linux-kernel - Re: [BUG] mmc_regulator_set_ocr can't cope with regulator-fixed

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:   Fri, 6 Aug 2021 13:44:45 +0530
From:   Ravikumar Kattekola <callmerk1986@...il.com>
To:     Mark Brown <broonie@...nel.org>, Peter Geis <pgwipeout@...il.com>
Cc:     Jaehoon Chung <jh80.chung@...sung.com>,
        Ulf Hansson <ulf.hansson@...aro.org>,
        Liam Girdwood <lgirdwood@...il.com>,
        Rob Herring <robh+dt@...nel.org>, linux-mmc@...r.kernel.org,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        "open list:ARM/Rockchip SoC..." <linux-rockchip@...ts.infradead.org>,
        devicetree@...r.kernel.org
Subject: Re: [BUG] mmc_regulator_set_ocr can't cope with regulator-fixed

Hi,
Resending my reply as my Mail client settings prevented delivery

On 05/08/21 6:38 pm, Mark Brown wrote:
> On Thu, Aug 05, 2021 at 08:58:58AM -0400, Peter Geis wrote:
>> On Thu, Aug 5, 2021 at 8:47 AM Mark Brown <broonie@...nel.org> wrote:
> 
>>> One thing to watch out for with this approach is if there's things that
>>> really need a specific voltage to be set then you'll have to stop those
>>> things happening if you've got a voltage regulator that can't deliver a
>>> voltage in the required range.  I don't know if this affects MMC or not,
>>> if it's just a case of being less efficient it's not such an issue.
> 
>> Yeah, but if this is a fixed regulator and it's a problem, then the
>> hardware is screwed anyways.
> 
> Well, the fact that the voltage is being changed at runtime indicates
> that we're changing something from whatever was in the fixed setup - it
> can sometimes be that we don't have access to some higher performance or
> lower power features for example.  That's not ideal but works perfectly
> safely.
> 
Suggested approach of checking "mmc->ocr_avail" might work.

But, IMO mmc core should check if the voltage can be changed or not

before trying to do regulator_set_voltage() in mmc_regulator_set_ocr().

Wouldn't that be better and solve this issue for other hosts as well.

Something like below in mmc_regulator_set_ocr ():

+               result = regulator_check_voltage_constraints(supply,
+                       min_uV, max_uV);
+               if(!result) {
+                       result = regulator_set_voltage(supply, min_uV, 
max_uV);
+                       if (result != -EINVAL && !mmc->regulator_enabled) {
+                               result = regulator_enable(supply);
+                               if (!result)
+                                       mmc->regulator_enabled = true;
+                       }

We could wrap the existing check_voltage function

+/* Check voltage constraints helper function */
+int regulator_check_voltage_constraints(struct regulator *regulator,
+                                       int min_uV, int max_uV)
+{
+       return regulator_check_voltage(regulator->rdev, &min_uV, &max_uV);
+}
+EXPORT_SYMBOL_GPL(regulator_check_voltage_constraints);

I hope this makes sense.

Regards,
RK