**** BEGIN LOGGING AT Tue Oct 14 02:59:57 2008 Oct 14 08:21:59 it seems that IR is not on PXA UART ?! :( Oct 14 08:22:01 gm Oct 14 08:29:41 hi Oct 14 08:30:55 <|miska|> hi Oct 14 08:31:34 BobOfDoom: have you found where is Ir attached to? Oct 14 08:34:06 Sleep_Walker: nope, I haven't looked at IR as I don't have any use for it Oct 14 08:34:26 I was assuming STUART Oct 14 08:34:32 but that's just by process of elimination Oct 14 08:35:05 I'm confused that there was no GPIO activity during sendign through IR Oct 14 08:35:10 but maybe just Marex Oct 14 08:35:26 but maybe just Marex's tool is not capable to be in the middle of sending Oct 14 08:35:57 yeah Oct 14 08:36:11 I think it's likely sending happens without scheduling happening Oct 14 08:36:19 ey guys Oct 14 08:36:26 can you help in one stupid error? Oct 14 08:36:28 blackcookie: hi Oct 14 08:36:49 can i paste a litlle error? Oct 14 08:37:03 i never use MySQL Oct 14 08:37:21 er what? Oct 14 08:37:21 paste? flood! flood!! flood!!!!! Oct 14 08:37:31 and i need to download plugins for msf (metasploid) Oct 14 08:37:39 blackcookie: read the topic Oct 14 08:37:45 this is not a cracking channel Oct 14 08:37:46 waht is flood? i dont speak inglish very well Oct 14 08:37:51 yeah Oct 14 08:37:55 so sorry Oct 14 08:37:58 no metasploit here Oct 14 08:37:58 but i dont wana crack nothing Oct 14 08:38:06 only i wana install plugins Oct 14 08:38:39 i am learning unix alone and is litlle fustred sometimes Oct 14 08:38:40 LOL Oct 14 08:38:49 msf is for segururity to Oct 14 08:38:51 well Oct 14 08:38:52 right, but you're in the wrong place Oct 14 08:38:55 we aren't security people Oct 14 08:39:15 but you know about linux? Oct 14 08:39:39 the problem is with the mysql Oct 14 08:39:46 then join #mysql Oct 14 08:40:02 this is not databases channel either Oct 14 08:43:26 :D Oct 14 08:46:38 Marex: do you have idea how to find GPIOs used (and thus UART) for BT and IR? Oct 14 08:46:48 (using your tool) Oct 14 08:48:44 btw. you should commit GPIOtrace into http://git.hackndev.com/?p=tools;a=summary Oct 14 08:48:55 it's quite usefull tool Oct 14 08:49:01 blackcookie: sorry to be harsh dude, we get a lot of people coming in asking how to crack gmail accounts and whatever. :-) We probably can't be of much use abuot metasploit/mysql so you're better off asking people who know it ;) Oct 14 08:50:11 i understand you Oct 14 08:50:28 always ask if somebody have a good exploi ft hotmail, etc Oct 14 08:50:49 now i found a chanel for this Oct 14 08:50:53 jejeje Oct 14 08:51:17 Sleep_Walker: simple way, mess them up and see if they stop working! Oct 14 08:51:21 but ubuntu is very dificult to install and configure Oct 14 08:52:10 BobOfDoom: hm, I've got some strange activity on GPIO47 Oct 14 08:52:21 Sleep_Walker: is it an input or output? Oct 14 08:52:26 Sleep_Walker: is it an alternate function? Oct 14 08:52:44 it's changing along with GPIO9 Oct 14 08:52:50 like it is connected Oct 14 08:52:58 9 is an output? Oct 14 08:53:09 gimme a second :) Oct 14 08:53:11 k Oct 14 08:53:19 I have to reparse GPIO.txt for another register :b Oct 14 08:54:59 direction of 47 is changing too Oct 14 08:55:22 for 9 it is stable Oct 14 08:55:30 oh.. Oct 14 08:55:40 that's very interesting Oct 14 08:55:42 0 Oct 14 08:55:48 maybe it's the battery Oct 14 08:55:56 AF = 0 you mean? Oct 14 08:56:02 DR Oct 14 08:56:32 for GARF_L and GAFR_U I have to reparse again :b Oct 14 08:56:32 actually Oct 14 08:56:40 that gpio trace may be just what I need Oct 14 08:56:53 Sleep_Walker: doh.. you should make a table of all the inputs and outputs and alternate functions Oct 14 08:57:11 I'm working on it Oct 14 08:57:21 Sleep_Walker: it's really useful.. cause otherwise you might be looking at an LCD pin or something useless like that Oct 14 08:57:44 BobOfDoom: I have list of already working and assigned GPIOs from Linux Oct 14 08:58:16 ah ok Oct 14 08:58:49 hm, I should enhance my parser for GAFR_L and GAFR_U to respect that two bit per GPIO stuff Oct 14 08:59:46 GAFR_{L,U} is not changing Oct 14 09:01:03 and it is probably set to 1 or 2 (I'm not sure with bit order for this now) Oct 14 09:01:07 gotta go working Oct 14 09:01:37 ok, cya Oct 14 09:09:36 hello Oct 14 09:11:50 Genesis: hi Oct 14 09:17:52 BobOfDoom: ha, GPIO57 is the right ONE, Ir first success Oct 14 09:21:33 Sleep_Walker: when you have a moment could you send me gpiotrace? Marex gave me a copy ages ago but I can't find it :( Oct 14 09:22:02 BobOfDoom: sutre Oct 14 09:22:04 sure Oct 14 09:22:08 IrDA works now Oct 14 09:22:11 Sleep_Walker: thanks Oct 14 09:22:21 Sleep_Walker: cool, well done! Oct 14 09:22:32 Sleep_Walker: now you are beating me :D Oct 14 09:22:53 hey, I don't have working suspend yet! Oct 14 09:22:57 stay calm :b Oct 14 09:23:11 I haven't fixed suspend in 2.6.27 yet either :p Oct 14 09:23:23 I'm working against -hnd Oct 14 09:23:46 but if you don't mind, I can join your playground Oct 14 09:24:09 http://sleepwalker.hackndev.com/gpiotrace.tar.bz2 Oct 14 09:24:23 http://sleepwalker.hackndev.com/gpiotrace.sh Oct 14 09:24:44 that shell script is slow parser, but producing quite nice output Oct 14 09:25:03 oh Oct 14 09:25:05 Sleep_Walker: thanks muchly Oct 14 09:25:27 http://sleepwalker.hackndev.com/hex2bin.sh Oct 14 09:25:39 my conversion library Oct 14 09:25:42 :b Oct 14 09:25:45 Sleep_Walker: sure if you like, feel free to commit into ato/linux-2.6-arm or create a new branch in linux-2.6 Oct 14 09:26:05 Sleep_Walker: I'd like to keep the "stable" branch in linux-2.6 one where nothing breaks Oct 14 09:26:06 * Sleep_Walker doesn't know GIT so much :b Oct 14 09:26:13 ah Oct 14 09:26:24 gotta really go now! Oct 14 09:26:26 well start playing with ato/linux-2.6-arm since it doesn't matter if you break stuff in there Oct 14 09:26:30 ok Oct 14 09:26:33 cya! Oct 14 09:35:51 ato tools * r1e9a84342cdd /gpiotrace/ (20 files in 5 dirs): Added Marex's nifty GPIO trace utility. Oct 14 09:35:52 ato tools * r8e2068ca9265 / (asic6-gpio.c devmem2.c dumpmem.c gpio-dir.c gpio-val.c): Merge branch 'master' of hackndev.com:/srv/git/tools Oct 14 10:35:53 Hi there Oct 14 10:36:40 hi here Oct 14 10:37:30 BobOfDoom: If you like that gpiotrace, you can add there notification remove if there already is Oct 14 10:37:38 so it will switch off itself Oct 14 10:38:11 is there any good tutorial about installing openmoko on a palm tx ? Oct 14 10:38:29 I found http://old.hackndev.com/node/701, but it's not detailled enough for me :/ Oct 14 10:44:16 BobOfDoom: how can I objdump BL? Oct 14 10:51:27 Sleep_Walker: get a dump of it... dumpmem 0x0 > bl.bin should do, or you can grab it from pdb in rom update, use unpdb.py in tools to chop the header off Oct 14 10:51:43 Sleep_Walker: then... Oct 14 10:51:43 I've got already bin Oct 14 10:51:53 (from custom ROM stuff :) Oct 14 10:52:16 btw. there is firmware for GSM :) can it be useful somehow? :b Oct 14 10:53:41 Sleep_Walker: objdump -b binary -m arm -D ipl.bin Oct 14 10:54:11 ohh, great! thx Oct 14 10:55:19 Sleep_Walker: I dunno whether GSM firmware is useful. Can run strings on it I suppose and get a command list or something Oct 14 10:55:33 but you also get a command list with AT* Oct 14 10:55:41 maybe it has hidden commands though I guess Oct 14 10:56:14 Sp4rKy: which openmoko? Oct 14 10:56:24 Sp4rKy: Om2008.9 ? Oct 14 10:57:48 Sp4rKy: raster has an openmoko/illume image for treo 650, it should work on TX as well, just replace the zImage with one for the TX: http://www.rasterman.com/ Oct 14 10:59:20 BobOfDoom: hmmm, 0xa0000000 is not there listed anywhere :/ Oct 14 10:59:43 which instruction can be used for leaving bootloader and continue booting? Oct 14 11:00:07 Sleep_Walker: ah really? it is in 650 one Oct 14 11:00:28 Sleep_Walker: do you mean on resume from suspend? Oct 14 11:00:37 yes Oct 14 11:01:07 you use `ldr' in palmt650_pm Oct 14 11:01:45 ah.. well it's no speicial instruction Oct 14 11:01:51 all it has to do is jump into ram Oct 14 11:02:04 hang on, lemme see what it looks like in 650 Oct 14 11:02:25 2a4: 13a004a0 movne r0, #-1610612736 ; 0xa0000000 Oct 14 11:02:25 2a8: 1b000133 blne 0x77c Oct 14 11:02:28 and then.. Oct 14 11:02:44 77c: e1a0f000 mov pc, r0 Oct 14 11:03:21 so that means load 0xa0000000 into r0, jump to 0x77c, which then puts r0 into the program counter (ie jumps to 0xa0000000) Oct 14 11:04:54 BobOfDoom: ok, i'll try Oct 14 11:05:06 BobOfDoom: does wifi works with openmoko + palm TX ? Oct 14 11:06:38 hm, when I have undefined instruction, it can mean that I have bad dump, right? :( Oct 14 11:06:44 Sp4rKy: I assume so, but there's probably no GUI to configure it Oct 14 11:07:05 Sp4rKy: most of the openmoko GUI stuff that controls wifi, bluetooth etc probably won't work without being tweaked for YX Oct 14 11:07:28 Sleep_Walker: nah Oct 14 11:07:33 Sleep_Walker: it just means there's data Oct 14 11:07:45 Sleep_Walker: cause objdump can't tell the difference between code and data Oct 14 11:07:56 so it tries to disassemble the data too Oct 14 11:07:59 BobOfDoom: how are your homies? Oct 14 11:08:15 I see Oct 14 11:09:03 raster: they're sleeping. So now would be a good time to muscle in on our turf Oct 14 11:09:29 * raster musters the peeps Oct 14 11:10:44 Sleep_Walker: raster and I are having a gang war: http://www.rasterman.com/files/the-gang-02.jpg http://meshy.org/~ato/tmp/rival-gang.jpg Oct 14 11:10:48 Sleep_Walker: care to join? :p Oct 14 11:10:52 * raster checks up on now to get nis shiny new oe+e+illume image on his n800 Oct 14 11:11:44 BobOfDoom: sure, you want me to send you photos and you'll copy & paste it to your table? :b Oct 14 11:11:53 Sleep_Walker: haha :D Oct 14 11:13:05 BobOfDoom: is there any step-to-step howto ? this is my first Om installation Oct 14 11:14:55 Sp4rKy: are you comfortable with Linux? Like can you follow instructions such as: partition an SD card into one small FAT partition and a larger ext2 partition? Oct 14 11:15:12 Sp4rKy: I don't think there's much in the way of user-friendly instructions Oct 14 11:15:52 Sp4rKy: http://download.enlightenment.org/misc/Illume/Treo-650/2008-10-12/README Oct 14 11:16:28 Sp4rKy: see if that makes any sense to you Oct 14 11:16:46 ok Oct 14 11:16:55 and yes, I'm comfortable with linux :) Oct 14 11:17:18 excellent Oct 14 11:17:43 so just follow that README but replace the zImage with one for a TX Oct 14 11:17:47 hmm, this readme should be useful Oct 14 11:17:51 ok Oct 14 11:18:32 and then, If I plug the card, my tx will boot on Om, and if I unplug it, the TX will boot on default PalmOs ? Oct 14 11:19:28 Sp4rKy: nope, it'll always boot to Palm OS and if you want to boot Om you just run the "Cocoboot" program (which following that README will install on the card) Oct 14 11:20:21 ok :D Oct 14 11:31:04 morning Oct 14 11:37:55 BobOfDoom: i'm sorry to ask you for so stupid question, but I don't find the zImage for TX Oct 14 11:46:05 Sp4rKy: http://atrey.karlin.mff.cuni.cz/~miska/ Oct 14 11:46:15 Sp4rKy: grab one of the kernel packages Oct 14 11:47:05 Sleep_Walker: http://meshy.org/~ato/tmp/gpiotrace.html Oct 14 11:47:38 BobOfDoom: cool Oct 14 11:47:59 what about dynamic version? I'll submit GPIO.txt and it will show this... Oct 14 11:49:23 ato tools * r8048b948c1b1 /gpiotrace/gpiotrace2html: gpiotrace2html: GPIO trace parser with coloured HTML table output. Oct 14 11:49:48 Sleep_Walker: grab that ^ run gpiotrace2html GPIO.txt > gpio.html Oct 14 11:49:53 BobOfDoom: it works :D Oct 14 11:50:00 Sp4rKy: congrats :D Oct 14 11:50:03 (for the moment I have the splash) Oct 14 11:50:17 fine, thanks :) Oct 14 11:50:32 I love enlightenment on my laptop, now I have it on my palm :D Oct 14 11:56:58 I love KDE3 :b Oct 14 11:57:01 damn Oct 14 11:58:23 damn? Oct 14 12:02:10 KDE 3.x required RAM: 512mb required CPU: 1 GHz Oct 14 12:02:15 ah, that is damn :p Oct 14 12:03:26 I've got working suspend! Oct 14 12:03:31 0xc0000000 Oct 14 12:03:32 :) Oct 14 12:04:13 c.. Oct 14 12:04:20 wtf is c? Oct 14 12:04:22 hm Oct 14 12:04:48 reserved (64 Mbyte) 0xC000_0000 Oct 14 12:04:56 what the heck? Oct 14 12:06:49 MUHAHAHAHHAHAHAHAHAHAHHAHAHA Oct 14 12:07:34 http://www.rasterman.com/files/the-gang-03.jpg Oct 14 12:09:17 * BobOfDoom quakes with fear Oct 14 12:10:06 my peeps are muscling in Oct 14 12:10:15 the arm farm grows Oct 14 12:14:43 BobOfDoom: so 32MB 0xa000_0000 and 32MB 0xc000_0000? Oct 14 12:14:52 or I can't understand that... Oct 14 12:15:03 or mem starts at 0xc000_0000?! Oct 14 12:15:08 I can't understand it Oct 14 12:15:15 nothinf should start at 0xc000 0000 Oct 14 12:15:19 writing there should do nothing Oct 14 12:15:21 it's reserved Oct 14 12:15:32 (if we're talking physical addresses) Oct 14 12:15:44 it's not even memmapped in PalmOS Oct 14 12:15:58 but you wrote there and it fixed suspend? Oct 14 12:16:01 in linux? Oct 14 12:16:02 BL can't use virtual address, can it? Oct 14 12:16:07 BobOfDoom: yes Oct 14 12:16:12 it simply works now Oct 14 12:16:23 Sleep_Walker: well.. probably not... it can turn on the MMU and setup virtual addreses Oct 14 12:16:33 but I doubt it'd do that Oct 14 12:16:39 hmm Oct 14 12:17:02 Sleep_Walker: can you show me the code you used in linux? Oct 14 12:20:35 it is like cat palmt650_pm.c | sed 's/t650/t680/' | sed 's/0xa0000000/0xc0000000/' Oct 14 12:20:50 .. Oct 14 12:20:52 waoh Oct 14 12:20:59 hmmmmmmmmmmmmmmmm Oct 14 12:22:38 I guess.. Oct 14 12:22:40 ah... Oct 14 12:22:43 I wonder Oct 14 12:22:55 resume_vector = phys_to_virt(0xa0000000); Oct 14 12:23:12 ...? Oct 14 12:23:50 . Oct 14 12:23:59 hmm Oct 14 12:24:16 I guess that must be returning something valid Oct 14 12:24:23 when you give it 0xc0000000 Oct 14 12:25:41 printk("resume vector: %x\n", virt_to_phys(phys_to_virt(0xc0000000))); Oct 14 12:25:44 might be enlightening Oct 14 12:25:48 but anyway it's great it works Oct 14 12:25:53 even if it defies logic :D Oct 14 12:26:40 wtf?! Oct 14 12:26:43 Counting objects: 13, done. Oct 14 12:26:45 Compressing objects: 100% (7/7), done. Oct 14 12:26:46 Writing objects: 100% (7/7), 678 bytes, done. Oct 14 12:26:48 Total 7 (delta 6), reused 0 (delta 0) Oct 14 12:26:49 error: hooks/update exited with error code 126 Oct 14 12:26:51 error: hook declined to update refs/heads/master Oct 14 12:26:52 To git+ssh://sleep_walker@git.hackndev.com/srv/git/linux-hnd Oct 14 12:26:54 ! [remote rejected] master -> master (hook declined) Oct 14 12:26:56 error: failed to push some refs to 'git+ssh://sleep_walker@git.hackndev.com/srv/git/linux-hnd' Oct 14 12:27:53 ooh Oct 14 12:27:59 I b0rked it I guess Oct 14 12:28:01 just a sec Oct 14 12:29:12 Sleep_Walker: try again.. Oct 14 12:29:17 Sleep_Walker: I re-enabled CIA Oct 14 12:29:30 if it fails again I'll turn it off Oct 14 12:30:37 WOOT! Oct 14 12:30:42 my zaurus data cable Oct 14 12:30:44 joy is me Oct 14 12:31:00 raster: congrats! :) Oct 14 12:31:12 no need to rely on ebay Oct 14 12:31:22 i'm runing out of usb ports... Oct 14 12:31:31 and i already have 10 Oct 14 12:33:12 ato tools * r335dd76ae314 /gpiotrace/ (gpiotrace2html palmt650.gpios): gpiotrace2html: read list of user-defined GPIO names from a file. Oct 14 12:40:41 ato tools * r649205eeea9e /gpiotrace/gpiotrace2html: gpiotrace2html: Show changes in AF. Oct 14 12:53:22 Sp4rKy: how did you go? did it get past the splash? Oct 14 13:06:03 yes Oct 14 13:06:08 it works pretty correctly Oct 14 13:07:06 Sp4rKy: awesome :) Oct 14 13:09:40 :) Oct 14 13:09:43 ato tools * r8589f20b4957 /gpiotrace/ (gpiotrace.prc gpiotrace2html include/regs.h m68k/gpiotrace.c): gpiotrace: monitor pxa27x extended GPIOs as well. Oct 14 13:10:17 Sleep_Walker: ^ may be useful Oct 14 13:13:26 BobOfDoom: false alarm, still not working Oct 14 13:13:33 It worked thrice Oct 14 13:13:42 that's weird Oct 14 13:16:37 yup Oct 14 13:16:56 I add there that address output Oct 14 13:17:15 btw. it's weird widely :b Oct 14 13:28:43 ato tools * r80b5d31ed0cc /gpiotrace/ (gpiotrace.prc m68k/gpiotrace.c): gpiotrace: allow disabling of the hack by re-executing it. Oct 14 13:29:07 Sleep_Walker: ^ gift for you Oct 14 13:29:21 great :o) Oct 14 13:34:50 ato tools * r09eb02194c27 /gpiotrace/ (gpiotrace2html palmt650.gpios): added USB pullup Oct 14 13:56:57 night all Oct 14 13:57:00 gn Oct 14 13:57:05 see you today :) Oct 14 13:58:35 <|miska|> gn Oct 14 14:21:42 * Genesis 's ebaying a treo650 after bye 2 palm Tungsten E Oct 14 14:29:19 yop rzr :) Oct 14 14:30:05 :) Oct 14 16:33:54 BobOfDoom: is possible in arm mode to access any memory in palm? Oct 14 16:34:16 BobOfDoom: isn't this way to suppress kernel+initrd size limit? Oct 14 16:34:43 just to write to some memory area and don't let PalmOS to get angry before boot? Oct 14 16:41:51 nah, in arm mode you dont have some areas of memory protected iirc Oct 14 16:42:19 you can happen to get reset if accessing memory in emu68k mode Oct 14 16:45:58 marex: and has this code stable address? Oct 14 16:46:33 I can imagine device specific memory map to find the largest hole... Oct 14 16:52:15 well you can dump whole ram Oct 14 16:52:21 even whole address space Oct 14 17:00:53 hm, if I wasn't associal guy, I would think it was some sort of sarcasm Oct 14 18:08:54 omfg Oct 14 18:09:15 * Sleep_Walker hits wall with his head Oct 14 19:01:58 <_hc> anyone here ever connected to a Palm TX using a serial terminal? Oct 14 19:02:16 <_hc> I am trying to find out to make my own cable, I have the athena connector and RS232... Oct 14 19:04:10 |miska|: ? ^ Oct 14 19:12:29 * |miska| didn't Oct 14 19:12:48 <|miska|> But I have connector scheme if you want Oct 14 19:15:19 BobOfDoom: wtf gsm_shutoff? Oct 14 19:27:01 <_hc> |miska|: that would be a great start Oct 14 19:27:38 <_hc> I am sitting here with the parts and a warmed up soldering iron, so once I get that pic, I'll start soldering! :D Oct 14 19:32:25 <_hc> hmm, wikipedia has a nice looking page on that connector: http://en.wikipedia.org/wiki/Athena_Connector Oct 14 19:34:34 <|miska|> hmmm, I'm affraid I've got nearly the same thing Oct 14 19:35:49 <|miska|> http://www.pdasoft.cz/files/root/obrazky_k_clankum/2005/tx_recorder/2.png Oct 14 19:37:51 omfg Oct 14 19:38:17 there is audio in/out in that connectore? Oct 14 19:38:31 s/e\?/\?/ Oct 14 19:38:37 <_hc> yup Oct 14 19:38:47 <_hc> that's how the various mic hacks work Oct 14 19:42:53 <|miska|> Sleep_Walker: yep, I've got handsfree for my TX Oct 14 19:43:15 |miska|: don't say me that you solder there BCM2132 :b Oct 14 19:43:32 <|miska|> It's analog audio Oct 14 19:43:44 <|miska|> Sleep_Walker: http://www.pdasoft.cz/modules.php?name=News&file=article&sid=2133 Oct 14 19:44:24 wow Oct 14 19:44:55 <|miska|> I think I still got it somewhere Oct 14 19:45:00 it reminds me English lessons and teach asking "What do you do in your free time?" :b Oct 14 19:45:08 s/teach/teacher/ Oct 14 19:46:15 <|miska|> snua has much cooler things to show Oct 14 19:47:16 Treo650 and Treo680 has completely different bootloader :( Oct 14 19:47:57 <|miska|> :-( Oct 14 20:12:05 <_hc> anyone got sound out of the speaker on the TX? Oct 14 20:12:13 <_hc> I haven't gotten it going yet, headphones yes Oct 14 20:13:08 <|miska|> In Linux? Oct 14 20:13:25 <_hc> right Oct 14 20:13:37 <_hc> works fine in Palm, I use it all the time Oct 14 20:13:46 <|miska|> It should be set right in linuxexpo release Oct 14 20:14:24 <|miska|> You need to adjust balance, I think left only with hackndev kernel Oct 14 20:14:55 <_hc> like with alsamixer? Oct 14 20:18:50 <_hc> also, random question, is there a way to power off the TX from the command line? Oct 14 20:19:04 <_hc> I use init 0 and power off, and neither actually turns it off in the ene Oct 14 20:19:06 <_hc> end Oct 14 20:19:17 <_hc> oops poweroff Oct 14 20:23:52 <|miska|> I think something like apm --suspend Oct 14 20:24:34 BobOfDoom: please, could you send me saved alsamixer state for playing with speaker? Oct 14 21:31:29 <_hc> BobOfDoom: alsastate for the speaker would rock! Oct 14 21:32:47 <_hc> |miska|: that suspends it, I mean more like shutdown and poweroff, so I can use PalmOS again Oct 14 22:28:24 _hc, push reset Oct 14 22:28:44 _hc, you cant poweroff the device since the CPU doesnt support it Oct 14 22:28:59 or rather ... the only way to do so is to desolder battery Oct 14 22:29:25 you can halt the cpu as well, but that wont be a real poweroff Oct 14 23:21:09 BobOfDoom: good morning :b Oct 14 23:27:52 Sleep_Walker: wakie wakie. hands off.... Oct 14 23:28:54 :b Oct 14 23:34:38 Sleep_Walker> BobOfDoom: isn't this way to suppress kernel+initrd size limit? Oct 14 23:34:39 nah Oct 14 23:34:43 well Oct 14 23:34:44 sort of Oct 14 23:35:00 if you want to write MMC driver from scratch Oct 14 23:35:13 cause we can't use PalmOS MMC driver from ARM mode Oct 14 23:35:22 if we're overwriting RAM randomly Oct 14 23:35:31 I thought that :b Oct 14 23:35:47 but I suppose you could load u-boot and then get that to read the MMC Oct 14 23:35:54 but I never got u-boot to work on pxa27x Oct 14 23:36:00 why bother Oct 14 23:36:06 palm os is a perfectly great bootloader Oct 14 23:36:11 Sleep_Walker> BobOfDoom: wtf gsm_shutoff? Oct 14 23:36:29 when high it shuts off the GSM :P Oct 14 23:36:41 might be a power control of some sort Oct 14 23:36:53 it's different to the other ones I've labeled gsm+power Oct 14 23:37:10 cause when you take it low again the chip is still dead Oct 14 23:37:23 raster: you don't use large kernels+initrd much, right? :b Oct 14 23:37:26 you have to toggle one of the other power ones to get it alive Oct 14 23:37:33 so that might mean shutoff is power and the others are reset Oct 14 23:37:34 but who knows Oct 14 23:37:34 Sleep_Walker: no. but why do u need a large initrd Oct 14 23:37:40 the only thing an initrd shoudl do Oct 14 23:37:43 BobOfDoom: I see Oct 14 23:37:45 if ANYTHING is get root mounted Oct 14 23:38:06 and thats not needed if mmc driver is in kernel and root is a partition on sd Oct 14 23:38:11 which frankly is double the performance for reads Oct 14 23:38:19 I finally got dump from putting phone on Oct 14 23:38:19 and 6-7x the perf for writes Oct 14 23:39:08 it made such a MASSIVe difference in speed - i'm never going back to ext2 on file image on fat16 Oct 14 23:39:10 raster: for example for kEdAR's multidevice kernel Oct 14 23:39:34 hmmmm Oct 14 23:39:49 * raster ponders the usefulness of a multi-device kernel Oct 14 23:39:52 one kernel rules them all :b Oct 14 23:39:57 * raster dismisses Oct 14 23:40:00 :) Oct 14 23:40:15 here's how i see it Oct 14 23:40:23 kernel probably shoudl be per-device Oct 14 23:40:24 imho Oct 14 23:40:59 * BobOfDoom finds it useful Oct 14 23:41:09 can just take the SD card from homie and give it to another Oct 14 23:41:12 sharing is caring :p Oct 14 23:41:23 but I guess most people don't have gngs Oct 14 23:41:27 gangs* Oct 14 23:41:30 BobOfDoom: do you remember what was the problem of inconsistency between http:// git and git:// ? Oct 14 23:41:43 BobOfDoom: HAHAHAHA Oct 14 23:41:47 tools repo suffers from that Oct 14 23:42:00 something bad on mr zeee Oct 14 23:42:02 well, I know about ... 5? ;b Oct 14 23:42:05 after a while he locks up Oct 14 23:42:14 Sleep_Walker: yeah http:// sucks. it's cause git-update-server-info isn't being run Oct 14 23:42:25 fcuk Oct 14 23:42:27 bad mr zeee! bad! Oct 14 23:42:34 Sleep_Walker: hang on, i'll fix it Oct 14 23:42:48 BobOfDoom: dunno. i just dont like the extra complexity of an initrd Oct 14 23:42:59 frankly.. i dont much like the separate kernel + rootfs really Oct 14 23:43:08 i like the idea of "one big blob" Oct 14 23:43:21 for example kernel could live at the start of it Oct 14 23:43:24 rotfs after that Oct 14 23:43:28 like kernel +initrd Oct 14 23:43:37 but loader would ned to know to only load kernel section into ram Oct 14 23:43:38 BobOfDoom: if you don't teach me how to fix it, I'll bother you again :b Oct 14 23:43:44 raster: I don't like initrds either.. so my gang-kernel has no initrd Oct 14 23:43:59 Sleep_Walker: alright Oct 14 23:44:09 Sleep_Walker: ssh hackndev.com Oct 14 23:44:17 Sleep_Walker: cd /srv/git/tools Oct 14 23:44:20 raster: well, fine you don't like it, but still it is weakness of palm bl Oct 14 23:44:38 * raster hugs plamboot Oct 14 23:44:38 Sleep_Walker: chmod a+x hooks/post-update Oct 14 23:44:43 best bootloader i've played with to date Oct 14 23:44:45 Sleep_Walker: GIT_DIR=. git-update-server-info Oct 14 23:44:45 palmos Oct 14 23:44:45 :) Oct 14 23:45:09 hooks/post-update just runs git-update-server-info after each push Oct 14 23:45:14 which makes http happy Oct 14 23:45:34 BobOfDoom: sorry to bother - it makes no effect since I have not rights for that... Oct 14 23:45:43 personally I prefer to avoid http, but I gues some people are stuck behind ridiculuous firewalls Oct 14 23:46:21 hmm Oct 14 23:47:25 http -> ultimae firewall-transversal-system Oct 14 23:47:27 Sleep_Walker: well.. I just did it Oct 14 23:47:31 thx Oct 14 23:47:31 https -> fuck you firewall Oct 14 23:47:32 :) Oct 14 23:47:45 firewalls -> stoopid Oct 14 23:48:01 (just stop allowing INCOMING connections you doofuses!) Oct 14 23:48:05 aye Oct 14 23:48:41