2020-08-15 22:03:46

by Phillip Lougher

Subject: [PATCH] squashfs: avoid bio_alloc() failure with 1Mbyte blocks

This is a regression introduced by the "migrate from ll_rw_block usage
to BIO" patch.

Bio_alloc() is limited to 256 pages (1 Mbyte). This can cause a
failure when reading 1 Mbyte block filesystems. The problem is
a datablock can be fully (or almost uncompressed), requiring 256
pages, but, because blocks are not aligned to page boundaries, it
may require 257 pages to read.

Bio_kmalloc() can handle 1024 pages, and so use this for the
edge condition.

Reported-by: Nicolas Prochazka <[email protected]>
Reported-by: Tomoatsu Shimada <[email protected]>
Signed-off-by: Phillip Lougher <[email protected]>
---
fs/squashfs/block.c | 6 +++++-
1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/fs/squashfs/block.c b/fs/squashfs/block.c
index 76bb1c846845..8a19773b5a0b 100644
--- a/fs/squashfs/block.c
+++ b/fs/squashfs/block.c
@@ -87,7 +87,11 @@ static int squashfs_bio_read(struct super_block *sb, u64 index, int length,
int error, i;
struct bio *bio;

- bio = bio_alloc(GFP_NOIO, page_count);
+ if (page_count <= BIO_MAX_PAGES)
+ bio = bio_alloc(GFP_NOIO, page_count);
+ else
+ bio = bio_kmalloc(GFP_NOIO, page_count);
+
if (!bio)
return -ENOMEM;

--
2.20.1

2020-08-16 01:45:45

by Guenter Roeck

Subject: Re: [PATCH] squashfs: avoid bio_alloc() failure with 1Mbyte blocks

On Fri, Aug 14, 2020 at 8:57 PM Phillip Lougher <[email protected]> wrote:
>
> This is a regression introduced by the "migrate from ll_rw_block usage
> to BIO" patch.
>
> Bio_alloc() is limited to 256 pages (1 Mbyte). This can cause a
> failure when reading 1 Mbyte block filesystems. The problem is
> a datablock can be fully (or almost uncompressed), requiring 256
> pages, but, because blocks are not aligned to page boundaries, it
> may require 257 pages to read.
>
> Bio_kmalloc() can handle 1024 pages, and so use this for the
> edge condition.
>
> Reported-by: Nicolas Prochazka <[email protected]>
> Reported-by: Tomoatsu Shimada <[email protected]>
> Signed-off-by: Phillip Lougher <[email protected]>

Fixes: 93e72b3c612a ("squashfs: migrate from ll_rw_block usage to BIO")
Reviewed-by: Guenter Roeck <[email protected]>

> ---
> fs/squashfs/block.c | 6 +++++-
> 1 file changed, 5 insertions(+), 1 deletion(-)
>
> diff --git a/fs/squashfs/block.c b/fs/squashfs/block.c
> index 76bb1c846845..8a19773b5a0b 100644
> --- a/fs/squashfs/block.c
> +++ b/fs/squashfs/block.c
> @@ -87,7 +87,11 @@ static int squashfs_bio_read(struct super_block *sb, u64 index, int length,
> int error, i;
> struct bio *bio;
>
> - bio = bio_alloc(GFP_NOIO, page_count);
> + if (page_count <= BIO_MAX_PAGES)
> + bio = bio_alloc(GFP_NOIO, page_count);
> + else
> + bio = bio_kmalloc(GFP_NOIO, page_count);
> +
> if (!bio)
> return -ENOMEM;
>
> --
> 2.20.1
>

2020-08-19 03:42:25

by Andrew Morton

Subject: Re: [PATCH] squashfs: avoid bio_alloc() failure with 1Mbyte blocks

On Fri, 14 Aug 2020 21:30:33 -0700 Guenter Roeck <[email protected]> wrote:

> On Fri, Aug 14, 2020 at 8:57 PM Phillip Lougher <[email protected]> wrote:
> >
> > This is a regression introduced by the "migrate from ll_rw_block usage
> > to BIO" patch.
> >
> > Bio_alloc() is limited to 256 pages (1 Mbyte). This can cause a
> > failure when reading 1 Mbyte block filesystems. The problem is
> > a datablock can be fully (or almost uncompressed), requiring 256
> > pages, but, because blocks are not aligned to page boundaries, it
> > may require 257 pages to read.
> >
> > Bio_kmalloc() can handle 1024 pages, and so use this for the
> > edge condition.
> >
> > Reported-by: Nicolas Prochazka <[email protected]>
> > Reported-by: Tomoatsu Shimada <[email protected]>
> > Signed-off-by: Phillip Lougher <[email protected]>
>
> Fixes: 93e72b3c612a ("squashfs: migrate from ll_rw_block usage to BIO")
> Reviewed-by: Guenter Roeck <[email protected]>

Thanks. I added cc:stable also.