From: Chengming Zhou <zhouchengming@bytedance.com>

We have to invalidate any duplicate entry even when !zswap_enabled
since zswap can be disabled anytime. If the folio store success before,
then got dirtied again but zswap disabled, we won't invalidate the old
duplicate entry in the zswap_store(). So later lru writeback may
overwrite the new data in swapfile.

Fixes: 42c06a0e8ebe ("mm: kill frontswap")
Cc: <stable@vger.kernel.org>
Signed-off-by: Chengming Zhou <zhouchengming@bytedance.com>
---
 mm/zswap.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/mm/zswap.c b/mm/zswap.c
index fe7ee2640c69..32633d0597dc 100644
--- a/mm/zswap.c
+++ b/mm/zswap.c
@@ -1516,7 +1516,7 @@ bool zswap_store(struct folio *folio)
 	if (folio_test_large(folio))
 		return false;
 
-	if (!zswap_enabled || !tree)
+	if (!tree)
 		return false;
 
 	/*
@@ -1531,6 +1531,10 @@ bool zswap_store(struct folio *folio)
 		zswap_invalidate_entry(tree, dupentry);
 	}
 	spin_unlock(&tree->lock);
+
+	if (!zswap_enabled)
+		return false;
+
 	objcg = get_obj_cgroup_from_folio(folio);
 	if (objcg && !obj_cgroup_may_zswap(objcg)) {
 		memcg = get_mem_cgroup_from_objcg(objcg);
-- 
2.40.1

On Thu,  8 Feb 2024 02:32:54 +0000 chengming.zhou@linux.dev wrote:

> From: Chengming Zhou <zhouchengming@bytedance.com>
> 
> We have to invalidate any duplicate entry even when !zswap_enabled
> since zswap can be disabled anytime. If the folio store success before,
> then got dirtied again but zswap disabled, we won't invalidate the old
> duplicate entry in the zswap_store(). So later lru writeback may
> overwrite the new data in swapfile.
> 
> ...
>
> --- a/mm/zswap.c
> +++ b/mm/zswap.c
> @@ -1516,7 +1516,7 @@ bool zswap_store(struct folio *folio)
>  	if (folio_test_large(folio))
>  		return false;
>  
> -	if (!zswap_enabled || !tree)
> +	if (!tree)
>  		return false;
>  
>  	/*
> @@ -1531,6 +1531,10 @@ bool zswap_store(struct folio *folio)
>  		zswap_invalidate_entry(tree, dupentry);
>  	}
>  	spin_unlock(&tree->lock);
> +
> +	if (!zswap_enabled)
> +		return false;
> +
>  	objcg = get_obj_cgroup_from_folio(folio);
>  	if (objcg && !obj_cgroup_may_zswap(objcg)) {
>  		memcg = get_mem_cgroup_from_objcg(objcg);

OK, thanks.

I saw only one reject from mm-unstable patches.  Your patch "mm/zswap:
make sure each swapfile always have zswap rb-tree" now does

--- a/mm/zswap.c~mm-zswap-make-sure-each-swapfile-always-have-zswap-rb-tree
+++ a/mm/zswap.c
@@ -1518,9 +1518,6 @@ bool zswap_store(struct folio *folio)
 	if (folio_test_large(folio))
 		return false;
 
-	if (!tree)
-		return false;
-
 	/*
 	 * If this is a duplicate, it must be removed before attempting to store
 	 * it, otherwise, if the store fails the old page won't be removed from

On 2024/2/9 05:09, Andrew Morton wrote:
> On Thu,  8 Feb 2024 02:32:54 +0000 chengming.zhou@linux.dev wrote:
> 
>> From: Chengming Zhou <zhouchengming@bytedance.com>
>>
>> We have to invalidate any duplicate entry even when !zswap_enabled
>> since zswap can be disabled anytime. If the folio store success before,
>> then got dirtied again but zswap disabled, we won't invalidate the old
>> duplicate entry in the zswap_store(). So later lru writeback may
>> overwrite the new data in swapfile.
>>
>> ...
>>
>> --- a/mm/zswap.c
>> +++ b/mm/zswap.c
>> @@ -1516,7 +1516,7 @@ bool zswap_store(struct folio *folio)
>>  	if (folio_test_large(folio))
>>  		return false;
>>  
>> -	if (!zswap_enabled || !tree)
>> +	if (!tree)
>>  		return false;
>>  
>>  	/*
>> @@ -1531,6 +1531,10 @@ bool zswap_store(struct folio *folio)
>>  		zswap_invalidate_entry(tree, dupentry);
>>  	}
>>  	spin_unlock(&tree->lock);
>> +
>> +	if (!zswap_enabled)
>> +		return false;
>> +
>>  	objcg = get_obj_cgroup_from_folio(folio);
>>  	if (objcg && !obj_cgroup_may_zswap(objcg)) {
>>  		memcg = get_mem_cgroup_from_objcg(objcg);
> 
> OK, thanks.
> 
> I saw only one reject from mm-unstable patches.  Your patch "mm/zswap:
> make sure each swapfile always have zswap rb-tree" now does

It's correct. Thanks!

The other patch that includes optimization and cleanup is updated based on
mm-unstable and just resend:

https://lore.kernel.org/all/20240209044112.3883835-1-chengming.zhou@linux.dev/

> 
> --- a/mm/zswap.c~mm-zswap-make-sure-each-swapfile-always-have-zswap-rb-tree
> +++ a/mm/zswap.c
> @@ -1518,9 +1518,6 @@ bool zswap_store(struct folio *folio)
>  	if (folio_test_large(folio))
>  		return false;
>  
> -	if (!tree)
> -		return false;
> -
>  	/*
>  	 * If this is a duplicate, it must be removed before attempting to store
>  	 * it, otherwise, if the store fails the old page won't be removed from
> 
>

On Thu, Feb 08, 2024 at 02:32:54AM +0000, chengming.zhou@linux.dev wrote:
> From: Chengming Zhou <zhouchengming@bytedance.com>
> 
> We have to invalidate any duplicate entry even when !zswap_enabled
> since zswap can be disabled anytime. If the folio store success before,
> then got dirtied again but zswap disabled, we won't invalidate the old
> duplicate entry in the zswap_store(). So later lru writeback may
> overwrite the new data in swapfile.
> 
> Fixes: 42c06a0e8ebe ("mm: kill frontswap")
> Cc: <stable@vger.kernel.org>
> Signed-off-by: Chengming Zhou <zhouchengming@bytedance.com>

Acked-by: Johannes Weiner <hannes@cmpxchg.org>

Nice, this is easier to backport and should be less disruptive to
mm-unstable as well. It makes sense to me to put the optimization and
cleanup that was cut out into a separate patch on top of mm-unstable.

>  mm/zswap.c | 6 +++++-
>  1 file changed, 5 insertions(+), 1 deletion(-)
> 
> diff --git a/mm/zswap.c b/mm/zswap.c
> index fe7ee2640c69..32633d0597dc 100644
> --- a/mm/zswap.c
> +++ b/mm/zswap.c
> @@ -1516,7 +1516,7 @@ bool zswap_store(struct folio *folio)
>  	if (folio_test_large(folio))
>  		return false;
>  
> -	if (!zswap_enabled || !tree)
> +	if (!tree)
>  		return false;
>  
>  	/*
> @@ -1531,6 +1531,10 @@ bool zswap_store(struct folio *folio)
>  		zswap_invalidate_entry(tree, dupentry);
>  	}
>  	spin_unlock(&tree->lock);
> +
> +	if (!zswap_enabled)
> +		return false;
> +
>  	objcg = get_obj_cgroup_from_folio(folio);
>  	if (objcg && !obj_cgroup_may_zswap(objcg)) {
>  		memcg = get_mem_cgroup_from_objcg(objcg);