This is a note to let you know that I've just added the patch titled
orangefs: fix deadlock; do not write i_size in read_iter
to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git%3Ba=su...
The filename of the patch is: orangefs-fix-deadlock-do-not-write-i_size-in-read_iter.patch and it can be found in the queue-4.14 subdirectory.
If you, or anyone else, feels it should not be added to the stable tree, please let stable@vger.kernel.org know about it.
From 6793f1c450b1533a5e9c2493490de771d38b24f9 Mon Sep 17 00:00:00 2001
From: Martin Brandenburg martin@omnibond.com Date: Thu, 25 Jan 2018 19:39:44 -0500 Subject: orangefs: fix deadlock; do not write i_size in read_iter
From: Martin Brandenburg martin@omnibond.com
commit 6793f1c450b1533a5e9c2493490de771d38b24f9 upstream.
After do_readv_writev, the inode cache is invalidated anyway, so i_size will never be read. It will be fetched from the server which will also know about updates from other machines.
Fixes deadlock on 32-bit SMP.
See https://marc.info/?l=linux-fsdevel&m=151268557427760&w=2
Signed-off-by: Martin Brandenburg martin@omnibond.com Cc: Al Viro viro@zeniv.linux.org.uk Cc: Mike Marshall hubcap@omnibond.com Signed-off-by: Linus Torvalds torvalds@linux-foundation.org Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org
--- fs/orangefs/file.c | 7 ++----- fs/orangefs/orangefs-kernel.h | 11 ----------- 2 files changed, 2 insertions(+), 16 deletions(-)
--- a/fs/orangefs/file.c +++ b/fs/orangefs/file.c @@ -446,7 +446,7 @@ ssize_t orangefs_inode_read(struct inode static ssize_t orangefs_file_read_iter(struct kiocb *iocb, struct iov_iter *iter) { struct file *file = iocb->ki_filp; - loff_t pos = *(&iocb->ki_pos); + loff_t pos = iocb->ki_pos; ssize_t rc = 0;
BUG_ON(iocb->private); @@ -486,9 +486,6 @@ static ssize_t orangefs_file_write_iter( } }
- if (file->f_pos > i_size_read(file->f_mapping->host)) - orangefs_i_size_write(file->f_mapping->host, file->f_pos); - rc = generic_write_checks(iocb, iter);
if (rc <= 0) { @@ -502,7 +499,7 @@ static ssize_t orangefs_file_write_iter( * pos to the end of the file, so we will wait till now to set * pos... */ - pos = *(&iocb->ki_pos); + pos = iocb->ki_pos;
rc = do_readv_writev(ORANGEFS_IO_WRITE, file, --- a/fs/orangefs/orangefs-kernel.h +++ b/fs/orangefs/orangefs-kernel.h @@ -566,17 +566,6 @@ do { \ sys_attr.mask = ORANGEFS_ATTR_SYS_ALL_SETABLE; \ } while (0)
-static inline void orangefs_i_size_write(struct inode *inode, loff_t i_size) -{ -#if BITS_PER_LONG == 32 && defined(CONFIG_SMP) - inode_lock(inode); -#endif - i_size_write(inode, i_size); -#if BITS_PER_LONG == 32 && defined(CONFIG_SMP) - inode_unlock(inode); -#endif -} - static inline void orangefs_set_timeout(struct dentry *dentry) { unsigned long time = jiffies + orangefs_dcache_timeout_msecs*HZ/1000;
Patches currently in stable-queue which might be from martin@omnibond.com are
queue-4.14/orangefs-fix-deadlock-do-not-write-i_size-in-read_iter.patch queue-4.14/orangefs-initialize-op-on-loop-restart-in-orangefs_devreq_read.patch queue-4.14/orangefs-use-list_for_each_entry_safe-in-purge_waiting_ops.patch