quetz-mod_python-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jim Gallacher (JIRA)" <j...@apache.org>
Subject [jira] Created: (MODPYTHON-182) Memory leak in request readline()
Date Fri, 11 Aug 2006 20:50:13 GMT
Memory leak in request readline()
---------------------------------

                 Key: MODPYTHON-182
                 URL: http://issues.apache.org/jira/browse/MODPYTHON-182
             Project: mod_python
          Issue Type: Bug
          Components: core
    Affects Versions: 3.2.10, 3.3, 3.2.x
         Environment: Apache 2.0.55 mpm-worker
            Reporter: Jim Gallacher
         Assigned To: Jim Gallacher


There is at least one memory leak request.readline(). I'm currently auditing the code so there
may well be others.

I think the leak will only occur when the request body is incompletely read with readline,
which is likely an unusual corner case. The following demostrates the leak:

Consider a request body consisting of:

'\n'.join([ 'a'*100 for i in xrange(0, 100) ])

and using this handler:


def handler(req):
    # Handler reads 20 lines but request contains 100 lines.
    # Since the body is not completely read, the buffer allocated
    # in requestobject.c req_readline is never freed.
    req.content_type = 'text/plain'
    count = 0
    for i in range(0, 20):
        line = req.readline()
        count += 1

    req.write('ok readline_partial: %d lines read' % count)
    return apache.OK

With this test setup I'm seeing a leak of approx 10k bytes per request.



-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

Mime
View raw message