[PATCH] mmc: mtk-sd: use div64_u64() instead of do_div()

[PATCH] mmc: mtk-sd: use div64_u64() instead of do_div()

[Qing Wang]

From: Wang Qing <wangqing@vivo.com>

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 <wangqing@vivo.com>
---
 drivers/mmc/host/mtk-sd.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/mmc/host/mtk-sd.c b/drivers/mmc/host/mtk-sd.c
index 65037e1..777c9a8
--- a/drivers/mmc/host/mtk-sd.c
+++ b/drivers/mmc/host/mtk-sd.c
@@ -766,7 +766,7 @@ static u64 msdc_timeout_cal(struct msdc_host *host, u64 ns, u64 clks)
 		clk_ns  = 1000000000ULL;
 		do_div(clk_ns, mmc->actual_clock);
 		timeout = ns + clk_ns - 1;
-		do_div(timeout, clk_ns);
+		div64_u64(timeout, clk_ns);
 		timeout += clks;
 		/* in 1048576 sclk cycle unit */
 		timeout = DIV_ROUND_UP(timeout, BIT(20));
-- 
2.7.4

Re: [PATCH] mmc: mtk-sd: use div64_u64() instead of do_div()

[Ulf Hansson]

On Wed, 9 Feb 2022 at 09:39, Qing Wang <wangqing@vivo.com> wrote:
>
> From: Wang Qing <wangqing@vivo.com>
>
> 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 <wangqing@vivo.com>

Applied for next, thanks!

Kind regards
Uffe


> ---
>  drivers/mmc/host/mtk-sd.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/mmc/host/mtk-sd.c b/drivers/mmc/host/mtk-sd.c
> index 65037e1..777c9a8
> --- a/drivers/mmc/host/mtk-sd.c
> +++ b/drivers/mmc/host/mtk-sd.c
> @@ -766,7 +766,7 @@ static u64 msdc_timeout_cal(struct msdc_host *host, u64 ns, u64 clks)
>                 clk_ns  = 1000000000ULL;
>                 do_div(clk_ns, mmc->actual_clock);
>                 timeout = ns + clk_ns - 1;
> -               do_div(timeout, clk_ns);
> +               div64_u64(timeout, clk_ns);
>                 timeout += clks;
>                 /* in 1048576 sclk cycle unit */
>                 timeout = DIV_ROUND_UP(timeout, BIT(20));
> --
> 2.7.4
>

Re: [PATCH] mmc: mtk-sd: use div64_u64() instead of do_div()

[Christophe JAILLET]

Le 17/02/2022 à 16:39, Ulf Hansson a écrit :
> On Wed, 9 Feb 2022 at 09:39, Qing Wang <wangqing@vivo.com> wrote:
>>
>> From: Wang Qing <wangqing@vivo.com>
>>
>> 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 <wangqing@vivo.com>
> 
> Applied for next, thanks!


This is wrong.

See [1].


Wang Qing, you should really warn all the people you have sent such patches.

CJ

[1]: 
https://lore.kernel.org/linux-kernel/19b96972-cee7-937f-21ce-c78982ed2048@linaro.org/


> 
> Kind regards
> Uffe
> 
> 
>> ---
>>   drivers/mmc/host/mtk-sd.c | 2 +-
>>   1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/drivers/mmc/host/mtk-sd.c b/drivers/mmc/host/mtk-sd.c
>> index 65037e1..777c9a8
>> --- a/drivers/mmc/host/mtk-sd.c
>> +++ b/drivers/mmc/host/mtk-sd.c
>> @@ -766,7 +766,7 @@ static u64 msdc_timeout_cal(struct msdc_host *host, u64 ns, u64 clks)
>>                  clk_ns  = 1000000000ULL;
>>                  do_div(clk_ns, mmc->actual_clock);
>>                  timeout = ns + clk_ns - 1;
>> -               do_div(timeout, clk_ns);
>> +               div64_u64(timeout, clk_ns);
>>                  timeout += clks;
>>                  /* in 1048576 sclk cycle unit */
>>                  timeout = DIV_ROUND_UP(timeout, BIT(20));
>> --
>> 2.7.4
>>
> 

Re: [PATCH] mmc: mtk-sd: use div64_u64() instead of do_div()

[Ulf Hansson]

On Thu, 17 Feb 2022 at 21:00, Christophe JAILLET
<christophe.jaillet@wanadoo.fr> wrote:
>
> Le 17/02/2022 à 16:39, Ulf Hansson a écrit :
> > On Wed, 9 Feb 2022 at 09:39, Qing Wang <wangqing@vivo.com> wrote:
> >>
> >> From: Wang Qing <wangqing@vivo.com>
> >>
> >> 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 <wangqing@vivo.com>
> >
> > Applied for next, thanks!
>
>
> This is wrong.
>
> See [1].

Thanks for reporting this, I am dropping the patch from my next branch!

See more comments below.

>
>
> Wang Qing, you should really warn all the people you have sent such patches.
>
> CJ
>
> [1]:
> https://lore.kernel.org/linux-kernel/19b96972-cee7-937f-21ce-c78982ed2048@linaro.org/
>
>
> >
> > Kind regards
> > Uffe
> >
> >
> >> ---
> >>   drivers/mmc/host/mtk-sd.c | 2 +-
> >>   1 file changed, 1 insertion(+), 1 deletion(-)
> >>
> >> diff --git a/drivers/mmc/host/mtk-sd.c b/drivers/mmc/host/mtk-sd.c
> >> index 65037e1..777c9a8
> >> --- a/drivers/mmc/host/mtk-sd.c
> >> +++ b/drivers/mmc/host/mtk-sd.c
> >> @@ -766,7 +766,7 @@ static u64 msdc_timeout_cal(struct msdc_host *host, u64 ns, u64 clks)
> >>                  clk_ns  = 1000000000ULL;
> >>                  do_div(clk_ns, mmc->actual_clock);
> >>                  timeout = ns + clk_ns - 1;
> >> -               do_div(timeout, clk_ns);
> >> +               div64_u64(timeout, clk_ns);

I guess a proper patch would be to convert clk_ns into an u32 instead!?

> >>                  timeout += clks;
> >>                  /* in 1048576 sclk cycle unit */
> >>                  timeout = DIV_ROUND_UP(timeout, BIT(20));
> >> --

Kind regards
Uffe