**** BEGIN LOGGING AT Sun Feb 05 10:59:59 2006 Feb 05 10:59:59 oh? Feb 05 11:00:10 it will be weeks before i get my serial cable Feb 05 11:00:23 You could try debugging with the LED Feb 05 11:00:33 happy to test images :) Feb 05 11:01:52 also i have done some digging Feb 05 11:02:12 shadows: I'll give you an example patch which might give you some clues as to what to look for Feb 05 11:02:20 okay hmm Feb 05 11:03:10 is it likely that there's not enough power being provided to the RAM where suspend info is? Feb 05 11:03:48 i don't even know Feb 05 11:04:06 shadows: No - its able to suspend and resume for the charging light Feb 05 11:04:17 yes up to a ponit Feb 05 11:04:19 up to a point Feb 05 11:04:23 after that point, nothing works Feb 05 11:04:30 the light is off, no lights at all Feb 05 11:04:41 but after you hit the power button? Feb 05 11:04:48 or does the charging code also hang? Feb 05 11:04:49 hmm Feb 05 11:05:02 the charging code is hung, AFAICT Feb 05 11:05:16 but after you press the power button? Feb 05 11:05:18 the device does not charge after being suspended for 2 minutes approximately Feb 05 11:05:59 what i do is i suspend the device, and then i can resume it if the device has been suspended for less than 2 minutes Feb 05 11:06:35 if i wait too long, the devices is suspended for more than about 2.5 minutes, the device has no lights on and does not respond to anything, not even plugging in the charge cord Feb 05 11:07:12 Can you leave the device for more than 2.5 minutes and then just unplug and replug the charger in? Feb 05 11:07:19 hi. does anybody know why irda-utils postinst doesn't work here? this is my do_rootfs log: http://en.pastebin.ca/40108. Feb 05 11:07:45 RP: that does not appear to do anything Feb 05 11:08:19 pH5: because it has to run on the target, not the build host? Feb 05 11:08:23 shadows: Plugging the charger in/out should cause the charging light to change state Feb 05 11:08:46 after approx 2.5 minutes in suspended state, the charging light remains off Feb 05 11:09:18 and now gnome-vfs is not compiling for me Feb 05 11:09:20 damn it! Feb 05 11:09:23 shadows: Ok, that's a lot more serious :-( Feb 05 11:10:31 RP: provide for me a rootfs and image to test, i will debug based on that. i'm not confident in my amd64 build environment Feb 05 11:10:40 pH5: what's the contents of the postinst in question? Feb 05 11:11:16 shadows: You're not using /dev/hda1 are you? Feb 05 11:11:25 i will use whatever you want, to debug Feb 05 11:11:41 i just today trashed my whole hdd partitioning, and re-did it to factory Feb 05 11:11:50 plus NAND restore just to be sure Feb 05 11:11:59 shadows: ok. I don't have any images handy so I'll have to build something Feb 05 11:12:23 the SharpROM suspended/resumed without trouble. so i am at least relieved my hardware may be safe Feb 05 11:12:34 Your hardware will be fine Feb 05 11:12:40 * shadows :) Feb 05 11:13:01 good. 'cause i was in this car crash today, and my expensive vehicle is messed up :/ Feb 05 11:13:23 something good happened today, at least my Z is not hopeless Feb 05 11:13:51 RP = my favorite person for Sunday Feb 05 11:15:02 Sorry to hear about the car crash :-( Feb 05 11:15:32 snow conditions, and the car is front wheel drive Feb 05 11:15:36 bad combination Feb 05 11:17:43 any particular partition scheme i should set up? Feb 05 11:17:57 shadows: Standard sharp one is fine Feb 05 11:18:03 excellent Feb 05 11:18:12 i wish i knew what the standard sharp scheme was exactly Feb 05 11:18:25 it's hard to find someone with an unmodified partition table for the C3000 Feb 05 11:18:57 though i doubt it matters, i am curious Feb 05 11:19:22 shadows: I think I have an image you can test - its from the openedhand hacked version of OE rather than OE itself but I'm guessing this doesn't bother you? Feb 05 11:19:42 nope, that's fine Feb 05 11:20:05 03koen 07org.oe.oz354fam083 * r88b0ab3f... 10/packages/matchbox-panel/ (2 files in 2 dirs): matchbox-panel: apply patch to stop wastefull battery polling, closes hh.org bug #1474 Feb 05 11:20:09 03koen 07org.oe.oz354fam083 * r11627509... 10/packages/libgtkstylus/libgtkstylus_0.3.bb: libgtkstylus: stop shipping development files, taken from .dev Feb 05 11:20:13 03koen 07org.oe.oz354fam083 * r894b997c... 10/packages/libgtkinput/ (files/gtkinput.sh libgtkinput_0.3.bb): libgtkinput: don't ship development files Feb 05 11:20:18 03jbowler 07org.oe.dev * r22dc7127... 10/packages/beep/beep_1.2.2.bb: beep: fix install to bindir in 1.2.2 Feb 05 11:20:23 03nail 07org.oe.dev * r911d3956... 10/packages/wget/wget_1.9.1.bb: Feb 05 11:20:23 wget: disable ipv6 for SlugOS Feb 05 11:20:23 slugos doesn't support IPV6 by default, and wget fails on IPV4 if IPV6 isn't Feb 05 11:20:23 supported in the kernel. Feb 05 11:20:27 03jbowler 07org.oe.dev * rd17c8965... 10/packages/ (5 files in 4 dirs): Feb 05 11:20:27 ixp4xx-kernel: update to new LEDs code in 2.6.15.2 Feb 05 11:20:28 - the meanings of various fields in /sys/class/leds/* have changed, Feb 05 11:20:30 slugos-init is updated to take account of these in 0.10 Feb 05 11:20:32 03jbowler 07org.oe.dev * r8703efdf... 10/packages/ (6 files in 5 dirs): Feb 05 11:20:34 explicit_merge of '911d39563f4bbb32b27acb504168de96ff1e68f7' Feb 05 11:20:36 and 'd17c8965045bca1d838e85c0b4579217b647682c' Feb 05 11:20:38 using ancestor '' Feb 05 11:20:40 to branch 'org.openembedded.dev' Feb 05 11:20:44 03jbowler 07org.oe.dev * r7caf52c1... 10/packages/ipkg/ (ipkg-0.99.155/upgrade-message-garbage.patch ipkg_0.99.155.bb): ipkg: fix spurious characters in ipkg upgrade messages in 0.99.155 Feb 05 11:20:47 03nail 07org.oe.dev * r253b3416... 10/packages/zd1211/ (zd1211-r52/firmware.patch zd1211_r52.bb): zd1211: Incorporate new version of firmware from original Zydas driver Feb 05 11:20:50 03florian 07org.oe.dev * r38239426... 10/packages/gpe-shield/gpe-shield_0.9.bb: gpe-shield: add 0.9:wq Feb 05 11:20:54 03nail 07org.oe.dev * rc6c0ed4e... 10/packages/meta/slugos-packages.bb: slugos-packages: Re-add groff to the feeds. It now works properly. Closes slugbug #264 Feb 05 11:20:57 03nail 07org.oe.dev * r61a885ef... 10/packages/beep/beep_1.2.2.bb: beep: Change {bindir} to {BINDIR} so the binary is installed in the right place Feb 05 11:21:00 03nail 07org.oe.dev * rc6e95012... 10/packages/linux/ (ixp4xx-kernel/2.6.15/defconfig ixp4xx-kernel_2.6.15.2.bb): ixp4xx-kernel: Add CONFIG_USB_LD=m, for USB thermometer Feb 05 11:21:05 03nail 07org.oe.dev * r5c10c326... 10/packages/meta/slugos-packages.bb: slugos-packages: Remove groff from ucslugc feed Feb 05 11:21:09 03justinp 07org.oe.dev * r4d5aa9c8... 10/packages/parted/parted_1.6.25.1.bb: parted: adding version 1.6.25.1 Feb 05 11:21:13 03justinp 07org.oe.dev * rbe42e683... 10/packages/meta/task-gpe.bb: task-gpe: remove redundant gpe-task-games Feb 05 11:21:17 03justinp 07org.oe.dev * r9f78e2cf... 10/packages/meta/e-image.bb: e-image: make it useful, remove non-working e packages and adding GPE tasks Feb 05 11:21:22 03justinp 07org.oe.dev * rbc5629f2... 10/packages/meta/task-e-x11.bb: task-e-x11: remove task-e-x11 (not -core) as it's non-working Feb 05 11:21:26 03justinp 07org.oe.dev * r0cdbcc7e... 10/packages/base-files/base-files/ (borzoi/fstab spitz/fstab): fstab (spitz, borzoi): hda3 is vfat, not ext3 Feb 05 11:21:30 03koen 07org.oe.dev * r96f72a9b... 10/packages/matchbox-panel/ (2 files in 2 dirs): matchbox-panel: apply patch to stop wastefull battery polling, closes hh.org bug #1474 Feb 05 11:21:35 03koen 07org.oe.dev * re07763d2... 10/packages/libgtkinput/ (files/gtkinput.sh libgtkinput_0.3.bb libgtkinput_cvs.bb): libgtkinput: add libgtkinput, a gtk module which enables maemostyle keyboard popup when using an inputmanager like the patch mb-applet-inputmgr in OE Feb 05 11:22:25 shadows: http://www.rpsys.net/openzaurus/temp/shadows/ Feb 05 11:22:38 shadows: I've not tested this yet either but will flash now ;-) Feb 05 11:23:18 * shadows snarfs Feb 05 11:23:25 shadows: I also have standard sharp partitioning btw Feb 05 11:24:00 okay Feb 05 11:24:29 the instructions at xlfag.com/zaurus are incorrect, i am pretty sure that 'sfdisk' built into the c3000 recovery console cannot produce the correct format Feb 05 11:27:46 03koen 07org.oe.oz354fam083 * rd253b0fa... 10/ (2 files in 2 dirs): add and activate gpe-shield 0.9 Feb 05 11:29:00 shadows: FWIW, it never touched hda1 or 2 on my Z, only made hda3 smaller and added more Feb 05 11:29:21 ooOH Feb 05 11:29:41 sorry, should haven mentioned that before Feb 05 11:29:58 it=I Feb 05 11:30:01 the trouble with info i have so far is that it only goes down to the cylinder size Feb 05 11:30:39 and the info does not indicate whether ms-dos compatibility is included as part of the first partition Feb 05 11:30:48 hmmm Feb 05 11:30:51 i.e. does that first partition start at 1kb, 16kb, or 32kb Feb 05 11:31:30 shadows: If its booting from the drive, I wouldn't worry about it Feb 05 11:31:40 yeah Feb 05 11:31:47 * JustinP boots back to 2.4/old old e image to test wifi Feb 05 11:32:16 i'm sort of exacting though, i want this info documented for historical reference Feb 05 11:32:22 in a language other than japanese ;) Feb 05 11:32:44 well if RP hasn't touched his.... Feb 05 11:32:53 hmm Feb 05 11:33:14 tools at http://www.cuodan.net/~jnc/files/buggy_recovery_supplements-cxx00_zaurii.tbz2 Feb 05 11:33:15 Can I dump the data from mine? (preferably within an image, not the recovery console) Feb 05 11:33:28 recovery console Feb 05 11:33:32 pb_: the postinst from irda-utils 0.9.16: http://en.pastebin.ca/40116. shouldn't it be run during first boot instead of do_rootfs as it writes to /etc/sysconfig/irda? Feb 05 11:33:58 would like output of running ./fdisk-2.12r /dev/hda and choosing 'x' for expert, then 'p' print Feb 05 11:36:18 and possibly also the output from ./sfdisk-2.12r -d /dev/hda Feb 05 11:36:57 RP: okay I have seen my first git 'race' Feb 05 11:37:56 zecke: a serious one? Feb 05 11:39:24 RP: that depends, git wanted to fetch a object and it vanished Feb 05 11:39:45 RP: and then git failed ;) This means automatic scripts could fail due the packing Feb 05 11:40:13 zecke: Surely you'd just need to rerun the fetch? Feb 05 11:40:36 The bug would be whether it returned an error code to say "rerun me" :) Feb 05 11:40:52 shadows: That image works for me Feb 05 11:40:57 RP: right, For humans this shouldn't be too problematic Feb 05 11:41:25 RP: but I'm thinking of a script/tinderbox that would automatically rebuild OE every once in a while Feb 05 11:41:39 RP: okay, flashing+unrolling now Feb 05 11:41:59 zecke: I can see the concern Feb 05 11:42:34 shit Feb 05 11:42:35 RP: I also see the problem (file locking over http) Feb 05 11:42:44 there goes my third wifi card Feb 05 11:43:01 zecke: Yes, and via rsync. Its why they created git daemon Feb 05 11:43:17 first got cracked, second disappeared, third is not responding to my Z... Feb 05 11:43:20 ::sigh:: Feb 05 11:44:33 JustinP: you have tried the D-Link DCF-660w ? Feb 05 11:45:09 shadows: no Feb 05 11:45:17 JustinP: highly reccommended Feb 05 11:45:21 Hmm. Firmware downloading looks broken on that image, perhaps due to the new udev it uses :-/ Feb 05 11:45:24 pH5: looks like it. alternatively, and perhaps better, move that section of code to /etc/init.d/irda-utils. Feb 05 11:45:25 it just needs a firmware update, and it's good to go Feb 05 11:45:29 I had a Linksys WL, then 2 Ambicoms..... Feb 05 11:45:40 ah Feb 05 11:45:49 they all worked pretty well Feb 05 11:45:53 I liked the ambicoms... Feb 05 11:45:58 don't know why it died on me Feb 05 11:46:13 cost me way too much, though Feb 05 11:46:22 :/ Feb 05 11:46:31 pb_: thanks, I'll try that Feb 05 11:47:18 pH5: that script also needs a patch for the hx according to mreimer Feb 05 11:50:19 shadows: http://pastebin.com/539859 Feb 05 11:50:20 and xerver-kdrive is breaking...and gnome-vfs-dbus....::sigh:: Feb 05 11:50:39 * RP -> back shortly Feb 05 11:51:24 RP: that shows cylinders though :/ Feb 05 11:51:34 cylinders leave a lot of wiggle room Feb 05 11:52:00 i'm sure it doesn't matter in practice Feb 05 11:53:21 * zecke just tried to compile Qtopia 2.2... Feb 05 11:53:42 zecke: install linux 2.4 and gcc 2.95.3 ;) Feb 05 11:53:43 gosh it wasn't enough that they fuck up Qtopia's buildsystem, now they even brainfucked Qt itself Feb 05 11:54:11 can it be added to OE? Feb 05 11:54:20 without losing your sanity? Feb 05 11:54:40 koen: well I'm better of rsyncing Qt and repacking it Feb 05 11:55:02 koen: the Australian Qtopia juice seems to be only in the Qt release :} Feb 05 11:55:14 heh Feb 05 11:56:27 * shadows boots RP image on c3000 Feb 05 11:59:26 lol not even ./configure --help is working Feb 05 12:01:41 Hi all. Feb 05 12:02:06 hey uv1 Feb 05 12:02:33 I have problems to pull (jan2006prebreakage). Merge suggested. Do we have a snapshot to download somewhere ? Feb 05 12:02:58 ~ewi Feb 05 12:02:59 somebody said ewi was ewi546.ewi.utwente.nl - main backup server for OE monotone. Hosts database snapshot: http://ewi546.ewi.utwente.nl/OE/OE.db.bz2 Feb 05 12:03:30 koen: Is that the prebreakage snapshot ? Feb 05 12:06:10 uv1: no, but you can checkout the tag Feb 05 12:07:58 koen: will try ... Feb 05 12:15:10 RP: I have my own linux2.6 git tree now :) Feb 05 12:15:43 zecke: nice :) Feb 05 12:25:24 03philippe 07org.oe.dev * r1a9bda89... 10/packages/xmonobut/xmonobut_0.4.1.bb: Feb 05 12:25:24 packages/xmonobut/xmonobut_0.4.1.bb Feb 05 12:25:24 * new file for more recent xmonobut Feb 05 12:25:24 * fixes handhelds.org bug #1459 Feb 05 12:26:28 Philippe: very good Feb 05 12:26:57 pb_, thanks. You're welcome :-) Feb 05 12:34:57 pb_: do you know if someone was stupid enough to implement xscale and iwmmxt instructions similiar to what one does with the floating point instructions? Feb 05 12:35:29 zecke: not that I know of Feb 05 12:35:45 too bad :} Feb 05 12:35:58 yeah Feb 05 12:36:19 the old xscale1 instructions (pre-iwmmxt) didn't get used for anything much in any case Feb 05 12:37:02 I get an illegal instruction (on my shark) and I can not recompile the software :} Feb 05 12:39:25 03koen 07org.oe.oz354fam083 * r67ba1645... 10/ (2 files in 2 dirs): add updated xmonobut from .dev, fixes hh.ort bug #1459 Feb 05 12:39:29 03koen 07org.oe.oz354fam083 * r319feb50... 10/ (2 files in 2 dirs): gpe-gallery: add 0.97, needed to fix hh.org bug #1462 Feb 05 12:39:33 03koen 07org.oe.dev * r0e478da2... 10/packages/gpe-gallery/gpe-gallery_0.97.bb: gpe-gallery: add 0.97 Feb 05 12:40:36 hi guys Feb 05 12:45:53 pH5: I've found a bug in udev 084 - the path to firmware_helper changed and udev.rules needs to be update to match Feb 05 12:47:16 an embedded system should be compiled with -fPIC to let shared libraries? Feb 05 12:47:44 always Feb 05 12:50:23 RP: hm.. debian has a /lib/udev/firmware.agent script instead of firmware_helper. perhaps we should follow suit? Feb 05 12:51:10 BleSS: any shared library code should (almost always) be compiled -fPIC, irrespective of embedded or not Feb 05 12:51:51 pH5: The udev package installs to firmware_helper so I think we just keep things simple and use that untill we have reason to change Feb 05 12:53:42 pb_, yes ok, i made wrong the question, i meant how embedded system is affected by shared libraries? is it more slow? Feb 05 12:53:57 compared to what? Feb 05 12:54:04 I don't really understand the question Feb 05 12:54:27 how affect shared libraries into an embedded system? Feb 05 12:54:44 just the same as a non-embedded system, really Feb 05 12:55:09 shared libraries and "embeddedness" are pretty much orthogonal issues Feb 05 12:55:39 ok Feb 05 12:56:48 and is valid building shared libraries for whatever architecture? Feb 05 12:57:05 sure Feb 05 12:58:57 mhhh which .bb file will create or build /linuxrc .... i got "<4>Failed to execute /linuxrc. Attempting defaults..." Feb 05 13:00:17 I used to have a funky-h2200-linuxrc.bb :) Feb 05 13:00:42 thanks to pb_ it had a cute nane Feb 05 13:01:10 name* Feb 05 13:01:27 mhhh koen what that mean that usually it don't exixts ? Feb 05 13:01:28 RP: ok thanks, fixed the path. I just noticed that firmware_helper is a 10k binary doing essentially the same as a 500 byte script. Feb 05 13:01:42 gremlin[it]: it usually doesn't exist Feb 05 13:03:47 mhhh so what can be ? Feb 05 13:03:53 "<4>Failed to execute /linuxrc. Attempting defaults..." Feb 05 13:04:04 /etc/init ? /sbin/init ? Feb 05 13:04:22 your kernel cmdline or your initrd try to execute /linuxrc Feb 05 13:05:50 yes is in the linuxargs of bootloader ... but before it worked so maybe "defaults" was ok ... Feb 05 13:08:28 now i try to specify /sbin/init.sysvinit on kernel args and try to find what is .. maybe i have a problem with libraries ... Feb 05 13:09:21 i don't have any libgcc on my rootfs :( :( :( :( Feb 05 13:15:15 that is extremely unfortunate Feb 05 13:18:07 yes ... kernel give me : error while loading shared libraries: libgcc_s.so.1: cannot open sh Feb 05 13:18:08 ared object file: No such file or directory Feb 05 13:19:06 mhhh any starting point about libgcc ? Feb 05 13:19:17 where i have to look ? Feb 05 13:19:37 gremlin[it]: http://ewi546.ewi.utwente.nl/tmp/viewmtn/getdiff.py?id1=bda4787c4d39c0c680a67a9406b6259986ebc180&id2=a2e3bdde97045586a369322e42c6a4e342cc87b0 Feb 05 13:19:46 and http://ewi546.ewi.utwente.nl/tmp/viewmtn/revision.psp?id=a2e3bdde97045586a369322e42c6a4e342cc87b0 Feb 05 13:20:54 thanks !!! Feb 05 13:21:23 the correct solution is to fix OE, but that patch works as a workaround Feb 05 13:25:44 i hope that is the last issue ... so 2.6.15 will boot on h3600 :) Feb 05 13:26:48 03pH5 07org.oe.dev * rc70fd3e0... 10/packages/udev/udev-084/udev.rules: udev_084: correct path to firmware_helper in udev.rules Feb 05 13:28:51 gremlin[it]: you shouldn't need to build a whole new rootfs just to boot 2.6.15 Feb 05 13:29:44 time to time i prefear try to build an image from begin :) ... Feb 05 13:30:47 I guess if that's what you want to do, fair enough. Feb 05 13:31:16 mhhh ERROR: Nothing provides runtime dependency libgcc1 Feb 05 13:31:54 that would mean gcc-cross wasn't built Feb 05 13:32:49 stick http://familiar.handhelds.org/releases/v0.8.3/feed/base/libgcc1_3.4.4-r3_arm.ipk in deploy/ipk Feb 05 13:33:07 I think it's more likely to mean that the inferred depends process is malfunctioning. Feb 05 13:33:30 tmp/stamps/x86_64-linux/gcc-cross-3.4.4-r3.do_* exist Feb 05 13:33:46 if gcc-cross genuinely wasn't built, you would have had a hard time compiling anything else Feb 05 13:34:03 tmp/deploy/ipk/libgcc1_3.4.4-r3_arm.ipk exist Feb 05 13:34:44 pb_ yes ... seem strange to be too :) :) :) Feb 05 13:48:49 mhhh koen the org.openembedded.oz354fam083 repo/branch is only for openzaurus or for all pda ? Feb 05 13:49:13 it will be used to build openzaurus 3.5.4 and familiar 0.8.3 Feb 05 13:49:46 so maybe is better if i sue thta one instead of .dev ..... Feb 05 13:53:31 shadows: How are you finding that image? Feb 05 13:53:58 (I updated the one I shared with a fixed udev and another issue I was aware of) Feb 05 13:55:15 gremlin[it]: 2.6 support is better in .dev (more modern udev etc) Feb 05 13:56:17 ok RP ... Feb 05 13:56:35 anyway ... after adding libgcc1 to familiar.conf ... all time i have Feb 05 13:56:48 ERROR: Nothing provides runtime dependency libgcc1 Feb 05 13:56:49 NOTE: no buildable providers for task-bootstrap Feb 05 14:22:13 grrrrr i have the libgcc1 ipkg package ... but how i can put in the rootfs ??? Feb 05 14:26:45 jbowler: The floating point frequency code is getting a bit ugly - I think I'm going to leave that out of mainline for now. Feb 05 14:31:26 gremlin[it]: ipkg should install it during do_rootfs. in my gpe-image build, ipkg first installs dropbear, then libz1 and then libgcc1 as dependencies. Feb 05 14:32:46 gremlin[it]: Its probably an idea to look at the do_rootfs log and see whether it considers installing it (and also at the depenencies of a few packages to see if libgcc is listed) Feb 05 14:32:49 mhhh seem .bb files for 3.4.4 is different ... there isn't PACKAGES = "libgcc ..." Feb 05 14:33:09 that can be a reason cause seem no one provide libgcc1 ... Feb 05 14:33:52 gremlin[it]: I guess you should just build task-bootstrap directly with "bitbake -b". Feb 05 14:35:26 RP: it's known OE bug that libgcc1 doesn't end up in depends Feb 05 14:35:44 koen: under what conditions? it seems to work here Feb 05 14:35:50 the giveaway is the 'can't find shared library provider for libgcc.so' Feb 05 14:35:54 pH5: random Feb 05 14:35:56 koen: That also works here Feb 05 14:36:00 uh :-) Feb 05 14:36:46 Random bugs are the painful ones :-( Feb 05 14:38:20 especially this one Feb 05 14:38:25 although it's fairly easy to spot Feb 05 14:38:26 RP building with -b ... don't put libgcc on rootfs :( Feb 05 14:38:38 perhaps somebody for whom it doesn't work could post his gcc-cross-3.4.4 build logs? Feb 05 14:39:27 pH5 i have to go ... but later i clean a bit and rebuild *gcc* *bootstrap* ... and let see :) Feb 05 14:39:43 gremlin[it]: great Feb 05 14:40:13 or just unbuild one ipkg in deploy/ipk, add a Depends: libgcc1 and bitbake bootstrap-image Feb 05 14:40:20 see u later Feb 05 15:18:28 /allchan say hi Feb 05 15:18:30 ... Feb 05 15:18:40 hi Feb 05 15:28:03 hi CoreDump|home :) Feb 05 15:28:58 * CoreDump|home waves at the local kernel god Feb 05 15:30:58 what the hell if suddenly wrong with my antivir. It prefixes all damn mails with an add of antivir Feb 05 15:32:08 03nail 07org.oe.dev * r1425776b... 10/packages/meta/slugos-packages.bb: slugos-packages: Remove groff from ucslugc feed properly Feb 05 15:41:07 that's just great. AntiVir no longer offers a freee-mail scanner Feb 05 15:45:22 great, the mail scanner costs 359.60? Feb 05 16:20:03 i'm back Feb 05 16:20:33 phew Feb 05 16:20:53 ping ph5 Feb 05 16:21:17 pong gremlin[it] Feb 05 16:22:15 ok... i remove tmp/stamps/*/*gcc* and rebuild with bitbake -v ... ok ? Feb 05 16:23:06 ok Feb 05 16:25:29 start Feb 05 16:36:23 done Feb 05 16:36:34 what u want to see pH5 ? Feb 05 16:38:36 mhh something happen i hale libgcc now !!!!!!!!!!!!!! Feb 05 16:39:06 damn, this is how I usually help people. Feb 05 16:39:31 ESP-debugging !!! Feb 05 16:45:13 gremlin[it]: did you use koen's workaround or put a libgcc in your ipkg dir manually? but if your install/gcc-cross.package contains PKG_libgcc: libgcc1 i don't see why it wouldn't have built it. Feb 05 16:45:42 pH5: the inferrend DEPENDS stuff gets in the way I guess Feb 05 16:46:13 pH5 no i don't use koen workaround Feb 05 16:46:42 koen: but wouldn't that show up as an error in the gpe-image do_rootfs log? Feb 05 16:47:21 pH5: or in the build, RDEPEND -> DEPENDS -> not found Feb 05 16:56:19 03pH5 07org.oe.dev * rf7d38069... 10/packages/libsexy/libsexy_0.1.5.bb: libsexy: add 0.1.5 Feb 05 17:00:38 03pH5 07org.oe.dev * rf2fb8c8a... 10/packages/libnotify/libnotify_0.3.2.bb: libnotify: add 0.3.2, a library to send desktop notifications to a notification daemon Feb 05 17:05:05 RP: i did a battery drain test, executing "cat /dev/urandom > /dev/null" and leaving the unit to run Feb 05 17:05:34 it suspended itself fine, and is blinking the LED green occasionally. i plug in the AC adapter, and the charge light comes on Feb 05 17:05:56 which is a lot different than before Feb 05 17:08:06 after letting it charge a minute, it is resuming fine Feb 05 17:08:07 :/ Feb 05 17:17:46 >>>>>>>>>> Linux h3600 2.6.15-hh0 #1 PREEMPT Sun Feb 5 17:59:35 CET 2006 armv4l unknown <<<<<<<<<< Feb 05 17:17:54 shadows: The green light is IDE activity Feb 05 17:18:18 gremlin[it]: !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! Feb 05 17:18:20 RP: it's *working*. i'm so confused Feb 05 17:18:30 gremlin[it]: ooooOooh. nice Feb 05 17:19:49 DHO!!! is an 'old' kernel without my recent patch ... uff uff ... Feb 05 17:20:44 RP: could my hdd partitioning somehow affect suspend/resume? Feb 05 17:20:47 this is so weird Feb 05 17:21:02 shadows: no, it wouldn't Feb 05 17:21:17 shadows: Its much more likely to be your AMD64 Feb 05 17:21:46 i tried out images and rootfs from someone who had an ia32 buildhost Feb 05 17:21:57 not your images, but... Feb 05 17:22:12 it would be nice to know why this failed previously Feb 05 17:22:19 i have no idea at the moment Feb 05 17:22:44 ~lart OZ for mounting SD cards sync Feb 05 17:22:44 * ibot pulls out his louisville slugger and uses OZ's head to break the homerun record for mounting SD cards sync Feb 05 17:26:35 against all logic and reasoning, i think it was the partition table Feb 05 17:27:40 to determine if the cause is an amd64 build host, we would need to both build the same code and compare the object output Feb 05 17:30:27 03pH5 07org.oe.dev * rc9c861f9... 10/packages/notification-daemon/notification-daemon_0.3.4.bb: notification-daemon: add 0.3.4, a service that listens to desktop notification requests and displays them Feb 05 17:31:39 RP: 'period' is more natural in this context than 'frequency', but with frequency it is certainly necessary to specific frequencies <1s - you could use mHz of course. Feb 05 17:36:23 jbowler: Is suspect I can't keep everyone happy with the timer. If I change to mHz, people will complain its not Hz which is the obvious unit for frequencies... Feb 05 17:38:37 Well, that's fine - you just need to handle values <1 Feb 05 17:39:04 It works for me any way - since I program a shell script round it, just so long as I can make the LED flash slow enough. Feb 05 17:39:32 RP: my results from testing your image on c3000 => http://oz.pastebin.com/540319 Feb 05 17:39:55 I got the cpu activity working consistently using a enum setting of 'on' 'flash' or 'off' for the 'active' and 'idle' states, but I've no SMP system to test on. Feb 05 17:40:24 (So I don't have the faintest idea whether the locking works, since it doesn't do anything on !SMP) Feb 05 17:41:44 jbowler: ok. For now I'm just going to concentrate on pushing the core - I've had another go at LKML with pressure to get it into -mm this time Feb 05 17:42:42 jbowler: If it makes it, we can then look at merging a cpu activity trigger and tweaking the timer if appropriate Feb 05 17:43:13 jbowler: I'm looking at using the rtc class in the zaurus tree - busy porting the sa1100 driver to it... Feb 05 17:45:26 You could provide both period and frequency - lots of things in sysfs present the same data in different ways. Feb 05 17:46:56 jbowler: I could. That would still need all the decimal place handling for frequency though so we just gain more code that way Feb 05 17:48:14 Well... not necessarily - they could both be integer, then period >1000ms would come out as frequency 1Hz Feb 05 17:49:04 Which would end up getting filed as a bug... Feb 05 17:50:38 The whole idea was to keep the led class simple, at least in mainline. If you need complicated, you can do it but ts perhaps best in your own trigger... Feb 05 17:51:27 In retrospect, a trigger with two delay parameters might have been nicer... Feb 05 17:57:42 i wonder why valid values are 1-100 and not 0-255 Feb 05 17:57:51 * shadows pokes kernel list archive Feb 05 17:59:24 RP: yes, when I first used it that's what I felt would have been easiest - 'mark' and 'space' in ms It sounds like there's some confusion over whether sysfs is an API or a UI... Feb 05 18:00:36 period in ms with the name 'frequency' doesn't give me any problem so long as there was adequate documentation of the interface... Feb 05 18:01:16 i thought "interval" is the right label for something in unit ms Feb 05 18:02:05 shadows: if it repeats then 'period' if it is one-shot then that would be wrong, I'd say interval or delay depend on the sematic. Feb 05 18:02:37 it is still correct for a one-shot Feb 05 18:02:41 I'd probably call it delay_on and delay_off :) Feb 05 18:02:51 IMO you just don't see the next interval Feb 05 18:03:17 it's imaginary :) Feb 05 18:04:05 "period" has the trouble where it is an english word used for sentence structure, and i think is confusing if you are not using it in an gramatical sentence Feb 05 18:04:53 i would choose interval over frequency, as it has about the same meaning, and is shorter length of an english word Feb 05 18:04:53 shadows: Only if you're american ;-) Feb 05 18:04:56 easier to type Feb 05 18:04:57 That doesn't make any sense, it's the correct term. Feb 05 18:05:26 hey i like to assert my western imperialist tendancies *achem* Feb 05 18:05:38 No, interval has no established meaning so far as I am aware... Feb 05 18:06:14 period is well defined as 1/frequency, and the SI units (s) are perhaps more familiar than Hz Feb 05 18:07:25 Hz is an SI unit though ;-) Feb 05 18:08:14 Yes, indeed, but it's not a fundamental one - it's derived from mass/lenght/time (+angle if you care). Feb 05 18:08:44 That's not what you said ;-) Feb 05 18:08:59 You put the pauses in the wrong place Feb 05 18:09:07 03coredump 07org.oe.dev * r9d35934e... 10/packages/altboot/ (9 files in 4 dirs): Feb 05 18:09:07 altboot: Feb 05 18:09:07 - New funcitions: init_rootfs, mount_sd, mount_cf and mount_home for smaller code and easier debugging / bugfixing. These functions can be used by all plugins. Feb 05 18:09:07 - Do not try to mount /sys, /proc, /media/cf and /media/home when they are already mounted (mainly for testing purposes) Feb 05 18:09:07 - Added plugin: install-tgz Feb 05 18:09:09 This plugin allows installation of an OE generated .tar.gz rootfs onto CF or SD as a real filesystem or into a loopfile of user-defined size. Still needs some work but is usable. Feb 05 18:09:53 and the (SI units (s)) are perhaps more familiar than [the SI unit] Hz Feb 05 18:10:21 I.e. everyone encounters T, fewer people encounter 1/T Feb 05 18:13:46 jbowler: I'm happy enough to change to use delay_on and delay_off to settle the problem once and forall. It might be easiest if you reply to the timer patch on LKML mentioning the advantages of this, then I can post a new patch. If we do it before the interface is established, we should be ok. Feb 05 18:14:36 k Feb 05 18:26:05 Ciao all Feb 05 18:27:11 ciao Pigi Feb 05 18:27:20 ciao gremlin[it] ! Feb 05 18:29:32 hi pigi Feb 05 18:29:40 hi pb_ Feb 05 18:41:28 ping pb_ Feb 05 18:41:44 gremlin[it]: hello Feb 05 18:42:10 can u please past me a couple of lines from your kernel tree ? Feb 05 18:42:34 arch/arm/mach-pxa/hx4700/hx4700.c Feb 05 18:43:32 the initialization of static struct platform_device hx4700_pxa_keys = { ... line 381 Feb 05 18:44:38 I don't think my tree is quite up to date, but here's what I have: Feb 05 18:44:39 static struct pxa_keys_button hx4700_button_table[] = { Feb 05 18:44:39 { KEY_POWER, GPIO_NR_HX4700_KEY_ON, 0 }, Feb 05 18:44:39 { KEY_F11 /* mail */, GPIO_NR_HX4700_KEY_AP3, 0 }, Feb 05 18:44:39 { KEY_F10 /* contacts */, GPIO_NR_HX4700_KEY_AP1, 0 }, Feb 05 18:44:39 }; Feb 05 18:44:58 er, whoops, missed Feb 05 18:45:00 static struct platform_device hx4700_pxa_keys = { Feb 05 18:45:00 .name = "pxa2xx-keys", Feb 05 18:45:00 .dev = { Feb 05 18:45:00 .platform_data = &hx4700_pxa_keys_data, Feb 05 18:45:00 }, Feb 05 18:45:01 }; Feb 05 18:45:05 no the platform_device dtruct Feb 05 18:45:27 mhhh and it compile ? Feb 05 18:45:32 I guess Feb 05 18:46:12 mhh 2.6.15 remove the .name from the struct ... and other chenges in platform_device ... Feb 05 18:46:18 i'll investigate ... Feb 05 18:48:06 the inteligence report i'm simply dumb :) :) :) Feb 05 19:01:26 ok resume from suspend work ... :) Feb 05 19:01:34 i go to feed ;) :) :) Feb 05 19:29:30 03coredump 07org.oe.dev * r40d10bcb... 10/packages/altboot/files/altboot.func: altboot: Fixed a type. I wonder how that got in there... Feb 05 19:30:06 *sigh* And the patch comment has a....typo... Feb 05 19:39:39 03koen 07org.oe.oz354fam083 * r67e13c0f... 10/packages/gpe-autostarter/ (2 files in 2 dirs): gpe-autostarter: add patch for dbus 0.60 from .dev Feb 05 19:39:43 03koen 07org.oe.oz354fam083 * r2a45d231... 10/packages/gpe-beam/ (gpe-beam-0.2.8/dbus-new-api.patch gpe-beam_0.2.8.bb): gpe-beam: add patch to enable dbus 0.60 from .dev Feb 05 20:00:37 gremlin[it]cena: you probably need to include platform_device.h Feb 05 20:10:03 bbl Feb 05 20:10:08 Anyone know anything about bbfile collections? Feb 05 20:11:21 yes Feb 05 20:11:23 ashak: yup ;) Feb 05 20:11:25 More specifically, do I simply set BBFILES_COLLECTIONS environment variable, or do I put it in the config file or something? Feb 05 20:11:39 ashak: I have it in the conf file Feb 05 20:11:43 zecke: Hiya :) Didn't know if you were awake or not. Feb 05 20:12:03 ashak: hehe I'm supposed to do homework Feb 05 20:12:17 Can BBFILES = "blah" go into conf/local.conf too then? Feb 05 20:12:27 yes Feb 05 20:12:43 Ah ok, cool, i've been setting it manually before using bb Feb 05 20:13:21 zecke: homework anything intersting? Feb 05 20:15:05 ashak: yeah, you can set it in either place. Bitbake inherits all the variables from your environment. Feb 05 20:15:46 the only catch with using the environment is that those values are frozen into the cache, so you have to remember to manually touch local.conf if you change the environment Feb 05 20:16:14 Ah I lied, i've been setting BBPATH manually Feb 05 20:16:32 pb_: Ok thanks Feb 05 20:32:29 Right... if I wanted to make my own modifications to a package, do i need to make a copy of an entire package's directory in my own tree, or just the relevent files that are changing? For example, if I wanted to modify the fstab file in base-files package, do I copy the entire base-files directory from packages to my own tree, or can I just create the relevent directories with an additional folder named the same as my machine name containing my modified fstab Feb 05 20:32:54 (like I have done on top of the default oe tree whilst I was testing) Feb 05 20:33:37 RP yes yes ... i did ... that why i said i ws dumb ... :) Feb 05 20:44:58 ashak: only relevant files and I think the entire basefiles dir Feb 05 20:56:57 koen: that's what I assumed, thanks Feb 05 21:01:58 Goodevening to all ! Feb 05 21:05:42 hey Ifaistos Feb 05 21:06:42 I have one question that came us i am browsing though the .bb files. In order to keep OE arch indepedant, shouldn't all patched be applied with a SRC_URI_append_ instead of the "usual" file directive ? Feb 05 21:07:29 especially for stuff like libc and uclibc Feb 05 21:08:44 i am trying to build a minimal distro that will one day cross-compile on different architectures ( i hope to...) Feb 05 21:10:32 Ifaistos: hi. that depends on the patch Feb 05 21:12:40 zecke: Yes i can undersand that... but for example the uclibc_0.9.28.bb has a lot of patches that are arm related and might fail on x86 or mispsel for example Feb 05 21:13:33 zecke: So instead of # Thumb interworking support Feb 05 21:13:33 SRC_URI += " file://thumb-mov-pc-bx.patch;patch=1" Feb 05 21:14:09 Ifaistos: is it likely that that this patch has a side effect on x86? Feb 05 21:14:16 should we write SRC_URI_append_arm=" file://thumb-mov-pc-bx.patch;patch=1" Feb 05 21:14:22 Ifaistos: if so, make it arm dependant Feb 05 21:14:30 zecke: No it's just an example :) Feb 05 21:14:45 Ifaistos: otherwise it is good to have this patch apply always Feb 05 21:14:57 Ifaistos: people are testing this patch (indirectly) Feb 05 21:16:23 zecke: Thanks. I am asking as i usually apply patches only for the arch that i am targeting (when using "standard" cross compiling not OE) Feb 05 21:16:54 Ifaistos: I find it good to know that the patches did not bitrot... Feb 05 21:18:19 so to make a patch arch dependand should i write SRC_URI_append_arm+="blah blah" Feb 05 21:18:34 for arm for example Feb 05 21:18:38 03koen 07org.oe.oz354fam083 * raac7456f... 10/ (5 files in 5 dirs): Feb 05 21:18:38 gpe: push latest gpe stuff Feb 05 21:18:38 * build libgpewidget with pkgconfig-native from .dev, rebuild pkgconfig-native from .oz354fam083 after that Feb 05 21:18:41 append and += Feb 05 21:18:52 is a bit... doubled ;) Feb 05 21:19:12 ok. got it. thanks ;) Feb 05 21:21:38 zecke on more: I am trying to build a .bb for quagga. it has 4 different daemons that i want to packet at different ipk. At which .bb i should look for an example on how to do it ? Feb 05 21:33:48 hi :) has anyone encountered a problem, that when you try to compile x11 or diet-x11 the configure script can not find keysymdefs.h, allthough it is present in the staging directory? Feb 05 21:34:01 03koen 07org.oe.dev * reedeac97... 10/packages/libgtkinput/libgtkinput_cvs.bb: libgtkinput: use SRCDATE Feb 05 21:35:58 Ifaistos: look foe FILES_ and PACKAGES ;) Feb 05 21:36:18 Ifaistos: e.g. xine makes heavy use of package splitting Feb 05 21:41:41 03tmbinc 07org.oe.dreambox * ra2dd3065... 10/packages/dreambox/dreambox-dvb-modules/dm7025/dream: dreambox-dvb-modules/dm7025: load rfmod module Feb 05 21:41:45 03tmbinc 07org.oe.dreambox * r5f48c6b9... 10/packages/netkit-base/netkit-base-0.17/opendreambox/inetd.conf: netkit-base/opendreambox: fix typo in inetd.conf Feb 05 21:43:12 ...ok I will ask differently: did anyone here manage to compile x11 or diet-x11? :> Feb 05 21:44:24 zecke: Thanks, i'll have a look Feb 05 21:45:06 Hmm, i have broken something, bitbake bootstrap-image isn't creating new output files anymore. It's also not giving me any error. Feb 05 21:51:05 ashak: try 'cleaning' it Feb 05 21:51:16 ashak: it might think it is done Feb 05 21:53:52 I have, bitbake -c clean bootstrap-image, and it seems to successfully clean, I even removed the files that it output and it seems to have recreated the exact same files the next time I bitbake bootstrap-image Feb 05 21:57:24 ashak: what is the problem then? Feb 05 21:59:06 it's building the previous version of the bootstrap-image, I changed base-files, cleaned base-files and bootstrap-image, then tried to built bootstrap-image again. Feb 05 22:05:41 NOTE: build 200602052204: completed, yet the image created is: bootstrap-image-20060205211352.rootfs.cramfs Feb 05 22:06:00 ashak: sounds like a 'cache' issue Feb 05 22:06:08 ashak: what happens if you touch conf/local.conf? Feb 05 22:07:54 zecke: i'll try Feb 05 22:09:36 my shark deadlocked :} Feb 05 22:11:58 zecke: Cool that fixed it Feb 05 22:12:17 ashak: well that is not too cool ;) Feb 05 22:12:25 ashak: what version of bitbake do you use? Feb 05 22:13:52 BitBake Build Tool Core version 1.3.2.1, bitbake version 1.3.2 Feb 05 22:16:55 is anyone running a .dev image from today? Feb 05 22:17:11 nope Feb 05 22:17:20 my openssh server is b0rked in a barebone image Feb 05 22:17:36 might be a missing dependency or general b0rkage Feb 05 22:18:00 it's bitching about missing openpty Feb 05 22:34:31 * chouimat is away: food Feb 05 22:35:28 found my ssh problem..it helps to actually INIT into a runlevel Feb 05 22:39:04 03jbowler 07org.oe.dev * rda977902... 10/packages/slugos-init/ (files/boot/disk slugos-init_0.10.bb): slugos-init: do not create recovery in boot/disk in 0.10 Feb 05 22:54:28 * Philippe is away: visual contact - real goth bed Feb 05 23:18:13 evening Feb 05 23:35:43 hmm im having problems with xlib in staging, certain things link wrong with xlib Feb 05 23:35:56 | xwin.c: undefined reference to `XOpenIM' Feb 05 23:36:05 and a few others Feb 05 23:37:05 simcop2387: you need a bigger xlib Feb 05 23:39:45 pb_ how do i do that Feb 05 23:40:05 build x11 rather than diet-x11 Feb 05 23:44:27 hmm lets figure out how Feb 05 23:48:26 how would i remove diet-x11 from the stage just build x11 Feb 05 23:57:05 nope... hmm wheres that manual Feb 06 00:00:03 -c clean is what i want Feb 06 00:00:46 you'll need to remove stuff from staging manually Feb 06 00:01:15 clean doesn't touch staging Feb 06 00:03:26 reenoo: shouldnt have to worry too much since x11 should replace all if not the important things Feb 06 00:07:22 03mgoebl 07org.oe.dev * rcfc5f212... 10/packages/libao/libao-alsa_0.8.6.bb: Feb 06 00:07:22 libao-alsa_0.8.6: Add libao plugin-package with alsa support Feb 06 00:07:22 - Make libao-alsa_0.8.6 from libao-alsa_0.8.6 and Feb 06 00:07:22 add alsa config and dependencies for plugin Feb 06 00:07:28 03mgoebl 07org.oe.dev * r65b0cd64... 10/packages/musicpd/mpd-alsa_0.11.5.bb: Feb 06 00:07:28 musicpd/mpd-alsa_0.11.5.bb: Add mpd-package with alsa support Feb 06 00:07:28 - Make mpd-alsa_0.11.5.bb from mpd_0.11.5.bb and Feb 06 00:07:28 add alsa-lib and libao dependency (./configure will add alsa support) Feb 06 00:07:32 (mpd not only uses alsa indirectly via libao, but also directly Feb 06 00:07:34 for mixer support) Feb 06 00:09:20 * simcop2387-zauru cant wait for alsa oncollie Feb 06 00:32:05 03rpurdie 07org.oe.dev * rfe920728... 10/packages/linux/ (7 files in 2 dirs): linux-oz-2.6: Switch to use the proposed RTC subsystem. Add an RTC subsystem driver for the sa11xx/pxa2xx. Update defconfigs to match. Feb 06 00:51:51 good nite Feb 06 01:27:26 RP: dev is not merged... Feb 06 01:29:19 JustinP: that's my recent commits Feb 06 01:29:59 they seem to automerge fine, I'll push that Feb 06 01:30:13 jbowler-away: is it? Feb 06 01:30:19 monotone: 65b0cd64eeef02b789460148d6bb958749417f11 mgoebl@openembedded.org 2006-02-05T23:24:19 Feb 06 01:30:23 monotone: fe9207286fe795f141cfced1550795b360e1e971 rpurdie@openembedded.org 2006-02-05T23:59:04 Feb 06 01:31:10 Well, I just pushed an unmerged head to monotone.nslu2-linux.org Feb 06 01:31:25 wonderful Feb 06 01:31:42 Where's the problem? Just do 'monotone merge' Feb 06 01:32:48 The merged head is on dominion.kabel.utwente.nl Feb 06 01:33:27 no wonder we have so many unmerged heads.... Feb 06 01:33:50 Eh? Rapid development being pushed to different servers Feb 06 01:34:11 ewi and vanille were not responding to me, so I pushed it elsewhere Feb 06 01:34:41 monotone.nslu2-linux.org automerges stuff anyway, but only about once an hour Feb 06 01:36:06 Now if monotone supported multiple simultaneous writers it would be a lot easier. Feb 06 03:26:49 * chouimat is back. Feb 06 08:14:34 hi Feb 06 08:16:21 good morning all Feb 06 08:16:41 haha! Feb 06 08:16:55 RP: with your image, my Z freezes up as before Feb 06 08:16:59 initially it worked Feb 06 08:17:02 recent posts to monotone-devel are pretty cute Feb 06 08:17:02 now it is non-working Feb 06 08:17:18 they keepusing the OE database as an example :) Feb 06 08:17:25 (http://lists.gnu.org/archive/html/monotone-devel/2006-02/msg00055.html) Feb 06 08:19:20 koen: good morning Feb 06 08:19:52 ah Feb 06 08:19:57 yes i suppose that is cute Feb 06 08:20:40 hey Ifaistos Feb 06 08:23:03 RP: did u try to reach me on saturday? Feb 06 08:55:05 morning Feb 06 08:59:36 cedric: I've made some changes to the poodle defconfig - I think it should work better now (and might not need that CMDLINE change) Feb 06 09:00:32 morning all btw :) Feb 06 09:01:07 hey RP Feb 06 09:04:14 heh Feb 06 09:04:36 mickeyl's buddy are slowing down vanille.de with their python stuff Feb 06 09:04:51 load average of 11 Feb 06 09:05:27 Cpu(s): 99.0% us, 0.7% sy, 0.0% ni, 0.0% id, 0.0% wa, 0.0% hi, 0.3% si Feb 06 09:06:33 nice... Feb 06 09:06:48 The LED subsystem made -mm btw Feb 06 09:06:53 cool Feb 06 09:08:33 morning all Feb 06 09:08:39 hey do13_ Feb 06 09:08:45 hi dirk Feb 06 09:08:58 RP: thx Feb 06 09:09:36 hi Richard, koen Feb 06 09:09:48 RP: the kernel initialise 13M of RAM instead of 32M Feb 06 09:10:08 try passing mem=32M in the CMDLINE Feb 06 09:10:24 cedric: If that works, we can hardcode it in poodle.c Feb 06 09:10:26 do13: I've heard that u have some interesting patch for me... Feb 06 09:11:05 RP: I'll have a look tonite Feb 06 09:11:37 cedric: The Sharp bootloader only sets up part of the ram on some machines - I know corgi.c has to set the memory size for the c7x0's Feb 06 09:13:47 RP: I'm bit lost for the touchscreen, do I need to contact LKM? I think he pushed some stuff for it Feb 06 09:14:50 RP: congrats, the leds patch is now mm tree Feb 06 09:15:45 cedric: Do you mean the sharpsl_pm? Feb 06 09:16:26 do13_: yep Feb 06 09:19:27 cedric: I've applied Dirk's patch to the kernel tree Feb 06 09:19:31 cedric: This patch should be oe. Fixed battery values can now be set in the macine specific xx_pm.c file. Feb 06 09:19:34 do13_: It seems to work fine btw Feb 06 09:20:04 do13_: Its nice to see the leds getting somewhere :) Feb 06 09:20:16 RP: Did you tested on corgi? Feb 06 09:20:21 cedric: LKM? Feb 06 09:20:28 do13_: I tested on spitz Feb 06 09:20:46 morning all Feb 06 09:20:50 RP: the guy at the top of the ARM developement Feb 06 09:20:56 morning ade|desk Feb 06 09:21:06 cedric: That's rmk Feb 06 09:21:19 RP: :-) Feb 06 09:21:41 cedric: I'm not sure what you'd contact him for - he's pushed some stuff but I think that's as much as he will do Feb 06 09:22:51 RP: You should do a test on corgi, because corgi use different charge off handling Feb 06 09:23:10 good morning Feb 06 09:28:26 do13_: I don't have access to a corgi - just husky Feb 06 09:29:06 RP: Ahhh Feb 06 09:29:10 RP / do13_ --> can one of u send me an email about give me guidelines around the led patch & the pm patch, so I can implement it for the poodle, there is so much stuff to do to be as the same level than the C7x0 machines, so if I can do the easy stuff for now then that'll be a start... Feb 06 09:29:46 cedric: You might want to see if any of Pavel Machek's code from his git tree didn't make it into mainline - that's where the touchscreen patches might be Feb 06 09:30:43 cedric: Basically you need to go through the 2.4 sharpsl_power.c code and find all the differences (which will be marked with ifdef(CONFIG_MACH_POODLE)'s ) Feb 06 09:31:19 cedric: Poodle uses the AD7846 for touchscreen or the ucb stuff? Feb 06 09:31:33 You then need to find a way to apply the differences neatly to the new code. A lot of of the hooks are preexisting Feb 06 09:32:04 I suspect it may have the 7846 Feb 06 09:32:24 RP: he sent me some stuff but I don't think is correct, I haven't applied his patches but I've just read his patches and I can't see anything for the touchscreen, only stuff for the backlight which are coded directly in the locomo.c Feb 06 09:32:42 RP: I though it ucb Feb 06 09:32:53 it was... Feb 06 09:33:23 cedric: poodle defconfig: # CONFIG_MCP_UCB1200 is not set Feb 06 09:33:43 cedric: and CONFIG_ADS7846_TS=y Feb 06 09:34:02 ok... Feb 06 09:34:08 cedric: See drivers/char/ads7846.c Feb 06 09:34:24 cedric: This means corgi_ts.c might work for you with some tweaks Feb 06 09:35:20 how can I be sure it's the 7846, I though John L. done some stuff the ucb... Feb 06 09:35:47 cedric: ucb is used on collie Feb 06 09:35:59 right Feb 06 09:37:42 RP / do13_ : will be nice to work as a pair programming... XP style... Feb 06 09:37:56 cedric: Looking at the Sharp 2.4 code for that, the poodle specific bits look like they can be ignored Feb 06 09:38:30 cedric: For poodle_pm.c you can use corgi_pm.c with some changes from the old sharpsl_battery.c Feb 06 09:38:35 RP: pm stuff? Feb 06 09:38:37 cedric: You'll just have to work out how corgi_ssp.c can be used Feb 06 09:38:57 cedric: touchscreen stuff Feb 06 09:39:06 RP: ok Feb 06 09:39:26 but corgi_ssp will also be needed for the _pm code Feb 06 09:40:18 oh lalala!!! :-) Feb 06 09:41:31 corgi_ssp deals with the max1111 voltage readings, the touchscreen and on corgi, the lcd. You will probably need 2 out of 3 of its functions Feb 06 09:42:17 03nail 07org.oe.dev * refc8bc6b... 10/packages/initscripts/ (3 files in 2 dirs): Feb 06 09:42:17 initscripts: Update SlugOS initscripts Feb 06 09:42:17 Update checkroot to include led status/beep on error, mount swap Feb 06 09:42:17 Update checkroot, move $rootcheck to $ROOTFSCK in rcS Feb 06 09:42:17 Update mountall, remove the swap mount, since it's already mounted Feb 06 09:42:21 03nail 07org.oe.dev * r59438d13... 10/packages/sysvinit/ (sysvinit/slugos/rcS-default sysvinit_2.86.bb): Feb 06 09:42:22 sysvinit: Update SlugOS rcS Feb 06 09:42:25 Add DOSWAP and ROOTFSCK, remove TICKADJ variable. Feb 06 09:42:27 03jbowler 07org.oe.dev * rad7d09e3... 10/packages/ (6 files in 6 dirs): Feb 06 09:42:28 initscripts-slugos: modify the slugos checkroot.sh not to hang on fsck failure in 1.0 Feb 06 09:42:30 slugos-init: add a single_user_ok function in /etc/default/functions in 0.10 Feb 06 09:42:32 sysvinit: update slugos rcS with new options in 2.86 Feb 06 09:42:34 - this is to fix the fsck of the rootfs so that it will actually work Feb 06 09:42:38 safely on machines with no console. Feb 06 09:42:43 03jbowler 07org.oe.dev * r2510e525... 10/packages/initscripts/initscripts-1.0/slugos/checkroot.sh: initscripts-slugos: correct checkroot.sh to default to DOSWAP in 1.0 Feb 06 09:42:45 03nail 07org.oe.dev * ref331acf... 10/packages/setpwc/ (files/makefile.diff setpwc_1.1.bb): setpwc: Upstream upgrade to 1.1 Feb 06 09:42:48 03jbowler 07org.oe.dev * re1959e48... 10/packages/meta/slugos-packages.bb: slugos-packages: setpwc version 0.9 is no longer available, remove from conf Feb 06 09:42:53 03jbowler 07org.oe.dev * r2eccf1b9... 10/packages/ (5 files in 4 dirs): Feb 06 09:42:56 initscripts-slugos: further fixes to checkroot.sh in 1.0 Feb 06 09:42:58 slugos-init: make beep silent, extend single_user_ok function in 0.10 Feb 06 09:42:59 03jbowler 07org.oe.dev * r930c14ed... 10/packages/initscripts/ (2 files in 2 dirs): initscripts-slugos: make checkroot.sh work with any sulogin in 1.0 Feb 06 09:43:03 03jbowler 07org.oe.dev * r19b7c1f5... 10/conf/distro/ (10 files): slugos: release 3.3-beta Feb 06 09:43:08 03jbowler 07org.oe.dev * rae1272ce... 10/conf/distro/slugos.conf: slugos: development branch 3.4 Feb 06 09:43:15 03koen 07org.oe.dev * r0dcf135f... 10/ (5 files in 5 dirs): GPE: push latest gpe bits Feb 06 09:43:25 RP: where do u leave? about a intensive kernel week-end session? :-) Feb 06 09:43:40 s/leave/live Feb 06 09:44:35 North East England Feb 06 09:44:56 cedric: I could suggest coming to FOSDEM but I suspect I'm going to be busy enough as it is :-/ Feb 06 09:45:29 I can't do FOSDEM this year Feb 06 09:46:20 cedric: Where do you live? Feb 06 09:46:36 next to geneva Feb 06 09:47:03 I've left UK in August 2004... Feb 06 09:48:53 do13_: and u? Feb 06 09:49:19 cedric: Germany Feb 06 09:51:48 and about another europe session, special zaurus kernel? Feb 06 09:57:52 morning Feb 06 09:57:57 hey hrw|work Feb 06 09:58:09 hey Marcin Feb 06 09:58:37 hrw|work: I suspect you want to do http://ewi546.ewi.utwente.nl/tmp/viewmtn/revision.psp?id=835b381694f63907d7e349c625eb909189a3bf73 for OZ as well Feb 06 09:59:28 hi hrw|work Feb 06 10:00:09 cedric: It might be possible to do something online sometime - if you're around to do some testing, it might be possible to sort the touchscreen/pm code out a bit. I just need to find the time to have a look at the code :-/ Feb 06 10:00:31 let's switch to 48 hour days Feb 06 10:01:00 koen: ok. Feb 06 10:01:30 RP: that'll be great, just let me know and I'll booked my night! Feb 06 10:02:31 cedric: Grep the old pxa_ssp.c and put this changes into corgi_ssp.c or create poodle_ssp.c Feb 06 10:04:00 do13_: I'll have a tonite, I don't have an OE envirnement at work Feb 06 10:04:16 s/a/a look Feb 06 10:04:24 morning Feb 06 10:04:48 cedric: I'd be interested to know how the kernel works after the defconfig updates (particularly SD cards) Feb 06 10:05:51 RP: I'll update it and let u know Feb 06 10:07:36 RP: drop an email when u think doing the XP evening Feb 06 10:09:09 hi Feb 06 10:09:23 ~curse kspread devels Feb 06 10:09:25 May the fleas of a thousand camels infest your most sensitive regions, kspread devels ! Feb 06 10:09:26 cedric: I'm not sure when it'll happen but i'll keep it in mind Feb 06 10:12:49 RP: and about next week, so we've got time to arrange something, sorry to be pushy... Feb 06 10:13:14 RP: sound problem, the problem when pausing and restarting in mplayer has returned Feb 06 10:13:31 XorA: Which problem was that? Feb 06 10:14:03 RP: you pause, unpause, then mplayer stops playing Feb 06 10:14:44 XorA: Which kernel version causes the regression? Feb 06 10:15:00 or more specifically, which ASoC version? :) Feb 06 10:15:31 RP: only with alsa , oss emulation works Feb 06 10:16:05 XorA: I suspect this is one for Liam when he returns... Feb 06 10:16:13 2.6.15-r1 Feb 06 10:16:35 RP: not serious, problem, but I remeber this was nailed once before Feb 06 10:16:43 2.6.15-r0 was ok? Feb 06 10:17:23 RP: Im not sure when it last worked, I dont pause video that often Feb 06 10:17:47 I don't remember this bug at all :-/ Feb 06 10:21:44 cedric: Now we have a generic wm8731 driver, you should be able to get sound on poodle easily btw - just need a poodle machine file for the sound Feb 06 10:24:09 RP: great, so just need to set the 8731 driver in the kernel? Feb 06 10:25:39 cedric: Its a little more complex than that - see sound/soc/pxa/corgi.c for an example - should be within your capabilities though Feb 06 10:26:26 RP: lol Feb 06 10:27:38 I'll print this irc log for reference... Feb 06 10:38:29 did bison-native change recently? Feb 06 10:40:43 ERROR: invalid load key, 'H'. while parsing /home/dp/zaurus/org.openembedded.dev/packages/bison/bison-native_2.0.bb Feb 06 10:47:48 XorA: load key? what is that :} Feb 06 10:49:29 zecke: I have no clue, something wierd is going on here, debugging Feb 06 10:49:47 zecke: something is corrupting the cache I think Feb 06 10:53:42 zecke: yeah, something blew my cache, a quick clean and its off working again **** ENDING LOGGING AT Mon Feb 06 10:59:56 2006