**** BEGIN LOGGING AT Thu Feb 22 02:59:58 2007 Feb 22 03:53:51 03bzhou * r5579 10optware/trunk/make/py-sqlalchemy.mk: py-sqlalchemy: 0.3.4 -> 0.3.5 Feb 22 04:00:58 03bzhou * r5580 10optware/trunk/make/netatalk.mk: netatalk: without-shadow Feb 22 11:16:53 hmm nfs-utils don't build in .dev Feb 22 14:35:20 ./.libs/libpcrecpp.so: undefined reference to `std::vector >::~vector()' Feb 22 14:35:23 ./.libs/libpcrecpp.so: undefined reference to `std::vector >::vector(std::allocator const&)' Feb 22 14:35:27 collect2: ld returned 1 exit status Feb 22 17:04:42 I forget who mentioned they'd tried webmin on a slug before. The issue I'm running into is building perl module Authen::PAM from CPAN. It doesn't seem to recognize the --host build arg, for cross compiling Feb 22 18:07:20 dwery: ping? Feb 22 18:08:43 dwery: When you get back, HopsNBarley and I would like your input on the qnap stuff Feb 22 18:17:28 NAiL, what happend with this crystal snafu you mentioned? Feb 22 18:17:42 The crystal is 33.000Mhz instead of 33.333Mhz. In itself, that isn't a disaster Feb 22 18:18:01 the problem is that if you don't hack the kernel a bit, your clock drifts 1% Feb 22 18:18:36 and %1 is a *lot* when we're talking a clock Feb 22 18:18:51 36 secs/hour Feb 22 18:18:58 sure. so this is a hardware design error? the clock doesn't match the cpu freq or something like that? Feb 22 18:19:42 Hardware design, yes. Feb 22 18:20:23 I don't know the exact details, but I assume that the linux kernel uses that crystal to keep time. Feb 22 18:20:38 with the assumption that it's ticking along at 33.333Mhz Feb 22 20:00:44 NAiL: pong Feb 22 20:02:02 dwery: /msg Feb 22 20:10:25 NAiL, just clipped 600k from the kernel image. Feb 22 20:10:50 lunch break.... Feb 22 20:11:22 600k?! Feb 22 20:11:52 HopsNBarley: btw, I've reworked the IDE-stuff. they now apply cleanly to 2.6.20.1 Feb 22 20:14:45 wait, didn't you say you didn't know C very well ? (-; Feb 22 20:14:51 do you mean the whole patch goes in? Feb 22 20:15:29 no, just the IDE-stuff. It's just a few lines. Feb 22 20:15:41 basically just adding pci-ids Feb 22 20:16:46 wtf? Feb 22 20:16:47 http://www.linux-ide.org/ Feb 22 20:19:37 NAiL: yeah, I noticed some ATA stuff in slugos. Is it only installed on non NSLU based boxes? Feb 22 20:19:45 yeah Feb 22 20:19:49 good. Feb 22 20:19:50 eg. dsmg600 and such Feb 22 20:19:54 ok Feb 22 20:19:55 or... hang on Feb 22 20:20:04 dunno, actually Feb 22 20:20:11 let me check... Feb 22 20:21:03 VoodooZ: btw, I'm working on a completely different kernel ;) Feb 22 20:21:38 oh. ok. Feb 22 20:21:50 It's for the QNAP ts101 Feb 22 20:21:51 openembedded/packages/tasks/task-slugos.bb does mention ata stuff. Feb 22 20:21:57 yeah Feb 22 20:22:12 The idea is to have support for all devices in one image Feb 22 20:22:52 that sucks space though. Feb 22 20:22:55 root@SRX2:/remote$ find /lib/modules/ | grep -i ata Feb 22 20:22:55 /lib/modules/2.6.20/kernel/drivers/ata Feb 22 20:22:55 /lib/modules/2.6.20/kernel/drivers/ata/pata_artop.ko Feb 22 20:22:55 /lib/modules/2.6.20/kernel/drivers/ata/libata.ko Feb 22 20:23:48 about 165K but still... Feb 22 20:26:11 I guess I can always just delete them. Feb 22 20:27:24 dwery: btw, do you know where I should submit the ide-patch? Feb 22 20:31:42 HopsNBarley: Do you have svn write access? Feb 22 20:34:47 nope. Feb 22 20:34:54 hmm... Feb 22 20:35:02 I guess we should set you up with that Feb 22 20:35:21 for what svn? Feb 22 20:35:32 I'm considering putting the patches in the svn repo, so we can have a central repo Feb 22 20:35:34 nslu2-linux Feb 22 20:35:46 in retrospect, that was a pretty stupid question. Feb 22 20:36:32 reflashing the TS now... Feb 22 20:38:24 NAiL: I guess to jeff/alan Feb 22 20:39:43 garzik/cox? Feb 22 20:40:40 yes Feb 22 20:41:54 * NAiL wishes rwhitby was here Feb 22 20:45:48 dwery, did NAiL trouble you to look at our turbostation patch? Feb 22 20:45:58 I did Feb 22 20:45:59 03nail * r705 10kernel/trunk/ts101/ (23 files in 2 dirs): kernel: Add patches for the qnap ts101 kernel Feb 22 20:46:18 And I just checked in the patches to svn Feb 22 20:46:29 Split them up first, of course Feb 22 20:46:48 so... should we try .13, and then work upward as previously suggested? Feb 22 20:50:00 I'd say just going right to the edge is better ;) Feb 22 21:00:42 have you looked at that yet? Feb 22 21:01:39 Just the pci-ids stuff Feb 22 21:01:42 man it takes forever to flash these things... Feb 22 21:02:22 Yeah, I'm wondering if that can be tuned on the second u-boot Feb 22 21:03:12 if the flash timings can be tuned, that can speed up stuff a fair bit Feb 22 21:07:12 HopsNBarley: We should get you a cert so you get write access to the svn repo Feb 22 21:08:10 ya know, rwhitby may have already done that. Feb 22 21:08:14 ah, ok Feb 22 21:08:24 If not, I'd ask him to Feb 22 21:09:08 dwery: you know where to check out to get the patches ;) Feb 22 21:09:39 yep ;) Feb 22 21:09:57 Feel free to improve :-P Feb 22 21:10:08 I've split up the diff by filename Feb 22 21:10:28 any finished patches should get a xxx- prefix and be added to series Feb 22 21:12:42 dwery: feel free to start working at any time :-P Feb 22 21:13:00 ehehe ;) Feb 22 21:13:10 dwery, any advice? Feb 22 21:13:27 topic? Feb 22 21:13:35 22:20 < HopsNBarley> so... should we try .13, and then work upward as previously suggested? Feb 22 21:13:50 uh, yeah, that. Feb 22 21:14:02 what's the latest working kernel? Feb 22 21:14:06 12.3 Feb 22 21:18:10 'kveld shrmps ;) Feb 22 21:18:22 halla Feb 22 21:19:03 nydelige greier dere mekker for oss ;) Feb 22 21:19:36 takker, men gjerne hold deg til engelsk ;) Feb 22 21:19:42 sure Feb 22 21:19:59 which studby? Feb 22 21:20:09 singsaker Feb 22 21:21:09 ok, I'm at Nardo Feb 22 21:21:33 sweet, I used to live there Feb 22 21:21:57 Fiolsvingen ;) Feb 22 21:22:10 ah, Norneveien here ;) Feb 22 21:23:57 So, how do you like your slug Feb 22 21:29:42 my slugs are awezome! one running openslug 3.10 with samba and proftpd, the other one is on unslung running cyrus, postfix and virtual webserver ;) Feb 22 21:29:59 nice Feb 22 21:30:07 oh i forgot proftpd on the unslung one too Feb 22 21:33:40 03nail * r5581 10optware/trunk/sources/buildroot/ (9 files in 2 dirs): buildroot: Update ts101-support Feb 22 21:33:41 03nail * r5582 10optware/trunk/platforms/ (packages-ts101.mk toolchain-ts101.mk): update ts101-platform files Feb 22 21:33:43 03nail * r5583 10optware/trunk/make/buildroot.mk: buildroot.mk: Remove a lot of ts101 ifeqs, update ts101 target Feb 22 22:08:00 hi Feb 22 22:09:18 morning Feb 22 22:10:15 * CoreDump|home decides today is NFS day Feb 22 22:10:42 god I hate the .dev branch =( Feb 22 22:12:49 hrw|gone: do you own a slug as well? Feb 22 22:13:08 CoreDump|home: nope Feb 22 22:13:50 hrw|gone: just wondering what brings you here ;) Feb 22 22:15:07 yet another OE channel Feb 22 22:15:50 and where other I will get infos which NAS to buy? :D Feb 22 22:17:08 hrw|gone: jesus, stay away from the NSLU2!! They are addictive as hell and replicate quickly Feb 22 22:17:35 ? Feb 22 22:18:21 CoreDump|home: I need box with ethernet (or wifi) and usb 2.0. 128MB ram would be nice Feb 22 22:18:23 after you get your first, you quickly find reasons to buy more =) Feb 22 22:18:53 CoreDump|home: what for? Feb 22 22:18:53 hrw|gone: hmm dunno of any Feb 22 22:19:10 CoreDump|home: nslu2 + ram expansion (to 256M) Feb 22 22:19:20 hrw|gone: well, think of the slug as a _really_ tiny PC w/ USB and HDD Feb 22 22:19:59 hrw|gone: yeah, RAM is the limiting factor. I'll do the upgrade sooner or later Feb 22 22:20:14 CoreDump|home: for 130 usd I can get pc with usb2.0 and 128mb ram Feb 22 22:20:17 and cfslot Feb 22 22:20:27 CoreDump|home: similar size as nslu2 Feb 22 22:20:41 hrw|gone: link? Feb 22 22:20:54 moment Feb 22 22:21:37 http://ewayco.com/51-embedded-systems-100-PC-mini-ITX-low-cost/100-TU-low-cost-pc-thin-client-embedded-system.html Feb 22 22:21:58 99 usd + 22 usd posting Feb 22 22:22:54 wow, interesting! Feb 22 22:24:09 03nail * r706 10kernel/trunk/ts101/2.6.20.1/ (6 files): ts101-kernel: Clean up i2c-s3531a a bit and move into one patch Feb 22 22:24:29 HopsNBarley: Ok, that's about as far as I get with the simple stuff Feb 22 22:25:36 does someone use USB Fax? Feb 22 22:25:52 blehh, my order for the serial parts for the slug got "lost"... Feb 22 22:26:27 * CoreDump|home doesn't Feb 22 22:26:41 dwery: You familiar with rs5c372? Feb 22 22:26:56 not familiar but shoul dbe able to cope with it Feb 22 22:27:11 uh Feb 22 22:27:32 http://lkml.org/lkml/2006/3/4/81 <-- not familiar? Feb 22 22:28:14 wrote it without the chio here iirc ;) Feb 22 22:28:58 ok, the ts101 uses that chip Feb 22 22:29:03 rs5c372a Feb 22 22:31:47 I thought it was in mainstream by now Feb 22 22:37:00 03bzhou * r5584 10optware/trunk/make/libextractor.mk: libextractor: 0.5.13 -> 0.5.17; fixed the build Feb 22 22:42:32 dwery: could you dig up the latest revision of the rs5c372 driver? Feb 22 22:46:09 it isn't mainstream? d'oh! I missed a driver somewhere! :) Feb 22 22:46:57 maybe it got merged with some other driver Feb 22 22:47:28 NAiL: it is in mainstream Feb 22 22:50:42 it is? odd Feb 22 22:50:58 ah Feb 22 22:50:59 found it Feb 22 22:51:04 just looking in the wrong place Feb 22 22:54:36 so.. I guess the stuff in arch/ppc/platforms should be cleaned up a bit. I don't know how the ppc folks do it... Should there be a separate set of files for the ts101? Feb 22 22:54:46 looks like qnap has just hacked the sandpoint stuff Feb 22 22:54:49 03bzhou * r5585 10optware/trunk/ (Makefile make/libextractor.mk): promoted libextractor to COMMON_CROSS_PACKAGES Feb 22 22:58:00 HopsNBarley: ping? Feb 22 23:01:20 yo. Feb 22 23:01:56 found another unneccessary patch Feb 22 23:01:57 linux-2.6.12.3-qnap/arch/ppc/syslib/todc_time.c Feb 22 23:02:08 +#obj-$(CONFIG_SANDPOINT) += i8259.o pci_auto.o todc_time.o Feb 22 23:02:08 +obj-$(CONFIG_SANDPOINT) += pci_auto.o Feb 22 23:02:12 It's not compiled at all Feb 22 23:02:29 yeah, i put some comments in there... and... yes. Feb 22 23:03:00 comments? Feb 22 23:03:10 03nail * r707 10kernel/trunk/ts101/2.6.20.1/linux-2.6.12.3_arch_ppc_syslib_todc_time.c: ts101-kernel: Remove another unneccessary patch Feb 22 23:04:17 the patches now clock in at 50k Feb 22 23:04:29 uh, printk i mean. then realized it wasn't even compiled. Feb 22 23:04:37 ah Feb 22 23:05:46 btw, how should we do it wrt. mtd maps. I don't think the physmap patch will make it upstream ;) Feb 22 23:05:57 I'm all for cmdline partitioning Feb 22 23:06:04 that should be in its own file, like the storcenter does. Feb 22 23:06:45 ...and just like that... i've got an openprotium login. Feb 22 23:06:52 man, that is just so cool. Feb 22 23:07:36 on the ts101? Feb 22 23:08:29 Btw, regarding flash partitioning... Feb 22 23:09:04 The original partition will suck with our setup Feb 22 23:09:13 I suggest repartitioning :) Feb 22 23:09:27 did that (-; Feb 22 23:09:27 ah Feb 22 23:09:34 how did you partition the stuff? Feb 22 23:09:36 actually, i have a 201, but same diff. Feb 22 23:10:08 i took the storcenter compatible layout so i could use the same image builder - 0x200K kernel, 0x590K root Feb 22 23:10:27 wait, make that 0x170K kernel Feb 22 23:11:03 hmm Feb 22 23:11:12 how does that work out with the original u-boot env? Feb 22 23:11:43 what do you mean? i made sure i didn't copy over u-boot (-; Feb 22 23:12:09 hehe Feb 22 23:12:24 no, but there's a u-boot config partition somewhere at the end of flash Feb 22 23:12:41 + .name = "U-Boot Config", Feb 22 23:12:41 + .offset = 0x00F40000, Feb 22 23:12:46 i think that stores the environment vars. Feb 22 23:12:57 yes Feb 22 23:15:27 hmm. how'd ethernet bonding get in there? i think it made eth0 go away. Feb 22 23:15:55 no idea ;) Feb 22 23:17:24 my fault of course. Feb 22 23:18:11 u-boot config is very small, 0x20K Feb 22 23:18:17 so that must be the environment. Feb 22 23:18:25 how much flash on the 101? Feb 22 23:18:30 16mb Feb 22 23:18:37 same as 201. tons. Feb 22 23:19:05 I'd like to get as much free space for a jffs2 as possible Feb 22 23:23:19 *duh* Feb 22 23:27:08 my kernel is around 1.4. i took out some stuff and added in jffs2, of course Feb 22 23:28:28 I think the kernel partition should be 2mb Feb 22 23:28:40 lemme through this non-bond kernel on there, and we might have it. Feb 22 23:28:51 non-bond? Feb 22 23:29:20 CoreDump|home: ping - how goes the altboot stuff? Feb 22 23:29:20 ah ha! that bonding ethernet driver that leaked in. Feb 22 23:29:36 rwhitby: slow due to still missing serial Feb 22 23:29:58 ah, yes. I'm waiting for a serial cable for my fsg-3 ... Feb 22 23:30:01 i'm in 1.35M, and most everything else could be module, except perhaps nfs if you want to boot from that. Feb 22 23:31:03 HopsNBarley: I'd like a kernel in flash for the rescue-fs. I won't flash the kernel more than once, since I'll be running debian kernels from disk. Feb 22 23:35:05 to wax philosophic, here is a problem i see in general all across OE. we all have our very legitimate wants and needs, yet, a single image/version of a package/bb file just won't solve all these issues. i think we need USE flags. Feb 22 23:35:40 there, i said it. Feb 22 23:36:03 hehe Feb 22 23:36:32 I'll be going to bed now Feb 22 23:36:35 nite all Feb 22 23:41:00 nite NAiL Feb 22 23:44:53 any idea why my e1000 doesn't show up? i.e. no eth0. yet it probes... Feb 23 00:04:16 anybody else working on the TS-x01? Feb 23 00:05:36 03bzhou * r5586 10optware/trunk/sources/perl/Cross/config.sh-powerpc-linux: perl: added config.sh for ts101 Feb 23 00:32:04 03bzhou * r5587 10optware/trunk/Makefile: Makefile: no touch host/.configured when adding platform Feb 23 00:34:49 HopsNBarley: Too much stuff removed? Feb 23 00:43:36 well, no. it's weird. i had to turn on the r8169 ethernet driver to get the net up. Feb 23 00:43:51 is the qnap config just wrong? Feb 23 00:49:19 quite possibl Feb 23 00:49:22 y Feb 23 00:52:11 i never cease to be amazed by the shoddiness of vendor code. sigh. Feb 23 00:54:22 yikes. scp is not fast, either. Feb 23 00:54:31 HopsNBarley: which nic is in your ts? Feb 23 00:54:46 apparently it's the r8169. Feb 23 00:55:06 what do you know about this? (-; Feb 23 00:55:52 Well, there's two different versions of the ts101 Feb 23 00:55:59 one with intel nic, and one with the r8192 Feb 23 00:56:17 okay. same for 201, i'm guessing. Feb 23 00:56:17 r8169 Feb 23 00:56:24 probably Feb 23 00:56:45 well, at any rate, i am ssh'ed into my TS. Feb 23 00:57:09 any experiment you are dying to see? Feb 23 00:57:16 hmm Feb 23 00:57:29 not that I can think of Feb 23 00:58:06 * CoreDump|home checks ts101 and shudders at the price Feb 23 00:58:17 hehe Feb 23 01:00:03 eno: thx for the ppc config.sh Feb 23 01:04:16 n8 Feb 23 01:08:20 Why the heck is python trying to pull in dietx11? Sigh, time to go digging in the bb files... Feb 23 01:09:51 the TS are still way too expensive. you can get a PC-based platform for the same price Feb 23 01:10:06 However, the Iomega storcenters are getting ever cheaper. Feb 23 01:14:38 TS101 w/o disk == 280 USD. Storcenter w/250gb disk == 360 Feb 23 01:14:44 That's here in .no Feb 23 01:15:12 what is amusing is in all that patching, it still says vendor: motorola, machine: sandpoint. should we fix this to something nicer? Feb 23 01:15:26 yes Feb 23 01:15:33 we've got storcenter with 2x 250Gdisk == $360 (i think) Feb 23 01:15:49 ohh Feb 23 01:16:02 should be more like $299 though, or even $249. Feb 23 01:16:08 "value pricing". Feb 23 01:16:08 oh well Feb 23 01:16:10 sigh Feb 23 01:16:10 hehe Feb 23 01:16:15 gotta try to sleep again Feb 23 01:16:16 later Feb 23 01:16:22 okay. good luck! Feb 23 01:16:28 slug + 250 disk was ~150¤ Feb 23 01:16:31 ;) Feb 23 01:28:41 Stupid python bbfile has a depends line that includes tk. Somehow I don't think that tk & X is desired on the slug. Feb 23 01:29:49 i've got that off in my branch too. just remove the dependency in python and it goes away. Feb 23 01:37:49 I know. Thanks! I'm just thinking of submitting it as a bug and letting the devs fix it, yell at me for being lame, or ignore it. :) Feb 23 01:38:29 we need use flags... Feb 23 01:40:06 That's why I run Gentoo on my other slug. :) Feb 23 01:41:30 yeah, how's that working out? Feb 23 01:42:05 I like it. Feb 23 01:42:13 if i could somehow crash ports and OE together.... Feb 23 01:42:20 I use distcc to pawn off the compiling onto bigger hardware. Feb 23 01:42:50 A slug with 64 or 128MB of ram would be about perfect. Feb 23 01:42:53 how's gentoo on cross compiling? does emerge, et al. know how to make packages for others? Feb 23 01:43:09 there is a useful tool called crossdev Feb 23 01:43:34 I just have to run "crossdev -t ' and it will build the environment for me. Feb 23 01:43:38 storcenter has 64M, and it is perfect. i'm running apache/modphp/gallery/mt-daap/dnsmasq/openntpd, all fine Feb 23 01:43:41 hey, nice. Feb 23 01:43:46 i'll check that out. Feb 23 01:44:27 My poor GentooSlug has spent the entire day building GCC 4.1.2 since it can't hand off the building of Stage2 and 3 to the big machines. :) Feb 23 01:45:12 we could probably twist OE to make gentoo binary tars easily enough. that would help alot. Feb 23 01:47:41 I've got my Gentoo install on a 1GB flash drive. I didn't bother installing the man pages, info docs, and other similar crap. I still have 350MB free. :) Feb 23 01:48:26 Ahoy all :) Feb 23 01:48:34 Howdy! Feb 23 01:49:15 bkero_: I just realized why my friend from OSU was so sad when duck season ended... :) Feb 23 01:49:32 Heh Feb 23 01:49:36 Bad joke. :P Feb 23 01:49:38 He takes far too much pride in shooting ducks. I told him to stay away from Corvalis. Feb 23 01:49:45 hey bkero_: did you get the webcam stuff sorted out? Feb 23 01:49:47 * joshin used to live in Medford. Feb 23 01:50:15 rwhitby:Not yet, got some new stuff on my plate. Feb 23 02:09:31 rwhitby:We got an EFIKA board, and we're fooling around with a couple of those too :) Feb 23 02:10:24 bkero_: cool Feb 23 02:10:32 bkero_: have you guys heard of openmoko? Feb 23 02:11:28 Those are the cell phones, yes? Feb 23 02:12:51 Platform :P Feb 23 02:13:21 A couple of the devs at the OSL know quite a bit about the project, so I wouldn't be surprised if a couple were involved. Feb 23 02:17:25 they were looking for mirrors - I told them they should contact osuosl ... Feb 23 02:24:16 Ah Feb 23 02:24:49 Wow, for being so 'open', they don't hav emuch on their site. :P Feb 23 02:28:18 bkero_: wiki.openmoko.org Feb 23 02:29:57 A bi tmore informative than their main page. Thanks. :) Feb 23 02:38:07 How are complete are thumb instructions on 2.6.16? **** ENDING LOGGING AT Fri Feb 23 02:59:57 2007