Dear Neil,

[I try to give feedback to achieve higher quality, but it may sound a bit pushy. Please read this with best intentions in mind.]

On 08-03-2022 13:36, Debian Bug Tracking System wrote:
Changed-By: Neil Williams<codeh...@debian.org>
Closes: 1003061
Changes:
  dmrgpp (6.02-3) unstable; urgency=medium
  .
    * Exclude armhf from autopkgtest (Closes: #1003061)

I'm a bit surprised to see you solved bug 1003061 in this way. It seems to me that the binary this package built is segfaulting under certain condition. If this was a bug in my package I think I would have cloned the bug, lowered the severity (as apparently it doesn't segfault on all input), report it upstream and documented the segfault until it's fixed. Or, if the architecture is unsupported and nobody will look into it, maybe request the removal on armhf and not build there anymore.

Also, (with my ci.debian.net maintainer hat on) I would have hoped you would improve the test (I think the ci.pl script) to abort instead of hang in case of a segfault, just in case it happens again in the future on other architectures. Or at least report the issue upstream so they know.

Paul

Attachment: OpenPGP_signature
Description: OpenPGP digital signature

Reply via email to