| Le-Chuck_ITA | Hi all, I want to draw your attention to the following bug, which has been "quickfixed" by a clever guy who found out a line to comment out, but requires kernel developers attention to be sure it's a good fix. It's a regression from feisty, and since in gutsy for the first time suspend to disk and to ram work flawlessly on toshiba laptops with dual core enabled, I supppose it is worth a look. Bug is: | 01:00 |
|---|---|---|
| Le-Chuck_ITA | https://bugs.edge.launchpad.net/ubuntu/+source/linux-source-2.6.22/+bug/152187/ | 01:01 |
| ubotu | Launchpad bug 152187 in linux-source-2.6.22 "Serial Wacom tablet fails to return from ACPI suspend to RAM" [Undecided,Confirmed] | 01:01 |
| Le-Chuck_ITA | I hope somebody sees this, and thanks a lot. | 01:02 |
| === \sh_away is now known as \sh | ||
| tjaalton | I've got frequent crashes with dapper on a HP DL380G5, and I've got a call trace that starts from bnx2_poll | 07:42 |
| tjaalton | that would suggest that there's a bug in the broadcom driver? | 07:42 |
| tjaalton | (I took a photo of the trace) | 07:43 |
| tjaalton | I found this while searching for similar issues http://bugs.centos.org/view.php?id=2103 | 07:44 |
| tjaalton | please see bug 75767, I've confirmed that one. seems similar enough to mine | 11:53 |
| ubotu | Launchpad bug 75767 in linux-source-2.6.15 "kernel panic: killing interrupt handler" [High,Confirmed] https://launchpad.net/bugs/75767 | 11:53 |
| andres | If I have made a kernel bugreport including a patch which is accepted upstream which fixes crashes and datacorruption, is there anything more I can do? (2.6.22 xen) | 12:49 |
| andres | (#164904 btw) | 13:13 |
| Kano | could some add a little patch for aufs? | 13:19 |
| Kano | patch/sec_perm-2.6.24.patch from aufs cvs | 13:19 |
| Kano | +EXPORT_SYMBOL(security_inode_permission); | 13:20 |
| Kano | mainly | 13:20 |
| Kano | http://kanotix.com/files/kernel/kernel-update-pack-generic-future-pure/source/sec_perm-2.6.24.patch | 13:23 |
| Kano | also i need one file more in the headers: | 13:25 |
| Kano | fs/aufs/Makefile:tmpfs = $(shell grep '\#.*define.*TMPFS_MAGIC' ${srctree}/mm/shmem.c | \ | 13:25 |
| Kano | so i need mm/shmem.c | 13:25 |
| abogani | Kano: Please use the right way: Clone hardy git repo, apply the patch, test resultant kernel, formatting patch as reported at https://wiki.ubuntu.com/KernelGitGuide and send a patch to kernel-team ml. | 13:26 |
| Kano | thats the original patch from aufs cvs | 13:26 |
| Kano | i would suggest to switch to aufs - unionfs is not working from lum | 13:26 |
| Kano | maybe the same issue,dont know | 13:27 |
| Kano | [ 4814.465576] aufs: Unknown symbol security_inode_permission | 13:28 |
| Kano | [ 5954.937595] unionfs: Unknown symbol security_file_ioctl | 13:28 |
| Kano | [ 5954.937662] unionfs: Unknown symbol security_inode_permission | 13:28 |
| Kano | it seems it is the same issue, just that unionfs needs even a 2nd export | 13:28 |
| abogani | andres: Depend from intrusive level of the patch and from type of bug which it resolve. Generally if the bug not meet https://wiki.ubuntu.com/StableReleaseUpdates requirements is very difficult that the patch will be accepted. | 13:29 |
| Kano | so you need to add even 2 export statements... | 13:29 |
| Kano | i do not need to send anything, if you want a working unionfs... | 13:29 |
| andres | abogani: As I said, it causes, sometimes silent, data corruption. Has very little impact, is accepted upstream, only affects xen kernels. | 13:29 |
| andres | Its the xfs not using vunmap() issue. | 13:30 |
| andres | And is very hard to diagnose as the backtraces are quite random, as it can happen at every place where vmap's are used. | 13:31 |
| andres | abogani: so I dont see why it shouldnt adhere to the stable release guidelines. | 13:31 |
| andres | abogani: May I nominate it for some release, or should that only be done by ubuntu devs? | 13:34 |
| abogani | andres: You should send email to kernel-team ml. | 13:36 |
| andres | Ok. | 13:36 |
| Kano | what file to patch for the headers? | 13:58 |
| Kano | i dont get it currently where to write the extra files | 13:58 |
| andres_f | abogani: done | 14:20 |
| andres_f | Ugh. The mailing list is moderated. | 14:22 |
| andres_f | Do you moderate the lists regulary, or should I subscribe and repost? | 14:22 |
| Kano | !24 | 14:22 |
| ubotu | Sorry, I don't know anything about 24 - try searching on http://ubotu.ubuntu-nl.org/factoids.cgi | 14:22 |
| Kano | updated | 14:22 |
| === asac_ is now known as asac | ||
| === \sh is now known as \sh_away | ||
| === \sh_away is now known as \sh | ||
| === \sh is now known as \sh_away | ||
| === henrix_ is now known as henrix | ||
| andres | Is it planned to activate paravirt_ops in 2.6.24-generic by default? | 17:58 |
| andres | I think some distros are planning that and it is quite convienient. | 17:58 |
| andres | (and afaik the overhead is marginal) | 17:58 |
| sommer | hey, all I'm documenting the differences between the server and desktop editions and I just wanted to confirm this article is accurate? | 20:25 |
| sommer | http://www.enterprisenetworkingplanet.com/netos/article.php/3710641 | 20:25 |
| sommer | basically just comaring /boot/config-version between a server and desktop install | 20:26 |
| sommer | comparing rather | 20:26 |
| ivoks | right, those are the differences | 20:26 |
| sommer | ivoks: cool I figured, but thought I'd double check | 20:27 |
| === asac_ is now known as asac | ||
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!