On Thu, Aug 03, 2023 at 09:38:20AM +0200, Thomas Huth wrote: > On 03/08/2023 09.10, Milan Zamazal wrote: > > Fabiano Rosas <[email protected]> writes: > > > > > Milan Zamazal <[email protected]> writes: > > > > > > > The QEMU CI fails in virtio-scmi test occasionally. As reported by > > > > Thomas Huth, this happens most likely when the system is loaded and it > > > > fails with the following error: > > > > > > > > qemu-system-aarch64: ../../devel/qemu/hw/pci/msix.c:659: > > > > msix_unset_vector_notifiers: Assertion > > > > `dev->msix_vector_use_notifier && dev->msix_vector_release_notifier' > > > > failed. > > > > ../../devel/qemu/tests/qtest/libqtest.c:200: kill_qemu() detected > > > > QEMU death from signal 6 (Aborted) (core dumped) > > > > > > > > As discovered by Fabiano Rosas, the cause is a duplicate invocation of > > > > msix_unset_vector_notifiers via duplicate vu_scmi_stop calls: > > > > > > > > msix_unset_vector_notifiers > > > > virtio_pci_set_guest_notifiers > > > > vu_scmi_stop > > > > vu_scmi_disconnect > > > > ... > > > > qemu_chr_write_buffer > > > > > > > > msix_unset_vector_notifiers > > > > virtio_pci_set_guest_notifiers > > > > vu_scmi_stop > > > > vu_scmi_set_status > > > > ... > > > > qemu_cleanup > > > > > > > > While vu_scmi_stop calls are protected by vhost_dev_is_started() > > > > check, it's apparently not enough. vhost-user-blk and vhost-user-gpio > > > > use an extra protection, see f5b22d06fb (vhost: recheck dev state in > > > > the vhost_migration_log routine) for the motivation. Let's use the > > > > same in vhost-user-scmi, which fixes the failure above. > > > > > > > > Fixes: a5dab090e142 ("hw/virtio: Add boilerplate for vhost-user-scmi > > > > device") > > > > Signed-off-by: Milan Zamazal <[email protected]> > > > > > > Reviewed-by: Fabiano Rosas <[email protected]> > > > > Please note that this bug fix should IMO definitely go to 8.1, to not > > have a bug in vhost-user-scmi and to not have broken tests. Any chance > > to get it merged? > > If nobody else is planning a pull request with this patch included, I can > take it for my next PR (since it is fixing the CI, I just saw another > failure here: > https://gitlab.com/qemu-project/qemu/-/jobs/4790457938#L4784 ) > > Thomas >
I picked it up but if you like I can drop it. -- MST
