On 2/12/19 3:37 PM, John Snow wrote: > > > On 2/12/19 3:16 PM, Eric Blake wrote: >> On 2/12/19 2:07 PM, John Snow wrote: >>> When bitmaps are persistent, they may incur a disk read or write when >>> bitmaps >>> are added or removed. For configurations like virtio-dataplane, failing to >>> acquire this lock will abort QEMU when disk IO occurs. >>> >>> We used to acquire aio_context as part of the bitmap lookup, so re-introduce >>> the lock for just the cases that have an IO penalty. >> >> It would be nice to call out which commit id dropped the aio_context >> acquisition during bitmap lookup (making it easier to analyze how long >> this has broken, and which downstream builds need the backport. >> > > OK, I will amend this. > > Looks like: > > commit 2119882c7eb7e2c612b24fc0c8d86f5887d6f1c3 > Author: Paolo Bonzini <pbonz...@redhat.com> > Date: Mon Jun 5 14:39:03 2017 +0200 > > since 2.10.
Hmm. block-dirty-bitmap-add's "persistent":true parameter was also added in 2.10 in commit fd5ae4cc. In fact, 2119882c was made at a time when there were not persistent bitmaps; so I guess that this means we have always been broken since fd5ae4cc. -- Eric Blake, Principal Software Engineer Red Hat, Inc. +1-919-301-3226 Virtualization: qemu.org | libvirt.org
signature.asc
Description: OpenPGP digital signature