**** BEGIN LOGGING AT Wed Oct 26 02:59:59 2016 Oct 26 04:11:49 denix: I am using master and doing nothing special just saw this warning Oct 26 04:11:59 hopefully it is harmless Oct 26 04:59:03 Hi all, I am trying to add a new packet (it++) but getting the following error during do_populate_sysroot: Oct 26 04:59:11 ERROR: itpp-4.3_8.2.0+gitAUTOINC+1553f9e2b0-r0 do_populate_sysroot: QA Issue: itpp.pc failed sanity test (tmpdir) in path /mnt/tmp/bitbake-glibc/work/armv7ahf-neon-oe-linux-gnueabi/itpp/4.3_8.2.0+gitAUTOINC+1553f9e2b0-r0/sysroot-destdir/usr/lib/pkgconfig [pkgconfig] Oct 26 04:59:36 Anyone any idea of what could be wrong? Oct 26 05:22:26 mmichael: seeme like it++ comes with a pkg-config file that needs fixing. without further looking and just guessing from the error message - it includes a temporary path from the build Oct 26 05:27:12 LetoThe2nd: thanks, did a grep for tmppath in the it++ checkout and found the following -> Oct 26 05:27:15 itpp.spec.in:BuildRoot: %{_tmppath}/%{name}-%{version}-%{release}-buildroot Oct 26 05:27:36 Could this be causing this issue? Oct 26 05:34:43 mmichael: it sounds possible, but i'm not really knowledgeable when it comes to QA nor pkg-config, sorry. Oct 26 05:35:15 mmichael: my guess would be to check other recipes that bring .pc files and see if they have some fixup mechanisms there. Oct 26 05:38:20 Thanks LetoThe2nd, I'll do that. Oct 26 06:41:50 mmichael: look for itpp.pc in Makefiles Oct 26 06:42:09 and most probably it is not fit for cross compliation Oct 26 06:42:14 so you have to hand hold it a bit Oct 26 06:42:36 some sed'ing or editing otherwise will help Oct 26 06:47:26 Thanks khem, had to take out a few lines from itpp.pc.in through a patch - all good now :-) Oct 26 07:07:15 how can I get my java build recipes to depend on icedtea7-native. If I add that to the DEPENDS it has no effect. All other uses of DEPEND seem to work. It's like the java system bypasses dependency checks Oct 26 10:44:24 mmichael, you the gnuradio guy? Oct 26 12:19:52 just cannot get the latest icedtea7-native to build properly Oct 26 12:22:31 Every other build server has to be baby sitted for a day to get that build not to fail with sha256 errors on an internal download of openjdk.tar.gz Oct 26 12:22:52 where the hell does that thing get cached Oct 26 12:22:54 how the hell Oct 26 12:24:01 somehow, the Makefile.am contains this openjdk.tar.gz as a variable that is to be downloaded by ./configure Oct 26 12:24:08 but somehow that file is cached by bitbake Oct 26 12:24:20 when I delete it, it still fails Oct 26 12:24:29 then when I loose my shit and mash the button... it comes right Oct 26 12:24:43 then on to the next build server Oct 26 12:26:11 THen to make things worse... bitbake simply ignores DEPENDS = "icedtea7-native" Oct 26 12:26:38 why do they give these java components these stupid names Oct 26 12:26:52 because when they fail, and they will, you dont know what the fuck is going on Oct 26 12:27:23 icead tea broke because jamvm is jammed because cacao is fucking out of stock because you entire god damn kitchen is in side the meta-java layer4 Oct 26 12:35:04 ln -sf /tc/work/ff8fd671fc1ba5b0/angstrom/sources/downloads/f89009ada191.tar.bz2 openjdk.tar.gz Oct 26 12:35:29 seeing shit like that this line above make me what to drive over there and just loose my shit at the guy who did this Oct 26 12:36:45 How can they allow makefiles to contain shit like this. Do they have a bunch of stupid moneies working on this java stuff Oct 26 12:48:35 If I can go back in time I will change one thing. Erase this abomonation called java from history... forever and pass a law that it may never be invented. What a fucking cluster fuck of a piece of software Oct 27 00:17:29 khem: per our conversation yesterday, are you saying to add a virtual/wayland-egl in mesa and then run with that... or did you mean something else? **** ENDING LOGGING AT Thu Oct 27 02:59:59 2016