Re: Potential spiped 1.6.2 release -- please test

[Graham Percival <gperciva@tarsnap.com>]

On Sat, Dec 11, 2021 at 11:55:06PM +0200, Peter Pentchev wrote:
> So about this particular Hurd patch... TBH, I would completely understand
> if you do not feel like accepting it;

Thanks!  I've rewritten the patches to follow our "house style", as well as
adding comments about why they're needed, and added one more patch that was
required.

We now have a branch which allows `make test` to pass on GNU/Hurd:
https://github.com/Tarsnap/spiped/tree/DO-NOT-MERGE-hurd

This currently has three commits:

- Allow ENOPROTOOPT for SO_REUSEADDR on AF_UNIX #341

  Might be merged to spiped master, likely after spiped 1.6.2.

- GNU/Hurd does not support cancellation points:

  This is on their list of "open issues", although much of the info is from
  2013 so I don't know how accurate it is:
  https://www.gnu.org/software/hurd/open_issues/libpthread_cancellation_points.html

- GNU/Hurd: shutdown(2) can produce EMIG_BAD_ID

  I think this is a kernel errno value, which their standard library should
  notice and replace with EBADF or ENOTSOCK.


The two latter commits are not candidates for merging to spiped master because
they're working around clear bugs (or missing features) in GNU/Hurd.  If you
would like to have a working spiped package on that system, please include
them as Debian-specific patches.
(Ideally as patches which are only applied on GNU/Hurd.)


Thanks for the work you've done testing and packaging spiped!

Cheers,
- Graham