diff mbox series

[13/20] filelock: convert __locks_insert_block, conflict and deadlock checks to use file_lock_core

Message ID 20240116-flsplit-v1-13-c9d0f4370a5d@kernel.org
State New
Headers show
Series filelock: split struct file_lock into file_lock and file_lease structs | expand

Commit Message

Jeff Layton Jan. 16, 2024, 7:46 p.m. UTC
Have both __locks_insert_block and the deadlock and conflict checking
functions take a struct file_lock_core pointer instead of a struct
file_lock one. Also, change posix_locks_deadlock to return bool.

Signed-off-by: Jeff Layton <jlayton@kernel.org>
---
 fs/locks.c | 132 ++++++++++++++++++++++++++++++++-----------------------------
 1 file changed, 70 insertions(+), 62 deletions(-)

Comments

NeilBrown Jan. 16, 2024, 10:32 p.m. UTC | #1
On Wed, 17 Jan 2024, Jeff Layton wrote:
> Have both __locks_insert_block and the deadlock and conflict checking
> functions take a struct file_lock_core pointer instead of a struct
> file_lock one. Also, change posix_locks_deadlock to return bool.
> 
> Signed-off-by: Jeff Layton <jlayton@kernel.org>
> ---
>  fs/locks.c | 132 ++++++++++++++++++++++++++++++++-----------------------------
>  1 file changed, 70 insertions(+), 62 deletions(-)
> 


>  
>  /* Must be called with the blocked_lock_lock held! */
> -static int posix_locks_deadlock(struct file_lock *caller_fl,
> -				struct file_lock *block_fl)
> +static bool posix_locks_deadlock(struct file_lock *caller_fl,
> +				 struct file_lock *block_fl)
>  {
> +	struct file_lock_core *caller = &caller_fl->fl_core;
> +	struct file_lock_core *blocker = &block_fl->fl_core;
>  	int i = 0;
> -	struct file_lock_core *flc = &caller_fl->fl_core;
>  
>  	lockdep_assert_held(&blocked_lock_lock);
>  
> @@ -1034,16 +1040,16 @@ static int posix_locks_deadlock(struct file_lock *caller_fl,
>  	 * This deadlock detector can't reasonably detect deadlocks with
>  	 * FL_OFDLCK locks, since they aren't owned by a process, per-se.
>  	 */
> -	if (IS_OFDLCK(flc))
> +	if (IS_OFDLCK(caller))
>  		return 0;

      return false;

Thanks,
NeilBrown
Jeff Layton Jan. 17, 2024, 12:42 p.m. UTC | #2
On Wed, 2024-01-17 at 09:32 +1100, NeilBrown wrote:
> On Wed, 17 Jan 2024, Jeff Layton wrote:
> > Have both __locks_insert_block and the deadlock and conflict checking
> > functions take a struct file_lock_core pointer instead of a struct
> > file_lock one. Also, change posix_locks_deadlock to return bool.
> > 
> > Signed-off-by: Jeff Layton <jlayton@kernel.org>
> > ---
> >  fs/locks.c | 132 ++++++++++++++++++++++++++++++++-----------------------------
> >  1 file changed, 70 insertions(+), 62 deletions(-)
> > 
> 
> 
> >  
> >  /* Must be called with the blocked_lock_lock held! */
> > -static int posix_locks_deadlock(struct file_lock *caller_fl,
> > -				struct file_lock *block_fl)
> > +static bool posix_locks_deadlock(struct file_lock *caller_fl,
> > +				 struct file_lock *block_fl)
> >  {
> > +	struct file_lock_core *caller = &caller_fl->fl_core;
> > +	struct file_lock_core *blocker = &block_fl->fl_core;
> >  	int i = 0;
> > -	struct file_lock_core *flc = &caller_fl->fl_core;
> >  
> >  	lockdep_assert_held(&blocked_lock_lock);
> >  
> > @@ -1034,16 +1040,16 @@ static int posix_locks_deadlock(struct file_lock *caller_fl,
> >  	 * This deadlock detector can't reasonably detect deadlocks with
> >  	 * FL_OFDLCK locks, since they aren't owned by a process, per-se.
> >  	 */
> > -	if (IS_OFDLCK(flc))
> > +	if (IS_OFDLCK(caller))
> >  		return 0;
> 
>       return false;
> 

Good catch. Fixed in my local branch.
diff mbox series

Patch

diff --git a/fs/locks.c b/fs/locks.c
index 6b8e8820dec9..280149860d5e 100644
--- a/fs/locks.c
+++ b/fs/locks.c
@@ -780,42 +780,41 @@  EXPORT_SYMBOL(locks_delete_block);
  * waiters, and add beneath any waiter that blocks the new waiter.
  * Thus wakeups don't happen until needed.
  */
-static void __locks_insert_block(struct file_lock *blocker,
-				 struct file_lock *waiter,
-				 bool conflict(struct file_lock *,
-					       struct file_lock *))
+static void __locks_insert_block(struct file_lock *blocker_fl,
+				 struct file_lock *waiter_fl,
+				 bool conflict(struct file_lock_core *,
+					       struct file_lock_core *))
 {
-	struct file_lock *fl;
-	struct file_lock_core *bflc;
-	BUG_ON(!list_empty(&waiter->fl_core.fl_blocked_member));
+	struct file_lock_core *blocker = &blocker_fl->fl_core;
+	struct file_lock_core *waiter = &waiter_fl->fl_core;
+	struct file_lock_core *flc;
+	BUG_ON(!list_empty(&waiter->fl_blocked_member));
 
 new_blocker:
-	list_for_each_entry(fl, &blocker->fl_core.fl_blocked_requests,
-			    fl_core.fl_blocked_member)
-		if (conflict(fl, waiter)) {
-			blocker =  fl;
+	list_for_each_entry(flc, &blocker->fl_blocked_requests, fl_blocked_member)
+		if (conflict(flc, waiter)) {
+			blocker =  flc;
 			goto new_blocker;
 		}
-	waiter->fl_core.fl_blocker = blocker;
-	list_add_tail(&waiter->fl_core.fl_blocked_member,
-		      &blocker->fl_core.fl_blocked_requests);
+	waiter->fl_blocker = file_lock(blocker);
+	list_add_tail(&waiter->fl_blocked_member,
+		      &blocker->fl_blocked_requests);
 
-	bflc = &blocker->fl_core;
-	if (IS_POSIX(bflc) && !IS_OFDLCK(bflc))
-		locks_insert_global_blocked(&waiter->fl_core);
+	if (IS_POSIX(blocker) && !IS_OFDLCK(blocker))
+		locks_insert_global_blocked(waiter);
 
 	/* The requests in waiter->fl_blocked are known to conflict with
 	 * waiter, but might not conflict with blocker, or the requests
 	 * and lock which block it.  So they all need to be woken.
 	 */
-	__locks_wake_up_blocks(&waiter->fl_core);
+	__locks_wake_up_blocks(waiter);
 }
 
 /* Must be called with flc_lock held. */
 static void locks_insert_block(struct file_lock *blocker,
 			       struct file_lock *waiter,
-			       bool conflict(struct file_lock *,
-					     struct file_lock *))
+			       bool conflict(struct file_lock_core *,
+					     struct file_lock_core *))
 {
 	spin_lock(&blocked_lock_lock);
 	__locks_insert_block(blocker, waiter, conflict);
@@ -872,12 +871,12 @@  locks_delete_lock_ctx(struct file_lock *fl, struct list_head *dispose)
 /* Determine if lock sys_fl blocks lock caller_fl. Common functionality
  * checks for shared/exclusive status of overlapping locks.
  */
-static bool locks_conflict(struct file_lock *caller_fl,
-			   struct file_lock *sys_fl)
+static bool locks_conflict(struct file_lock_core *caller_fl,
+			   struct file_lock_core *sys_fl)
 {
-	if (sys_fl->fl_core.fl_type == F_WRLCK)
+	if (sys_fl->fl_type == F_WRLCK)
 		return true;
-	if (caller_fl->fl_core.fl_type == F_WRLCK)
+	if (caller_fl->fl_type == F_WRLCK)
 		return true;
 	return false;
 }
@@ -885,20 +884,23 @@  static bool locks_conflict(struct file_lock *caller_fl,
 /* Determine if lock sys_fl blocks lock caller_fl. POSIX specific
  * checking before calling the locks_conflict().
  */
-static bool posix_locks_conflict(struct file_lock *caller_fl,
-				 struct file_lock *sys_fl)
+static bool posix_locks_conflict(struct file_lock_core *caller_flc,
+				 struct file_lock_core *sys_flc)
 {
+	struct file_lock *caller_fl = file_lock(caller_flc);
+	struct file_lock *sys_fl = file_lock(sys_flc);
+
 	/* POSIX locks owned by the same process do not conflict with
 	 * each other.
 	 */
-	if (posix_same_owner(&caller_fl->fl_core, &sys_fl->fl_core))
+	if (posix_same_owner(caller_flc, sys_flc))
 		return false;
 
 	/* Check whether they overlap */
 	if (!locks_overlap(caller_fl, sys_fl))
 		return false;
 
-	return locks_conflict(caller_fl, sys_fl);
+	return locks_conflict(caller_flc, sys_flc);
 }
 
 /* Determine if lock sys_fl blocks lock caller_fl. Used on xx_GETLK
@@ -907,28 +909,31 @@  static bool posix_locks_conflict(struct file_lock *caller_fl,
 static bool posix_test_locks_conflict(struct file_lock *caller_fl,
 				      struct file_lock *sys_fl)
 {
+	struct file_lock_core *caller = &caller_fl->fl_core;
+	struct file_lock_core *sys = &sys_fl->fl_core;
+
 	/* F_UNLCK checks any locks on the same fd. */
-	if (caller_fl->fl_core.fl_type == F_UNLCK) {
-		if (!posix_same_owner(&caller_fl->fl_core, &sys_fl->fl_core))
+	if (caller->fl_type == F_UNLCK) {
+		if (!posix_same_owner(caller, sys))
 			return false;
 		return locks_overlap(caller_fl, sys_fl);
 	}
-	return posix_locks_conflict(caller_fl, sys_fl);
+	return posix_locks_conflict(caller, sys);
 }
 
 /* Determine if lock sys_fl blocks lock caller_fl. FLOCK specific
  * checking before calling the locks_conflict().
  */
-static bool flock_locks_conflict(struct file_lock *caller_fl,
-				 struct file_lock *sys_fl)
+static bool flock_locks_conflict(struct file_lock_core *caller_flc,
+				 struct file_lock_core *sys_flc)
 {
 	/* FLOCK locks referring to the same filp do not conflict with
 	 * each other.
 	 */
-	if (caller_fl->fl_core.fl_file == sys_fl->fl_core.fl_file)
+	if (caller_flc->fl_file == sys_flc->fl_file)
 		return false;
 
-	return locks_conflict(caller_fl, sys_fl);
+	return locks_conflict(caller_flc, sys_flc);
 }
 
 void
@@ -1006,27 +1011,28 @@  EXPORT_SYMBOL(posix_test_lock);
 
 #define MAX_DEADLK_ITERATIONS 10
 
-/* Find a lock that the owner of the given block_fl is blocking on. */
-static struct file_lock *what_owner_is_waiting_for(struct file_lock *block_fl)
+/* Find a lock that the owner of the given @blocker is blocking on. */
+static struct file_lock_core *what_owner_is_waiting_for(struct file_lock_core *blocker)
 {
-	struct file_lock *fl;
+	struct file_lock_core *flc;
 
-	hash_for_each_possible(blocked_hash, fl, fl_core.fl_link, posix_owner_key(&block_fl->fl_core)) {
-		if (posix_same_owner(&fl->fl_core, &block_fl->fl_core)) {
-			while (fl->fl_core.fl_blocker)
-				fl = fl->fl_core.fl_blocker;
-			return fl;
+	hash_for_each_possible(blocked_hash, flc, fl_link, posix_owner_key(blocker)) {
+		if (posix_same_owner(flc, blocker)) {
+			while (flc->fl_blocker)
+				flc = &flc->fl_blocker->fl_core;
+			return flc;
 		}
 	}
 	return NULL;
 }
 
 /* Must be called with the blocked_lock_lock held! */
-static int posix_locks_deadlock(struct file_lock *caller_fl,
-				struct file_lock *block_fl)
+static bool posix_locks_deadlock(struct file_lock *caller_fl,
+				 struct file_lock *block_fl)
 {
+	struct file_lock_core *caller = &caller_fl->fl_core;
+	struct file_lock_core *blocker = &block_fl->fl_core;
 	int i = 0;
-	struct file_lock_core *flc = &caller_fl->fl_core;
 
 	lockdep_assert_held(&blocked_lock_lock);
 
@@ -1034,16 +1040,16 @@  static int posix_locks_deadlock(struct file_lock *caller_fl,
 	 * This deadlock detector can't reasonably detect deadlocks with
 	 * FL_OFDLCK locks, since they aren't owned by a process, per-se.
 	 */
-	if (IS_OFDLCK(flc))
+	if (IS_OFDLCK(caller))
 		return 0;
 
-	while ((block_fl = what_owner_is_waiting_for(block_fl))) {
+	while ((blocker = what_owner_is_waiting_for(blocker))) {
 		if (i++ > MAX_DEADLK_ITERATIONS)
-			return 0;
-		if (posix_same_owner(&caller_fl->fl_core, &block_fl->fl_core))
-			return 1;
+			return false;
+		if (posix_same_owner(caller, blocker))
+			return true;
 	}
-	return 0;
+	return false;
 }
 
 /* Try to create a FLOCK lock on filp. We always insert new FLOCK locks
@@ -1098,7 +1104,7 @@  static int flock_lock_inode(struct inode *inode, struct file_lock *request)
 
 find_conflict:
 	list_for_each_entry(fl, &ctx->flc_flock, fl_core.fl_list) {
-		if (!flock_locks_conflict(request, fl))
+		if (!flock_locks_conflict(&request->fl_core, &fl->fl_core))
 			continue;
 		error = -EAGAIN;
 		if (!(request->fl_core.fl_flags & FL_SLEEP))
@@ -1167,7 +1173,7 @@  static int posix_lock_inode(struct inode *inode, struct file_lock *request,
 	 */
 	if (request->fl_core.fl_type != F_UNLCK) {
 		list_for_each_entry(fl, &ctx->flc_posix, fl_core.fl_list) {
-			if (!posix_locks_conflict(request, fl))
+			if (!posix_locks_conflict(&request->fl_core, &fl->fl_core))
 				continue;
 			if (fl->fl_lmops && fl->fl_lmops->lm_lock_expirable
 				&& (*fl->fl_lmops->lm_lock_expirable)(fl)) {
@@ -1469,23 +1475,25 @@  static void time_out_leases(struct inode *inode, struct list_head *dispose)
 	}
 }
 
-static bool leases_conflict(struct file_lock *lease, struct file_lock *breaker)
+static bool leases_conflict(struct file_lock_core *lc, struct file_lock_core *bc)
 {
 	bool rc;
+	struct file_lock *lease = file_lock(lc);
+	struct file_lock *breaker = file_lock(bc);
 
 	if (lease->fl_lmops->lm_breaker_owns_lease
 			&& lease->fl_lmops->lm_breaker_owns_lease(lease))
 		return false;
-	if ((breaker->fl_core.fl_flags & FL_LAYOUT) != (lease->fl_core.fl_flags & FL_LAYOUT)) {
+	if ((bc->fl_flags & FL_LAYOUT) != (lc->fl_flags & FL_LAYOUT)) {
 		rc = false;
 		goto trace;
 	}
-	if ((breaker->fl_core.fl_flags & FL_DELEG) && (lease->fl_core.fl_flags & FL_LEASE)) {
+	if ((bc->fl_flags & FL_DELEG) && (lc->fl_flags & FL_LEASE)) {
 		rc = false;
 		goto trace;
 	}
 
-	rc = locks_conflict(breaker, lease);
+	rc = locks_conflict(bc, lc);
 trace:
 	trace_leases_conflict(rc, lease, breaker);
 	return rc;
@@ -1495,12 +1503,12 @@  static bool
 any_leases_conflict(struct inode *inode, struct file_lock *breaker)
 {
 	struct file_lock_context *ctx = inode->i_flctx;
-	struct file_lock *fl;
+	struct file_lock_core *flc;
 
 	lockdep_assert_held(&ctx->flc_lock);
 
-	list_for_each_entry(fl, &ctx->flc_lease, fl_core.fl_list) {
-		if (leases_conflict(fl, breaker))
+	list_for_each_entry(flc, &ctx->flc_lease, fl_list) {
+		if (leases_conflict(flc, &breaker->fl_core))
 			return true;
 	}
 	return false;
@@ -1556,7 +1564,7 @@  int __break_lease(struct inode *inode, unsigned int mode, unsigned int type)
 	}
 
 	list_for_each_entry_safe(fl, tmp, &ctx->flc_lease, fl_core.fl_list) {
-		if (!leases_conflict(fl, new_fl))
+		if (!leases_conflict(&fl->fl_core, &new_fl->fl_core))
 			continue;
 		if (want_write) {
 			if (fl->fl_core.fl_flags & FL_UNLOCK_PENDING)