lwip-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[lwip-devel] [bug #27316] netconn: Possible deadlock in err_tcp


From: Simon Goldschmidt
Subject: [lwip-devel] [bug #27316] netconn: Possible deadlock in err_tcp
Date: Tue, 25 Aug 2009 14:25:27 +0000
User-agent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.4; de; rv:1.9.1.2) Gecko/20090729 Firefox/3.5.2

URL:
  <http://savannah.nongnu.org/bugs/?27316>

                 Summary: netconn: Possible deadlock in err_tcp
                 Project: lwIP - A Lightweight TCP/IP stack
            Submitted by: goldsimon
            Submitted on: Di 25 Aug 2009 14:25:26 GMT
                Category: TCP
                Severity: 3 - Normal
              Item Group: Crash Error
                  Status: None
                 Privacy: Public
             Assigned to: None
             Open/Closed: Open
         Discussion Lock: Any
         Planned Release: 1.4.0
            lwIP version: CVS Head

    _______________________________________________________

Details:

As reported by Eran Rundstein on lwip-users, there may be a deadlock when
err_tcp is called and the recvmbox or acceptmbox is full.

The reason is sys_mbox_post is used instead of sys_mbox_trypost. This might
block the tcpip_thread, which should be avoided.

Possible solutions are
- Require the platform port to always allow NULL to be posted at least once,
then we could use sys_mbox_trypost
- Use another mechanism to pass the info (that the connection is closed) to
the application - could be harder to implement as we may not block reading on
the mbox in the application thread (since there might be no more data coming)




    _______________________________________________________

Reply to this item at:

  <http://savannah.nongnu.org/bugs/?27316>

_______________________________________________
  Nachricht geschickt von/durch Savannah
  http://savannah.nongnu.org/





reply via email to

[Prev in Thread] Current Thread [Next in Thread]