On (06/20/16 13:55), Byungchul Park wrote:
> In order to use lockdep-enabled bit_spin_lock, we have to call
> bit_spin_init() when a instance including the bit used as lock
> creates, and bit_spin_free() when the instance including the bit
> destroys.
> 
> The zram is one of bit_spin_lock users. And this patch adds
> bit_spin_init() and bit_spin_free() properly to apply the lock
> correctness validator to bit_spin_lock the rzam is using.

Hello,

just for information

there was a proposal from -rt people to use normal spinlocks
for table's entries, rather that bitspinlock. I had a patch
some time ago, will obtain some performance data and post
RFC [may be tomorrow].

        -ss

> Signed-off-by: Byungchul Park <byungchul.p...@lge.com>
> ---
>  drivers/block/zram/zram_drv.c | 10 ++++++++++
>  1 file changed, 10 insertions(+)
> 
> diff --git a/drivers/block/zram/zram_drv.c b/drivers/block/zram/zram_drv.c
> index 370c2f7..2bc3bde 100644
> --- a/drivers/block/zram/zram_drv.c
> +++ b/drivers/block/zram/zram_drv.c
> @@ -495,6 +495,11 @@ static void zram_meta_free(struct zram_meta *meta, u64 
> disksize)
>       }
>  
>       zs_destroy_pool(meta->mem_pool);
> +
> +     for (index = 0; index < num_pages; index++) {
> +             bit_spin_free(ZRAM_ACCESS, &meta->table[index].value);
> +     }
> +
>       vfree(meta->table);
>       kfree(meta);
>  }
> @@ -503,6 +508,7 @@ static struct zram_meta *zram_meta_alloc(char *pool_name, 
> u64 disksize)
>  {
>       size_t num_pages;
>       struct zram_meta *meta = kmalloc(sizeof(*meta), GFP_KERNEL);
> +     int index;
>  
>       if (!meta)
>               return NULL;
> @@ -520,6 +526,10 @@ static struct zram_meta *zram_meta_alloc(char 
> *pool_name, u64 disksize)
>               goto out_error;
>       }
>  
> +     for (index = 0; index < num_pages; index++) {
> +             bit_spin_init(ZRAM_ACCESS, &meta->table[index].value);
> +     }
> +
>       return meta;
>  
>  out_error:
> -- 
> 1.9.1
> 

Reply via email to