[00:27:14] --- kaj has become available [00:28:21] --- haba has left [01:00:26] --- haba has become available [03:03:26] --- jaltman has left: Disconnected [03:04:30] --- abo has become available [03:57:02] --- abo has left [03:57:25] --- abo has become available [05:22:38] --- haba has left [05:25:56] --- abo has left [05:26:18] --- abo has become available [05:27:17] --- jaltman has become available [05:59:50] --- abo has left [06:00:03] --- abo has become available [07:01:13] --- haba has become available [07:01:26] --- deason has become available [07:11:04] --- abo has left [07:11:50] --- abo has become available [07:11:53] --- dwbotsch has left [07:14:20] --- RedBear has become available [07:22:17] --- meffie has become available [07:29:51] --- abo has left [07:30:31] --- abo has become available [08:33:19] --- abo has left [08:33:31] --- reuteras has left [08:33:31] --- haba has left [08:33:55] --- haba has become available [08:33:58] --- reuteras has become available [08:34:17] --- abo has become available [08:35:00] --- abo has left [08:35:01] --- reuteras has left [08:35:01] --- kaj has left [08:35:54] --- abo has become available [08:36:14] --- reuteras has become available [08:36:18] --- kaj has become available [08:45:42] --- reuteras has left [08:51:57] --- haba has left [09:06:21] --- abo has left [09:06:32] --- abo has become available [09:29:36] --- sxw mobile has become available [09:31:58] andersk: I can fix your gerrit issue. I won't have decent net access till tomorrow night. [09:32:48] As Derrick says, I also want to look at the locking patch in more detail. [09:34:06] Our conversion to struct flock should just be the inverse of a conversion the VFS has already done. [10:22:59] --- Russ has become available [10:26:46] --- kaj has left [10:48:12] --- haba has become available [11:02:56] --- sxw mobile has left [11:14:57] It is; see posix_lock_to_flock in fs/locks.c. [11:18:27] --- haba has left [11:20:00] --- haba has become available [12:05:03] --- kaj has become available [12:23:08] --- Russ has left: Disconnected [12:50:37] is gerrit 1317 (GetSomeSpace_r thing) something that should be ported to 1.4 for 1.4.12pre4, or should it wait? [12:51:26] i'm unsure, honestly [12:53:22] in theory you should not hit it often at all if you know what you're doing to a certain degree... but I already know of one place where that wasn't the case, so... [12:54:42] i can probably guess... [12:56:58] the question is 1.4.12 versus 1.4.13, really [13:04:36] do you mind if I push to 1.4 gerrit either way? so it can be seen what would actually go in, and for general review regardless (for 1.4 it'll be a little different) [13:12:50] push to gerrit, yes [13:13:07] (sorry, upstream router here seems to have fallen over) [13:24:27] --- dev-zero@jabber.org has left [14:07:01] --- Russ has become available [14:14:15] --- jaltman has left: Disconnected [14:30:50] --- sxw mobile has become available [14:32:10] --- sxw mobile has left [15:08:46] --- Kevin Sumner has left [15:36:46] --- deason has left [16:11:14] --- jaltman has become available [18:11:15] --- deason has become available [20:03:37] cvsweb.cgi on openafs.org is not superseded by gerrit, right? [20:04:11] no [20:04:55] (trying to figure out why my FBSD/osi_machdep.h has 151 lines and therefore where it came from, along the way to figuring out who broke it and when, and thereby the correct fix) [20:05:04] git annotate [20:05:26] But that means I have to install git [20:05:32] emo tear [20:05:43] It's not hard. [20:06:12] rhel, install EPEL and git. fedora, install git. debian, install git. the "hard" one is macos. i told macports to install git [20:06:13] (I do actually have it installed; I just haven't used it, yet.) [20:06:22] then, other emo tear [20:06:27] Derrick. Remember the "FBSD" part? [20:06:35] fine, ports then [20:07:22] if the freebsd developers hadn't botched my "no really you've dropped kern/75710 on the floor, if you don't plan to fix it just remove it from the supported hardware list" i'd care more about freebsd [20:07:37] --- abo has left [20:08:17] like, they asked for a backtrace. which i provided. it went into a moderation queue where someone threw it away. so i provided a patch which made things at least not panic. that made the ticket, and it's still there. like 2 years later. [20:08:37] --- abo has become available [20:09:05] So I see. [20:09:21] this isn't hard: plug it in. ifconfig it up. panic. if you don't have one? then it's not really supported hardware, is it... [20:09:46] Sure sounds like not supported hardware. Though maybe I could bug #freebsd-bugbusters ... [20:09:54] i don't care now. [20:10:06] and they won't either: the answer will be "try 8-current" [20:10:27] which when pfsense works on top of it, i will: i want 2 axe drivers to not panic! [20:11:31] (in all seriousness, i will probably have a freebsd fileserver and client in my router whenever i get the time to deal with it. i understand pretty well all the non-hardware code now) [20:12:14] Exciting. [20:12:38] very. all because i didn't want any linux in my house. [20:13:50] Sometimes I wonder if there are any operating systems left that have not kicked your puppy at one point or another. [20:14:10] mine? probably not [20:14:23] solaris: hey, who needs IDE controller support? everyone uses SCSI, right? [20:14:29] openbsd: theo [20:15:10] aix: has streams issues. in 4.x you need to work around bugs. in 5.x the bug workarounds cause bugs. [20:15:54] hpux: 11i v3 appears to have a random and semi-undocumented kernel extension interface. 11.0 and 11i v1 were missing the VFS interface header entirely [20:16:26] irix: i'd tell you my complaint if i ever got that far. i think the compiler is still busy warning me about something [20:17:07] tru64: there's this buffer cache system apparently. as long as you don't want documentation for it, an interface to it, or expect it to not panic or eat your files, all is well [20:17:42] i suppose sunos 4 was mostly ok [20:18:07] --- abo has left [20:18:11] --- deason has left [20:18:36] --- deason has become available [20:18:53] --- abo has become available [20:19:27] --- haba has left [20:22:51] --- haba has become available [20:33:29] kaduk: http://git.openafs.org/?p=openafs.git;a=blame;f=src/afs/FBSD/osi_machdep.h [20:46:19] deason: thanks Looks like 1a027a3e added an #if without matching #end And I even got the fix right beforehand ... I think. [20:47:17] ah, simon failed to split his ifdef ladder patch from the glock init patch correctly [20:48:53] i'll fix it if you don't feel like pushing to gerrit; we'd obviously rather you pushed to gerrit if you plan to do more [20:53:27] I don't anticipate having time to figure out how to push to gerrit in the next two days, but probably in the next week. "I bet it's not urgent, since no one else is complaining" [20:53:33] (I really should be asleep, now) [20:54:48] make your change. git commit (file). it gives you a sha1. git push ssh://gerrit.openafs.org:29418/openafs sha1:refs/for/master [20:55:13] this presumes you created an account in gerrit and gave it your ssh key, but that's all [20:55:22] and the wiki has a page with the details, too [20:59:30] Thw wiki is still down, it seems, but I note that you can substitute any commitish for sha1, including “HEAD” (the thing you just committed). [21:01:35] "still". [21:01:38] try "again" [21:02:40] some clever crawler keeps invoking a lot of cgi scripts concurrently, pushing the machine into swap [21:47:23] --- deason has left [22:07:58] --- reuteras has become available [23:07:25] --- abo has left [23:08:06] --- abo has become available [23:08:29] --- reuteras has left [23:08:48] --- reuteras has become available [23:12:38] --- dev-zero@jabber.org has become available [23:16:23] --- reuteras has left [23:32:54] --- Russ has left: Disconnected