Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
From:   Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To:     linux-kernel@vger.kernel.org
Cc:     Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        stable@vger.kernel.org,
        Trond Myklebust <trond.myklebust@hammerspace.com>,
        Sasha Levin <sashal@kernel.org>
Subject: [PATCH 5.12 219/292] NFSv4/pnfs: Fix the layout barrier update
Date:   Mon, 19 Jul 2021 16:54:41 +0200
Message-Id: <20210719144950.183875874@linuxfoundation.org>
In-Reply-To: <20210719144942.514164272@linuxfoundation.org>
References: <20210719144942.514164272@linuxfoundation.org>
User-Agent: quilt/0.66
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Precedence: bulk

From: Trond Myklebust <trond.myklebust@hammerspace.com>

[ Upstream commit aa95edf309ef31e2df4a37ebf0e5c2ca2a6772ab ]

If we have multiple outstanding layoutget requests, the current code to
update the layout barrier assumes that the outstanding layout stateids
are updated in order. That's not necessarily the case.

Instead of using the value of lo->plh_outstanding as a guesstimate for
the window of values we need to accept, just wait to update the window
until we're processing the last one. The intention here is just to
ensure that we don't process 2^31 seqid updates without also updating
the barrier.

Fixes: 1bcf34fdac5f ("pNFS/NFSv4: Update the layout barrier when we schedule a layoutreturn")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 fs/nfs/pnfs.c | 30 +++++++++++++++---------------
 1 file changed, 15 insertions(+), 15 deletions(-)

diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c
index a5e3628a5273..cd041ba6f0aa 100644
--- a/fs/nfs/pnfs.c
+++ b/fs/nfs/pnfs.c
@@ -966,10 +966,8 @@ void
 pnfs_set_layout_stateid(struct pnfs_layout_hdr *lo, const nfs4_stateid *new,
 			const struct cred *cred, bool update_barrier)
 {
-	u32 oldseq, newseq, new_barrier = 0;
-
-	oldseq = be32_to_cpu(lo->plh_stateid.seqid);
-	newseq = be32_to_cpu(new->seqid);
+	u32 oldseq = be32_to_cpu(lo->plh_stateid.seqid);
+	u32 newseq = be32_to_cpu(new->seqid);
 
 	if (!pnfs_layout_is_valid(lo)) {
 		pnfs_set_layout_cred(lo, cred);
@@ -979,19 +977,21 @@ pnfs_set_layout_stateid(struct pnfs_layout_hdr *lo, const nfs4_stateid *new,
 		clear_bit(NFS_LAYOUT_INVALID_STID, &lo->plh_flags);
 		return;
 	}
-	if (pnfs_seqid_is_newer(newseq, oldseq)) {
+
+	if (pnfs_seqid_is_newer(newseq, oldseq))
 		nfs4_stateid_copy(&lo->plh_stateid, new);
-		/*
-		 * Because of wraparound, we want to keep the barrier
-		 * "close" to the current seqids.
-		 */
-		new_barrier = newseq - atomic_read(&lo->plh_outstanding);
-	}
-	if (update_barrier)
-		new_barrier = be32_to_cpu(new->seqid);
-	else if (new_barrier == 0)
+
+	if (update_barrier) {
+		pnfs_barrier_update(lo, newseq);
 		return;
-	pnfs_barrier_update(lo, new_barrier);
+	}
+	/*
+	 * Because of wraparound, we want to keep the barrier
+	 * "close" to the current seqids. We really only want to
+	 * get here from a layoutget call.
+	 */
+	if (atomic_read(&lo->plh_outstanding) == 1)
+		 pnfs_barrier_update(lo, be32_to_cpu(lo->plh_stateid.seqid));
 }
 
 static bool
-- 
2.30.2