All of lore.kernel.org
 help / color / mirror / Atom feed
From: Chuck Lever <chuck.lever@oracle.com>
To: linux-rdma@vger.kernel.org, linux-nfs@vger.kernel.org
Subject: [PATCH v1 05/11] xprtrdma: Allocate Protection Domain in rpcrdma_ep_create()
Date: Fri, 21 Feb 2020 17:00:33 -0500	[thread overview]
Message-ID: <20200221220033.2072.22880.stgit@manet.1015granger.net> (raw)
In-Reply-To: <20200221214906.2072.32572.stgit@manet.1015granger.net>

Make a Protection Domain (PD) a per-connection resource rather than
a per-transport resource. In other words, when the connection
terminates, the PD is destroyed.

Thus there is one less HW resource that remains allocated to a
transport after a connection is closed.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
---
 net/sunrpc/xprtrdma/verbs.c |   26 ++++++++++----------------
 1 file changed, 10 insertions(+), 16 deletions(-)

diff --git a/net/sunrpc/xprtrdma/verbs.c b/net/sunrpc/xprtrdma/verbs.c
index f361213a8157..36fe7baea014 100644
--- a/net/sunrpc/xprtrdma/verbs.c
+++ b/net/sunrpc/xprtrdma/verbs.c
@@ -363,14 +363,6 @@ static void rpcrdma_update_cm_private(struct rpcrdma_xprt *r_xprt,
 		rc = PTR_ERR(ia->ri_id);
 		goto out_err;
 	}
-
-	ia->ri_pd = ib_alloc_pd(ia->ri_id->device, 0);
-	if (IS_ERR(ia->ri_pd)) {
-		rc = PTR_ERR(ia->ri_pd);
-		pr_err("rpcrdma: ib_alloc_pd() returned %d\n", rc);
-		goto out_err;
-	}
-
 	return 0;
 
 out_err:
@@ -403,9 +395,6 @@ static void rpcrdma_update_cm_private(struct rpcrdma_xprt *r_xprt,
 
 	rpcrdma_ep_destroy(r_xprt);
 
-	ib_dealloc_pd(ia->ri_pd);
-	ia->ri_pd = NULL;
-
 	/* Allow waiters to continue */
 	complete(&ia->ri_remove_done);
 
@@ -423,11 +412,6 @@ static void rpcrdma_update_cm_private(struct rpcrdma_xprt *r_xprt,
 	if (ia->ri_id && !IS_ERR(ia->ri_id))
 		rdma_destroy_id(ia->ri_id);
 	ia->ri_id = NULL;
-
-	/* If the pd is still busy, xprtrdma missed freeing a resource */
-	if (ia->ri_pd && !IS_ERR(ia->ri_pd))
-		ib_dealloc_pd(ia->ri_pd);
-	ia->ri_pd = NULL;
 }
 
 static int rpcrdma_ep_create(struct rpcrdma_xprt *r_xprt,
@@ -514,6 +498,12 @@ static int rpcrdma_ep_create(struct rpcrdma_xprt *r_xprt,
 	ep->rep_remote_cma.flow_control = 0;
 	ep->rep_remote_cma.rnr_retry_count = 0;
 
+	ia->ri_pd = ib_alloc_pd(id->device, 0);
+	if (IS_ERR(ia->ri_pd)) {
+		rc = PTR_ERR(ia->ri_pd);
+		goto out_destroy;
+	}
+
 	rc = rdma_create_qp(id, ia->ri_pd, &ep->rep_attr);
 	if (rc)
 		goto out_destroy;
@@ -540,6 +530,10 @@ static void rpcrdma_ep_destroy(struct rpcrdma_xprt *r_xprt)
 	if (ep->rep_attr.send_cq)
 		ib_free_cq(ep->rep_attr.send_cq);
 	ep->rep_attr.send_cq = NULL;
+
+	if (ia->ri_pd)
+		ib_dealloc_pd(ia->ri_pd);
+	ia->ri_pd = NULL;
 }
 
 /* Re-establish a connection after a device removal event.


  parent reply	other threads:[~2020-02-21 22:00 UTC|newest]

Thread overview: 22+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-02-21 22:00 [PATCH v1 00/11] NFS/RDMA client side connection overhaul Chuck Lever
2020-02-21 22:00 ` [PATCH v1 01/11] xprtrdma: Invoke rpcrdma_ep_create() in the connect worker Chuck Lever
2020-02-21 22:00 ` [PATCH v1 02/11] xprtrdma: Refactor frwr_init_mr() Chuck Lever
2020-02-21 22:00 ` [PATCH v1 03/11] xprtrdma: Clean up the post_send path Chuck Lever
2020-02-21 22:00 ` [PATCH v1 04/11] xprtrdma: Refactor rpcrdma_ep_connect() and rpcrdma_ep_disconnect() Chuck Lever
2020-02-21 22:00 ` Chuck Lever [this message]
2020-03-01 18:11   ` [PATCH v1 05/11] xprtrdma: Allocate Protection Domain in rpcrdma_ep_create() Tom Talpey
2020-03-01 18:29     ` Chuck Lever
2020-03-01 18:38       ` Tom Talpey
2020-02-21 22:00 ` [PATCH v1 06/11] xprtrdma: Invoke rpcrdma_ia_open in the connect worker Chuck Lever
2020-02-21 22:00 ` [PATCH v1 07/11] xprtrdma: Remove rpcrdma_ia::ri_flags Chuck Lever
2020-02-21 22:00 ` [PATCH v1 08/11] xprtrdma: Disconnect on flushed completion Chuck Lever
2020-02-21 22:00 ` [PATCH v1 09/11] xprtrdma: Merge struct rpcrdma_ia into struct rpcrdma_ep Chuck Lever
2020-02-21 22:01 ` [PATCH v1 10/11] xprtrdma: Extract sockaddr from struct rdma_cm_id Chuck Lever
2020-02-24 16:15   ` Anna Schumaker
2020-02-24 16:18     ` Chuck Lever
2020-02-24 16:23       ` Anna Schumaker
2020-02-21 22:01 ` [PATCH v1 11/11] xprtrdma: kmalloc rpcrdma_ep separate from rpcrdma_xprt Chuck Lever
2020-03-01 18:09 ` [PATCH v1 00/11] NFS/RDMA client side connection overhaul Tom Talpey
2020-03-01 18:12   ` Chuck Lever
2020-03-11 15:27 ` Chuck Lever
2020-03-11 17:16   ` Schumaker, Anna

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=20200221220033.2072.22880.stgit@manet.1015granger.net \
    --to=chuck.lever@oracle.com \
    --cc=linux-nfs@vger.kernel.org \
    --cc=linux-rdma@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 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.