13512 2019-09-10 21:09:30 +0000 reproducible QA: Build core-image-sato reproducibly 2020-02-22 00:28:22 +0000 1 1 7 Build System, Metadata & Runtime OE-Core core unspecified x86 Multiple RESOLVED FIXED Medium normal 3.1 1 JPEWhacker sakib.sajal meta.mr.watcher meta.watcher randy.macleod richard.purdie ross.burton --- --- No (bug/feature does not impact docs) oldest_to_newest 84916 0 JPEWhacker 2019-09-10 21:09:30 +0000 core-image-sato should be built as part of the reproducible build QA test for better recipe coverage. 84991 1 ross.burton 2019-09-17 09:15:48 +0000 I did a test build. These are the recipes that fail: jpeg-tools_2.0.3-r0_amd64.deb libatspi-src_2.32.1-r0_amd64.deb libical-dbg_3.0.5-r0_amd64.deb libical_3.0.5-r0_amd64.deb libjpeg-dbg_2.0.3-r0_amd64.deb libjpeg62_2.0.3-r0_amd64.deb librsvg-2-2_2.40.20-r0_amd64.deb librsvg-2-dbg_2.40.20-r0_amd64.deb libtag-c0_1.11.1-r0_amd64.deb libtag-dbg_1.11.1-r0_amd64.deb libturbojpeg0_2.0.3-r0_amd64.deb pango-src_1.44.6-r0_amd64.deb 86399 2 randy.macleod 2020-02-11 03:34:01 +0000 I think that libical is the only remaining reproducible build problem package when I run: oe-selftest -r reproducible.ReproducibleTests.test_reproducible_builds but I only see this one unassigned defect mentioning it. I noticed that Richard and Armin were celebrating the abundance of working builds. Am I missing something? I did check in master-next for a change to libical. 86400 3 JPEWhacker 2020-02-11 04:14:09 +0000 I think this bug predates us making individual bugs for each non reproducible package. I also thought that we were already testing core-image-sato on the autobuilder and probably haven't seen if any iCal issues there. If you are seeing problems, I think a new bug would need to be made. 86410 4 randy.macleod 2020-02-12 04:16:03 +0000 Sakib is going to try to reproduce the libical problem that I saw and debug it if needed. 86495 5 randy.macleod 2020-02-19 21:57:31 +0000 We finally figured out the local.conf entries needed to reproduce the issue in another account. In the -dev package, a file: /usr/lib/cmake/LibIcal/LibIcalTargets.cmake has some host contamination: INTERFACE_COMPILE_OPTIONS "-I/ala-lpggp31/rmacleod/src/distro/yocto/b/rust-jan6/reproducibleA/tmp-glibc/work/core2-64-oe-linux/libical/3.0.7-r0/recipe-sysroot/usr/include/glib-2.0;-I/usr/lib/glib-2.0/include;-DG_LOG_DOMAIN=\"libical-glib\";-DLIBICAL_GLIB_COMPILATION" So Sakib is going to extend the sed filter in libical's recipe to fix that. 86552 6 randy.macleod 2020-02-22 00:28:22 +0000 The last known issue, with libical-dev is fixed by: https://git.openembedded.org/openembedded-core/commit/?id=9b3aaa03ace3975b2d6fb9a707b4b791d4f0f916