I am not sure, maybe you are right, we will never the case.
But I known this patch can avoid the case when probing a FS
fail with current magics.
rosys...@rosinson.com
From: John Crispin
Date: 2018-04-02 23:25
To: rosys...@rosinson.com; lede-dev
Subject: Re: [LEDE-DEV] [LEDE-DEV,2/3] fstools: fix potential memory leak
On 29/03/18 10:25, rosys...@rosinson.com wrote:
> Fix potential memory leak in fstools
>
> Signed-off-by: Rosy Song <rosys...@rosinson.com>
Hi,
the description is wrong, this fixes a potential null pointer deref and
i dont really see how probing of a FS would work without a magic so we
probably wont ever hit this case ?
John
> ---
> libblkid-tiny/libblkid-tiny.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/libblkid-tiny/libblkid-tiny.c b/libblkid-tiny/libblkid-tiny.c
> index ccfaf63..12cab6f 100644
> --- a/libblkid-tiny/libblkid-tiny.c
> +++ b/libblkid-tiny/libblkid-tiny.c
> @@ -282,7 +282,7 @@ int probe_block(char *block, struct blkid_struct_probe
> *pr)
>
> mag = &idinfos[i]->magics[0];
>
> - while (mag->magic) {
> + while (mag && mag->magic) {
> int off = (mag->kboff * 1024) + mag->sboff;
> char magic[32] = { 0 };
>
> --
> 2.13.3
_______________________________________________
Lede-dev mailing list
Lede-dev@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/lede-dev
