Le Goater
2024-02-07 20:12   ` Philippe Mathieu-Daudé
2024-02-08  4:30   ` Peter Xu
2024-02-09  9:35     ` Cédric Le Goater
2024-02-07 13:33 ` [PATCH 03/14] memory: Add Error** argument to .log_global*() handlers Cédric Le Goater
2024-02-08  5:48   ` Peter Xu
2024-02-09 10:14     ` Cédric Le Goater
2024-02-12  8:43       ` Avihai Horon
2024-02-12 16:36         ` Cédric Le Goater
2024-02-08 15:59   ` Philippe Mathieu-Daudé
2024-02-07 13:33 ` [PATCH 04/14] migration: Modify ram_init_bitmaps() to report dirty tracking errors Cédric Le Goater
2024-02-07 20:15   ` Philippe Mathieu-Daudé
2024-02-12  8:51   ` Avihai Horon
2024-02-12 16:37     ` Cédric Le Goater
2024-02-07 13:33 ` [PATCH 05/14] vfio: Add Error** argument to .set_dirty_page_tracking() handler Cédric Le Goater
2024-02-07 20:16   ` Philippe Mathieu-Daudé
2024-02-07 13:33 ` [PATCH 06/14] vfio: Add Error** argument to vfio_devices_dma_logging_start() Cédric Le Goater
2024-02-07 20:17   ` Philippe Mathieu-Daudé
2024-02-07 13:33 ` [PATCH 07/14] vfio: Add Error** argument to vfio_devices_dma_logging_stop() Cédric Le Goater
2024-02-07 20:18   ` Philippe Mathieu-Daudé
2024-02-07 13:33 ` [PATCH 08/14] vfio: Use new Error** argument in vfio_save_setup() Cédric Le Goater
2024-02-07 20:21   ` Philippe Mathieu-Daudé
2024-02-12  9:17   ` Avihai Horon
2024-02-12 17:54     ` Cédric Le Goater [this message]
2024-02-13 13:57       ` Avihai Horon
2024-02-07 13:33 ` [PATCH 09/14] vfio: Add Error** argument to .vfio_save_config() handler Cédric Le Goater
2024-02-07 20:22   ` Philippe Mathieu-Daudé
2024-02-12  9:21   ` Avihai Horon
2024-02-07 13:33 ` [PATCH 10/14] vfio: Also trace event failures in vfio_save_complete_precopy() Cédric Le Goater
2024-02-07 13:33 ` [PATCH 11/14] vfio: Extend vfio_set_migration_error() with Error* argument Cédric Le Goater
2024-02-07 20:25   ` Philippe Mathieu-Daudé
2024-02-12  9:35   ` Avihai Horon
2024-02-16 13:12     ` Cédric Le Goater
2024-02-07 13:33 ` [PATCH 12/14] migration: Report error when shutdown fails Cédric Le Goater
2024-02-07 20:26   ` Philippe Mathieu-Daudé
2024-02-08  5:52   ` Peter Xu
2024-02-07 13:33 ` [PATCH 13/14] migration: Use migrate_has_error() in close_return_path_on_source() Cédric Le Goater
2024-02-08  5:52   ` Peter Xu
2024-02-08 13:07   ` Fabiano Rosas
2024-02-08 13:45     ` Cédric Le Goater
2024-02-08 13:57       ` Fabiano Rosas
2024-02-12 13:03         ` Cédric Le Goater
2024-02-14 16:00           ` Fabiano Rosas
2024-02-16 15:17             ` Cédric Le Goater
2024-02-23  4:14     ` Peter Xu
2024-02-07 13:33 ` [RFC PATCH 14/14] migration: Fix return-path thread exit Cédric Le Goater
2024-02-08  5:57   ` Peter Xu
2024-02-12 16:04     ` Cédric Le Goater
2024-02-23  4:25       ` Peter Xu
2024-02-08 13:29   ` Fabiano Rosas
2024-02-12 15:44     ` Cédric Le Goater
2024-02-14 20:35       ` Fabiano Rosas
2024-02-16 15:08         ` Cédric Le Goater
2024-02-16 17:35           ` Fabiano Rosas
2024-02-23  4:31             ` Peter Xu
2024-02-23 14:05               ` Fabiano Rosas
2024-02-26  8:44                 ` Cédric Le Goater

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=01f3e1e5-75e0-474f-91fc-27d20a4c96ed@redhat.com \
    --to=clg@redhat.com \
    --cc=alex.williamson@redhat.com \
    --cc=avihaih@nvidia.com \
    --cc=farosas@suse.de \
    --cc=peterx@redhat.com \
    --cc=qemu-devel@nongnu.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).