Re: [PATCH 1/2] exec: load_script: don't blindly truncate shebang string

[Michal Hocko]

On Mon 12-11-18 17:09:31, Oleg Nesterov wrote:
> load_script() simply truncates bprm->buf and this is very wrong if the
> length of shebang string exceeds BINPRM_BUF_SIZE-2. This can silently
> truncate i_arg or (worse) we can execute the wrong binary if buf[2:126]
> happens to be the valid executable path.
> 
> Change load_script() to return ENOEXEC if it can't find '\n' or zero in
> bprm->buf. Note that '\0' can come from either prepare_binprm()->memset()
> or from kernel_read(), we do not care.
> 
> Signed-off-by: Oleg Nesterov <oleg@xxxxxxxxxx>

A bit cryptic to my taste but it looks correct. I have tried to come up
with something more tasty but I am afraid it would be just a matter of
taste.

Acked-by: Michal Hocko <mhocko@xxxxxxxx>

> ---
>  fs/binfmt_script.c | 10 +++++++---
>  1 file changed, 7 insertions(+), 3 deletions(-)
> 
> diff --git a/fs/binfmt_script.c b/fs/binfmt_script.c
> index 7cde3f4..d0078cb 100644
> --- a/fs/binfmt_script.c
> +++ b/fs/binfmt_script.c
> @@ -42,10 +42,14 @@ static int load_script(struct linux_binprm *bprm)
>  	fput(bprm->file);
>  	bprm->file = NULL;
>  
> -	bprm->buf[BINPRM_BUF_SIZE - 1] = '\0';
> -	if ((cp = strchr(bprm->buf, '\n')) == NULL)
> -		cp = bprm->buf+BINPRM_BUF_SIZE-1;
> +	for (cp = bprm->buf+2;; cp++) {
> +		if (cp >= bprm->buf + BINPRM_BUF_SIZE)
> +			return -ENOEXEC;
> +		if (!*cp || (*cp == '\n'))
> +			break;
> +	}
>  	*cp = '\0';
> +
>  	while (cp > bprm->buf) {
>  		cp--;
>  		if ((*cp == ' ') || (*cp == '\t'))
> -- 
> 2.5.0
> 
> 

-- 
Michal Hocko
SUSE Labs