[openib-general] [PATCH] user_mad: Add receive side RMPP support

[Hal Rosenstock]

On Thu, 2005-06-30 at 14:45, Roland Dreier wrote:
> Looking at the code, I have to say that I don't like the strategy of
> returning the actual length of the MAD but not copying anything if the
> MAD is too big.  It seems error prone to return a length bigger than
> the user passed in to read(), and it doesn't feel right either.
> 
> I understand and agree with the sentiment of not wanting to add
> another ioctl() to get the length.  Instead, how about returning a
> ib_user_mad_hdr with a status of ENOSPC and putting the actual length
> somewhere.  I'm not sure if it's better to change the ABI and add a
> length field to ib_user_mad_hdr, or if we want to return the first 36
> bytes of an RMPP MAD so the user can figure out the correct length.
> 
> Also, since we check if the MAD at the head of the receive queue won't
> fit in the user's buffer when we're about to pop it off the queue,
> instead of dequeuing and then requeuing it, we can just leave it at
> the head of the queue.  (I'm not sure if this buys us anything but it
> seems there might be a race MADs get out of order)
> 
> Sorry for not bringing this up when you asked about how to handle
> too-big MADs a few days ago -- I didn't understand your question
> really.

You understand it now and answered it well :-) I'm not wedded to the
read approach and you have offered 2 alternatives. 

Unfortunately, I don't think the length in the first segment is reliable
(it works for OpenIB but may not for other RMPP implementations) so I
would rather not rely on that. So I guess this means an ABI change :-(

Do you concur ? If so, I will work on that approach.

-- Hal