All of lore.kernel.org
 help / color / mirror / Atom feed
From: Hannes Reinecke <hare@suse.de>
To: "Martin K. Petersen" <martin.petersen@oracle.com>
Cc: Christoph Hellwig <hch@lst.de>,
	Doug Gilbert <dgilbert@interlog.com>,
	Daniel Wagner <daniel.wagner@suse.com>,
	James Bottomley <james.bottomley@hansenpartnership.com>,
	linux-scsi@vger.kernel.org, Hannes Reinecke <hare@suse.de>
Subject: [PATCH 6/6] scsi: avoid pointless memory allocation in scsi_alloc_target()
Date: Tue,  2 Jun 2020 13:33:11 +0200	[thread overview]
Message-ID: <20200602113311.121513-7-hare@suse.de> (raw)
In-Reply-To: <20200602113311.121513-1-hare@suse.de>

As we already know the index of the requested target in
scsi_alloc_target() we can try to fetch it first before trying
to allocate a new one. This will save us a pointless memory
allocation in case the target is already present and uncontended.

Signed-off-by: Hannes Reinecke <hare@suse.de>
---
 drivers/scsi/scsi_scan.c | 21 ++++++++++++++++++---
 1 file changed, 18 insertions(+), 3 deletions(-)

diff --git a/drivers/scsi/scsi_scan.c b/drivers/scsi/scsi_scan.c
index 8afb71c036d5..41818111808e 100644
--- a/drivers/scsi/scsi_scan.c
+++ b/drivers/scsi/scsi_scan.c
@@ -387,7 +387,7 @@ static void scsi_target_reap_ref_put(struct scsi_target *starget)
  * is responsible for both reaping and doing a last put
  */
 static struct scsi_target *scsi_alloc_target(struct device *parent,
-					     int channel, uint id)
+					     u16 channel, u16 id)
 {
 	struct Scsi_Host *shost = dev_to_shost(parent);
 	struct device *dev = NULL;
@@ -397,7 +397,23 @@ static struct scsi_target *scsi_alloc_target(struct device *parent,
 	struct scsi_target *starget;
 	struct scsi_target *found_target;
 	int error, ref_got;
-	unsigned long tid;
+	unsigned long tid = (channel << 16) | id;
+
+	/*
+	 * Try if the target is already present, and save us
+	 * a pointless memory allocation if so.
+	 */
+	spin_lock_irqsave(shost->host_lock, flags);
+	found_target = xa_load(&shost->__targets, tid);
+	if (found_target) {
+		get_device(&found_target->dev);
+		if (kref_get_unless_zero(&found_target->reap_ref)) {
+			spin_unlock_irqrestore(shost->host_lock, flags);
+			return found_target;
+		}
+		put_device(&found_target->dev);
+	}
+	spin_unlock_irqrestore(shost->host_lock, flags);
 
 	starget = kzalloc(size, GFP_KERNEL);
 	if (!starget) {
@@ -418,7 +434,6 @@ static struct scsi_target *scsi_alloc_target(struct device *parent,
 	starget->state = STARGET_CREATED;
 	starget->scsi_level = SCSI_2;
 	starget->max_target_blocked = SCSI_DEFAULT_TARGET_BLOCKED;
-	tid = scsi_target_index(starget);
  retry:
 	spin_lock_irqsave(shost->host_lock, flags);
 	found_target = xa_load(&shost->__targets, tid);
-- 
2.16.4


  parent reply	other threads:[~2020-06-02 11:34 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-06-02 11:33 [PATCHv4 0/6] scsi: use xarray for devices and targets Hannes Reinecke
2020-06-02 11:33 ` [PATCH 1/6] scsi: convert target lookup to xarray Hannes Reinecke
2020-06-02 11:33 ` [PATCH 2/6] target_core_pscsi: use __scsi_device_lookup() Hannes Reinecke
2020-06-02 11:33 ` [PATCH 3/6] scsi: move target device list to xarray Hannes Reinecke
2020-06-02 11:33 ` [PATCH 4/6] scsi: remove direct device lookup per host Hannes Reinecke
2020-06-02 11:33 ` [PATCH 5/6] scsi_error: use xarray lookup instead of wrappers Hannes Reinecke
2020-06-02 11:33 ` Hannes Reinecke [this message]
2020-06-03 12:53 ` [PATCHv4 0/6] scsi: use xarray for devices and targets Christoph Hellwig
2020-06-03 18:23   ` Douglas Gilbert
2020-06-04 16:12   ` Hannes Reinecke

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=20200602113311.121513-7-hare@suse.de \
    --to=hare@suse.de \
    --cc=daniel.wagner@suse.com \
    --cc=dgilbert@interlog.com \
    --cc=hch@lst.de \
    --cc=james.bottomley@hansenpartnership.com \
    --cc=linux-scsi@vger.kernel.org \
    --cc=martin.petersen@oracle.com \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.