From: Bart Van Assche <bart.vanassche@wdc.com>
To: target-devel@vger.kernel.org
Subject: [PATCH 07/20] target: Document when CMD_T_STOP and CMD_T_COMPLETE are set
Date: Fri, 22 Jun 2018 21:52:54 +0000 [thread overview]
Message-ID: <20180622215307.8758-8-bart.vanassche@wdc.com> (raw)
Document those aspects of transport_cmd_check_stop_to_fabric() and
transport_generic_free_cmd() of which it is nontrivial to derive
these from their implementation.
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com>
Reviewed-by: Mike Christie <mchristi@redhat.com>
Cc: Hannes Reinecke <hare@suse.com>
Cc: Christoph Hellwig <hch@lst.de>
---
drivers/target/target_core_transport.c | 34 +++++++++++++++++++++-----
1 file changed, 28 insertions(+), 6 deletions(-)
diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c
index 087dddfc5c35..47318f64296d 100644
--- a/drivers/target/target_core_transport.c
+++ b/drivers/target/target_core_transport.c
@@ -632,6 +632,13 @@ static void target_remove_from_state_list(struct se_cmd *cmd)
spin_unlock_irqrestore(&dev->execute_task_lock, flags);
}
+/*
+ * This function is called by the target core after the target core has
+ * finished processing a SCSI command or SCSI TMF. Both the regular command
+ * processing code and the code for aborting commands can call this
+ * function. CMD_T_STOP is set if and only if another thread is waiting
+ * inside transport_wait_for_tasks() for t_transport_stop_comp.
+ */
static int transport_cmd_check_stop_to_fabric(struct se_cmd *cmd)
{
unsigned long flags;
@@ -2624,6 +2631,27 @@ static void target_wait_free_cmd(struct se_cmd *cmd, bool *aborted, bool *tas)
spin_unlock_irqrestore(&cmd->t_state_lock, flags);
}
+/*
+ * This function is called by frontend drivers after processing of a command
+ * has finished.
+ *
+ * The protocol for ensuring that either the regular flow or the TMF
+ * code drops one reference is as follows:
+ * - Calling .queue_data_in(), .queue_status() or queue_tm_rsp() will cause
+ * the frontend driver to drop one reference, synchronously or asynchronously.
+ * - During regular command processing the target core sets CMD_T_COMPLETE
+ * before invoking one of the .queue_*() functions.
+ * - The code that aborts commands skips commands and TMFs for which
+ * CMD_T_COMPLETE has been set.
+ * - CMD_T_ABORTED is set atomically after the CMD_T_COMPLETE check for
+ * commands that will be aborted.
+ * - If the CMD_T_ABORTED flag is set but CMD_T_TAS has not been set
+ * transport_generic_free_cmd() skips its call to target_put_sess_cmd().
+ * - For aborted commands for which CMD_T_TAS has been set .queue_status() will
+ * be called and will drop a reference.
+ * - For aborted commands for which CMD_T_TAS has not been set .aborted_task()
+ * will be called. transport_cmd_finish_abort() will drop the final reference.
+ */
int transport_generic_free_cmd(struct se_cmd *cmd, int wait_for_tasks)
{
int ret = 0;
@@ -2652,12 +2680,6 @@ int transport_generic_free_cmd(struct se_cmd *cmd, int wait_for_tasks)
if (!aborted || tas)
ret = target_put_sess_cmd(cmd);
}
- /*
- * If the task has been internally aborted due to TMR ABORT_TASK
- * or LUN_RESET, target_core_tmr.c is responsible for performing
- * the remaining calls to target_put_sess_cmd(), and not the
- * callers of this function.
- */
if (aborted) {
pr_debug("Detected CMD_T_ABORTED for ITT: %llu\n", cmd->tag);
wait_for_completion(&cmd->cmd_wait_comp);
--
2.17.1
reply other threads:[~2018-06-22 21:52 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20180622215307.8758-8-bart.vanassche@wdc.com \
--to=bart.vanassche@wdc.com \
--cc=target-devel@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).