**** BEGIN LOGGING AT Sun Jul 08 02:59:58 2012 Jul 08 08:58:57 nully: you have to request an account Jul 08 08:59:26 Ah, I looked for that, but i did not see it written anywhere, I probably misseed it I guess. Jul 08 08:59:32 mrmoko should be able to give you one Jul 08 09:00:14 it is on the front page, though too small Jul 08 09:00:58 wow :/ i tried doing searches for `account' and `request' and `login'... All I had to do was read the front page, all the way. Jul 08 09:03:20 i must say it's not the best solution Jul 08 09:03:47 but we had so much spam.. Jul 08 09:03:56 It really makes the project feel very closed :/ But yes, SPAM does stink. Jul 08 09:06:07 yes, that's very unfortunate Jul 08 09:07:11 but trac seems not to have a feature like email verification for new accounts Jul 08 09:07:58 that would minimize spam Jul 08 09:39:33 freesmartphone.org: 03morphis 07utilities * r4f32a7f637c9 10/android/image-utils/fastboot/Makefile.am: android: image: install headers into own sub directory Jul 08 09:39:34 freesmartphone.org: 03morphis 07utilities * r1f438c8a1d9a 10/android/image-utils/ (ChangeLog NEWS configure.ac): android: image-utils: prepare for next release 0.1.1 Jul 08 09:39:34 freesmartphone.org: 03morphis 07utilities * r25a5a3b67b6a 10/android/image-utils/ (fastboot/Makefile.am mkbootimg/Makefile.am): android: image-utils: use static linking Jul 08 09:39:35 freesmartphone.org: 03morphis 07utilities * r86445f408535 10/android/image-utils/ (libmincrypt/Makefile.am libzipfile/Makefile.am): android: image-utils: don't install libzipfile and libmincrypt libraries Jul 08 09:39:35 freesmartphone.org: 03morphis 07utilities * r1097722a3e5c 10/android/image-utils/ (fastboot/Makefile.am mkbootimg/Makefile.am): android: image-utils: don't install any header files Jul 08 09:39:36 freesmartphone.org: 03morphis 07utilities * r4566e120fc80 10/android/image-utils/ (ChangeLog NEWS configure.ac): android: image-utils: prepare for next release 0.1.2 Jul 08 10:12:39 Moin. Is there any page which describes building SHR for samsung devices? Jul 08 10:18:27 playya: not really, there are some bits in the SHR and FSO wiki Jul 08 10:18:37 ok Jul 08 10:18:51 http://wiki.freesmartphone.org/index.php/Hardware/Nexus_S Jul 08 10:19:04 it think I only need the point where to put the meta-samsung layer Jul 08 10:19:14 you're using shr-chroot? Jul 08 10:20:07 no. is it recommended now? Jul 08 10:20:10 yes Jul 08 10:20:15 it's possible without Jul 08 10:20:27 but you will encounter problems Jul 08 10:20:43 (depends on which host system you're using) Jul 08 10:21:10 hehe. yes Jul 08 10:22:48 otherwise the relevant OE configs are here: http://git.shr-project.org/git/?p=shr-makefile.git;a=tree;f=conf/shr-core;h=5dd578dfab351e337ead4eeffe0cde9ecc244d7b;hb=HEAD Jul 08 10:23:09 bblayers.conf is the one where meta-samsung is configured for bitbake Jul 08 10:26:55 yes. but i need the priority to avoid overriding things or missing overrides Jul 08 10:28:00 lunch!! Jul 08 10:35:03 playya: where do you need to override something in detail? Jul 08 11:05:35 I don't know, yet morphis Jul 08 11:06:12 in general that should work without any problems with the default configuraiton Jul 08 11:09:57 ok, let me know if you need some help Jul 08 11:18:04 i have to leave. see you later Jul 08 15:41:37 freesmartphone.org: 03GNUtoo 07cornucopia * rd4ae83909b87 10/fsodeviced/conf/GTA04/alsa-3.4/stereoout: Jul 08 15:41:37 freesmartphone.org: fsodeviced: complete the GTA04 stereoout scenario. Jul 08 15:41:37 freesmartphone.org: Signed-off-by: Denis 'GNUtoo' Carikli Jul 08 17:06:32 Alex[sp3dev], hello, I still don't know who to boot lk Jul 08 17:08:50 ayaka: ^_^ Jul 08 17:08:56 paulk-desktop: hi Jul 08 17:09:32 Alex[sp3dev], hi Jul 08 17:09:33 Alex[sp3dev], 0x0 or base_address can't work at all, it will reboot Jul 08 17:10:06 ayaka: what are you trying to do and what are the addresses again? Jul 08 17:11:04 paulk-desktop: so, I tried using a socket in init.rc and it returned "already in use". I'll try to see how rild uses its own socket. Do you have any ideas how the blobs magically have sockets working? Jul 08 17:11:27 BASE_ADDR := 0x40200000, 0x40200000 is from htc android img Jul 08 17:11:29 mhh I didn't really look at it since then Jul 08 17:12:21 paulk-desktop: good news is that ril is stable. I've paid for the service on the sim card in nexus today and as soon as the money arrived, 3G connected and gtalk logged on Jul 08 17:12:21 Alex[sp3dev], I have try 0x0 ,0x40200000, 0x30000000, 0x10000000 in MEMBASE Jul 08 17:12:31 great Jul 08 17:12:48 I have reworked a lot of things regarding data during the last few days Jul 08 17:12:53 now it's rock solid Jul 08 17:13:07 paulk-desktop: I think we need to add wakelocks and polling for the link ONLINE status to ipc read/write. Looks like ril crashes in suspend sometimes because of something, probably that Jul 08 17:13:40 ok, good idea Jul 08 17:14:16 but I only see HTC logo, after a while it reboot Jul 08 17:14:17 I need to clone the ril sources. I'm at home pc now, and all previous work was.. at work :) Jul 08 17:14:46 ok Jul 08 17:15:01 it's pushed to morphis git for libsamsung-ipc and replicant for samsung-ril Jul 08 17:15:21 finally switched to using a 22'' external monitor at home. no more staring at 1600x900@13.3 Jul 08 17:17:07 good :) Jul 08 17:18:17 http://higgs.rghost.ru/private/39097068/64bec57c3ed78e285437d3afcfec7dda/image.png my laptop has a GPLv3 sticker. I got it when Stallman gave a lecture in Moscow in december Jul 08 17:18:45 ok, back to coding Jul 08 17:18:52 looks great: Jul 08 17:19:06 is that an apple keyboard? Jul 08 17:19:30 nope, some cheap usb keyboard. It has a rubber detachable cover to protect from dust Jul 08 17:19:42 ah ok Jul 08 17:21:00 if you're interested, it's Oklick 555S. Probably some russian-branded chinese OEM Jul 08 17:21:56 paulk-desktop: what should I check out? your git or replicant? Jul 08 17:22:11 should be the same Jul 08 17:22:32 ah maybe replicant is one commit ahead Jul 08 17:22:55 though it's a pain to make it work with ICS Jul 08 17:23:09 why Jul 08 17:23:38 this is still for GB, and they changed the structures I introduced the use of with these new commits Jul 08 17:24:09 if you don't want to adapt stuff, wait a bit, I'll probably adapt it tomorrow Jul 08 17:24:48 won't makefile hackery help us like we did the previous time? Jul 08 17:25:20 nah, here it's a bit deeper Jul 08 17:25:38 ok Jul 08 17:26:31 I'll wait for your fixes. We can probably make a compatibility header and some ifdef to avoid polluting makefiles Jul 08 17:34:52 paulk-desktop: we have a problem now Jul 08 17:34:59 external/libsamsung-ipc/samsung-ipc/device/xmm6260/xmm6260_modemctl.h:44: error: field 'radio_stat' has incomplete type Jul 08 17:35:03 some header missing? Jul 08 17:35:21 possible Jul 08 17:35:34 probably you removed when refucktoring Jul 08 17:35:42 must be it Jul 08 17:40:30 ok, I forgot to hard reset my local changes. Let's hope the rest is fine Jul 08 18:03:11 GNUtoo-desktop: ping Jul 08 18:05:07 paulk-desktop: ping Jul 08 18:05:13 pong Jul 08 18:05:55 paulk-desktop: I have pushed some sort of, well, fixes. I've moved all those compatibility ril defines to a header. I've built your ril master on ics Jul 08 18:05:57 https://gitorious.org/~ast/replicant/asts-hardware_ril_samsung-ril/commits/08jul_fixes Jul 08 18:06:43 and it works? Jul 08 18:07:02 well, will try now Jul 08 18:07:08 I see no reason why it won't Jul 08 18:07:10 I doubt data will work Jul 08 18:07:36 at least some of it won't Jul 08 18:07:49 (data_call_list probably won't) Jul 08 18:08:06 I still have to think about how I will deal with maintaining both GB/ICS Jul 08 18:10:36 so by saying data is rock solid you meant that it doesn't work at all? Jul 08 18:11:21 indeed it does not work at all Jul 08 18:12:03 no, I mean that on GB, it is now full-featured, takes care of all the error cases and various other stuff that makes it very reliable Jul 08 18:22:09 paulk-desktop: found this in ril.cpp http://pastebin.com/4YnjhV8W . looks like we're on the right track Jul 08 18:22:29 oh nice! Jul 08 18:39:14 GNUtoo-desktop: I tried on my own with my Nexus S and fsogsmd and got a little bit further: http://pastie.org/4221546 Jul 08 18:39:30 look at the line: "g_dbus_connection_real_closed: Remote peer vanished with error: Underlying GIOStream returned 0 bytes on an async read (g-io-error-quark, 0). Exiting." Jul 08 18:39:50 somehow our internal dbus connection gets vanished Jul 08 18:40:05 ok, this way of getting socket works for rild side. still I think patching init.rc (not the board-specific one) is not extremely good and I wonder why the blob can obtain the socket the old way. The problem is that the client (mediaserver) gets permission denied, though the socket is radio:system Jul 08 18:40:34 GNUtoo-desktop: and I hopefully finally fixed the usbnet problem Jul 08 18:41:23 GNUtoo-desktop: if you want to test: http://pastie.org/4221559 Jul 08 18:48:55 ok one second Jul 08 18:52:59 ok back Jul 08 18:53:07 I was testing my headphones on gta04 Jul 08 18:54:09 ah ok Jul 08 18:54:17 the patch for usbnet should help Jul 08 18:54:23 it fixes the problem but is a hack Jul 08 18:54:27 but who cares Jul 08 18:54:31 indeed Jul 08 18:55:19 the dbus problem is a little bit more hard to resolve Jul 08 18:55:36 ok Jul 08 18:55:43 did you push the usbnet patch? Jul 08 18:56:26 paulk-desktop, hi, now I've resoldered the headphones, it was way harder that it seems but at the end I've headphones + mic Jul 08 18:56:34 GNUtoo-desktop: will do in a minute Jul 08 18:56:47 hm the dbus problem of fsogsmd seems to be related to this one here: Jul 08 18:56:47 fsotdld: symbol lookup error: /usr/lib/cornucopia/modules/fsotdl/provider_location_cellidwifi.so: undefined symbol: fso_tdl_abstract_location_provider_get_type Jul 08 18:57:01 ah ok Jul 08 18:57:01 GNUtoo-desktop, ok Jul 08 18:57:03 thanks a lot Jul 08 18:57:12 (the thanks was for morphis) Jul 08 18:57:17 I saw :) Jul 08 18:58:11 SHR: 03morphis 07meta-smartphone * r20df5dff22b1 10/meta-samsung/recipes-core/systemd/ (3 files in 2 dirs): meta-samsung: systemd-machine-units: force rndis0 interface to power up on boot Jul 08 19:01:41 GNUtoo-desktop: ^^ Jul 08 19:02:26 ok Jul 08 19:03:24 morphis, I was more thinking of that: Jul 08 19:03:47 http://shr-project.org/trac/ticket/2019 Jul 08 19:04:01 it contains that: Jul 08 19:04:02 http://shr-project.org/trac/attachment/ticket/2019/usb_cable.rules Jul 08 19:04:25 basically it does ifup usb0 when you plug a cable in Jul 08 19:04:33 and ifdown usb0 when you remove the cable Jul 08 19:05:02 GNUtoo-desktop: yeah, I like that Jul 08 19:05:19 if you want and have time, please get this working :) Jul 08 19:05:39 I tough that JaMa|Off would have picked it Jul 08 19:06:19 really? Jul 08 19:06:23 JaMa|Off: ping Jul 08 19:07:01 I was concentrating on other stuff Jul 08 19:07:09 for instance make the gta04 audio work flawlessly Jul 08 19:07:19 it's not totally ok but much better than before Jul 08 19:07:23 it needs testing Jul 08 19:07:33 I had crashes of fsoaudiod long time ago Jul 08 19:07:41 and it was not resolved Jul 08 19:07:48 so I suspect they are still there Jul 08 19:08:39 apart that it's much better than before Jul 08 19:09:09 * the two (stereoout, gsm )scenarios are now complete. Jul 08 19:09:31 * fsoaudiod should work after ringing( I didn't test myself but slyon tested) Jul 08 19:10:00 my next blocking issue is the headphones detection Jul 08 19:10:13 great Jul 08 19:10:26 paulk-desktop, had issues with headphones detection Jul 08 19:10:31 like noise in the mic Jul 08 19:10:45 so if you detect the headphones in a call..... Jul 08 19:10:57 or simply while recoriding Jul 08 19:12:47 freesmartphone.org: 03morphis 07cornucopia * rba004d18ee86 10/fsotdld/src/plugins/provider_location_cellidwifi/Makefile.am: fsotdld: provider_location_cellidwifi: link against internal library; fixes #682 Jul 08 19:20:22 paulk-desktop: ok, I'm closer to figuring out how it works. the proprietary RIL creates a socket in NAMESPACE_ABSTRACT as opposed to NAMESPACE_RESERVED. patching init.rc and grabbing socket from zygote works as well. iirc, the proprietary ril then sets permissions on its socket. I'll try to investigate that. I think the easiest thing is to patch init.rc and set permissions 666 on our socket. that'd be a compromise, but we Jul 08 19:22:01 okay Jul 08 19:23:03 why the hell is this so complicated? why couldn't they've used ACL for socket permissions and set multiple groups to a socket? or just use dbus Jul 08 19:25:50 GNUtoo-desktop: was there a ticket for the the modem_samsung bug? Jul 08 19:33:45 yes old ticket Jul 08 19:35:57 freesmartphone.org: 03morphis 07cornucopia * r13a887bae111 10/fsogsmd/src/lib/mediatorhelpers.vala: (log message trimmed) Jul 08 19:35:57 freesmartphone.org: fsogsmd: lib: disable lookup of network provider names via odatad Jul 08 19:35:57 freesmartphone.org: Aquiring a dbus service at runtime while in context of a mediator causes the folllowing Jul 08 19:35:57 freesmartphone.org: behaviour: Jul 08 19:35:57 freesmartphone.org: 2012-07-08T19:16:55.326085Z [WARN] fsogsmd : Could not find and valid provider name for MCC/MNC 26202: The connection is closed Jul 08 19:35:58 freesmartphone.org: 2012-07-08T19:16:55.326902Z [DEBUG] fsogsmd : Enqueue a new command: type = IPC_TYPE_GET, command = IPC_NET_REGIST Jul 08 19:35:59 freesmartphone.org: 2012-07-08T19:16:55.329576Z [WARN] subsystem : Can't claim busname org.freesmartphone.ogsmd Jul 08 19:36:29 GNUtoo: it should now work again Jul 08 19:36:37 GNUtoo: but we're missing APN resolving right now Jul 08 19:36:45 I will fix that in the next days Jul 08 19:37:56 ok nice Jul 08 19:38:02 i'll build Jul 08 19:39:02 please report if it's not fixed Jul 08 19:39:16 I tested it on my device and it was working Jul 08 19:40:40 ok Jul 08 19:41:25 paulk-desktop: so. it kinda works now. as far as I've understand, they're now using NAMESPACE_ABSTRACT everywhere. Looks like it's an official security hole or a way to circumvent any checks and create a socket between daemons with different GID Jul 08 19:41:54 ok Jul 08 19:42:07 I really wonder why they introduced that socket "security" thing Jul 08 19:42:11 that's a pain in the ass Jul 08 19:42:48 I'll clean the code and push a hack. my policy about android development is that if it works, just leave it as is. it will break with every release anyway Jul 08 19:55:11 paulk-desktop: https://gitorious.org/~ast/replicant/asts-hardware_ril_samsung-ril-client/commit/8400f499c44482d3c73a1b837f583fe320a0ed9a and https://gitorious.org/~ast/replicant/asts-hardware_ril_samsung-ril/commit/4935eac2f82042ad662a16fbf8058f2139ac3843 Jul 08 19:55:22 stupid but works Jul 08 20:15:53 thanks and good bye Jul 08 20:44:57 morphis, hmmm Jul 08 20:45:17 2012-07-08T20:48:59.901798Z [INFO] fsogsmd : crespo_ipc_fmt_client_send: IPC request (mseq=0x6d command=IPC_NET_REGIST (0x0805) type=GET) Jul 08 20:45:24 I get the logs full of theses now Jul 08 20:45:34 I'll try manually Jul 08 20:48:33 hmmm same issue manually Jul 08 20:48:36 is it pushed? Jul 08 20:48:51 because I've autorev and just rebuilt + adb pushed the rootfs Jul 08 20:49:59 hmmm Jul 08 20:50:01 fsogsmd - 2:0.11.99+gitr5+719f5882bea02367a20242a90956460c31765b3f-r2.1 Jul 08 20:50:06 not lastest rev Jul 08 20:53:13 2012-07-08T20:56:39.461187Z [WARN] fsogsmd : Could not find and valid provider name for MCC/MNC 22210 Jul 08 20:53:27 should I change sim card? Jul 08 21:02:19 hmm Jul 08 21:02:22 that's just a warning. try adding the necessary string, though Jul 08 21:02:27 it's still not at HEAD Jul 08 21:02:39 I'll rebuild once again Jul 08 21:03:43 /lastlog Jul 08 21:04:14 | pdphandler.c:697:2: error: too few arguments to function 'ipc_gprs_pdp_context_setup' Jul 08 21:04:21 that was the issue Jul 08 21:06:01 http://pastie.org/private/v26qlyd4liocxdbjfng Jul 08 21:06:51 morphis,^^^ I'll try again when it will compile again Jul 08 21:07:56 haha. gprs is a PITA for everyone recently Jul 08 21:08:59 I'll try to rebuild everything(fsogsmd + libsamsung-ipc) Jul 08 21:14:04 hmmm Jul 08 21:15:48 it fails the same way Jul 08 21:17:44 GNUtoo-desktop: you should not compile with $AUTOREV for libsamsung-ipc Jul 08 21:17:56 yes because of the recent changes? Jul 08 21:17:59 thats not supported atm Jul 08 21:18:14 like Alex[sp3dev] 's changes Jul 08 21:18:48 GNUtoo-desktop: my changes were only adding the new device. gprs is paulk-desktop's work Jul 08 21:18:55 ok Jul 08 21:19:20 GNUtoo-desktop, ah yes, that's because of me Jul 08 21:19:28 what should I do ? Jul 08 21:19:34 maybe better fixing it? Jul 08 21:19:39 ok Jul 08 21:19:44 I can fix the vapi Jul 08 21:19:46 GNUtoo-desktop: don't use $AUTOREV for libsamsung-ipc right now Jul 08 21:19:49 ok thanks a lot Jul 08 21:19:50 but someone has to fix fsogsmd Jul 08 21:19:57 ok Jul 08 21:20:02 paulk-desktop: yes, thats on my list Jul 08 21:20:05 maybe I could try to fix it Jul 08 21:20:09 once the vapi is fixed Jul 08 21:20:11 but for now we should stick with 0.1 in SHR Jul 08 21:20:13 the missing argument is CID that should be set to 0 when there is only one connection at the time Jul 08 21:20:19 because it would give the right verbose error Jul 08 21:20:55 but we could do that: create a stable branch that remains 0.1 and make SHR use that branch Jul 08 21:21:13 yeah, we need to tag working revisions Jul 08 21:21:18 paulk-desktop: we don't need Jul 08 21:21:39 http://downloads.freesmartphone.org/sources/libsamsung-ipc/0.1/ Jul 08 21:21:54 ah, so you can use the archive directly? Jul 08 21:22:03 that is what fsogsmd should use currently Jul 08 21:22:12 paulk-desktop, I am the only one using autorev it seems Jul 08 21:22:17 ok Jul 08 21:22:18 GNUtoo-desktop: yes .) Jul 08 21:22:34 JaMa|Off, may use autorev too in his jansa/test branch Jul 08 21:22:35 GNUtoo-desktop: I am just enabling it for some components here Jul 08 21:22:41 ok Jul 08 21:22:49 I need it for fsodeviced at least Jul 08 21:22:50 for gta04 Jul 08 21:23:10 shat libsamsung-ipc rev should I use? Jul 08 21:23:23 I just remove autorev for it? Jul 08 21:24:31 I'll retry Jul 08 21:24:55 yes, I think you should remove autorev for libsamsung-ipc at least Jul 08 21:25:05 with that: Jul 08 21:25:06 http://pastie.org/private/eenwlavmotfzm33nnrcq Jul 08 21:25:15 74664ce8d19f747e89a38d1c5fe439c6afb155f6 should work fine Jul 08 21:25:15 we should remove it from the inc file maybe Jul 08 21:25:32 in replicant libsamsung-ipc works fine right? Jul 08 21:25:42 yes Jul 08 21:25:44 it does Jul 08 21:25:45 ok Jul 08 21:27:37 maybe morphis should push what's in http://pastie.org/private/eenwlavmotfzm33nnrcq because I am unable to justify well such commit in the commit message Jul 08 21:28:11 GNUtoo-desktop: no, we should not adjust fso-autorev.inc for this Jul 08 21:28:19 ah? Jul 08 21:28:29 adjust and push I mean Jul 08 21:28:37 why? Jul 08 21:29:09 because it will be fixed later? Jul 08 21:31:34 yes Jul 08 21:31:57 if you want it compiling right now, don't use $AUTOREV Jul 08 21:32:05 ok Jul 08 21:32:09 in a week or later you can switch back Jul 08 21:32:21 (if I find the time to fix it) **** ENDING LOGGING AT Mon Jul 09 02:59:58 2012