Re: [PATCH mm-hotfixes-unstable] mm/zswap: invalidate duplicate entry when !zswap_enabled

[Johannes Weiner]

On Thu, Feb 08, 2024 at 02:32:54AM +0000, chengming.zhou@xxxxxxxxx wrote:
> From: Chengming Zhou <zhouchengming@xxxxxxxxxxxxx>
> 
> 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@xxxxxxxxxxxxxxx>
> Signed-off-by: Chengming Zhou <zhouchengming@xxxxxxxxxxxxx>

Acked-by: Johannes Weiner <hannes@xxxxxxxxxxx>

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);