**** BEGIN LOGGING AT Mon Dec 04 02:59:57 2006 Dec 04 03:55:11 !oebug 929 Dec 04 03:55:27 !oebug 929 Dec 04 03:55:32 * * Bug 929, Status: NEW, Created: 2006-05-01 07:50 Dec 04 03:55:34 * * pb(AT)reciva.com: staged quilt scripts contain absolute path to gawk Dec 04 03:55:36 * * http://bugs.openembedded.org/show_bug.cgi?id=929 Dec 04 04:01:11 03koen 07org.oe.dev * rd6c00f5b... 10/ (3 files in 3 dirs): gcc 4.1.1: add patch to fix an ICE with building glibc with certain CFLAGS, if you get the ICE, you should update to binutils 2.17.50.0.5 as well Dec 04 06:20:59 !oebug 1495 Dec 04 06:21:00 * * Bug 1495, Status: NEW, Created: 2006-10-16 13:49 Dec 04 06:21:02 * * likewise(AT)gmx.net: icecc.bbclass fails to create a proper tar.bz2 file with the cross-compiler Dec 04 06:21:02 * * http://bugs.openembedded.org/show_bug.cgi?id=1495 Dec 04 06:35:15 !oebug 1553 Dec 04 06:35:16 * * Bug 1553, Status: RESOLVED, Created: 2006-11-04 00:23 Dec 04 06:35:17 * * hvontres(AT)sbcglobal.net: Patch to get oss volume working on poodle Dec 04 06:35:18 * * http://bugs.openembedded.org/show_bug.cgi?id=1553 Dec 04 07:56:18 morning Dec 04 08:03:26 XorA : Hi Dec 04 08:06:02 morning Dec 04 08:12:37 hrw|work, morning Dec 04 08:21:21 hi hrw|work Dec 04 08:24:34 good morning all Dec 04 08:28:46 Hello Dec 04 08:29:18 I would like to know if there is a clean way for bitbake not to consider some packages when building gpe-image Dec 04 08:29:41 There are a couple of packages that fail for me, and that I do not consider important at this time Dec 04 08:30:33 I see that gpe-image depends upon task-gpe, but removing the said packages brings a error when building rootfs Dec 04 08:31:49 ERROR: Cannot satisfy the following dependencies for gpe-task-pim: gpesyncd Dec 04 08:32:14 Where does bitbake gets it's data for that? Dec 04 08:32:35 I have already (cleanly) deleted the entry from task-gpe... Dec 04 08:34:14 add them to MASK Dec 04 08:34:46 In local.conf Dec 04 08:34:52 ? Dec 04 08:35:00 yes Dec 04 08:35:06 Aye, I'll try that Dec 04 08:35:07 ex: BBMASK = "(opie|qt|qpe)" Dec 04 08:35:08 Thanks Dec 04 08:35:10 Aye Dec 04 08:47:17 Nope, still get the same error Dec 04 08:47:25 It isn't the cache, it was rebuilt Dec 04 08:47:44 it isn't task-gpe, I have removed the entries Dec 04 08:52:43 adding them to BBMASK wouldn't help you a bit Dec 04 08:53:02 Ah? Dec 04 08:53:10 since BBMASK masks them for the parser, while they keep being referenced in the task files Dec 04 08:53:18 Why am I getting such an error? Dec 04 08:53:27 I have removed them from the task files Dec 04 08:53:33 so as usual: "ignore anything emte says" Dec 04 08:53:39 :) Dec 04 08:53:45 B_Lizzard: and you bumped PR on the tasks? Dec 04 08:53:56 PR? Dec 04 08:56:34 As far as I know, I have removed all references to such packages Dec 04 08:56:53 At least from the tasks referenced in gpe-image Dec 04 08:57:26 right, but if you don't bump PR, you still have the old task-foo.ipk in deploy referencing those packages Dec 04 08:58:06 Excuse me, but I don't understand what "bump PR" means. Dec 04 08:58:22 Sorry if I'm being stupid or uncooperative Dec 04 08:58:53 as usual "dont expect useful help from koen" Dec 04 08:59:01 Might you be so kind as to explain what you mean in that? Dec 04 08:59:22 You complete each other, in a way, eh? Dec 04 08:59:27 emte and koen Dec 04 08:59:31 :) Dec 04 08:59:55 Maybe if you merge, you can be the ultimate help-bot Dec 04 09:00:47 I'm affraid not Dec 04 09:01:00 nothing can counter the cluelessness displayed by emte Dec 04 09:01:16 :| Dec 04 09:01:35 and if you were anywhere near as unbalanced as koen you would be in a hospital Dec 04 09:01:57 B_Lizzard: check the task recipe and see if you find a variable called 'PR' Dec 04 09:02:15 PR = "r3" Dec 04 09:02:17 r4? Dec 04 09:02:21 yes Dec 04 09:02:33 Ok, thanks Dec 04 09:02:43 You could have just said revision Dec 04 09:02:46 :) Dec 04 09:02:58 'revision' isn't mentioned usually Dec 04 09:03:32 Ok, in any case, thank you for your helpful information Dec 04 09:03:39 Both of you ;) Dec 04 09:04:03 B_Lizzard: http://www.openembedded.org/bonsai/query/description/bump%20PR/?scm_id=3 Dec 04 09:04:46 Ah, I see. I could have checked on google, as it seems Dec 04 09:05:13 Well, no harm done I guess. Only a couple of broken hearts. LOVE WILL PREVAIL Dec 04 09:13:11 * koen suspects Ifaistos, Crofton and likewise have an email from fed-ex in their inbox Dec 04 09:16:05 Ah, it all makes sense now, koen Dec 04 09:16:11 It worked like a charm Dec 04 09:16:14 Thanks a lot Dec 04 09:32:49 Is bugzilla.handhelds.org still used by OE? I mean bugs in ipkg should still go there, I guess? Dec 04 09:34:24 Laibsch: OE never used it Dec 04 09:34:43 good morning hrw|work Dec 04 09:34:57 Question remains where bugs in ipkg et al should go. Dec 04 09:37:52 Ifaistos, likewise: icecc works for you? It is now horribly broken for me. The error I get is rather strange: | Error: no such file "ICECC_VERSION=/export/interim/openembedded/tmp/openzaurus/cross/ice/arm-linux-openzaurusarm-linuxhard3.4.4Rie.tar.bz2" Dec 04 09:42:08 Laibsch: ipkg is hh.org Dec 04 09:43:40 So, I guess we have four bugtrackers now? http://wiki.openzaurus.org/Bugtrackers Dec 04 09:46:16 Laibsch : I am having problems getting icecc to work reliably even if the distro is different i.e ubuntu and debian Dec 04 09:46:50 Ifaistos: For me it already fails with a single computer. It completely breaks compilation after all. Dec 04 09:46:51 Laibsch : i am still testing "nornal" builds, havent touched OE yet Dec 04 09:47:03 Ifaistos: I see. Dec 04 09:48:35 Laibsch : I believe we first need to set up a working icecc network and then see what happens with OE Dec 04 09:48:54 Laibsch : or we will not know where the problem is from Dec 04 09:50:20 hrw|work: Would you approve of my latest changes to http://wiki.openzaurus.org/Bugtrackers ? Dec 04 09:50:40 good morning Dec 04 09:50:46 Ifaistos: Makes sense. I thought the icecc base problems had been worked out already. Dec 04 09:51:02 Good morning florian_kc. What do you say about current http://wiki.openzaurus.org/Bugtrackers Dec 04 09:56:30 Laibsch: That's good... but i wonder which apps use the hh.org bugzilla. iirc this one is for familiar packages only now, Dec 04 09:57:50 Laibsch : btw koen yesterday checked in a bug fix, did you tried with that ? Dec 04 10:08:15 I've read some where that I need to compile my kernel image for my specific device.... at the moment I'm using the hx4700 profile but I have a rx4540 device... will the hx4700 profile work for me? Dec 04 10:10:25 Ifaistos: not sure. but I will do without icecc for now until somebody says it works at least partially for him. Dec 04 10:10:40 florian_kc: ipkg is an example. Dec 04 10:11:26 Laibsch: ipkg sufferes from the same problem that gpe did untils some weeks ago: it does not have its own bugtracker. Dec 04 10:12:15 Suffering or not. It currently is the place to report things to. BTW, did you move over gpe bugs reported in hh.org to ltg? Dec 04 10:13:35 What is the current QA setup? I read about bittest in the wiki. I wonder if there is any kind of process behind it like we have the bug days. Is it only zecke working on this (time permitting I guess) or are other people making a coordinated effort? Dec 04 10:15:09 03koen 07org.oe.dev * rcef334c7... 10/ (1 packages/tslib/tslib_1.0.bb): tslib: remove some more detect-stylus depends for devices that can use a recent kernel and udev to get the touchscreen0 symlink Dec 04 10:15:46 morning all Dec 04 10:16:17 hey likewise Dec 04 10:16:24 koen: hey Dec 04 10:18:55 i use device.sh :P Dec 04 10:23:28 03koen 07org.oe.dev * r425cf1b3... 10/ (1 packages/tslib/tslib_1.0.bb): tslib: fix HOMEPAGE Dec 04 10:24:21 koen: judging from your commit, the bug report was correct and my database isn't broken? Dec 04 10:24:53 ArjanS: actually, no Dec 04 10:25:16 *something* depends on detect-stylus, but it's not xserver-kdrive Dec 04 10:25:38 morning folks Dec 04 10:25:40 hey mickeyl Dec 04 10:26:12 koen: interesting blog entry. this sounds as 2.10.6 now performs better than 2.6.10 ? Dec 04 10:30:34 mickeyl: that still isn't the case Dec 04 10:30:55 mickeyl: displaying complex svgs would be a lot faster now Dec 04 10:31:37 mickeyl: but with the xft patch in OE and a recent cairo gtk 2.10 is not a lot slower as 2.6, but a bit Dec 04 10:31:57 ok, fair enough. Dec 04 10:32:52 mickeyl: the tesselate fuction got a *thousand* times faster from 1.2.6 -> 1.3.5 Dec 04 10:33:06 educate me... what's a tesselater anyway? Dec 04 10:33:06 so you can see that cairo was doing something *very* wrong in 1.2.x Dec 04 10:33:50 mickeyl: a tesselator breaks down the complex geometry (like bezier curves) into triangles Dec 04 10:34:10 understood Dec 04 10:34:38 "of the plane is a collection of plane figures that fills the plane with no overlaps and no gaps." from wikipedia Dec 04 10:34:51 "A tessellation or tiling of the plane is a collection of plane figures that fills the plane with no overlaps and no gaps. " Dec 04 10:34:52 even Dec 04 10:35:36 http://en.wikipedia.org/wiki/Tesselate Dec 04 10:37:12 hrw|meeting: I believe you like increasing the fixed bug count ;-) Dec 04 10:37:27 * Laibsch points to bug 172 Dec 04 10:45:34 morning all Dec 04 10:46:27 hey Liam Dec 04 10:48:23 This could be an intresting application to have in OE.dev. I have tried to make it, but it failes building. http://www.umts-tools.com/ Dec 04 10:49:08 koen: so you are sure the speed upgrade is not bogus? (there was still doubt on the cairo list)? Dec 04 10:49:50 likewise: the speed upgrade isn't bogus, but I do share the doubts of the thousandfold increase Dec 04 10:50:38 koen: some ppl reported that various different runs produced very different results. Well, we will soon see. The overall progress is *good*! Dec 04 11:09:49 Laibsch : well the good news is that x86_64 can compile using i686, the bad news is that it takes all the compile requests... Dec 04 11:11:00 Laibsch : for some reason i have not figured out is shows a load of 16 to icecc while the i686 machines show a load of 1000 Dec 04 11:11:43 Ifaistos: Load of 1000? Puh! Never got above ~20ish Dec 04 11:14:46 Laibsch : actually this is icecc's unit of mesurment to see if a machine can accepts another compile job. And in the case of the i686 machines is wrong... as they are idle Dec 04 11:15:16 Laibsch : looking at icecc code to see how it computes that at the moment Dec 04 11:24:56 crazy Ipaq users, just got an email asking me to enable win32 support. Dec 04 11:25:04 in mplayer Dec 04 11:25:41 XorA: lol Dec 04 11:26:25 no qemu-i386 on your ipaq then ? Dec 04 11:26:35 * XorA doesnt even own an Ipaq Dec 04 11:27:42 Laibsch: I see that my idea to enable utf8 in bugzilla worked out :) Dec 04 11:28:01 is there a GPS software for linux? Dec 04 11:28:11 gpsd, gpsbabel, etc Dec 04 11:28:17 koen: Thanks for that. Not the first time I use it though. Dec 04 11:28:22 and routeplanners? Dec 04 11:28:24 ~google gps linux Dec 04 11:28:30 tnx Dec 04 11:32:44 i can't find any software like tomtom or so ... positions, roadmaps, but no garmin/tomtom like software that tells you how to drive etc Dec 04 11:34:11 re Dec 04 11:38:47 BTW: updated the BitBake manual on berlios Dec 04 11:41:28 koen, you have an email regarding EfIKA? Dec 04 11:41:43 Crofton: yes Dec 04 11:41:56 argh Dec 04 11:42:01 what does it say? Dec 04 11:42:29 "Bill Buck van GENESI-USA, INC heeft aan Koen Kooi 1 FedEx International Dec 04 11:42:29 Economy pakket(ten) gezonden." Dec 04 11:42:39 it's a mail from the fed-ex tracker Dec 04 11:42:57 hmm Dec 04 11:43:23 none yet Dec 04 11:43:33 maybe they are atching in order of project approval Dec 04 11:45:55 * Laibsch points hrw|meeting to bugs 172, 716, 1314 and 1155 as possibly easy picks. Dec 04 11:46:34 Crofton: or packaging it up for shipping takes a while Dec 04 11:46:52 Crofton: or domestic shipments use USPS Dec 04 11:47:13 yeah Dec 04 11:47:22 * Crofton crosses his fingers Dec 04 11:47:47 any idea what we need to power it and what to use for mass storage? Dec 04 11:48:11 atx and PATA ide drive iirc Dec 04 11:49:41 Are there ppl regularly running bittest besides zecke? Dec 04 11:49:42 I should get one of the flash to ATA adapters at some point Dec 04 11:50:04 Subject line on libc-ports : (ARM EABI) ports-20061127 + kernel-headers-2.6.18 == broken system Dec 04 11:51:12 Crofton: right, that's why angstrom still uses 2.6.15.99 Dec 04 11:51:33 Crofton: the fix should be simple, just hook up the *at syscalls for arm Dec 04 11:51:49 how do you control wich headers are used? Dec 04 11:52:40 argos-wlan:~/Projects/OpenEmbedded/org.openembedded.dev koen$ grep headers conf/distro/angstrom-2007.1.conf Dec 04 11:52:40 PREFERRED_VERSION_linux-libc-headers_i686 ?= "2.6.18" Dec 04 11:52:41 PREFERRED_VERSION_linux-libc-headers_powerpc ?= "2.6.18" Dec 04 11:52:41 PREFERRED_VERSION_linux-libc-headers ?= "2.6.15.99" Dec 04 11:54:18 wew Dec 04 11:54:25 I copied that line from angstrom Dec 04 11:56:07 Crofton : Got an email also for the Efika Dec 04 11:56:51 you are also in EU? When did you sign up? Dec 04 11:56:54 Crofton : i am crossing fingers they have not declared a value > 50$ or i the custom officials will be very happy Dec 04 11:57:26 Crofton : Yes EU. I signed up and then told koen ;) Dec 04 11:57:51 heh Dec 04 12:02:02 Crofton : Fedex has the worst service here... if the declared value is .5 cents over the import limit they charge something like 100-120Euro to clear it from customs Dec 04 12:04:34 I need to work harder getting my USRP to work on Full speed usb ... Dec 04 12:05:49 man I hate my new toy Dec 04 12:06:22 bummer Dec 04 12:07:29 zecke: which new toy? Dec 04 12:07:39 T830 of Fujitsu-Siemens Dec 04 12:07:43 zecke: is it green and unethical? Dec 04 12:08:23 koen: it is sort of green and unethical but different to what you think Dec 04 12:09:02 which is the best routeplanner for linux? Dec 04 12:16:23 koen: Is it http://www.ppcnux.com/modules.php?name=News&file=article&sid=6614 you guys are talking about? Makes me druel. How much did you pay? Dec 04 12:18:07 pay? Dec 04 12:18:53 pay? Dec 04 12:19:33 Laibsch: it was (is) a developer program, so we get them for free Dec 04 12:21:18 Wow, that is awesome. I have long used low-power devices rather than THz machines which keep a nuclear power plant busy. But I wonder if that is for mass-market. I am sure there will be some ppl who like it, though. Dec 04 12:33:02 Laibsch: http://designer.spreadshirt.jp/ does not resolve here Dec 04 12:33:53 Should we sign up for this: http://www.powerdeveloper.org/8641d.php Dec 04 12:33:58 Not really embedded though Dec 04 12:35:07 koen: Thanks for letting me know. I should remove that URL. The project is being phased out. Dec 04 12:41:10 Laibsch: I want to inform you that my realname is Marcin Juszkiewicz Dec 04 12:41:59 Crofton: ppc is cool, so the more people having ppc hw, the better Dec 04 12:42:05 hrw|meeting: I know. Did I piss you off in any way using your nick or something? Dec 04 12:42:21 koen, that is why we need PS3's :) Dec 04 12:45:41 Laibsch: look at bugs which you pointed me to Dec 04 12:47:54 Laibsch: first 2 ones have my note from yesterday Dec 04 12:47:58 hrw|tv: I know you commented on them. But do you monitor them? Just wanted to send you a poke that I guess most of them can be closed. Dec 04 12:48:12 yesterday? Dec 04 12:48:49 Oh, I misread the date as being February and March Dec 04 12:49:08 I thought they had been sitting idle for much longer. Dec 04 12:51:59 Laibsch: remember that anyone can close bugs... Dec 04 12:53:55 hrw|tv: Don't worry. But I rather tread carefully. I do close bugs when I understand them 100% and if I do not interfere with anybody's work. In doubt I usually err on the side of asking one times too many. Dec 04 12:54:14 s/In doubt/When in doubt/ Dec 04 12:56:48 Laibsch: generate laibsch@openembedded.org key (or other userid but recognizable one please) and send it to koen and mickeyl via mail Dec 04 12:57:37 like http://www.openembedded.org/wiki/MonotonePhraseBook says :) Dec 04 12:58:54 yep Dec 04 13:02:09 wb mickeyl Dec 04 13:02:14 hey hrw|work Dec 04 13:02:21 enjoyed lunch? Dec 04 13:03:46 ping lrg Dec 04 13:04:12 hey greentux Dec 04 13:04:25 morning all Dec 04 13:04:26 hrw|work: Will do. Dec 04 13:04:27 hi lrg ... :) Dec 04 13:04:32 morning Dec 04 13:04:33 hrw: absolutely. visited my mother for lunch Dec 04 13:04:39 yo RP, lrg, Laibsch, greentux Dec 04 13:04:50 lrg: koen told me you are working @IMX31? Dec 04 13:04:50 hey Richard, Mickey Dec 04 13:05:13 greentux: yep, it's one of my many things to do Dec 04 13:05:25 hey lrg Dec 04 13:05:26 lrg: you are using the gda board? Dec 04 13:05:34 hey vlo|work Dec 04 13:05:43 greentux: no, the IMX31ADS Dec 04 13:06:07 mickeyl: that explain all Dec 04 13:06:08 lrg: gda has a nice smartphone reference kit... Dec 04 13:06:12 hrw|work: :)) Dec 04 13:06:18 koen: I had been looking for that page for some time this morning. Thanks for the URL. Dec 04 13:06:36 greentux: I saw at in Florida in June and spoke with their director Dec 04 13:06:36 lrg: do you have a link to the board you are using? Dec 04 13:06:56 greentux: it's on the Freescale site Dec 04 13:07:26 lrg: http://www.gdatech.com/ref_IMX31smartphone.shtml Dec 04 13:09:51 greentux: from what I see, not so different than ADS board Dec 04 13:10:47 vlo|work: ok, the gde has all the smartphone features gprs/wlan etc Dec 04 13:13:42 yep ... so you should be able to get all the work from ADS work and integrate it on this one Dec 04 13:14:41 vlo|work, lrg: ok nice. Or I finde some devs and send a gda board to them :) Dec 04 13:15:03 vlo|work: what is supported with oe and what ha sproblems??? Dec 04 13:15:59 greentux: we have oe support for mx31ads :) Dec 04 13:16:26 lrg: ok, with all componentes? usb, lan etc? Dec 04 13:16:58 lrg: apm? pcmcia? mmc,sd? so many features :) Dec 04 13:17:29 greentux: most of them, It's based on the Freescale kernel (2.6.16), although I've upgraded it to 2.6.19. Dec 04 13:17:53 lrg: fine. what is the intention for you to support this cpu? Dec 04 13:18:33 greentux: I'm only interested in audio Dec 04 13:18:44 only? audio rocks! Dec 04 13:19:05 * mickeyl can't wait to work on a Synthesizer based on Wolfson chips... Dec 04 13:19:19 mickeyl: :) Dec 04 13:19:36 lrg: cant find on the freescale site how many audio in/outs analog/digital available ? Dec 04 13:19:53 mickeyl: My main interest is really in successor to mx31, cant say anymore Dec 04 13:19:54 * koen can't wait for a DMX512 controller built into a pda :) Dec 04 13:20:18 lrg: *nod* did you see marvel announcing their commitment to the pxa series recently? Dec 04 13:20:47 greentux: thay have atlas PMU and audio chip. I think it has an NDA before you can get deatils Dec 04 13:21:01 mickeyl: not yet, but I think Graeme did Dec 04 13:21:27 mickeyl: fwiw I have a Zylonite on order Dec 04 13:21:50 lrg: ah cool. tell me how it performs please Dec 04 13:22:11 lrg: you're going to make it to FOSDEM 07 ? Dec 04 13:22:33 lrg: the zylonite boards still don't have a cpu with vfp, right? Dec 04 13:22:48 mickeyl: performs well. I was with Stanley last Feb in Shanghai testing and coding on the board. We will do a comparison to pxa27x when it arrives Dec 04 13:22:56 mickeyl: yep Dec 04 13:23:00 lrg: excellent! Dec 04 13:23:07 lrg: book the Astrid this time :)) Dec 04 13:23:14 koen: yep, no fpu, but it does have mmx2 Dec 04 13:23:50 ah, another "performance by marketing" thing Dec 04 13:23:56 mickeyl: will do, we must remember the AP and extension power sockets :) Dec 04 13:24:28 lrg: details on how many input output interfaces ??? 8) Dec 04 13:24:41 lrg: righto Dec 04 13:25:25 greentux: my mx31 board has 11 3.5mm jacks Dec 04 13:25:39 lrg: oh nice. Dec 04 13:25:45 greentux: so does mine ;) Dec 04 13:26:29 * XorA denies all Dec 04 13:26:42 mickeyl: can you edit openembedded.org/events too? Dec 04 13:27:02 vlo|work: XorA ran into problem whereby init didn't boot with recent images. This might be why your mx31 board didn't boot Dec 04 13:27:16 hrw|work: sure Dec 04 13:27:17 * mickeyl does Dec 04 13:28:06 lrg: nice to hear. so the upcoming mx31 devices will be good supported Dec 04 13:28:41 greentux:yep, although Freescale need to do some cleaning up work on their kernel source before it will go into mainline Dec 04 13:28:46 lrg: hum, very interesting ... I've not been able to move any further. Angstrom kernel is able to boot the LTIB image, but it cannot boot the Angstrom image Dec 04 13:29:31 lrg: so I think there has always been something wrong with the Angstrom images I've generated Dec 04 13:29:41 vlo|work: I'm sure XorA has fixed this glibc issue now, it probably worth rebuilding from scratch Dec 04 13:31:35 lrg: Ok, let's go for a mtn pull and a nice complete rebuild .... ouch, my old proc is already scared :) Dec 04 13:35:48 ~lart objdump Dec 04 13:35:48 * ibot offers objdump some herring Dec 04 13:37:09 aaaah, herring Dec 04 13:37:55 morning Dec 04 13:38:03 hey chouimat Dec 04 13:38:10 Xora: what was this weird init problem lrg talked about ? Dec 04 13:39:21 koen http://ars.userfriendly.org/cartoons/?id=20061203 <--- why linux on a cell phone is not a good idea ;) Dec 04 13:39:43 vlo|work: no glibc in /lib Dec 04 13:41:10 ;DDD Dec 04 14:07:31 RP: do you have linux-rp_2.6.19.bb floating somewhere? Dec 04 14:19:18 I received a request from one of the main devs of FOSS in Japan about "Cleanly buildable revision information of dev tree" and "Cleanly buildable local.conf settings" Dec 04 14:19:48 I think he is frustrated by OE changing so fast and requiring PREFERRED_VERSION settings etc. Dec 04 14:20:01 depends on your DISTRO setting Dec 04 14:20:32 If I am correct, he would be looking for a "known good" combo of revision id and local.conf. I doubt there even is such a thing. Dec 04 14:20:39 Or is there? Dec 04 14:20:52 today, openzaurus-unstable Dec 04 14:20:58 today, Angstrom Dec 04 14:21:01 today angstrom-2007.1 Dec 04 14:21:15 today, generic Dec 04 14:22:20 I see. Would the answer be different tomorrow? I mean we have to admit that at times things break horribly and without warning. I understand this can be frustrating. Dec 04 14:23:28 * XorA does oz-unstable and generic builds every work day, and sees none of these catastrophics failures that you speak of Dec 04 14:23:29 Has OE been relatively stable lately? I think QA has paid off and I am not experiencing so many troubles lately. Dec 04 14:23:52 XorA: Not recently. Me neither. They have been there, though. Dec 04 14:24:22 Laibsch: you are just use to it, QA is as bad or as good as it was Dec 04 14:24:36 Laibsch: and the best thing to describe it is: "We have no idea" Dec 04 14:25:23 * koen builds angstrom from scratch every day and hasn't encountered catastrophic failures Dec 04 14:25:38 most of the problems newbs seem to have with breakages seems to stem from panic and the thought that .dev is somehow stable. Alerted on this channel during EU daytime and most stuff is fixed there and then Dec 04 14:26:23 and the inability to read the actual error message Dec 04 14:26:27 next year we'll start doing stable snapshots and things should be better Dec 04 14:26:29 like "Disk full" Dec 04 14:26:32 OK, so basically I will tell him "no such thing. Try latest tree. Most devs build daily w/o probs. If there are problems, they are usually dealt with quickly" Dec 04 14:27:15 if there are problems, report them to the bugtracker Dec 04 14:27:24 Laibsch: yes, and if he speaks even moderate english then he can come here, and we promise we will lease zecke :-) Dec 04 14:27:34 s/lease/leash/ Dec 04 14:27:52 koen: He is not an absolute newbie and has already contributed to OE BTS and thus OE itself. He is very knowledgeable but time-restraint. Of course, he is still at the low end of the OE learning curve. Dec 04 14:28:06 angstrom/.dev has one problem: few toolchains per week Dec 04 14:28:10 His English is way above normal, too. Dec 04 14:28:23 XorA: haha, just send me pizza and I'm quiet Dec 04 14:28:26 that's irrelevant for reporting bugs Dec 04 14:28:37 Laibsch: I personally feel there is no substitute for chatting here for getting the feel Dec 04 14:29:04 * koen likes rule #5 of http://www.angstrom-distribution.org/ Dec 04 14:29:41 XorA: I will mention that, too. But he is involved with lots of projects. So I am not sure if he can spare the time. Japan is +8 hours which does not help with all the European guys not waking up before 3PM ;-) Dec 04 14:29:48 koen: 1st rule of the bug tracker, the bugtracker does not exist? Dec 04 14:29:49 What are the differences between abiword and abiword embedded? Dec 04 14:30:09 I can't seem to find anything relevant on the net Dec 04 14:30:21 He uses SL-6000. Could that make a difference? It is not so wide-spread. Dec 04 14:31:01 B_Lizzard: less icons on toolbar in embedded Dec 04 14:31:15 Great, thanks a lot! Dec 04 14:31:36 koen: bug 1110 is in the BTS! ;-) Dec 04 14:34:34 koen: so is bug 1409. I think you will see these are high-quality bug reports from a knowledgeable person. Dec 04 14:36:36 btw - does 1409 still happen? we got rid of all dbus-glib providers other then dbus-glib Dec 04 14:36:49 it doesn't happen for angstrom Dec 04 14:37:11 I am not sure. I will make a note in the BTS. Dec 04 14:37:15 koen: angstrom has pref_prov Dec 04 14:37:24 Laibsch: I made one already Dec 04 14:37:26 hrw|work: right Dec 04 14:37:29 OK Dec 04 14:37:31 so it doesn't happen on angstrom Dec 04 14:40:26 I guess I should advise him to build for angstrom instead of openzaurus-unstable, right? Dec 04 14:41:25 yep Dec 04 14:52:21 hrw|work: Sadly I don't have a 2.6.19 kernel yet :-( Dec 04 14:52:24 when I have file://defconfig in my bb file, is it taken from "files" directory or from "bbfilename-version" directory? Dec 04 14:52:29 Anybody using qemu successfully with angstrom? That is what he will use at first. I don't want to promise it will work if there are problems. Dec 04 14:52:54 I mean the SRC_URI stuff Dec 04 14:53:50 Jin^eLD: . files/ PN/ PN-PV/ Dec 04 14:53:53 Jin^eLD: this order Dec 04 14:54:04 thx Dec 04 14:55:10 Laibsch: it works over here Dec 04 14:56:46 cool. Thanks for letting me know. Dec 04 15:06:18 Laibsch: FWIW, Poky should also always build and is easy to setup although it doesn't support tosa atm (but does support qemuarm+qemux86) Dec 04 15:15:44 OT: anyone know how to change widget Z order using X or Gtk? Dec 04 15:15:52 can't find anything in the API Dec 04 15:16:43 you mean 'raise' or 'lower' them? Dec 04 15:16:48 yeah Dec 04 15:16:52 e.g. Dec 04 15:16:54 i show widget A Dec 04 15:16:56 i show widget B Dec 04 15:17:03 B overlaps A and is on top Dec 04 15:17:06 but i want A to be on top Dec 04 15:17:13 always the one that's shown last is on top Dec 04 15:17:22 which is not what i want Dec 04 15:18:30 aah Dec 04 15:18:31 heh Dec 04 15:18:32 thanks Dec 04 15:18:34 gdk_window_raise Dec 04 15:18:35 etc. Dec 04 15:18:42 it's part of Gdk (of course) Dec 04 15:19:02 you have devhelp installed? Dec 04 15:22:20 no. guess i should Dec 04 15:27:06 hm Dec 04 15:27:23 is it possible to create for 1 distro/machine combo two images with different sizes? Dec 04 15:27:45 yeap Dec 04 15:27:46 easy Dec 04 15:27:46 i.e., a pivot_root image for use as initrd and an image for the flash? Dec 04 15:27:53 it's on the wiki Dec 04 15:27:59 i can't seem to find it Dec 04 15:28:19 but if you say it's there i'll look harder ;) Dec 04 15:28:33 2sec Dec 04 15:29:24 http://www.openembedded.org/user-manual&dpage=ch07s06 Dec 04 15:31:23 giel, Dec 04 15:31:54 thanks, but i think that's not exactly what i'm looking for Dec 04 15:32:34 ha Dec 04 15:32:47 if i type bb pivotboot-image i want an ext2 image of 8MB, and if i type bb foo-image i want one of 64MB, or something like that Dec 04 15:33:00 i use it to make 2 images : one for my nand & one for my nor flash Dec 04 15:34:00 ah but it's not the same image oki , i misunderstood , sorry Dec 04 15:34:12 np Dec 04 15:34:26 i'll try to figure something out Dec 04 15:36:56 !oebugs 1647 Dec 04 15:42:26 RP: Laibsch : Your ping is there for ages, copy DL_DIR into the DL_DIR of bittest and be done Dec 04 15:43:21 RP: Laibsch : Thanks to wget -c it will not download a single line of new code, for QA in theory this helps none, in practice it actually helps a bit... Dec 04 15:43:57 ping? Dec 04 15:44:16 * lrg has just realised he has been driving his car with no MOT for about 1 month now. Oops. Dec 04 15:44:30 checking for the status code of HTTP1.1 Dec 04 15:47:38 zecke: Are you referring to the bittest that I am running? Can I just symlink bittest DL_DIR to OE DL_DIR? Dec 04 15:47:53 Laibsch: not symlink, better copy Dec 04 15:48:33 Laibsch: and to say it this way "Ich habe Pferde kotzen sehen" and for a true QA you really to want to check the md5/sha sums Dec 04 15:48:39 where bittest it in svn? Dec 04 15:49:09 Laibsch: and at my two internships the first question regarding OE was. Can we forbid fetching from external trees? Dec 04 15:49:14 zecke: Sure. That is why I suggested splitting the test and doing the whole thing only from time to time. Dec 04 15:49:33 zecke: cool :) Dec 04 15:50:11 zecke: fetching from external trees? You mean, they were asking if they can circumvent GPL and use OE stuff but not publish their own work? Dec 04 15:50:11 So our policy was to download the tarball by hand, sha256 it, add the sum to the bb file, put the downloaded tar on the internal server Dec 04 15:50:35 Oh, I see Dec 04 15:50:39 Laibsch: no, this has nothing to do with licenses at all :) Dec 04 15:50:43 zecke: internal mirror ... nice and boring job ;) Dec 04 15:52:30 Genesis: dirty solution, but works: you can just put IMAGE_ROOTFS_SIZE_ext2 in the foo-image.bb file Dec 04 15:52:31 zecke: Is there anything in the works on how you want ppl help with automatic QA? Dec 04 15:52:55 Laibsch: sadly most stuff is varpoware :( Dec 04 15:53:06 Who is vaporizing? Dec 04 15:53:11 ACTION  Dec 04 15:53:43 crappit, power out Dec 04 15:54:16 Can you specify a) task to test and b) where to report results? I am not sure about this tinderbox and how get and interpret results from it. Dec 04 15:54:25 * chouimat wonders at which temperature zecke starts to vaporize :) Dec 04 15:54:41 Should one run the complete bittest weekly? Dec 04 15:54:53 chouimat: It is quite cold here in .de Dec 04 15:54:58 Laibsch: what we can do now: Send good reports to the tinderboxserver Dec 04 15:55:06 Laibsch: it still runs on ewi and receives build logs Dec 04 15:55:12 Laibsch: here it's a warm -5C Dec 04 15:55:15 most of sources are fetchable currently Dec 04 15:55:26 Laibsch: it can receive results of BitTest as well Dec 04 15:55:37 Laibsch: the representation of these results is quite bad though Dec 04 15:55:41 But where can I see the bittest results? Dec 04 15:56:23 Laibsch: but Building is only a minor part of QA, what we can't do now is create good reports, bitch people, and actually test the stuff that was built Dec 04 15:56:38 Laibsch: this depends on your cli (bittest --help) Dec 04 15:58:17 I hope you do not mind me differing. My feeling is you concentrate too much on the tool side (and you have done an awesome job!). What is more important IMHO is to get ppl use it and report stuff. So you need to define objectives, tasks, responsibilities. Dec 04 15:58:37 Like, we want to check availability of source once weekly. Dec 04 15:58:47 And hashes monthly Dec 04 15:58:59 For all of OE. Dec 04 15:59:09 Laibsch: right (tools side) Dec 04 15:59:27 Laibsch: the question is how to make people look at the results and act on errors Dec 04 15:59:43 Laibsch: and the best answer I have is we need a tool Dec 04 15:59:55 Answer: Not by making them invisible and ugly (hint, hint) Dec 04 15:59:56 zeckeBot which will be able to bitch about errors Dec 04 16:00:12 or community for a first go. Dec 04 16:00:13 zeckeBot: God damn, koen you broke OpenZaurus builds Dec 04 16:00:19 I could help. Dec 04 16:00:26 But I would not know what to do. Dec 04 16:00:39 There are lots others like me. Willing to contribute but no coders. Dec 04 16:00:59 Laibsch: e.g. what we can do now is look t build results Dec 04 16:01:02 Laibsch: e.g. http://ewi546.ewi.utwente.nl/tinderbox/showbuilds.pl?tree=poky Dec 04 16:01:03 Most of them will be put off by stuff not working. So QA benefits us doubly. Dec 04 16:01:20 Laibsch: poky fails for two reasons, one is missing setserial source Dec 04 16:01:48 I think there is more we can do. Dec 04 16:01:50 Laibsch: one is something weird with GnuTls. RP could fix it but we don't want to put too much preassure on his shoulders Dec 04 16:02:00 http://ewi546.ewi.utwente.nl/tinderbox/showbuilds.pl?tree=OpenZaurus-354x Dec 04 16:02:09 OZ fails due qemu on Ubuntu Edgy Dec 04 16:02:22 http://ewi546.ewi.utwente.nl/tinderbox/showbuilds.pl?tree=Angstrom Dec 04 16:02:24 angstrom rocks Dec 04 16:02:48 See what you are doing? Dec 04 16:02:52 no Dec 04 16:02:55 ah.. I forgot to push qemu... Dec 04 16:03:03 You are sharing your knowledge about how to use this tool. Dec 04 16:03:23 I think that is something that is more valuable ATM than creating an even better mousetrap. Dec 04 16:03:31 Laibsch: Green == Good Dec 04 16:03:33 Separate tool and process Dec 04 16:03:37 Laibsch: Yellow == In Progress Dec 04 16:03:43 Laibsch: Red == Bad -> click on it Dec 04 16:03:48 Well, I would not have even known where to look. Dec 04 16:03:53 Or what to test. Dec 04 16:04:03 Laibsch: a long ugly page will be shown Dec 04 16:04:11 I think there is value in you defining and guiding the process. Dec 04 16:04:14 Laibsch: on the upper side you will see wgat got built Dec 04 16:04:19 at the bottom you see the failure Dec 04 16:04:33 zecke: will push fix for oz354x Dec 04 16:04:33 be back in a minute Dec 04 16:04:36 :) Dec 04 16:08:57 * hvontres|poodle finally gets a clue as to what tinderbox does :) Dec 04 16:09:50 Laibsch: all the build results you see are generated from a dual xeon system Dec 04 16:09:58 Laibsch: which is still mostly idle Dec 04 16:10:01 zecke: fix pushed Dec 04 16:10:22 03hrw 07org.oe.oz354x * r5b114bed... 10/ (4 files in 3 dirs): qemu-native: fix problem with 2.6.17 linux-libc-headers on host Dec 04 16:11:36 zecke: So there is really no need for others to help with QA? I mean I am cool with that and I believe a central responsibility for QA is quite well. Dec 04 16:12:04 Laibsch: autobuilders like those run by zecke and nslu2 people are only part of QA rather Dec 04 16:12:06 Laibsch: oh sorry, there is enough to do and help is welcome Dec 04 16:12:31 Laibsch: as OE supports more than infinite configurations there more we build the better Dec 04 16:12:34 OK, then it would be a good idea to spell things out. Dec 04 16:13:08 * hrw|work started build for #1525 Dec 04 16:13:16 my first sh3 build Dec 04 16:13:44 Is it only building (inherit tinderbox) or also linting (man bittest)? What are the priorities. It seems that a (human) gateway from error to BTS might be something? Dec 04 16:14:15 Laibsch: human gateway is mostly missing Dec 04 16:14:36 Laibsch: but do not assign bugs to people, communicate with them :) Dec 04 16:14:51 Laibsch: and that is really difficult Dec 04 16:15:08 communicate with the bugs? Dec 04 16:15:12 http://tieguy.org/talks/LCA-2005-paper-html/ Dec 04 16:15:45 zecke: maybe we could put koen in charge of "communications" ;) Dec 04 16:16:00 zecke: But all the stuff from both bittest and bitbake are at http://ewi546.ewi.utwente.nl/tinderbox/showbuilds.pl?tree=OpenEmbeddedBuilds et al? Dec 04 16:16:01 hvontres|poodle: no no, I will handle it ;) Dec 04 16:16:19 Laibsch: This tbox has the concept of a 'Tree' Dec 04 16:16:47 Laibsch: and we have four tree's defined: OpenEmbeddedBuilds,Poky,OZ354x,Angstrom Dec 04 16:17:13 Laibsch: and you can send anything you want to the box Dec 04 16:17:27 build reports, bittest results, serial console logs of booted images... Dec 04 16:18:10 http://ewi546.ewi.utwente.nl/tinderbox/showbuilds.pl?tree=OpenEmbeddedBuilds Dec 04 16:18:12 I know I can send the stuff. But I can also send stuff to /dev/null ;-) Pulling stuff is where I am still learning. Especially pulling goal-oriented. Dec 04 16:18:39 Laibsch: That is easy: Everything must be GREEN Dec 04 16:18:47 Laibsch: if it is Yellow: Wait Dec 04 16:18:52 OK OK Dec 04 16:19:03 Laibsch: if it is Red: Check the log and find someone to solve it Dec 04 16:19:11 But is all the info the tbox has in one of those links? Dec 04 16:19:30 Laibsch: I'm tempted to say yes Dec 04 16:19:43 Then I would devote some time in the future to look through the links ever so often. Dec 04 16:20:32 I stopped sending stuff since I was having REDs but the problem was local IOW misconfiguration. Dec 04 16:20:50 Separating that cruft out would be nice. Dec 04 16:22:46 BTW, the tieguy.org stuff sounds like what I have been trying to do to the best of my ability and which I plan to continue doing in the future. Dec 04 16:23:01 Luis rocks :) Dec 04 16:23:13 last time I read this, I agreed Dec 04 16:24:07 zecke: I have small diff to tinderclient.bbclass Dec 04 16:24:24 + server = data.getVar('TINDER_HOST', d, True ) or "ewi546.ewi.utwente.nl" Dec 04 16:24:24 + url = data.getVar('TINDER_URL', d, True ) or "/tinderbox" Dec 04 16:24:48 okay... Dec 04 16:24:56 zecke: what do you think about setting most of TINDER_ vars to some defaults? Dec 04 16:25:04 will pastebin full diff when finish Dec 04 16:25:22 okay, will go home now and into the gym... Dec 04 16:35:00 cu Dec 04 16:39:03 !oebug 1525 Dec 04 16:39:04 * * Bug 1525, Status: NEW, Created: 2006-10-23 12:16 Dec 04 16:39:05 * * kristoffer_e1(AT)hotmail.com: Crosscompiling GCC 4.1.1 fails at end Dec 04 16:39:06 * * http://bugs.openembedded.org/show_bug.cgi?id=1525 Dec 04 16:40:10 :-) Dec 04 16:40:27 :D Dec 04 16:40:29 !oebug florian Dec 04 16:40:31 * * Bug florian, Status: InvalidBugId Dec 04 16:40:51 *gg* Dec 04 16:41:01 * florian_kc is too big for a bug ;) Dec 04 16:45:14 bye Dec 04 16:46:19 florian_kc: not 400 millions ago :) Dec 04 16:46:35 chouimat: hehe Dec 04 17:20:37 03kristoffer 07org.oe.dev * r87eec792... 10/ (1 conf/distro/jlime-donkey.conf): Dec 04 17:20:37 conf/distro/jlime-donkey.conf : Added parted & Dialog to distro_rdepends Dec 04 17:20:37 * Added parted & dialog programs to distro_rdepends since Dec 04 17:20:37 they are needed for installer. Dec 04 17:25:04 hi cbrake how his the .il stuff going ;-) Dec 04 17:26:24 rob_w: pretty well. No bit problems so far. Dec 04 17:26:50 mine was now replaced because of that crc error on loading the kernel Dec 04 17:27:21 rob_w: I had to enable kernel preemption due to the noddy NAND flash driver sucking resources, and set a critical process to RT. Dec 04 17:27:51 rob_w: ouch, I sure hope this is not a pattern as we had to send 1 of our systems back as well. Dec 04 17:27:53 interesting Dec 04 17:28:11 rob_w: kernel preemption broke the SD driver, so now that needs fixed. Dec 04 17:29:03 i also had strange sighitings on using preempt Dec 04 17:30:00 re Dec 04 17:30:13 wb Dec 04 17:31:23 hmm, does anyone know how to diff, ignoring certain changes? the -I regexp option did not seem to do anything at all for me Dec 04 17:32:10 this really drives me crazy, I am trying to make a clean kernel patch from a vendor's version, and they checked the kernel into their CVS which somehow messed up $Revision tags, but only on certain files Dec 04 17:32:34 Jin^eLD: do diff and then clean it? Dec 04 17:33:00 Jin^eLD: diff --exclude Dec 04 17:33:11 hrw: well, the kernel is a "work in progress", so I may have to repeat the process when updates come Dec 04 17:33:20 cbrake: --exclude only excludes certain files, afaik? Dec 04 17:33:54 I'm more looking for something "if result of diff is only a change in $Revision" - ignore it Dec 04 17:34:05 Jin^eLD: ahh, revision tags -- what a pain :-\ Dec 04 17:34:10 yeah.. Dec 04 17:34:47 $Revision, $Id, several things messed up Dec 04 17:35:08 Jin^eLD: do diff and clean it with script Dec 04 17:37:18 that's what I am trying now.. I just hoped there would be an easier way Dec 04 17:38:02 -I RE --ignore-matching-lines=RE Ignore changes whose lines all match RE. Dec 04 17:38:10 yep, tried that - had no success at all Dec 04 17:38:52 I have no idea why but it seemed to completely ignore my RE Dec 04 17:38:57 ;( Dec 04 17:41:31 19:06 hrw@home:tmp$ diff -u 1 2 --ignore-matching-lines="Revision" Dec 04 17:41:43 it ignore Revision lines Dec 04 17:41:47 hmm Dec 04 17:41:51 I did not have the 1 2 thing Dec 04 17:41:58 maybe that was my mistake Dec 04 17:42:07 oh wait Dec 04 17:42:09 1 2 are files Dec 04 17:42:13 :) Dec 04 17:42:20 I used -Naur Dec 04 17:43:03 and it did not do anything Dec 04 17:43:09 I mean - in regard to -I Dec 04 17:43:59 hmmm Dec 04 17:44:08 the -I comes after the files Dec 04 17:44:11 that must be it Dec 04 17:44:15 then it works Dec 04 17:44:20 oh man Dec 04 17:44:34 if size matters for some of you: http://www.geek-fr.com/blog/what_size_is_yours Dec 04 17:45:18 Jin^eLD: works? Dec 04 17:45:50 nicolasfr: mine wifi card is smaller then smallest of yours Dec 04 17:45:55 hrw: testing... Dec 04 17:46:27 no Dec 04 17:46:48 hrw: argh! I want it then ;) Dec 04 17:47:02 hrw: you can't post picture? Dec 04 17:47:14 when I tested a short example, it worked Dec 04 17:47:20 but on the actual files it failed again, I have no idea why Dec 04 17:47:34 let's say I want to ignore that: *! Import of linux kernel 2.6.16.19 Dec 04 17:48:02 I do --ignore-matching-lines="Import of linux kernel", right? or just "Import" Dec 04 17:48:12 but it still ends up in the diff Dec 04 17:48:59 ok I think thats the trick - it only ignores stuff that is different, not stuff that is new Dec 04 17:49:30 hrw: thank you! you helped me with a little breaktrhough Dec 04 17:49:35 Jin^eLD: Time to break out sed ? Dec 04 17:49:46 nicolasfr: http://ewi546.ewi.utwente.nl/tmp/hrw/t/DSC00106.JPG Dec 04 17:49:47 hvontres|poodle: I think I'll need that in addition... Dec 04 17:50:58 nicolasfr: small enough? Dec 04 17:51:29 bb in 10-15 minutes Dec 04 17:51:40 hrw : cool... how is the driver coming along? and how much will one of those set me back ? Dec 04 17:51:45 time to relocate my body... l8r Dec 04 17:52:36 hvontres|poodle: do not know how much they will cost. driver will be when it will be ready Dec 04 17:52:39 afk Dec 04 17:58:59 * mrz80 is away: Metaphysically I'm still here; physically, I'm elsewhere Dec 04 17:59:33 hrw: http://www.geek-fr.com/_detail/blog/ambicom-wl1100c-cf.jpg Dec 04 17:59:45 hrw: comparable! Dec 04 18:02:54 i'm trying to get a few extra packages in a bootstrap-image using a task-base configured machine. If I define MACHINE_EXTRA_RDEPENDS = "gtk+", and then run bitbake bootstrap-image, all dependencies get built, but gtk and friends do not end up in the image?? Dec 04 18:04:15 any ideas what I might be doing wrong. Dec 04 18:12:57 cbrake, strange .. Dec 04 18:14:09 * cbrake unpacks task-base ipk ... Dec 04 18:15:18 hey cbrake i just noticed that we could make the worlds first upgradeable pda, ".. wanna more ram , no problem " Dec 04 18:16:01 rob_w: how is that? Dec 04 18:16:50 say u get the standard with 64 nand 64 ram 300mhz then you want more and get the 512 /128 520mhz Dec 04 18:18:17 rob_w: heh, yeah its nice to have upgrade options :-) Dec 04 18:18:42 rob_w: how you want to provide upgrades? Dec 04 18:18:55 rob_w: send old, get new? Dec 04 18:19:29 pay the new , keep ur old ;-) Dec 04 18:19:51 ahh, just figured out that I need to clean task-base timestamps if you change MACHINE_EXTRA_RDEPENDS to force the ipk to get rebuilt -- makes sense. Dec 04 18:20:34 cbrake: bitbake task-base -crebuild Dec 04 18:20:37 easier Dec 04 18:20:45 just don't try "clean task-base". It seems to gleefully start deleting your build. Dec 04 18:21:00 hrw: ahh, excellent. I think its time for a new faq entry ... Dec 04 18:21:12 * hrw have to go Dec 04 18:21:17 hrw: later Dec 04 18:21:21 have a nice evening Dec 04 18:27:49 hello Dec 04 18:27:59 umi tu nekdo cesky? Dec 04 18:31:05 i search the os , wroted in python or assembler Dec 04 18:31:43 i search the os , wroted in python or assembler Dec 04 18:31:43 i search the os , wroted in python or assembler Dec 04 18:31:43 i search the os , wroted in python or assembler Dec 04 18:31:44 i search the os , wroted in python or assembler Dec 04 18:31:45 i search the os , wroted in python or assembler Dec 04 18:31:46 i search the os , wroted in python or assembler Dec 04 18:31:57 wroted? Dec 04 18:32:27 i a lite eadglish Dec 04 18:32:30 i czech Dec 04 18:32:48 write Dec 04 18:33:38 hi Dec 04 18:33:48 does the "${@[ " notation turn on python inline evaluation? Dec 04 18:34:30 Sleep|Palm: ty jsi czech? Dec 04 18:34:42 HopsNBarley: i dont understant Dec 04 18:35:20 jsem,snad nevadi Dec 04 18:36:03 it's in the slugos-image.bb file. Dec 04 18:36:52 Sleep|Palm: ne nevadi !ja taky. prosim te mohl bys mi prekladat to co mi pisou? neumim moc ang. Dec 04 18:38:20 tomeff: spis ne,jedu z palmu Dec 04 18:38:39 Sleep|Palm: aha Dec 04 18:39:18 Sleep|Palm: pls. jen ze hledam neakej os napsanej v pythonu Dec 04 18:39:57 wth? Dec 04 18:40:07 are your keyboards broken? Dec 04 18:40:19 nobody expects the czech inquisition! Dec 04 18:40:50 Sleep|Palm: pls. jen ze hledam neakej os napsanej v pythonu Dec 04 18:40:51 Sleep|Palm: pls. jen ze hledam neakej os napsanej v pythonu Dec 04 18:40:52 Sleep|Palm: pls. jen ze hledam neakej os napsanej v pythonu Dec 04 18:41:27 flood Dec 04 18:41:31 slaughter Dec 04 18:41:35 kickban! Dec 04 18:41:37 :p Dec 04 18:42:00 aha Dec 04 18:42:00 to mi nic nerika Dec 04 18:42:02 ale v pascalu uz jsem videl Dec 04 18:42:20 * Marex taky Dec 04 18:42:26 * Marex dokonce jeden skoro napsal Dec 04 18:42:37 ja ted newm presne co to bylo za jazyk ale vim na 100% ze je to na P Dec 04 18:42:54 Marex: nechces skusit v PHP :-D Dec 04 18:43:01 this is an english channel Dec 04 18:43:07 Perl? Dec 04 18:43:08 Sleep|Palm, j/w how'd you manage to write that "pipe" | on palm < Dec 04 18:43:14 behave, or find your own channel to talk gibberish in Dec 04 18:43:25 Sleep|Palm: mozdna Dec 04 18:44:05 sorry for czech,I'm already gone ;) Dec 04 18:44:57 Only official languages here are swedish and english Dec 04 18:44:59 in that order Dec 04 18:45:33 svenska? Dec 04 18:45:55 * lrg -> pub Dec 04 18:46:01 cu Dec 04 18:46:12 ああっ女神さまっ :-D Dec 04 18:47:40 I'm here again Dec 04 18:47:53 sorry for czech inquisition here :D Dec 04 18:49:13 Jag är ledsen. Dec 04 18:49:20 :D Dec 04 18:49:22 koen, yes svenska :D Dec 04 18:50:15 that was for another, more friendly beginning here ;) Dec 04 18:53:31 re Dec 04 18:54:49 mickey|dinner: Mahlzeit :) Dec 04 18:59:03 03kristoffer 07org.oe.dev * ra2c70fdc... 10/ (4 files in 2 dirs): Dec 04 18:59:03 initscripts/initscripts-1.0/jornada6xx/checkroot.sh : Dec 04 18:59:03 * We dont want fsck on boot since our clock resets Dec 04 18:59:03 on every reboot. Dec 04 19:18:14 does anyone have the USB gadget support for the kernel working with the PXA270? Dec 04 19:19:26 morning all Dec 04 19:21:00 hi Dec 04 19:37:34 morning all Dec 04 19:37:53 HopsNBarley: hi there Dec 04 19:38:05 howdy to you! Dec 04 19:38:44 likewise, our iomega storcenter download is now available, if you had any interest in that. Dec 04 19:39:02 HopsNBarley: thank you kindly, yes, I could learn from that :-) Dec 04 20:01:44 hrw|gone: which bitbake are you using? Dec 04 20:04:11 has anyone noticed memory leaks in taglib? Dec 04 20:09:57 Jin^eLD: not using it, where is it used? Dec 04 20:09:59 koen: hi Dec 04 20:10:54 likewise: for parsing mp3/ogg/ape tags Dec 04 20:10:56 hey likewise Dec 04 20:11:42 likewise: If you work with icecc get the lastest version from the kde.org svn Dec 04 20:12:03 likewise: packaged ones with debian/ubuntu seem to have problems Dec 04 20:12:13 Ifaistos: that one works? Dec 04 20:12:53 Ifaistos: what is broken? I got icecc so far that my cross toolchain were tarred and sent to remote hosts, extracted there, but failed during execution. Dec 04 20:12:57 likewise: i am able to do simple compiles (not OE) with 3 machines 2 i686 and 1amd64 Dec 04 20:13:28 Ifaistos: could you quote me the revision number of your subversion checkout? (svn info in the dir, if I am not mistaken) Dec 04 20:15:20 likewise: i found out that icecc is very sensitive to version numbers.... The packaged version with debian etch and ubuntu edgy were misreporting cpu usage at the i685 clients and because of this only the x86_64 host was used Dec 04 20:16:48 likewise: its also importand to set the ICECC_VERSION corectly for every arch i.e i686:package_with_enviroment,x86_64:oackage etc Dec 04 20:17:44 Ifaistos: OK, I am not sure if icecc.bbclass does that right, but I think it assumes only a single arch. Dec 04 20:18:15 likewise: One thing i have not solved... is that it "chokes" if -j is set.... not sure why though Dec 04 20:18:30 -j to the Makefile options? Dec 04 20:18:36 likewise: yes Dec 04 20:18:44 koen, Hi! Dec 04 20:18:56 hey psokolovsky Dec 04 20:19:07 psokolovsky: did you get my mail about the n30 stuff? Dec 04 20:19:08 koen, would like to discuss couple of issues with you Dec 04 20:19:27 sure, go ahead Dec 04 20:19:40 likewise: it seems that if you build the cross-compiler for i686 arch then x86_64 can also run it (at least in ubuntu/amd64) Dec 04 20:19:42 koen, Yes, I got, good news! I didn't get any response to device donation proposal yet though. Dec 04 20:20:09 koen, What about upgrading linux-handhelds-2.6 version in use by Angstrom? Dec 04 20:20:21 likewise: if you specify it at ICECC_VERSION Dec 04 20:20:31 psokolovsky: to which version? Dec 04 20:20:40 koen, recenently, CVS was upgraded to 2.6.19, and well, .18 & .18 is theer for long time ;-) Dec 04 20:21:01 last I heard .18 doesn't work for h2200 Dec 04 20:21:19 koen, that's good q, I'd say follwoing: if you're comfortable to use 2.6.18 for htcuniversal, then it should be good for most other machines as well ;-) Dec 04 20:21:29 koen, it was fixed and tested Dec 04 20:21:49 latest of .18 is -hh2 (tagged before upgarde to .19) Dec 04 20:22:19 what would angstrom gain with moving to .18? Dec 04 20:22:58 koen, either way, .16 is plain too old. hx4700, h4000 undergone more or less good testing for hx4700, h4000, so we really should just use it, and allow for even more testing Dec 04 20:23:17 'more or less'? Dec 04 20:23:23 likewise: Here is a part of the log while building samba (not through OE) -> http://pastebin.ca/267387 Dec 04 20:23:28 I want something a bit more tangible Dec 04 20:23:44 koen, there's work happening on kernel all teh time and .16 is just too old now. Dec 04 20:24:08 koen@h2200:~$ uptime Dec 04 20:24:08 20:48:56 up 20 days, 5:54, 3 users, load average: 1.10, 1.12, 1.09 Dec 04 20:24:18 .16 isn't broken, so why move? Dec 04 20:24:19 Ifaistos: yes, that's what I recognize. I did have OE/icecc work for non-cross stuff already. Dec 04 20:24:25 koen, I usually post things which changed in each -hh release in commit msgs, you can have a look there. Dec 04 20:24:45 psokolovsky: I'm still banned from using any hh.org service, remember? Dec 04 20:25:02 likewise: Do you get a problem with -j ? Dec 04 20:25:04 koen, I mean, to OE.dev commits Dec 04 20:25:24 psokolovsky: that still doesn't give me tangible test results Dec 04 20:25:43 Ifaistos: no, it worked for me Dec 04 20:25:54 psokolovsky: upgrading because of OMG SHINY!!!! is nonsense Dec 04 20:26:51 koen, interesting idea ;-). but HH.org kernels are work in progress, and each tag brings something new. But well, if you want 1 good reason to move to 2.6.16-hh8, or 2.6.17, or 2.6.18 is - common buttonmap refactor is there, which makes sure that all buttons on all suported devices are actually supported. Dec 04 20:27:13 koen, what test results are you looking for? Dec 04 20:28:33 psokolovsky: someone testing basic stuff like suspend/resume, battery life, mmc/cf/etc, battery reporting and sending a summary to the mailinglist Dec 04 20:29:21 koen, wow, I do that all the time here on the devices I have! I don't tag -hh without such basic testing! Dec 04 20:29:42 koen, but I got idea, before switching to .17 or .18 you want RFC on mailing list. ok. Dec 04 20:30:06 koen, I'd still want to settle switching to common-buttonmap kernel, at least 2.6.16-hh8 Dec 04 20:30:30 koen, any concerns with that? that would require using svn version of xserver-common Dec 04 20:30:36 (with patch) Dec 04 20:30:58 poky florian to make a release :) Dec 04 20:32:00 koen, would this be prerequisite for switch? or maybe we switch, and I use the argument that it was sufficiently tested in Angstrom snapshot to push it upstream? ;-) Dec 04 20:33:01 psokolovsky: consensus on the angstrom devel mailinglist would be a good start Dec 04 20:33:15 koen, ok, will do. thanks Dec 04 20:33:23 * koen heads back to the tv Dec 04 20:38:24 ooooooooooooooooohhhhhh no, I have this weird gcc cross-cross-initial problem again with today update: /home/valentin/EPFL/iMXBoard/linux/openembedded/build/mx31ADS/conf/tmp/cross/arm-angstrom-linux-gnueabi/bin/as: unrecognized option `-Qy' Dec 04 20:45:03 can anyone recommend a light xterm like program for OE base devices (GTK based preferred)? Dec 04 21:00:34 Any good way to follow a file in the repository? changes and such? Dec 04 21:01:30 svn log filename? Dec 04 21:01:49 groovey thx Dec 04 21:01:49 oh, you mean with monotone? Dec 04 21:01:52 yeah Dec 04 21:01:57 uhm :) Dec 04 21:02:51 I see mtn has some "log" option as well Dec 04 21:03:11 so mt log Dec 04 21:03:13 then? Dec 04 21:03:32 you could try it.. unfortunately the only mtn commands I used were pull and update :) Dec 04 21:03:41 I am not familiar with it... Dec 04 21:03:53 :) I'll give that a try Dec 04 21:04:55 tell me if it worked :> may be useful to know Dec 04 21:06:20 Kristoffer: mtn log --diffs packages/linux/ixp4xx-kernel_2.6.19.bb Dec 04 21:06:33 likewise, thx Dec 04 21:11:30 hmm building .oz354x gpe-image i get -- No file to patch. Skipping patch. Dec 04 21:11:30 1 out of 1 hunk ignored Dec 04 21:11:30 Patch poodle-asound.state.patch does not apply (enforce with -f) -- known issue by any chance? Dec 04 21:12:32 indeed i don't see the files it want's to patch in the tree anyplace Dec 04 21:40:49 hi florian__ Dec 04 21:46:39 gremlin[it]: know beer wuhrer? ;-> Dec 04 21:49:55 hi univac ... no don't know such beer Dec 04 21:50:37 i drunk it, it's from italy Dec 04 21:50:41 hrhr Dec 04 21:51:10 wuhrer ... a close to .at place ... isn't an italian name ;) Dec 04 21:52:10 yep] Dec 04 22:03:06 re Dec 04 22:03:11 hey pb_ Dec 04 22:27:55 hi people, should I add one to the PR after changing some h1940 specific stuff in the tslib .bb? Dec 04 22:28:03 http://bbrv.blogspot.com/2006/12/efika-power.html Dec 04 22:29:00 ArjanS, is change would affect how .ipk is generated, yes Dec 04 22:29:29 psokolovsky: it does, but only for h1940 builds. I'm removing the detect-stylus dependency Dec 04 22:29:59 ArjanS, if it affects build for one machine, it affects build at all ;-) Dec 04 22:30:16 ok, thanks, I'll bump it :) Dec 04 22:30:17 so yes, please bump, so it was propogated everywhere Dec 04 22:43:26 * mickeyl reads http://bugzilla.gnome.org/show_bug.cgi?id=172535 Dec 04 22:43:32 interesting stuff coming up in Gtk 2.12 Dec 04 22:43:45 (hopefully Dec 04 22:43:46 ) Dec 04 22:44:53 mickeyl: lets hope so :) Dec 04 22:50:43 03arjan 07org.oe.dev * rff1adcdb... 10/ (3 files in 3 dirs): tslib: remove h1940 machine dependency on detect-stylus and switch to using the default /dev/input/touchscreen0 created by udev Dec 04 23:21:11 RP: mickeyl suggested that you can perhaps answer me a short question about arm assembly. I need to write 0x12345678 at the address 0xA0000000. Dec 04 23:27:06 hello Dec 04 23:27:13 sorry to bother Dec 04 23:27:31 stefan_schmidt: adr r0, foo \n mov r1, #0xa0000000 \n str r0, [r1] \n foo: .word 0x12345678 Dec 04 23:27:34 ned some help crosscompiling to arm Dec 04 23:27:39 fro intel Dec 04 23:27:46 from intel@gentoo Dec 04 23:28:45 anyone? Dec 04 23:29:07 mreimer: Thanks a lot. I start with this. Dec 04 23:30:24 denis: using OE? Dec 04 23:30:31 what is your target (arm) platform? Dec 04 23:30:38 not yet, I have a zipit Dec 04 23:31:14 i put a 64mb SD card on it, so now I want to get all I can on it Dec 04 23:31:28 netcat, freebase, mp4players, etc Dec 04 23:31:31 mp3 Dec 04 23:31:59 I hear that OE has a lot of utilities than I could use Dec 04 23:32:52 I downloaded and compiled buildroot Dec 04 23:33:31 denis_, they might be more knowledgable about buildroot in #elinux also Dec 04 23:33:45 oh, thanxs Dec 04 23:34:43 #edev actually Dec 04 23:35:12 I get those two confused Dec 04 23:35:21 :P thx Dec 04 23:35:31 what exactly is the difference? Dec 04 23:35:41 the same guys seem to hang out in both Dec 04 23:35:42 the people who still flog buildroot only listen in #edev Dec 04 23:35:51 :) Dec 04 23:36:37 weird, I met a guy from Aeronix a week ago Dec 04 23:36:45 well a few weeks back Dec 04 23:38:01 they don't seem like a company that would do consumer junk ... Dec 04 23:39:04 absolutely weird Dec 04 23:39:14 don't call my zipit junk :P Dec 04 23:39:21 :) Dec 04 23:39:31 http://www.aeronix.com/products Dec 04 23:40:10 lots of folks in #edev hack on zipit Dec 04 23:40:21 that and the juicebox Dec 04 23:41:22 FYI: zipit internal photos: http://www.nslu2-linux.org/gallery/zipit Dec 04 23:41:58 I've just opened mine and soldered a sd card, what a pain........ Dec 04 23:42:48 stefan_schmidt: I think you have your answer :) Dec 04 23:44:16 RP: Yes. I just have second short one. :) Is there a command to power off the device? Dec 04 23:45:29 stefan_schmidt: suspend (apm -s) or halt (shutdown -h)? :) (also, you can use sysrq) Dec 04 23:46:13 RP: No system running. I only execute the binary from the bootloader. Dec 04 23:47:06 stefan_schmidt: Its normally hardware specific - some CPU instuction or manipulation of a GPIO line Dec 04 23:47:34 RP: Nevermind. Power off by the power swicth should be fine for this hack. Thanks. Dec 04 23:48:46 psokolovsky: is 2.6.19 resuming ok on your hx4700? Dec 04 23:49:21 mreimer, sure. on bare kernel. no, on opie from fam0.8.4 Dec 04 23:50:30 'night all Dec 04 23:51:28 good night RP Dec 04 23:51:52 psokolovsky: I'm on GPE. I'll try without, but it shouldn't make a difference. I'm using your static defconfig Dec 04 23:52:32 mreimer, For me, fam0.8.4 gpe doesn't boot at all. what build do you use? Dec 04 23:52:47 psokolovsky: .dev from a couple of months ago Dec 04 23:53:24 mreimer, wow ;-). I built current angstrom snaphost, but didn't yet have chance to try it. Dec 04 23:53:34 mreimer, I guess fam is finally too old for .19. Dec 04 23:53:58 mreimer, no explaination why care kernel resumes, and with userspace - no, so far ;-( Dec 04 23:54:03 psokolovsky: but what could it be? userland shouldn't be affected. my serial console isn't even resuming Dec 04 23:54:45 mreimer, if I new... maybe some broken module is still loaded by userspace, something like that %) Dec 04 23:54:54 all I had loaded was ipv6 Dec 04 23:54:57 we hsould haev asked how it works for RP ;-) Dec 04 23:55:33 so, it's weird, I agree ;-( Dec 04 23:58:11 psokolovsky: even with nothing running in userland (except portmap and a couple of others) it doesn't work Dec 04 23:58:20 psokolovsky: I wonder if the pm.c changes got messed up again Dec 04 23:59:00 mreimer, worked for me yesterday. can try now again. Dec 04 23:59:36 psokolovsky: this is HEAD with your static defconfig with gcc 3.4.1. I even deleted the build directory Dec 05 00:00:14 mreimer, what do you use as initrd? Dec 05 00:00:33 psokolovsky: I boot from CF Dec 05 00:00:52 mreimer, do you use bootstrap image? Dec 05 00:00:58 so actually it's your static config plus I made a few more modules built-in like ide-cs Dec 05 00:01:27 psokolovsky: no, I start from haret in wm2003 without initrd, booting right from /dev/hda2 Dec 05 00:01:44 mreimer, you may not believe, but userspace matters ;-I Dec 05 00:01:55 mreimer, what's exactly on /dev/hda2? that gpe image? Dec 05 00:02:05 psokolovsky: ext2 gpe image Dec 05 00:02:18 psokolovsky: I'll start removing drivers Dec 05 00:02:24 mreimer, wanna try initrd from here: http://celib.sourceforge.net/linux-ppc/ Dec 05 00:02:49 psokolovsky: yes, I'll give that a try in a few minutes. Thanks for your help Dec 05 00:02:52 mreimer, it's magic inird from ancient times, which works with anything ;-I Dec 05 00:02:56 :-) Dec 05 00:03:16 mreimer, well, let's have that settled, I'm not less puzzled than you. Dec 05 00:03:24 psokolovsky: yes, let's Dec 05 00:03:38 mreimer, or, if you consider that all too boring, there's 2.6.18 branch ;-) Dec 05 00:03:58 psokolovsky: is HEAD more stable than .19? Dec 05 00:04:12 mreimer, HEAD is .19 ;-) Dec 05 00:04:20 psokolovsky: heh Dec 05 00:04:28 psokolovsky: I mean, is .18 more stable than HEAD/.19 Dec 05 00:05:00 mreimer, yes, sure it is. .18 is almost month old, .19 is 3 days ;-) Dec 05 00:05:10 but is anyone working on .18? Dec 05 00:06:40 mreimer, I guess, the answer is no. I do a bit though - eleaborating h4000 in .18, and fighting .19 issues ;-) Dec 05 00:06:55 psokolovsky: I'm guessing the same problem is in .18 Dec 05 00:07:39 mreimer, nack. .18 works pretty fine with fam0.8.4 gpe. Dec 05 00:08:31 psokolovsky: hmm. I haven't had any version newer than .15 suspend/resume for me reliably Dec 05 00:09:07 mreimer, any stats? Dec 05 00:09:26 psokolovsky: no, I didn't record anything. Bad Matt. :-( Dec 05 00:09:28 lol. I just suspended my desktop machine instead of hx4700 ;-) Dec 05 00:10:33 psokolovsky: :-D Dec 05 00:10:34 mreimer, damn. confirmed hx4700 with hx4700_static_defocnfig doesn't resume ;-\ Dec 05 00:10:42 psokolovsky: at least it's reproducible Dec 05 00:10:47 I'll check pm.c Dec 05 00:10:58 it'll be so good to get mainstreamed Dec 05 00:11:46 psokolovsky: was that with your initrd? or should I still try it? Dec 05 00:11:57 mreimer, yes, that was that initrd Dec 05 00:12:10 ok Dec 05 00:25:20 psokolovsky: one problem may be that PXA_SUSPEND_SAVE_EXTRA_REGS is missing from mach-pxa/Kconfig vs 2.6.15 Dec 05 00:25:32 psokolovsky: so some extra code for hx4700 doesn't get assembled in sleep.S Dec 05 00:25:48 mreimer, I dodn't remember hx4700 needing it Dec 05 00:26:09 I think pH5 added it for the case where we boot from haret Dec 05 00:26:53 psokolovsky: http://handhelds.org/cgi-bin/cvsweb.cgi/linux/kernel26/arch/arm/mach-pxa/sleep.S.diff?r1=1.12&r2=1.13 Dec 05 00:27:39 psokolovsky: "save CCCR register on the stack during suspend to work around the magician bootloader overwriting it during resume" Dec 05 00:28:08 mreimer, so, where's issue? Dec 05 00:28:21 psokolovsky: CCCR isn't getting saved/restored Dec 05 00:28:35 mreimer, why? and how this related to hx4700? Dec 05 00:28:59 psokolovsky: maybe the htc bootloader isn't setting it properly? Dec 05 00:29:46 mreimer, I sware resume worked in -hh1. some mindless commits broke it - that's what I think Dec 05 00:30:19 psokolovsky: you may be right Dec 05 00:36:50 psokolovsky: _EXTRA_REGS didn't help Dec 05 00:51:08 mreimer, ok, h4000 does resume with CVS HEAD, hx4700 does not Dec 05 00:51:19 now I wonder, If I test hx4700 at all actuall... Dec 05 00:51:35 bd2 told it works, that's all I'm sure in 100% Dec 05 00:51:56 ok Dec 05 00:54:27 psokolovsky: a kernel built without hx4700_lcd + w100fb resumes Dec 05 00:54:33 and very fast :-) Dec 05 00:54:49 or maybe it was because I didn't run Xw100. I'll narrow that down Dec 05 00:55:28 mreimer, good work! Dec 05 00:56:27 psokolovsky: I'd guess that the platform_driver conversion was not complete Dec 05 00:56:42 psokolovsky: what does h4000 use for video? Dec 05 00:56:50 mreimer, pxafb Dec 05 01:00:28 psokolovsky: the second hunk looks wrong: http://handhelds.org/cgi-bin/cvsweb.cgi/linux/kernel26/drivers/video/w100fb.c.diff?r1=1.25&r2=1.26 Dec 05 01:00:53 mreimer, see next commit Dec 05 01:00:58 ah yes Dec 05 01:03:04 mreimer, I have strng suspicion that the whole that commit is wrong. Dec 05 01:03:52 psokolovsky: I'm wondering if there's a mixup between platform_data and driver_data Dec 05 01:05:06 mreimer, let's revert that commit locally and test Dec 05 01:05:10 psokolovsky: ok Dec 05 01:06:23 psokolovsky: 2.6.15 is already a platform_driver Dec 05 01:07:31 mreimer, of course. just another bullshit commit from spyro. you see, I'm really tired of this crap. Dec 05 01:09:05 mreimer, we need to setup commit mails with diffs asap. or this firefighting not going to stop. Dec 05 01:09:07 psokolovsky: he has made quite a few mistakes Dec 05 01:09:17 psokolovsky: good idea Dec 05 01:09:17 yep Dec 05 01:09:30 it will be so nice to get mainstreamed and be using git Dec 05 01:10:17 mreimer, yes. let's now hope that spyro will be the one who will bother with git (at least he said that he's working on that) Dec 05 01:11:32 psokolovsky: rev 1.25 doesn't fix it Dec 05 01:11:34 mreimer, reverting those changes didn't help here ;-( Dec 05 01:11:40 yep ;-( Dec 05 01:12:29 maybe w100fb.h changes Dec 05 01:14:15 no changes there Dec 05 01:14:27 I mean rev 1.9 -> 1.10 Dec 05 01:16:50 psokolovsky: nope, that didn't help Dec 05 01:17:16 mreimer, I sign off of my changes ;-) Dec 05 01:17:28 psokolovsky: :-) Dec 05 01:17:31 mreimer, but are you sure w100fb.c is culprit? Dec 05 01:17:50 psokolovsky: good point. I'll try adding just w100fb and no hx4700_lcd Dec 05 01:18:26 ok. I'm doing vice-versa ;-) Dec 05 01:19:26 two-man bisect Dec 05 01:21:02 heh, it doesn't boot here in such way Dec 05 01:21:20 yup Dec 05 01:24:33 mreimer, nack regarding resume w/o w100fb & hx4700_lcd ;-( Dec 05 01:25:35 psokolovsky: try adding PXA_SUSPEND_SAVE_EXTRA_REGS=y to your .config Dec 05 01:25:46 psokolovsky: and make sure sleep.S gets recompiled Dec 05 01:29:26 psokolovsky: do you see any point to platform_driver_register(&hx4700_lcd_driver ) ? we could just call lcd_hw_init() in hx4700_lcd_init(). Dec 05 01:30:26 mreimer, well, I don't know. We could have just monolitic code at all ;-). I think that should be decided on cross-port level... Dec 05 01:30:57 psokolovsky: I'm saying that because it only does lcd_hw_init() on probe, and doesn't do anything else, even in remove() Dec 05 01:31:43 psokolovsky: of course the other two drivers are essential Dec 05 01:31:50 mreimer, I know that there're quite a few drivers which do just the same Dec 05 01:32:43 mreimer, "Boot + suspend/resume confirmed to work on hx4700." that's from bd2. unfortunately, he's not irc now Dec 05 01:32:49 not on Dec 05 01:32:58 psokolovsky: I wonder if he's booting from flash rather than haret Dec 05 01:33:09 mreimer, yes Dec 05 01:33:19 hmm Dec 05 01:35:23 psokolovsky: I'm out of time for today. I'll pick away at it again when I get a chance. Dec 05 01:35:56 mreimer, Ok! Thanks much for you peeking at this! I'll try to catch bd2 too. Dec 05 01:36:08 psokolovsky: thanks Paul Dec 05 01:51:37 'night everyone **** ENDING LOGGING AT Tue Dec 05 02:59:57 2006