Comment 10 for bug 1862926

Thank you for the second upload, it fixed the general FTBFS but some architectures still show problems:

i386 (the only remaining 32-bit architecture?):

The genesys backend fails the unit tests due to a problem with floating point arithmetics on 32/64-bit. There are bug reports upstream, but all unsolved:

https://gitlab.com/sane-project/backends/issues/241
https://gitlab.com/sane-project/backends/issues/157

ppc64el:

Fails on symbol table generation:

dh_makeshlibs -- -v1.0.29 -Pdebian/libsane -plibsane
dpkg-gensymbols: warning: new libraries appeared in the symbols file: libsane-ricoh2.so.1 libsane-escl.so.1
dpkg-gensymbols: warning: some new symbols appeared in the symbols file: see diff output below
dpkg-gensymbols: error: some symbols or patterns disappeared in the symbols file: see diff output below
dpkg-gensymbols: warning: debian/libsane/DEBIAN/symbols doesn't match completely debian/libsane.symbols.ppc64el

Do we really need to generate symbol tables for each scanner driver? No one develops against the drivers, only against the main library.

To fix, one simply needs to copy the patch from the buildlog and apply it.

And why do we need these scanner drivers on server-only architectures?

s390x:

Bog known and fixed upstream:

https://gitlab.com/sane-project/backends/-/merge_requests/329

So the biggest problem is the i386, as upstream has no solution for that, perhaps skip this one test (genesys backend) i386-only.