[<prev] [next>] [day] [month] [year] [list]
Message-Id: <20200214154854.6746-32-sashal@kernel.org>
Date: Fri, 14 Feb 2020 10:40:24 -0500
From: Sasha Levin <sashal@...nel.org>
To: linux-kernel@...r.kernel.org, stable@...r.kernel.org
Cc: Ritesh Harjani <riteshh@...ux.ibm.com>, Jan Kara <jack@...e.cz>,
Matthew Bobrowski <mbobrowski@...browski.org>,
Joseph Qi <joseph.qi@...ux.alibaba.com>,
Theodore Ts'o <tytso@....edu>, Sasha Levin <sashal@...nel.org>,
linux-ext4@...r.kernel.org
Subject: [PATCH AUTOSEL 5.5 032/542] ext4: fix ext4_dax_read/write inode locking sequence for IOCB_NOWAIT
From: Ritesh Harjani <riteshh@...ux.ibm.com>
[ Upstream commit f629afe3369e9885fd6e9cc7a4f514b6a65cf9e9 ]
Apparently our current rwsem code doesn't like doing the trylock, then
lock for real scheme. So change our dax read/write methods to just do the
trylock for the RWF_NOWAIT case.
This seems to fix AIM7 regression in some scalable filesystems upto ~25%
in some cases. Claimed in commit 942491c9e6d6 ("xfs: fix AIM7 regression")
Reviewed-by: Jan Kara <jack@...e.cz>
Reviewed-by: Matthew Bobrowski <mbobrowski@...browski.org>
Tested-by: Joseph Qi <joseph.qi@...ux.alibaba.com>
Signed-off-by: Ritesh Harjani <riteshh@...ux.ibm.com>
Link: https://lore.kernel.org/r/20191212055557.11151-2-riteshh@linux.ibm.com
Signed-off-by: Theodore Ts'o <tytso@....edu>
Signed-off-by: Sasha Levin <sashal@...nel.org>
---
fs/ext4/file.c | 10 ++++++----
1 file changed, 6 insertions(+), 4 deletions(-)
diff --git a/fs/ext4/file.c b/fs/ext4/file.c
index 6a7293a5cda2d..977ac58dc718d 100644
--- a/fs/ext4/file.c
+++ b/fs/ext4/file.c
@@ -88,9 +88,10 @@ static ssize_t ext4_dax_read_iter(struct kiocb *iocb, struct iov_iter *to)
struct inode *inode = file_inode(iocb->ki_filp);
ssize_t ret;
- if (!inode_trylock_shared(inode)) {
- if (iocb->ki_flags & IOCB_NOWAIT)
+ if (iocb->ki_flags & IOCB_NOWAIT) {
+ if (!inode_trylock_shared(inode))
return -EAGAIN;
+ } else {
inode_lock_shared(inode);
}
/*
@@ -487,9 +488,10 @@ ext4_dax_write_iter(struct kiocb *iocb, struct iov_iter *from)
bool extend = false;
struct inode *inode = file_inode(iocb->ki_filp);
- if (!inode_trylock(inode)) {
- if (iocb->ki_flags & IOCB_NOWAIT)
+ if (iocb->ki_flags & IOCB_NOWAIT) {
+ if (!inode_trylock(inode))
return -EAGAIN;
+ } else {
inode_lock(inode);
}
--
2.20.1
Powered by blists - more mailing lists