[12:10] <LocutusOfBorg> hi folks, pretty please update virtualbox kernel modules for xenial?
[12:14] <rtg> LocutusOfBorg, it is currently at 5.0.16-dfsg-2 - Is there some serious bug ?
[12:15] <LocutusOfBorg> not sure
[12:15] <LocutusOfBorg> kernel fixes for 4.6
[12:16] <rtg> LocutusOfBorg, are you talking about the DKMS package ?
[12:16] <LocutusOfBorg> yes, and guest-dkms
[12:16] <LocutusOfBorg> uploading a new virtualbox now
[12:17] <rtg> ok - so the kernel team does not maintain that package. I'm mostly concerned about the in-kernel version.
[12:18] <LocutusOfBorg> oops
[12:18] <LocutusOfBorg> the dkms and guest-dkms are provided by me
[12:18] <LocutusOfBorg> I mean virtualbox-modules
[12:18] <LocutusOfBorg> and virtualbox-guest-modules
[12:29] <rtg> smb, re: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1570441/comments/3 - are you saying all of those patches are relevant and should be applied as well ?
[12:29] <ubot5`> Launchpad bug 1570441 in linux (Ubuntu Xenial) "Kernel Panic in Ubuntu 16.04 netboot installer" [Undecided,In progress]
[12:40] <smb> rtg, I was before I realized that somehow the patch which is claimed to break things is unlikely applied to Xenial
[12:40] <rtg> smb, we do have 'x86/topology: Create logical package id' in Xenial as of -18
[12:41] <smb> rtg, oh ok... was I looking at the wrong branch then? I did pull master this morning...
[12:41] <rtg> smb, always master-next
[12:42] <smb> rtg, Ok, sorry. So if we have that one we probably want the whole bunch of fixups
[12:42] <rtg> smb, it was applied for bug #1397880
[12:42] <smb> :/
[12:42] <ubot5`> bug 1397880 in intel "[Feature] Memory Bandwidth Monitoring" [Undecided,New] https://launchpad.net/bugs/1397880
[12:43] <rtg> looks like it was a scaffold patch
[12:43] <smb> If scaffold means pretty bad, yeah
[12:44] <smb> Only looking at the commit messages of the follow-up it feels like a case where they realized only in hindsight how much breakage this caused
[12:45] <rtg> smb, well, lemme look at those other patches as well
[12:47] <smb> rtg, The Xen one was the only one that did not have a fixes lin in the commit. But then reading through the description I had a feeling it might still be better to proactively apply it too
[12:48] <rtg> smb, or I could just try to revert that topology patch and fixup the memory bandwidth monitoring code.
[12:48] <smb> rtg, Hm, do you think that is possible?
[12:49] <rtg> smb, dunno, I'll try it out. Seems like a safer move at this point
[12:49] <smb> rtg, I agree. If the monitoring can be made working without this it seems safer. 
[12:54] <smb> rtg, amazing. for the amount of follow-ups this produced the change looks pretty harmless
[12:54] <rtg> smb, well, there is a fair bit of code movement in that patch
[12:55] <rtg> whoops, looking at the wrong one.
[12:55] <rtg> you are right, it is simple
[12:55] <smb> rtg, ah ok. I only saw additions
[12:56] <rtg> smb, it reverts cleanly, which means the compile will likely fail 'cause there is a missing function or header
[12:57] <smb> rtg, Right there was a new structure field and a bunch of functions in it
[12:59] <rtg> smb, topology_max_packages is missing. that might be simple enough.
[13:00] <smb> rtg, except if the monitoring is based on assumptions about available packages... 
[13:00] <rtg> just checking when that appeared
[13:02] <rtg> smb, drat, it all seems kind of interdependent
[13:02] <cking> that's a hairball
[13:04] <smb> Yeah I could imagine that bandwith monitoring is depening on some more specific topology info
[13:05] <smb> So lets try all currently known follow-ups and pray/hope for the best
[17:00] <manjo> apw, rtg, I am looking for an Ubuntu-4.5 tag on unstable tree ... and I am not seeing one.. if  I apply the 3 patches in http://kernel.ubuntu.com/~kernel-ppa/mainline/v4.5.1-wily/ will that be close enough to an Ubuntuized kernel ?