[Pkg-iscsi-maintainers] dh(1)-based packaging pushed to unstable, signed my old git tags

Christian Seiler christian at iwakd.de
Sat May 2 23:31:35 UTC 2015

Hash: SHA256

Hi Ritesh,

just so you know: I've now pushed 3 things to git:

1. I've signed my old git tags.

Since the commits they reference remained the same, you should be able
to just fetch them (git fetch --tags), but just as a heads-up in case
git doesn't like it and you have to force-fetch the tags.

2. I've re-added the gbp.conf you added previously (I had removed it
during the Jessie freeze because the release team would never have
accepted that change), with the additional setting of meta = True in the
[dch] section - that makes Closes: #xxx work properly.

and finally:

On 04/24/2015 04:12 PM, Ritesh Raj Sarraf wrote:
> On Thursday 23 April 2015 05:19 PM, Christian Seiler wrote:
>> - switch to dh(1) instead of manual calls for dh_*, and override 
>> where necessary
>> This should make the rules file quite a bit easier, and
>> immediately fix #777602 (dh_installinit call deprecated). Also,
>> future improvements to debhelper will automatically be picked
>> up.
> Yes. We should transition to the newer tools.

3. I've completely rewritten debian/rules to use dh(1) instead of manual
calls to dh_*. Also, I'm now using debian/*.install instead of manual
calls to install in the rules file. This simplifies things quite a bit,
since one now doesn't have to check if we are building against an arch
that supports udebs or not (debhelper does that for us).

I've compared the generated binary packages from the version in Jessie
and my changes with dh(1) on amd64 using the debbindiff tool from the
reproducible build people. Since I'm not really running a reproducible
build toolchain, timestamps and build ids change all over the place, but
otherwise (except for postinst where dh_installinit's generated version
is slightly different but has the same effect, see #777602) everything
is completely identical.

Therefore, I'm sure that my packaging changes don't affect the generated

If you have any comments / objections to my changes to debian/rules,
please tell me.

The next things I'll do are going to be:

 - While playing around with the build, I've noticed that upstream's
   Makefiles don't properly respect external build-flags in a sufficient
   manner. I've seen that you have already patched some parts to make
   hardening work better, but it's not complete. I'll try to fix the
   Makefiles so that all build flags are properly respected.

 - Do the other stuff I mentioned in my original email (systemd,
   umounting logic, but triaging, etc.)


PS: I found Bug #784092 while switching to dh(1), you might want to take
a look at that. It's fixed with the current changes in git, but as I
mentioned in that bug, it might be a good idea to do a jessie-pu.
Version: GnuPG v2


More information about the Pkg-iscsi-maintainers mailing list