Re: [Qemu-block] [PATCH 1/3] migration: Call blk_resume_after_migration() for postcopy

[Kevin Wolf]

Am 13.04.2017 um 20:03 hat Eric Blake geschrieben:
> On 04/13/2017 12:54 PM, Kevin Wolf wrote:
> > Am 13.04.2017 um 19:39 hat Eric Blake geschrieben:
> >> On 04/13/2017 12:23 PM, Kevin Wolf wrote:
> >>> Commit d35ff5e6 ('block: Ignore guest dev permissions during incoming
> >>> migration') added blk_resume_after_migration() to the precopy migration
> >>> path, but neglected to add it to the duplicated code that is used for
> >>> postcopy migration. This means that the guest device doesn't request the
> >>> necessary permissions, which ultimately led to failing assertions.
> >>>
> >>> Add the missing blk_resume_after_migration() to the postcopy path.
> >>>
> >>> Signed-off-by: Kevin Wolf <address@hidden>
> >>> ---
> >>>  migration/savevm.c | 8 ++++++++
> >>>  1 file changed, 8 insertions(+)
> >>
> >> Reviewed-by: Eric Blake <address@hidden>
> >>
> >> Are we targetting this for 2.9-rc5, or is it 2.10 material?
> > 
> > At this point, I think it's clearly 2.10.
> 
> Okay.  Restating, to make sure I got your reasoning: the removed
> assertions of commit e3e0003 imply that 2.9 is not regressing in
> behavior, and at this point the worst the code can do without this patch
> applied is behave like it's done previously; therefore this patch is not
> fixing an observable 2.9 behavior and therefore not worth holding up the
> release.

Right, basically the new op blockers become ineffective for guest
devices after postcopy migration.

> But for 2.10, it's absolutely essential, as we have another patch
> pending to revert e3e0003 at which point we have a behavior break
> without this patch.

Correct.

Kevin

pgpbItRy6eLoF.pgp
Description: PGP signature