Re: [PATCH]mm:Correctly update number of rotated pages on active list.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Thu 03-12-15 10:08:11, Pradeep Goswami (Pradeep Kumar Goswami) wrote:
> This patch corrects the number of pages which are rotated on active list.
> The counter for rotated pages effects the number of pages
> to be scanned on active pages list in  low memory situations.

Why this should be changed?

This seems to be deliberate:
        /*
         * Count referenced pages from currently used mappings as rotated,
         * even though only some of them are actually re-activated.  This
         * helps balance scan pressure between file and anonymous pages in
         * get_scan_count.
         */
        reclaim_stat->recent_rotated[file] += nr_rotated;

What kind of problem are you trying to fix?

> 
> Signed-off-by: Pradeep Goswami <pradeep.goswami@xxxxxxxxxxxxx>
> Cc: Rebecca Schultz Zavin <rebecca@xxxxxxxxxxx>
> Cc: Vladimir Davydov <vdavydov@xxxxxxxxxxxxx>
> Cc: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
> ---
> --- a/mm/vmscan.c       2015-11-18 20:55:38.208838142 +0800
> +++ b/mm/vmscan.c       2015-11-19 14:37:31.189838998 +0800
> @@ -1806,7 +1806,6 @@ static void shrink_active_list(unsigned
>  
>                 if (page_referenced(page, 0, sc->target_mem_cgroup,
>                                     &vm_flags)) {
> -                       nr_rotated += hpage_nr_pages(page);
>                         /*  
>                          * Identify referenced, file-backed active pages and 
>                          * give them one more trip around the active list. So
> @@ -1818,6 +1817,7 @@ static void shrink_active_list(unsigned
>                          */  
>                         if ((vm_flags & VM_EXEC) && page_is_file_cache(page)) {
>                                 list_add(&page->lru, &l_active);
> +                               nr_rotated += hpage_nr_pages(page);
>                                 continue;
>                         }   
>                 }   
> 
> Thanks,
> Pradeep.
> --
> To unsubscribe, send a message with 'unsubscribe linux-mm' in
> the body to majordomo@xxxxxxxxx.  For more info on Linux MM,
> see: http://www.linux-mm.org/ .
> Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>

-- 
Michal Hocko
SUSE Labs

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxx.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>



[Index of Archives]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [IETF Annouce]     [Bugtraq]     [Linux]     [Linux OMAP]     [Linux MIPS]     [ECOS]     [Asterisk Internet PBX]     [Linux API]