call for testers

Brad Fitzpatrick brad at danga.com
Tue Sep 5 00:10:29 UTC 2006


Okay:

http://danga.com/memcached/dist/experimental/memcached-1.2.x-svn359.tar.gz

That should solve your libevent detection problem, as well as fix the two
warnings.

Will look into the test failures....


On Tue, 5 Sep 2006, Jacques Caron wrote:

> Hi,
>
> Tried on FreeBSD 5.4 i386 and amd64 and I can't even compile the
> thing out of the box. Configure complains it can't find libevent even
> when it's present, and log seems to indicate prefix is "" even if I
> try to force it to /usr/local, which is supposed to be the default anyway:
>
> configure:2996: checking for event_set in -levent
> configure:3026: gcc -o conftest -g -O2 -I/include   -L/lib conftest.c
> -levent   >&5
> /usr/bin/ld: cannot find -levent
>
> I guess the problem is somewhere around:
>
> # Check whether --with-libevent or --without-libevent was given.
> if test "${with_libevent+set}" = set; then
>    withval="$with_libevent"
>
> fi;
> if test "$with_libevent" != "no"; then
>          CFLAGS="$CFLAGS -I$with_libevent/include"
>          LDFLAGS="$LDFLAGS -L$with_libevent/lib"
> fi
>
> ./configure --with-libevent=/usr/local works, though, but it should
> default to $prefix...
>
> Compilation gives these warnings:
>
> items.c: In function `item_link_q':
> items.c:143: warning: comparison is always true due to limited range
> of data type
> items.c: In function `item_unlink_q':
> items.c:161: warning: comparison is always true due to limited range
> of data type
>
> Once compiled, prove test fails (and hangs), and prove -v test says:
>
>  > prove -v test
> test/00-startup.........1..1
> ok 1 - started the server
> ok
> test/getset.............1..14
> not ok 1 - stored foo
>
> #     Failed test (test/getset.t at line 14)
> #          got: undef
> #     expected: 'STORED
> # '
> not ok 2 - foo == 'fooval'
>
> #     Failed test
> (/usr/home/jc/memcached-1.2.x-svn357/test/lib/MemcachedTest.pm at line 52)
> #          got: undef
> #     expected: 'VALUE foo 0 6
> # fooval
> # END
> # '
> (and hangs there forever)
>
> Everything exactly the same on i386 and amd64.
>
> Let me know if you need any further testing...
>
> Jacques.
>
> At 23:37 04/09/2006, Brad Fitzpatrick wrote:
> >Before I do a release, can I get some people running the test suite on
> >this snapshot?
> >
> >http://danga.com/memcached/dist/experimental/memcached-1.2.x-svn357.tar.gz
> >
> >Instructions:
> >
> >   -- ./configure
> >   -- make
> >   -- prove test
> >      -- if all good (you see at bottom:  "All tests successful, ...."),
> >         let me know.
> >      -- if bad, reply with "prove -v test", which will show verbose
> >         per-test output
> >
> >Looking for:
> >
> >   -- Various Linux Distros (I've tried Debian and CentOS)
> >   -- Solaris 10 (my Nexenta VM is failing to boot again)
> >   -- FreeBSD
> >   -- OS X
> >   -- 64-bit whatevers (I did test it on 64-bit CentOS)
> >
> >Let me know of any quirks, even if they're not part of the tests.
> >
> >- Brad
>
>
>


More information about the memcached mailing list