2022-02-09 12:58:17

by 王擎

Subject: [PATCH] net: dsa: ocelot: use div64_u64() instead of do_div()

From: Wang Qing <[email protected]>

do_div() does a 64-by-32 division.
When the divisor is u64, do_div() truncates it to 32 bits, this means it
can test non-zero and be truncated to zero for division.

fix do_div.cocci warning:
do_div() does a 64-by-32 division, please consider using div64_u64 instead.

Signed-off-by: Wang Qing <[email protected]>
---
drivers/net/dsa/ocelot/felix_vsc9959.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/dsa/ocelot/felix_vsc9959.c b/drivers/net/dsa/ocelot/felix_vsc9959.c
index bf8d382..5c2482f
--- a/drivers/net/dsa/ocelot/felix_vsc9959.c
+++ b/drivers/net/dsa/ocelot/felix_vsc9959.c
@@ -1178,7 +1178,7 @@ static void vsc9959_new_base_time(struct ocelot *ocelot, ktime_t base_time,
if (base_time < current_time) {
u64 nr_of_cycles = current_time - base_time;

- do_div(nr_of_cycles, cycle_time);
+ div64_u64(nr_of_cycles, cycle_time);
new_base_time += cycle_time * (nr_of_cycles + 1);
}

--
2.7.4

2022-02-09 19:58:42

by Vladimir Oltean

Subject: Re: [PATCH] net: dsa: ocelot: use div64_u64() instead of do_div()

Hi Wang,

On Wed, Feb 09, 2022 at 12:39:02AM -0800, Qing Wang wrote:
> From: Wang Qing <[email protected]>
>
> do_div() does a 64-by-32 division.
> When the divisor is u64, do_div() truncates it to 32 bits, this means it
> can test non-zero and be truncated to zero for division.
>
> fix do_div.cocci warning:
> do_div() does a 64-by-32 division, please consider using div64_u64 instead.
>
> Signed-off-by: Wang Qing <[email protected]>
> ---
> drivers/net/dsa/ocelot/felix_vsc9959.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/net/dsa/ocelot/felix_vsc9959.c b/drivers/net/dsa/ocelot/felix_vsc9959.c
> index bf8d382..5c2482f
> --- a/drivers/net/dsa/ocelot/felix_vsc9959.c
> +++ b/drivers/net/dsa/ocelot/felix_vsc9959.c
> @@ -1178,7 +1178,7 @@ static void vsc9959_new_base_time(struct ocelot *ocelot, ktime_t base_time,
> if (base_time < current_time) {
> u64 nr_of_cycles = current_time - base_time;
>
> - do_div(nr_of_cycles, cycle_time);
> + div64_u64(nr_of_cycles, cycle_time);
> new_base_time += cycle_time * (nr_of_cycles + 1);
> }
>
> --
> 2.7.4
>

I would prefer that you teach your scripts that, if a range check exists
for the divisor prior to the division, it gets taken into consideration.

vsc9959_qos_port_tas_set:

if (taprio->cycle_time > NSEC_PER_SEC ||
taprio->cycle_time_extension >= NSEC_PER_SEC)
return -EINVAL;

vsc9959_new_base_time(ocelot, taprio->base_time,
taprio->cycle_time, &base_ts);

vsc9959_psfp_sgi_set:

if (sgi->cycletime < VSC9959_PSFP_GATE_CYCLETIME_MIN ||
sgi->cycletime > NSEC_PER_SEC)
return -EINVAL;

vsc9959_new_base_time(ocelot, sgi->basetime, sgi->cycletime, &base_ts);

So all callers provide a cycle_time argument that is smaller than
NSEC_PER_SEC (1000000000L = 0x3B9ACA00 => fits on 32 bits).