linux-btrfs.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH] btrfs: remove warnings superseded by refcount_t usage
@ 2018-06-25 16:38 David Sterba
  2018-06-26 14:00 ` Nikolay Borisov
  0 siblings, 1 reply; 4+ messages in thread
From: David Sterba @ 2018-06-25 16:38 UTC (permalink / raw)
  To: linux-btrfs; +Cc: David Sterba

There are several WARN_ON calls that catch incrorrect reference counter
updates, but this is what the refcount_t does already:

* refcount_inc from 0 will warn
* refcount_dec_and_test from 0 will warn

Signed-off-by: David Sterba <dsterba@suse.com>
---
 fs/btrfs/delayed-ref.h | 1 -
 fs/btrfs/extent_map.c  | 2 +-
 fs/btrfs/transaction.c | 1 -
 fs/btrfs/volumes.c     | 1 -
 4 files changed, 1 insertion(+), 4 deletions(-)

diff --git a/fs/btrfs/delayed-ref.h b/fs/btrfs/delayed-ref.h
index ea1aecb6a50d..16ee92c541bf 100644
--- a/fs/btrfs/delayed-ref.h
+++ b/fs/btrfs/delayed-ref.h
@@ -210,7 +210,6 @@ btrfs_free_delayed_extent_op(struct btrfs_delayed_extent_op *op)
 
 static inline void btrfs_put_delayed_ref(struct btrfs_delayed_ref_node *ref)
 {
-	WARN_ON(refcount_read(&ref->refs) == 0);
 	if (refcount_dec_and_test(&ref->refs)) {
 		WARN_ON(ref->in_tree);
 		switch (ref->type) {
diff --git a/fs/btrfs/extent_map.c b/fs/btrfs/extent_map.c
index 6648d55e5339..6808ad4ed3c9 100644
--- a/fs/btrfs/extent_map.c
+++ b/fs/btrfs/extent_map.c
@@ -72,7 +72,7 @@ void free_extent_map(struct extent_map *em)
 {
 	if (!em)
 		return;
-	WARN_ON(refcount_read(&em->refs) == 0);
+
 	if (refcount_dec_and_test(&em->refs)) {
 		WARN_ON(extent_map_in_tree(em));
 		WARN_ON(!list_empty(&em->list));
diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c
index 4485eae41e88..f6b68099b767 100644
--- a/fs/btrfs/transaction.c
+++ b/fs/btrfs/transaction.c
@@ -41,7 +41,6 @@ static const unsigned int btrfs_blocked_trans_types[TRANS_STATE_MAX] = {
 
 void btrfs_put_transaction(struct btrfs_transaction *transaction)
 {
-	WARN_ON(refcount_read(&transaction->use_count) == 0);
 	if (refcount_dec_and_test(&transaction->use_count)) {
 		BUG_ON(!list_empty(&transaction->list));
 		WARN_ON(!RB_EMPTY_ROOT(&transaction->delayed_refs.href_root));
diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
index e034ad9e23b4..f64d26b4f278 100644
--- a/fs/btrfs/volumes.c
+++ b/fs/btrfs/volumes.c
@@ -5330,7 +5330,6 @@ static struct btrfs_bio *alloc_btrfs_bio(int total_stripes, int real_stripes)
 
 void btrfs_get_bbio(struct btrfs_bio *bbio)
 {
-	WARN_ON(!refcount_read(&bbio->refs));
 	refcount_inc(&bbio->refs);
 }
 
-- 
2.17.0


^ permalink raw reply related	[flat|nested] 4+ messages in thread

* Re: [PATCH] btrfs: remove warnings superseded by refcount_t usage
  2018-06-25 16:38 [PATCH] btrfs: remove warnings superseded by refcount_t usage David Sterba
@ 2018-06-26 14:00 ` Nikolay Borisov
  2018-06-29 14:05   ` David Sterba
  0 siblings, 1 reply; 4+ messages in thread
From: Nikolay Borisov @ 2018-06-26 14:00 UTC (permalink / raw)
  To: David Sterba, linux-btrfs



On 25.06.2018 19:38, David Sterba wrote:
> There are several WARN_ON calls that catch incrorrect reference counter
> updates, but this is what the refcount_t does already:
> 
> * refcount_inc from 0 will warn
> * refcount_dec_and_test from 0 will warn
> 

But these warnings are only going to be triggered in the case of
CONFIG_REFCOUNT_FULL otherwise refcount falls back to plain atomic
operations.

> Signed-off-by: David Sterba <dsterba@suse.com>
> ---
>  fs/btrfs/delayed-ref.h | 1 -
>  fs/btrfs/extent_map.c  | 2 +-
>  fs/btrfs/transaction.c | 1 -
>  fs/btrfs/volumes.c     | 1 -
>  4 files changed, 1 insertion(+), 4 deletions(-)
> 
> diff --git a/fs/btrfs/delayed-ref.h b/fs/btrfs/delayed-ref.h
> index ea1aecb6a50d..16ee92c541bf 100644
> --- a/fs/btrfs/delayed-ref.h
> +++ b/fs/btrfs/delayed-ref.h
> @@ -210,7 +210,6 @@ btrfs_free_delayed_extent_op(struct btrfs_delayed_extent_op *op)
>  
>  static inline void btrfs_put_delayed_ref(struct btrfs_delayed_ref_node *ref)
>  {
> -	WARN_ON(refcount_read(&ref->refs) == 0);
>  	if (refcount_dec_and_test(&ref->refs)) {
>  		WARN_ON(ref->in_tree);
>  		switch (ref->type) {
> diff --git a/fs/btrfs/extent_map.c b/fs/btrfs/extent_map.c
> index 6648d55e5339..6808ad4ed3c9 100644
> --- a/fs/btrfs/extent_map.c
> +++ b/fs/btrfs/extent_map.c
> @@ -72,7 +72,7 @@ void free_extent_map(struct extent_map *em)
>  {
>  	if (!em)
>  		return;
> -	WARN_ON(refcount_read(&em->refs) == 0);
> +
>  	if (refcount_dec_and_test(&em->refs)) {
>  		WARN_ON(extent_map_in_tree(em));
>  		WARN_ON(!list_empty(&em->list));
> diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c
> index 4485eae41e88..f6b68099b767 100644
> --- a/fs/btrfs/transaction.c
> +++ b/fs/btrfs/transaction.c
> @@ -41,7 +41,6 @@ static const unsigned int btrfs_blocked_trans_types[TRANS_STATE_MAX] = {
>  
>  void btrfs_put_transaction(struct btrfs_transaction *transaction)
>  {
> -	WARN_ON(refcount_read(&transaction->use_count) == 0);
>  	if (refcount_dec_and_test(&transaction->use_count)) {
>  		BUG_ON(!list_empty(&transaction->list));
>  		WARN_ON(!RB_EMPTY_ROOT(&transaction->delayed_refs.href_root));
> diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c
> index e034ad9e23b4..f64d26b4f278 100644
> --- a/fs/btrfs/volumes.c
> +++ b/fs/btrfs/volumes.c
> @@ -5330,7 +5330,6 @@ static struct btrfs_bio *alloc_btrfs_bio(int total_stripes, int real_stripes)
>  
>  void btrfs_get_bbio(struct btrfs_bio *bbio)
>  {
> -	WARN_ON(!refcount_read(&bbio->refs));
>  	refcount_inc(&bbio->refs);
>  }
>  
> 

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH] btrfs: remove warnings superseded by refcount_t usage
  2018-06-26 14:00 ` Nikolay Borisov
@ 2018-06-29 14:05   ` David Sterba
  2018-07-05 15:55     ` David Sterba
  0 siblings, 1 reply; 4+ messages in thread
From: David Sterba @ 2018-06-29 14:05 UTC (permalink / raw)
  To: Nikolay Borisov; +Cc: David Sterba, linux-btrfs

On Tue, Jun 26, 2018 at 05:00:46PM +0300, Nikolay Borisov wrote:
> On 25.06.2018 19:38, David Sterba wrote:
> > There are several WARN_ON calls that catch incrorrect reference counter
> > updates, but this is what the refcount_t does already:
> > 
> > * refcount_inc from 0 will warn
> > * refcount_dec_and_test from 0 will warn
> > 
> 
> But these warnings are only going to be triggered in the case of
> CONFIG_REFCOUNT_FULL otherwise refcount falls back to plain atomic
> operations.

Right, I'm for REFCOUNT_FULL on by default obviously, but we need to
take handle the case where it's not. And there isn't a refcount type
with the full semantics.

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH] btrfs: remove warnings superseded by refcount_t usage
  2018-06-29 14:05   ` David Sterba
@ 2018-07-05 15:55     ` David Sterba
  0 siblings, 0 replies; 4+ messages in thread
From: David Sterba @ 2018-07-05 15:55 UTC (permalink / raw)
  To: dsterba, Nikolay Borisov, David Sterba, linux-btrfs

On Fri, Jun 29, 2018 at 04:05:40PM +0200, David Sterba wrote:
> On Tue, Jun 26, 2018 at 05:00:46PM +0300, Nikolay Borisov wrote:
> > On 25.06.2018 19:38, David Sterba wrote:
> > > There are several WARN_ON calls that catch incrorrect reference counter
> > > updates, but this is what the refcount_t does already:
> > > 
> > > * refcount_inc from 0 will warn
> > > * refcount_dec_and_test from 0 will warn
> > > 
> > 
> > But these warnings are only going to be triggered in the case of
> > CONFIG_REFCOUNT_FULL otherwise refcount falls back to plain atomic
> > operations.
> 
> Right, I'm for REFCOUNT_FULL on by default obviously, but we need to
> take handle the case where it's not. And there isn't a refcount type
> with the full semantics.

There refcount helpers will be extended to allow the full check
semantics so I'll use them and drop the WARN_ONs once the patch gets
merged.

https://lore.kernel.org/lkml/20180703100102.16615-1-mark.rutland@arm.com/

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2018-07-05 15:55 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2018-06-25 16:38 [PATCH] btrfs: remove warnings superseded by refcount_t usage David Sterba
2018-06-26 14:00 ` Nikolay Borisov
2018-06-29 14:05   ` David Sterba
2018-07-05 15:55     ` David Sterba

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).