> -----Original Message----- > From: linux-kernel-owner@xxxxxxxxxxxxxxx [mailto:linux-kernel- > owner@xxxxxxxxxxxxxxx] On Behalf Of Spencer Baugh > Sent: Wednesday, July 22, 2015 5:08 PM > Subject: [PATCH] target: Drop iSCSI use of mutex around max_cmd_sn > increment ... > diff --git a/drivers/target/iscsi/iscsi_target_configfs.c > b/drivers/target/iscsi/iscsi_target_configfs.c > index c1898c8..29d5930 100644 > --- a/drivers/target/iscsi/iscsi_target_configfs.c > +++ b/drivers/target/iscsi/iscsi_target_configfs.c > @@ -706,8 +706,8 @@ static ssize_t lio_target_nacl_show_info( > rb += sprintf(page+rb, " 0x%08x 0x%08x 0x%08x 0x%08x" > " 0x%08x 0x%08x\n", > sess->cmdsn_window, > - (sess->max_cmd_sn - sess->exp_cmd_sn) + 1, > - sess->exp_cmd_sn, sess->max_cmd_sn, > + ((u32) atomic_read(&sess->max_cmd_sn) - sess- > >exp_cmd_sn) + 1, > + sess->exp_cmd_sn, (u32) atomic_read(&sess->max_cmd_sn), > sess->init_task_tag, sess->targ_xfer_tag); > rb += sprintf(page+rb, "----------------------[iSCSI" > " Connections]-------------------------\n"); Two calls to atomic_read could pick up different values; calling it once and using the same value twice would ensure the arguments are consistent with each other. > diff --git a/drivers/target/iscsi/iscsi_target_device.c > b/drivers/target/iscsi/iscsi_target_device.c > index 5fabcd3..a526904 100644 > --- a/drivers/target/iscsi/iscsi_target_device.c > +++ b/drivers/target/iscsi/iscsi_target_device.c ... > @@ -57,9 +57,7 @@ void iscsit_increment_maxcmdsn(struct iscsi_cmd *cmd, > struct iscsi_session *sess > > cmd->maxcmdsn_inc = 1; > > - mutex_lock(&sess->cmdsn_mutex); > - sess->max_cmd_sn += 1; > - pr_debug("Updated MaxCmdSN to 0x%08x\n", sess->max_cmd_sn); > - mutex_unlock(&sess->cmdsn_mutex); > + atomic_inc(&sess->max_cmd_sn); > + pr_debug("Updated MaxCmdSN to 0x%08x\n", atomic_read(&sess- > >max_cmd_sn)); > } > EXPORT_SYMBOL(iscsit_increment_maxcmdsn); If there is another change before the atomic_read, this would print a value unrelated to the increment done by this thread. atomic_inc_return would provide the appropriate value to print. --- Robert Elliott, HP Server Storage -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
