On Mon, 2017-01-23 at 11:44 -0500, Paul Gortmaker wrote: > After updating this morning (everything on master, re-using the > existing > build dir that was built OK within the last 48h) I saw this: > > ERROR: pseudo-native-1.8.1-r0 do_populate_sysroot: pseudo-native: > chrpath command failed with exit code 7: > b'/home/paul/poky/build/tmp/work/x86_64-linux/pseudo-native/1.8.1- > r0/sysroot-destdir/home/paul/poky/build/tmp/work/x86_64-linux/pseudo- > native/1.8.1-r0/recipe-sysroot-native/usr/bin/pseudodb: > RPATH=/home/paul/poky/build/tmp/sysroots/x86_64- > linux/usr/lib:/home/paul/poky/build/tmp/sysroots/x86_64- > linux/lib\n'b"new > rpath > '$ORIGIN/../../../../../../../sysroots/x86_64- > linux/usr/lib:$ORIGIN/../../../../../../../sysroots/x86_64-linux/lib' > too large; maximum length 107\n" > ERROR: pseudo-native-1.8.1-r0 do_populate_sysroot: Function failed: > relocatable_binaries_preprocess > ERROR: Logfile of failure stored in: > /home/paul/poky/build/tmp/work/x86_64-linux/pseudo-native/1.8.1- > r0/temp/log.do_populate_sysroot.11
To follow up we've figured out this is the error you see if you take an existing build directory, pre recipe specific sysroot and then update to post recipe specific sysroot. It seems that pseudo-native, openssl-native and subversion-native have issues where you can pass in new configuration but they don't recompile the binaries, despite those binaries depending on the paths you pass in. We're seeing how widespread the problem is before deciding exactly how to handle this. It might be a case of some PR bumps, obvioulsy wiping TMPDIR is the comprehensive solution. Cheers, Richard -- _______________________________________________ Openembedded-core mailing list Openembedded-core@lists.openembedded.org http://lists.openembedded.org/mailman/listinfo/openembedded-core