Re: [PATCH v3] target: Drop iSCSI use of mutex around max_cmd_sn increment

2015-08-03 Thread Nicholas A. Bellinger
On Thu, 2015-07-23 at 14:53 -0700, Spencer Baugh wrote: > From: Roland Dreier > > In a performance profile, taking a mutex in iscsit_increment_maxcmdsn() > shows up very high. However taking a mutex around "sess->max_cmd_sn += 1" > seems pretty silly: we're not serializing against other

Re: [PATCH v3] target: Drop iSCSI use of mutex around max_cmd_sn increment

2015-08-03 Thread Nicholas A. Bellinger
On Thu, 2015-07-23 at 14:53 -0700, Spencer Baugh wrote: From: Roland Dreier rol...@purestorage.com In a performance profile, taking a mutex in iscsit_increment_maxcmdsn() shows up very high. However taking a mutex around sess-max_cmd_sn += 1 seems pretty silly: we're not serializing against

Re: [PATCH v3] target: Drop iSCSI use of mutex around max_cmd_sn increment

2015-07-23 Thread Spencer Baugh
From: Roland Dreier In a performance profile, taking a mutex in iscsit_increment_maxcmdsn() shows up very high. However taking a mutex around "sess->max_cmd_sn += 1" seems pretty silly: we're not serializing against other contexts in any useful way. I did a quick audit and there don't appear

Re: [PATCH v3] target: Drop iSCSI use of mutex around max_cmd_sn increment

2015-07-23 Thread Spencer Baugh
From: Roland Dreier rol...@purestorage.com In a performance profile, taking a mutex in iscsit_increment_maxcmdsn() shows up very high. However taking a mutex around sess-max_cmd_sn += 1 seems pretty silly: we're not serializing against other contexts in any useful way. I did a quick audit and