From: Chuck Lever <chuck.lever@oracle.com>
To: linux-rdma@vger.kernel.org, linux-nfs@vger.kernel.org
Subject: [PATCH v1 01/14] xprtrdma: Fix DMAR failure in frwr_op_map() after reconnect
Date: Wed, 09 Nov 2016 14:04:56 -0500 [thread overview]
Message-ID: <20161109190456.15007.87544.stgit@manet.1015granger.net> (raw)
In-Reply-To: <20161109184735.15007.96507.stgit@manet.1015granger.net>
When a LOCALINV WR is flushed, the frmr is marked STALE, then
frwr_op_unmap_sync DMA-unmaps the frmr's SGL. These STALE frmrs
are then recovered when frwr_op_map hunts for an INVALID frmr to
use.
All other cases that need frmr recovery leave that SGL DMA-mapped.
The FRMR recovery path unconditionally DMA-unmaps the frmr's SGL.
To avoid DMA unmapping the SGL twice for flushed LOCAL_INV WRs,
alter the recovery logic (rather than the hot frwr_op_unmap_sync
path) to distinguish among these cases. This solution also takes
care of the case where multiple LOCAL_INV WRs are issued for the
same rpcrdma_req, some complete successfully, but some are flushed.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Tested-by: Vasco Steinmetz <linux@kyberraum.net>
---
net/sunrpc/xprtrdma/frwr_ops.c | 37 ++++++++++++++++++++++---------------
net/sunrpc/xprtrdma/xprt_rdma.h | 3 ++-
2 files changed, 24 insertions(+), 16 deletions(-)
diff --git a/net/sunrpc/xprtrdma/frwr_ops.c b/net/sunrpc/xprtrdma/frwr_ops.c
index 2109495..26b26be 100644
--- a/net/sunrpc/xprtrdma/frwr_ops.c
+++ b/net/sunrpc/xprtrdma/frwr_ops.c
@@ -44,18 +44,20 @@
* being done.
*
* When the underlying transport disconnects, MRs are left in one of
- * three states:
+ * four states:
*
* INVALID: The MR was not in use before the QP entered ERROR state.
- * (Or, the LOCAL_INV WR has not completed or flushed yet).
- *
- * STALE: The MR was being registered or unregistered when the QP
- * entered ERROR state, and the pending WR was flushed.
*
* VALID: The MR was registered before the QP entered ERROR state.
*
- * When frwr_op_map encounters STALE and VALID MRs, they are recovered
- * with ib_dereg_mr and then are re-initialized. Beause MR recovery
+ * FLUSHED_FR: The MR was being registered when the QP entered ERROR
+ * state, and the pending WR was flushed.
+ *
+ * FLUSHED_LI: The MR was being invalidated when the QP entered ERROR
+ * state, and the pending WR was flushed.
+ *
+ * When frwr_op_map encounters FLUSHED and VALID MRs, they are recovered
+ * with ib_dereg_mr and then are re-initialized. Because MR recovery
* allocates fresh resources, it is deferred to a workqueue, and the
* recovered MRs are placed back on the rb_mws list when recovery is
* complete. frwr_op_map allocates another MR for the current RPC while
@@ -177,12 +179,15 @@
static void
frwr_op_recover_mr(struct rpcrdma_mw *mw)
{
+ enum rpcrdma_frmr_state state = mw->frmr.fr_state;
struct rpcrdma_xprt *r_xprt = mw->mw_xprt;
struct rpcrdma_ia *ia = &r_xprt->rx_ia;
int rc;
rc = __frwr_reset_mr(ia, mw);
- ib_dma_unmap_sg(ia->ri_device, mw->mw_sg, mw->mw_nents, mw->mw_dir);
+ if (state != FRMR_FLUSHED_LI)
+ ib_dma_unmap_sg(ia->ri_device,
+ mw->mw_sg, mw->mw_nents, mw->mw_dir);
if (rc)
goto out_release;
@@ -262,10 +267,8 @@
}
static void
-__frwr_sendcompletion_flush(struct ib_wc *wc, struct rpcrdma_frmr *frmr,
- const char *wr)
+__frwr_sendcompletion_flush(struct ib_wc *wc, const char *wr)
{
- frmr->fr_state = FRMR_IS_STALE;
if (wc->status != IB_WC_WR_FLUSH_ERR)
pr_err("rpcrdma: %s: %s (%u/0x%x)\n",
wr, ib_wc_status_msg(wc->status),
@@ -288,7 +291,8 @@
if (wc->status != IB_WC_SUCCESS) {
cqe = wc->wr_cqe;
frmr = container_of(cqe, struct rpcrdma_frmr, fr_cqe);
- __frwr_sendcompletion_flush(wc, frmr, "fastreg");
+ frmr->fr_state = FRMR_FLUSHED_FR;
+ __frwr_sendcompletion_flush(wc, "fastreg");
}
}
@@ -308,7 +312,8 @@
if (wc->status != IB_WC_SUCCESS) {
cqe = wc->wr_cqe;
frmr = container_of(cqe, struct rpcrdma_frmr, fr_cqe);
- __frwr_sendcompletion_flush(wc, frmr, "localinv");
+ frmr->fr_state = FRMR_FLUSHED_LI;
+ __frwr_sendcompletion_flush(wc, "localinv");
}
}
@@ -328,8 +333,10 @@
/* WARNING: Only wr_cqe and status are reliable at this point */
cqe = wc->wr_cqe;
frmr = container_of(cqe, struct rpcrdma_frmr, fr_cqe);
- if (wc->status != IB_WC_SUCCESS)
- __frwr_sendcompletion_flush(wc, frmr, "localinv");
+ if (wc->status != IB_WC_SUCCESS) {
+ frmr->fr_state = FRMR_FLUSHED_LI;
+ __frwr_sendcompletion_flush(wc, "localinv");
+ }
complete(&frmr->fr_linv_done);
}
diff --git a/net/sunrpc/xprtrdma/xprt_rdma.h b/net/sunrpc/xprtrdma/xprt_rdma.h
index 0d35b76..6e1bba3 100644
--- a/net/sunrpc/xprtrdma/xprt_rdma.h
+++ b/net/sunrpc/xprtrdma/xprt_rdma.h
@@ -216,7 +216,8 @@ struct rpcrdma_rep {
enum rpcrdma_frmr_state {
FRMR_IS_INVALID, /* ready to be used */
FRMR_IS_VALID, /* in use */
- FRMR_IS_STALE, /* failed completion */
+ FRMR_FLUSHED_FR, /* flushed FASTREG WR */
+ FRMR_FLUSHED_LI, /* flushed LOCALINV WR */
};
struct rpcrdma_frmr {
next prev parent reply other threads:[~2016-11-09 19:04 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-11-09 19:04 [PATCH v1 00/14] client-side NFS/RDMA patches proposed for v4.10 Chuck Lever
2016-11-09 19:04 ` Chuck Lever [this message]
2016-11-09 19:05 ` [PATCH v1 02/14] xprtrdma: Cap size of callback buffer resources Chuck Lever
2016-11-09 19:05 ` [PATCH v1 03/14] xprtrdma: Make FRWR send queue entry accounting more accurate Chuck Lever
2016-11-09 19:05 ` [PATCH v1 04/14] xprtrdma: Support for SG_GAP devices Chuck Lever
2016-11-09 19:05 ` [PATCH v1 05/14] SUNRPC: Proper metric accounting when RPC is not transmitted Chuck Lever
2016-11-09 19:05 ` [PATCH v1 06/14] xprtrdma: Address coverity complaint about wait_for_completion() Chuck Lever
2016-11-09 19:05 ` [PATCH v1 07/14] xprtrdma: Avoid calls to ro_unmap_safe() Chuck Lever
2016-11-09 19:05 ` [PATCH v1 08/14] xprtrdma: Squelch "max send, max recv" messages at connect time Chuck Lever
2016-11-09 19:06 ` [PATCH v1 09/14] xprtrdma: Shorten QP access error message Chuck Lever
2016-11-09 19:06 ` [PATCH v1 10/14] xprtrdma: Update dprintk in rpcrdma_count_chunks Chuck Lever
2016-11-09 19:06 ` [PATCH v1 11/14] xprtrdma: Relocate connection helper functions Chuck Lever
2016-11-09 19:06 ` [PATCH v1 12/14] xprtrdma: Simplify synopsis of rpcrdma_ep_connect() Chuck Lever
2016-11-09 19:06 ` [PATCH v1 13/14] xprtrdma: Refactor FRMR invalidation Chuck Lever
2016-11-09 19:06 ` [PATCH v1 14/14] xprtrdma: Update documenting comment Chuck Lever
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=20161109190456.15007.87544.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 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).