**** BEGIN LOGGING AT Wed Dec 14 02:59:57 2011 **** ENDING LOGGING AT Wed Dec 14 06:20:29 2011 **** BEGIN LOGGING AT Wed Dec 14 06:22:15 2011 Dec 14 09:49:37 hello, I just released my work on a multiboot firmware for the Toshiba AC100 based on tegra 2 chip Dec 14 09:50:18 You may want to check my pull on https://gitorious.org/~tequila/ac100/tequilas-ac100-multiboot Dec 14 09:51:06 I had to modify a little kexecboot to make it useful for ac100 device Dec 14 09:52:11 Patch is there: https://gitorious.org/~tequila/ac100/tequilas-ac100-multiboot/blobs/master/patches/kexecboot-122708bc.patch Dec 14 09:52:38 with a readme at the end to explain why I made these modifications Dec 14 10:01:01 Tequila: thanks Dec 14 10:01:35 hey Jay7, thanks to you and your work ;) Dec 14 10:01:54 I'll review this and mainline your patches Dec 14 10:02:19 I think you may want to adapt it at little Dec 14 10:02:51 the main limitation is the 256 bytes cmdline limit Dec 14 10:02:58 I see Dec 14 10:03:15 and the second the forced "root=" Dec 14 10:03:44 you may override root in your config file Dec 14 10:04:09 second one should win Dec 14 10:04:22 hum okay Dec 14 10:04:42 btw, did kernel sources for toshiba define COMMAND_LINE_SIZE somewhere? Dec 14 10:04:46 I remember the 256 bytes limitation made me think the root= was forced... Dec 14 10:05:30 arm do it and the limitation is 1024 Dec 14 10:05:44 no, 'root=' is just first item in cmdline passed to kexec Dec 14 10:06:33 btw, what is root for toshiba? Dec 14 10:06:39 yes of course, I lost time to find the 256 limitations and I did the root= related patch before Dec 14 10:06:45 and why it doensn't detected properly? Dec 14 10:06:57 no it is detected properly Dec 14 10:07:13 my real trouble was the 256 bytes limitation Dec 14 10:07:16 ah Dec 14 10:07:27 only that fianally ;) Dec 14 10:07:46 try just increase default COMMAND_LINE_SIZE then in util.h Dec 14 10:07:51 you may forget the patch related to root= part Dec 14 10:08:16 ok :) Dec 14 10:08:37 of course, this should have been a simple fix Dec 14 10:09:05 hehe :) Dec 14 10:09:29 we should consider to increase it by default Dec 14 10:09:40 but I prefered a more flexible solution Dec 14 10:10:26 in my patch this becomes just a buffer size limit to read lines from conf and /proic/cmdline Dec 14 10:16:23 Tequila: btw, is GUI working ok on Toshiba? Dec 14 10:17:29 no... Dec 14 10:17:34 it is just perfect ;) Dec 14 10:17:43 fine :) Dec 14 10:18:11 I just modified the "KEXECBOOT" string in gui.c Dec 14 10:18:27 to extend it Dec 14 10:18:28 I see.. I'm thinking about making this configurable too.. Dec 14 10:18:52 It would be really nice Dec 14 10:19:49 anyway, I would ask you about testing kexecboot just with increased default COMMAND_LINE_SIZE in util.h Dec 14 10:19:57 or even with default Dec 14 10:20:26 if kernel headers provide right C_L_S it should work out of the box Dec 14 10:30:28 well as it is cross-compiled, I confirm it doesn't find the right value Dec 14 10:30:52 I really saw truncated cmdlines when I started to test Dec 14 10:31:36 and kernels didn't boot correctly until I extended the used buffer Dec 14 10:36:58 oups, in fact we really don't mind about COMMAND_LINE_SIZE in util.h as a short default one is sufficient for the ac100 Dec 14 10:39:16 btw we can also specify a cmdline to have debug output on the screen during the used kernel boot... Dec 14 10:40:13 but effectively I've to check if finally it takes the 1024 value Dec 14 10:40:37 so I'll put that test on on my TODO list... Dec 14 10:41:53 so to resume, at that time, only the 256 bytes limitation to read boot.cfg file lines is really critical for us Dec 14 10:45:20 ok Dec 14 10:45:35 we will consider to increase this default to 1024 Dec 14 10:45:55 and just an info if you have someone asking how our tegra can kexec with a 2.6.38 based kernel Dec 14 10:47:02 I essentialy delayed D-cache disabling and disabled SMP & THUMBEE Dec 14 10:47:54 so should user recompile kernel to use kexec? Dec 14 10:48:09 have you some manual? Dec 14 10:48:25 It would be nice to place link to it on kexecboot.org site Dec 14 10:54:10 still not Dec 14 10:54:47 but there will be one at http://ac100.grandou.net/ Dec 14 10:55:29 I'm currently building the firmware again from scratch to verify I didn't pushed any mistake Dec 14 10:55:41 and it tooks few hours Dec 14 10:56:49 oups, it's just finished :) just 50 minutes with a AMD x4 Dec 14 11:10:14 hehe :) Dec 14 11:10:28 I should go now Dec 14 11:10:40 * Jay7 afk Dec 14 17:09:39 Jay7: Here is the page related to my success with ac100 & tegra ;) http://ac100.grandou.net/multiboot Dec 14 19:30:37 Tequila: thanks **** ENDING LOGGING AT Thu Dec 15 02:59:57 2011