| bryceh | tjaalton, looks like we got a slew of xserver bug reports about 7 days ago - http://www.bryceharrington.org/Arsenal/ubuntu-x-swat/Reports/totals-quantal-workqueue.svg | 00:25 |
|---|---|---|
| bryceh | tjaalton, probably those SIGABRT bugs I'm guessing, somehow related to the X stack update? | 00:25 |
| RAOF | bryceh: Those SIGABRT bugs are likely related to dropping the rethrow-signals patch and lightdm passing -core instead. | 01:11 |
| RAOF | bryceh: ie: we'll now be getting more apport crashes than before (because -core is more reliable at dumping core) but they'll be marked as SIGABRT because that's how -core dies. If that turns out to be annoying we can post-process the dumps in the xorg-server apport hook to extract the right signal. | 01:13 |
| bryceh | RAOF, makes sense, although isn't it odd there was a sudden jump over the period of 1-2 days then it went back to the previous rate | 01:15 |
| RAOF | Yeah, that is a bit odd. | 01:16 |
| RAOF | ...actually, not *that* odd. | 01:16 |
| RAOF | Because whoopsie's duplicate detection would be kicking in, and wouldn't be letting people file new bugs. | 01:16 |
| bryceh | skimming traces, it looks like we're getting good stack traces. I'm curious why the titles aren't being filled in though. Would be more helpful if they specified the symbol | 01:17 |
| bryceh | hmm, reasonable hypothesis re whoopsie. I don't think it was noticing X crashes previously though; if it is now then I can cross off a work item! :-) | 01:19 |
| bryceh | interesting; we have a topic page for xorg. http://status.ubuntu.com/ubuntu-quantal/group/topic-quantal-desktop-xorg.html | 01:21 |
| RAOF | So, to extract that, we'd want to post-process the backtrace and extract the signo out of OsSigHandler; that's got the right signal. | 01:21 |
| RAOF | Hm. Actually, so does Xorg.0.log; the log should say ?Server crashed with signal 11 (SIGSEGV)? | 01:22 |
| RAOF | That might be more reliable. | 01:22 |
| bryceh | hrm, errors.ubuntu.com seems not to be loading the problems list | 01:23 |
| bryceh | RAOF, yeah agreed those should be improved so the titles are clearer | 01:27 |
| RAOF | In my experience errors.ubuntu.com takes *ages* to load. | 01:27 |
| RAOF | Probably because it's scraping launchpad. | 01:28 |
| bryceh | aw well | 01:30 |
| bryceh | here we go | 01:35 |
| bryceh | ok one bug against xserver-xorg-core this month, ranked at 28: https://bugs.launchpad.net/ubuntu/+source/xorg-server/+bug/1033533 | 01:35 |
| ubottu | Launchpad bug 1033533 in xorg-server (Ubuntu) "Xorg crashed with SIGABRT" [Medium,Confirmed] | 01:35 |
| bryceh | RAOF, so I'll take that as confirmation of that hypothesis | 01:37 |
| bryceh | src=0x0 | 01:38 |
| RAOF | Yeah, that's not going to end well :) | 01:39 |
| RAOF | Also, hurray! Working backtraces from xserver bugs! | 01:40 |
| bryceh | indeed :-) | 01:43 |
| bryceh | if I had time I could even totally triage all these | 01:43 |
| bryceh | bet that one is an easy fix though | 01:44 |
| tjaalton | bryceh: some of them seem to be due to nouveau dri/ddx | 03:32 |
| bjsnider | ok, so there's going to be a fourth legacy blob series, the 304 | 05:11 |
| bjsnider | so there will need to be a special package, perhaps nvidia-304 or something, created for that | 05:12 |
| tjaalton | fourth yes, but -71 isn't alive anymore | 05:16 |
| seb128 | can the xorg-edger nvidia's binary drivers be installed on a quantal stack? | 09:03 |
| shadeslayer | hi, the last X upgrade on quantal has rendered my X useless, xorg.log says it can't find any device | 09:37 |
| tjaalton | log please | 09:37 |
| tjaalton | seb128_: probably | 09:38 |
| seb128_ | tjaalton, thanks | 09:38 |
| shadeslayer | yeah, I will need a couple of minutes :) | 09:38 |
| tjaalton | pastebinit /var/log/.. | 09:39 |
| shadeslayer | http://paste.kde.org/540530 | 09:41 |
| tjaalton | a hybrid laptop? | 09:44 |
| tjaalton | pastebin dmesg too | 09:44 |
| shadeslayer | yep | 09:44 |
| shadeslayer | sec | 09:44 |
| shadeslayer | paste.kde.org/540536 | 09:45 |
| tjaalton | macbook.. | 09:46 |
| shadeslayer | uh yeah, it worked just fine till about a hour ago | 09:47 |
| tjaalton | and why the heck do you have nomodeset on the kernel cmdline? | 09:47 |
| shadeslayer | won't boot without it | 09:47 |
| tjaalton | well that dmesg is useless | 09:47 |
| shadeslayer | uh ok | 09:48 |
| tjaalton | what do you mean "until an hour ago"? mesa is the only update since couple of weeks | 09:48 |
| shadeslayer | lemme pastebin last couple of lines from apt logs | 09:50 |
| shadeslayer | paste.kde.org/540542 | 09:51 |
| tjaalton | do you have the previous xserver-xorg-core in /var/cache/apt/archives? | 09:53 |
| shadeslayer | nope | 09:54 |
| shadeslayer | but I can download stuff since I have my phone tethered | 09:54 |
| shadeslayer | also have that package installed already | 09:54 |
| tjaalton | it shouldn't matter, there was an update to an arm-specific patch | 09:55 |
| tjaalton | mesa got an update to fix an intel bug | 09:56 |
| tjaalton | so when was your previous reboot before the upgrade? | 09:56 |
| shadeslayer | 4 days ago | 09:56 |
| tjaalton | and didn't upgrade anything between? | 09:57 |
| shadeslayer | I most likely did | 09:58 |
| shadeslayer | checking | 09:58 |
| tjaalton | do you get the plymouth splash? | 09:59 |
| shadeslayer | mesa upgrade 2 days ago, that's from 8.0.4 to 9.0 | 09:59 |
| shadeslayer | uh, was out of the room for that part, lemme reboot and check | 10:00 |
| tjaalton | without nomodeset | 10:00 |
| shadeslayer | I am certain I won't be able to boot without it | 10:01 |
| shadeslayer | yep, only a black screen | 10:01 |
| tjaalton | so kernel is broken somehow | 10:01 |
| shadeslayer | hm ... | 10:02 |
| shadeslayer | I have always booted with nomodeset | 10:03 |
| tjaalton | ha | 10:03 |
| tjaalton | so no regression | 10:03 |
| shadeslayer | since without it it never works | 10:03 |
| tjaalton | ok, so probably your X got upgraded within the four days, meaning now your ati doesn't support non-kms anymore | 10:04 |
| shadeslayer | and with nomodeset I get Plymouth | 10:04 |
| shadeslayer | oh | 10:04 |
| tjaalton | dunno why it doesn't fall back on vesa then | 10:04 |
| tjaalton | well, tries and fails | 10:05 |
| shadeslayer | ok, any suggestions on what I can do ? | 10:06 |
| tjaalton | nope.. | 10:06 |
| shadeslayer | blergh | 10:07 |
| shadeslayer | tjaalton: thanks anyway | 10:07 |
| tjaalton | try a newer kernel? | 10:07 |
| tjaalton | 3.6-rcx | 10:07 |
| tjaalton | without nomodeset | 10:07 |
| shadeslayer | hm ... | 10:07 |
| shadeslayer | looking at the mainline debs | 10:08 |
| shadeslayer | ok, Plymouth came up with 3.6 and not passing nomodeset | 10:22 |
| shadeslayer | wat | 10:22 |
| shadeslayer | a error occured while mounting boot | 10:22 |
| === seb128_ is now known as seb128 | ||
| === yofel_ is now known as yofel | ||
| tjaalton | uh, how to install the backport stack on precise? enable the ppa, then what? | 15:44 |
| tjaalton | thought there was one package that would pull it all, including the kernel | 15:48 |
| === ogra_ is now known as ogra | ||
| === ogra is now known as ogra_ | ||
| ricotz | tseliot, hi | 18:05 |
| ricotz | tseliot, please drop the 3.6 buildfix patch which isnt needed anymore and break the build | 18:06 |
| ricotz | also please lower xserver-xorg-core (>= 2:1.11.99.901) to (>= 2:1.10.99.901) to make the lts-backport easier | 18:07 |
| tseliot | ricotz: I'll do that tomorrow. Thanks | 18:42 |
| Prf_Jakob | Sarvatt_, bryceh: Regarding the vmwgfx bug where it didn't start, I think the patches Dave Airlie posted will fix those. | 20:45 |
| bryceh | Prf_Jakob, thanks | 21:11 |
| shadeslayer | tjaalton: if you're awake, using 3.6 rc3 without nomodeset gives me a orange screen ... not sure how else to describe it | 22:27 |
| bryceh | shadeslayer, check dmesg? | 22:37 |
| shadeslayer | bryceh: uh ... I kind of hacked my way around stuff and got everything working | 22:38 |
| shadeslayer | using nomodeset, and specifying X to use fbdev in /etc/X11/Xorg.conf | 22:38 |
| shadeslayer | and I can't post dmesg because I can't switch to a tty when not using nomodeset | 22:39 |
| bjsnider | an orange screen | 22:40 |
| bjsnider | don't think i've heard of that before | 22:40 |
| shadeslayer | :D | 22:40 |
| shadeslayer | I've never seen one before :P | 22:41 |
| bjsnider | i guess ubuntu now has the orange screen o' death | 22:41 |
| shadeslayer | lol | 22:41 |
| shadeslayer | lemme reboot without nomodeset and no xorg.conf and try to switch to a tty | 22:41 |
| shadeslayer | new color this time | 22:43 |
| bjsnider | purple would make sense | 22:44 |
| shadeslayer | http://db.tt/b8KiiYRu | 22:44 |
| shadeslayer | not exactly purple :P | 22:45 |
| bjsnider | you should be able to get to a tty from that | 22:45 |
| shadeslayer | nope | 22:46 |
| shadeslayer | no amount of ctrl-alt-f1 gives me a tty | 22:47 |
| bjsnider | f1-f6 | 22:47 |
| shadeslayer | tried all of them | 22:47 |
| bjsnider | if you touch the lock keys do the lights on the keyboard light up? | 22:47 |
| shadeslayer | sure | 22:48 |
| shadeslayer | on a external keyboard | 22:48 |
| bjsnider | the keyboad is working then | 22:48 |
| shadeslayer | caps lock lights up on the onboard keyboard, so that works as well | 22:49 |
| shadeslayer | bjsnider: why the aversion to nomodeset ? | 23:01 |
| shadeslayer | hm, dmesg.2 is empty | 23:07 |
| shadeslayer | that's the boot without nomodeset | 23:07 |
| bjsnider | the modern drivers need kms | 23:12 |
| shadeslayer | bjsnider: oh ok | 23:33 |
| shadeslayer | well, like I said, no dmesg logs from the boot without nomodeset | 23:33 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!
