**** BEGIN LOGGING AT Mon Jan 23 02:59:57 2012 Jan 23 03:36:58 SHR: 03shr-devel 07buildhistory * r3bc8e5f043b0 10/packages/om_gta04-oe-linux-gnueabi/ (50 files in 50 dirs): Build 201201230336 of shr 20120123 for machine om-gta04 on shr-chroot Jan 23 04:57:09 SHR: 03morphis 07meta-smartphone * r37e7ef612c1a 10/meta-aurora/recipes-aurora/aurora/aurora-base.inc: meta-aurora: aurora-base: bump SRCREV Jan 23 05:54:51 a happy new year of the dragon to all Chinese users! :-) Jan 23 07:42:28 SHR: 03shr-devel 07buildhistory * r6f500f79105b 10/packages/armv7a-vfp-neon-oe-linux-gnueabi/libsamsung-ipc/ (8 files in 8 dirs): Build 201201230839 of shr 20120123 for machine om-gta04 on shr-chroot Jan 23 07:47:25 SHR: 03Martin.Jansa 07meta-smartphone * rb14c7cba1968 10/meta-fso/recipes-freesmartphone/freesmartphone/libsamsung-ipc_git.bb: libsamsung-ipc: add mtd-utils to DEPENDS Jan 23 09:11:24 JaMa: you have remarks about my efl sdk patch ? Jan 23 09:11:46 i need to increase my backlogs .... Jan 23 09:13:08 captainigloo: I've moved those recipes to right directories, but build was failing in gdb-canadian (not your fault) Jan 23 09:13:45 captainigloo: current version: http://git.openembedded.org/meta-openembedded-contrib/commit/?h=jansa/test&id=bed1e86d02b38f0da4c663b0d41d2a005c12ec58 Jan 23 09:14:56 hum ok Jan 23 09:15:36 so i guess that others meta-toolchiain should failed with current oe-core ? Jan 23 09:15:56 wrt gdb yes Jan 23 09:16:22 it's related to this http://git.openembedded.org/meta-openembedded-contrib/commit/?h=jansa/test&id=bed1e86d02b38f0da4c663b0d41d2a005c12ec58 Jan 23 09:16:25 http://patchwork.openembedded.org/patch/17129/ Jan 23 09:16:29 this ^ :) Jan 23 09:18:04 but this patch alone wasn't enough to fix it here, I'll try again today Jan 23 09:18:33 ok thanks :) Jan 23 09:36:49 Yippie, build fixed! Jan 23 09:36:49 Project shr-core-om-gta02-aurora-image build #13: FIXED in 28 min: http://norman-schleicher.de/jenkins/job/shr-core-om-gta02-aurora-image/13/ Jan 23 09:44:05 JaMa: I wasn't sure what your question was and we keep missing each other on IRC (I'm at work right now) Jan 23 09:44:24 JaMa: can you reply to the email on shr-devel? Jan 23 09:45:51 captainigloo: why is there empty task-efl-standalone-sdk-target-dbg which doesn't RDEPEND on anything pulled to meta-toolchain-efl? Jan 23 09:45:57 rah_: ok Jan 23 09:47:42 captainigloo: nvm it RRECOMMENDS something :) Jan 23 09:52:10 SHR: 03shr-devel 07buildhistory * r68c22c8b9ca2 10/packages/ (66 files in 66 dirs): Build 201201231015 of shr 20120123 for machine om-gta02 on shr-chroot Jan 23 10:02:05 captainigloo: after reverting gdb-cross-canadian support for python I ended with similar error like you http://paste.pocoo.org/show/539160/ Jan 23 10:02:22 rebuilding them to see why they doesn't exist Jan 23 10:05:50 hi, I cant clone from git.openmoko.org but I can pull from it Jan 23 10:06:11 any one knows a workarround Jan 23 10:07:23 Yippie, build fixed! Jan 23 10:07:24 Project shr-core-nokia900-aurora-image build #13: FIXED in 30 min: http://norman-schleicher.de/jenkins/job/shr-core-nokia900-aurora-image/13/ Jan 23 10:17:53 jluis: clones of some repos are on shr buildhost if it's enough for you Jan 23 10:20:27 SHR: 03shr-devel 07buildhistory * r15004eb509ba 10/packages/ (61 files in 61 dirs): Build 201201231052 of shr 20120123 for machine nokia900 on shr-chroot Jan 23 10:24:24 JaMa, could you give me the path for u-boot and qi? Jan 23 10:26:08 http://build.shr-project.org/sources/git2_git.openmoko.org.git.qi.git.tar.gz Jan 23 10:26:50 and OM's u-boot is not built by SHR so I have only tarball for upstream version.. Jan 23 10:30:17 Jama thanks qi is way more important than uboot (at least for me ) Jan 23 10:40:12 SHR: 03shr-devel 07buildhistory * r8e18f65cd6ac 10/packages/palmpre-oe-linux-gnueabi/fsogsmd/ (44 files in 44 dirs): Build 201201231120 of shr 20120123 for machine palmpre on shr-chroot Jan 23 10:59:18 SHR: 03shr-devel 07buildhistory * rfb582621bd37 10/packages/palmpre2-oe-linux-gnueabi/fsogsmd/ (44 files in 44 dirs): Build 201201231140 of shr 20120123 for machine palmpre2 on shr-chroot Jan 23 11:17:13 SHR: 03shr-devel 07buildhistory * r07b5b63afb8a 10/packages/crespo-oe-linux-gnueabi/fsogsmd/ (44 files in 44 dirs): Build 201201231159 of shr 20120123 for machine crespo on shr-chroot Jan 23 11:34:57 Project shr-core-deploy-cleanup build #6: SUCCESS in 1 min 45 sec: http://norman-schleicher.de/jenkins/job/shr-core-deploy-cleanup/6/ Jan 23 11:35:07 SHR: 03shr-devel 07buildhistory * r13bf9bd85f67 10/packages/om_gta04-oe-linux-gnueabi/fsogsmd/ (44 files in 44 dirs): Build 201201231216 of shr 20120123 for machine om-gta04 on shr-chroot Jan 23 11:37:28 Hi guys! Jan 23 11:38:52 GNUtoo: hi thx for the repport about pre kernel not starting! I should have done that a while ago but was tired of this bug that I though I was only to hit Jan 23 11:43:53 GarthPS, hi Jan 23 11:44:23 GarthPS, I tried as hard as I could to get a log from the kenrel but no log were produced Jan 23 11:44:33 GarthPS, if we get a log we could fix it Jan 23 11:44:37 GNUtoo: what pre do you hav e? Jan 23 11:44:41 else we have to drop it Jan 23 11:44:44 pre1 Jan 23 11:44:50 the first model Jan 23 11:44:54 the 8GB model Jan 23 11:45:19 ... it would be a shame as i personnaly put my interested to my phone under shr/auror in suspĂȘnd cause of this bug Jan 23 11:45:25 I have a pre2 Jan 23 11:45:38 then we need to fix that bug Jan 23 11:45:44 I don't know well the pre Jan 23 11:45:50 I tried with klog + bootie Jan 23 11:45:51 it failed Jan 23 11:45:59 but maybe I didn't do the right thing Jan 23 11:46:05 maybe there are other options Jan 23 11:46:10 like mtdoops, Jan 23 11:46:18 or if bootie can read memory regions Jan 23 11:46:26 that would permit to read the log buffer Jan 23 11:46:49 the big problem is taht the serial is unknown and seem to be behind the USB port Jan 23 11:46:49 I was hit since nearly the begining of shr-core by this bug. morphis tried a bit to look at it but switch back then to other devs (it was befor the FSOSHRCON the last time I know he worked on it) Jan 23 11:47:09 ok Jan 23 11:47:17 GNUtoo: GarthPS: which phone and linux version is having this suspend issue? please try to cite bug reports or kernel versions when you discuss these so that people who read the irc log can understand :) Jan 23 11:47:41 I think it's a toolchain problem Jan 23 11:47:47 with an old toolchain it works fine Jan 23 11:48:08 and the kenrel was already patched for the alignements issues Jan 23 11:48:26 basically there are 2 problems Jan 23 11:48:28 one is fixed Jan 23 11:48:33 but rest the other one at least Jan 23 11:48:38 lindi-: yop lindi thx but it is not a suspend issue but a but issue : last kenerl from SHR buil host does not boot on pre devices Jan 23 11:48:40 which we don't know why Jan 23 11:49:11 GNUtoo: did you speak about this with morphis ? Jan 23 11:49:22 yes we are about to drop it Jan 23 11:49:28 (the pre devices) Jan 23 11:49:37 so if you can help us fixing that bug.... Jan 23 11:49:43 we won't drop it Jan 23 11:50:23 GNUtoo: I would love to but I don't know how (no skills but can learn) Jan 23 11:50:42 I don't know the specificities of the palm pre enough Jan 23 11:50:46 that's my problem Jan 23 11:50:50 I know a bit bootie Jan 23 11:51:16 but I really wonder how to extract that kernel ring buffer/messages Jan 23 11:52:33 GNUtoo: I will try to ask some help to webosinternal Jan 23 11:52:39 ok Jan 23 12:21:14 GNUtoo: did you eventually work out that uSD card issue (Error -16 or whatever it was)? Jan 23 12:21:58 pabs3, yes I did Jan 23 12:22:09 but with a workarround Jan 23 12:22:18 here's my use case Jan 23 12:22:25 I've an SHR install on NAND Jan 23 12:22:32 with a vfat microsd Jan 23 12:22:52 so I've hardcoded the boot arguments in a modified QI Jan 23 12:23:06 the boot arguments are sd_max_clk Jan 23 12:23:18 which instead of putting to 1M Jan 23 12:23:22 I've put it to 2M Jan 23 12:25:54 GarthPS, what are the arguments to boot webos once SHR is installed? Jan 23 12:26:22 oops wrong channel Jan 23 12:26:29 (it's to dump the bootloader) Jan 23 12:28:01 GNUtoo: hhm, higher glamo clock rate usually prevents booting for me (Debian/SHR on uSD) Jan 23 12:28:22 yes but 2M was good for me Jan 23 12:28:31 it was the in-between right clock Jan 23 12:29:31 GNUtoo: how did you find the right one? bisect? Jan 23 12:30:02 1M did -16 Jan 23 12:30:12 5M didnt' work well Jan 23 12:30:17 so I tried 2M and it worked Jan 23 12:30:34 so yes kind of bisect Jan 23 12:48:20 SHR: 03shr-devel 07buildhistory * r4b27844d58be 10/images/om_gta02/eglibc/shr-image/ (10 files): Build 201201231234 of shr 20120123 for machine om-gta02 on shr-chroot Jan 23 12:49:15 GNUtoo: hmm, I guess I have to bisect too. sucks that there isn't a way for glamo to automatically use the right freq Jan 23 12:50:22 pabs3, note that glamo docs are officially released Jan 23 12:51:12 GNUtoo: yeah I helped get that done, by prodding Sean lots for some months :) Jan 23 12:51:45 and still nothing on the announce list, grrr Jan 23 12:52:24 yes but it's there: Jan 23 12:52:54 http://wiki.openmoko.org/wiki/Community_Updates/2011-12-01 Jan 23 12:53:01 thanks for that Jan 23 12:54:06 I personally doubt any code will result from this, I imagine Mali/etc will get an open driver before glamo Jan 23 13:13:13 pabs3, it could ease mainlining of glamo Jan 23 14:41:53 looks like git.openmoko.org & bugtraker are working again Jan 23 14:43:52 \o/ Jan 23 14:43:56 SHR: 03shr-devel 07buildhistory * r34e50b92ed9c 10/images/nokia900/eglibc/shr-image/ (10 files): Build 201201231348 of shr 20120123 for machine nokia900 on shr-chroot Jan 23 15:03:06 jluis_: if you know it, why are you trying to clone from old tarbals when newer are available? :) Jan 23 15:27:25 captainigloo: can you try efl toolchain from here http://jama.dyndns-home.com/org.openembedded.shr-core.sdk/ ? Jan 23 15:27:39 captainigloo: if it works for you I'll push it to shr branches Jan 23 15:31:03 JaMa: it's the sdk itself right ? Jan 23 15:31:22 i will test it tonight at home with my x86_64 machine Jan 23 15:31:32 i have only 32bits machine here Jan 23 15:32:05 ok, not sure if jama.dyndns-home.com will be online in the evening, but I can scp it to buildhost Jan 23 15:32:46 and I'll build 32bit version for test too Jan 23 15:36:45 JaMa: what did you modify to build this files ? Jan 23 15:37:01 i would like to build sdk here Jan 23 15:38:25 captainigloo: see jansa/test branch for those patches Jan 23 15:38:45 captainigloo: the rest was just rebuilding stuff with wrong RDEPS Jan 23 15:45:44 JaMa: did you get access to intone upstream? Jan 23 15:46:19 pabs3: yes Jan 23 15:46:38 excellent Jan 23 15:47:02 time to start using it sometimes :) Jan 23 15:47:33 not that access, I mean intone as app Jan 23 15:56:52 JaMa: what do you mean with rebuilding with wrong RDEPS ? Jan 23 15:58:56 grep Packages file for recipe names depending on not existing stuff and then -c cleansstate them Jan 23 15:59:34 from #yocto: Jan 23 15:59:35 15:52:31 < JaMa> is it just me, or are runtime dependency names badly changed when using debian.bblass? Jan 23 15:59:38 15:52:56 < JaMa> for cross-canadian (otherwise it works fine sofar) Jan 23 15:59:41 15:55:29 < JaMa> I had to rebuild gdb-cross-canadian-arm manually to fix runtime dependency from expat-native to libexpat1, but expat wasn't changed in a while afaik, so not sure how gdb-cross-canadian-arm got wrong runtime provider Jan 23 15:59:45 from shlibs Jan 23 15:59:47 that's the problem ^ Jan 23 16:00:28 ah ok Jan 23 16:00:39 so a build from scratch would works fine ? Jan 23 16:02:44 SHR: 03Martin.Jansa 07shr-chroot * re39f182cd59a 10/ (12 files in 7 dirs): bitbake upgrade Jan 23 16:02:54 probably no as there is something wrong with it.. I've built gdb-cross-canadian-arm for first time and it picked wrong expat name, so I had to rebuild it again Jan 23 16:03:52 but fwiw on other TMPDIR, it worked for me (with other issues but not related to debian.bbclass) Jan 23 16:06:15 it's not really clear to me, but i will try with your patches Jan 23 16:06:43 SHR: 03Martin.Jansa 07meta-smartphone * r35ef66a50791 10/meta-shr/recipes-support/picocom/picocom_1.6.bb: picocom: moved to meta-oe Jan 23 16:17:07 JaMa: what's shr-2012.01? Jan 23 16:20:05 dos1: name of upcomming release Jan 23 16:20:41 JaMa: ok, ping me when i should add it to download wizard Jan 23 16:20:59 JaMa: btw. yesterday i've added warning about --numeric-owner Jan 23 16:23:02 dos1: http://wiki.shr-project.org/trac/wiki/Release Jan 23 16:23:48 good about --numeric-owner, still many people didn't know about it Jan 23 16:26:25 hmm I've removed n900-oe repo from gitweb and still someone (google probably) is downloading it :/ Jan 23 16:30:36 JaMa: you know if there is an effort to port SHR on n950 ? Jan 23 16:31:29 I know only one person with n950 and he is not working on it afaik Jan 23 16:32:06 i have one Jan 23 16:32:30 so now I know 2 :) Jan 23 16:32:44 i tried an old SHR image i had build 1year ago Jan 23 16:33:01 i was not able to display anything Jan 23 16:33:16 n900 image? Jan 23 16:33:21 yes Jan 23 16:33:37 but it needs a specific xorg driver Jan 23 16:34:02 the xorg-omapfb build for n900 is not working Jan 23 16:34:13 and of course the kernel is specific Jan 23 16:34:17 SHR: 03shr-devel 07buildhistory * r8691e8275eb1 10/images/palmpre/eglibc/shr-image/ (10 files): Build 201201231544 of shr 20120123 for machine palmpre on shr-chroot Jan 23 16:34:33 but mer guys have a repo for it Jan 23 16:34:41 yup that's what I expected Jan 23 16:35:17 if i find the time, i will try, i know how to make recipes now :) Jan 23 16:35:29 even if they are not perfect ! Jan 23 16:36:16 great.. maybe some parts ie from kernel could be backported to n900 Jan 23 16:36:49 they are using a 2.6.32 kernel Jan 23 16:37:05 iirc n900 has a newer kernel ? Jan 23 16:37:51 ah :/ and yes 2.6.37 Jan 23 17:43:40 captainigloo: I have 32bit sdk ready but my upload is so slow that it estimates > hour to upload it :/ probably will be faster to build it on buildhost :) Jan 23 17:45:12 JaMa: ahaha :) Jan 23 17:45:19 i will try Jan 23 17:45:22 captainigloo: it will be here http://build.shr-project.org/tests/jama/sdk/ (tomorrow) Jan 23 17:46:07 * JaMa going to read book until 64bit version is uploaded and then shutdown jama.dyndns.. Jan 23 17:47:12 JaMa: ok thanks Jan 23 17:47:23 but i need to buildmy own sdk for my customer Jan 23 17:48:01 it seems that since i use angstrom-bleeding and specially gcc 4.6 i get a lot of problems Jan 23 17:48:22 shr is also using gcc Jan 23 17:48:25 4.6 Jan 23 17:49:21 yes i know, but for example i'm not able to launch my specific kernel build with gcc 4.6 anymore Jan 23 17:49:43 and the sdk is not building fine Jan 23 17:50:41 what's problem with kernel? some old version? Jan 23 17:52:45 i think it's a linker problem Jan 23 17:53:30 i get this kind of printk Jan 23 17:53:32 [ 0.000000] unwind: Unknown symbol address c000c504 Jan 23 17:53:32 [ 0.000000] unwind: Index not found c000c504 Jan 23 17:53:37 at stratup Jan 23 17:53:45 oldest we have is 2.6.24, but it works too with gcc-4.6 Jan 23 17:54:50 it's why i would like the sdk with gcc 4.6, to try the build outside oe Jan 23 17:55:06 ah Jan 23 17:55:12 and see if it's a .config problem, or a build problem Jan 23 17:55:26 or i don't know, maybe i missed something Jan 23 17:57:26 hi, who knows tsmd? Jan 23 18:55:33 SHR: 03shr-devel 07buildhistory * rcc4d626cc70d 10/ (62 files in 53 dirs): Build 201201231748 of shr 20120123 for machine palmpre2 on shr-chroot Jan 23 18:59:42 captainigloo: upload finished, bbl Jan 23 19:18:22 GNUtoo: hi Jan 23 19:20:23 Alex[sp3dev], hi Jan 23 19:22:35 GNUtoo: so, I've fixed my uSD regulator. i've untarred the crespo aurora image, edited xorg.conf, but touchscreen is not working, although evtest works fine Jan 23 19:22:52 Alex[sp3dev], did you add evdev instead of mtev? Jan 23 19:23:11 GNUtoo: no. why won't mdev work? Jan 23 19:23:15 Alex[sp3dev], btw for crespo we have a wifi problem Jan 23 19:23:22 Alex[sp3dev], because it's modified for palmpre Jan 23 19:23:24 oops Jan 23 19:23:29 s/palmpre/crespo/ Jan 23 19:23:46 the wifi problem is with bcm4329 Jan 23 19:23:53 there is brcm80211 Jan 23 19:23:56 GNUtoo: indeed, that worked. thanks Jan 23 19:24:18 more precisely brcmfmac Jan 23 19:24:28 what's wrong with brcmfmac? Jan 23 19:24:33 basically there is an interupt issue Jan 23 19:24:41 so I think I will take the 3.0 kernel Jan 23 19:24:49 and do the regilator thing Jan 23 19:25:04 but do you have any hints on how to do regulator stuff (runtime PM) Jan 23 19:26:30 GNUtoo: i've disabled most PM for now. just enabling power domain makes my kernel freeze. highmem won't work as well. as for the rest, there are not many changes from android 2.6.35. You should probably take linaro tree or linux 3.3 when it is ready and port stuff from ics 3.0 kernel. there are major fb changes Jan 23 19:27:11 yes I know about fb Jan 23 19:27:22 but I would be happy even with 3.0 kernel Jan 23 19:27:57 sgs2 seems like a tough thing. looks like i'll have to rewrite the alsa codec stuff Jan 23 19:28:08 so the only thing that I want to know is how to hookup in the board file the ifconfig wlan0 up to the sdio power/carddetect/etc... Jan 23 19:28:41 GNUtoo: i've not yet set up the card to the point wlan0 appears, but at least mmc is detected and ids for bcm4330 are displayed Jan 23 19:29:06 ok but you know how to do that from the kernel for your older msm/qsd device Jan 23 19:29:33 you added support for wifi there Jan 23 19:29:37 how did you do it? Jan 23 19:29:43 runtime PM? Jan 23 19:29:51 or the activator module thing? Jan 23 19:29:53 well, actually... on tegra, we can supply power gpio to sd via pdata, and rfkill just works. but on sgs2, for now, i'm just leaving it always on Jan 23 19:30:31 ok Jan 23 19:31:00 i think there's a gpio rfkill driver by nvidia in mainline now Jan 23 19:31:08 Alex[sp3dev], btw you may want to go in #replicant channel too for your modem Jan 23 19:31:25 they know how to make samsung modem works Jan 23 19:31:29 *work Jan 23 19:31:46 ask paulk or morphis Jan 23 19:32:02 GNUtoo: not now. I'm not going to devote more than an hour per day to this port, and today I have spent it fixing mmc. I think I need to implement charger and usb host/client switching first Jan 23 19:32:04 so basically it can work with android or GNU/Linux with fsogsmd Jan 23 19:32:17 ok Jan 23 19:32:25 btw, my usb client does not work when i enable either ohci or ehci, so probably i really need that usb route setup Jan 23 19:32:36 Alex[sp3dev], we(replicant) are insterested in samsung galaxy SII too Jan 23 19:32:51 just that no one bought it yet Jan 23 19:32:55 how is it? Jan 23 19:32:59 GNUtoo: actually, I want to mainline it, but that'll need some work Jan 23 19:33:04 I mean how hackable is it? Jan 23 19:33:28 any proprietary kernel modules like for storage? Jan 23 19:33:33 what about serial and bootloader? Jan 23 19:33:49 altough I am not going to buy it before finishing some other ports Jan 23 19:33:51 like gta04 Jan 23 19:34:23 it is pretty oss-friendly (i mean, all v4l2 stuff, camera sensors etc is in mainline. except, for, probably, accelerometers). serial is not available (or is not hacked yet). bootloader sucks - supplies the smdkpv310 mtype. we can chainload uboot, but that'd require some work Jan 23 19:35:22 ok Jan 23 19:35:25 nice for the mainlining Jan 23 19:35:43 serial is kind of a blocker Jan 23 19:35:44 as for storage, it uses emmc for internal storage. there's the jjjj (j4fs) partition with configs. we don't really need it, but sgs2 kernel has sources for j4fs Jan 23 19:35:58 ok Jan 23 19:36:12 GNUtoo: you can take a look at my board here https://gitorious.org/i9100-proper-linux/i9100-proper-linux-kernel/blobs/i9100_linux/arch/arm/mach-exynos/mach-sgs_i9100.c Jan 23 19:36:18 because for instance the galaxy tab has a proprietary kernel module Jan 23 19:36:31 ok thanks Jan 23 19:36:53 GNUtoo: it even has si4709 fm radio. vanilla i2c driver seems to probe correctly, although scan fails, but I haven't looked deeper Jan 23 19:37:29 fsa9480 Jan 23 19:37:38 yes, and why Jan 23 19:37:45 why is it not found out Jan 23 19:37:47 ? Jan 23 19:37:54 what do you mean Jan 23 19:37:55 no fsa9480 datasheets? Jan 23 19:38:07 I mean the serial should be behind the USB right? Jan 23 19:38:48 with the correct resistors values it could be doable Jan 23 19:38:59 well, probably. i didn't take a look myself, cause i didn't have a spare cable. however, there's some uboot log dumped so it should be doable Jan 23 19:39:02 but it's strange that xda has nothing Jan 23 19:39:21 ah the cable is non-standard? Jan 23 19:39:42 or is it a standard micro-usb cable? Jan 23 19:39:51 GNUtoo: standard micro-usb/mhl cable Jan 23 19:40:31 ok nice Jan 23 19:42:11 paulk-desktop, hi Jan 23 19:42:15 paulk-desktop, ^^^ Jan 23 19:52:23 hi Jan 23 19:54:01 I doubt galaxys2 is s5pc110 so serial setup might change Jan 23 19:54:24 maybe you should look at service manual Jan 23 19:56:39 paulk-desktop, ok Jan 23 19:57:07 but it could be an interesting target, at least for SHR, specially if someone works on it for mainline Jan 23 19:57:27 ok Jan 23 19:58:18 i can't stop getting amazed at how cool and easy git is for moving commits between branches and projects Jan 23 20:02:34 GNUtoo: i'm getting those "Too large timeout requested for CMD53!" on mmc2 now, after updating bcm80211 to the latest version which supports bcm4330 Jan 23 20:03:20 hmmm Jan 23 20:03:30 personally I've a problem with function 2 Jan 23 20:03:34 it needs an IRQ Jan 23 20:03:42 I think it was merged Jan 23 20:03:48 but didn't solve my problem Jan 23 20:13:32 hum, but wlan0 shows up. it's either interrupt problems or radio regulator is off Jan 23 20:13:46 but ifup causes lots of this warnings Jan 23 20:15:12 ok Jan 23 20:15:23 what happens on nexus s? Jan 23 20:16:23 Alex[sp3dev], because of function 2 it freeze the device Jan 23 20:16:30 but I used an activator Jan 23 20:16:59 because I think wifi/SDIO runtime PM was introduced in a kernel > 2.6.35 Jan 23 20:17:02 but we have 3.0 too Jan 23 20:17:10 we will eventually migrate to 3.0 Jan 23 20:19:12 actually i don't like that hackery with wifi power callbacks that got to mainline from android trees. on tegra and msm, wifi was just controlled using sdio vdd, but here it looks convoluted Jan 23 20:20:48 ok Jan 23 20:20:52 hmmm Jan 23 20:31:07 hi mickeyl Jan 23 21:02:17 SHR: 03shr-devel 07buildhistory * r7c48e8a0e27d 10/ (12 files in 3 dirs): Build 201201232005 of shr 20120123 for machine crespo on shr-chroot Jan 23 21:30:51 hi, is there anyone with palmpre here? Jan 23 21:38:00 hmmm Jan 23 21:38:08 it seem that there are some flags passed by oe Jan 23 21:38:13 the no thumb and -marm Jan 23 21:41:48 -mno-thumb-interwork -marm to be more exact Jan 23 22:07:43 GarthPS, now I'm fighting with oe Jan 23 22:08:00 for some reason ./run.do_compile produces a working kernel Jan 23 22:08:12 but uImage-palmpre.bin isn't ok Jan 23 22:13:25 GarthPS, at what point is your build? Jan 23 22:14:25 GNUtoo: Jan 23 22:15:21 well I did build a kernel before you push your patch, did not check if it boot but it should not. I can build the new one if you want Jan 23 22:15:50 GarthPS, the patch is now in the palm kernel but not in oe Jan 23 22:16:00 so just use last rev of the kernel Jan 23 22:17:42 GarthPS, let me find the rev for you Jan 23 22:17:51 GNUtoo: I edit the palm kernel recipe directly or wha t ? Jan 23 22:18:07 dc2ff991a816e51847cc5edd0bbbf19d1600ad49 Jan 23 22:18:15 use that rev in the palm pre kernel recipe Jan 23 22:18:20 and make clean Jan 23 22:18:28 in the kenrel git dir Jan 23 22:18:40 then ../temp/run.do_compile.* Jan 23 22:18:44 and it should work Jan 23 22:18:53 you should have an arm/boot/uImage Jan 23 22:18:59 that should boot Jan 23 22:19:04 if I did things right Jan 23 22:19:23 I'll try something Jan 23 22:20:06 since there is no serial it's hard to find the why Jan 23 22:21:51 arrrg it recompiles the toolchain Jan 23 22:29:28 GNUtoo: :) already done here the toolchain Jan 23 22:30:15 ok Jan 23 22:30:27 I don't know why...but it recompiles me the toolchain each time Jan 23 22:30:31 at each bitbake Jan 23 22:31:46 GarthPS, so try bitbake -c compile virtual/kernel Jan 23 22:33:09 GNUtoo: hmm bitbake is hanging here .. Jan 23 22:33:24 ok Jan 23 22:34:58 ah finally do_compile started Jan 23 22:40:26 hi JaMa Jan 23 22:40:56 JaMa, at each build(bitbake foo) it tries to rebuild the toolchain, still no fix for that ? Jan 23 22:41:57 it was never the case here Jan 23 22:42:05 ah ok Jan 23 22:42:09 I tought it was Jan 23 22:42:11 GNUtoo: ok time to bed! cu tomorrow ! Jan 23 22:42:14 and oe-core has some patches for some unexpected rebuilds now Jan 23 22:42:16 GarthPS, Linux palmpre 2.6.24-palm #1 Mon Jan 23 23:34:45 CET 2012 armv7l GNU/Linux Jan 23 22:42:20 gn all Jan 23 22:42:30 GNUtoo: the problem was only after switching machines Jan 23 22:42:35 ok Jan 23 22:42:37 GNUtoo: when gcc was changed Jan 23 22:43:08 JaMa, the difference between switching machines and not switching machine is that it still rebuild toolchain when I don't switch machines but at least it doesn't fail Jan 23 22:43:42 here's my shell: Jan 23 22:44:08 (it's pasting the data....) Jan 23 22:44:18 http://www.pastie.org/private/mpq0saowojqkurufja77q Jan 23 22:45:45 and are you sure it was built during first build (which failed?) Jan 23 22:46:08 it failed because do_fetch for xev failed Jan 23 22:46:12 else it succedded Jan 23 22:46:17 xev is not an image dep Jan 23 22:46:20 image were built fine Jan 23 22:47:38 well it doesn't do that here and on buildhost.. so if it's reproducible for you, you have to debug what happens with stamps or with sstate siginfo files Jan 23 22:47:49 btw is that ok: Jan 23 22:47:51 but first upgrade to latest oe-core/shr Jan 23 22:47:54 uboot-mkimage -A arm -O linux -T kernel -C none -a 20008000 -e 20008000 -n shr/2.6.24+gitr2+dc2ff991a816e51847cc5edd0bbbf19d1600ad49/palmpre -d linux.bin arch/arm/boot/uImage Jan 23 22:48:18 it's already an uImage Jan 23 22:48:26 why does it mkimage on it again? Jan 23 22:48:28 qi can boot vmlinuz directly too btw Jan 23 22:48:34 wow nice Jan 23 22:48:51 always has :) Jan 23 22:49:11 ok Jan 23 22:49:20 the documentation just isn't that great :/ Jan 23 22:49:26 lol ok Jan 23 22:50:12 JaMa, should I ask in oe for the mkImage? Jan 23 22:50:35 ah wait a bit Jan 23 22:50:46 the arch/arm/boot/uImage is the destination??? Jan 23 22:52:05 what you mean by "ask in oe for the mkImage" ? Jan 23 22:52:20 I am wondering if the command was right Jan 23 22:52:33 I mean do_compile produces an uImage Jan 23 22:52:43 what's the purpose of uboot-mkimage? Jan 23 22:53:53 if uImage is used everytime for palmpre, then just add KERNEL_IMAGETYPE = "uImage" to palmpre.inc Jan 23 22:53:57 + ENTRYPOINT=20008000 Jan 23 22:53:59 that seem wrong Jan 23 22:54:11 and specify UBOOT_ENTRYPOINT + UBOOT_LOADADDRESS Jan 23 22:54:47 Load Address: 80008000 Jan 23 22:54:48 Entry Point: 80008000 Jan 23 22:54:53 that are the right values I guess Jan 23 22:54:57 not 20008000 Jan 23 22:55:11 thanks a lot!!! Jan 23 22:55:19 sorry don't know about palmpre, and yw :) Jan 23 22:55:34 * JaMa off to bed, gnight Jan 23 22:55:38 ok Jan 23 22:55:44 * GNUtoo finishes the fix and go to bed Jan 23 23:13:12 SHR: 03GNUtoo 07meta-smartphone * r57c7360f5a41 10/meta-palm/ (2 files in 2 dirs): meta-palm: linux-palmpre: fix booting and bump SRCREV Jan 23 23:13:15 SHR: 03shr-devel 07buildhistory * rc604a35b3bcc 10/ (12 files in 3 dirs): Build 201201232214 of shr 20120123 for machine om-gta04 on shr-chroot Jan 23 23:39:49 SHR: 03GNUtoo 07meta-smartphone * rced0c400c954 10/meta-palm/recipes-graphics/tslib/tslib_git.bbappend: meta-palm: make tslib bbappend MACHINE_ARCH for palmpre Jan 24 01:12:26 SHR: 03shr-devel 07buildhistory * raf52e3798db7 10/ (184 files in 174 dirs): Build 201201240022 of shr 20120123 for machine om-gta02 on shr-chroot **** ENDING LOGGING AT Tue Jan 24 02:59:57 2012