Bug#856337: systemd: please support kernel 4.4, or don't hardcode dm interface versions

2017-03-04 Thread Michael Biebl
I've uploaded test packages to https://people.debian.org/~biebl/systemd/stretch/ You can install them by first installing apt-transport-https, then adding to your sources.list: deb [trusted=yes] https://people.debian.org/~biebl/systemd/stretch/ ./ Please report back if those packages fix your

Bug#856337: systemd: please support kernel 4.4, or don't hardcode dm interface versions

2017-02-28 Thread Marc Lehmann
On Tue, Feb 28, 2017 at 06:35:16PM +0100, Michael Biebl wrote: > Can you elaborate where and how such data corruption can happen? > Not being able to shutdown/detach all DM devices has been the case for > basically forever in Debian (most prominent example if / is on LVM). > I've never seen data

Bug#856337: systemd: please support kernel 4.4, or don't hardcode dm interface versions

2017-02-28 Thread Marc Lehmann
On Tue, Feb 28, 2017 at 02:55:30PM +0100, Michael Biebl wrote: > >> Can you please elaborate why you filed this against systemd? > > > > Because I thought systemd is what cleans up dm-targets at shutdown. > > From your initial bug report, it was not clear where exactly you got > this message.

Bug#856337: systemd: please support kernel 4.4, or don't hardcode dm interface versions

2017-02-28 Thread Michael Biebl
Am 28.02.2017 um 02:19 schrieb Marc Lehmann: > I upgraded to stretch, using kernel 4.4.47. When rebooting, system could > not clean up device mapper targets, leading to a hard crash for all cache > targets, requiring many hours of heavy I/O after bootup and creating a > high-risk window for data co

Bug#856337: systemd: please support kernel 4.4, or don't hardcode dm interface versions

2017-02-28 Thread Michael Biebl
Control: forwarded -1 https://github.com/systemd/systemd/issues/5492 I've forwarded this upstream. Please followup there if you have further information. Thanks again Christian for your valuable input. -- Why is it that all of the instruments seeking intelligent life in the universe are pointe

Bug#856337: systemd: please support kernel 4.4, or don't hardcode dm interface versions

2017-02-28 Thread Christian Hofstaedtler
* Michael Biebl [170228 14:57]: > TBH, I find it rather broken having to embed a local copy of dm-ioctl.h > to work around this. My feeling as well. > Christian, this version mismatch, does that happen for minor version > differences as well, like say 4.10.0 vs 4.10.1 or only major versions > li

Bug#856337: systemd: please support kernel 4.4, or don't hardcode dm interface versions

2017-02-28 Thread Michael Biebl
Am 28.02.2017 um 14:27 schrieb Marc Lehmann: > On Tue, Feb 28, 2017 at 02:25:34AM +0100, Michael Biebl > wrote: >>> The specific error I got was something like (from memory): >>> >>>Could not detach DM dm-11: ioctl mismatch, kernel(4.34.4), user(4.35.4) >>> >> >> That doesn't look like a syst

Bug#856337: systemd: please support kernel 4.4, or don't hardcode dm interface versions

2017-02-28 Thread Marc Lehmann
On Tue, Feb 28, 2017 at 02:25:34AM +0100, Michael Biebl wrote: > > The specific error I got was something like (from memory): > > > >Could not detach DM dm-11: ioctl mismatch, kernel(4.34.4), user(4.35.4) > > > > That doesn't look like a systemd problem, systemd should work fine with > ker

Bug#856337: systemd: please support kernel 4.4, or don't hardcode dm interface versions

2017-02-28 Thread Marc Lehmann
On Tue, Feb 28, 2017 at 10:11:10AM +0100, Christian Hofstaedtler wrote: > > Alternatively, since lvm and dmsetup from stretch do not have any issues > > with kernel 4.4, I suspect systemd hardcodes version numbers in direct > > calls instead of going through e.g. libdevmapper - going through a li

Bug#856337: systemd: please support kernel 4.4, or don't hardcode dm interface versions

2017-02-28 Thread Christian Hofstaedtler
* Marc Lehmann [170228 02:21]: > Alternatively, since lvm and dmsetup from stretch do not have any issues > with kernel 4.4, I suspect systemd hardcodes version numbers in direct > calls instead of going through e.g. libdevmapper - going through a library > such as libdevmapper that hanmdles the k

Bug#856337: systemd: please support kernel 4.4, or don't hardcode dm interface versions

2017-02-27 Thread Michael Biebl
Am 28.02.2017 um 02:19 schrieb Marc Lehmann: > Package: systemd > Version: 232-18 > Severity: normal > > Dear Maintainer, > > I upgraded to stretch, using kernel 4.4.47. When rebooting, system could > not clean up device mapper targets, leading to a hard crash for all cache > targets, requiring m

Bug#856337: systemd: please support kernel 4.4, or don't hardcode dm interface versions

2017-02-27 Thread Marc Lehmann
Package: systemd Version: 232-18 Severity: normal Dear Maintainer, I upgraded to stretch, using kernel 4.4.47. When rebooting, system could not clean up device mapper targets, leading to a hard crash for all cache targets, requiring many hours of heavy I/O after bootup and creating a high-risk wi