Re: [PATCH v9 12/25] mm: Move end_index check out of readahead loop

From: Eric Biggers
Date: Fri Mar 20 2020 - 12:58:33 EST


On Fri, Mar 20, 2020 at 07:22:18AM -0700, Matthew Wilcox wrote:
> From: "Matthew Wilcox (Oracle)" <willy@xxxxxxxxxxxxx>
>
> By reducing nr_to_read, we can eliminate this check from inside the loop.
>
> Signed-off-by: Matthew Wilcox (Oracle) <willy@xxxxxxxxxxxxx>
> Reviewed-by: John Hubbard <jhubbard@xxxxxxxxxx>
> Reviewed-by: William Kucharski <william.kucharski@xxxxxxxxxx>
> ---
> mm/readahead.c | 17 +++++++++++------
> 1 file changed, 11 insertions(+), 6 deletions(-)
>
> diff --git a/mm/readahead.c b/mm/readahead.c
> index d01531ef9f3c..a37b68f66233 100644
> --- a/mm/readahead.c
> +++ b/mm/readahead.c
> @@ -167,8 +167,6 @@ void __do_page_cache_readahead(struct address_space *mapping,
> unsigned long lookahead_size)
> {
> struct inode *inode = mapping->host;
> - struct page *page;
> - unsigned long end_index; /* The last page we want to read */
> LIST_HEAD(page_pool);
> loff_t isize = i_size_read(inode);
> gfp_t gfp_mask = readahead_gfp_mask(mapping);
> @@ -178,22 +176,29 @@ void __do_page_cache_readahead(struct address_space *mapping,
> ._index = index,
> };
> unsigned long i;
> + pgoff_t end_index; /* The last page we want to read */
>
> if (isize == 0)
> return;
>
> - end_index = ((isize - 1) >> PAGE_SHIFT);
> + end_index = (isize - 1) >> PAGE_SHIFT;
> + if (index > end_index)
> + return;
> + /* Avoid wrapping to the beginning of the file */
> + if (index + nr_to_read < index)
> + nr_to_read = ULONG_MAX - index + 1;
> + /* Don't read past the page containing the last byte of the file */
> + if (index + nr_to_read >= end_index)
> + nr_to_read = end_index - index + 1;

There seem to be a couple off-by-one errors here. Shouldn't it be:

/* Avoid wrapping to the beginning of the file */
if (index + nr_to_read < index)
nr_to_read = ULONG_MAX - index;
/* Don't read past the page containing the last byte of the file */
if (index + nr_to_read > end_index)
nr_to_read = end_index - index + 1;

I.e., 'ULONG_MAX - index' rather than 'ULONG_MAX - index + 1', so that
'index + nr_to_read' is then ULONG_MAX rather than overflowed to 0.

Then 'index + nr_to_read > end_index' rather 'index + nr_to_read >= end_index',
since otherwise nr_to_read can be increased by 1 rather than decreased or stay
the same as expected.

- Eric