On 10/21/19 2:48 PM, Matthew Bobrowski wrote:
This patch updates the lock pattern in ext4_dio_read_iter() to only
perform the trylock in IOCB_NOWAIT cases.
Signed-off-by: Matthew Bobrowski <mbobrowski@xxxxxxxxxxxxxx>
This shall need a rebase (as mentioned in patch-1 discussion).
It will be good to mention that the locking condition
is kept similar to 942491c9e6d6 ("xfs: fix AIM7 regression").
Also, I think it may also need a fixes tag.
It seems unconditional inode_lock_shared was added by below commit.
Fixes: 16c54688592c ("ext4: Allow parallel DIO reads")
Otherwise, this patch looks good to me. You may add:
Reviewed-by: Ritesh Harjani <riteshh@xxxxxxxxxxxxx>
---
fs/ext4/file.c | 8 +++++++-
1 file changed, 7 insertions(+), 1 deletion(-)
diff --git a/fs/ext4/file.c b/fs/ext4/file.c
index 6ea7e00e0204..8420686b90f5 100644
--- a/fs/ext4/file.c
+++ b/fs/ext4/file.c
@@ -52,7 +52,13 @@ static int ext4_dio_read_iter(struct kiocb *iocb, struct iov_iter *to)
ssize_t ret;
struct inode *inode = file_inode(iocb->ki_filp);
- inode_lock_shared(inode);
+ if (iocb->ki_flags & IOCB_NOWAIT) {
+ if (!inode_trylock_shared(inode))
+ return -EAGAIN;
+ } else {
+ inode_lock_shared(inode);
+ }
+
if (!ext4_dio_supported(inode)) {
inode_unlock_shared(inode);
/*
