Re: [PATCH] IB/srp receive buffer handling robustness improvement

2010-08-04 Thread Roland Dreier
Thanks, I combined these patches and also got rid of __srp_post_recv() since we don't need the split anymore... so I queued the patch below. (Bart, by the way, this patch didn't have a "---" after the changelog part, and also didn't have a diffstat for the patch. If you use git, then "git format-p

Re: [PATCH] IB/srp receive buffer handling robustness improvement

2010-07-30 Thread David Dillow
On Fri, 2010-07-30 at 12:59 +0200, Bart Van Assche wrote: > The current strategy in ib_srp for posting receive buffers is as follows: > * Post one buffer after channel establishment. > * Post one buffer before sending an SRP_CMD or SRP_TSK_MGMT to the target. > As a result, only the first non-SRP_R

[PATCH] IB/srp receive buffer handling robustness improvement

2010-07-30 Thread Bart Van Assche
The current strategy in ib_srp for posting receive buffers is as follows: * Post one buffer after channel establishment. * Post one buffer before sending an SRP_CMD or SRP_TSK_MGMT to the target. As a result, only the first non-SRP_RSP information unit from the target will be processed. If that fir