/srv/irclogs.ubuntu.com/2016/07/12/#ubuntu-kernel.txt

=== JanC is now known as Guest37740
=== JanC_ is now known as JanC
mowthegrassHi There - Anyone facing issues with 3.19 /4.2 kernel, All DELL servers running on 3.19/4.2 kernel on reboot does not boot and gets stuck with message random: LVM unrandom read with 33 bits of entropy available 09:37
mowthegrassis it a BUG 09:37
apwmowthegrass, i'd say it is likely a bug indeed.  that last message doesn't look to be anything other than a warning about early use of entropy before it is really very random; from my reading it is not a blocker09:44
apwmowthegrass, well it _might_ indicate we are triggering transfer of ral randomness into urandom, so it might indicate a block, very hard to be sure09:46
apwmowthegrass, do the machines come out of it after some time? 09:46
apwmowthegrass, but either way, you should file a bug "ubuntu-bug linux" and add the details09:47
mowthegrassapw:No they dont come up 10:22
apwmowthegrass, do they come up from cold, or do you have to boot an older kernel10:23
mowthegrassit just drops to switched clocksource tsc10:23
=== ogra_ is now known as ogra
mowthegrassapw,Thanks on your input 12:33
fg_afaict, xenial's kernel is not affected by CVE-2016-6187 (the apparmor setprocattr oops) because the causing commit bb646cdb12e75d82258c2f2e7746d5952d3e321a is not contained in the 4.4 stable kernels - is that correct?13:14
argeshttps://people.canonical.com/~ubuntu-security/cve/2016/CVE-2016-6187.html looks like it13:36
fg_arges: thanks, I only looked at / knew about http://kernel.ubuntu.com/reports/kernel-cves.html (which does not list CVE's that don't affect)13:39
fg_arges: although your link also says "needs-triage" for some of the entries (e.g., source: linux) ;)13:43
argestyhicks: ^^^13:45
tyhicksfg_: hello14:40
tyhicksfg_: that's correct (xenial is not affected)14:40
fg_tyhicks: thanks!14:41
tyhicksfg_: since I populated the break-fix line with the git hashes of the commit that introduced the flaw and the commit that fixed the flaw, a bot will come through and update the status of all the kernels14:41
tyhicksfg_: I left it as 'needs-triage' intentionally because of that14:42
fg_tyhicks: okay - that sounds reasonable ;) thanks for clearing it up!14:49
tyhicksno problem :)14:53
dmj_s76is it still possible to get a patch for a nasty bug in the nouveau driver?15:56
dmj_s76in time for the kernel cycle on the 16.04.1 iso?15:57
apwdmj_s76, is it in the kernel itself ?  the last kernel has been uploaded (in theory)15:57
apwdmj_s76, though depending on severity and recoverablility, you never know15:58
dmj_s76apw: https://git.launchpad.net/~ubuntu-kernel/ubuntu/+source/linux/tree/drivers/gpu/drm/nouveau/nouveau_drm.c15:58
apwdmj_s76, what is the bug number15:58
dmj_s76apw: It prevents installation and booting until you install the nvidia proprietary drivers with Pascal (There is a workaround, but it involves the user knowing about modelines and vga modes)15:59
dmj_s76apw, looking that up now16:00
=== infinity_ is now known as infinity
dmj_s76apw: couldn't find a bug report so filed one: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/160234016:30
ubot5Launchpad bug 1602340 in linux (Ubuntu) "nouveau: boot hangs at blank screen with unsupported graphics cards" [Undecided,Incomplete]16:30
apwdmj_s76, and is there a fix yet ?16:30
apwdmj_s76, i assume the code thinks it supports those cards16:31
dmj_s76apw: I believe so.  Ben Skeggs worked on one yesterday, which I'm planning to test today.16:31
apwdmj_s76, could you add any information you have on the fix to the bug as well pls16:31
dmj_s76apw: https://github.com/skeggsb/nouveau/commit/11fcd1624b0a1c73fe3b2fa15c3cc45816da047016:32
dmj_s76apw: added16:35
apwdmj_s76, thanks ...16:38
apwdmj_s76, and ... keep that bug updated with your testing too pls16:39
dmj_s76Will do, really hoping to get this in the iso, since it's an issue that will make most users consider Ubuntu uninstallable if they have a recent GPU.16:42
=== JanC is now known as Guest1477
=== JanC_ is now known as JanC
apwdmj_s76, i assume you have a range of affected and non-affected GPUs you could test this on to make sure it does not regress older kit17:00
apwdmj_s76, would a test kernel help your testing ?17:02
dmj_s76apw: Yes, we have 9 series and 7 series to test for regression17:13
apwdmj_s76, cool, could you document that in the bug as well17:13
kamaldmj_s76, I've built a Xenial test kernel with the nouveau patch applied: http://people.canonical.com/~kamal/lp1602340/ ... does that advance your testing efforts?17:42
dmj_s76apw: Thanks for the build.  I've tested the patched kernel on our hardware and it allows clean installs to boot into unity without trouble.20:16
dmj_s76No regressions on the older nvidia hardware we have.20:16
apwkamal, ^21:51
dmj_s76kamal: thanks for the build...fixes boot issues with current gen nvidia gpus nicely22:08
dmj_s76If we can get that into the iso, it'll save quite a bit of systems from being effectively uninstallable.22:09
kamaldmj_s76, yes, saw your feedback, thanks -- we are indeed going to get it into the Xenial ISO for ya22:11
dmj_s76kamal: Thanks...that'll really help make the Ubuntu install experience better.22:22
kamaldmj_s76, we agree!  thanks very much for your work on this.22:24

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!