[00:09:27] --- Russ has left: Disconnected [00:09:37] --- dev-zero@jabber.org has left [00:40:14] --- dev-zero@jabber.org has become available [00:45:11] --- haba has become available [01:36:01] --- brantgurga has left [02:05:08] --- brantgurga has become available [03:46:44] --- sxw has become available [04:35:11] --- RedBear has become available [04:41:24] --- sxw has left [05:31:12] --- Jeffrey Altman has left: Replaced by new connection [05:33:28] --- Jeffrey Altman has become available [05:44:54] --- sxw has become available [06:31:57] --- stevenjenkins has left [06:35:52] --- stevenjenkins has become available [06:58:29] --- deason has become available [07:04:27] --- sxw has left [07:46:01] --- brantgurga has left [08:35:52] --- haba has left [08:40:15] --- Russ has become available [09:02:23] --- brantgurga has become available [10:02:30] --- dev-zero@jabber.org has left [10:48:22] --- dev-zero@jabber.org has become available [13:03:16] derrick, simon, are either of you around? [13:03:20] no [13:03:28] Not at all. [13:03:39] However, my evil twin Skippy may be able to answer your question. [13:03:56] how do we tell the evil twin from the non-evil one? [13:03:57] would either/both of you look at afs/tproa.net/users/s/u/summatusmentis/test.diff and tell me if I'm doing my export of rx_GetLocalPeers correctly? [13:04:20] kula: when you find out, its generally too late. [13:04:47] uh... put a slash on the front of that path, obviously [13:05:05] "rx.h" should probably be rx/rx.h [13:05:14] but really, afsincludes.h should do it for you [13:05:34] (and is included in afs_server.c already) [13:05:41] so, get rid of that [13:06:02] -#include +#include "../src/afs/UKERNEL/afs_usrops.h" [13:06:06] Yuck? Why? [13:06:15] it didn't like [13:06:23] "it" [13:06:33] The cat's mother, obviously. [13:06:44] it obviously built before... [13:06:53] I'm sure it's me [13:07:18] the GetLocalPeers exports look fine, but are windows-only, so you don't actually need them for the unix cache manager [13:07:35] (the .def ones) [13:07:36] Yeh. How do we allocate those numbers - is it just 'pick the next available one'? [13:07:47] i believe so. but ideally from head, not 1.4 [13:08:07] The mapfile stuff looks fine, with the exception that I'm not sure about just adding things to the end. [13:08:11] What's the ordering in that file? [13:08:19] There isn't any. [13:08:20] i don't think it's ordered [13:08:30] Okay, so just add it at the end ... [13:08:31] simon, I tried to at least line my stuff up, which is why I added to the end [13:08:32] I think I alphabetized it originally so that I could analyze it more readily, but the linker doesn't care. [13:08:48] hang on, i am going to edit your diff [13:08:51] --- stevenjenkins has left [13:09:24] the reason for the linktest.c edit was because I was getting "undefined reference" errors [13:09:52] ehich were undefined? [13:10:09] that's also the reason for the exporting of rx_GetLocalPeers [13:10:17] lemme go look, but it wasn't any of my code [13:10:31] --- deason has left [13:11:00] uafs_SetRxPort, uafs_Init, uafs_RxServerProc, and uafs_Shutdown were undefined [13:11:08] before I changed the include [13:11:13] you should have gotten an error before that [13:11:21] Okay, so what have you changed that might have affected those? [13:11:38] that patch has everything I've changed [13:11:59] Try reverting the linktest.c portion, and making clean. Do you still get the error? [13:11:59] try this: /afs/andrew.cmu.edu/usr/shadow/test.diff [13:12:06] i didn't change the uafs linktest thing. [13:12:15] you should change it back, get the real error and see what it is [13:12:25] simon, I'm just looking now [13:14:28] --- deason has become available [13:15:07] --- stevenjenkins has become available [13:21:35] alright, the linktest thing didn't re-appear, it may have been something dirty left over [13:22:20] Cool. [13:22:47] /afs/tproa.net/users/s/u/summatusmentis/error_for_derrick_simon [13:23:11] that's the undefined reference error I'm getting because of my code, I haven't tried derrick's patch, one moment [13:23:39] I think Derrick's patch only changes the number of the export on Windows, but I could be wrong ... [13:24:20] linktest doesn't seem to link against rx. [13:24:24] that's all I saw, and that doesn't do anything [13:24:33] no. my version also had some changes to formatting [13:24:37] Ah no, ignore me. [13:25:20] and yes, that is "link -lafsrpc" [13:25:35] (since you're using -lpthread, you want it and not -lrx" [13:26:42] sorry, do I need to be doing something special with make? I've not specified -lpthread anywhere, just a standard ./regen.sh, ./configure, make [13:26:59] no, you need yo edit the makefile rule for linktest [13:27:17] and then at the top, ./config.status so the makefile gets regenerated (edit Makefile.in, not Makefile) [13:27:40] Is your entire new function in rx.c not inside a #ifdef RXDEBUG ? [13:27:47] --- brantgurga has left [13:32:05] you know what, it is inside #ifdef RXDEBUG [13:32:10] that's the issue then? [13:33:33] Well, that's why nothing can link against it. You aren't building it. [13:34:03] I guess that didn't register. Thanks [13:35:11] looks like the build finished. Thanks guys [13:42:52] #error can be useful in situations like these. [13:43:21] #error? [13:43:23] If you aren't sure whether something is being built or not, just stick a #error before it. If it is being built then the compiler will complain. [13:43:31] oh [13:43:34] Yeh. It's a C preprocessor directive. [13:43:37] So you can do [13:43:41] #error Got here [13:57:14] --- dev-zero@jabber.org has left [13:59:04] DKMS is so nice. [13:59:08] I wish Debian would have done that years ago. [13:59:20] Good idea from Red Hat there. [13:59:24] agreed. [13:59:49] * Russ has already gotten totally used to having openafs-modules-dkms in Debian unstable and just always having the module there. [14:00:29] The main pain with dkms is if you have a lab with single machine with a dodgy tool chain. Then things become tedious. [14:00:47] Yeah. [14:01:09] * Russ really needs to fix make libafs_tree to be arch-independent again too. Before squeeze releases. [14:01:45] Since not being able to build kernel modules for x86_64 kernels with x86 userspace is kind of a big limitation right now. [14:01:58] Plus, it would be nice to have the source packages be arch-independent again. [14:02:43] Wh [14:03:40] Sorry, bluetooth stack just crashed. [14:04:21] It was playing Jenga with your keyboard driver. [14:04:22] It lost. [14:04:39] I guess we only build x86_64 + x86_64 for Redhat - the build process doesn't types, so I don't hit that problem. [14:07:05] > Sorry, bluetooth stack just crashed. i have this habit of panicing the mac with bluetooth a2dp. far more than with afs even though i do afs testing on my lap [14:07:30] --- cclausen has become available [14:07:49] > Good idea from Red Hat there. Dell [14:08:22] Oh, okay. [14:08:26] Red Hat just adopted it first. [14:08:32] Or at least before Debian by a lot. [14:08:52] I suspect that it was one of those places where Debian didn't feel quite the same amount of pain since module-assistant was a fairly decent half-way solution. [14:09:07] "only kinda sucked" [14:09:25] the first time i used m-a i did in fact only kinda want to spork someone at the end [14:09:50] I'm pretty sure Fedora will be de-adopting it soon. They're pretty much comitted to the idea that everything that gets loaded into their kernel should be in the kernel tree. [14:10:05] That would be ironic. [14:10:15] Given that Debian gets all the grief about being licensing fantatics. [14:10:32] Debian ship OpenAFS. Fedora refuse to. Need I say more. [14:10:41] But Debian actually has pretty solid support for out-of-tree modules, including non-free ones. [14:10:58] Which is good, since whenever I get a new machine, ALSA never works. [14:11:05] I always have to build the current out-of-tree modules. [14:11:08] the whole state of affairs is why i have written off linux for my own personal use [14:12:00] --- mmeffie has become available [14:12:38] --- stevenjenkins has left [14:13:29] Fortunately, I don't run Fedora's kernel. [14:16:45] 2^119 key recovery attack on AES-256. Not horribly exciting yet, but interesting. [14:17:20] Bleah; is it time for another symmetric cipher alreawdy? [14:17:24] oh man. I just setup TrueCrypt. [14:17:25] Already, that is. [14:17:35] I'll have to re-do everything with TwoFish... [14:17:40] 2^119 is still pretty safe. [14:17:59] i highly recommend reading bruce schneier's comments about this before running off to re-encrypt everything. [14:18:28] des3 the aes'd stuff. yeah! [14:19:02] actually, true crypt does support using two ciphers on data [14:19:09] not des3 though [14:19:18] Yeah, Schneier is expecting it to mostly be an issue for the AES-based SHA-3 entrants. [14:19:33] --- stevenjenkins has become available [14:19:42] basically a "yeah, you folks trying to come up with SHA-3 should think about this" kind of deal [14:20:34] * Russ wonders how far we're off from a quantum factoring computer. [14:20:56] The qubit folks keep producing steadily improving results. [14:38:15] --- brantgurga has become available [14:38:20] --- brantgurga has left: Lost connection [14:38:20] --- brantgurga has become available [14:38:20] --- brantgurga has left: Lost connection [14:38:21] --- brantgurga has become available [14:38:21] --- brantgurga has left: Lost connection [14:38:21] --- brantgurga has become available [14:38:21] --- brantgurga has left: Lost connection [14:39:31] --- brantgurga has become available [14:48:04] --- brantgurga has left [15:13:05] Turns out dkms isn't in RHEL at all. Fedora ship with it, and it's available as an extra, but it's not part of the Enterprise Linux distribution. [15:18:58] Oh, huh. [15:30:47] --- brantgurga has become available [15:44:55] --- mmeffie has left [15:51:55] --- mdionne has left [15:53:46] --- mdionne has become available [15:57:19] --- Jeffrey Altman has left: Disconnected [16:02:25] --- dev-zero@jabber.org has become available [16:58:20] --- cclausen has left [17:10:27] --- mdionne has left [17:12:19] --- Jeffrey Altman has become available [17:12:30] --- brantgurga has left [17:20:18] --- brantgurga has become available [17:29:18] --- dev-zero@jabber.org has left [17:56:59] --- mdionne has become available [19:22:02] --- mdionne has left [19:23:10] --- brantgurga has left [19:23:40] --- brantgurga has become available [19:31:03] --- deason has left [19:38:44] --- deason has become available [21:01:22] --- dev-zero@jabber.org has become available [21:45:22] --- deason has left [22:49:31] --- dev-zero@jabber.org has left: Replaced by new connection [22:49:32] --- dev-zero@jabber.org has become available [23:44:41] --- Russ has left: Disconnected