[pacman-dev] [PATCH] Reject files larger than INT_MAX in read_sigfile.

[Andrew Gregory]

On 06/05/16 at 07:49pm, Tobias Stoeckmann wrote:
> Signature files larger than INT_MAX are already suspicious,
> but if they are larger than SIZE_MAX, this code couldn't even
> copy them into memory, accepting them as "blank" files at worst.
> 
> While adding the INT_MAX check, I also rearranged the code to
> avoid a quite harmless TOCTOU race condition between
> stat() and fopen().
> 
> Signed-off-by: Tobias Stoeckmann <tobias at stoeckmann.org>
> ---
> Thanks for pointing out the flaw Florian!
> ---
>  lib/libalpm/be_package.c | 10 +++++-----
>  1 file changed, 5 insertions(+), 5 deletions(-)
> 
> diff --git a/lib/libalpm/be_package.c b/lib/libalpm/be_package.c
> index c9ed770..055fb1e 100644
> --- a/lib/libalpm/be_package.c
> +++ b/lib/libalpm/be_package.c
> @@ -700,17 +700,17 @@ static int read_sigfile(const char *sigpath, unsigned char **sig)
>  	struct stat st;
>  	FILE *fp;
>  
> -	if(stat(sigpath, &st) != 0) {
> +	if((fp = fopen(sigpath, "rb")) == NULL) {
>  		return -1;
>  	}
>  
> -	MALLOC(*sig, st.st_size, return -1);
> -
> -	if((fp = fopen(sigpath, "rb")) == NULL) {
> -		free(*sig);
> +	if(fstat(fileno(fp), &st) != 0 || st.st_size > INT_MAX) {

limits.h should be included for INT_MAX.  Is there not a more
meaningful limit we can use for this than INT_MAX?

> +		fclose(fp);
>  		return -1;
>  	}
>  
> +	MALLOC(*sig, st.st_size, return -1);

Needs fclose(fp)

> +
>  	if(fread(*sig, st.st_size, 1, fp) != 1) {
>  		free(*sig);
>  		fclose(fp);
> -- 
> 2.8.3