Bug#852013: Patch to prevent segfaults on signal

[Mattia Rizzolo]

On Sat, Jan 28, 2017 at 04:35:47PM -0500, Brett Smith wrote:
> I'm not complaining, just checking: my patch got reverted about twenty
> minutes after it got committed to master.  Are you sure you were testing
> in that time before it got reverted?

Yes.
Chris reverted it after my email.

> Earlier I was able to reproduce what was in your original report: same
> basic flow of debug messages, same issue of tempfiles not going away.
> The only difference I saw was that it segfaulted.  I thought maybe that
> wasn't included in the original report just because your shell didn't
> report it, or something like that.

As I wrote somewhere (but only on IRC maybe), I also noticed segfaults
while trying to reproduce it, but not so often.  It appeared to not
cleanup also when it didn't segfault.

> And it seemed to explain why the
> cleanup code wasn't running: the whole Python process was dying before
> it could get to the `finally` block that does the cleanup.  On my host,
> fixing this segfault and fixing the cleanup code went hand in hand,
> because now diffoscope could reliably get to the cleanup code.


Yesterday I tried several times with what was HEAD (with your patch) and
before that, and appeared to not clean up in both cases, without
sefualt reported.  Today it seem to always clean up with your patches
but always segfault (which is weird, as it never segfaulted reliably for
me) without.  meh.
OTOH, I now realized that somehow my shell might have eaten the
"Segmentation fault" message (although it doesn't do that usually), as I
just had a run terminated with code 139 without said message.
I'll report back once it does not segfault and not clean up.


BTW, I've committed your patche in a separated branch, as it deserves
to get in at some point (pity also because Debian is now freezing, and
we need to get the package into stretch, etc...

-- 
regards,
                        Mattia Rizzolo

GPG Key: 66AE 2B4A FCCF 3F52 DA18  4D18 4B04 3FCD B944 4540      .''`.
more about me:  https://mapreri.org                             : :'  :
Launchpad user: https://launchpad.net/~mapreri                  `. `'`
Debian QA page: https://qa.debian.org/developer.php?login=mattia  `-
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <http://lists.alioth.debian.org/pipermail/reproducible-builds/attachments/20170129/e0fe9b66/attachment.sig>