On Fri, Apr 24, 2020 at 11:12 PM Ondrej Mosnacek
<omosnace(a)redhat.com> wrote:
> On Fri, Apr 24, 2020 at 8:50 PM Ondrej Mosnacek <omosnace(a)redhat.com> wrote:
> > On Wed, Apr 22, 2020 at 10:12 AM Daiki Ueno <ueno(a)fedoraproject.org>
wrote:
> > > Hello,
> > >
> > > I am not sure if this deserves a Fedora Change proposal, so I'd like
to
> > > hear any opinions first before proceeding with the process.
> > >
> > > NSS (the crypto library used by Firefox) historically supports 2
> > > database formats: SQLite and DBM. The latter is considered legacy and
> > > we switched the default database format to SQLite in F28[1]. Since then
> > > I presume most of the applications have switched to the new format.
> > > Therefore we are planning to phase out the support of DBM, targetting
> > > F33+.
> > >
> > > Please let me know if there is any concern.
> >
> > It seems this broke the kernel build. I did some scratch build today
> > to test some patches, but it failed with this:
> >
> > + /usr/bin/pesign -c 'Red Hat Test Certificate' --certdir
> > /etc/pki/pesign-rh-test -i arch/x86/boot/bzImage -o vmlinuz.signed -s
> > pesign: Could not initialize nss.
> > NSS says "The certificate/key database is in an old, unsupported
> > format." errno says "No such file or directory"
> > error: Bad exit status from /var/tmp/rpm-tmp.YKqoK0 (%build)
> > RPM build errors:
> > Bad exit status from /var/tmp/rpm-tmp.YKqoK0 (%build)
> > Child return code was: 1
>
> Probably related:
https://github.com/rhboot/pesign/issues/34
I filed a bug against pesign here:
https://bugzilla.redhat.com/show_bug.cgi?id=1827902
Good catch, and thank you for filing the bug. For the meantime I
reverted the DBM disablement to unblock the kernel package build: