Re: [PATCH v4 6/8] tcp: Cast operands of sequence comparison macros to uint32_t before using them

[Jon Maloy <jmaloy@redhat.com>]

On 2025-09-09 14:16, Stefano Brivio wrote:
> Otherwise, passing signed types causes automatic promotion of the
> result of the subtractions as well, which is not what we want, as
> these macros rely on unsigned 32-bit arithmetic.
> 
> The next patch introduces a ssize_t operand for SEQ_LE, illustrating
> the issue.
> 
> Signed-off-by: Stefano Brivio <sbrivio@redhat.com>

Reviewed-by: Jon Maloy <jmaloy@redhat.com>

> ---
>   tcp_internal.h | 12 ++++++++----
>   1 file changed, 8 insertions(+), 4 deletions(-)
> 
> diff --git a/tcp_internal.h b/tcp_internal.h
> index d0009f8..ce6fee2 100644
> --- a/tcp_internal.h
> +++ b/tcp_internal.h
> @@ -21,10 +21,14 @@
>   						   sizeof(struct ipv6hdr), \
>   						   sizeof(uint32_t))
>   
> -#define SEQ_LE(a, b)			((b) - (a) < MAX_WINDOW)
> -#define SEQ_LT(a, b)			((b) - (a) - 1 < MAX_WINDOW)
> -#define SEQ_GE(a, b)			((a) - (b) < MAX_WINDOW)
> -#define SEQ_GT(a, b)			((a) - (b) - 1 < MAX_WINDOW)
> +#define SEQ_LE(a, b)			\
> +	((uint32_t)(b) - (uint32_t)(a) < MAX_WINDOW)
> +#define SEQ_LT(a, b)			\
> +	((uint32_t)(b) - (uint32_t)(a) - 1 < MAX_WINDOW)
> +#define SEQ_GE(a, b)			\
> +	((uint32_t)(a) - (uint32_t)(b) < MAX_WINDOW)
> +#define SEQ_GT(a, b)			\
> +	((uint32_t)(a) - (uint32_t)(b) - 1 < MAX_WINDOW)
>   
>   #define FIN		(1 << 0)
>   #define SYN		(1 << 1)