[PATCH v3] erofs: free pclusters if no cached folio attached

[Gao Xiang]

Hi Chunhai,

On 2024/11/11 16:25, Chunhai Guo wrote:
> Once a pcluster is fully decompressed and there are no attached cached
> folios, its corresponding `struct z_erofs_pcluster` will be freed. This
> will significantly reduce the frequency of calls to erofs_shrink_scan()
> and the memory allocated for `struct z_erofs_pcluster`.
> 
> The tables below show approximately a 96% reduction in the calls to
> erofs_shrink_scan() and in the memory allocated for `struct
> z_erofs_pcluster` after applying this patch. The results were obtained
> by performing a test to copy a 4.1GB partition on ARM64 Android devices
> running the 6.6 kernel with an 8-core CPU and 12GB of memory.
> 
> 1. The reduction in calls to erofs_shrink_scan():
> +-----------------+-----------+----------+---------+
> |                 | w/o patch | w/ patch |  diff   |
> +-----------------+-----------+----------+---------+
> | Average (times) |   11390   |   390    | -96.57% |
> +-----------------+-----------+----------+---------+
> 
> 2. The reduction in memory released by erofs_shrink_scan():
> +-----------------+-----------+----------+---------+
> |                 | w/o patch | w/ patch |  diff   |
> +-----------------+-----------+----------+---------+
> | Average (Byte)  | 133612656 | 4434552  | -96.68% |
> +-----------------+-----------+----------+---------+
> 
> Signed-off-by: Chunhai Guo <guochunhai at vivo.com>
> ---
> v1 -> v2:
>   - rename erofs_prepare_to_release_pcluster() to __erofs_try_to_release_pcluster()
>   - use trylock in z_erofs_put_pcluster() instead of erofs_try_to_release_pcluster()
> 
> v1: https://lore.kernel.org/linux-erofs/588351c0-93f9-4a04-a923-15aae8b71d49@linux.alibaba.com/
> change since v1:
>   - rebase this patch on "sunset z_erofs_workgroup` series
>   - remove check on pcl->partial and get rid of `be->try_free`
>   - update test results base on 6.6 kernel
> ---
>   fs/erofs/zdata.c | 56 ++++++++++++++++++++++++++++++++----------------
>   1 file changed, 38 insertions(+), 18 deletions(-)
> 
> diff --git a/fs/erofs/zdata.c b/fs/erofs/zdata.c
> index 6b73a2307460..c1cb08d536c0 100644
> --- a/fs/erofs/zdata.c
> +++ b/fs/erofs/zdata.c
> @@ -885,14 +885,11 @@ static void z_erofs_rcu_callback(struct rcu_head *head)
>   			struct z_erofs_pcluster, rcu));
>   }
>   
> -static bool erofs_try_to_release_pcluster(struct erofs_sb_info *sbi,
> +static bool __erofs_try_to_release_pcluster(struct erofs_sb_info *sbi,
>   					  struct z_erofs_pcluster *pcl)
>   {
> -	int free = false;
> -
> -	spin_lock(&pcl->lockref.lock);
>   	if (pcl->lockref.count)
> -		goto out;
> +		return false;
>   
>   	/*
>   	 * Note that all cached folios should be detached before deleted from
> @@ -900,7 +897,7 @@ static bool erofs_try_to_release_pcluster(struct erofs_sb_info *sbi,
>   	 * orphan old pcluster when the new one is available in the tree.
>   	 */
>   	if (erofs_try_to_free_all_cached_folios(sbi, pcl))
> -		goto out;
> +		return false;
>   
>   	/*
>   	 * It's impossible to fail after the pcluster is freezed, but in order
> @@ -909,8 +906,16 @@ static bool erofs_try_to_release_pcluster(struct erofs_sb_info *sbi,
>   	DBG_BUGON(__xa_erase(&sbi->managed_pslots, pcl->index) != pcl);
>   
>   	lockref_mark_dead(&pcl->lockref);
> -	free = true;
> -out:
> +	return true;
> +}
> +
> +static bool erofs_try_to_release_pcluster(struct erofs_sb_info *sbi,
> +					  struct z_erofs_pcluster *pcl)
> +{
> +	bool free = false;

Just few nits.

bool free; ?

> +
> +	spin_lock(&pcl->lockref.lock);
> +	free = __erofs_try_to_release_pcluster(sbi, pcl);
>   	spin_unlock(&pcl->lockref.lock);
>   	if (free) {
>   		atomic_long_dec(&erofs_global_shrink_cnt);
> @@ -942,16 +947,26 @@ unsigned long z_erofs_shrink_scan(struct erofs_sb_info *sbi,
>   	return freed;
>   }
>   
> -static void z_erofs_put_pcluster(struct z_erofs_pcluster *pcl)
> +static void z_erofs_put_pcluster(struct erofs_sb_info *sbi,
> +		struct z_erofs_pcluster *pcl, bool try_free)
>   {
> +	bool free = false;
> +
>   	if (lockref_put_or_lock(&pcl->lockref))
>   		return;
>   
>   	DBG_BUGON(__lockref_is_dead(&pcl->lockref));
> -	if (pcl->lockref.count == 1)
> -		atomic_long_inc(&erofs_global_shrink_cnt);
> -	--pcl->lockref.count;
> +	if (--pcl->lockref.count == 0) {
> +		if (try_free && xa_trylock(&sbi->managed_pslots)) {
> +			free = __erofs_try_to_release_pcluster(sbi, pcl);
> +			xa_unlock(&sbi->managed_pslots);
> +		}
> +		if (!free)

		atomic_long_add(!free, &erofs_global_shrink_cnt); ?



> +			atomic_long_inc(&erofs_global_shrink_cnt);
> +	}
>   	spin_unlock(&pcl->lockref.lock);
> +	if (free)
> +		call_rcu(&pcl->rcu, z_erofs_rcu_callback);
>   }
>   
>   static void z_erofs_pcluster_end(struct z_erofs_decompress_frontend *fe)
> @@ -972,7 +987,7 @@ static void z_erofs_pcluster_end(struct z_erofs_decompress_frontend *fe)
>   	 * any longer if the pcluster isn't hosted by ourselves.
>   	 */
>   	if (fe->mode < Z_EROFS_PCLUSTER_FOLLOWED_NOINPLACE)
> -		z_erofs_put_pcluster(pcl);
> +		z_erofs_put_pcluster(EROFS_I_SB(fe->inode), pcl, false);
>   
>   	fe->pcl = NULL;
>   }
> @@ -1274,6 +1289,7 @@ static int z_erofs_decompress_pcluster(struct z_erofs_decompress_backend *be,
>   	int i, j, jtop, err2;
>   	struct page *page;
>   	bool overlapped;
> +	bool try_free = true;
>   
>   	mutex_lock(&pcl->lock);
>   	be->nr_pages = PAGE_ALIGN(pcl->length + pcl->pageofs_out) >> PAGE_SHIFT;
> @@ -1332,8 +1348,10 @@ static int z_erofs_decompress_pcluster(struct z_erofs_decompress_backend *be,
>   		for (i = 0; i < pclusterpages; ++i) {
>   			page = be->compressed_pages[i];
>   			if (!page ||
> -			    erofs_folio_is_managed(sbi, page_folio(page)))
> +			    erofs_folio_is_managed(sbi, page_folio(page))) {
> +				try_free = false;
>   				continue;
> +			}
>   			(void)z_erofs_put_shortlivedpage(be->pagepool, page);
>   			WRITE_ONCE(pcl->compressed_bvecs[i].page, NULL);
>   		}
> @@ -1379,6 +1397,12 @@ static int z_erofs_decompress_pcluster(struct z_erofs_decompress_backend *be,
>   	/* pcluster lock MUST be taken before the following line */
>   	WRITE_ONCE(pcl->next, Z_EROFS_PCLUSTER_NIL);
>   	mutex_unlock(&pcl->lock);
> +
> +	if (z_erofs_is_inline_pcluster(pcl))
> +		z_erofs_free_pcluster(pcl);
> +	else
> +		z_erofs_put_pcluster(sbi, pcl, try_free);
> +

unnecessary blank line.

Thanks,
Gao Xiang

>   	return err;
>   }