From: "Matthew Wilcox (Oracle)" <willy@xxxxxxxxxxxxx> Subject: proc: optimise smaps for shmem entries Avoid bumping the refcount on pages when we're only interested in the swap entries. Link: https://lkml.kernel.org/r/20200910183318.20139-5-willy@xxxxxxxxxxxxx Signed-off-by: Matthew Wilcox (Oracle) <willy@xxxxxxxxxxxxx> Acked-by: Johannes Weiner <hannes@xxxxxxxxxxx> Cc: Alexey Dobriyan <adobriyan@xxxxxxxxx> Cc: Chris Wilson <chris@xxxxxxxxxxxxxxxxxx> Cc: Huang Ying <ying.huang@xxxxxxxxx> Cc: Hugh Dickins <hughd@xxxxxxxxxx> Cc: Jani Nikula <jani.nikula@xxxxxxxxxxxxxxx> Cc: Matthew Auld <matthew.auld@xxxxxxxxx> Cc: William Kucharski <william.kucharski@xxxxxxxxxx> Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> --- fs/proc/task_mmu.c | 8 +------- 1 file changed, 1 insertion(+), 7 deletions(-) --- a/fs/proc/task_mmu.c~proc-optimise-smaps-for-shmem-entries +++ a/fs/proc/task_mmu.c @@ -520,16 +520,10 @@ static void smaps_pte_entry(pte_t *pte, page = device_private_entry_to_page(swpent); } else if (unlikely(IS_ENABLED(CONFIG_SHMEM) && mss->check_shmem_swap && pte_none(*pte))) { - page = find_get_entry(vma->vm_file->f_mapping, + page = xa_load(&vma->vm_file->f_mapping->i_pages, linear_page_index(vma, addr)); - if (!page) - return; - if (xa_is_value(page)) mss->swap += PAGE_SIZE; - else - put_page(page); - return; } _
