[PATCH -next] ethernet: renesas: Use div64_ul instead of do_div

[PATCH -next] ethernet: renesas: Use div64_ul instead of do_div

[Yang Li]

do_div() does a 64-by-32 division. Here the divisor is an
unsigned long which on some platforms is 64 bit wide. So use
div64_ul instead of do_div to avoid a possible truncation.

Eliminate the following coccicheck warning:
./drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c:2742:1-7: WARNING:
do_div() does a 64-by-32 division, please consider using div64_ul
instead.

Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Signed-off-by: Yang Li <yang.lee@linux.alibaba.com>
---
 drivers/net/ethernet/renesas/ravb_main.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/renesas/ravb_main.c b/drivers/net/ethernet/renesas/ravb_main.c
index b4c597f..2b89710 100644
--- a/drivers/net/ethernet/renesas/ravb_main.c
+++ b/drivers/net/ethernet/renesas/ravb_main.c
@@ -2489,7 +2489,7 @@ static int ravb_set_gti(struct net_device *ndev)
 		return -EINVAL;
 
 	inc = 1000000000ULL << 20;
-	do_div(inc, rate);
+	inc = div64_ul(inc, rate);
 
 	if (inc < GTI_TIV_MIN || inc > GTI_TIV_MAX) {
 		dev_err(dev, "gti.tiv increment 0x%llx is outside the range 0x%x - 0x%x\n",
-- 
1.8.3.1

Re: [PATCH -next] ethernet: renesas: Use div64_ul instead of do_div

[Sergey Shtylyov]

Hello!

    Why you didn't Cc me (as a reviewer)?

On 18.11.2021 5:50, Yang Li wrote:

> do_div() does a 64-by-32 division. Here the divisor is an
> unsigned long which on some platforms is 64 bit wide. So use
> div64_ul instead of do_div to avoid a possible truncation.
> 
> Eliminate the following coccicheck warning:
> ./drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c:2742:1-7: WARNING:
> do_div() does a 64-by-32 division, please consider using div64_ul
> instead.
> 
> Reported-by: Abaci Robot <abaci@linux.alibaba.com>
> Signed-off-by: Yang Li <yang.lee@linux.alibaba.com>
> ---
>   drivers/net/ethernet/renesas/ravb_main.c | 2 +-
>   1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/net/ethernet/renesas/ravb_main.c b/drivers/net/ethernet/renesas/ravb_main.c
> index b4c597f..2b89710 100644
> --- a/drivers/net/ethernet/renesas/ravb_main.c
> +++ b/drivers/net/ethernet/renesas/ravb_main.c
> @@ -2489,7 +2489,7 @@ static int ravb_set_gti(struct net_device *ndev)
>   		return -EINVAL;
>   
>   	inc = 1000000000ULL << 20;
> -	do_div(inc, rate);
> +	inc = div64_ul(inc, rate);

    Why not just:

	inc = div64_ul(1000000000ULL << 20, rate);

>   	if (inc < GTI_TIV_MIN || inc > GTI_TIV_MAX) {
>   		dev_err(dev, "gti.tiv increment 0x%llx is outside the range 0x%x - 0x%x\n",

MBR, Sergey

RE: [PATCH -next] ethernet: renesas: Use div64_ul instead of do_div

[David Laight]

From: Sergey Shtylyov
> Sent: 18 November 2021 09:08
> On 18.11.2021 5:50, Yang Li wrote:
> 
> > do_div() does a 64-by-32 division. Here the divisor is an
> > unsigned long which on some platforms is 64 bit wide. So use
> > div64_ul instead of do_div to avoid a possible truncation.
> >
> > Eliminate the following coccicheck warning:
> > ./drivers/net/ethernet/marvell/mvpp2/mvpp2_main.c:2742:1-7: WARNING:
> > do_div() does a 64-by-32 division, please consider using div64_ul
> > instead.
> >
> > Reported-by: Abaci Robot <abaci@linux.alibaba.com>
> > Signed-off-by: Yang Li <yang.lee@linux.alibaba.com>
> > ---
> >   drivers/net/ethernet/renesas/ravb_main.c | 2 +-
> >   1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/drivers/net/ethernet/renesas/ravb_main.c b/drivers/net/ethernet/renesas/ravb_main.c
> > index b4c597f..2b89710 100644
> > --- a/drivers/net/ethernet/renesas/ravb_main.c
> > +++ b/drivers/net/ethernet/renesas/ravb_main.c
> > @@ -2489,7 +2489,7 @@ static int ravb_set_gti(struct net_device *ndev)
> >   		return -EINVAL;
> >
> >   	inc = 1000000000ULL << 20;
> > -	do_div(inc, rate);
> > +	inc = div64_ul(inc, rate);
> 
>     Why not just:
> 
> 	inc = div64_ul(1000000000ULL << 20, rate);
> 
> >   	if (inc < GTI_TIV_MIN || inc > GTI_TIV_MAX) {
> >   		dev_err(dev, "gti.tiv increment 0x%llx is outside the range 0x%x - 0x%x\n",

Even with hardware divide a 64/32 divide is likely to be faster that a 64/64 one.

Maybe the coccicheck warning message should suggest checking the domain
of the divisor and then changing the type??

	David

-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
Registration No: 1397386 (Wales)