**** BEGIN LOGGING AT Tue Aug 16 02:59:57 2011 Aug 16 03:00:44 anyone even here Aug 16 08:24:32 morning all Aug 16 09:03:46 hi bluelightning Aug 16 09:04:04 hi pb_ Aug 16 09:23:43 * pb_ stabs cmake Aug 16 09:24:53 early in the morning already? Aug 16 09:25:14 yeah, never too early to hate cmake Aug 16 09:25:29 ah, here we go Aug 16 09:25:32 NOTE: package webkit-clutter-1.3.11+git1+563e5fa32621f74b7238673032ebdae82ae45a0b-r0: task do_compile: Started Aug 16 09:25:33 phew Aug 16 09:34:36 wow..GMail opened in firefox 5.0 is eating 230MB on this poor Win box Aug 16 09:35:32 41MB for a blank page Aug 16 09:35:35 not bad Aug 16 10:16:27 03Paul Menzel  07master * r9a13ec3a5b 10openembedded.git/recipes/ckermit/ckermit_211.bb: Aug 16 10:16:27 ckermit: Order meta fields according to style guide [1] Aug 16 10:16:27 [1] http://wiki.openembedded.org/index.php/Styleguide Aug 16 10:16:27 Signed-off-by: Paul Menzel Aug 16 10:16:37 03Paul Menzel  07master * r5350bbef01 10openembedded.git/recipes/ckermit/ckermit_211.bb: Aug 16 10:16:37 ckermit: Fix typos in comments Aug 16 10:16:37 Signed-off-by: Paul Menzel Aug 16 11:36:52 hi Aug 16 12:24:03 is this gcc ICE known: http://pastebin.com/1pV9JqDB ? Aug 16 12:25:58 I don't think I've seen that one before. What version of gcc is that? Aug 16 12:27:08 gcc-cross-4.6.1+svnr175454-r3 Aug 16 12:27:21 (recent oe-core head) Aug 16 12:27:30 gcc-cross-4.6.1+svnr175454-r9 Aug 16 12:30:05 imagine gcc version 4.6.1 20110627 (prerelease) seems curing a misterious Alignment Fault happening with gcc4.5 on armv5 Aug 16 12:30:32 still under inspection Aug 16 12:35:04 ICE reproducable with gcc-cross-4.6.1+svnr175454-r2 too, but not with gcc-cross-4.5-r33.1+svnr170443 Aug 16 12:38:49 03Martin JaMa Jansa  07master * re02237d7e4 10openembedded.git/recipes/shr/ (4 files): Aug 16 12:38:49 libphone-ui-shr, libphone-ui, libphone-utils, e-wm-config-illume2-shr: bump SRCREVs Aug 16 12:38:49 Signed-off-by: Martin JaMa Jansa Aug 16 12:45:38 ensc|w: worth filing a report in bugzilla if there isn't one already Aug 16 12:47:55 in which bugzilla? gcc, linaro? Aug 16 12:50:18 gcc Aug 16 12:50:27 afaik, we are not applying any of the crazy linaro stuff Aug 16 12:51:06 ensc|w: while building gcc-cross? Aug 16 12:52:12 JaMa|Off: no, while building lttng-ust (which is a new dependency) Aug 16 12:52:26 ensc|w: I've sent ICE test case for latest linaro patches (applied in meta-oe) while building lttng-ust Aug 16 12:52:44 ensc|w: I have it only for armv7a, armv4t built fine Aug 16 12:52:46 critical options are '-march=armv7-a -O1' Aug 16 12:52:49 ensc|w: -O0 builds also fine Aug 16 12:52:55 yes, -O0 is ok Aug 16 12:53:20 so khem knows about it, but I haven't filled it myself Aug 16 12:56:30 there is arecent message Aug 16 12:56:31 http://www.spinics.net/lists/arm-kernel/msg136454.html Aug 16 12:56:47 but about kernel Aug 16 13:03:00 I think that report on the kernel list is bogus. Aug 16 13:03:24 or, well, it might be true that the kernel is being miscompiled, but if that's true then it is the kernel's fault Aug 16 13:03:56 can't say, I'm lost after so many pointers in 2 lines... Aug 16 13:09:18 pb_: I'm starting to wonder about the meta-openembedded layer Aug 16 13:09:59 what are you wondering about it? Aug 16 13:10:21 it is not needed for a console build Aug 16 13:10:37 pb_: the volatile should not be optimized away, so I think the report is valid Aug 16 13:10:39 if we merge meta-oe in oe-core the rest are only windows managers Aug 16 13:11:37 ensc|w: well, it works if you add "volatile" to the declaration of var. I'm not convinced that casting in (volatile) on those specific accesses is a legitimate thing to do. Aug 16 13:11:53 have to check the standard to be sure though, I guess, and I don't have it on hand here Aug 16 13:13:27 ant_work: that's probably true, but (afaik) there is no plan to merge meta-oe into oe-core, and the fact that the rest of the layers are "just" GUI bits doesn't seem like a bad thing Aug 16 13:15:23 ensc|w: so, regarding your ice, are you actually using the gcc from meta-oe? Aug 16 13:16:52 pb_: gcc people seem to accept the volatile thing as a bug in gcc: http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50078 Aug 16 13:16:57 we need to look carefully at what's in meta-oe IMHO... there should always be a good reason (not just "we need it and couldn't think of a better place") Aug 16 13:17:47 there is still part of toolchain and find that a bad thing Aug 16 13:17:48 ensc|w: ah, fair enough. I guess it is indeed meant to work then. Aug 16 13:17:58 pb_: reported my ICE as http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50099 I think it is gcc from oe-core but I am not sure whether/what is bbappended from meta Aug 16 13:18:35 if you have the meta-oe layer enabled then I think you will get meta-oe's gcc Aug 16 13:18:50 (unless you've taken special measures to avoid that) Aug 16 13:19:14 yesterday's oe-core shows 4.6.1 20110627 (prerelease) Aug 16 13:19:24 but, anyway, might as well wait and see what the gcc folks say now you've filed the ice Aug 16 13:19:42 the same with gcc from meta-oe :) gcc version 4.6.1 20110627 (prerelease) (GCC) Aug 16 13:19:45 I'm not sure whether meta-oe actually changes the version string Aug 16 13:20:20 it might be useful if it did add "(1337 meta-oe)" or something to the end to avoid confusion Aug 16 13:20:48 pb_: yes, you are right. it is from meta-oe with the linaro patches Aug 16 13:20:48 well, in an ideal world I suppose meta-oe would refrain from providing its own eleet gcc at all, but still. Aug 16 13:21:12 or (linaro 07.2011) Aug 16 13:22:46 doh, webkit now failing at 99% Aug 16 13:23:02 * pb_ stabs cmake again Aug 16 13:23:33 did somebody succeeded in building openssl-1.0.0d from meta-oe? Aug 16 13:23:45 it fails when expanding INC_PR here Aug 16 13:23:56 * JaMa|Off yes, but had to apply this http://git.openembedded.org/cgit.cgi/meta-openembedded-contrib/commit/?h=shr&id=bad363ba09589e470921d6cd437ba27b6a0082f5 Aug 16 13:24:41 ah... this openssl.inc, I removed the INC_PR and make stuck in an endless recursion then Aug 16 13:24:54 (resp. failed after 255 sub-makes) Aug 16 16:13:02 Damn native xsltproc is checking _host_ /etc/xml/catalog Aug 16 16:14:15 Is somebody building oe in really clean and small chroot? Aug 16 17:03:21 lumag_gone: I usually do my builds in a relatively small chroot, but it varies Aug 16 17:05:28 lumag_gone: swabber is designed to pick up those sorts of things Aug 16 17:50:30 did someone got problems of mouse inverting axes when changing resolution using xrandr? Aug 16 17:50:35 I am seeing it using geode driver Aug 16 17:50:39 with oe-core Aug 16 18:45:18 how do I get around "ERROR: libmad was skipped: because it requires commercial license to ship product" Aug 16 18:47:02 comment COMMERCIAL_LICENSE in distro/include/default-distrovars.inc ? Aug 16 18:47:36 joelangle disable mp3 support Aug 16 18:48:13 woglinde, angle? :P even angel would've been better Aug 16 18:49:23 yes yes Aug 16 18:50:15 woglinde, thanks . there are a few recipes in COMMERCIAL_LICENSE. I will comment it in conf/ instead of removing from DEPS in the recipe Aug 16 18:50:21 hopefully that should be clean enough Aug 16 18:56:12 hi effem Aug 16 18:56:25 hi woglinde, all Aug 16 18:56:38 all good ? Aug 16 19:18:41 what's the difference between openembedded and openembedded-core? Aug 16 19:20:06 openembedded-core is the future Aug 16 19:21:50 rah: http://wiki.openembedded.org/index.php/OpenEmbedded-Core Aug 16 19:22:56 thanks Aug 16 19:57:56 joelagnel: you can also set COMMERCIAL_LICENSE = '"" in your local.conf **** ENDING LOGGING AT Wed Aug 17 02:59:57 2011