Package: release.debian.org Severity: normal Tags: stretch User: release.debian....@packages.debian.org Usertags: pu
Hi stable release managers, The last linux upload via the stretch point release unfortunately uncovered #928125, easily triggerble when forcing the kernel to scan the partition table on a newly created loop device. In the upstream development of the 4.9 series between the 4.9.140 upload in the previous stretch point release, some backports to changes related to the loopback were again later reverted due to regressions. One commit though was keept up to 4.9.168, which relates to the thread in https://lore.kernel.org/stable/20190320125806.gd9...@quack2.suse.cz/ . I seek input/advice/opinion from you stable release managers: Do you want to issue a SUA and fast-track a fix for this issue via stretch-updates? For completeness, the isolated revert is at https://salsa.debian.org/kernel-team/linux/commit/5f66def673547b7d1e7a841937fbd0ab10a925a0 . If not, the issue would be fixed when rebasing to the 4.9.173 or later kernel as well, work for that is ongoin as well in https://salsa.debian.org/kernel-team/linux/merge_requests/141 . Ideally kernels which are pending for a point relase and which we upload in advance for the point release seem to not get as much wider testing as it would make such regressions reported after the point release less frequent, I do not have a good proposal for that though. Regards, Salvatore
