[openib-general] RFC: detecting duplicate MAD requests

[Hal Rosenstock]

On Mon, 2006-06-12 at 19:43, Sean Hefty wrote:
> Hal Rosenstock wrote:
> >>This brings up a concern.  There doesn't seem to be a limit to the number of 
> >>received MADs that can be queued for a user.  Should we have such a limit?
> > 
> > 
> > How are MADs counted ? Is a multisegment MAD 1 MAD or multiple MADs ? If
> > the latter, it seems problematic to limit this as the response to a get
> > response might be very large.
> 
> I could go either way, or use a hybrid of some sort  Counting a multisegment MAD 
> as 1 MAD might be a little easier.  We could also allow something like 100 
> segments, or at least 1 reassembled MAD.  So, the user could have 100 single 
> segment MADs, 50 2-segment MADs, etc.

This seems prone to introducing a different problem and that this number
would need to scale with the size of the subnet.

> Without some sort of restriction, a userspace app that's slow to pull receive 
> MADs from the kernel would result in consuming a large amount of kernel memory.

Understood but dropping a MAD after acknowledging also seems like a bad
thing to me. Couldn't this be controlled on the request side (assuming
the request has a response as opposed to unsolicited sends/receives) ?

-- Hal

> - Sean