win32

["David N. Welton" <da...@dedasys.com>]

Pat was nice enough to have another look at what was going with
windows, and filed this bug report:

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=21422

    The mod_rivet.so binary that is built by the patches submitted as
    bug 21336 crashes under apache 1.3.27. The symptoms are that under
    normal use the server is stuck in a loop. What is happening is
    that the master apache server launches the child process which
    will create a number of threads to service the incoming
    requests. This child process is then crashing due to a
    configuration problem and the master notices this and then creates
    a new child and on and on.

    The crash is occurring in Rivet_InitTclStuff at mod_rivet.c:714
    where memory is allocated for the cache using ap_pcalloc. The
    default value of rsc->cache_size * sizeof(char *) overflows a
    signed int and becomes negative. ap_pcalloc takes a signed int and
    passes this int to memset which raises an access violation.

    The workaround is to use the RivetServerConf CacheSize
    directive. Values up to 50 appear to be fine.

    The error seems to be in mod_rivet.c:703 where the value
    ap_max_requests_per_child appears to have some gibberish value in
    it.

Interesting (and an excellent bug report).  Do you have
'MaxRequestsPerChild' defined in your config?  If not, the default
ought to be 0.

Thanks again,
-- 
David N. Welton
   Consulting: http://www.dedasys.com/
     Personal: http://www.dedasys.com/davidw/
Free Software: http://www.dedasys.com/freesoftware/
   Apache Tcl: http://tcl.apache.org/

---------------------------------------------------------------------
To unsubscribe, e-mail: rivet-dev-unsubscribe@tcl.apache.org
For additional commands, e-mail: rivet-dev-help@tcl.apache.org