Re: [PATCH] um: proc/exitcode: fix simple_strtol() out-of-bounds read

["Shengzhuo Wei" <me@cherr.cc>]

On 2026-04-23 01:39, Shengzhuo Wei wrote:
> write(2) should report the number of bytes consumed. Returning the original
> count would claim success even when the input was truncated, so userspace
> cannot detect it.
> 
> Clamp the copy length to sizeof(buf)-1, add a terminator, and return the
> consumed length.
> 
> Fixes: 201f99f170df ("uml: check length in exitcode_proc_write()")
> Fixes: e16f5350d4cf ("uml: get declaration of simple_strtoul")
> Signed-off-by: Shengzhuo Wei <me@cherr.cc>
> ---
>  arch/um/kernel/exitcode.c | 5 +++--
>  1 file changed, 3 insertions(+), 2 deletions(-)
> 
> diff --git a/arch/um/kernel/exitcode.c b/arch/um/kernel/exitcode.c
> index 43edc2aa57e4fbd4a3d24f96878c76f9f8fd4eaa..8de404ff21a213918c5351bc20a6e047bf1b93f5 100644
> --- a/arch/um/kernel/exitcode.c
> +++ b/arch/um/kernel/exitcode.c
> @@ -43,16 +43,17 @@ static ssize_t exitcode_proc_write(struct file *file,
>  	size_t size;
>  	int tmp;
>  
> -	size = min(count, sizeof(buf));
> +	size = min(count, sizeof(buf) - 1);
>  	if (copy_from_user(buf, buffer, size))
>  		return -EFAULT;
> +	buf[size] = '\0';
>  
>  	tmp = simple_strtol(buf, &end, 0);
>  	if ((*end != '\0') && !isspace(*end))
>  		return -EINVAL;
>  
>  	uml_exitcode = tmp;
> -	return count;
> +	return size;

There is another problem that have been talked about in another patch.
Not consuming the whole string passed by user may (and will) break
some programs that want to ensure "FULL WRITE" that use some function
like "while (len > 0) { len -= write(); } ".

Should we just let the return count as-is and fix the OOB problem only?

Link: https://lore.kernel.org/all/20260424063531.d1508a3c79c4b7808ed04420@linux-foundation.org/t/#u
Link: https://lore.kernel.org/all/ef2301c4-bbc5-496e-b9d2-9adcff8b8d42@p183/

Regards,
Shengzhuo Wei

>  }
>  
>  static const struct proc_ops exitcode_proc_ops = {
> 
> ---