From: Qianggui Song <qianggui.song@amlogic.com>
To: Marc Zyngier <maz@kernel.org>
Cc: Hanjie Lin <hanjie.lin@amlogic.com>,
Jason Cooper <jason@lakedaemon.net>,
Jianxin Pan <jianxin.pan@amlogic.com>,
Neil Armstrong <narmstrong@baylibre.com>,
Kevin Hilman <khilman@baylibre.com>,
linux-kernel@vger.kernel.org,
linux-arm-kernel@lists.infradead.org,
linux-amlogic@lists.infradead.org,
Thomas Gleixner <tglx@linutronix.de>,
Xingyu Chen <xingyu.chen@amlogic.com>,
Jerome Brunet <jbrunet@baylibre.com>
Subject: Re: [PATCH 2/4] irqchip/meson-gpio: rework meson irqchip driver to support meson-A1 SoCs
Date: Thu, 12 Dec 2019 19:13:16 +0800 [thread overview]
Message-ID: <4b892b12-4ffb-7fff-ba27-9e606c958257@amlogic.com> (raw)
In-Reply-To: <0cbbb895b50a838fd1dfa9e59528367d@www.loen.fr>
On 2019/12/12 1:26, Marc Zyngier wrote:
> On 2019-12-10 02:08, Qianggui Song wrote:
>> Hi, Marc
>> Thank you for your review
>>
>> On 2019/12/6 21:13, Marc Zyngier wrote:
>>> On 2019-12-06 12:17, Qianggui Song wrote:
>>>> Since Meson-A1 Socs register layout of gpio interrupt controller
>>>> have
>>>> difference with previous chips, registers to decide irq line and
>>>> offset
>>>> of trigger method are all changed, the current driver should be
>>>> modified.
>>>>
>>>> Signed-off-by: Qianggui Song <qianggui.song@amlogic.com>
>>>> ---
>>>> drivers/irqchip/irq-meson-gpio.c | 79
>>>> ++++++++++++++++++++++++--------
>>>> 1 file changed, 60 insertions(+), 19 deletions(-)
>>>>
>>>> diff --git a/drivers/irqchip/irq-meson-gpio.c
>>>> b/drivers/irqchip/irq-meson-gpio.c
>>>> index 829084b568fa..1824ffc30de2 100644
>>>> --- a/drivers/irqchip/irq-meson-gpio.c
>>>> +++ b/drivers/irqchip/irq-meson-gpio.c
>>>> @@ -30,44 +30,74 @@
>>>> * stuck at 0. Bits 8 to 15 are responsive and have the expected
>>>> * effect.
>>>> */
>>>> -#define REG_EDGE_POL_EDGE(x) BIT(x)
>>>> -#define REG_EDGE_POL_LOW(x) BIT(16 + (x))
>>>> -#define REG_BOTH_EDGE(x) BIT(8 + (x))
>>>> -#define REG_EDGE_POL_MASK(x) ( \
>>>> - REG_EDGE_POL_EDGE(x) | \
>>>> - REG_EDGE_POL_LOW(x) | \
>>>> - REG_BOTH_EDGE(x))
>>>> +#define REG_EDGE_POL_EDGE(params,
>>>> x) BIT((params)->edge_single_offset + (x))
>>>> +#define REG_EDGE_POL_LOW(params, x) BIT((params)->pol_low_offset +
>>>> (x))
>>>> +#define REG_BOTH_EDGE(params, x) BIT((params)->edge_both_offset +
>>>> (x))
>>>> +#define REG_EDGE_POL_MASK(params, x) ( \
>>>> + REG_EDGE_POL_EDGE(params, x) | \
>>>> + REG_EDGE_POL_LOW(params, x) | \
>>>> + REG_BOTH_EDGE(params, x))
>>>> #define REG_PIN_SEL_SHIFT(x) (((x) % 4) * 8)
>>>> #define REG_FILTER_SEL_SHIFT(x) ((x) * 4)
>>>>
>>>> +#define INIT_MESON8_COMMON_DATA \
>>>> + .edge_single_offset = 0, \
>>>> + .pol_low_offset = 16, \
>>>> + .pin_sel_mask = 0xff, \
>>>> + .ops = { \
>>>> + .gpio_irq_sel_pin = meson8_gpio_irq_sel_pin, \
>>>> + },
>>>
>>> Please place the #defines that operate on the various data
>>> structures
>>> *after* the definition of the structures. It would greatly help
>>> reading the changes.
>>>
>> OK, will place it below the definition of struct
>> meson_gpio_irq_params
>> in the next patch.
>>>> +
>>>> +struct meson_gpio_irq_controller;
>>>> +static void meson8_gpio_irq_sel_pin(struct
>>>> meson_gpio_irq_controller
>>>> *ctl,
>>>> + unsigned int channel, unsigned long hwirq);
>>>> +struct irq_ctl_ops {
>>>> + void (*gpio_irq_sel_pin)(struct meson_gpio_irq_controller *ctl,
>>>> + unsigned int channel,
>>>> + unsigned long hwirq);
>>>> + void (*gpio_irq_init)(struct meson_gpio_irq_controller *ctl);
>>>> +};
>>>> +
>>>> struct meson_gpio_irq_params {
>>>> unsigned int nr_hwirq;
>>>> bool support_edge_both;
>>>> + unsigned int edge_both_offset;
>>>> + unsigned int edge_single_offset;
>>>> + unsigned int pol_low_offset;
>>>> + unsigned int pin_sel_mask;
>>>> + struct irq_ctl_ops ops;
>>>> };
>>>>
>>>> static const struct meson_gpio_irq_params meson8_params = {
>>>> .nr_hwirq = 134,
>>>> + INIT_MESON8_COMMON_DATA
>>>> };
>>>>
>>>> static const struct meson_gpio_irq_params meson8b_params = {
>>>> .nr_hwirq = 119,
>>>> + INIT_MESON8_COMMON_DATA
>>>> };
>>>>
>>>> static const struct meson_gpio_irq_params gxbb_params = {
>>>> .nr_hwirq = 133,
>>>> + INIT_MESON8_COMMON_DATA
>>>> };
>>>>
>>>> static const struct meson_gpio_irq_params gxl_params = {
>>>> .nr_hwirq = 110,
>>>> + INIT_MESON8_COMMON_DATA
>>>> };
>>>>
>>>> static const struct meson_gpio_irq_params axg_params = {
>>>> .nr_hwirq = 100,
>>>> + INIT_MESON8_COMMON_DATA
>>>> };
>>>>
>>>> static const struct meson_gpio_irq_params sm1_params = {
>>>> .nr_hwirq = 100,
>>>> .support_edge_both = true,
>>>> + .edge_both_offset = 8,
>>>> + INIT_MESON8_COMMON_DATA
>>>> };
>>>
>>> OK, this isn't great. The least you could do is to make
>>> your initializer parametric, so that it takes the nr_hwirq as
>>> a parameter.
>>>
>>> Then, any additional member that overrides common behaviour
>>> should come after the main initializer.
>>>
>>> Also, do you need 'support_edge_both'? Isn't a non-zero
>>> 'edge_both_offset' enough to detect the feature?
>>>
>>
>> Sorry, but I am not very clear that "make your initializer
>> parametric,
>> so that it takes the nr_hwirq as a parameter". Is that
>> initializer(initial function in .ops ? ) as a parameter of struct
>> meson_gpio_irq_params ? If nr_hwirq as a parameter of init function
>> of
>> .ops then will make lot of init function for each platform.
>>
>> How about move .ops from macro like below:
>> #define INIT_MESON8_COMMON_DATA \
>> .edge_single_offset = 0, \
>> .pol_low_offset = 16, \
>> .pin_sel_mask = 0xff,
>>
>> static const struct meson_gpio_irq_params sm1_params = {
>> .nr_hwirq = 100,//main initializer
>> .ops = {
>> .gpio_irq_sel_pin = meson8_gpio_irq_sel_pin,
>> /*in below to assign support_edge_both
>> * edge_both_offset
>> * call after main initializer to additional
>> * member
>> */
>> .gpio_irq_init = meson_sm1_irq_init,
>> },
>> INIT_MESON8_COMMON_DATA// m8 to sm1 are the same.
>> };
>
> No, what I'm suggesting is something like this:
>
> diff --git a/drivers/irqchip/irq-meson-gpio.c
> b/drivers/irqchip/irq-meson-gpio.c
> index 8478100706a6..27a3207a944d 100644
> --- a/drivers/irqchip/irq-meson-gpio.c
> +++ b/drivers/irqchip/irq-meson-gpio.c
> @@ -43,24 +43,27 @@
> /* Below is used for Meson-A1 series like chips*/
> #define REG_PIN_A1_SEL 0x04
>
> -#define INIT_MESON8_COMMON_DATA \
> - .edge_single_offset = 0, \
> - .pol_low_offset = 16, \
> - .pin_sel_mask = 0xff, \
> - .ops = { \
> - .gpio_irq_sel_pin = meson8_gpio_irq_sel_pin, \
> - },
> -
> -#define INIT_MESON_A1_COMMON_DATA \
> - .support_edge_both = true, \
> - .edge_both_offset = 16, \
> - .edge_single_offset = 8, \
> - .pol_low_offset = 0, \
> - .pin_sel_mask = 0x7f, \
> - .ops = { \
> - .gpio_irq_sel_pin = meson_a1_gpio_irq_sel_pin, \
> - .gpio_irq_init = meson_a1_gpio_irq_init, \
> - },
> +#define INIT_MESON_COMMON(irqs, init, sel) \
> + .nr_hwirq = irqs, \
> + .ops = { \
> + .gpio_irq_sel_pin = sel, \
> + .gpio_irq_init = init, \
> + }
> +
> +#define INIT_MESON8_COMMON_DATA(irqs) \
> + INIT_MESON_COMMON(irqs, NULL, \
> + meson8_gpio_irq_sel_pin), \
> + .pol_low_offset = 16, \
> + .pin_sel_mask = 0xff,
> +
> +#define INIT_MESON_A1_COMMON_DATA(irqs) \
> + INIT_MESON_COMMON(irqs, meson_a1_gpio_irq_init, \
> + meson_a1_gpio_irq_sel_pin), \
> + .support_edge_both = true, \
> + .edge_both_offset = 16, \
> + .edge_single_offset = 8, \
> + .pol_low_offset = 0, \
> + .pin_sel_mask = 0x7f,
>
> struct meson_gpio_irq_controller;
> static void meson8_gpio_irq_sel_pin(struct meson_gpio_irq_controller
> *ctl,
> @@ -89,40 +92,33 @@ struct meson_gpio_irq_params {
> };
>
> static const struct meson_gpio_irq_params meson8_params = {
> - .nr_hwirq = 134,
> - INIT_MESON8_COMMON_DATA
> + INIT_MESON8_COMMON_DATA(134),
> };
>
> static const struct meson_gpio_irq_params meson8b_params = {
> - .nr_hwirq = 119,
> - INIT_MESON8_COMMON_DATA
> + INIT_MESON8_COMMON_DATA(119),
> };
>
> static const struct meson_gpio_irq_params gxbb_params = {
> - .nr_hwirq = 133,
> - INIT_MESON8_COMMON_DATA
> + INIT_MESON8_COMMON_DATA(133),
> };
>
> static const struct meson_gpio_irq_params gxl_params = {
> - .nr_hwirq = 110,
> - INIT_MESON8_COMMON_DATA
> + INIT_MESON8_COMMON_DATA(110),
> };
>
> static const struct meson_gpio_irq_params axg_params = {
> - .nr_hwirq = 100,
> - INIT_MESON8_COMMON_DATA
> + INIT_MESON8_COMMON_DATA(100),
> };
>
> static const struct meson_gpio_irq_params sm1_params = {
> - .nr_hwirq = 100,
> + INIT_MESON8_COMMON_DATA(100),
> .support_edge_both = true,
> .edge_both_offset = 8,
> - INIT_MESON8_COMMON_DATA
> };
>
> static const struct meson_gpio_irq_params a1_params = {
> - .nr_hwirq = 62,
> - INIT_MESON_A1_COMMON_DATA
> + INIT_MESON_A1_COMMON_DATA(62),
> };
>
Thanks, will try it in later patch
> static const struct of_device_id meson_irq_gpio_matches[] = {
>
>
> Thanks,
>
> M.
>
_______________________________________________
linux-amlogic mailing list
linux-amlogic@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-amlogic
next prev parent reply other threads:[~2019-12-12 11:13 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-12-06 12:17 [PATCH 0/4] irqchip/meson-gpio: Add support for Meson-A1 SoC Qianggui Song
2019-12-06 12:17 ` [PATCH 1/4] dt-bindings: interrupt-controller: New binding for Meson-A1 SoCs Qianggui Song
2019-12-06 12:17 ` [PATCH 2/4] irqchip/meson-gpio: rework meson irqchip driver to support meson-A1 SoCs Qianggui Song
2019-12-06 13:13 ` Marc Zyngier
2019-12-10 2:08 ` Qianggui Song
2019-12-11 17:26 ` Marc Zyngier
2019-12-12 11:13 ` Qianggui Song [this message]
2019-12-06 12:17 ` [PATCH 3/4] irqchip/meson-gpio: Add support for meson a1 SoCs Qianggui Song
2019-12-06 12:17 ` [PATCH 4/4] arm64: dts: meson: a1: add gpio interrupt controller support Qianggui Song
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4b892b12-4ffb-7fff-ba27-9e606c958257@amlogic.com \
--to=qianggui.song@amlogic.com \
--cc=hanjie.lin@amlogic.com \
--cc=jason@lakedaemon.net \
--cc=jbrunet@baylibre.com \
--cc=jianxin.pan@amlogic.com \
--cc=khilman@baylibre.com \
--cc=linux-amlogic@lists.infradead.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=maz@kernel.org \
--cc=narmstrong@baylibre.com \
--cc=tglx@linutronix.de \
--cc=xingyu.chen@amlogic.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®