**** BEGIN LOGGING AT Thu Apr 05 02:59:56 2007 Apr 05 07:32:33 <[florian]> armijn: I flashed it via serial console, using adam2flash.pl Apr 05 08:25:11 [florian]: hmkay Apr 05 09:36:39 [florian]: any idea how I would do that *without* a serial console? :) Apr 05 09:37:08 ADAM2 is not listening on port 21 it seems Apr 05 09:37:18 at least, I get a connection, but it is closed immediately Apr 05 11:30:02 nbd * r6869 /trunk/ (3 files in 2 dirs): add missing modules for usb sound support on 2.6 (#1535) Apr 05 13:23:18 hcg * r6870 /trunk/target/linux/at91-2.6/ (10 files in 6 dirs): Apr 05 13:23:18 Added module to generate u-boot config file. Added platform overrides Apr 05 13:23:18 for base files. Apr 05 14:29:39 <[florian]> armijn: no idea Apr 05 14:29:46 <[florian]> armijn: I just interupted the boot with serial console Apr 05 14:31:02 armijn: On my D-Link DSL-G604T, you need to FTP in and issue the USER command within 7 seconds of ADAM2 starting to stop the boot. Apr 05 14:36:34 farnz: that doesn't work on the C54APRA unfortunately Apr 05 14:36:34 [florian]: i think i have the 2.6 map driver ready, except for the Kconfig additions -- will test Apr 05 14:36:54 [florian]: ok, will have to go all serial then :) Apr 05 14:39:20 armijn: do you have shell access with stock firmware? then cat /proc/ticfg/env and see if autoload and autoload_timeout are set Apr 05 14:40:43 hm Apr 05 14:40:49 I have shell access Apr 05 14:40:53 well, not right now :) Apr 05 15:15:04 <[florian]> sn9_: ok, test it and tell me :) Apr 05 15:22:34 http://uejikov.g-service.ru/rootfs_split.diff <- any objections? Apr 05 15:22:39 nbd: ^^ Apr 05 15:23:11 ejka: looks good Apr 05 20:25:49 [florian]: well, i get a null dereference somewhere Apr 05 20:26:22 there are no symbols in the backtrace, so i can only guess where Apr 05 22:08:38 nbd * r6871 /packages/utils/mksh/Makefile: fix mksh compile with ccache Apr 05 22:09:32 seems [mbm]'s fixes to find -exec cause much breakage Apr 05 22:15:59 hm Apr 05 22:16:13 mtd0 0x90091000,0x903f0000 Apr 05 22:16:13 mtd1 0x90010090,0x90090000 Apr 05 22:16:13 mtd2 0x90000000,0x90010000 Apr 05 22:16:13 mtd3 0x903f0000,0x90400000 Apr 05 22:16:13 mtd4 0x90010000,0x903f0000 Apr 05 22:16:27 the start value of mtd1 is somewhat odd...it seems Apr 05 22:16:29 this is AR7 btw Apr 05 22:17:05 I mean, I would have expected that mtd1 would start at 0x90010000 Apr 05 22:17:19 at least, if I read the AR7 installation instructions on the weiki Apr 05 22:17:19 mtd4 Apr 05 22:17:22 ehr, wiki Apr 05 22:17:56 it goes 2, 4, 3 Apr 05 22:18:23 for booting? Apr 05 22:18:27 1 and 0 are overlays Apr 05 22:18:39 hmkay Apr 05 22:18:45 this is all kinda new for me :) Apr 05 22:19:02 right, will try Apr 05 22:19:13 i would guess 2 to be adam2, and 4 to be the nvram config Apr 05 22:19:55 1 is likely the kernel, and 0 squashfs Apr 05 22:20:19 both read-only, with 3 used for flashing Apr 05 22:21:09 at least that's what i gather from your paste. i have not looked at any reference for your model Apr 05 22:21:55 oops, i meant 3 is the nvram config, not 4 Apr 05 22:22:15 4 is for flashing Apr 05 22:22:45 * sn9_ considers making the font bigger Apr 05 22:53:17 ok Apr 05 22:53:25 after a lot of fuzzing: Apr 05 22:53:42 change the start value of mtd1 to the start value of mtd4 Apr 05 22:54:32 now I don't know if mtd0 should start where mtd1 ends Apr 05 22:54:43 hoping [florian] can help with some insight there :) Apr 05 22:55:06 it all depends on your adam2 settings Apr 05 22:55:13 and the firmware, I suppose Apr 05 22:55:27 not really Apr 05 22:55:31 why not? Apr 05 22:55:43 I flashed the result of the OpenWrt build process to mtd4 Apr 05 22:55:46 the bootloader decides where in the flash the kernel is Apr 05 22:56:32 so that would be mtd1 (very default adam2) Apr 05 22:56:50 but it doesn't use mtd0 then at all? Apr 05 22:56:54 adam2flash.pl flashes openwrt on devices that don't have a combined kernel+rootfs partition Apr 05 22:57:13 nbd: this one apparently does Apr 05 22:57:19 mtd4 Apr 05 22:57:21 yeah Apr 05 22:57:22 well, there already was a mtd4 partition Apr 05 22:57:22 i see it Apr 05 22:57:34 this is the conceptronic c54apra btw Apr 05 22:59:08 I'm officially confused now Apr 05 22:59:30 why would I need to change the offsets for mtd1 for it to work? Apr 05 22:59:38 because that is what I had to do Apr 05 23:01:01 you could change adam2 settings instead Apr 05 23:01:09 i doubt it matters which Apr 05 23:01:22 well, I changed it in ADAM2 via FTP Apr 05 23:01:37 i mean the variables Apr 05 23:03:04 *rewind* Apr 05 23:03:43 which partition is used by ADAM2 to find the kernel? Apr 05 23:03:58 is that mtd4, or mtd1? Apr 05 23:04:16 i would assume that depends on the variables Apr 05 23:04:28 it can probably use either Apr 05 23:04:33 ok Apr 05 23:05:28 then the question is, where can I see those variables? Apr 05 23:06:08 and specifically, which variable is of interest Apr 05 23:07:11 hm Apr 05 23:07:12 http://ar7.wikispaces.com/ADAM2 Apr 05 23:07:14 quote Apr 05 23:07:29 "ADAM2 executes "go" command automatically to run a mtd1 content if there is no activity on the Serial console, "autoload" is true and "autoload_timeout" exceeded. Usually, the mtd1 (0x90010000) contains a compressed Linux kernel image." Apr 05 23:07:45 Engrish is (c) of that page Apr 05 23:09:34 you might still be able to use adam2flash.pl Apr 05 23:09:44 well, it works now, so I guess I'm happy Apr 05 23:10:16 just not entirely sure if mtd0 should start where mtd1 ends with the OpenWrt image Apr 05 23:10:57 do you want to use a squashfs image or a jffs2-only one? Apr 05 23:11:22 well, [florian] said I should just use squasfs for now Apr 05 23:12:35 gah...brainrot Apr 05 23:13:14 with squash, mtd0 can probably start anywhere Apr 05 23:13:48 it's read-only anyway Apr 05 23:14:45 well, let's see if a netgear dg834g also wants to work Apr 05 23:15:05 i have one of those -- a v2 Apr 05 23:15:37 its partition layout is much easier to handle Apr 05 23:20:18 is it? Apr 05 23:20:30 mtd0 0x900d0000,0x903e0000 Apr 05 23:20:30 mtd1 0x90020000,0x900d0000 Apr 05 23:20:30 mtd2 0x90000000,0x90020000 Apr 05 23:20:30 mtd3 0x903e0000,0x903f0000 Apr 05 23:20:30 mtd4 0x903f0000,0x90400000 Apr 05 23:20:45 that is from the device Apr 05 23:20:46 see? no weird offsets Apr 05 23:20:48 but it is a v3 Apr 05 23:20:52 ah Apr 05 23:21:01 the v3 is different Apr 05 23:21:16 so it seems :) Apr 05 23:22:44 ah, the dg834g page on the wiki explains it **** ENDING LOGGING AT Fri Apr 06 03:00:00 2007