linux-kernel - Re: [PATCH v7 03/13] hwmon: pwm-fan: Use 64-bit division macros for period and duty cycle

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Tue, 10 Mar 2020 08:05:58 -0700
From:   Guenter Roeck <linux@...ck-us.net>
To:     Uwe Kleine-König <u.kleine-koenig@...gutronix.de>
Cc:     Guru Das Srinagesh <gurus@...eaurora.org>,
        linux-pwm@...r.kernel.org,
        Thierry Reding <thierry.reding@...il.com>,
        Subbaraman Narayanamurthy <subbaram@...eaurora.org>,
        linux-kernel@...r.kernel.org, Kamil Debski <kamil@...as.org>,
        Bartlomiej Zolnierkiewicz <b.zolnierkie@...sung.com>,
        Jean Delvare <jdelvare@...e.com>,
        Liam Girdwood <lgirdwood@...il.com>,
        Mark Brown <broonie@...nel.org>, linux-hwmon@...r.kernel.org
Subject: Re: [PATCH v7 03/13] hwmon: pwm-fan: Use 64-bit division macros for
 period and duty cycle

On 3/10/20 5:08 AM, Uwe Kleine-König wrote:
> Hello Guenter,
> 
> On Mon, Mar 09, 2020 at 02:48:22PM -0700, Guenter Roeck wrote:
>> On Mon, Mar 09, 2020 at 12:35:06PM -0700, Guru Das Srinagesh wrote:
>>> Because period and duty cycle are defined in the PWM framework structs
>>> as ints with units of nanoseconds, the maximum time duration that can be
>>> set is limited to ~2.147 seconds. Redefining them as u64 values will
>>> enable larger time durations to be set.
>>>
>>> As a first step, prepare drivers to handle the switch to u64 period and
>>> duty_cycle by replacing division operations involving pwm period and duty cycle
>>> with their 64-bit equivalents as appropriate. The actual switch to u64 period
>>> and duty_cycle follows as a separate patch.
>>>
>>> Where the dividend is 64-bit but the divisor is 32-bit, use *_ULL
>>> macros:
>>> - DIV_ROUND_UP_ULL
>>> - DIV_ROUND_CLOSEST_ULL
>>> - div_u64
>>>
>>> Where the divisor is 64-bit (dividend may be 32-bit or 64-bit), use
>>> DIV64_* macros:
>>> - DIV64_U64_ROUND_CLOSEST
>>> - div64_u64
>>
>> There is no explanation why this is necessary. What is the use case ?
>> It is hard to imagine a real-world use case with a duty cycle of more
>> than 2 seconds.
> 
> When my Laptop is in suspend there is an LED that blinks with a period
> of approximately 5 seconds. (To be fair, the brightness is more a sinus
> than a rectangle, but still.)
> 

I don't see support in the LED subsystem to utilize the PWM framework directly
for blinking. Plus, you say yourself that it isn't a _real_ use case, just a
theoretic one.

Either case, the reason / use case for this series should be explained
in the summary patch. That is what it is for. That case is not made.

Guenter