fnctl: f_modown should call write_lock_irqsave/restore

Commit 703625118069f9f8960d356676662d3db5a9d116 exposed that f_modown()
should call write_lock_irqsave instead of just write_lock_irq so that
because a caller could have a spinlock held and it would not be good to
renable interrupts.

Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Tavis Ormandy <taviso@google.com>
Cc: stable <stable@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

authored by Greg Kroah-Hartman and committed by Linus Torvalds b04da8bf 9a3cbe32

+4 -2
+4 -2
fs/fcntl.c
··· 199 static void f_modown(struct file *filp, struct pid *pid, enum pid_type type, 200 int force) 201 { 202 - write_lock_irq(&filp->f_owner.lock); 203 if (force || !filp->f_owner.pid) { 204 put_pid(filp->f_owner.pid); 205 filp->f_owner.pid = get_pid(pid); ··· 213 filp->f_owner.euid = cred->euid; 214 } 215 } 216 - write_unlock_irq(&filp->f_owner.lock); 217 } 218 219 int __f_setown(struct file *filp, struct pid *pid, enum pid_type type,
··· 199 static void f_modown(struct file *filp, struct pid *pid, enum pid_type type, 200 int force) 201 { 202 + unsigned long flags; 203 + 204 + write_lock_irqsave(&filp->f_owner.lock, flags); 205 if (force || !filp->f_owner.pid) { 206 put_pid(filp->f_owner.pid); 207 filp->f_owner.pid = get_pid(pid); ··· 211 filp->f_owner.euid = cred->euid; 212 } 213 } 214 + write_unlock_irqrestore(&filp->f_owner.lock, flags); 215 } 216 217 int __f_setown(struct file *filp, struct pid *pid, enum pid_type type,