f2fs:Fix locking requirements in the function recover_dentry
PATCH 1/1 f2fs:Fix locking requirements in the function recover_dentry
This fixes locking requirements in the function recover_dentry to properly lock the critical region between the labels retry and out with the read/write semaphone of the f2fs's superblock representation in order to protect against other concurrent threads of execution accessing the same regions of memory used by the code in this critical
--- fs/f2fs/recovery.c | 4 ++++ 1 file changed, 4 insertions(+)
fs/f2fs/recovery.c +4 -0
@@ -75,6 +75,7 @@ static int recover_dentry(struct inode *inode, struct page *ipage)struct qstr name;struct page *page;struct inode *dir, *einode; +struct f2fs_sb_info *sbi =F2FS_I_SB (inode);int err =0 ; dir =f2fs_iget (inode->i_sb, pino);
@@ -97,6 +98,7 @@ static int recover_dentry(struct inode *inode, struct page *ipage) goto out_err; } retry: +f2fs_lock_op (sbi); de =f2fs_find_entry (dir, &name, &page);if (de && inode->i_ino ==le32_to_cpu (de->ino)) goto out_unmap_put;
@@ -117,6 +119,7 @@ retry: }f2fs_delete_entry (de, page, dir, einode);iput (einode); +f2fs_unlock_op (sbi); goto retry; } err =__f2fs_add_link (dir, &name, inode, inode->i_ino, inode->i_mode);
@@ -138,6 +141,7 @@ out_unmap_put: out_err:iput (dir); out: +f2fs_unlock_op (sbi); f2fs_msg(inode->i_sb, KERN_NOTICE, "%s: ino = %x, name = %s, dir = %lx, err = %d", __func__, ino_of_node(ipage), raw_inode->i_name,
Signature · 4 lines
2.1.4 ------------------------------------------------------------------------------
Quoted 8 lines
-----Original Message-----
From: Nicholas Krause [mailto:[email protected]]
Sent: Sunday, September 06, 2015 1:27 AM
To: [email protected]
Cc: [email protected]; [email protected]; [email protected]; [email protected]
Subject: [PATCH] f2fs:Fix locking requirements in the functionrecover_dentry
Quoted 4 lines
This fixes locking requirements in the function
recover_dentryto properly lock the critical region between the labels retry and out with the read/write semaphone of the f2fs's superblock representation in order to protect against other concurrent threads of execution
The protecting is not needed, I've already relayed in last thread, please refer to the link below:
http://www.gossamer-threads.com/lists/linux/kernel/2234556?do=post_view_threaded#2234556
Thanks,