**** BEGIN LOGGING AT Sun Oct 23 02:59:57 2011 Oct 23 05:46:24 moin Oct 23 07:10:32 <[Rui]> moin Oct 23 07:18:15 moin Oct 23 07:18:57 [Rui]: hmm... does it work meanwhile? Oct 23 07:19:12 <[Rui]> nyet Oct 23 07:19:26 hmm Oct 23 07:19:26 <[Rui]> I'm going to try the lite Oct 23 07:19:53 [Rui]: http://build.shr-project.org/shr-core/images/om-gta02/shr-full-20111021-om-gta02.rootfs.tar.gz Oct 23 07:19:56 is what you tried? Oct 23 07:20:10 <[Rui]> yup Oct 23 07:20:28 ok... try the lite one... I will try that one... Oct 23 07:21:09 * [Rui] nods Oct 23 07:21:32 <[Rui]> any particular reason I should also upgrade the qi? Oct 23 07:21:43 no Oct 23 07:21:57 not that I know of Oct 23 07:22:03 mine is quite olde Oct 23 07:22:21 <[Rui]> oh it's the same hash, so it's the same Oct 23 07:22:37 [Rui]: did you try in SD or NAND Oct 23 07:22:42 <[Rui]> d Oct 23 07:22:43 <[Rui]> sd Oct 23 07:22:50 ok Oct 23 07:23:27 bbiab with the result Oct 23 07:25:05 <[Rui]> starting boot Oct 23 07:25:30 <[Rui]> hms... same. I see entering (old_state=off, new_state=normal) then nothing else Oct 23 07:26:32 I've seen segfault on n900, but on om-gta02 it works fine for me Oct 23 07:28:21 <[Rui]> diaper, brb Oct 23 07:40:02 <[Rui]> meh, the lite didn't go much further. one line more, only, kernel panic - not syncing: VFS: :Unable to mount root fs on unknown-block(31,6) Oct 23 07:40:03 <[Rui]> brb Oct 23 07:54:05 [Rui]: accidently I downloaded the lite one... but that one works fine for me Oct 23 07:55:09 [Rui]: at least it boots and the first start wizard works Oct 23 07:55:20 * mrmoku puts in a SIM Oct 23 07:57:10 [Rui]: PIN works too Oct 23 07:58:08 [Rui]: it registered Oct 23 07:58:36 [Rui]: and a call to the voice box worked too Oct 23 08:00:42 <[Rui]> maybe it needs rootwait (adding it to append-GTA02) Oct 23 08:02:07 <[Rui]> seems to not help Oct 23 08:02:10 [Rui]: yeah, maybe something like that Oct 23 08:02:30 rootwait should be there already Oct 23 08:02:52 <[Rui]> JaMa: not on append-GTA02 it isn't :| Oct 23 08:02:56 indeed, my /proc/cmdline has it Oct 23 08:03:09 <[Rui]> maybe it's a default, now Oct 23 08:03:13 [Rui]: from Qi Oct 23 08:03:17 <[Rui]> ok Oct 23 08:03:50 [Rui]: but worth trying adding glamo.sd_max_clk=1000000 to append-GTA02 Oct 23 08:04:27 <[Rui]> ok, so let's try that :) Oct 23 08:05:31 <[Rui]> now I have the shr splash logo Oct 23 08:06:06 [Rui]: first boot takes quite some time Oct 23 08:06:08 <[Rui]> is that slowing the µsd card? Oct 23 08:06:12 <[Rui]> mrmoku: yeah Oct 23 08:06:17 [Rui]: yes it is Oct 23 08:06:36 <[Rui]> JaMa: so I could try using bigger numbers to finetune it, then Oct 23 08:07:00 yes Oct 23 08:07:06 <[Rui]> will give it a few minutes more Oct 23 08:07:33 [Rui]: also try to boot with POWER pressed for a bit longer to enable debug mode from Qi Oct 23 08:07:45 so you'll see output and know what's taking so long Oct 23 08:07:53 <[Rui]> wizard from e Oct 23 08:08:12 good :) Oct 23 08:08:24 <[Rui]> sizing? Oct 23 08:08:33 <[Rui]> what should I choose for gta02? Oct 23 08:08:39 biggest is good Oct 23 08:08:40 I choose bottom left normally Oct 23 08:08:42 :) Oct 23 08:09:09 <[Rui]> pin is compressed to the left Oct 23 08:09:17 wait a moment Oct 23 08:09:30 <[Rui]> before being able to use it, the shr wizard started Oct 23 08:09:34 first comes the shr part of the wizard Oct 23 08:09:35 yeah Oct 23 08:09:50 that's ok Oct 23 08:09:59 [Rui]: don't skip the root password Oct 23 08:11:00 mrmoku: isn't it fixed already (root password?) Oct 23 08:11:16 JaMa: in the image from 21st? Oct 23 08:11:21 <[Rui]> I think I see what people meant by the bug in the softkey Oct 23 08:11:39 <[Rui]> hms, icon theme isn't set by default Oct 23 08:11:41 <[Rui]> it should Oct 23 08:12:06 right click -> reload Oct 23 08:12:11 (or refresh) Oct 23 08:12:29 <[Rui]> ok Oct 23 08:12:40 mrmoku: debug-tweaks were added 21 Oct 2011 09:19:43 so I hope yes, but not sure Oct 23 08:12:56 JaMa: ok, did not check Oct 23 08:12:58 <[Rui]> I'm probably going to extract that right click buttun into some code that can be mapped to a key Oct 23 08:13:03 <[Rui]> for e it's probably easy Oct 23 08:13:07 time is passing fast :P Oct 23 08:13:12 <[Rui]> but I'd like it working in gnome too Oct 23 08:13:27 * [Rui] is thinking of his tablet :) Oct 23 08:13:52 <[Rui]> I wonder if the sms bug is still there Oct 23 08:13:59 I'll build another image today with BQ27x00 enabled, testing it now Oct 23 08:14:17 mrmoku: and I've merged those linux.inc changes.. so your THUMBEE should be fixed now Oct 23 08:14:32 <[Rui]> ah my multi-shift patch is there :) Oct 23 08:15:00 efl is very recent :) Oct 23 08:15:14 <[Rui]> that's also helpful. Oct 23 08:15:25 <[Rui]> now I need to follow the instructions to set up my shr-core build environment Oct 23 08:17:09 <[Rui]> this is already setup for shr-core? http://shr-project.org/trac/wiki/Building%20SHR Oct 23 08:17:20 yes Oct 23 08:17:30 it's for all versions Oct 23 08:18:25 <[Rui]> ok, setting up chroot Oct 23 08:19:12 JaMa: Am I skipped something? No more updates for GTA02? Oct 23 08:19:48 Q-Master: using shr-u? Oct 23 08:20:17 yep Oct 23 08:20:33 Q-Master: we're all working on shr-core.. Oct 23 08:20:44 ah. ok. Oct 23 08:21:01 I was just interested Oct 23 08:30:18 JaMa: ok, thx Oct 23 08:30:50 [Rui]: incoming call worked too :) Oct 23 08:31:21 <[Rui]> mrmoku: I set it up into a new µsd card, so if things go bad, I just swap it :) Oct 23 08:31:24 hmm BQ27xx00 failed to register, so I'll push only updated linux.inc Oct 23 08:31:33 [Rui]: good plan :-) Oct 23 08:31:48 <[Rui]> mrmoku: I was wondering if the fscking "sms storing in sim as well" bug was still there Oct 23 08:32:00 I fear yes :/ Oct 23 08:32:05 SHR: 03Martin.Jansa 07meta-smartphone * rbf7e4102ba56 10/ (5 files in 5 dirs): linux.inc: sync with cleaned version from meta-handheld Oct 23 08:32:07 <[Rui]> I already need to plug my sim into a phone to clean it up :( Oct 23 08:32:10 SHR: 03Martin.Jansa 07meta-smartphone * raee870d71b34 10/ (5 files in 5 dirs): linux.inc: be more strict in CONF_SED_SCRIPT Oct 23 08:32:53 [Rui]: last news I know about is that dos1 wanted to dig into opimd code for that bug... some months ago Oct 23 08:33:05 <[Rui]> opim... hms... Oct 23 08:33:15 <[Rui]> I'm sorry but I don't have many good words for it :) Oct 23 08:33:24 my problem with sms is that they're quite expensive to test for me Oct 23 08:33:36 <[Rui]> hms... Oct 23 08:33:59 <[Rui]> I might be able to test sending from my job's phone costs next to nothing for me, as long a it's not in the hundreds or something :) Oct 23 08:40:40 SHR: 03Martin.Jansa 07shr-chroot * r7a6eccd91997 10/ (303 files in 32 dirs): system upgrade Oct 23 09:02:16 * mrmoku off for family trip Oct 23 09:02:17 bbl Oct 23 10:02:57 <[Rui]> well, so far it's compiling, so I'll just let it finish trying to bake shr-lite-image :) Oct 23 10:03:00 <[Rui]> see ya Oct 23 10:53:19 jama included upstream bq27x00? it's just crap for all I know Oct 23 10:55:33 DocScrutinizer: shr 2.6.39 does not compile it Oct 23 10:56:26 DocScrutinizer: it uses the platform battery driver. noticed this since I olny have voltage_now and not current_now Oct 23 10:57:15 DocScrutinizer: is the non-upstream bq27x00 driver better? Oct 23 10:57:48 lindi-: OM's bq27200 driver supports more important registers iirc. Oct 23 10:58:02 And also works with HDQ (that we need). Oct 23 10:58:26 PaulFertser: ok Oct 23 10:58:50 but at the moment neither is used in shr 2.6.39 Oct 23 11:03:37 honestly this is a clear case for sending upstream a "patch" that nukes the original bq27x00 drive with --834 and replaces a slightly beautified and augmented version of OM driver for it Oct 23 11:05:03 I guess we should first get it working with 2.6.39 Oct 23 11:07:02 I guess there can't be much that blocks this, for bq27200/i2c. Not sure for bq27000/HDQ and the FIQ it uses Oct 23 11:07:41 SHR: 03morphis 07meta-smartphone * rc806a77afb39 10/meta-fso/recipes-freesmartphone/freesmartphone/fso-specs_git.bb: meta-fso: fso-specs: bump SRCREV and PV Oct 23 11:07:42 SHR: 03morphis 07meta-smartphone * r129f7cb9bc3a 10/meta-aurora/recipes-aurora/aurora/aurora-daemon_git.bb: meta-aurora: aurora-daemon: add libfso-qt as dependency Oct 23 11:07:42 SHR: 03morphis 07meta-smartphone * r2b89c366c91b 10/meta-fso/recipes-freesmartphone/freesmartphone/qfsodbusxml2cpp_git.bb: meta-fso: qfsodbusxml2cpp: import from oe-dev and add BBCLASSEXTEND = "native" Oct 23 11:07:42 SHR: 03morphis 07meta-smartphone * rbb31fd6ffcb2 10/meta-fso/recipes-freesmartphone/freesmartphone/libfso-glib_git.bb: meta-fso: libfso-glib: bump SRCREV and PV Oct 23 11:07:43 SHR: 03morphis 07meta-smartphone * r5081edecd55b 10/meta-fso/recipes-freesmartphone/freesmartphone/libfso-qt_git.bb: meta-fso: libfso-qt: import from oe-dev Oct 23 11:47:45 freesmartphone.org: 03morphis 07cornucopia * rb8c7dbef9169 10/fsogsmd/src/plugins/lowlevel_samsung_crespo/plugin.vala: Oct 23 11:47:45 freesmartphone.org: Revert "fsogsmd: lowlevel_samsung_crepso: remove unnecessary sleep phase" Oct 23 11:47:45 freesmartphone.org: Without a sleep we're sending commands to the modem but the modem is not ready set to Oct 23 11:47:45 freesmartphone.org: receive and process them. Oct 23 11:47:46 freesmartphone.org: This reverts commit 5130de4552c1c1b597a3ab48cb6430284a7faec0. Oct 23 11:47:46 freesmartphone.org: 03morphis 07cornucopia * rf88a975ded97 10/fsogsmd/src/plugins/modem_samsung/ (6 files): Oct 23 11:47:47 freesmartphone.org: fsogsmd: modem_samsung: several updates Oct 23 11:47:47 freesmartphone.org: Signed-off-by: Simon Busch Oct 23 12:09:18 hi. don't know where else to ask. about palm pre kernel, is there a newer kernel than 2.6.24? if not, is it because of binary blobs not working on more recent kernel or is it because of lack of manpower (devs)? Oct 23 12:16:10 morphis, ping Oct 23 12:20:06 good morning people Oct 23 13:52:54 pespin: I've located the genlist issue Oct 23 14:10:57 zub, I was having lunch ;) Oct 23 14:11:03 zub, nice, which was the problem? Oct 23 14:12:33 http://linux.fjfi.cvut.cz/~zub/SHR/libeflvala/0004-elementary-fix-genlist-argument-passing.patch Oct 23 14:12:41 passing address of a local... Oct 23 14:12:57 recent valac started copying the args Oct 23 14:13:03 that's why it broke Oct 23 14:13:07 so I marked it as ref Oct 23 14:13:10 and it works Oct 23 14:13:24 but there are likely more such issues throughout libeflvala Oct 23 14:13:43 and when I ctrl-c the example I see a different SIGSEGV, so there are still issues Oct 23 14:14:00 (but this sigsegv happens even if I don't run the genlist example) Oct 23 14:14:25 zub, hmm ref? I didn' know about that word, but afaik efl classes doesn't have reference count, so that doesn't make much sense to me :P Oct 23 14:14:33 zub, maybe tey should be set as unowned Oct 23 14:14:50 maybe you should have a look at what I'm talking about :-P Oct 23 14:15:21 yeah, that's true too.. sorry a bit sleepy after eating lol Oct 23 14:15:55 I'll look at thar ref stuff in doc Oct 23 14:17:22 http://fpaste.org/2TG6/ Oct 23 14:17:41 the old code just passes &self->priv->m_item_class Oct 23 14:17:49 the new code copies that, pass address of that Oct 23 14:18:06 if elementary keeps the pointer, as it seems to do... sigsegv Oct 23 14:18:30 as it keeps pointer to something on stack, that disappeared meanwhile Oct 23 14:19:13 <[Rui]> hey Oct 23 14:19:15 (the trouble is about the Elm_Genlist_Item_Class arg (second) of elm_genlist_item_append) Oct 23 14:20:12 zub, so you are passing a copy of the object to elm internals? Oct 23 14:20:28 <[Rui]> Looking good, so far ==> NOTE: Running task 1412 of 6641 (ID: 2221, /OE/shr-core/meta-openembedded/meta-oe/recipes-graphics/cairo/cairo_1.10.2.bb, do_configure) Oct 23 14:20:29 nö Oct 23 14:20:40 I'm just making elm keep pointer to the real thing Oct 23 14:20:50 not to some gon-by-the-time-elm-wants-to-use-it temporary Oct 23 14:20:57 zub, ah with new/old code you mean the generated by valac right? Oct 23 14:21:01 yes Oct 23 14:21:04 ok ;) Oct 23 14:22:33 http://docs.enlightenment.org/auto/elementary/group__Genlist.html#gaaff65b97fbdb4b47862ed0a2b85839db - in the spirit of crappy doc, it doesn't say a wor about the lifetime of the "itc" argument Oct 23 14:22:56 but I'v always seen that the actual Item class is kept throughout the lifetime Oct 23 14:23:08 and form behaviour I do think they just keep the pointer and assume it's alive Oct 23 14:24:37 zub, the question I was asking before is, wouldn't be more correct to pass ownership of the var to Genlist? as in setting the param as owned. That should avoid the copying at C level and should be more correct conceptually as far as I understand it. But I'm not sure Oct 23 14:25:17 well... how would you "pass the ownership" then? Oct 23 14:25:31 zub, using "owned" keyword Oct 23 14:25:40 go ahead, and let me watch Oct 23 14:25:46 like you do in HashTable Oct 23 14:26:17 as I was saying I'm not sure, but you could try calling func( ... , (owned) param_to_pass_ownerhsip, ...); Oct 23 14:26:32 that should avoid unrefing the param in the func Oct 23 14:26:38 well, what would the generated code do? how would it "pass ownership" to elm who doesn't care? Oct 23 14:27:05 zub, basically, if you are in func2 and call that func() passing ownesership Oct 23 14:27:14 afaik then you are not deleting/unrefing the param Oct 23 14:27:20 as you are not the owner anymore Oct 23 14:27:29 which I suppouse it's the problem you were having Oct 23 14:27:31 do youknow C? Oct 23 14:27:40 yeah Oct 23 14:27:48 so... how would that work with a plain stuid struct? Oct 23 14:28:14 wherte the called function wants a pointer... and keeps it, and doesn't give a ... Oct 23 14:28:33 anyway, do it and let me know, /me is interested Oct 23 14:29:22 zub, are you saving the GenItemClass objects somewhere? ie. in a HashTable or a array? Oct 23 14:29:38 attribute/member var Oct 23 14:29:45 (however vala calls it) Oct 23 14:30:12 after all, it's reused for all items in this case, and you often want to reuse it I believe Oct 23 14:30:28 zub, all of them? are you transfering ownership of objects to this container? Oct 23 14:30:42 what container are you talking about? Oct 23 14:30:51 hashtable/array Oct 23 14:30:58 class MyWindow { private Elm.GenlistItemClass m_item_class; } Oct 23 14:31:04 as I said: 16:24:42 < zub> attribute/member var Oct 23 14:31:12 I'm not keepig it in any container Oct 23 14:31:22 zub, ah sorry my fault, I was talking about GenItem Oct 23 14:31:46 I never used the Genlist api, but I fighted many times with list and other apis ;) Oct 23 14:32:02 * zub doesn't like EFL Oct 23 14:32:31 <[Rui]> WARNING: For recipe dbus-glib, the following files were installed but not shipped in any package: Oct 23 14:32:31 <[Rui]> WARNING: /usr/libexec/dbus-bash-completion-helper Oct 23 14:32:31 <[Rui]> WARNING: /etc/bash_completion.d/dbus-bash-completion.sh Oct 23 14:32:51 zub, the problem is that efl classes doesn't have referene count, so you need to "save" the efl items somewhere passing ownership to that data structure container Oct 23 14:33:19 zub, so in case you are not doing that, I guess there's part of your problem Oct 23 14:33:29 sigh... Oct 23 14:33:35 you have to pass items to Genlist API Oct 23 14:33:36 I disagree Oct 23 14:33:38 and I'm frustrated Oct 23 14:33:53 but you need to keep them alive manually so vala doesn't free them automatically Oct 23 14:34:01 ok, once again Oct 23 14:34:12 we're talking of the item class argument Oct 23 14:34:20 I keep that alive Oct 23 14:34:32 but valac generates C code that passes adress-of-a-temporary-copy to efl Oct 23 14:34:45 so no matter what I do with my original, efl gets destroyed copy Oct 23 14:34:59 after all, the same issue is happenign with the genlist example in libeflvala Oct 23 14:35:16 and IIRC you have seen the same issue yourself Oct 23 14:36:29 * pespin trying to look at the genlist api you posted here but it's not loading u.u Oct 23 14:37:03 http://docs.enlightenment.org/auto/elementary/group__Genlist.html + search for elm_genlist_item_append() Oct 23 14:37:49 it's not loading here :S Oct 23 14:39:36 !google efl documentation Oct 23 14:40:47 it's down... I'll look at Elementary.h Oct 23 14:41:51 indeed, it's dead now Oct 23 14:41:56 <[Rui]> pespin: server is having trouble, raster's looking into it now. Oct 23 14:42:07 [Rui], ok :) Oct 23 14:42:24 <[Rui]> apparently i broke while I was svn updating on my vm :) Oct 23 14:42:35 <[Rui]> not due to it, of course. Oct 23 14:44:40 pespin: just look at the generated code; if efl assumes it can use the pointer (which is does, I just discussed it in #e), then it will not work -> certainly somehing has to be changed in libeflvala Oct 23 14:44:59 using ref looks rational to me (and it works), though if you know of better fix, go ahead Oct 23 14:45:09 yeah, reading the doc in Elementary.h before looking at the example vala and C code :) Oct 23 14:45:43 zub, I'm not saying it's a wrong fix :) Oct 23 14:46:37 <[Rui]> back online Oct 23 14:47:11 ok thanks Oct 23 14:57:06 zub, ok, I thought the append() func was similar to the one used in List and it's not, sorry ;) Oct 23 14:58:24 na ja Oct 23 14:58:28 so, now you like the fix? Oct 23 15:00:31 still looking :) Oct 23 15:14:54 morphis: ping Oct 23 15:21:46 zub, ok, it seems correct to me, let me apply the patch and try + look at the C code :) Oct 23 15:25:37 wunderbar Oct 23 15:28:04 zub, wunderbar? Oct 23 15:28:29 https://secure.wikimedia.org/wiktionary/en/wiki/wunderbar Oct 23 15:29:21 ok :P Oct 23 15:31:34 zub, ok, you get ack for my part if that's useful for something hehe Oct 23 15:31:51 zub, you can ask mickeyl or playya to push the patch :) Oct 23 15:36:25 mickeyl, playya: would someone be willing to commit this http://linux.fjfi.cvut.cz/~zub/SHR/libeflvala/0004-elementary-fix-genlist-argument-passing.patch ? Oct 23 15:37:14 AFAIK it fixes genlist for current (0.14) valac Oct 23 15:37:26 though I have not tested with 0.12, /me is gonna try it out Oct 23 15:39:31 ok, it seems non-destructive, works both with 0.12.1 and 0.14.0.36-3d153 Oct 23 15:42:31 I'd also push for http://linux.fjfi.cvut.cz/~zub/SHR/libeflvala/0002-elementary-rename-elm-to-elementary-to-match-EFL-cha.patch but unfortunately nobody cares that it's broken and works by accident Oct 23 15:44:08 as for the ecore vs. ecore_evas thing, I gave up; it's complicated, I think libeflvala should expolicitly request it via pkg-config, but whatever Oct 23 16:03:41 this would probably better be discussed at some mailing list Oct 23 16:03:55 what would be the appropriate place for libeflvala? Oct 23 16:04:53 morphis: hi. So you asked me a question yesterday that I did not understaood. "whats with the kernel from shr buildhost?" Oct 23 16:05:36 Plus, finaly what is the diff between your working kernel and thoses building by me or by shr-core buildhost Oct 23 16:05:49 also, w.r.t. ecore-evas, it's needed only in the one example, so better solution is to put it into that one example's deps, no need to put it into libeflvala itself - btu then again, it would make sense to split ecore-evas from ecore.vapi Oct 23 17:02:48 mickeyl: ping Oct 23 17:07:57 morphis: ping Oct 23 17:08:50 mrmoku: ping Oct 23 17:08:54 s/ping/pong/ Oct 23 17:12:47 morphis: hey :) Oct 23 17:12:53 one question Oct 23 17:13:12 the gta04 gives the provider name in UCS Oct 23 17:13:31 and the atmediator doesn't handle that Oct 23 17:13:56 UCS was some character encoding format? Oct 23 17:14:02 is the correct fix to fix AtNetworkGetStatus? Oct 23 17:14:03 yup Oct 23 17:14:12 or should I override it for gta04 only Oct 23 17:14:13 ? Oct 23 17:14:18 hm Oct 23 17:14:40 can we retrieve the information about the encoding format from the modem? Oct 23 17:14:52 wasn't there some command for this? Oct 23 17:14:55 there is modem.charset Oct 23 17:15:09 who sets it? Oct 23 17:16:15 lib/at/atchannel.vala: Oct 23 17:20:58 morphis: there is decodeString which handles the charset correctly... but is not used for all atcommands Oct 23 17:21:42 ok Oct 23 17:21:57 then try to use in AtNetworkGetSTatus Oct 23 17:22:08 good, thx Oct 23 17:27:44 morphis: hmm... damn... that would probably break gta02 Oct 23 17:28:09 it is configured for UCS2 too... but does not seem to send provider name in +COPS as UCS2 Oct 23 17:28:53 mickeyl, playya I just requested commit access for eflvala at e-devel ml, it would be nice if you could answer there :) Oct 23 17:31:03 hi! Oct 23 17:32:00 hey Slyon Oct 23 17:39:00 mrmoku, did you get a serial connection on you gta04? Oct 23 17:39:46 s/on you/to your/ Oct 23 17:49:17 mrmoku: we can make a config option for it Oct 23 17:57:23 morphis: hmmm Oct 23 17:57:34 Slyon: yup, serial is working Oct 23 17:59:41 morphis: would be interesting to hear what mickeyl thinks... maybe at modems out there are too different Oct 23 18:04:29 morphis: looking at the gsm standard there is nothing written about oper being in the selected charsets for +COPS Oct 23 18:04:52 mrmoku, hmm strage i don't get a login promt at the serial console Oct 23 18:05:04 and i had to remove u-boot, as it failed to build Oct 23 18:05:08 so probably what fsogsmd does right now is correct and maybe overriding for gta04 is the correct thing to do Oct 23 18:05:33 Slyon: possible I did not push the adjusted inittab Oct 23 18:05:45 mrmoku, and my e crashes Oct 23 18:05:49 or better... I probably adjusted it on my phone only :) Oct 23 18:06:09 hmm Oct 23 18:06:56 Slyon: did you *un*check composite? Oct 23 18:07:24 mrmoku, i didn't get to any screen Oct 23 18:07:35 except the "e crashed" message Oct 23 18:07:36 ahh, not even the first start wizard? Oct 23 18:07:40 nope Oct 23 18:07:46 heh, ouch Oct 23 18:08:24 what did you adjust in you inittab ttyO2 > ttyS2 ? Oct 23 18:08:27 S:2345:respawn:/sbin/getty 115200 ttyS2 Oct 23 18:08:36 ttyO2 is for the new kernel Oct 23 18:09:11 mrmoku, hmm i changed this in the om-gta04.conf machine file but it doesn't seem to be applied to the image Oct 23 18:10:14 ok now i have serial login Oct 23 18:10:45 hmm... interesting... I have a signal strength and a provider name without any SIM inserted :P Oct 23 18:11:40 heh Oct 23 18:11:43 mrmoku: when mickeyl is available again, yes :) Oct 23 18:12:24 morphis: I'm a bit frightened that I would break other modems touching stuff in lib/at Oct 23 18:12:40 so I think I will do it for gta04 only Oct 23 18:12:43 hm, then you should override the mediator for now Oct 23 18:12:45 yes Oct 23 18:12:47 ok Oct 23 18:17:21 Slyon: anything in Xorg.log or Xsession.log that gives some insight about e crashing? Oct 23 18:17:45 mrmoku, yes... all modules except omapfb can't be loaded Oct 23 18:18:14 Slyon: yeah, that's an interesting thing... the extension modules like dbe, dri ... are not installed Oct 23 18:18:24 but it works for me without them Oct 23 18:18:30 hm... ok Oct 23 18:19:53 except of this I cant see anything in Xorg.log but: Oct 23 18:20:20 if i start the enlightenment_start.oe script i get this line: Enlightenment cannot initialize Ecore_X! Oct 23 18:20:51 hmm Oct 23 18:25:41 mrmoku, did you change anything else? Oct 23 18:38:33 Slyon: hmm Oct 23 18:38:41 Slyon: which kernel are you using? Oct 23 18:39:01 the one compiled by OE Oct 23 18:39:43 that one does *not* work for the modem for the very least Oct 23 18:40:02 maybe it has other problems as well Oct 23 18:40:12 might be some nasty toolchain issue :/ Oct 23 18:40:26 I'm using the one from goldelico/unstable Oct 23 18:40:41 hmm... i'll try with goldelicos kernel Oct 23 18:40:50 that's very interesting Oct 23 18:41:24 leviathan: I don't remember... is hso working in 3.1 ? Oct 23 18:41:38 leviathan: do you get /dev/ttyHSx ? Oct 23 18:42:37 hmm Oct 23 18:42:38 wait Oct 23 18:42:52 Slyon: btw. I had fixed u-boot to build from the correct tree... did not commit because that one did not work as well... and lost my fix by make update :/ Oct 23 18:43:20 mrmoku: no, I don't get ttyHS* Oct 23 18:43:24 ok Oct 23 18:43:36 while it says it would have been detected Oct 23 18:43:42 maybe I just don't have the driver Oct 23 18:43:43 mrmoku, ok. i just commented u-boot for now Oct 23 18:43:43 :-) Oct 23 18:43:58 mrmoku, with the goldelico/unstable kernel it's the same: e crashes Oct 23 18:44:03 ok Oct 23 18:44:22 leviathan: with the old kernel modprobe hso is enought to make them appear Oct 23 18:44:25 -t Oct 23 18:45:01 | /OE/shr-core/tmp/sysroots/x86_64-linux/usr/bin/gperf: error while loading shared libraries: libstdc++.so.6: cannot open shared object file: No such file or directory Oct 23 18:45:07 gah... getting that one again :/ Oct 23 18:47:55 mrmoku: ah Oct 23 18:47:56 :-D Oct 23 18:47:59 there they are Oct 23 18:48:00 ^^ Oct 23 18:48:35 ttyHS0 ttyHS1 ttyHS2 ttyHS3 ttyHS4 ttyHS5 Oct 23 18:48:38 strange Oct 23 18:48:46 leviathan: great Oct 23 18:48:55 6 devices? O_o Oct 23 18:48:59 hmm Oct 23 18:49:01 but anyway Oct 23 18:49:02 yeah Oct 23 18:49:04 I've got them Oct 23 18:49:05 :-) Oct 23 18:49:33 Slyon: what do you think... should we go for the new kernel? help leviathan a bit? Oct 23 18:49:59 yeeepi, someone helps me ^^ Oct 23 18:51:09 mrmoku, i guess we should go for 3.x, what's missing there? Oct 23 18:51:11 leviathan, ^^^ Oct 23 18:51:33 I've got an error in fbtimings for some unknown reason Oct 23 18:51:52 I'd like to merge TI OMAP direct rendering manager into our kernel Oct 23 18:51:58 for getting 3D acceleration Oct 23 18:52:06 in order to use plasma active Oct 23 18:52:08 :-) Oct 23 18:52:45 mrmoku, my display alternates balck/white about every 30s... :/ Oct 23 18:53:16 Slyon: that's normal :P Oct 23 18:53:20 ok^^ Oct 23 18:53:33 well... for me it alternates between X display and white Oct 23 18:53:42 yes. it's the same here Oct 23 18:53:47 white because backlight is not handled yet Oct 23 18:53:53 neither in old nor in new kernel Oct 23 18:54:06 so omapfb goes off... and backlight keeps on Oct 23 18:54:23 at least that's how *I* understand it :) Oct 23 18:54:49 heh, at least it's not only my device :) Oct 23 18:57:39 Slyon: so what do you think about the kernel? Oct 23 18:57:56 mrmoku, i'd go for the new one Oct 23 18:58:06 ok Oct 23 18:58:13 forward is the way :-P Oct 23 18:58:18 yup Oct 23 18:58:25 (which is one of the reasons there is no SHR stable ;) Oct 23 19:01:07 hmm.. Oct 23 19:08:38 Slyon: not to be misunderstood... I'm for the new kernel too :) Oct 23 19:16:09 freesmartphone.org: 03morphis 07cornucopia * r60e42d883eef 10/fsogsmd/src/plugins/modem_samsung/ (channel.vala modemstate.vala unsolicited.vala utils.vala): Oct 23 19:16:09 freesmartphone.org: fsogsmd: modem_samsung: use network state information for user information and state update Oct 23 19:16:09 freesmartphone.org: Signed-off-by: Simon Busch Oct 23 19:16:09 freesmartphone.org: 03morphis 07cornucopia * r76daa9930d63 10/fsogsmd/src/plugins/modem_samsung/ (Makefile.am mediators_device.vala plugin.vala): Oct 23 19:16:09 freesmartphone.org: fsogsmd: modem_samsung: register our device subsystem mediators Oct 23 19:16:09 freesmartphone.org: Signed-off-by: Simon Busch Oct 23 19:16:10 freesmartphone.org: 03morphis 07cornucopia * rc746ef30d4d4 10/fsogsmd/src/plugins/modem_samsung/ (mediators_sim.vala plugin.vala): Oct 23 19:16:10 freesmartphone.org: fsogsmd: modem_samsung: remove sim mediators we currently not support Oct 23 19:16:11 freesmartphone.org: Signed-off-by: Simon Busch Oct 23 19:16:11 freesmartphone.org: 03morphis 07cornucopia * r7b28bbb2435d 10/fsogsmd/src/plugins/modem_samsung/ (Makefile.am mediators_network.vala plugin.vala): Oct 23 19:16:12 freesmartphone.org: fsogsmd: modem_samsung: implement network mediators Oct 23 19:16:12 freesmartphone.org: Signed-off-by: Simon Busch Oct 23 19:16:13 freesmartphone.org: 03morphis 07cornucopia * r31c4ea3dbd75 10/fsogsmd/src/plugins/modem_samsung/ (Makefile.am mediators_pdp.vala plugin.vala): Oct 23 19:16:14 freesmartphone.org: fsogsmd: modem_samsung: add mediators for pdp subsystem Oct 23 19:17:10 freesmartphone.org: Signed-off-by: Simon Busch Oct 23 19:20:56 freesmartphone.org: 03morphis 07cornucopia * r29ed07f6aa0f 10/fsogsmd/src/plugins/modem_samsung/utils.vala: Oct 23 19:20:56 freesmartphone.org: fsogsmd: modem_samsung: save network reg state in global modem state Oct 23 19:20:56 freesmartphone.org: Signed-off-by: Simon Busch Oct 23 19:29:38 freesmartphone.org: 03morphis 07cornucopia * re83149dbaa52 10/fsogsmd/src/plugins/modem_samsung/unsolicited.vala: Oct 23 19:29:38 freesmartphone.org: fsogsmd: modem_samsung: don't report signal strength update when not registered with network Oct 23 19:29:38 freesmartphone.org: Signed-off-by: Simon Busch Oct 23 19:35:58 morphis: ^^ that one is interesting Oct 23 19:36:08 I have the same on gta04... even without any SIM inserted Oct 23 19:39:49 mrmoku: yes, seems like modern modems are connecting to network wether you don't told them to do so Oct 23 19:41:06 heh, yeah Oct 23 19:42:06 hmm but how do they know which network to connect to?? without a SIM? Oct 23 19:45:16 Slyon: don't know if connects to a network or just reports the best signal it get's from some network Oct 23 19:45:28 i see Oct 23 19:46:22 the modem of the gta04 internally is from qualcomm someone told me Oct 23 19:55:56 is that correct? Oct 23 19:57:26 morphis, yes Oct 23 19:57:39 morphis, http://www.option.com/en/products/products/embedded-solutions/lgamodulegtm601-609/specifications/#start Oct 23 19:57:51 it's an option gtm601w Oct 23 19:58:31 "Being a Qualcomm and Gobi licensee, Option has full access to the firmware that is running on the Qualcomm baseband processor. " Oct 23 20:07:51 argh, I just removed all shr-chroot, clond it, and whenever I do make setup-shr-core, I get this error: Oct 23 20:07:51 git: res_query.c:251: __libc_res_nquery: Assertion `hp != hp2' failed. Oct 23 20:08:00 same as before -.- Oct 23 20:18:10 Slyon: trying a fresh image now... lets see if it segfaults for me too now Oct 23 20:23:31 so Iam off Oct 23 20:23:32 gn8 Oct 23 20:23:48 gn8 morphis Oct 23 20:25:25 me too. gn8! Oct 23 20:25:32 gb8 Slyon Oct 23 20:25:37 heh Oct 23 20:25:41 ? Oct 23 20:25:46 does it work or not? Oct 23 20:27:03 Slyon: putting the SD in right now Oct 23 20:27:21 ok. i'll wait before i leave :) Oct 23 20:27:49 Slyon: ahh... your fixed boot screen :-) Oct 23 20:27:57 jop Oct 23 20:28:40 hmm Oct 23 20:28:54 Slyon: yup... segfault Oct 23 20:29:30 * mrmoku searches his guitar pick once again :P Oct 23 20:29:36