Failing test cases (and builds) on other architectures in Debian

This is merely a historical archive of years 2008-2021, before the migration to mailman3.

A maintained and still updated list archive can be found at https://lists.osmocom.org/hyperkitty/list/OpenBSC@lists.osmocom.org/.

Ruben Undheim ruben.undheim at gmail.com
Fri Nov 16 07:18:39 UTC 2018


> I actually hacked up one of those borderline-insane py scripts to mangle our C
> code into big-endian-reversed structs.
>
> See
> https://gerrit.osmocom.org/#/c/libosmocore/+/11786 for the script and
> https://gerrit.osmocom.org/#/c/libosmocore/+/11787 for what it did to libosmocore
>
> The script implementation itself is mad and convoluted, of course. How could it
> be different when handling and mangling C code.

Awesome, Neels!  This can become a very useful tool.

After all structs are fixed, we also have a few places with "ntohs"
and its friends to fix.

I've gotten quite a lot of test suites to pass now for several of the
packages. See here:
https://buildd.debian.org/status/package.php?p=Debian-mobcom-maintainers%40lists.alioth.debian.org&comaint=yes

However, in some cases I am unsure if I have fixed the problem or if I
have just hidden the problem...

For instance, this one:
 - https://sources.debian.org/patches/libosmocore/0.12.1-2/0006-Fix-some-byte-ordering-for-big-endian-architectures.patch/

It makes the test pass.

Even more hacky, and probably wrong is this one:
 - https://browse.dgit.debian.org/osmo-bsc.git/commit/?id=656937f42ab08ee206ddf4cd3c943d36916b0dcc
 (it also makes the test pass)

Could you please have a look at these two patches and see if they just
hide the problem, and if they do, what the correct fix is?

Best regards
Ruben



More information about the OpenBSC mailing list