[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: rsyslog 7.6.2 and liblogging-stdlog0 (unstable)



On Fri, Mar 28, 2014 at 08:49:38PM +1300, Michael Cree wrote:
> On Mon, Mar 24, 2014 at 07:31:26AM -0500, Bob Tracy wrote:
> > It could be as simple as liblogging-stdlog0 not having been put in the
> > queue to build, but the upgrade of rsyslog to version 7.6.2-1 is being
> > held pending the availability of the liblogging-stdlog0 package. 
> 
> That's strange.  According to the wanna-build server liblogging is built
> and installed on Alpha, but download the Alpha packages file and
> liblogging-stdlog0 is not present in it.
> 
> And browsing to http://ftp.debian-ports.org/debian/pool-alpha/main/libl/
> one can see that there is no liblogging directory.  

Now fixed.  You should be able to upgrade rsyslog and other packages. The
rebuild of software against libdb5.3 is finished so you should be able to
remove libdb5.1 without any problems now.

However there are a number of problems holding up rebuilding the Gnome
software.  The longest standing one is libpeas fails to build [1] with a
failure in the test suite.  Gedit depends on libpeas thus is no longer
installable.  But more recently libav and glib2.0 have started to FTBFS
and they are holding up the rebuilding of quite a few Gnome packages.

libav fails with a link time relocation error [2]. But it is not the known
bug with relaxation as --no-relax does not fix it.  There does not appear
to be any missing -fPIC in building shareable libraries either. So I am
a little bit mystified by the failure.  The version in experimental builds
okay, but I do not know how long we have to wait before that is uploaded
to unstable.

Glib2.0 fails in the test suite on the async-splice-output-stream
test [3].

And eglibc version 2.18 FTBFS with test suite failures in tests such as
tst-rwlock6 [4].  Interestingly upstream glibc v2.18 builds okay and
passes the test suite, and I can't get it to fail, even if I apply as
many as Debian patches as possible and use as many of the Debian
configure options as possible. Even if I remove as many of the Debian
patches as possible from the Debian source I can't get it to pass the
test suite.  So I am a bit stumped on this one too.

Maybe those many "Alpha Porters" who responded to the Release Teams'
Architecture Roll Call, would like to take a serious look at these
failures.

Cheers
Michael.

[1] http://buildd.debian-ports.org/libpeas
[2] http://buildd.debian-ports.org/libav
[3] http://buildd.debian-ports.org/glib2.0
[4] http://buildd.debian-ports.org/eglibc


Reply to: