[fpc-devel] status of "haltproc" issue tracked by 17383, 14958, 18831

[Marco van de Voort]

In our previous episode, Seth Grover said:
> Jonas (or whoever else might care to answer),
> 
> I've been scratching my head trying to figure out which issue to
> follow in mantis regarding this bug. I *think* from what I've reading
> the issue has been resolved completely (I haven't got a trunk testing
> environment set up to test it), so hopefully you can straighten me
> out.
> 
> Issue 14958 ("Cannot unload shared library") was logged and then
> fixed, but it caused a regression which I logged as issue 17383
> ("unhandled exception caught by .so's handler causes infinite loop").
> You reverted the regression, and then it appears later committed
> revision 16418 to once again resolve issue 14958.
> 
> I can see however that there are still relationships between 14958 and
> 14959 (closed) and 12492 (still open). But I don't think the issue of
> the .so unloading is directly affected by that issue.
> 
> Anyway, there were a lot of commits and revisions and
> interdependencies between these issues, but I think what I'm taking
> away from all of this is that the issue has been, in your estimation,
> fully resolved: that shared libraries can now be successfully unloaded
> (14958), and that the unhandled exception regression no longer occurs
> (17383). Is that correct? And it appears that these fixes were
> implemented in 2.5.1+ and not ported back to 2.4.4, is that also
> correct?

Revisions that are not merged are here:

http://www.stack.nl/~marcov/mergelogs/resultset.txt

As far as I can see none of them are merged.

Mantis items that have been merged to 2.4.4 should have had their targets
set to 2.4.4, but that is not entirely fool-proof since it happens using
information in commitmessages, and not all commits will mention the mantis
item.