[00:20:09] --- reuteras has become available [00:56:42] --- reuteras has left [01:26:07] --- Russ has left: Disconnected [02:04:53] --- Simon Wilkinson has left [02:43:10] --- Simon has become available [03:06:25] --- Simon has left [03:18:18] --- Simon has become available [03:18:51] --- Simon has left [03:19:33] --- Simon has become available [03:24:48] --- Simon has left [04:20:30] --- Simon has become available [04:30:22] --- Simon has left [04:43:11] --- Simon has become available [04:47:53] --- Simon has left [06:20:17] --- lama has become available [06:34:47] --- lama has left [06:36:08] --- lars.malinowsky has become available [06:57:10] --- lars.malinowsky has left [06:57:43] --- lars.malinowsky has become available [07:31:38] --- deason has become available [08:30:14] --- lars.malinowsky has left [09:22:58] --- Simon has become available [09:27:20] --- lars.malinowsky has become available [09:32:25] --- jaltman/FrogsLeap has left: Disconnected [09:32:34] --- jaltman/FrogsLeap has become available [09:42:17] --- Simon has left [10:42:20] --- jaltman/FrogsLeap has left: Replaced by new connection [10:42:21] --- jaltman/FrogsLeap has become available [10:46:38] --- rra has become available [10:49:42] --- Simon has become available [11:01:00] --- Simon has left [11:01:14] --- jaltman/FrogsLeap has left: Disconnected [11:56:02] --- Simon Wilkinson has become available [11:57:00] hello, do we want to put the repo rpm in the 1.4.14 directory? i made one some time ago. [11:57:24] it's still in /afs/sinenomine.net/user/mmeffie/public/rpm [11:58:23] we need 2 [11:58:34] oh? [11:58:44] repo and repo-rhel [11:58:53] see previous releases [11:59:19] ok, thanks, i'll look. [12:46:54] ok, the other rpm is there now. (and the srpms in case you want them) [13:08:34] --- mfelliott has become available [14:04:10] --- pod has left [14:35:34] Looks like both 1.6 and 1.4 fails to build on Linux ppc64 2.6.32. [14:35:46] See http://bugs.debian.org/618496 [14:35:56] not new breakage then. yay! [14:35:56] I'll forward a note to openafs-bugs. [14:37:21] no flock64 on ppc64? [14:37:35] The 1.6 build failure is in sys_setgroups, which is weird. [14:37:42] That stuff should be dead on hosts with keyrings, no? [14:37:59] isn't ppc64 from a pecial, non-linus tree? [14:38:14] setgroups isn't dead, merely resting. [14:38:28] I don't know very much about it, but that would surprise me since I think he's using Debian packages, and Debian tends to not be willing to do that sort of thing. [14:38:38] bring out your dead [14:39:29] It's only syscall probing that we disable when we find we've got keyrings. I think if your syscall table isn't obfuscated, we'll happily use it. [14:39:45] So we still build all of that code. [14:41:12] The problem here is trivial, though. We're missing a semicolon [14:42:12] /var/lib/dkms/openafs/1.4.14/build/src/afs/afs_prototypes.h:976: warning: `struct flock64' declared inside parameter list ? [14:42:43] Ah, no, I'm looking at the 1.6 failure. It's different. [14:42:47] ah, ok [14:42:53] because that's the downfall for /var/lib/dkms/openafs/1.4.14/build/src/libafs/MODLOAD-2.6.32-5-powerpc64-MP/afs_vnop_flock.c:43: error: conflicting types for `lockIdSet' [14:43:02] I'm pretty sure that we fixed the flock64 issue [14:43:31] conceivable. [14:43:49] oh, more loaded. hang on [14:43:57] yeah, ok [14:44:04] you gonna push that? [14:44:15] I'll push a fix for the semicolon. That's the kind of bug I can handle this evening. [14:49:01] flock64 was the problem that also affected sparc64 and is fixed already in 1.6, I *think*. [14:49:13] There are two different issues. [14:49:51] yeah, looking, looks like it [14:50:48] Hmmm. gerrit is still dog slow, even after I garbage collected last night. [14:51:13] Did you gc both repos? [14:51:35] Hmmm. No. [14:51:50] There's the one in /srv/git and there's another under Gerrit's home directory, and they both seem to have an effect. [14:52:12] I did srv/gerrit, but not the other. [14:52:14] Hang on. [14:52:21] 4234, btw. [14:53:17] Make sure you have an 002 umask before gc'ing the one in /srv/git, or things get unhappy. [14:53:28] I'm just about to do it as gerrit. [14:53:32] That should work. [14:53:36] The ownership in that directory is a little screwed. [14:53:39] Yes. [14:53:40] It is. [14:53:56] I haven't been able to figure out the right thing to set it to and how to set it so that it stays unscrewed. [14:54:54] Yeah. I'm really not sure why stuff in there is owned by me. [14:56:09] Anyway, that one now garbage collected too. [14:56:20] If I wasn't so tired, I'd write a cronjob. Instead, it's now on the list. [14:57:49] like so many times in my life, i will say "we need a sysadmin" [14:58:12] the words have lost all meaning [14:58:42] In theory, I was one. I'm trying to give it up. [14:58:42] I would be one except that our e-mail system is looking like a Japanese nuclear reactor at the moment. [14:59:05] i was one when i started at cmu in 1995. i hve been giving it up for years. it's like smoking [14:59:41] i gave it up and then something happened and i lit up again. and installed whatever and scripted it and rebooted and traced and... [15:00:21] I like being a sysadmin. [15:00:54] more power to you. you are underappreciated for it [15:01:20] my problem is i wish to solve each problem exactly once. if i cannot do it in a form which means i can syndicate that solution, i hate everyone [15:02:03] Sometimes I think it would be easier to just have hating everyone as a default state, rather than needing to find reasons ... [15:02:51] Yeah, that's the part about being a sysadmin that I like. [15:03:00] That's why I hate hardware. You can't ever fix hardware problems only once. [15:04:22] Surely a hammer lets you resolve intermittent faults exactly once. [15:19:58] --- deason has left [15:29:20] --- jaltman/FrogsLeap has become available [20:55:26] > The problem here is trivial, though. We're missing a semicolon Oops! I sent cherry-picks to 4235 and 4236. [22:41:26] --- lars.malinowsky has left