linux-kernel - Re: [PATCH V2] gpio: brcmstb: support gpio-line-names property

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]

Message-ID: <7fcc5cb2-5fdb-d1cf-e55b-c0f2d407e072@gmail.com>
Date:   Wed, 11 Mar 2020 12:02:54 -0700
From:   Doug Berger <opendmb@...il.com>
To:     Bartosz Golaszewski <bgolaszewski@...libre.com>
Cc:     Gregory Fong <gregory.0xf0@...il.com>,
        Linus Walleij <linus.walleij@...aro.org>,
        Florian Fainelli <f.fainelli@...il.com>,
        bcm-kernel-feedback-list@...adcom.com,
        linux-gpio <linux-gpio@...r.kernel.org>,
        arm-soc <linux-arm-kernel@...ts.infradead.org>,
        LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH V2] gpio: brcmstb: support gpio-line-names property

On 3/11/20 5:44 AM, Bartosz Golaszewski wrote:
> pon., 9 mar 2020 o 20:02 Doug Berger <opendmb@...il.com> napisał(a):
>>
>> The default handling of the gpio-line-names property by the
>> gpiolib-of implementation does not work with the multiple
>> gpiochip banks per device structure used by the gpio-brcmstb
>> driver.
>>
>> This commit adds driver level support for the device tree
>> property so that GPIO lines can be assigned friendly names.
>>
>> Signed-off-by: Doug Berger <opendmb@...il.com>
>> ---
>>  drivers/gpio/gpio-brcmstb.c | 44 ++++++++++++++++++++++++++++++++++++++++++++
>>  1 file changed, 44 insertions(+)
>>
>> diff --git a/drivers/gpio/gpio-brcmstb.c b/drivers/gpio/gpio-brcmstb.c
>> index 05e3f99ae59c..fcfc1a1f1a5c 100644
>> --- a/drivers/gpio/gpio-brcmstb.c
>> +++ b/drivers/gpio/gpio-brcmstb.c
>> @@ -603,6 +603,49 @@ static const struct dev_pm_ops brcmstb_gpio_pm_ops = {
>>         .resume_noirq = brcmstb_gpio_resume,
>>  };
>>
>> +static void brcmstb_gpio_set_names(struct device *dev,
>> +                                  struct brcmstb_gpio_bank *bank)
>> +{
>> +       struct device_node *np = dev->of_node;
>> +       const char **names;
>> +       int nstrings, base;
>> +       unsigned int i;
>> +
>> +       base = bank->id * MAX_GPIO_PER_BANK;
>> +
>> +       nstrings = of_property_count_strings(np, "gpio-line-names");
>> +       if (nstrings <= base)
>> +               /* Line names not present */
>> +               return;
>> +
>> +       names = devm_kcalloc(dev, MAX_GPIO_PER_BANK, sizeof(*names),
>> +                            GFP_KERNEL);
>> +       if (!names)
>> +               return;
>> +
>> +       /*
>> +        * Make sure to not index beyond the end of the number of descriptors
>> +        * of the GPIO device.
>> +        */
>> +       for (i = 0; i < bank->width; i++) {
>> +               const char *name;
>> +               int ret;
>> +
>> +               ret = of_property_read_string_index(np, "gpio-line-names",
>> +                                                   base + i, &name);
>> +               if (ret) {
>> +                       if (ret != -ENODATA)
>> +                               dev_err(dev, "unable to name line %d: %d\n",
>> +                                       base + i, ret);
>> +                       break;
>> +               }
> 
> This bit is confusing to me. If we can't read the property we break
> the loop and leave the remaining line names null but at the same time
> it isn't considered a probe failure? Would you mind at least
> commenting on this in the code?
> 
> Bart
> 
The label names are viewed as a convenience for the user and are not
necessary for the proper functionality of the driver and device, so we
don't want to prevent the driver from succeeding at probe due to an
error in the gpio-line-names property. The bank->gc.names member is
still made non-NULL which is what we really care about to prevent the
misapplication of label names by devprop_gpiochip_set_names().

In fact, it is expected that the device-tree will only include label
strings up to the last GPIO of interest so the ENODATA error is
considered a valid result to terminate any further labeling so there is
no need for an error message in that case.

Other error results are unexpected so an error message indicating the
consequence of the error is appropriate here.

I'm not sure which aspect is confusing to you, so it's not clear to me
how best to comment the code. I can hazard a guess, but if you have a
suggestion I'm happy to submit a v3.

Thanks for taking the time to review this,
    Doug