I think it makes more sense to handle this the same way I handled
max_sectors: make it a per-target parameter passed in when connecting
to the target.  We could make cmds_per_lun a similar parameter, but
are there likely to be any SRP targets that need this to be limited?
Also, what is max_targets?

  
OK we can do the same way that you handled max_sectors.
SRP targets may prefer a specific cmds_per_lun to reach max sequential performance.
max_targets == max_id

  
    > + fix the bug of reuse the iu while it's still in_use

I think I see the bug: a send may complete and have its IU recycled
before the corresponding command is completed, and end up screwing
things up.  Is this right?
  
Yes
If so I would prefer to fix things in a slightly different way.
Rather than a TX ring, we should just keep a list of free IUs ready to
send and only add IUs to the end of the list when we're really done
with the IU.
  
A free list of IUs is fine also

Vu

_______________________________________________
openib-general mailing list
[email protected]
http://openib.org/mailman/listinfo/openib-general

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

Reply via email to