[feature] remove support for libdecaf

[miodvallat]

Short description

As suggested in #12953, there is need to support libdecaf any longer.

Checklist

I have:

• read the CONTRIBUTING.md document
• compiled this code
• tested this code (quickly)
• included documentation (including possible behaviour changes)
• documented the code
• added or modified regression test(s)
• added or modified unit test(s)

[coveralls]

Pull Request Test Coverage Report for Build 12182335989

Details

• 1 of 1 (100.0%) changed or added relevant line in 1 file are covered.
• 42 unchanged lines in 11 files lost coverage.
• Overall coverage increased (+0.3%) to 64.712%

---

Files with Coverage Reduction	New Missed Lines	%
pdns/validate.cc	1	68.14%
pdns/dnssecinfra.cc	2	76.3%
pdns/recursordist/sortlist.cc	2	72.94%
pdns/sstuff.hh	2	56.83%
modules/lmdbbackend/lmdbbackend.cc	2	73.83%
pdns/recursordist/test-syncres_cc2.cc	3	88.85%
pdns/stubresolver.cc	3	77.58%
pdns/recursordist/rec-main.cc	3	62.4%
pdns/shuffle.cc	4	39.71%
pdns/recursordist/syncres.cc	10	79.34%

Totals
Change from base Build 12178254343:	0.3%
Covered Lines:	125741
Relevant Lines:	163437

---

💛 - Coveralls

[omoerbeek]

Recursor build failure due to missed auto* case: https://github.com/PowerDNS/pdns/actions/runs/12156498965/job/33900317485?pr=14926

[miodvallat]

I suppose I need to keep an empty m4/pdns_with_libdecaf.m4 for autoreconf to operate correctly.

[omoerbeek]

Hmm, that should not be needed, I'll take a look why autoconf wants that file later today.

[miodvallat]

I agree that this shouldn't be needed, but if keeping a mostly empty file helps, then it doesn't hurt to carry that empty file for a while.

[omoerbeek]

I agree that this shouldn't be needed, but if keeping a mostly empty file helps, then it doesn't hurt to carry that empty file for a while.

Sure, if I don't find the solution quickly, an empty file it is.

[omoerbeek]

Of course the issue doesn't reproduce locally. I avoid gaining too much auto* knowledge, so the proposed workaround it is.

[miodvallat]

I avoid gaining too much auto* knowledge, so the proposed workaround it is.

Of course! You can't have both auto* knowledge and otto* knowledge.

[Habbie]

Recursor build failure due to missed auto* case: https://github.com/PowerDNS/pdns/actions/runs/12156498965/job/33900317485?pr=14926

this is -so weird-

[omoerbeek]

Recursor build failure due to missed auto* case: https://github.com/PowerDNS/pdns/actions/runs/12156498965/job/33900317485?pr=14926

this is -so weird-

caching is indeed hard... does the weirdness prevent merging?

[Habbie]

I figured it out. The right fix is git rm pdns/recursordist/m4/pdns_with_libdecaf.m4, I am quite sure :)

[Habbie]

(no caching was involved at all)

[miodvallat]

Ah, I missed it due it being a symbolic link. Good catch!

[miodvallat]

(no caching was involved at all)

It was DNS.

[omoerbeek]

I suspect I did my test with the workaround already in and didn't pay attention.