| 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
| ||
|
Message-ID: <54C90AAC.3010907@collabora.co.uk>
Date: Wed, 28 Jan 2015 17:13:32 +0100
From: Javier Martinez Canillas <javier.martinez@...labora.co.uk>
To: Srinivas Kandagatla <srinivas.kandagatla@...aro.org>,
Ulf Hansson <ulf.hansson@...aro.org>
CC: devicetree@...r.kernel.org, linux-samsung-soc@...r.kernel.org,
linux-mmc@...r.kernel.org, Doug Anderson <dianders@...omium.org>,
linux-kernel@...r.kernel.org, Kukjin Kim <kgene@...nel.org>,
Olof Johansson <olof@...om.net>,
linux-arm-kernel@...ts.infradead.org
Subject: Re: [PATCH 2/5] mmc: pwrseq_simple: Extend to support more pins
Hello Srinivas,
Thanks a lot for your feedback.
On 01/28/2015 03:01 PM, Srinivas Kandagatla wrote:
> Hi Javier,
>
> You are in a lead of 3 hrs from me..
> Surprisingly I send very much same patch just few Mins ago :-)
:-)
I didn't find the posted patch you are referring too though, did you cc
linux-mmc?
> May be we can merge goods in both :-)
>
Sure, I want $subject to be generic enough to be useful for other platforms.
> On 28/01/15 10:10, Javier Martinez Canillas wrote:
>> Many WLAN attached to a SDIO/MMC interface, needs more than one pin for
>> their reset sequence. For example, is very common for chips to have two
>> pins: one for reset and one for power enable.
>>
>> This patch adds support for more reset pins to the pwrseq_simple driver
>> and instead hardcoding a fixed number, it uses the of_gpio_named_count()
>> since the MMC power sequence is only built when CONFIG_OF is enabled.
>>
>> Signed-off-by: Javier Martinez Canillas <javier.martinez@...labora.co.uk>
>> ---
>> drivers/mmc/core/pwrseq_simple.c | 54 ++++++++++++++++++++++++++++++----------
>> 1 file changed, 41 insertions(+), 13 deletions(-)
>>
>> diff --git a/drivers/mmc/core/pwrseq_simple.c b/drivers/mmc/core/pwrseq_simple.c
>> index 0958c696137f..9e51fe1051c5 100644
>> --- a/drivers/mmc/core/pwrseq_simple.c
>> +++ b/drivers/mmc/core/pwrseq_simple.c
>> @@ -11,6 +11,7 @@
>> #include <linux/slab.h>
>> #include <linux/device.h>
>> #include <linux/err.h>
>> +#include <linux/of_gpio.h>
>> #include <linux/gpio/consumer.h>
>>
>> #include <linux/mmc/host.h>
>> @@ -19,34 +20,44 @@
>>
>> struct mmc_pwrseq_simple {
>> struct mmc_pwrseq pwrseq;
>> - struct gpio_desc *reset_gpio;
>> + struct gpio_desc **reset_gpio;
>
> May be renaming it to reset_gpios makes more sense..
>
Ok
> If you make this struct gpio_desc *reset_gpios[0]; You can aviod an
> extra kmalloc and free ..
>
>
That's a very good idea, thanks.
>> + int nr_gpios;
>> };
>>
>> static void mmc_pwrseq_simple_pre_power_on(struct mmc_host *host)
>> {
>
> [...
>> struct mmc_pwrseq_simple *pwrseq = container_of(host->pwrseq,
>> struct mmc_pwrseq_simple, pwrseq);
>> + int i;
>>
>> - if (!IS_ERR(pwrseq->reset_gpio))
>> - gpiod_set_value_cansleep(pwrseq->reset_gpio, 1);
>> + for (i = 0; i < pwrseq->nr_gpios; i++)
>> + if (!IS_ERR(pwrseq->reset_gpio[i]))
>> + gpiod_set_value_cansleep(pwrseq->reset_gpio[i], 1);
>
> ...]
>
>> }
>>
>> static void mmc_pwrseq_simple_post_power_on(struct mmc_host *host)
>> {
>
> [...
>
>> struct mmc_pwrseq_simple *pwrseq = container_of(host->pwrseq,
>> struct mmc_pwrseq_simple, pwrseq);
>> + int i;
>>
>> - if (!IS_ERR(pwrseq->reset_gpio))
>> - gpiod_set_value_cansleep(pwrseq->reset_gpio, 0);
>> + for (i = 0; i < pwrseq->nr_gpios; i++)
>> + if (!IS_ERR(pwrseq->reset_gpio[i]))
>> + gpiod_set_value_cansleep(pwrseq->reset_gpio[i], 0);
> ...]
>
> Now that we have more code in mmc_pwrseq_simple_post_power_on() and
> mmc_pwrseq_simple_pre_power_on(), Just move most of them into a common
> function like:
>
> static void __mmc_pwrseq_simple_power_on_off(struct mmc_host *host,
> bool on)
> {
> struct mmc_pwrseq_simple *pwrseq = container_of(host->pwrseq,
> struct mmc_pwrseq_simple, pwrseq);
> int i;
>
> if (!IS_ERR(pwrseq->reset_gpios)) {
> for (i = 0; i < pwrseq->ngpios; i++)
> gpiod_set_value_cansleep(pwrseq->reset_gpios[i],
> on ? : 0);
> }
> }
>
> static void mmc_pwrseq_simple_pre_power_on(struct mmc_host *host)
> {
> __mmc_pwrseq_simple_power_on_off(host, true);
> }
>
> static void mmc_pwrseq_simple_post_power_on(struct mmc_host *host)
> {
> __mmc_pwrseq_simple_power_on_off(host, false);
> }
>
>
Sure, will do.
>> }
>>
>> static void mmc_pwrseq_simple_free(struct mmc_host *host)
>> {
>> struct mmc_pwrseq_simple *pwrseq = container_of(host->pwrseq,
>> struct mmc_pwrseq_simple, pwrseq);
>> + int i;
>>
>> - if (!IS_ERR(pwrseq->reset_gpio))
>> - gpiod_put(pwrseq->reset_gpio);
>> + if (pwrseq->nr_gpios > 0) {
>> + for (i = 0; i < pwrseq->nr_gpios; i++)
>> + if (!IS_ERR(pwrseq->reset_gpio[i]))
>> + gpiod_put(pwrseq->reset_gpio[i]);
>> + kfree(pwrseq->reset_gpio);
>> + }
>>
>> kfree(pwrseq);
>> host->pwrseq = NULL;
>> @@ -63,17 +74,27 @@ int mmc_pwrseq_simple_alloc(struct mmc_host *host, struct device *dev)
>> {
>> struct mmc_pwrseq_simple *pwrseq;
>> int ret = 0;
>> + int i;
>>
>> pwrseq = kzalloc(sizeof(struct mmc_pwrseq_simple), GFP_KERNEL);
>> if (!pwrseq)
>> return -ENOMEM;
>>
>> - pwrseq->reset_gpio = gpiod_get_index(dev, "reset", 0, GPIOD_OUT_HIGH);
>> - if (IS_ERR(pwrseq->reset_gpio) &&
>> - PTR_ERR(pwrseq->reset_gpio) != -ENOENT &&
>> - PTR_ERR(pwrseq->reset_gpio) != -ENOSYS) {
>> - ret = PTR_ERR(pwrseq->reset_gpio);
>> - goto free;
>> + pwrseq->nr_gpios = of_gpio_named_count(dev->of_node, "reset-gpios");
>> + if (pwrseq->nr_gpios > 0) {
>
> What happens if there are no gpios? This fuction should return -ENOENT
> and should not even try to allocate pwrseq?
Not quite, the DT binding states that the GPIOs are optional so it should
not fail if no GPIOs are defined.
> Probably you should do of_gpio_named_count before allocating memory.
>
I didn't do that because patch #4 "mmc: pwrseq_simple: Add optional reference
clock support" will need the struct mmc_pwrseq_simple even if no GPIOs are
defined.
A SDIO attached chip could require only an external clock or someone could
extend the pwrseq_simple driver to support an external regulator for example.
>> + pwrseq->reset_gpio = kzalloc(sizeof(struct gpio_desc *) *
>> + pwrseq->nr_gpios, GFP_KERNEL);
>> +
>> + for (i = 0; i < pwrseq->nr_gpios; i++) {
>> + pwrseq->reset_gpio[i] = gpiod_get_index(dev, "reset", i,
>> + GPIOD_OUT_HIGH);
>> + if (IS_ERR(pwrseq->reset_gpio[i]) &&
>> + PTR_ERR(pwrseq->reset_gpio[i]) != -ENOENT &&
>> + PTR_ERR(pwrseq->reset_gpio[i]) != -ENOSYS) {
>> + ret = PTR_ERR(pwrseq->reset_gpio[i]);
>
> is simple to add:
> while(--i)
> gpiod_put(pwrseq->reset_gpio[i])
>
>
>
That's true, will change.
>> + goto free;
>> + }
>> + }
>
>
>> }
>>
>> pwrseq->pwrseq.ops = &mmc_pwrseq_simple_ops;
>> @@ -81,6 +102,13 @@ int mmc_pwrseq_simple_alloc(struct mmc_host *host, struct device *dev)
>>
>> return 0;
>> free:
>> + if (pwrseq->nr_gpios > 0) {
>> + for (i = 0; i < pwrseq->nr_gpios; i++)
>> + if (!IS_ERR_OR_NULL(pwrseq->reset_gpio[i]))
>> + gpiod_put(pwrseq->reset_gpio[i]);
>> + kfree(pwrseq->reset_gpio);
>> + }
>> +
>> kfree(pwrseq);
>> return ret;
>> }
>>
>
> I get a feeling that am just dumping my patch here.. If possible could
> you have look at it too.
>
Of course, do you have a link archive since I can't find it on my inbox.
> Thanks,
> srini
>
Again, thanks a lot and best regards,
Javier
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists