Re: [liblo] [ANN] liblo version 0.28 released

On Wed, Feb 5, 2014 at 8:19 AM, Stephen Sinclair <rad...@gm...> wrote:
> Wait a second.. I've just reviewed the source and the build log in a
> bit more detail.

For which version? 0.28-1 or 0.28-2?

>
> In server.c, indeed the result of setting SO_REUSEPORT is actually
> ignored -- and is only tried in a multicast context in any case.

I'm not actually sure about that... that would depend on what does lo_throw do.

>
> Secondly, the build log doesn't seem to indicate anything to me about
> SO_REUSEPORT.  The errors are actually quite varied.  I think the
> SO_REUSEPORT thing may be a red herring, or at least I don't see any
> evidence for it.
>
> Let's see:
>
> armel: seems successful.
>
> armhf: everything seems to go as expected without errors, but then the
> process exits with Error 2.  Any way of getting more information?
>
> ia64: the error is a C++ compile error, which version of g++ is in
> use?  I think the C++11 features may fail on g++4.6, where C++11 was
> only partially implemented yet the lambda test passes in configure.
> That's the only case I
> know of that could cause this problem.

ia64 is going away (inability to update gcc among other issues) so we
can ignore it.

>
> kfreebsd: yes, testlo failed for me when i tested on freebsd in qemu
>
> i386, s390x, powerpc
>> liblo server error 9912 in /: Invalid message received (expected)
>> E: Caught signal ‘Terminated’: terminating immediately
>
> What's that all about?  Does the debian build server send a terminate
> signal to the test process for some reason?

I think you are watching the build logs for 0.28-2, where I patched
out the SO_REUSEPORT section. I actually had a draft reply noting that
the builds still fail, although this seems to be a different reason.
At [1] you can find all build logs.

The build logs for 0.28-1 (on i386 to eliminate strange arch issues)
seem to be regular failures:

===
liblo server error 9912 in /: Invalid message received (expected)

test run not completed
liblo test FAILED
Reply received from osc.udp://206.12.19.115:15954/
Reply received from osc.udp://206.12.19.115:15954/
make[2]: *** [test] Error 1
liblo server error 92 in setsockopt(SO_REUSEPORT): Protocol not
availablemake[2]: Leaving directory `/«PKGBUILDDIR»/src'
make[1]: *** [test] Error 2
make[1]: Leaving directory `/«PKGBUILDDIR»'
make: *** [debian/stamp-makefile-check] Error 2
dpkg-buildpackage: error: debian/rules build-arch gave error exit status 2
===

However for version 0.28-2 I removed the SO_REUSEPORT code and this is
what we get (again i386):

===
Reply sent to osc.udp://206.12.19.115:18934/

liblo server error 9912 in /: Invalid message received (expected)
make[1]: *** wait: No child processes.  Stop.
make[1]: *** Waiting for unfinished jobs....
make[1]: *** wait: No child processes.  Stop.
make: *** [debian/stamp-makefile-check] Error 2
make[2]: *** [test] Terminated
Reply received from osBuild killed with signal TERM after 150 minutes
of inactivity
===

At the end of the log we see the reason: "Build killed with signal
TERM after 150 minutes of inactivity". Perhaps testlo is waiting for a
message that never gets sent? I have to note that the build daemons
have very restricted internet access so it may well not be possible to
access the computer itself over the external IP address. This failure
is different, as the build daemon killed the build after a timeout,
instead of the build actually failing.

Is it possible to tell testlo to only use localhost? That interface is
known to be usable for builds.


[1] https://buildd.debian.org/status/logs.php?pkg=liblo


-- 

Saludos,
Felipe Sateler