Bug#1062956: sysprof: NMU diff for 64-bit time_t transition

[Simon McVittie]

Control: tags -1 + wontfix

On Sat, 24 Feb 2024 at 02:10:45 -0800, Steve Langasek wrote:
> +sysprof (46~beta-1.1~exp1) experimental; urgency=medium
> +
> +  * Non-maintainer upload.
> +  * Rename libraries for 64-bit time_t transition.

I'm pretty sure this one is actually unnecessary...

> --- sysprof-46~beta/debian/libsysprof-6-6t64.symbols	1970-01-01 00:00:00.000000000 +0000
> +++ sysprof-46~beta/debian/libsysprof-6-6t64.symbols	2024-02-24 10:09:00.000000000 +0000
> @@ -0,0 +1,5 @@
> +libsysprof-6.so.6 libsysprof-6-6t64 #MINVER#
> +* Build-Depends-Package: libsysprof-6-dev
> + sysprof_callgraph_category_get_type at Base 45~beta
> + sysprof_callgraph_flags_get_type at Base 45~beta
> + sysprof_recording_phase_get_type at Base 45~beta

... because these symbols clearly have no time_t in them.

Looking at build logs in
https://adrien.dcln.fr/misc/armhf-time_t/2024-02-22T21%3A03%3A00/logs/libsysprof-6-dev/
it seems like the test-build is onky failing because sysprof has headers that
are valid as C but not valid as C++. This is new since libsysprof-4, which
was the version that was tested successfully during the early phases of
this transition.

This unconditional use of restrict seems like a bug, I'll report it
upstream.

https://salsa.debian.org/vorlon/armhf-time_t/-/merge_requests/138 makes
this package compile successfully during the ABI check.

    smcv