linux-nfs.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Chuck Lever <chuck.lever@oracle.com>
To: bfields@fieldses.org
Cc: linux-rdma@vger.kernel.org, linux-nfs@vger.kernel.org
Subject: [PATCH v2 3/9] svcrdma: svc_rdma_put_context() is invoked twice in Send error path
Date: Wed, 04 May 2016 10:53:05 -0400	[thread overview]
Message-ID: <20160504145305.6602.70505.stgit@klimt.1015granger.net> (raw)
In-Reply-To: <20160504144730.6602.89073.stgit@klimt.1015granger.net>

Get a fresh op_ctxt in send_reply() instead of in svc_rdma_sendto().
This ensures that svc_rdma_put_context() is invoked only once if
send_reply() fails.

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
---
 net/sunrpc/xprtrdma/svc_rdma_sendto.c |   28 +++++++++++++---------------
 1 file changed, 13 insertions(+), 15 deletions(-)

diff --git a/net/sunrpc/xprtrdma/svc_rdma_sendto.c b/net/sunrpc/xprtrdma/svc_rdma_sendto.c
index 4f1b1c4..54d53330 100644
--- a/net/sunrpc/xprtrdma/svc_rdma_sendto.c
+++ b/net/sunrpc/xprtrdma/svc_rdma_sendto.c
@@ -463,25 +463,21 @@ static int send_reply(struct svcxprt_rdma *rdma,
 		      struct svc_rqst *rqstp,
 		      struct page *page,
 		      struct rpcrdma_msg *rdma_resp,
-		      struct svc_rdma_op_ctxt *ctxt,
 		      struct svc_rdma_req_map *vec,
 		      int byte_count)
 {
+	struct svc_rdma_op_ctxt *ctxt;
 	struct ib_send_wr send_wr;
 	u32 xdr_off;
 	int sge_no;
 	int sge_bytes;
 	int page_no;
 	int pages;
-	int ret;
-
-	ret = svc_rdma_repost_recv(rdma, GFP_KERNEL);
-	if (ret) {
-		svc_rdma_put_context(ctxt, 0);
-		return -ENOTCONN;
-	}
+	int ret = -EIO;
 
 	/* Prepare the context */
+	ctxt = svc_rdma_get_context(rdma);
+	ctxt->direction = DMA_TO_DEVICE;
 	ctxt->pages[0] = page;
 	ctxt->count = 1;
 
@@ -565,8 +561,7 @@ static int send_reply(struct svcxprt_rdma *rdma,
  err:
 	svc_rdma_unmap_dma(ctxt);
 	svc_rdma_put_context(ctxt, 1);
-	pr_err("svcrdma: failed to send reply, rc=%d\n", ret);
-	return -EIO;
+	return ret;
 }
 
 void svc_rdma_prep_reply_hdr(struct svc_rqst *rqstp)
@@ -585,7 +580,6 @@ int svc_rdma_sendto(struct svc_rqst *rqstp)
 	int ret;
 	int inline_bytes;
 	struct page *res_page;
-	struct svc_rdma_op_ctxt *ctxt;
 	struct svc_rdma_req_map *vec;
 
 	dprintk("svcrdma: sending response for rqstp=%p\n", rqstp);
@@ -598,8 +592,6 @@ int svc_rdma_sendto(struct svc_rqst *rqstp)
 	rp_ary = svc_rdma_get_reply_array(rdma_argp, wr_ary);
 
 	/* Build an req vec for the XDR */
-	ctxt = svc_rdma_get_context(rdma);
-	ctxt->direction = DMA_TO_DEVICE;
 	vec = svc_rdma_get_req_map(rdma);
 	ret = svc_rdma_map_xdr(rdma, &rqstp->rq_res, vec, wr_ary != NULL);
 	if (ret)
@@ -635,7 +627,12 @@ int svc_rdma_sendto(struct svc_rqst *rqstp)
 		inline_bytes -= ret;
 	}
 
-	ret = send_reply(rdma, rqstp, res_page, rdma_resp, ctxt, vec,
+	/* Post a fresh Receive buffer _before_ sending the reply */
+	ret = svc_rdma_post_recv(rdma, GFP_KERNEL);
+	if (ret)
+		goto err1;
+
+	ret = send_reply(rdma, rqstp, res_page, rdma_resp, vec,
 			 inline_bytes);
 	if (ret < 0)
 		goto err1;
@@ -648,7 +645,8 @@ int svc_rdma_sendto(struct svc_rqst *rqstp)
 	put_page(res_page);
  err0:
 	svc_rdma_put_req_map(rdma, vec);
-	svc_rdma_put_context(ctxt, 0);
+	pr_err("svcrdma: Could not send reply, err=%d. Closing transport.\n",
+	       ret);
 	set_bit(XPT_CLOSE, &rdma->sc_xprt.xpt_flags);
 	return -ENOTCONN;
 }


  parent reply	other threads:[~2016-05-04 14:53 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-05-04 14:52 [PATCH v2 0/9] NFS/RDMA server patches proposed for v4.7 Chuck Lever
2016-05-04 14:52 ` [PATCH v2 1/9] svcrdma: Support IPv6 with NFS/RDMA Chuck Lever
2016-05-04 14:52 ` [PATCH v2 2/9] svcrdma: Do not add XDR padding to xdr_buf page vector Chuck Lever
2016-05-04 14:53 ` Chuck Lever [this message]
2016-05-04 14:53 ` [PATCH v2 4/9] svcrdma: Remove superfluous line from rdma_read_chunks() Chuck Lever
2016-05-04 14:53 ` [PATCH v2 5/9] svcrdma: Post Receives only for forward channel requests Chuck Lever
2016-05-04 14:53 ` [PATCH v2 6/9] svcrdma: Drain QP before freeing svcrdma_xprt Chuck Lever
2016-05-04 14:53 ` [PATCH v2 7/9] svcrdma: Eliminate code duplication in svc_rdma_recvfrom() Chuck Lever
2016-05-04 14:53 ` [PATCH v2 8/9] svcrdma: Generalize svc_rdma_xdr_decode_req() Chuck Lever
2016-05-04 14:53 ` [PATCH v2 9/9] svcrdma: Simplify the check for backward direction replies Chuck Lever
2016-05-13 20:12 ` [PATCH v2 0/9] NFS/RDMA server patches proposed for v4.7 J. Bruce Fields

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=20160504145305.6602.70505.stgit@klimt.1015granger.net \
    --to=chuck.lever@oracle.com \
    --cc=bfields@fieldses.org \
    --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 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).