[00:33:05] --- lama has become available [03:33:55] --- jaltman has become available [03:35:07] --- haba has become available [04:55:00] --- meffie has become available [06:34:05] --- jaltman has left: Disconnected [06:34:14] --- jaltman has become available [06:51:43] --- jaltman has left: Disconnected [07:20:09] --- deason has become available [07:55:01] --- reuteras has left [08:28:22] --- lama has left [09:53:33] --- rra has become available [09:58:31] --- haba has left [10:32:30] --- sxw has become available [10:36:30] --- sxw has left [10:40:24] I pushed out a new Debian 1.4.14 package that pulls up the 2.6.38 kernel support fixes plus some other stuff that went onto stable. [10:40:34] However, I'm very confused by our stable branches. [10:40:54] The release work for 1.4.14 seemed to happen in a branch called openafs-stable-1_4_12_1? [10:41:01] And that branch is still active? [10:41:14] I was assuming that openafs-stable-1_4 would be the stable branch that I should use for packaging going forward. [10:41:56] openafs-stable-1_4_x, rather. [10:44:56] 1.4.14.1 will be from openafs-stable-1_4_12_1; 1.4.15 will be from openafs-stable-1_4_x [10:45:37] yeah, I was confused by the solaris shutdown stuff going in 14_12_1; I thought that was a 1.4.15 thing [10:46:14] > yeah, I was confused by the solaris shutdown stuff going in 14_12_1 i was just going to pull them all up and broach the topic. [10:46:36] probably still "too ambitious" for what should be in 1.4.14.1 [10:51:20] Oh, okay. So if i pull things from 1_4_x for the Debian package, I may end up in a weird state for packaging 1.4.14.1. Ah well. [10:51:36] Things are kind of a mess for 1.4.14 anyway since I'm also carrying Debian packages. This should be much simpler for for 1.6. [10:51:51] Debian *patches*, rather. [10:57:39] --- jaltman has become available [11:08:14] rra: for which versions of debian do you build packages? [11:14:40] I upload new packages to unstable (or experimental for 1.6) and generally backport the current unstable version to stable once it migrates to testing. I normally don't build backports to oldstable once a new stable is out. [11:16:23] I also take care of security updates for any version of Debian with active security support. [11:17:28] --- summatusmentis has left [11:37:15] --- jaltman has left: Disconnected [11:41:56] Russ: I think you mis-cherry-picked “linux: 2.6.38: New d_op handling”. Your version makes extra whitespace changes and has an unmatched +#if defined(AFS_LINUX26_ENV) in src/afs/LINUX/osi_vfsops.c. [11:45:52] I highly recommend ‘git config --global merge.conflictstyle diff3’ to avoid these kinds of mistakes. This way you see <<<<<<< HEAD ||||||| parent of 0d95e6b... linux: 2.6.38: New d_op handling #if defined(AFS_LINUX26_ENV) ======= #if defined(STRUCT_SUPER_BLOCK_HAS_S_D_OP) sb->s_d_op = &afs_dentry_operations; #endif #if defined(AFS_LINUX26_ENV) >>>>>>> 0d95e6b... linux: 2.6.38: New d_op handling which gives you enough information to correctly resolve the conflict, instead of just <<<<<<< ours ======= #if defined(STRUCT_SUPER_BLOCK_HAS_S_D_OP) sb->s_d_op = &afs_dentry_operations; #endif #if defined(AFS_LINUX26_ENV) >>>>>>> theirs which doesn’t. [11:48:10] --- jaltman has become available [11:57:16] I see: you noticed and added the ‘#endif’ instead of deleting the extra ‘#if defined(AFS_LINUX26_ENV)’, which means you effectively pulled in half of commit 4219bfd5 “libafs: afs_backing_dev_info is 2.6-only”, but not all of it. This means that on !AFS_LINUX26_ENV, you #ifdef’d away the osi_Alloc of a struct backing_dev_info but not the osi_Free. [11:58:08] --- jaltman has left: Disconnected [11:58:43] --- jaltman has become available [12:00:41] I pushed the diff between your incorrect cherry-pick and the correct one as git://andersk.mit.edu/openafs-ubuntu.git fix-debian-merge (2b700d2f). [12:03:59] --- jaltman has left: Disconnected [12:51:15] Ah, thank you. I got lost in trying to resolve those merge conflicts (and was very confused as to why I was getting merge conflicts in the first place). [12:51:39] Sounds like I broke pre-2.6 and non-Linux, which probably doesn't matter for Debian immediately, but I will grab your fix. [12:52:48] Yeah, diff3 mode really ought to be the default, I think. [13:47:05] have the docs.openafs.org man pages been regenerated since that windows =include thing was fixed? because... they're still broken [15:05:06] --- sxw has become available [15:05:37] --- sxw has left [15:05:55] --- sxw has become available [15:27:21] --- sxw has left [15:42:41] --- Simon Wilkinson has become available [16:27:51] Don't suppose anyone knows where uint32_t is defined in the AIX kernel headers? [16:28:11] Trying to figure out which include is missing to fix the problem with gerrit 4101 [16:42:40] stdint.h gives it to me on the aix box I have access to [16:45:11] --- deason has left [16:46:07] --- deason has become available [17:36:52] --- Jeffrey Altman has left: Replaced by new connection [17:36:58] --- Jeffrey Altman has become available [17:37:18] --- steven.jenkins has left: Lost connection [17:37:18] --- mfelliott has left: Lost connection [17:37:18] --- phalenor has left: Lost connection [17:37:18] --- shadow@gmail.com/owl1FDC9F18 has left: Lost connection [17:38:22] --- rra has left: Disconnected [18:06:14] --- Russ has become available [18:10:06] --- jaltman/FrogsLeap has become available [18:37:40] --- mfelliott has become available [18:40:44] --- shadow@gmail.com/owl1FDC9F18 has become available [18:41:37] --- steven.jenkins has become available [18:42:00] --- phalenor has become available [19:37:06] --- jaltman/FrogsLeap has left: Disconnected [21:42:19] --- deason has left [21:57:31] --- deason has become available [22:01:40] --- deason has left [23:22:47] --- reuteras has become available [23:28:43] --- steven.jenkins has left [23:38:03] --- steven.jenkins has become available