Your message dated Fri, 07 Feb 2014 16:21:57 +0000
with message-id <52f50825.1000...@pwned.gg>
and subject line undo inappropriate changes
has caused the Debian Bug report #721544,
regarding libboost-thread1.54.0: should link against Boost.Atomics on some
architectures
to be marked as done.
This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.
(NB: If you are a system administrator and have no idea what this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact ow...@bugs.debian.org
immediately.)
--
721544: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=721544
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems
--- Begin Message ---
Package: libboost-thread1.54.0
Version: 1.54.0-2
Severity: serious
Justification: Policy 10.2
On some architectures (at least ia64 and s390[x]), libboost_thread
needs to link against libboost_atomics to avoid winding up with an
undefined reference to boost::atomics::detail::lockpool::get_lock_for.
Otherwise, some packages (such as gnuradio) will FTBFS:
/usr/lib/gcc/ia64-linux-gnu/4.6/../../../libboost_thread.so: undefined
reference to `boost::atomics::detail::lockpool::get_lock_for(void const
volatile*)'
collect2: ld returned 1 exit status
make[3]: *** [gnuradio-runtime/lib/gr_runtime_test] Error 1
It may be simplest to have it do so unconditionally.
Could you please take a look?
Thanks!
--- End Message ---
--- Begin Message ---
Control: unmerge 721577
Control: unblock 704032
Control: reassign -1 boost1.54
Control: found -1 1.54.0-2
Control: notfound -1 1.54.0-3
Control: fixed 1.54.0-3
OK, I've sorted out this clusterfuck of merges.
#721544 and #721577 are strictly separate bugs.
#721544 was about how a Boost internal library itself didn't link to one of its
dependencies within Boost. This was boost ticket #9041 and has been fixed:
https://svn.boost.org/trac/boost/ticket/9041 and was fixed in Debian, the
original reason why this bug was closed.
#721577 is how cpp-netlib has a faulty build system that links some tests
statically. However, some Boost internal libs (namely the one mentioned above
from #721544) have architecture-dependant dependencies that need to be linked
differently. The build system for cpp-netlib does not do this, causing the bug.
This is an upstream issue.
I think it's this latter bug that should actually be blocking #704032 - I will
make this change separately.
--
GPG: 4096R/1318EFAC5FBBDBCE
git://github.com/infinity0/pubkeys.git
signature.asc
Description: OpenPGP digital signature
--- End Message ---
