Re: [PATCH v3 07/11] dt-bindings: clock: Add StarFive JH7110 system clock and reset generator

From: Hal Feng <hal.feng@starfivetech.com>
To: Conor Dooley <conor@kernel.org>, Stephen Boyd <sboyd@kernel.org>
Cc: <linux-riscv@lists.infradead.org>, <devicetree@vger.kernel.org>,
	<linux-clk@vger.kernel.org>, Palmer Dabbelt <palmer@dabbelt.com>,
	Rob Herring <robh+dt@kernel.org>,
	Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org>,
	Michael Turquette <mturquette@baylibre.com>,
	Philipp Zabel <p.zabel@pengutronix.de>,
	"Emil Renner Berthing" <emil.renner.berthing@canonical.com>,
	<linux-kernel@vger.kernel.org>
Subject: Re: [PATCH v3 07/11] dt-bindings: clock: Add StarFive JH7110 system clock and reset generator
Date: Wed, 22 Feb 2023 21:27:37 +0800	[thread overview]
Message-ID: <c0472d7f-56fe-3e91-e0a0-49ee51700b5d@starfivetech.com> (raw)
In-Reply-To: <Y/VWNPfApsfm3/UD@spud>

On Tue, 21 Feb 2023 23:39:32 +0000, Conor Dooley wrote:
> On Tue, Feb 21, 2023 at 02:17:17PM -0800, Stephen Boyd wrote:
>> Quoting Conor Dooley (2023-02-16 10:20:34)
>> > On Thu, Feb 16, 2023 at 10:42:20PM +0800, Hal Feng wrote:
>> > > On Tue, 27 Dec 2022 20:15:20 +0000, Conor Dooley wrote:
>> > > > On Mon, Dec 26, 2022 at 12:26:32AM +0800, Hal Feng wrote:
>> > > Please see the picture of these external clocks in clock tree.
>> > > 
>> > > # mount -t debugfs none /mnt
>> > > # cat /mnt/clk/clk_summary
>> > >                                  enable  prepare  protect                                duty  hardware
>> > >    clock                          count    count    count        rate   accuracy phase  cycle    enable
>> > > -------------------------------------------------------------------------------------------------------
>> > >  *mclk_ext*                             0        0        0    12288000          0     0  50000         Y
>> > >  *tdm_ext*                              0        0        0    49152000          0     0  50000         Y
>> > >  *i2srx_lrck_ext*                       0        0        0      192000          0     0  50000         Y
>> > >  *i2srx_bclk_ext*                       0        0        0    12288000          0     0  50000         Y
>> > >  *i2stx_lrck_ext*                       0        0        0      192000          0     0  50000         Y
>> > >  *i2stx_bclk_ext*                       0        0        0    12288000          0     0  50000         Y
>> > >  *gmac1_rgmii_rxin*                     0        0        0   125000000          0     0  50000         Y
>> > >     gmac1_rx                          0        0        0   125000000          0     0  50000         Y
>> > >        gmac1_rx_inv                   0        0        0   125000000          0   180  50000         Y
>> > >  *gmac1_rmii_refin*                     0        0        0    50000000          0     0  50000         Y
>> > >     gmac1_rmii_rtx                    0        0        0    50000000          0     0  50000         Y
>> > >        gmac1_tx                       0        0        0    50000000          0     0  50000         N
>> > >           gmac1_tx_inv                0        0        0    50000000          0   180  50000         Y
>> > >  *osc*                                  4        4        0    24000000          0     0  50000         Y
>> > >     apb_func                          0        0        0    24000000          0     0  50000         Y
>> > >  ...
>> > > 
>> > > The clock "gmac1_rgmii_rxin" and the clock "gmac1_rmii_refin" are
>> > > actually used as the parent of other clocks.
>> > 
>> > > The "dummy" clocks
>> > > you said are all internal clocks.
>> > 
>> > No, what I meant by "dummy" clocks is that if you make clocks "required"
>> > in the binding that are not needed by the hardware for operation a
>> > customer of yours might have to add "dummy" clocks to their devicetree
>> > to pass dtbs_check.
>> 
>> They can set the phandle specifier to '<0>' to fill in the required
>> property when there isn't anything there. If this is inside an SoC, it
>> is always connected because silicon can't change after it is made
>> (unless this is an FPGA). Therefore, any and all input clocks should be
>> listed as required.
> 
>> If the clk controller has inputs that are
>> pads/balls/pins on the SoC then they can be optional if a valid design
>> can leave those pins not connected.
> 
> From the discussion on the dts patches, where the clocks have been put
> (intentionally) into board.dts, I've been under the impression that we
> are in this situation.

For the system (sys) clock controller, we are in this situation.
For the always-on (aon) clock controller, we are not, because some input
clocks are inside the SoC.

> Up to Hal to tell us if the hardware is capable of having those inputs
> left unfilled!

The situation is different for v1.2A and v1.3B boards.

For the v1.2A board,
gmac1 only requires "gmac1_rmii_refin", which support 100MHz
gmac0 only requires "gmac0_rgmii_rxin", which support 1000MHz

For the v1.3B board,
gmac1 only requires "gmac1_rgmii_rxin", which support 1000MHz
gmac0 only requires "gmac0_rgmii_rxin", which support 1000MHz

So we should set the "required" property depending on different
boards.

Best regards,
Hal

> 
> FWIW, there's a v4 [1] of this series - but the question has yet to be
> resolved.
> 
> Thanks,
> Conor.
> 
> 1 - https://lore.kernel.org/all/20230221024645.127922-1-hal.feng@starfivetech.com/