From: Masahiro Yamada <yamada.masahiro@socionext.com>
To: Dinh Nguyen <dinguyen@kernel.org>
Cc: Mark Rutland <mark.rutland@arm.com>,
"devicetree@vger.kernel.org" <devicetree@vger.kernel.org>,
Marek Vasut <marex@denx.de>, Dinh Nguyen <dinh.linux@gmail.com>,
Richard Weinberger <richard@nod.at>,
Rob Herring <robh+dt@kernel.org>,
"linux-arm-kernel@lists.infradead.org"
<linux-arm-kernel@lists.infradead.org>
Subject: Re: [PATCH 1/2] ARM: dts: arria10: update NAND clocking
Date: Mon, 23 Jul 2018 10:48:33 +0900 [thread overview]
Message-ID: <CAK7LNAQfcnugbUxJpMyJC5FT0ejtk0iA4+vXZ=UmX5rYRJs2tA@mail.gmail.com> (raw)
In-Reply-To: <5a06f316-8251-08b9-491a-912425cf204a@kernel.org>
2018-07-21 8:16 GMT+09:00 Dinh Nguyen <dinguyen@kernel.org>:
>
>
> On 07/18/2018 09:22 PM, Masahiro Yamada wrote:
>> 2018-07-11 23:51 GMT+09:00 Dinh Nguyen <dinh.linux@gmail.com>:
>>> Hi Masahiro,
>>>
>>> On Tue, Jul 10, 2018 at 9:45 PM Masahiro Yamada
>>> <yamada.masahiro@socionext.com> wrote:
>>>>
>>>> 2018-07-10 22:35 GMT+09:00 Dinh Nguyen <dinguyen@kernel.org>:
>>>>>
>>>>>
>>>>> On 07/09/2018 08:31 PM, Masahiro Yamada wrote:
>>>>>> 2018-07-10 7:41 GMT+09:00 Dinh Nguyen <dinguyen@kernel.org>:
>>>>>>> The NAND IP needs 2 clocks(nand_x_clk and nand_clk). This patch adds a
>>>>>>> nand_x_clk, which is derived from the nand_clk, but has a fixed divider
>>>>>>> of 4.
>>>>>>>
>>>>>>> Update the NAND node to use the additional clock.
>>>>>>>
>>>>>>> Signed-off-by: Dinh Nguyen <dinguyen@kernel.org>
>>>>>>> ---
>>>>>>> arch/arm/boot/dts/socfpga_arria10.dtsi | 13 +++++++++++--
>>>>>>> 1 file changed, 11 insertions(+), 2 deletions(-)
>>>>>>>
>>>>>>> diff --git a/arch/arm/boot/dts/socfpga_arria10.dtsi b/arch/arm/boot/dts/socfpga_arria10.dtsi
>>>>>>> index a4dcb68..558b5ea 100644
>>>>>>> --- a/arch/arm/boot/dts/socfpga_arria10.dtsi
>>>>>>> +++ b/arch/arm/boot/dts/socfpga_arria10.dtsi
>>>>>>> @@ -377,13 +377,21 @@
>>>>>>> clk-gate = <0xC8 11>;
>>>>>>> };
>>>>>>>
>>>>>>> - nand_clk: nand_clk {
>>>>>>> + nand_x_clk: nand_x_clk {
>>>>>>> #clock-cells = <0>;
>>>>>>> compatible = "altr,socfpga-a10-gate-clk";
>>>>>>> clocks = <&l4_mp_clk>;
>>>>>>> clk-gate = <0xC8 10>;
>>>>>>> };
>>>>>>>
>>>>>>> + nand_clk: nand_clk {
>>>>>>> + #clock-cells = <0>;
>>>>>>> + compatible = "altr,socfpga-a10-gate-clk";
>>>>>>> + clocks = <&nand_x_clk>;
>>>>>>> + fixed-divider = <4>;
>>>>>>> + clk-gate = <0xC8 10>;
>>>>>>> + };
>>>>>>> +
>>>>>>> spi_m_clk: spi_m_clk {
>>>>>>> #clock-cells = <0>;
>>>>>>> compatible = "altr,socfpga-a10-gate-clk";
>>>>>>> @@ -650,7 +658,8 @@
>>>>>>> reg-names = "nand_data", "denali_reg";
>>>>>>> interrupts = <0 99 4>;
>>>>>>> dma-mask = <0xffffffff>;
>>>>>>> - clocks = <&nand_clk>;
>>>>>>> + clocks = <&nand_clk>, <&nand_x_clk>;
>>>>>>> + clock-names = "nand", "nand_x";
>>>>>>
>>>>>>
>>>>>> Doesn't SOCFPGA really need the ECC clock?
>>>>>>
>>>>>
>>>>> I don't see any mentions of an ECC clock in any of the SoCFPGA
>>>>> documentation for the NAND IP.
>>>>
>>>>
>>>> I'd like you to check the IP documentation
>>>> published by Denali (or Cadence),
>>>> not published by Altera.
>>>>
>>>> Is it available for you?
>>>>
>>>>
>>>>
>>>> In my SoCs, the same clock signal (nand-clk-200m)
>>>> is connected to clk_x and ecc_clk, like this:
>>>>
>>>> -------------
>>>> -------------- | DENALI IP |
>>>> | | | |
>>>> | clk |---(nand-clk-50m)--------->| clk |
>>>> | controller | | |
>>>> | |---(nand-clk-200m)-------->| clk_x |
>>>> -------------- | | |
>>>> --->| ecc_clk |
>>>> | |
>>>> -------------
>
> Here's a diagram of the clocking for the NAND IP on SOCFPGA.
>
> -------------
> -------------- | DENALI IP |
> | | | |
> | clk | | |-->clk(/4) |
> | controller | | | |
> | |---(nand_clk)-------------|-->clk_x |
> ------------- | | |
> | |-->ecc_clk |
> | |
> ---------------
>
> From the clock controller, it provides a single 200MHz clock to the NAND
> IP. Inside the NAND IP, there is a /4 for the clk. The 200MHz clock is
> used for the clk_x and ecc_clk.
>
> The nand-clk is the only clock that can be gated.
The Denali IP does not contain the divisor in it.
My best guess is Altera implemented a simple
wrapper with the 1/4 divisor, like this?
----------------------------
| Wrapper |
| ------------- |
| | DENALI IP | |
| | | |
| |-(1/4)-->| clk | |
| | | | |
---(nand_clk)--|--|-------->| clk_x | |
| | | | |
| |-------->| ecc_clk | |
| | | |
| ------------- |
----------------------------
>>>>
>>>>
>>>> Some clock lines are often tied-up,
>>>> but such low-level information does not appear in
>>>> user documents.
>>>> I talked to a hardware engineer to get this information.
>>>>
>>>>
>>>>
>>>> I guess the situation is similar for your case,
>>>> but I do not have information enough to argue this.
>>>>
>>>
>>> I don't disagree with this representation. But after asking around,
>>> I'm getting the impression that the ECC clock is not exposed to the
>>> clock manager. This implies that the clock is always turned on and
>>> there doesn't seem to be any software control for it. I tested ECC
>>> with this patch and it looks to be working correctly, without any ecc
>>> clock representation.
>>>
>>> I'm not sure if I should expose a dummy ecc_clk to satisfy the binding?
>>
>>
>> DT should represent the real HW.
>> If the Denali IP in SOCFPGA needs ecc_clk input,
>> DT should describe it.
>>
>>
>> I do not have knowledge about the SOCFPGA internals.
>> I do not know which is correct:
>>
>> - The Denali IP in SOCFPGA needs ecc_clk,
>> but there is no software control for it.
>>
>> - The Denali IP in SOCFPGA does not need ecc_clk.
>>
>
> So should I be representing the clocks even though it doesn't come from
> the clock controller?
Anyway, you know the same clock is connected to nand_x to ecc_clk.
So, there is no problem for adding "ecc" as well, right?
--
Best Regards
Masahiro Yamada
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
WARNING: multiple messages have this Message-ID (diff)
From: yamada.masahiro@socionext.com (Masahiro Yamada)
To: linux-arm-kernel@lists.infradead.org
Subject: [PATCH 1/2] ARM: dts: arria10: update NAND clocking
Date: Mon, 23 Jul 2018 10:48:33 +0900 [thread overview]
Message-ID: <CAK7LNAQfcnugbUxJpMyJC5FT0ejtk0iA4+vXZ=UmX5rYRJs2tA@mail.gmail.com> (raw)
In-Reply-To: <5a06f316-8251-08b9-491a-912425cf204a@kernel.org>
2018-07-21 8:16 GMT+09:00 Dinh Nguyen <dinguyen@kernel.org>:
>
>
> On 07/18/2018 09:22 PM, Masahiro Yamada wrote:
>> 2018-07-11 23:51 GMT+09:00 Dinh Nguyen <dinh.linux@gmail.com>:
>>> Hi Masahiro,
>>>
>>> On Tue, Jul 10, 2018 at 9:45 PM Masahiro Yamada
>>> <yamada.masahiro@socionext.com> wrote:
>>>>
>>>> 2018-07-10 22:35 GMT+09:00 Dinh Nguyen <dinguyen@kernel.org>:
>>>>>
>>>>>
>>>>> On 07/09/2018 08:31 PM, Masahiro Yamada wrote:
>>>>>> 2018-07-10 7:41 GMT+09:00 Dinh Nguyen <dinguyen@kernel.org>:
>>>>>>> The NAND IP needs 2 clocks(nand_x_clk and nand_clk). This patch adds a
>>>>>>> nand_x_clk, which is derived from the nand_clk, but has a fixed divider
>>>>>>> of 4.
>>>>>>>
>>>>>>> Update the NAND node to use the additional clock.
>>>>>>>
>>>>>>> Signed-off-by: Dinh Nguyen <dinguyen@kernel.org>
>>>>>>> ---
>>>>>>> arch/arm/boot/dts/socfpga_arria10.dtsi | 13 +++++++++++--
>>>>>>> 1 file changed, 11 insertions(+), 2 deletions(-)
>>>>>>>
>>>>>>> diff --git a/arch/arm/boot/dts/socfpga_arria10.dtsi b/arch/arm/boot/dts/socfpga_arria10.dtsi
>>>>>>> index a4dcb68..558b5ea 100644
>>>>>>> --- a/arch/arm/boot/dts/socfpga_arria10.dtsi
>>>>>>> +++ b/arch/arm/boot/dts/socfpga_arria10.dtsi
>>>>>>> @@ -377,13 +377,21 @@
>>>>>>> clk-gate = <0xC8 11>;
>>>>>>> };
>>>>>>>
>>>>>>> - nand_clk: nand_clk {
>>>>>>> + nand_x_clk: nand_x_clk {
>>>>>>> #clock-cells = <0>;
>>>>>>> compatible = "altr,socfpga-a10-gate-clk";
>>>>>>> clocks = <&l4_mp_clk>;
>>>>>>> clk-gate = <0xC8 10>;
>>>>>>> };
>>>>>>>
>>>>>>> + nand_clk: nand_clk {
>>>>>>> + #clock-cells = <0>;
>>>>>>> + compatible = "altr,socfpga-a10-gate-clk";
>>>>>>> + clocks = <&nand_x_clk>;
>>>>>>> + fixed-divider = <4>;
>>>>>>> + clk-gate = <0xC8 10>;
>>>>>>> + };
>>>>>>> +
>>>>>>> spi_m_clk: spi_m_clk {
>>>>>>> #clock-cells = <0>;
>>>>>>> compatible = "altr,socfpga-a10-gate-clk";
>>>>>>> @@ -650,7 +658,8 @@
>>>>>>> reg-names = "nand_data", "denali_reg";
>>>>>>> interrupts = <0 99 4>;
>>>>>>> dma-mask = <0xffffffff>;
>>>>>>> - clocks = <&nand_clk>;
>>>>>>> + clocks = <&nand_clk>, <&nand_x_clk>;
>>>>>>> + clock-names = "nand", "nand_x";
>>>>>>
>>>>>>
>>>>>> Doesn't SOCFPGA really need the ECC clock?
>>>>>>
>>>>>
>>>>> I don't see any mentions of an ECC clock in any of the SoCFPGA
>>>>> documentation for the NAND IP.
>>>>
>>>>
>>>> I'd like you to check the IP documentation
>>>> published by Denali (or Cadence),
>>>> not published by Altera.
>>>>
>>>> Is it available for you?
>>>>
>>>>
>>>>
>>>> In my SoCs, the same clock signal (nand-clk-200m)
>>>> is connected to clk_x and ecc_clk, like this:
>>>>
>>>> -------------
>>>> -------------- | DENALI IP |
>>>> | | | |
>>>> | clk |---(nand-clk-50m)--------->| clk |
>>>> | controller | | |
>>>> | |---(nand-clk-200m)-------->| clk_x |
>>>> --------------? | | |
>>>> --->| ecc_clk |
>>>> | |
>>>> -------------
>
> Here's a diagram of the clocking for the NAND IP on SOCFPGA.
>
> -------------
> -------------- | DENALI IP |
> | | | |
> | clk | | |-->clk(/4) |
> | controller | | | |
> | |---(nand_clk)-------------|-->clk_x |
> ------------- | | |
> | |-->ecc_clk |
> | |
> ---------------
>
> From the clock controller, it provides a single 200MHz clock to the NAND
> IP. Inside the NAND IP, there is a /4 for the clk. The 200MHz clock is
> used for the clk_x and ecc_clk.
>
> The nand-clk is the only clock that can be gated.
The Denali IP does not contain the divisor in it.
My best guess is Altera implemented a simple
wrapper with the 1/4 divisor, like this?
----------------------------
| Wrapper |
| ------------- |
| | DENALI IP | |
| | | |
| |-(1/4)-->| clk | |
| | | | |
---(nand_clk)--|--|-------->| clk_x | |
| | | | |
| |-------->| ecc_clk | |
| | | |
| ------------- |
----------------------------
>>>>
>>>>
>>>> Some clock lines are often tied-up,
>>>> but such low-level information does not appear in
>>>> user documents.
>>>> I talked to a hardware engineer to get this information.
>>>>
>>>>
>>>>
>>>> I guess the situation is similar for your case,
>>>> but I do not have information enough to argue this.
>>>>
>>>
>>> I don't disagree with this representation. But after asking around,
>>> I'm getting the impression that the ECC clock is not exposed to the
>>> clock manager. This implies that the clock is always turned on and
>>> there doesn't seem to be any software control for it. I tested ECC
>>> with this patch and it looks to be working correctly, without any ecc
>>> clock representation.
>>>
>>> I'm not sure if I should expose a dummy ecc_clk to satisfy the binding?
>>
>>
>> DT should represent the real HW.
>> If the Denali IP in SOCFPGA needs ecc_clk input,
>> DT should describe it.
>>
>>
>> I do not have knowledge about the SOCFPGA internals.
>> I do not know which is correct:
>>
>> - The Denali IP in SOCFPGA needs ecc_clk,
>> but there is no software control for it.
>>
>> - The Denali IP in SOCFPGA does not need ecc_clk.
>>
>
> So should I be representing the clocks even though it doesn't come from
> the clock controller?
Anyway, you know the same clock is connected to nand_x to ecc_clk.
So, there is no problem for adding "ecc" as well, right?
--
Best Regards
Masahiro Yamada
next prev parent reply other threads:[~2018-07-23 1:48 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-07-09 22:41 [PATCH 1/2] ARM: dts: arria10: update NAND clocking Dinh Nguyen
2018-07-09 22:41 ` Dinh Nguyen
2018-07-09 22:41 ` [PATCH 2/2] ARM: dts: socfpga: update NAND clocking for c5/a5 Dinh Nguyen
2018-07-09 22:41 ` Dinh Nguyen
2018-07-10 1:31 ` [PATCH 1/2] ARM: dts: arria10: update NAND clocking Masahiro Yamada
2018-07-10 1:31 ` Masahiro Yamada
2018-07-10 13:35 ` Dinh Nguyen
2018-07-10 13:35 ` Dinh Nguyen
2018-07-11 2:43 ` Masahiro Yamada
2018-07-11 2:43 ` Masahiro Yamada
2018-07-11 14:51 ` Dinh Nguyen
2018-07-11 14:51 ` Dinh Nguyen
2018-07-19 2:22 ` Masahiro Yamada
2018-07-19 2:22 ` Masahiro Yamada
2018-07-20 23:16 ` Dinh Nguyen
2018-07-20 23:16 ` Dinh Nguyen
2018-07-23 1:48 ` Masahiro Yamada [this message]
2018-07-23 1:48 ` Masahiro Yamada
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='CAK7LNAQfcnugbUxJpMyJC5FT0ejtk0iA4+vXZ=UmX5rYRJs2tA@mail.gmail.com' \
--to=yamada.masahiro@socionext.com \
--cc=devicetree@vger.kernel.org \
--cc=dinguyen@kernel.org \
--cc=dinh.linux@gmail.com \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=marex@denx.de \
--cc=mark.rutland@arm.com \
--cc=richard@nod.at \
--cc=robh+dt@kernel.org \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.