**** BEGIN LOGGING AT Mon Apr 24 09:59:58 2006 Apr 24 10:49:46 03rwhitby * 10kernel/2.6.16/ (7 files): Removed obsolete patches, and added the copypage-xscale patch from OE. Apr 24 12:22:30 03rwhitby * 10kernel/patches/ixp400-eth/series: Removed the mac-address patch (we do it via a netdev event now). Apr 24 15:55:12 03bzhou * 10unslung/ (make/py-constraint.mk Makefile): new py-constraint ipk Apr 24 17:42:38 which kernel options are required to make use of the uart as a console? Apr 24 17:46:22 all I'm seeing is the "Uncrompessing kernel..........................." line Apr 24 19:07:51 has anyone a clue why my slug keeps getting one hour behind all the time? Apr 24 19:08:11 ? Apr 24 19:09:44 actually never mind, it was the hwclock -s in crontab Apr 24 20:16:24 03bzhou * 10unslung/make/py-constraint.mk: added README and examples Apr 24 23:35:26 03rwhitby * 10kernel/2.6.16/92-nslu2-mac.patch: Fixed the nslu2 mac patch to read the correct partition Apr 25 00:00:53 03rwhitby * 10unslung/.cvsignore: Added ds101j. Apr 25 00:12:07 is there a common mistake that would lead to no noticable output after "Uncompressing Linux.... done, booting the kernel"? Apr 25 00:13:49 tsowell: yes - incorrect machine id is usually the cause of that Apr 25 00:13:56 do you have the correct shim prepended? Apr 25 00:14:42 I have no clue, so I guess I'll look into that Apr 25 00:15:02 Look at the devio calls in the kernel cvs repo makefile Apr 25 00:15:31 redboot gives an ixdp425 id, which is incorrect for an nslu2 kernel. so we prepend some assembly which fixes it. Apr 25 00:16:08 sigh, all because some ppl are squeamish about upgrading their bootloader ;-) Apr 25 00:18:17 03rwhitby * 10kernel/2.6.16/92-nslu2-mac.patch: Fixed the mac address offset into the redboot partition Apr 25 00:19:28 jacques: heh. well, we now have a confirmed case of a hard brick because someone used the Linksys EraseAll utility to inadvertently "upgrade" their bootloader. Apr 25 00:20:00 So for the small number of people who have tried, one failure out of maybe less than 50 is a high percentage for total brickage risk. Apr 25 00:20:17 Hence the advise not to upgrade the bootloader :-) Apr 25 00:25:57 I'd be willing to do it.. though I'd probably brick it further if I had to solder onto the jtag pads Apr 25 00:26:41 that's the only recovery if you loose the bootloader ... Apr 25 00:29:26 yeah. I could do it if the required soldering wasn't so precise Apr 25 00:29:32 oh well. Apr 25 00:38:00 03bzhou * 10unslung/make/nmap.mk: upstream upgrade to 4.03 Apr 25 00:38:44 rwhitby: thanks for the help. got it working Apr 25 00:42:14 was it the prepend? Apr 25 00:42:32 yep Apr 25 00:59:14 cool Apr 25 00:59:34 when it hangs there, it's usually the machine id Apr 25 01:37:59 03rwhitby * 10kernel/Makefile: Removed madwifi-ng modules before rebuilding them, so you don't get asked the question. Apr 25 02:49:17 03bzhou * 10unslung/make/py-cherrypy.mk: upstream upgrade to 2.2.1 Apr 25 03:30:06 <[g2]> tsowell why would you want to upgrade the bootloader ? Apr 25 03:31:46 redboot doesn't seem to bad from an end-user perspective... my only complaint is that it takes a while to get from power-on to booting Apr 25 03:32:45 perhaps that's unavoidable, but it doesn't happen on other bootloaders on other embedded systems I've used Apr 25 03:39:25 <[g2]> tsowell so you simply want faster boot times ? Apr 25 03:40:17 yeah Apr 25 03:40:39 <[g2]> there are easy options for that Apr 25 03:40:58 oh, cool. I guess I should read about it Apr 25 03:41:22 <[g2]> APEX with the slug if you are a little brave with a properly clocked slug Apr 25 03:41:29 <[g2]> or I've got a Loft Apr 25 03:42:28 <[g2]> it's really just a matter of how much you want to play with it versus how much you want it to be done and what one can affored Apr 25 03:42:38 <[g2]> s/affored/afford/ Apr 25 03:42:38 [g2] meant: it's really just a matter of how much you want to play with it versus how much you want it to be done and what one can afford Apr 25 03:43:15 yeah. I'm probably not going to try it because it would be too much of a pain to get a jtag attached if I screw up Apr 25 03:44:55 <[g2]> tsowell what sw are you running ? Apr 25 03:46:07 so far just busybox. I'm still futzing with the kernel stuff right now Apr 25 03:46:30 <[g2]> tsowell so you are building an openslug from source ? Apr 25 03:46:49 <[g2]> or running in a chroot ? Apr 25 03:47:02 I'm using buildroot Apr 25 03:47:38 <[g2]> wow Apr 25 03:48:16 ? Apr 25 03:48:37 <[g2]> that's so 20 months ago Apr 25 03:49:21 <[g2]> iirc Aug/Sept. '04 Apr 25 03:49:30 hehe. I just prefer make over python for a build system Apr 25 03:49:44 so I'm rolling my own Apr 25 03:49:53 <[g2]> there kernel make files are in CVS Apr 25 03:49:59 <[g2]> s/there/the/ Apr 25 03:50:00 [g2] meant: the kernel make files are in CVS Apr 25 03:50:39 oh, all I have left is the ethernet Apr 25 03:51:07 <[g2]> well you haven't run anything yet either :) Apr 25 03:51:17 I've run busybox, hehe Apr 25 03:51:41 <[g2]> meaning, like passing the 800+ perl tests, or several hundred python tests Apr 25 03:52:35 I don't need no testing Apr 25 03:52:43 <[g2]> :) Apr 25 03:55:27 hoookaaaaay Apr 25 03:56:45 heh - jacques is the king of toolchain testing Apr 25 04:01:01 well, I think it'll be fine for me. I'm only going to run mpd plus some custom stuff Apr 25 04:02:43 tsowell: have you seen ucslugc? Apr 25 04:04:11 yeah, that would still be too easy. it's no fun without the frustration Apr 25 04:27:14 <[g2]> rwhitby sounds like tsowell is just tinkering Apr 25 04:29:02 good stuff. we need more tinkerers. Apr 25 09:30:24 03rwhitby * 10kernel/2.6.16/ (92-nas100d-mac.patch 92-nslu2-mac.patch): Updated the mac event patches to print the new mac address after the ixp module loads **** ENDING LOGGING AT Tue Apr 25 09:59:57 2006