I can always reproduce the by creating the race condition. I am trying to
hit a deadline this week. I will try to create a test code that reproduces
the same which you can run to test it. I will have it to you hopefully by
the weekend.

Bharath 

-----Original Message-----
From: Roland Dreier [mailto:[EMAIL PROTECTED] 
Sent: Wednesday, November 07, 2007 1:09 PM
To: Bharath Ramesh
Cc: OFA-General
Subject: Re: [ofa-general] IBV_WC_STATUS status set as 135

 > I think the problem that was my SRQ buffer was hitting low watermark and
> before I could post additional buffers to the SRQ my QP was going into  >
an error state, a race condition that I had missed.

There's still something fishy -- I don't see how ibv_poll_cq() could ever
set wc.status to 135.  Do you understand why you weren't getting a valid
error code?

 - R.

_______________________________________________
general mailing list
[email protected]
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general

To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general

Reply via email to