sdeziel | while testing the retpoline enabled kernel ( linux-image-4.4.0-113-generic ) from xenial-proposed, I found a regression LP: #1748671 which I'd like to be fixed before this kernel hits -updates | 16:20 |
---|---|---|
ubot5 | Launchpad bug 1748671 in linux (Ubuntu) "BUG: unable to handle kernel NULL pointer dereference at 0000000000000009" [Undecided,Confirmed] https://launchpad.net/bugs/1748671 | 16:20 |
sdeziel | oh, I just noticed linux-image-4.4.0-115-generic in -proposed, I'll see if it helps with the above bug | 16:29 |
apw | sdeziel, do you know the last version that was not trivially reproduciblw | 16:32 |
sdeziel | apw: 4.4.0-112.135 didn't have the issue | 16:32 |
sdeziel | apw: it's what I had to roll back to | 16:33 |
sdeziel | apw: 4.4.0-115.139 is also affected | 17:09 |
apw | sdeziel, is that triggered by ipv6 in ipv4 | 17:11 |
sdeziel | apw: I don't know as the artful container tries to get both. I'll try to disable one and will report back. Unfortunately, the affected machine is this very laptop :( | 17:14 |
sdeziel | apw: I just re-read an you said "ipv6 in ipv4" which I'm not using. Did you mean ipv6 OR ipv4 like I understood the first time? | 17:18 |
apw | sdeziel, i had gotten the impression there was ipv6 in there | 17:18 |
apw | sdeziel, this is 100% reproducible for you i believe right? | 17:19 |
sdeziel | apw: yes, I'm assigning both v4 and v6 natively | 17:19 |
sdeziel | yes, 100% | 17:19 |
apw | sdeziel, as there is only one change in 112->113 for that sort fo thing, which has one fix since | 17:19 |
apw | sdeziel, though that fix claims to be hyperthetical ... so ... hmmm | 17:19 |
apw | sdeziel, so ... if you can test them I will throw both those options up for you | 17:20 |
sdeziel | apw: sure, I'm here to test stuff :) | 17:20 |
apw | sdeziel, amd64 ? | 17:22 |
sdeziel | apw: yes | 17:22 |
sdeziel | apw: "net: ipv4: fix for a race condition in raw_sendmsg" ? | 17:23 |
apw | right ... whcih is also where your stack strace is in | 17:24 |
apw | sdeziel, ^ will have something to test in about 20m | 17:24 |
sdeziel | apw: awesome thanks | 17:25 |
sdeziel | apw: I don't want to cloud the picture but I also hit LP: #1748513 with a similar looking trace to my untrained eye. It's on a different machine and not easily reproducible but maybe they both are related? | 17:26 |
ubot5 | Launchpad bug 1748513 in linux (Ubuntu) "BUG: unable to handle kernel paging request at 00007f29ec101010" [Undecided,Confirmed] https://launchpad.net/bugs/1748513 | 17:26 |
apw | sdeziel, ok i think we may have found an issue here which would explain the issue ... so could you test: http://people.canonical.com/~apw/lp1748671/fix-not-xenial/ | 17:47 |
sdeziel | apw: sure, give me a couple of minutes | 17:48 |
apw | sdeziel, and don't forget the linux-image-extra | 18:15 |
sdeziel | apw: no I won't :) | 18:15 |
sdeziel | apw: I'm missing linux-signed-image-4.4.0-115-generic (https://paste.ubuntu.com/=wFyMVxCK2G/). Not sure if it matters | 18:19 |
sdeziel | rebooting | 18:20 |
apw | sdeziel, yeah i didn't make one, i can't make one with a signature that you could boot anyhow | 18:20 |
sdeziel | apw: the test kernel fixed the problem | 18:24 |
sdeziel | thanks a lot! | 18:25 |
apw | sdeziel, ok great, will get that applied | 18:26 |
sdeziel | apw: any idea if that could also fix LP: #1748513? | 18:26 |
ubot5 | Launchpad bug 1748513 in linux (Ubuntu) "BUG: unable to handle kernel paging request at 00007f29ec101010" [Undecided,Confirmed] https://launchpad.net/bugs/1748513 | 18:26 |
=== ayan is now known as Guest19170 | ||
FjodorAE | Was looking forward to vboxguest module with 4.16, yet config patch seems to suggest it isn't set (config based on 4.15) - is someone aware, or where should I file a report? | 21:16 |
apw | FjodorAE, file a bug against the package linux for now, and let me know the bug number | 21:26 |
FjodorAE | apw: Thank you - https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1749015 | 21:41 |
ubot5 | Ubuntu bug 1749015 in linux (Ubuntu) "4.16 - VBOXGUEST not built" [Undecided,New] | 21:41 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!