**** BEGIN LOGGING AT Fri Sep 04 02:59:57 2009 Sep 04 03:01:30 hmm Sep 04 03:01:48 what image/distro are you building? Sep 04 03:01:56 http://pastebin.com/dc236ef6 heh kind of lengthy and filled extra files Sep 04 03:02:21 angstrom 2008.1 base-image Sep 04 03:07:05 the most relevant lines of that pastebin are near the bottom Sep 04 03:09:21 when it boots, do you have udevd process running, /dev mounted and anything in it? Sep 04 03:09:54 no i can't do anything Sep 04 03:10:18 because console is unavailble, getty can't see /dev/ttyS0 Sep 04 03:10:37 i did init=/bin/sh to try to debug it a bit further Sep 04 03:14:50 ah, I see Sep 04 03:14:53 okay cool, the logs are available via logread Sep 04 03:15:01 Jan 1 00:36:46 (none) daemon.err udevd[76]: error in select: Invalid argument Sep 04 03:15:48 quite a few of those Sep 04 03:16:23 its logging that continuously Sep 04 03:16:30 what kernel, what libc? Sep 04 03:17:48 2.6.24.7 ppc405gp, gcc 4.4.1, uclibc 9.30.1 Sep 04 03:19:13 google gives this thread: http://lists.impactlinux.com/pipermail/firmware-impactlinux.com/2009-April/thread.html#166 Sep 04 03:19:31 which discuss this problem with udev-140 and uclibc Sep 04 03:19:40 heh i'm there Sep 04 03:20:02 I guess you can try udev-124 instead Sep 04 03:20:53 sounds like a plan Sep 04 03:23:38 thanks Sep 04 03:26:55 what about udev-146 Sep 04 03:26:56 hrm Sep 04 03:27:09 * m4t checks to see if it has relevant change(s) Sep 04 03:30:51 oh its definitely changed Sep 04 03:31:39 141: fdcount = ppoll(pfd, nfds, NULL, &orig_mask) Sep 04 03:31:56 146: fdcount = ppoll(&pmon, 1, NULL, &sigmask); Sep 04 03:36:25 ChangeLog: configure.ac: print error if gperf is missing Sep 04 03:36:27 :| Sep 04 03:37:25 somebody knows how to bake angstorm without build a kernel? Sep 04 03:47:36 here's a fun one: udev/udevadm-info.c:55: error: 'AT_SYMLINK_NOFOLLOW' undeclared (first use in this function) Sep 04 03:48:00 even though staging has : Sep 04 03:48:01 linux/fcntl.h:#define AT_SYMLINK_NOFOLLOW0x100 Sep 04 03:48:13 fcntl.h:# define AT_SYMLINK_NOFOLLOW0x100 Sep 04 03:48:46 and fcntl.h is included in udevadm-info.c Sep 04 03:52:18 problem seems to be in fcntl.h itself Sep 04 03:52:19 #if 0 /*def __USE_GNU*/ Sep 04 03:52:21 ... Sep 04 03:53:31 are you trying udev-146? Sep 04 03:54:00 ye Sep 04 03:54:07 ran into another one now: undefined reference to `openat' Sep 04 03:57:12 looks like openat has been intentionally removed Sep 04 03:57:20 * m4t has no idea wtf it does Sep 04 03:58:26 any specific reason not to try udev-124? Sep 04 04:01:32 http://bugs.gentoo.org/show_bug.cgi?id=266290 Sep 04 04:02:11 http://bugs.gentoo.org/attachment.cgi?id=188500&action=view should fix it, heh Sep 04 04:02:28 * m4t stops destroying his header files Sep 04 04:02:46 undestroying? Sep 04 04:12:19 81-netlink-owner-check.patch 82-encoding-overflow.patch should fix the issues, i will test out the build and try it out Sep 04 04:23:38 Starting udev Sep 04 04:23:38 and populating dev cache Sep 04 04:23:43 :) Sep 04 04:23:44 124 Sep 04 04:23:47 03Khem Raj  07org.openembedded.dev * rc26fc5db90 10openembedded.git/recipes/dropbear/ (5 files): Sep 04 04:23:47 dropbear: Enable allow empty password patch. Sep 04 04:23:47 * Switch to using INC_PR Sep 04 04:23:47 * Enable allow-empty_passed.patch without this Sep 04 04:23:47 its not possible to login unless root password Sep 04 04:23:49 is changed to be non empty Sep 04 04:23:53 Signed-off-by: Khem Raj Sep 04 04:39:04 m4t: congratulations! Sep 04 04:39:24 thanks Sep 04 04:39:26 http://pastebin.com/d2e8c9ec9 :-] Sep 04 04:40:37 egh so the only thing left really is to dremel the rear panel of this enclosure Sep 04 04:40:37 m4t: I'd suggest mentioning this problem with udev>=140 and uclibc on the mailing list. it maybe also platform or kernel specific though... Sep 04 04:48:17 * denix leaves now Sep 04 07:54:52 goodmorning Sep 04 07:55:21 Is it correct that the tinylogin addgroup utility can not be used to add an existing user to an existing group? Sep 04 08:03:04 does gnash run in framebuffer (without having x installed)?` Sep 04 08:03:42 yes, there are recipes for gnash on the fb Sep 04 08:04:50 they are called gnash-minimal in oe Sep 04 08:05:01 ok, thanks. I will check for that Sep 04 08:14:06 looks like gnash-minimal doesn't build Sep 04 08:17:17 ERROR: '/home/jovox/oe/openembedded/recipes/agg/agg_2.5.bb' failed Sep 04 08:17:48 Hi Sep 04 08:19:38 I've just attempted a build, we have included groff in our image. The QA during configuration is picking up a build dependancy on gawk. Should "DEPEND += gawk" be added to the groff recipe? Sep 04 08:44:02 good morning Sep 04 09:17:15 florian: good morning Sep 04 09:17:36 MWelchUK: probably, yes. Sep 04 09:19:11 good morning Sep 04 09:24:05 swfdec finally installed but doesnt seem to work very well. I tried it using the standalone client swfplay.. Some of the swf files loads and it looks like the swf is starting but then it just stops loading it.. Anyone tried swfplay? Sep 04 09:44:08 03Koen Kooi  07org.openembedded.dev * r00705178d3 10openembedded.git/recipes/linux/ (23 files in 4 dirs): linux-omap git: add dss2 and cache patches Sep 04 09:44:19 03Koen Kooi  07org.openembedded.dev * reff28b3b4e 10openembedded.git/recipes/dropbear/ (5 files): Sep 04 09:44:19 Revert "dropbear: Enable allow empty password patch." Sep 04 09:44:19 dropbear.inc needs the patch non-applied, see the thread in http://thread.gmane.org/gmane.comp.handhelds.openembedded/26302 Sep 04 09:44:19 This reverts commit c26fc5db90702b035bd545cff3ee7575a0f9b70f. Sep 04 10:11:04 holly sh*t, abi bumped again! :) Sep 04 10:11:22 pb__, Looking at some LFS docs (at the bottom) it looks like there are a number of "installation dependences". should these all be in it? Sep 04 10:23:36 Link helps! : http://tinyurl.com/m54mv7 Sep 04 10:27:44 MWelchUK_work: I think most of those are part of the base dependency set that must always be present. Sep 04 10:28:01 If it genuinely does require bash (as opposed to any other sh) then that should probably be declared. Sep 04 10:28:33 binutils and gcc are implicit dependencies (see base.bbclass); make is, I think, always assumed to be there although oe doesn't depend on it explicitly. Sep 04 10:28:55 likewise sed, cat, tr: I think the requirement for those is fairly deeply ingrained into the oe core. Sep 04 10:31:23 pb__,Interestingly the two warnings I see in the config.log are for gawk and gs. Ghostscript isn't even mentioned on that list, so I guess that document is a little out of date. Sep 04 10:32:24 pb__, it could be that we have already built some of the other dependencies, I'm not sure whether that would be enough for them not to show up as errors? Sep 04 10:36:28 MWelchUK_work: what are the actual warnings you see? Sep 04 10:36:49 if you hadn't built the other deps yet then I would have expected them to be errors not warnings, but you never know Sep 04 10:42:18 mornig Sep 04 10:43:06 booxter: thats why I put updating on hold for some time Sep 04 10:43:42 I cannot afford two days for rebuilding tmp Sep 04 11:12:50 pb__, Ah! Just looked at insane.class, that's completely wrong. Sep 04 11:14:09 It's using -I/usr/include for some reason. I also had ccache installed, which I thought I'd removed. It works fine on one of our other machines (that definitely doesn't have ccache) so I'm going to remove it and try again... Sep 04 11:18:19 Yeah, it seems like ccache is generally more trouble than it's worth for a lot of people. Sep 04 11:19:14 I really think it should be removed as a "required package" in OEAndYourDistro wiki page. It's clearly not required as OE builds fine without it. Sep 04 11:19:23 and it causes problems Sep 04 11:19:38 Yeah, that page is probably just out of date. Feel free to fix it. Sep 04 11:25:02 Shall I remove all mentions of ccache from that page? Sep 04 11:25:30 It seems to be fingered for breaking lots of builds in random ways Sep 04 11:26:07 Might as well. Current thinking seems to be that you're better off without it. Sep 04 11:30:08 has anyone got swfdec work properly? I just tried the mozilla plugin ant it loads the swfs but just wont "run" them Sep 04 12:04:42 good morning Sep 04 12:20:06 hi rkirti Sep 04 12:22:07 hello florian Sep 04 13:01:20 can someone please help me out with the error I am getting when bitbaking http://pastebin.com/m19fedcd4 Sep 04 13:21:41 morning Sep 04 13:22:06 whitefox: did you bitbake glib-2.0 ? Sep 04 13:22:25 looks like the recipe doesnt state this dependency Sep 04 13:26:05 * * OE Bug 5318 has been created by juanbackson(AT)gmail.com Sep 04 13:26:07 * * unable to bitbake ortp Sep 04 13:26:09 * * http://bugs.openembedded.org/show_bug.cgi?id=5318 Sep 04 13:29:29 rkirti: no, but should I try that? Sep 04 13:31:57 whitefox: yep Sep 04 13:32:02 that should solve the issue Sep 04 13:34:11 03Koen Kooi  07org.openembedded.dev * r881f92e340 10openembedded.git/recipes/linux/ (9 files in 5 dirs): Sep 04 13:34:11 linux-omap git: another checkpoint Sep 04 13:34:11 * add V4l2 interfaces to DSS2 Sep 04 13:34:11 * add back MADC driver Sep 04 13:34:11 * add zippy support for beagleboard Sep 04 13:34:13 * add DSS2 for omap3evm Sep 04 13:38:32 <_whitefox> rkirti: it still sas "glib.h: No such file or directory" Sep 04 13:39:22 glib.h? Sep 04 13:39:25 bitbake glib-1.2 Sep 04 13:40:02 <_whitefox> hrw: thanks let me try again Sep 04 13:41:14 hrw: how did you determine ? both 1.2 and 2.0 provide the file right ? Sep 04 13:41:44 no Sep 04 13:42:03 ortp really uses glib-2.0 - I did not chekced pastebin before Sep 04 13:47:09 <_whitefox> i am still getting the same error even after bitbake 1.2 Sep 04 13:47:28 <_whitefox> hrw: still the same Sep 04 13:56:24 the recipe doesnt have glib-2.0 as dep. Sep 04 13:56:28 thats a bug Sep 04 14:12:13 * florian shoots opkg Sep 04 14:12:54 florian: I forgot to remind you about 1.32 Sep 04 14:12:57 I wonder why this stupid piece of softtware to decided to prefer armv4 over armv5 Sep 04 14:13:06 hrw: right! Sep 04 14:13:09 florian: check arch.conf? Sep 04 14:14:08 hrw: its during image build... so its ipkg.conf Sep 04 14:14:20 eh opkg.conf Sep 04 14:14:32 same thing Sep 04 14:14:32 arch armv4t 26 Sep 04 14:14:38 arch armv5te 31 Sep 04 14:14:46 and there is armv5te package? Sep 04 14:18:03 morning kergoth Sep 04 14:18:17 hrw: should be... just rebuilding the thingy Sep 04 14:18:23 crap Sep 04 14:18:31 we need to increase inodes amount Sep 04 14:18:35 hey Sep 04 14:21:01 hi kergoth, hi Crofton Sep 04 14:30:31 have a nice rest of day Sep 04 14:32:31 i ♥ my psgrep & ppsgrep zsh functions. wraps around a ps | grep. psgrep shows only processes with the grepped text in their commandlines, and ppsgrep shows those, plus the full ps tree view of their parents Sep 04 14:38:52 heh, very good Sep 04 14:40:57 got the shlibdeps missing dep checking working last night, it caught 8 of them in a standard mvl6 build. some would never have caused a problem in the wild though, since their deps got pulled in indirectly, but even so... amazing how many aren't recorded in the metadata Sep 04 14:41:03 heh Sep 04 14:41:42 yeah, I can imagine there's a lot of that going on. Sep 04 14:51:34 i hate it when i can't *quite* wrap my head around a theoretical problem case with some code Sep 04 14:51:42 i have this itch, i know there's a case I'm missing, that it could fail in Sep 04 14:51:48 but i can't quite work out the exact circumstances Sep 04 14:52:28 erk Sep 04 14:52:42 fyi, never store a generator in the datastore Sep 04 14:52:46 things get very unhappy Sep 04 14:52:50 copy.copy() explodes Sep 04 14:52:52 hehe Sep 04 14:55:02 oh jesus Sep 04 14:55:06 shoot me now Sep 04 14:55:11 stupid anonymous python concatenation Sep 04 14:55:26 you can't use set(), because some other snippet is shadowing the builtin Sep 04 14:55:31 * kergoth runs __builtins__["set"]() Sep 04 15:11:15 ugh Sep 04 15:11:20 note: REPENDS != RDEPENDS Sep 04 15:11:22 stupid typo Sep 04 15:12:36 this is something it would be nice to be easier to catch. downside to this flexible way of handling the metadata Sep 04 15:12:45 kergoth: We need to stop that concatenation Sep 04 15:12:58 would kind of rather it said jackass, that variable doesn't exist Sep 04 15:13:00 kergoth: Wrap each snippet of code in a function, then call each function in turn Sep 04 15:13:08 yeah, thats what i was thinking too Sep 04 15:13:21 its actually on that task list.. low priority though Sep 04 15:13:27 heh. isn't that how it used to be? Sep 04 15:15:01 pb__: No, each was compiled as a seperate function Sep 04 15:15:04 pb__: i think it eval'd each function and exec'd it indepedently Sep 04 15:15:11 hence the speed gain by concatenation Sep 04 15:15:25 as opposed to wrapping each bit in a function definition, and appending calls to those functions at the end fo the string, and eval'ing that Sep 04 15:15:57 ah, right, I see. Sep 04 15:16:27 * kergoth isn't entirely clear on why eval'ing them separately is such a huge loss, performance wise, but *shrug*, whatever work Sep 04 15:16:28 s Sep 04 15:16:45 yeah, it does seem weird Sep 04 15:16:57 but, like you say. oh well. Sep 04 15:33:19 whats diff between base_bindir and bindir Sep 04 15:33:31 bitbake -e shows they both point to /bin Sep 04 15:34:10 I thought one of them prolly bindir should be /usr/bin Sep 04 15:34:28 bindir is /usr/bin unless you're on micro, which flattens the structure Sep 04 15:39:06 ah thats the problem I am on micro Sep 04 15:39:23 trying to build procps Sep 04 15:39:41 some recipes/projects assume they're different, probably just another one to fix eh Sep 04 15:39:43 and it depends on bindir and base_bindir Sep 04 15:46:13 kergoth: so whats the norm ? it needs to peep into /usr/bin as well as /bin Sep 04 15:46:33 its doing it correctly unless its micro Sep 04 15:48:18 not sure what you mean. if it puts somet binaries in one and some in the other, it needs to not choke when they're different, and put all the files in the same location Sep 04 15:48:46 actually it needs to pick them from different location Sep 04 15:48:55 hi kergoth and khem Sep 04 15:49:15 and since bindir is /bin it does not find the binary to copy and dies Sep 04 15:49:30 woglinde: howdy Sep 04 15:49:46 khem your ssh patch wasn so good *g* Sep 04 15:49:52 t Sep 04 15:50:25 khem: so fix it to not die if they're the same. shouldn't be that difficult Sep 04 15:50:29 woglinde: yeah actually I messed my commit got myself tangled into git Sep 04 15:50:35 it was meant to be local Sep 04 15:50:55 I wish there was something like git push Sep 04 15:51:10 I did a push from a not_meant_to_be branch Sep 04 15:51:31 oh Sep 04 15:51:32 you can push a specific commit id Sep 04 15:51:39 but remember that that pushes that commit -and its parents- Sep 04 15:51:42 kergoth git push Sep 04 15:51:43 just wouldnt push whats on top of that commit Sep 04 15:51:53 git push : Sep 04 15:51:58 yeah yeah the problem is below :) Sep 04 15:51:59 i use it on occasion to avoid pushing the top commits Sep 04 15:52:09 then rebase -i is your friend Sep 04 15:52:10 :) Sep 04 15:52:36 I think I have way too many local branches that I mess them quickly and more when kid is running around with messy diapers Sep 04 15:53:01 yeah I have fallen for rebase lately Sep 04 15:54:46 kergoth: say binary 'a' is installed into /bin and b into /usr/bin now it copied them like cp ${D}${base_bindir}/a .. cp {D}${bindir}/b .. Sep 04 15:55:15 i still don't see the problem. if they're the same, it's ${D}/bin/a and ${D}/bin/b Sep 04 15:56:10 NOTE: Task failed: Missing dependencies for gnutls-2.4.2-r4: ncurses, readline, libgpg-error Sep 04 15:56:11 hehe Sep 04 15:56:14 khem I see Sep 04 15:58:00 dare i attempt a world build? Sep 04 15:58:10 * kergoth is a glutton for punishment, clearly Sep 04 16:00:03 kergoth: they are installed into bin/ and usr/bin actually Sep 04 16:00:35 its the following steps after install which wants to copy them around and fail Sep 04 16:00:39 if the recipe isn't obeying our path variables, it should be fixed to pass them in Sep 04 16:06:26 kergoth: can you see recipes/procps/procps_3.2.7.bb whats missing in it Sep 04 16:07:45 i don't see anything wrong with that. what's the error? Sep 04 16:07:56 the same one I explained earlier Sep 04 16:08:13 no, the exact error Sep 04 16:08:15 not your summary of it Sep 04 16:08:43 | mv: cannot stat `/home/kraj/work/oe/tmp/work/armv5te-oe-linux-uclibceabi/procps-3.2.7-r8/image/bin/uptime': No such file or directory Sep 04 16:08:46 | ERROR: Function do_install failed Sep 04 16:10:11 do_install did not honor bindir and base_bindir being same Sep 04 16:10:30 yeah, probably need to patch their makefile, or pass in the bindir/base_bindir explicitly Sep 04 16:10:48 well, either that or do an if .. check where it is, then move it, but thats .. ick Sep 04 16:10:50 hehe Sep 04 16:12:37 i c they dont use configure Sep 04 16:12:56 figured.. gotta love custom buildsystems.... Sep 04 16:13:28 you can tell the people who hate autoconf never had to do a lot of distro work Sep 04 16:13:40 very true Sep 04 16:15:57 hi florian Sep 04 16:19:54 kergoth: i think love it or hate it but autotools is one reason of many reasons linux is widely deployed Sep 04 16:20:03 agreed Sep 04 16:20:09 everyone knows to ./configure; make; make install Sep 04 16:20:17 hm Sep 04 16:20:21 not every one Sep 04 16:20:22 *g* Sep 04 16:20:26 hehe Sep 04 16:20:46 I work on bsd these days and its ugly there too although they use makefiles Sep 04 16:20:53 but have a look at the make files yuck Sep 04 16:21:15 when it comes to supporting multiarch Sep 04 16:21:29 problem with autotools is, that most people dont update their configure.in/ac with the new features Sep 04 16:22:08 problem with autotools is, most developers don't understand it, or care to Sep 04 16:22:12 or using the old doucs Sep 04 16:22:21 kergoth that too Sep 04 16:22:45 saw it many times shippe aclocal stuff without the .m4 Sep 04 16:23:44 anyone want to check out the final (well, mostly) missing dep code i did for MV? I'm sure they wont mind, since we'll be pushing it upstream anyway Sep 04 16:24:07 could use another set of eyes Sep 04 16:24:11 http://kergoth.pastey.net/124651 Sep 04 16:24:55 should probably add a .rstrip() on the 'yield line' in the all_shlibdeps Sep 04 16:26:02 should move the getvar of pkgdest out of the loop in shlibdeps too Sep 04 16:26:07 * kergoth shakes head Sep 04 16:37:54 and autotools is smarters than 95% of humans Sep 04 16:38:09 thats the biggest problem Sep 04 16:38:45 kergoth, the old "more self-review now that I put it up for public review" problem, eh? :) Sep 04 16:38:53 indeed Sep 04 16:39:05 only a couple minor things though this time :) Sep 04 16:39:23 think I'm getting slightly less crappy at python, this is almost comprehensible ;) Sep 04 16:41:41 one thing i wonder.. Sep 04 16:42:05 filter(lambda p: foo(p), mylist) or (item for item in mylist if foo(item)) Sep 04 16:42:24 'item for item' annoys me, but the lambda isn't particularly pretty either Sep 04 16:44:02 * kergoth wanders off to clean his garage... moving sucks Sep 04 16:45:33 kergoth hm where do you move? Sep 04 16:48:23 isn't it already too hot to do that? :) Sep 04 17:18:09 Tartarus: how is AZ ? Sep 04 17:18:41 Tartarus: I has planned a visit to Chandler this long weekend but have to cancel Sep 04 17:23:00 03Khem Raj  07org.openembedded.dev * r955c5032b2 10openembedded.git/conf/distro/include/ (4 files): Sep 04 17:23:00 angstrom: Use linux-uclibceabi instead of linux-uclibcgnueabi Sep 04 17:23:00 * For uclibc based ARM EABI systems switch to using linux-uclibceabi Sep 04 17:23:00 as TARGET_OS Sep 04 17:23:00 Signed-off-by: Khem Raj Sep 04 17:23:01 Acked-by: Koen Kooi Sep 04 17:25:00 * khem reviews his grammar Sep 04 17:26:59 Does anyone know the best way to force linking with libstdc++ when using oe_runmake? Sep 04 17:27:53 radhermit: what specific package are you referring to Sep 04 17:27:57 The situation is that I'm trying to build a program that only uses a static makefile and sets compiler flags and linking flags in the makefile so I have to unset those variables before calling oe_runmake Sep 04 17:28:14 khem: cc_flasher, that is not in the standard repo Sep 04 17:29:06 Since libstdc++ looks to be in the cross directory the linking process doesn't find it since it probably only looks in the staging directories, right? Sep 04 17:31:06 radhermit: if you need libstdc++ one of easy and best ways it to link using g++ Sep 04 17:31:09 radhermit set LDFLAGS Sep 04 17:31:27 woglinde: You mean don't unset them? Sep 04 17:32:03 woglinde: that may not be good thing to do when gcc is used to link Sep 04 17:32:14 khem: Ok, so just set LD or something similar to g++? Sep 04 17:32:31 yes Sep 04 17:32:40 khem: Thanks, I'll give it a try Sep 04 17:34:36 Thanks again, that worked well Sep 04 17:35:40 with gcc you -lsupc++ or something Sep 04 17:39:07 yeah so its extra pain that g++ already takes care for you Sep 04 17:40:22 pb__: gcc 4.4.1 looks ok with C++ with few apps I tried Sep 04 17:40:44 I am working on getting g++ regression testsuite executing on uclibc/thumb Sep 04 17:41:12 if that works reasonably then we can be bit more confident Sep 04 17:50:42 Tartarus: heh, I think I'd have to do my cleaning at 3am or something for it to not be too hot to clean the garage.. Sep 04 18:00:04 khem: okay, excellent Sep 04 18:00:54 I am having trouble compiling perl (target) Sep 04 18:00:54 Tartarus: you're in AZ now? Sep 04 18:01:21 yes, running the testsuite would be good. I guess you'd probably need to do a side by side comparison run against 4.2.4 since we are some way from having zero failures anyway. Sep 04 18:02:02 target perl is a bit of a nightmare, you're probably best off avoiding that. I would suggest running the testsuite driver on the host and just execute the test binaries on the target, don't try to run any of the infrastructure there. Sep 04 18:02:12 pb____: I already have results of 4.2.4 from couple of months back but I should run it again once I have the image building should be matter of time Sep 04 18:02:38 pb____: yeah thats what I will resort to Sep 04 18:02:53 iirc, all the g++ tests are dejagnu so that should be fairly easy to arrange Sep 04 18:03:03 I think the problems are because /bin and /usr/bin are same for micro but some apps dont think so Sep 04 18:03:19 pb____: yes I have the cross setup Sep 04 18:03:33 possibly, though I think I have had perl running in a micro setup in the past. Sep 04 18:03:39 I was also aiming for uclibc tests Sep 04 18:03:48 which unfortunately needs to be run on target Sep 04 18:04:02 only on x86 though, I haven't done it on arm in ages. Sep 04 18:04:22 the only perl app that I ever wanted to run on arm was frozen-bubble anyway :-} Sep 04 18:04:23 yeah for x86 things are a bit nicer Sep 04 18:04:29 heh Sep 04 18:05:34 I think I started rewriting f-b in C although I never finished it. nowadays I guess perl-on-target is a bit less of a horror than it was in 2005, so probably not so much worth porting anymore anyway. Sep 04 18:05:52 bbiab, gotta go dig a hole Sep 04 18:55:13 re Sep 04 19:42:49 re Sep 04 19:43:07 hi woglinde Sep 04 19:43:11 hello woglinde Sep 04 19:43:37 that's better, tail was getting out of hand Sep 04 20:15:08 re Sep 04 20:17:01 wb Sep 04 20:20:35 hi Martin-B Sep 04 21:42:00 03Stanislav Brabec  07org.openembedded.dev * r69ae87f9c9 10openembedded.git/ (2 files in 2 dirs): xf86-video-fbdev: Updated to version 0.4.1. Sep 04 21:42:10 03Stanislav Brabec  07org.openembedded.dev * r907417094c 10openembedded.git/ (conf/checksums.ini recipes/xorg-app/xinput_1.4.2.bb): xinput: Updated to version 1.4.2. Sep 04 21:42:11 03Stanislav Brabec  07org.openembedded.dev * r4c4bb699ac 10openembedded.git/ (2 files in 2 dirs): xf86-input-evdev: Update to version 2.2.5. Sep 04 22:48:21 pb_: ping Sep 04 22:48:33 he ant Sep 04 22:48:39 hi woglinde Sep 04 22:49:10 I get arch/arm/mach-pxa/kernelsw.c:985: error: implicit declaration of function '__ioremap' Sep 04 22:49:37 arch/arm/mm/ioremap.c * Hacked for ARM by Phil Blundell Sep 04 22:50:07 seems gone between 2.6.16 and 2.6.26 Sep 04 22:50:12 hmm Sep 04 23:08:05 hm.. Sep 04 23:08:19 linuxtogo.org looking strange Sep 04 23:08:35 jay7 yeah Sep 04 23:08:41 planet is down too Sep 04 23:09:05 geil Sep 04 23:09:17 well.. time to sleep then :) Sep 04 23:09:23 hehe Sep 04 23:09:26 good nite Sep 04 23:09:34 I will follow soon Sep 04 23:22:57 * kergoth yawns Sep 04 23:23:38 hehe Sep 04 23:23:40 too hot? Sep 04 23:24:01 worn out, and all i did was clean a garage, and didn't even finish it. amazing how much heat can just seem to drain your energy Sep 04 23:24:44 haha Sep 04 23:24:53 be lucky to dont have to much humidity Sep 04 23:25:03 yeah no doubt Sep 04 23:25:25 where are you located? Sep 04 23:25:42 *g* I am dont suffer to much from both in berlin Sep 04 23:26:07 ahh right :) Sep 04 23:47:17 good nite Sep 05 00:41:07 hi how can i put my own full screen app on beagleboard on startup instead of login screen ? Sep 05 01:38:58 i want to disable login screen of angstrom ? how can i do it ? Sep 05 01:58:50 how can i add my java app on startup ? shows fullscreen window i ve installed java on beagle **** ENDING LOGGING AT Sat Sep 05 02:59:58 2009