Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:18 as permitted sender) client-ip=2620:137:e000::1:18;
From:   =?UTF-8?q?Lu=C3=ADs=20Henriques?= <lhenriques@suse.de>
To:     Jeff Layton <jlayton@kernel.org>, Xiubo Li <xiubli@redhat.com>,
        Ilya Dryomov <idryomov@gmail.com>
Cc:     ceph-devel@vger.kernel.org, linux-kernel@vger.kernel.org,
        =?UTF-8?q?Lu=C3=ADs=20Henriques?= <lhenriques@suse.de>
Subject: [PATCH v3] ceph: invalidate pages when doing direct/sync writes
Date:   Thu,  7 Apr 2022 15:38:34 +0100
Message-Id: <20220407143834.7516-1-lhenriques@suse.de>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Precedence: bulk

When doing a direct/sync write, we need to invalidate the page cache in
the range being written to.  If we don't do this, the cache will include
invalid data as we just did a write that avoided the page cache.

Signed-off-by: Luís Henriques <lhenriques@suse.de>
---
 fs/ceph/file.c | 9 +++++++++
 1 file changed, 9 insertions(+)

Ok, here's a new attempt.  After discussion in this thread and on IRC, I
think this is the right fix.  generic/647 now passes with and without
encryption.  Thanks!

Changes since v2:
- Invalidation needs to be done after a write

Changes since v1:
- Replaced truncate_inode_pages_range() by invalidate_inode_pages2_range
- Call fscache_invalidate with FSCACHE_INVAL_DIO_WRITE if we're doing DIO

diff --git a/fs/ceph/file.c b/fs/ceph/file.c
index 5072570c2203..63e67eb60310 100644
--- a/fs/ceph/file.c
+++ b/fs/ceph/file.c
@@ -1938,6 +1938,15 @@ ceph_sync_write(struct kiocb *iocb, struct iov_iter *from, loff_t pos,
 			break;
 		}
 		ceph_clear_error_write(ci);
+		ret = invalidate_inode_pages2_range(
+				inode->i_mapping,
+				pos >> PAGE_SHIFT,
+				(pos + len - 1) >> PAGE_SHIFT);
+		if (ret < 0) {
+			dout("invalidate_inode_pages2_range returned %d\n",
+			     ret);
+			ret = 0;
+		}
 		pos += len;
 		written += len;
 		dout("sync_write written %d\n", written);