[10:58] <jtaylor> apw: hi, any news on the 4.8 lvm raid patch?
[11:05] <smb> jtaylor, last status I got is that he got something that seems to work but wanted to let upstream review for not breaking the case of having feature flags set. and also in case there are similar broken checks hidden in other parts of the code. Don't think there was feedback, yet. But not sure
[12:02] <arvind> ioria, from reading the https://wiki.ubuntu.com/KernelTeam/GitKernelBuild it seems like there are no debs created for linux-firmware. With the current linux-firmware installed will that cause complications. The wiki does not really discuss this it seems
[12:03] <arvind> s/complications./complications?/g
[12:20] <apw> arvind, no separate debs for firmware, firmware included in the kernel is in the linux-image files and overrides any in linux-firmware
[12:20] <apw> though in theory they are no overlapping as well
=== ghostcube_ is now known as ghostcube
[14:43] <apw> jtaylor, i think i have a fix for that, but i do want to make sure that i am not missing something before letting it go on data one might care about
[14:44] <apw> jtaylor, i have submitted it upstream to get some feedback
[14:44] <jtaylor> apw: can you link me the post?
[14:45] <apw> jtaylor, https://www.redhat.com/archives/dm-devel/2016-October/msg00161.html
[14:47] <jtaylor> thanks
=== mjeanson_ is now known as mjeanson
[14:48] <apw> jtaylor, it feels right to make that change, but one also has to have a healthy paranoia on his kind of thing -- you are raid'ing that data i assume because one wants it
[14:54] <apw> jtaylor, that initial feedback is looking promising
[14:54] <om26er> jsalisbury: any news from upstream on that regression ?
[14:55] <om26er> Seems it'll pass the 16.10 release timeline.
[14:57] <jtaylor> apw: patch looks reasonable too me
[14:57] <jtaylor> apw: but I am quite scared nobody noticed this before release, so this has never been tested on older raids
[14:59] <apw> jtaylor, that does seem to be the implication, i have tested with that applied and things seem to be unaffected.  luckily the new flags are mostly to do with reshaping support which is an opt-in exercise
[14:59] <jsalisbury> om26er, No update yet.  They want to collect some additional data, so I'll post the commands in the bug.
[14:59] <jtaylor> apw: probably, though there is some stuff changing the flag on super_sync
[15:00] <jtaylor> I wonder what happens when you update your superblocks and there is something else that is incompatible
[15:00] <om26er> jsalisbury: if its not catastrophic maybe we can distro patch that before release and then wait for upstream to fix the issue ? I fear there are going to be lots of people with similar issue.
[15:00] <jtaylor> I better make sure my backups are in order before upgrading :)
[15:01] <apw> jtaylor, right now i don't think there are any incompatible options so that can't happen, but if there were, it is suspect
[15:03] <jsalisbury> om26er, That might be an option.  For some reason, others with similar hardware cannot reproduce the issue, so upstream is digging deeper
[15:03] <apw> jtaylor, yeah i am just doing mine too :
[15:03] <apw> :)
[15:03] <om26er> hmm