[00:12:37] --- Russ has left: Disconnected [00:16:42] --- abo has become available [04:52:41] --- meffie has become available [06:56:30] --- shadow@gmail.com/barnowl67590625 has left [07:03:20] --- shadow@gmail.com/barnowl67590625 has become available [07:20:57] --- Russ has become available [07:29:49] --- deason/gmail has become available [07:58:55] --- Russ has left: Disconnected [08:00:54] --- meffie has left [08:16:44] --- summatusmentis has become available [08:56:21] --- mho has become available [10:46:01] --- mfelliott has become available [13:53:35] --- Russ has become available [14:03:44] --- Russ has left [15:29:45] --- deason/gmail has left [15:46:24] --- jaltman/FrogsLeap has left: Replaced by new connection [15:46:25] --- jaltman/FrogsLeap has become available [18:00:09] Is adding a new function to afs_osi.c a reasonable thing, or should such things go elsewhere? [18:00:23] depends on the function [18:01:04] shutdown_osisleep() [18:01:13] (Should probably get renamed, really.) [18:19:13] I guess LINUX{,24} calls it osi_event_shutdown [18:20:38] --- jaltman/FrogsLeap has left: Disconnected [18:21:40] --- jaltman/FrogsLeap has become available [18:25:04] Also, why does LINUX24/osi_sleep.c have code conditional on AFS_LINUX24_ENV? Does that really ever matter? [18:25:38] its probably left over from before the code fork [18:25:56] --- jaltman/FrogsLeap has left: Disconnected [18:28:23] --- jaltman/FrogsLeap has become available [18:29:30] Okay, so a relatively sane reason. [19:11:21] Anyone familiar with the linux shutdown sequence care to look at whether afs_evhasht/getevent() are still being used up through to osi_module.c:afs_cleanup()? DARWIN, for example, can kill the hashtable in shutdown_osi() ... [23:09:57] ... do I need to do something dumb like AFS_SGI_ENV instead of AFS_IRIX_ENV?