If I compare two siginfo files with bitbake-diffsigs I expect this to travel the dependencies recursively. This is also confirmed by the examples I found on the mailing list. However this does not work. Example: pokyuser@0c1d32109027:/data/bitbake$ bitbake-diffsigs sstate-cache/82/bb/sstate:opencv:armv8a_tegra210-sorama-linux:4.1.0:r0:armv8a_tegra210:3:82bb97e29f39be34384adb0dd8d1561c096b6cecd50ceeebe3352d9ad20e9f9f_compile.tgz.siginfo sstate-cache/8c/11/sstate:opencv:armv8a_tegra210-sorama-linux:4.1.0:r0:armv8a_tegra210:3:8c11a5c8dfc7ba7a36a0c88caed504f2a2443d019915bd7582b54ca071c0c171_compile.tgz.siginfo NOTE: Starting bitbake server... NOTE: Started PRServer with DBfile: /workdir/build-nano/cache/prserv.sqlite3, IP: 127.0.0.1, PORT: 44171, PID: 16650 Hash for dependent task opencv/opencv_4.1.0.bb:do_configure changed from 86faeda9154b5fd9bed99627fce732cc19d9d4202be205c5f826e3e17ec28a4a to 1ae7154fd0468cca964a7204e2ed2fc14ed17352f88840c8702a714cc4a492ab Unable to find matching sigdata for /workdir/layers/meta-openembedded/meta-oe/recipes-support/opencv/opencv_4.1.0.bb:do_configure with hashes 86faeda9154b5fd9bed99627fce732cc19d9d4202be205c5f826e3e17ec28a4a or 1ae7154fd0468cca964a7204e2ed2fc14ed17352f88840c8702a714cc4a492ab dependend files do exist: -rwxrwxrwx 2 1000 pokyuser 31053 Aug 13 13:02 sstate-cache/86/fa/sstate:opencv:armv8a_tegra210-sorama-linux:4.1.0:r0:armv8a_tegra210:3:86faeda9154b5fd9bed99627fce732cc19d9d4202be205c5f826e3e17ec28a4a_configure.tgz.siginfo -rwxrwxrwx 2 1000 pokyuser 31053 Aug 16 14:49 sstate-cache/1a/e7/sstate:opencv:armv8a_tegra210-sorama-linux:4.1.0:r0:armv8a_tegra210:3:1ae7154fd0468cca964a7204e2ed2fc14ed17352f88840c8702a714cc4a492ab_configure.tgz.siginfo So I had expected that these files are examined too.
Bulk move from 4.99 or 0.00 to 5.99
There was work done in this area so create a new bug with specific reproducer if it is still a problem.