[PATCH] readahead: Correct the start and size in ondemand_readahead()

[PATCH] readahead: Correct the start and size in ondemand_readahead()

[Yin Fengwei]

The commit
9425c591e06a ("page cache: fix page_cache_next/prev_miss off by one")
updated the page_cache_next_miss() to return the index beyond
range.

But it breaks the start/size of ra in ondemand_readahead() because
the offset by one is accumulated to readahead_index. As a consequence,
not best readahead order is picked.

Tracing of the order parameter of filemap_alloc_folio() showed:
     page order    : count     distribution
        0          : 892073   |                                        |
        1          : 0        |                                        |
        2          : 65120457 |****************************************|
        3          : 32914005 |********************                    |
        4          : 33020991 |********************                    |
with 9425c591e06a9.

With parent commit:
     page order    : count     distribution
        0          : 3417288  |****                                    |
        1          : 0        |                                        |
        2          : 877012   |*                                       |
        3          : 288      |                                        |
        4          : 5607522  |*******                                 |
        5          : 29974228 |****************************************|

Fix the issue by set correct start/size of ra in ondemand_readahead().

Reported-by: kernel test robot <oliver.sang@intel.com>
Closes: https://lore.kernel.org/oe-lkp/202306211346.1e9ff03e-oliver.sang@intel.com
Fixes: 9425c591e06a ("page cache: fix page_cache_next/prev_miss off by one")
Signed-off-by: Yin Fengwei <fengwei.yin@intel.com>
---
 mm/readahead.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/mm/readahead.c b/mm/readahead.c
index 47afbca1d122e..a1b8c628851a9 100644
--- a/mm/readahead.c
+++ b/mm/readahead.c
@@ -614,11 +614,11 @@ static void ondemand_readahead(struct readahead_control *ractl,
 				max_pages);
 		rcu_read_unlock();
 
-		if (!start || start - index > max_pages)
+		if (!start || start - index - 1 > max_pages)
 			return;
 
-		ra->start = start;
-		ra->size = start - index;	/* old async_size */
+		ra->start = start - 1;
+		ra->size = start - index - 1;	/* old async_size */
 		ra->size += req_size;
 		ra->size = get_next_ra_size(ra, max_pages);
 		ra->async_size = ra->size;
-- 
2.39.2

Re: [PATCH] readahead: Correct the start and size in ondemand_readahead()

[Yin Fengwei]

On 6/27/23 13:07, Yin Fengwei wrote:
> The commit
> 9425c591e06a ("page cache: fix page_cache_next/prev_miss off by one")
> updated the page_cache_next_miss() to return the index beyond
> range.
> 
> But it breaks the start/size of ra in ondemand_readahead() because
> the offset by one is accumulated to readahead_index. As a consequence,
> not best readahead order is picked.
> 
> Tracing of the order parameter of filemap_alloc_folio() showed:
>      page order    : count     distribution
>         0          : 892073   |                                        |
>         1          : 0        |                                        |
>         2          : 65120457 |****************************************|
>         3          : 32914005 |********************                    |
>         4          : 33020991 |********************                    |
> with 9425c591e06a9.
> 
> With parent commit:
>      page order    : count     distribution
>         0          : 3417288  |****                                    |
>         1          : 0        |                                        |
>         2          : 877012   |*                                       |
>         3          : 288      |                                        |
>         4          : 5607522  |*******                                 |
>         5          : 29974228 |****************************************|
> 
> Fix the issue by set correct start/size of ra in ondemand_readahead().
> 
> Reported-by: kernel test robot <oliver.sang@intel.com>
> Closes: https://lore.kernel.org/oe-lkp/202306211346.1e9ff03e-oliver.sang@intel.com
> Fixes: 9425c591e06a ("page cache: fix page_cache_next/prev_miss off by one")
> Signed-off-by: Yin Fengwei <fengwei.yin@intel.com>
> ---
>  mm/readahead.c | 6 +++---
>  1 file changed, 3 insertions(+), 3 deletions(-)
> 
> diff --git a/mm/readahead.c b/mm/readahead.c
> index 47afbca1d122e..a1b8c628851a9 100644
> --- a/mm/readahead.c
> +++ b/mm/readahead.c
> @@ -614,11 +614,11 @@ static void ondemand_readahead(struct readahead_control *ractl,
>  				max_pages);
>  		rcu_read_unlock();
>  
> -		if (!start || start - index > max_pages)
> +		if (!start || start - index - 1 > max_pages)
>  			return;
The offset by one only happens when no gaps in the range. So this patch need an update.
I will send out v2 soon. Thanks.


Regards
Yin, Fengwei

>  
> -		ra->start = start;
> -		ra->size = start - index;	/* old async_size */
> +		ra->start = start - 1;
> +		ra->size = start - index - 1;	/* old async_size */
>  		ra->size += req_size;
>  		ra->size = get_next_ra_size(ra, max_pages);
>  		ra->async_size = ra->size;