On 8/29/25 04:12, Rong Tao wrote:
> From: Rong Tao <[email protected]>
>
> strnstr should not treat the ending '\0' of s2 as a matching character
> if the parameter 'len' equal to s2 string length, for example:
A good catch, thanks!
But this doesn't fix just the `len == strlen(s2)` case but a more
general case when s2 is a suffix of the first len characters of s1,
right? The commit message should reflect that.
>
> 1. bpf_strnstr("openat", "open", 4) = -ENOENT
> 2. bpf_strnstr("openat", "open", 5) = 0
>
> This patch makes (1) return 0, indicating a successful match.
>
> Fixes: e91370550f1f ("bpf: Add kfuncs for read-only string operations")
> Signed-off-by: Rong Tao <[email protected]>
> ---
> kernel/bpf/helpers.c | 10 +++++++++-
> 1 file changed, 9 insertions(+), 1 deletion(-)
>
> diff --git a/kernel/bpf/helpers.c b/kernel/bpf/helpers.c
> index 401b4932cc49..bf04881f96ec 100644
> --- a/kernel/bpf/helpers.c
> +++ b/kernel/bpf/helpers.c
> @@ -3672,10 +3672,18 @@ __bpf_kfunc int bpf_strnstr(const char *s1__ign,
> const char *s2__ign, size_t len
>
> guard(pagefault)();
> for (i = 0; i < XATTR_SIZE_MAX; i++) {
> - for (j = 0; i + j < len && j < XATTR_SIZE_MAX; j++) {
> + for (j = 0; i + j <= len && j < XATTR_SIZE_MAX; j++) {
> __get_kernel_nofault(&c2, s2__ign + j, char, err_out);
> if (c2 == '\0')
> return i;
> + /**
> + * corner case i+j==len to ensure that we matched
> + * entire s2. for example, param len=3:
> + * s1: A B C D E F -> i==1
> + * s2: B C D -> j==2
> + */
This is not really a good example as it's not clear whether D is a null
byte or not. How about something like:
/**
* We allow reading an extra byte from s2 (note the
* `i + j <= len` above) to cover the case when s2 is
* a suffix of the first len chars of s1.
*/
> + if (i + j == len)
> + break;
Viktor
> __get_kernel_nofault(&c1, s1__ign + j, char, err_out);
> if (c1 == '\0')
> return -ENOENT;
