**** BEGIN LOGGING AT Sun Jun 01 02:59:58 2014 Jun 01 16:21:59 Does anyone know if thumb2 ISA is backwards compatible with ARM Jun 01 16:22:22 ISA Jun 01 16:22:47 IrGGu^^: define "backwards compatible" Jun 01 16:23:08 With packages compiled with ARM ISA Jun 01 16:23:34 yes, it is Jun 01 16:24:30 that is how cssu-thumb works - only some packages are thumb2 compiled, the rest are stock ARM Jun 01 16:25:27 I see... So only system components do have to really make use of Thumb2 right? Jun 01 16:27:03 As the low level stuff handles memory management AFAIK Jun 01 16:27:41 not really, applications also benefit from reduced code size and newer gcc Jun 01 16:28:26 Despite not being compiled for Thumb2? Jun 01 16:30:29 I meant - if they are compiled with thumb2 ISA. It doesn't make much of a sense to use gcc4.7.2 from cssu-thumb "sdk" and to not thumb2 compile them Jun 01 16:30:51 anyway you'll have to run those on cssu-thumb, because of the libstdc++ version Jun 01 16:31:01 and libgcc1 Jun 01 16:31:44 Thanks Jun 01 16:35:27 freemangordon: giff gcc4.9 Jun 01 16:36:11 kerio: toldya, feel free to build and package it for SB :P Jun 01 16:36:21 I'll take it from there Jun 01 16:36:42 wget gcc49.tar.gz && tar zxvf gcc49.tar.gz && cd gcc49 && ./configure && make && make install Jun 01 16:36:50 i bet nothing will go wrong Jun 01 16:38:11 haha Jun 01 16:38:15 right? Jun 01 16:38:37 ever tried to build gcc? Jun 01 16:38:51 once Jun 01 16:38:52 keep in mind, we need it for ScratchBox Jun 01 16:38:55 everything went wrong Jun 01 16:39:17 :)) Jun 01 16:44:35 how weird :D Jun 01 19:03:39 * DocScrutinizer05 *eg* Jun 01 19:04:28 freemangordon: congrats for latest kernel build and maemo test run Jun 01 19:05:18 freemangordon: could you port the stock maemo kernel cam drivers? Jun 01 19:05:49 N900 camera interface is pretty non-default by design, with the mux and all Jun 01 19:55:07 DocScrutinizer05: we'd better fix maemo to use the upstream drivers. which still doesn't work, but that's another story :) Jun 01 19:55:21 *don't work Jun 01 20:04:44 maemo cannot use "upstream drivers" since there's no other device like N900, so most likely no upstream drivers existing fro the cam hw interface Jun 01 20:06:11 N900 has a mux on the single existing cam if of SoC, and the kernel drivers for cam0 and cam1 are tweaked to switch that mux according to which of both device nodes got opened atm - AFAIK Jun 01 20:06:23 (afaik for the latter bit) Jun 01 20:06:35 (the mux is for sure) Jun 01 20:07:30 so I guess you will need special N900 cam drivers to handle that mux, or it never will work as expected Jun 01 20:21:20 lemme put it this way: you'll have *severe* trouble initializing the cam sensor chip on module load when the mux isn't connecting the cam chip to the appropriate interface. And your drivers for cam0 and cam1 are fighting to set different parameters to same SoC interface Jun 01 20:21:43 this *can't* work in my book Jun 01 20:23:26 you might be able to use default aka upstream drivers for the one cam that by default is connected to SoC via mux (when mux control is in power on [reset] default state) if and only if you don't dare to load the kernel driver for the other cam as well Jun 01 20:28:03 for your reference: the mux consists of chips N5801 and N5802 (page 2), and the whole thing is controlled (cam A / cam B) via GPIO97 Jun 01 20:28:35 which as well serves as "CAM_B_EN" for the front cam Jun 01 20:29:33 main cam also sits on I2C bus3 Jun 01 20:35:59 btw totally nuts is the touchscreen controller which (data lines) is connected to the LCD(!), not to the SoC Jun 01 20:47:07 front cam also sits on I2C3 Jun 01 20:52:56 freemangordon: ^^^ Jun 01 20:56:48 btw page5 suggests that it's I2C_2 (two!) that goes to flex, no I2C_3 anywhere Jun 01 20:57:51 page12 claims it's I2C_3 that's running to the front cam module Jun 01 20:59:13 one of both must be incorrect Jun 01 21:02:01 I'd rather think they placed front cam on I2C_2 so it doesn't collide with the other (main) cam on I2C_3 Jun 01 22:07:34 sixwheeledbeast: bme chargefloat2 log has been created on 23:46, 1 June 2014 by sixwheeledbeast ?? Jun 01 22:08:15 I made it registered users only for a few months. Jun 01 22:09:03 and why you renamed it? Jun 01 22:09:14 I didn't AFAIK Jun 01 22:09:24 then I got that message wrong Jun 01 22:09:30 Only clicked protect Jun 01 22:11:13 If a page gets continuous spam from different IP's I have started making them registered only for a few months. Jun 01 22:12:26 >...has been *created* ...<< Jun 01 22:12:30 weird Jun 01 22:18:16 anyway a autoblocker&autoreverter on anonymous edits with subject: SUBJ1 would beter serve the purpose. No idea how to implement that, though Jun 01 22:19:55 or finally find the dang broken script in the internets and nuke or even fix it (though I hate fixing scriptkidde "tools" for messing with mediawiki content) Jun 01 22:57:39 <_jester_> hello world Jun 01 23:02:44 hello jester Jun 01 23:02:48 i am the world Jun 01 23:57:29 LOL **** ENDING LOGGING AT Mon Jun 02 02:59:59 2014