**** BEGIN LOGGING AT Sat Dec 18 03:00:01 2010 Dec 18 07:43:05 yann: sorry... I'm using Qi since a long time... dunno what is old and what recommended Dec 18 07:43:21 yann: probably gena2x is one of the u-boot profs :) Dec 18 07:56:30 GNUtoo|laptop: yeah, having always problems with gdb too Dec 18 08:00:40 yawn, moin Dec 18 09:50:01 mrmoku: hm, there are 2 qi images in the rc1 dl dir - any prefered one ? Dec 18 10:28:29 mrmoku: any news with n900-uboot & wd? Dec 18 10:46:41 DocScrutinizer: no, JaMa was trying to port the n900 patches to current u-boot Dec 18 10:46:54 DocScrutinizer: the old u-boot used by maemo/meego does not build with our toolchain Dec 18 10:47:02 yann: shr-testing that is, right? Dec 18 10:47:58 mrmoku: blargh Dec 18 10:47:59 * mrmoku can't find no rc1 dl dir Dec 18 10:48:22 too new a toolchain - that's really nasty Dec 18 10:48:49 at least it' something keeping us busy :P Dec 18 10:48:54 +s Dec 18 10:51:22 DocScrutinizer: and then... I have wd problems without u-boot too Dec 18 10:51:39 mrmoku: I had long debates on #maemo about n900 uboot's Alzheimer, caused by no proper env partition. I suggested a few tackling vectors: a)CAL partition data seems to me like it's quite identically structured to uboot env partition. b) uboot could learn to store env to a file on eMMC. c) uboot could sote env to own program text, an empty data structure there Dec 18 10:51:43 sometimes boot is fast enough... sometimes not Dec 18 10:52:24 DocScrutinizer: probably not enough interest in u-boot ? Dec 18 10:52:34 this should be fixable by uboot dealing with wd Dec 18 10:52:52 mrmoku: nah, just too exaotic Dec 18 10:53:29 pretty few low level coders on #maemo Dec 18 10:53:41 heh, ok Dec 18 10:54:29 #meego-arm otoh is somewhat hostile. They always know better what to do and what's needed and *NOT NEEDED* :-( Dec 18 10:55:07 and generally not interested in talkers like me - standard answer: send patches! Dec 18 10:56:09 "but I'm no real decent sw-devel, I'm kinda EE and general system architecture" - "then dummy up!" Dec 18 10:57:38 "may I at least remark your driver for xyz chip is braindamaged and doesn't support suspend mode, and even uses polling" - "so what? it's upstream like that" Dec 18 10:58:39 * DocScrutinizer learnt to hate meego "community" Dec 18 10:59:27 * mrmoku feels the same :/ Dec 18 11:02:10 mrmoku: could you toss me a html (git?) URL to OM/OE's lis302dl driver? Dec 18 11:02:33 DocScrutinizer: let me try to find it :) Dec 18 11:02:38 thanks Dec 18 11:03:03 also a ls -l of the related sysfs directory would be welcome Dec 18 11:05:04 hmm lis3lv02d ? Dec 18 11:05:13 ouch, yes Dec 18 11:05:14 http://gitorious.org/~jama/htc-msm-2-6-32/openmoko-kernel/trees/shr-2.6.34/drivers/hwmon Dec 18 11:05:19 in there Dec 18 11:05:34 * mrmoku checks if upgrade finished to do a ls -l Dec 18 11:05:36 the lv302d is the borked one iirc Dec 18 11:07:44 we poll the sensor only at 20Hz... enough for the * joystick Dec 18 11:08:53 yeah, that's a joystick driver, not a driver for a generic accelerometer for general purpose low power IRQ driven use in a mobile battery powered device Dec 18 11:10:07 ""Note that 40Hz input device can eat up about 10% CPU at 800MHZ"" Dec 18 11:14:11 ""* The sensor can generate interrupts for free-fall and direction detection (distinguishable with FF_WU_SRC and DD_SRC) but to keep the things simple and _fast_ we activate it only for free-fall, Dec 18 11:16:11 394 int lis3lv02d_joystick_enable(void) Dec 18 11:16:52 yann: qi-ubi is needed if you flash ubi image in your nand flash, qi (without ubi) is for jffs2 image there Dec 18 11:17:17 is3_dev.idev->poll = lis3lv02d_joystick_poll; //uerrrgh Dec 18 11:18:12 mrmoku: I don't need a sysfs ls for this driver Dec 18 11:19:35 printk(KERN_ERR DRIVER_NAME ": joystick initialization failed\n"); Dec 18 11:20:10 mv $this /dev/null Dec 18 11:22:00 hehe Dec 18 11:22:26 *this* is a rather nice driver for lis302: http://mxr.maemo.org/fremantle/source/kernel/drivers/i2c/chips/lis302dl.c - has 1 or two flaws that are easily fixed, but isn't as broken by concept as the lis3lv02d Dec 18 11:26:56 please don't tell me SHR needs to use lis3lv02d just because the nokia driver isn't "upstream" Dec 18 11:27:59 morning Dec 18 11:28:53 one flaw of that driver is http://mxr.maemo.org/fremantle/source/kernel/drivers/i2c/chips/lis302dl.c#165 /* Control High Pass filter selection. not used */ Dec 18 11:28:58 hi mickeyl Dec 18 11:30:55 DocScrutinizer: you mean use it for OM devices too? Dec 18 11:31:05 yes Dec 18 11:31:20 it's clearly the better concept Dec 18 11:31:32 ok, we should look at porting it into the om kernel then Dec 18 11:32:48 http://mxr.maemo.org/fremantle/source/kernel/drivers/i2c/chips/lis302dl.c#187 /* Enable interrupt wakeup on x and y axis */ is another flaw of that driver, related to the first one Dec 18 11:35:11 you better enable IRQ for X,Y,Z all three and use HP-FILTER to only wake up on *movement*. The way they do it, the only non-alarming position of device is when lying flat Dec 18 11:36:29 also a few sysnodes missing, to expose these chip config options to userland (threshold, HP-Filter parameters & enable, IRQ enable) Dec 18 11:37:04 all very easy and straighforward, even I could code it during an afternoon :-) Dec 18 11:45:26 oops THS even has a synode Dec 18 11:45:33 sysnode even Dec 18 11:47:08 mrmoku: for OM devices this driver is essential for proper wakeup from suspend Dec 18 11:47:36 DocScrutinizer: the one we have now? or the one from maemo? Dec 18 11:47:40 (on device movement that is) Dec 18 11:47:47 the maemo one Dec 18 11:47:51 ok Dec 18 11:48:44 I actually suggested to implement the driver that way on OM, like some 2 years ago Dec 18 11:49:01 when I wasn't aware Nokia had that nice piece of sw Dec 18 12:00:56 does the nokia one also has the ondemand low power mode when noone is listening? Dec 18 12:03:17 errr, lis302 chip is by itself rather lowpower. And that driver isn't supposed to be polled, thus nobody "listening" usually. Rather it gets serviced by kernel (I2C transmit) only when somebody is reading from the coords(?) sysnode, or when the chip itself detects a IRQ-condition Dec 18 12:03:59 this concept can't provide a "noone listening" semantic, as the IRQs are signalled via kevent Dec 18 12:05:15 you could power down the chip if all IRQ WU sources are disabled, though the driver in its current form doesn't support this disabling (as mentioned above it needs a few easy to add sysnodes for that) Dec 18 12:05:38 interesting Dec 18 12:05:43 i wander what the OM driver then does Dec 18 12:05:47 err, wonder, actually Dec 18 12:06:07 aiui it is polling the chip Dec 18 12:06:12 @ 20Hz Dec 18 12:06:22 right Dec 18 12:06:34 braindead concept Dec 18 12:06:51 didn't andy do it because of the hardware bug with the missing CS? Dec 18 12:06:55 and of course won't work in suspend Dec 18 12:07:17 err, I'm not aware about such hw bug. Dec 18 12:08:36 the (c) in the borked lis3lv02d driver used by OM doesn't suggest Andy did anything to it Dec 18 12:10:48 I know in A5 we had bot IRQ pins of both chips connected, while later on that got changed to single dedicated IRQ / chip Dec 18 12:10:56 both* Dec 18 12:11:12 yes Dec 18 12:11:15 that was what i was remembering Dec 18 12:11:37 so maybe the problem was the CPU can't distinguish which of the chips fired an IRQ, in A5 Dec 18 12:11:44 *nod* Dec 18 12:12:01 that's however not exactly true Dec 18 12:13:16 you can configure which chip uses line 1 and which uses line 2 (given these IRQ lines are wired-OR and not a totem-pole output, which I think is the case) Dec 18 12:13:49 you also can read WU-SOURCE of both chips to see which one was the one that fired Dec 18 12:16:13 heh small meego upgrade :) Overall download size: 219.1 MiB. After the operation, additional 378.8 MiB will be used. Dec 18 12:16:32 mrmoku: that is in http://build.shr-project.org/tests/shr-testing/images/om-gta02/ Dec 18 12:16:39 mickeyl, I looked at nexus S defconfig and I saw that: CONFIG_PHONET=y Dec 18 12:16:42 hi btw Dec 18 12:16:50 I do not have such phone Dec 18 12:16:57 but maybe they use a nokia_isi modem too Dec 18 12:17:17 GNUtoo|laptop: interesting. that would mean our work on that would be applicable to more than just the n900 Dec 18 12:17:18 nice Dec 18 12:17:25 s/would/could/ Dec 18 12:17:47 maybe Dec 18 12:17:53 http://android.git.kernel.org/?p=kernel/samsung.git;a=blob;f=arch/arm/configs/herring_defconfig;h=28cce1ecf97134ba30928b71a4699c83c8f76b29;hb=refs/heads/android-samsung-2.6.35 Dec 18 12:18:55 mickeyl, how do you debug fsogsmd? Dec 18 12:19:06 I mean I tried with gdb+gdbserver and failed Dec 18 12:19:21 try directly on the device Dec 18 12:19:28 failed there too Dec 18 12:19:33 that's bad, why? Dec 18 12:19:40 gdbserver: Dec 18 12:19:48 warning: no loadable sections found in added symbol-file /home/gnutoo/embedded/sshfs/usr/lib/.debug/libz.so.1.2.3 Dec 18 12:19:50 on target: Dec 18 12:20:01 hmm Dec 18 12:20:03 what's gdbserver again? Dec 18 12:20:17 remode gdb Dec 18 12:20:22 remote Dec 18 12:20:23 ah, remove debugging never worked for me Dec 18 12:20:27 why don't you debug locally? Dec 18 12:20:58 warning: Unable to find dynamic linker breakpoint function. Dec 18 12:21:22 it worked for me in the past(wesnoth) Dec 18 12:21:22 my question is about the difference between qi-s3c2442.udfu and qi-s3c2442-1.0.2-r5+gitr0+....udfu Dec 18 12:21:43 yann, look if qi-s3c2442.udfu is a symlink Dec 18 12:22:11 /usr/lib/libfsotransport.so.0: No such file or directory. Dec 18 12:22:12 etc... Dec 18 12:22:14 can't see that from http Dec 18 12:22:19 I simlynk them Dec 18 12:22:21 ? Dec 18 12:22:28 the problem often is on html interface you can't tell links from normal files Dec 18 12:22:28 yann, you can with md5sum Dec 18 12:22:33 right Dec 18 12:22:50 if it's a symlink it should be a symlink to the lastest version Dec 18 12:24:14 /usr/lib/libfsotransport.so.0 -> libfsotransport.so.0.0.0 Dec 18 12:24:19 strange the file is there Dec 18 12:25:01 when I do: Dec 18 12:25:03 set sysroot / Dec 18 12:25:06 I've that now: Dec 18 12:25:07 warning: no loadable sections found in added symbol-file /usr/lib/.debug/libnl-route.so.2.0.0 Dec 18 12:28:06 and of course the output is unusable Dec 18 12:28:26 #0 0x4097d2e8 in ?? () Dec 18 12:28:34 #1 0x4097a4c0 in ?? () Dec 18 12:28:43 Backtrace stopped: previous frame identical to this frame (corrupt stack?) Dec 18 12:28:58 hmm Dec 18 12:29:02 i have no idea Dec 18 12:29:06 i always run gdb directly Dec 18 12:29:11 I can't even debug in assembly Dec 18 12:29:18 mickeyl, that's what I'm talking about Dec 18 12:29:26 make sure you build the stuff with --enable-debug btw. Dec 18 12:29:36 I was talking about running gdb directly Dec 18 12:29:36 otherwise you're missing line number info everywhere Dec 18 12:29:52 ok Dec 18 12:29:57 I built from stock oe Dec 18 12:30:18 hmm Dec 18 12:30:29 I'll build manually Dec 18 12:30:32 yeah Dec 18 12:30:39 --enable-debug will give you -ggdb and -O0 Dec 18 12:30:42 but oe *should* be debuggable Dec 18 12:30:48 so line numbers in vala code and C code Dec 18 12:30:50 ah no -ggdb3? Dec 18 12:30:59 3 probably as well Dec 18 12:31:01 ok Dec 18 12:31:36 i will try to get somethign going in pass through mode Dec 18 12:31:48 if we seriously want to work on that, we need it Dec 18 12:32:42 should I add --enable-debug to the oe recipes? Dec 18 12:32:42 thankfully wifi works Dec 18 12:32:58 not by default, please Dec 18 12:33:05 this will enable even more debugging output and code Dec 18 12:33:10 which will slow down our stuff Dec 18 12:34:32 mickeyl, I mean -ggdb3 should be the default Dec 18 12:34:43 ok, but not induced by the recipe itself Dec 18 12:34:47 * GNUtoo|laptop looks in bitbake.conf Dec 18 12:34:49 but rather by OPTIMIZATION_FLAGS Dec 18 12:34:52 yes indeed Dec 18 12:36:03 also will give you bloated imgs Dec 18 12:36:11 ouch it's not there Dec 18 12:36:21 DocScrutinizer, no, because we strip them after Dec 18 12:36:27 aaah Dec 18 12:36:40 yay for seperate packing of debugging symbols ;) Dec 18 12:36:47 YEP Dec 18 12:38:12 mrmoku: you'll take care about new 302 driver? please ping me about the needed improvements to the Nokia one, whenever you get at it Dec 18 12:38:23 JaMa, hi where are -ggdb3 in the shr configs, were they removed? Dec 18 12:40:37 mrmoku: ping Dec 18 12:43:13 still that: Dec 18 12:43:14 /lib/libc.so.6: No such file or directory. Dec 18 12:43:20 I do set sysroot / Dec 18 12:43:24 to see if it's better Dec 18 12:44:27 [ 4372.122711] Out of memory: kill process 19165 (gdb) score 3409 or a child Dec 18 12:44:37 that's why I use gdbserver + gdb Dec 18 12:44:50 I'll try again with gdbserver + gdb Dec 18 12:46:27 hmm Dec 18 12:46:38 cdc_phonet is loaded on the PC Dec 18 12:46:53 now how can make the interface come up? Dec 18 12:48:52 use that: Dec 18 12:49:04 https://elektranox.org/n900/libisi/power.html Dec 18 12:49:10 or use the fso plugin for that Dec 18 12:49:15 the lowlevel nokia900 Dec 18 12:49:23 hmm Dec 18 12:49:41 then the interface will be visible on the PC? Dec 18 12:50:10 I'll install a ggdb3 libisi too Dec 18 12:50:33 I don't know how to use the phone interface on a pc Dec 18 12:50:37 ah, ok :D Dec 18 12:50:38 but I tried Dec 18 12:50:42 so I know some stuff Dec 18 12:50:47 only that I didn't succedded Dec 18 12:50:51 basically you need to: Dec 18 12:50:54 rmmod g_ether Dec 18 12:50:58 modprobe g_nokia Dec 18 12:51:02 that creates 2 interfaces Dec 18 12:51:07 one on the phone side Dec 18 12:51:12 and one on the computer side Dec 18 12:51:26 g_nokia on the PC? Dec 18 12:51:27 + some other interfaces such as acm or usb0 but that's not important here Dec 18 12:51:33 no, on the phone Dec 18 12:51:37 ya, did that Dec 18 12:51:52 ok Dec 18 12:51:58 i guess the missing link is Dec 18 12:51:59 then I know how to power on phonet0 Dec 18 12:52:00 I activated the phonet0 with the command: Dec 18 12:52:00 /mnt/initfs/sbin/phonet -a 0x6C -i phonet0 Dec 18 12:52:08 this is what nicola wrote Dec 18 12:52:14 what's that? Dec 18 12:52:17 so we need the interface to come up Dec 18 12:52:38 ifconfig interface up? Dec 18 12:52:47 it's not there Dec 18 12:52:50 hmmm Dec 18 12:52:52 no phonet0 yet Dec 18 12:52:55 that's why i'm asking Dec 18 12:53:01 let's recap a bit Dec 18 12:53:08 let me try to turn on the modem Dec 18 12:53:11 do you have the binary offhand? Dec 18 12:53:29 we should build it for debugging purposes... Dec 18 12:53:31 arm-something-gcc should do the trick Dec 18 12:53:34 k Dec 18 12:54:29 mickeyl: I gather you know a bit about phonet and libisi. Any idea if some network status info along %EM is available? Dec 18 12:54:32 so back to the recap stuff Dec 18 12:54:41 DocScrutinizer: i don't know anything yet Dec 18 12:54:42 modem interface on the phone side: phonet0 Dec 18 12:54:50 GNUtoo|laptop: ./sane-toolchain-eglibc.inc:FULL_OPTIMIZATION_pn-eglibc_arm Dec 18 12:54:52 DocScrutinizer: no time to investigate anything yet. just starting Dec 18 12:54:55 g_nokia interface on the phone side: upnlink0 Dec 18 12:55:21 g_nokia interface on the laptop/desktop side: usbpn0 Dec 18 12:55:25 mickeyl: k, if you run into anything related, please let me know Dec 18 12:56:27 DocScrutinizer: will do Dec 18 12:56:56 GNUtoo|laptop: for OOM use rm -f /lib/.debug/libc.so.6 Dec 18 12:57:00 JaMa, ok so it should be there....but why is debuggong broken Dec 18 12:57:09 *debugging Dec 18 12:57:16 because it's damn big Dec 18 12:57:34 I'll try if alignment traps are gone in current eglibc rev Dec 18 12:57:35 I mean with gdb+gdbserver Dec 18 12:58:17 GNUtoo|laptop: i have the upnlink0, but i don't have any interface on the PC side Dec 18 12:58:58 err, usppn0 i meant Dec 18 12:59:29 do you have phonet on pc side? Dec 18 12:59:33 no Dec 18 12:59:41 i mean the kenel module Dec 18 12:59:57 mickey@saphir:~$ lsmod|grep phonet Dec 18 12:59:57 cdc_phonet 5146 0 Dec 18 12:59:57 phonet 23295 1 cdc_phonet Dec 18 13:00:05 ok Dec 18 13:00:13 so you have it Dec 18 13:00:18 hmmm Dec 18 13:00:36 ok, wait... i see Dec 18 13:00:44 ok Dec 18 13:00:49 it's also called usbpn0 on the PC Dec 18 13:00:52 i was looking for phonet0 Dec 18 13:01:05 so i have the interface on both sides Dec 18 13:01:08 * mickeyl ifconfig up Dec 18 13:02:37 ok Dec 18 13:02:51 but then after that...I'm lost Dec 18 13:03:33 ok Dec 18 13:03:37 ofonod does not see any modem Dec 18 13:06:35 do you have the right ofono config? Dec 18 13:06:55 there is something like fsofsmd.conf for ofono Dec 18 13:07:04 hmm Dec 18 13:07:14 they have a generic isi driver Dec 18 13:07:17 and a n900 drvier Dec 18 13:10:48 ~lart ofonod Dec 18 13:10:49 * apt gets a hotmal account and SPAMs ofonod Dec 18 13:11:05 oh well Dec 18 13:11:07 let me use libisi Dec 18 13:11:34 hi there Dec 18 13:11:51 moin Kensan Dec 18 13:12:52 how are things? Dec 18 13:13:14 * Kensan is looking for a "smallish" Python starter project Dec 18 13:13:17 ;) Dec 18 13:13:48 yeah isn't phonet funny? has a virtual modem via network, on PC Dec 18 13:13:58 GNUtoo: is there a keepalive test or something? Dec 18 13:14:04 or at least detection Dec 18 13:14:19 authenticate and network are already doing too much Dec 18 13:14:23 a ping would do Dec 18 13:14:31 just want to check whether the passthrough works Dec 18 13:15:00 where is sre when you need him... Dec 18 13:17:18 reading... Dec 18 13:18:04 there are some vala ecamples in tests Dec 18 13:18:33 * GNUtoo forgott to eat....and should do it soon Dec 18 13:20:25 hmmm Dec 18 13:21:27 else we could use maemo...but their stuff would interfeer with the initialization...not counting the proprietaryness and the "sms service" Dec 18 13:22:03 hmm Dec 18 13:22:08 upnlink0 vs. phonet0 Dec 18 13:22:17 what is what for? Dec 18 13:23:03 I described the interfaces not so long ago...what do you mean exactly? Dec 18 13:23:55 ok, the PC sends out packages Dec 18 13:23:56 look up... Dec 18 13:24:05 the device gets the packages Dec 18 13:24:14 but doesn't answer Dec 18 13:24:22 ok Dec 18 13:24:22 so the communication does work Dec 18 13:24:28 but the modem is not answering yet Dec 18 13:24:34 ok Dec 18 13:24:47 my concern were Dec 18 13:25:13 how to bind phonet0 and the g_nokia phone side interface Dec 18 13:25:17 ? Dec 18 13:25:26 or route Dec 18 13:25:32 or something Dec 18 13:25:36 hmm Dec 18 13:26:01 we should find out what /mnt/initfs/sbin/phonet -a 0x6C -i phonet0 does Dec 18 13:26:28 ok Dec 18 13:26:34 on maemo you get asked which mode you want to connect via usb Dec 18 13:26:47 so i wonder whether we have to explicitly switch to pcsuite mode Dec 18 13:26:52 and if so, how Dec 18 13:27:23 mrmoku knows Dec 18 13:44:56 mickeyl, any luck? Dec 18 13:45:15 for me debugging still doesn't work Dec 18 13:45:51 #0 0x4095c2e8 in g_isi_client_destroy () from embedded/sshfs/usr/lib/libisi.so.0 Dec 18 13:45:57 it's a from, not an in Dec 18 13:51:40 mickeyl: pong Dec 18 13:52:08 what steps did you do to enable pass through and where did it hang for you? Dec 18 13:52:30 just made sure g_nokia is loaded and powered the modem on Dec 18 13:52:47 and then the tests in libisi timeout Dec 18 13:52:58 instead when booted to maemo they don't Dec 18 13:53:02 ok Dec 18 13:53:20 so the missing link is probably to enable pass through on the device side Dec 18 13:53:30 or rather connect the pass through interface to the actual modem Dec 18 13:53:33 maybe because of the different kernel? Dec 18 13:53:42 meego uses 2.6.35 and ofono supports that Dec 18 13:54:09 for instance the initialization is different between both kernel but according to the source code it should autodetect Dec 18 13:54:12 why are we not using the meego kernel again? Dec 18 13:54:19 maemo kernel Dec 18 13:54:23 maemo power kernel to be exact Dec 18 13:54:26 is what we use Dec 18 13:54:27 dvfs is not there Dec 18 13:54:43 yes Dec 18 13:54:47 the state is unknown(we don't know what is lacking) Dec 18 13:54:51 why are we _not_ using th e meego kernel was the question Dec 18 13:54:56 ah, we never tried Dec 18 13:55:54 when we asked to the meego devs they didn't know what was lacking in their kernel Dec 18 13:56:10 heh Dec 18 13:56:12 no DVFS means huge power consumation Dec 18 13:56:14 ok, what a chaos :D Dec 18 13:56:39 the question is rather why we don't use it yet => because it's not ready yet Dec 18 13:57:18 mickeyl: if I do LD_LIBRARY_PATH=../src ./network Dec 18 13:57:22 because of course we want to use it, but when it'll be ready Dec 18 13:57:25 while g_ether is loaded Dec 18 13:57:26 I get Dec 18 13:57:31 ** (process:5989): DEBUG: g_pn_netlink_set_address: Operation not permitted Dec 18 13:57:34 ** (process:5989): DEBUG: g_pn_netlink_add_route: Operation not permitted Dec 18 13:57:42 i'm afraid you need to be root Dec 18 13:58:02 probably Dec 18 13:58:06 i see packets coming in Dec 18 13:58:11 though when doing that with g_nokia loaded Dec 18 13:58:12 on the device side Dec 18 13:58:14 I get in addition Dec 18 13:58:15 ** (process:5989): DEBUG: Phonet: usbpn0 is down, idx=0xa Dec 18 13:58:15 Modem Reachable Status: down Dec 18 13:58:28 did you change phonet0 to usbpn0? Dec 18 13:58:43 on the host side it's not called phonet0 as far as i can see Dec 18 13:59:05 hi morphis Dec 18 13:59:17 GNUtoo|laptop: heyho Dec 18 13:59:21 after doing ifconfig phonet0 up Dec 18 13:59:28 and running it as root on my lapotp Dec 18 13:59:40 ** (process:6004): DEBUG: Phonet: usbpn0 is up, idx=0xa Dec 18 13:59:40 ** (process:6004): WARNING **: Retry COMM_ISI_VERSION_GET_REQ Dec 18 13:59:40 ** (process:6004): WARNING **: Timeout COMM_ISI_VERSION_GET_REQ Dec 18 13:59:40 ** (process:6004): WARNING **: Unable to bootstrap mtc driver Dec 18 13:59:45 yep Dec 18 13:59:51 that's exactly where it's hanging for me Dec 18 13:59:59 it sends the REQ out Dec 18 14:00:03 the REQ arrives on the device Dec 18 14:00:07 but noone answers on the device Dec 18 14:00:10 yup Dec 18 14:00:24 brb Dec 18 14:00:30 ok Dec 18 14:00:39 we might have to forward that to phonet0 Dec 18 14:03:00 could someone still running maemo check what Dec 18 14:03:08 effect choosing "PC Suite" mode has? Dec 18 14:03:19 would be interesting to see whether it calls a binary Dec 18 14:03:31 or anything else Dec 18 14:09:03 hmm, nnokia pc suite may be completely uninteresting Dec 18 14:09:11 this probably just talks AT Dec 18 14:10:00 ah, you guys trying to work with the phonet-stuff from your local pc? Dec 18 14:10:10 yes Dec 18 14:10:29 it's no real serial interface? Dec 18 14:10:37 no fake serial Dec 18 14:10:51 basically a proprietary userland program called pnatd emulate AT Dec 18 14:11:40 mickeyl: the most obvious effect is g_nokia being loaded Dec 18 14:11:45 dunno what else it doesn Dec 18 14:11:47 -n Dec 18 14:11:57 and how to check what it does Dec 18 14:12:04 * mrmoku boots to maemo Dec 18 14:13:14 yeah Dec 18 14:13:20 lets compare the output of ifconfig -a Dec 18 14:13:24 in a working scenario and a non-working Dec 18 14:13:43 and route Dec 18 14:15:11 mickeyl: I have tx/rx bytes on both... phonet0 and upnlink0 Dec 18 14:15:19 after issuing the network command Dec 18 14:15:50 right Dec 18 14:15:54 that shows the missing link Dec 18 14:16:06 "shows" Dec 18 14:16:20 how's your ifconfig -a and route -rn looking? Dec 18 14:16:46 invalid option -r :P Dec 18 14:16:54