| 00:00.20 | *** join/#oe ljp (~lpotter@203.94.178.46) |
| 00:00.47 | zecke | CosmicPenguin: I reported the bug, but I already was a duplicate :} |
| 00:01.13 | zecke | ah yeah Keyboard Driver... 3/4 done |
| 00:01.27 | mickeyl | ok, it isn't a problem in the data. using the standard updater.sh/zimage.c750/initrd.bin it works |
| 00:01.45 | mickeyl | so something with the updater.sh or the .data files |
| 00:06.44 | CIA-8 | 03kergoth 07 * r1.797 10oe/ (bin/oe/build.py BitKeeper/etc/logging_ok): BUGFIX: oe.build.mkstamp had a hardcoded path to the stamps dir. |
| 00:06.48 | CIA-8 | 03kergoth 07 * r1.798 10oe/BitKeeper/triggers/ (ciabot_bk.sh ciabot.projects): Update BK triggers. |
| 00:07.12 | kergoth | mickeyl: the new updater uses a different flash eraser. its more recent mtd utils, which fixes a number of bugs. the one sharp ships doesnt properly handle bad blocks |
| 00:07.15 | kergoth | heh |
| 00:07.47 | mickeyl | hmm well |
| 00:07.48 | mickeyl | nice, but |
| 00:07.52 | mickeyl | nonworky here |
| 00:08.00 | mickeyl | any other c750 users around? |
| 00:08.08 | kergoth | well, dunno how to diagnose. i dont have one, cant further diagnose |
| 00:08.20 | mickeyl | yeah i know |
| 00:08.27 | mickeyl | where is it known to work? |
| 00:08.28 | kergoth | and why does it work flawlessly on the 6k? |
| 00:08.37 | mickeyl | 6k so, ok. |
| 00:08.42 | CosmicPenguin | heh - my boss is writing the "How to upgrade vanilla Redhat 9 to work with OE" document... :) |
| 00:08.44 | mickeyl | mithro--: can you try on your c700 ? |
| 00:08.46 | CosmicPenguin | Good luck with all of that |
| 00:08.49 | kergoth | this is pdaxrom based, and they use their isntaller on the c7x0 just fine |
| 00:08.52 | kergoth | dunno |
| 00:09.32 | kergoth | mickeyl: can you try installing pdaxrom? and if that works, i'll compare their stuff to ours and try to find the issue |
| 00:09.38 | mickeyl | eek |
| 00:09.44 | mickeyl | i don't think i want to do that |
| 00:09.52 | mickeyl | besides, i really need a break from zaurus stuff |
| 00:09.57 | mickeyl | a couple of days |
| 00:10.14 | kergoth | well, itd be good to get this release functional and announced before we take breaks |
| 00:10.18 | kergoth | we're so close |
| 00:10.20 | kergoth | heh |
| 00:10.29 | kergoth | guess we could revert the installer for now |
| 00:10.34 | kergoth | use teh old method |
| 00:10.37 | mickeyl | yep, until things sorted out |
| 00:10.37 | kergoth | do the new installer for 3.5.3 |
| 00:10.46 | mickeyl | leave the new installer for the 6k |
| 00:11.32 | kergoth | okay, reverted |
| 00:11.45 | kergoth | can you upload your working updater.sh into the c7x0 & 5600 dirs? |
| 00:11.56 | mickeyl | yes |
| 00:12.06 | kergoth | thatll be it then. can announce it |
| 00:12.14 | mickeyl | ok, will do. |
| 00:12.21 | zecke | kergoth: kick ban mickey for ~2 days please |
| 00:12.32 | mickeyl | heh |
| 00:12.40 | zecke | mickeyl: otherwise you won't leave |
| 00:13.05 | kergoth | :) |
| 00:13.08 | kergoth | just get it announced, then take off |
| 00:13.12 | zecke | my levels goes down into the cellar/basement I need to sleep now |
| 00:13.44 | mickeyl | g'night zecke |
| 00:17.00 | Pendalar | ok mickeyl for now I just copied the 3.5.2 area |
| 00:17.05 | mickeyl | ok thanks |
| 00:17.12 | Pendalar | I'll probably copy all, leave as root until we can work out something |
| 00:20.42 | RP | kergoth: I'm a bit worried about the change to the erase tool. Sharp's tools have some "special" bad block handling :-/ |
| 00:22.20 | kergoth | RP: for the rootfs? |
| 00:22.25 | kergoth | we're still using nandlogical for mtd1 |
| 00:22.35 | kergoth | this wasnt done for no reasno |
| 00:22.46 | kergoth | there were users that couldnt flash or use anything but sharp until this change was made |
| 00:22.53 | kergoth | thats why sash made the change |
| 00:24.32 | RP | What exactly was the change? |
| 00:24.48 | RP | Sharp use two badblock markers in the OOB data instead of the one.... |
| 00:25.03 | RP | which affects the rootfs |
| 00:25.07 | kergoth | dont know. ask sash |
| 00:25.40 | zecke | 3,2,1 suspend |
| 00:26.53 | kergoth | RP: interesting, that could definately be causing issues. if thats the cause, i dont see why using stock upstream eraseall wouldve helped anything |
| 00:27.26 | RP | neither do I. Unless it just masks the problem :-/ |
| 00:27.42 | RP | I added something to the 2.6 sharpsl driver to keep things compatible... |
| 00:28.45 | pb_ | eraseall shouldn't need to care about the bad block markers. an erased block is not bad, by definition. |
| 00:29.17 | RP | a bad block is one that won't erase properly isn't it? |
| 00:29.25 | pb_ | no, not necessarily |
| 00:30.00 | kergoth | pb_: mind if i switch PACKAGE_ARCH's default to be HOST rather than TARGET? |
| 00:30.02 | pb_ | a bad block is either one that was marked bad from the factory, or one that has experienced a program or erase error |
| 00:30.05 | pb_ | kergoth: go for it |
| 00:30.46 | pb_ | RP: for example, a block that had a bit stuck high would definitely be bad, but would appear to erase successfully |
| 00:30.48 | RP | but you shouldn't attepmpt to erase them then? |
| 00:31.13 | pb_ | correct, but that's not eraseall's problem. the mtd driver blocks attempts to erase bad blocks. |
| 00:31.42 | RP | ah. I assumed it was a low level util accessing the device directly |
| 00:32.04 | pb_ | no, it just does ioctl(MEMERASE) for each block. |
| 00:32.28 | RP | ok, fair enough... |
| 00:32.40 | RP | It could have been an issue... |
| 00:32.42 | RP | :) |
| 00:33.12 | RP | I've posted the dilema about the machine business to arm-linux... |
| 00:33.19 | pb_ | ok, jolly good |
| 00:33.25 | RP | I've tried to remain neutral ish |
| 00:33.35 | RP | I can see a problem though |
| 00:33.46 | mickeyl | Pendalar: one thing before I go to bed, can you check permissions on the mirror? i now get a 403 for everything below official/unstable/ |
| 00:33.58 | mickeyl | 3.5.2 announcement sent and made on website |
| 00:34.00 | mickeyl | g'night |
| 00:34.02 | Pendalar | I'm pulling as root so it's probably wack at |
| 00:34.05 | Pendalar | doh.. don't go heh |
| 00:34.15 | RP | The problem is memory size. We can't check early enough to work it out without different machine definitons AFAICT... |
| 00:34.21 | pb_ | Why not? |
| 00:35.04 | pb_ | Surely you can just do that in your fixup() routine. |
| 00:35.18 | RP | Is that called early enough? |
| 00:35.21 | pb_ | Yes. |
| 00:35.49 | RP | fair enough |
| 00:37.16 | pb_ | Check the code in setup_arch(). fixup() is almost the first thing that's called: it happens before commandline parsing, and definitely before the memory map is set up. |
| 00:37.32 | RP | You're right. I'm looking at it now |
| 00:38.11 | RP | I've actually tried that once before. It triggers a bug in the hdr handling code... |
| 00:38.45 | pb_ | What's the hdr? |
| 00:39.07 | RP | sorry, in the tag handlers |
| 00:39.21 | pb_ | ah |
| 00:39.25 | pb_ | okay, well, let's fix that bug. |
| 00:39.58 | RP | well, its a "bug". I think its a quick fix deliberate hack |
| 00:40.42 | Pendalar | [root@srv01 unstable]# du -sh -> 858M . |
| 00:40.46 | RP | squash_mem_tags in compat.c changes the mem tag to ATAG_NONE but leaves it with a length. The later code doesn't like that... |
| 00:40.58 | RP | I doubt it'll get fixed as its compat code |
| 00:41.01 | Pendalar | [pendalar@sc8-pr-shell1 unstable]$ du -sh => 763M . |
| 00:41.04 | Pendalar | ... |
| 00:41.08 | Pendalar | I'm a bit confused |
| 00:41.27 | RP | different sector size? |
| 00:41.45 | RP | Its a lot of difference for that though... |
| 00:41.57 | Pendalar | yeah |
| 00:42.05 | pb_ | stupid laptop |
| 00:42.08 | pb_ | later all |
| 00:42.09 | Pendalar | if sf.net's was bigger, I forgot to mirror something |
| 00:42.13 | Pendalar | but how could mine be bigger.. |
| 00:43.24 | MSpin | links not copied right? |
| 00:44.45 | emte | followed too many links |
| 00:44.53 | Pendalar | that was my first thought, I'm not sure how scp handles symlinks |
| 00:45.01 | Pendalar | it doesn't matter too much I guess |
| 00:45.15 | emte | might have duplicates as well |
| 00:45.26 | emte | if it hardlinks the remot symlinks |
| 00:45.31 | emte | remote* |
| 00:53.36 | RP | 'night all |
| 01:41.34 | *** join/#oe Virusmaster (~Thomas@p5091B93E.dip.t-dialin.net) |