linux-kernel - Re: [PATCH 5/8] clk: rockchip: Add clock type GATE_NO_SET

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 for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <9365795.CDJkKcVGEf@diego>
Date: Wed, 02 Oct 2024 10:08:36 +0200
From: Heiko Stübner <heiko@...ech.de>
To: Michael Turquette <mturquette@...libre.com>,
 Stephen Boyd <sboyd@...nel.org>, Rob Herring <robh@...nel.org>,
 Krzysztof Kozlowski <krzk+dt@...nel.org>, Conor Dooley <conor+dt@...nel.org>,
 Philipp Zabel <p.zabel@...gutronix.de>, Yao Zi <ziyao@...root.org>
Cc: linux-clk@...r.kernel.org, devicetree@...r.kernel.org,
 linux-arm-kernel@...ts.infradead.org, linux-rockchip@...ts.infradead.org,
 linux-kernel@...r.kernel.org, Celeste Liu <CoelacanthusHex@...il.com>,
 Yao Zi <ziyao@...root.org>
Subject: Re: [PATCH 5/8] clk: rockchip: Add clock type GATE_NO_SET_RATE

Hi,

Am Dienstag, 1. Oktober 2024, 06:23:59 CEST schrieb Yao Zi:
> This clock type is similar to GATE, but doesn't allow rate setting,
> which presents on RK3528 platform.

this definitly needs more explanation in the commit message.

I.e. regular individual gates always set the CLK_SET_RATE_PARENT flag
because of course the gates themselfs cannot influence the rate.


But in general, I'm also not convinced yet. Yes if some driver tries to
change the rate on those, it may affect the parent rate, but that is also
true for the other individual gates.

So what makes aclk_emmc (as GATE_NO_SET_RATE) more special than
"hclk_emmc" (as regular GATE). [Same for the other clocks of course] .


So this either needs more explanation, or for the sake of simplicity
use regular GATE for now for those and we revisit when it becomes
necessary.


Heiko


> Signed-off-by: Yao Zi <ziyao@...root.org>
> ---
>  drivers/clk/rockchip/clk.c |  8 ++++++++
>  drivers/clk/rockchip/clk.h | 14 ++++++++++++++
>  2 files changed, 22 insertions(+)
> 
> diff --git a/drivers/clk/rockchip/clk.c b/drivers/clk/rockchip/clk.c
> index 73d2cbdc716b..7d233770e68b 100644
> --- a/drivers/clk/rockchip/clk.c
> +++ b/drivers/clk/rockchip/clk.c
> @@ -521,6 +521,14 @@ void rockchip_clk_register_branches(struct rockchip_clk_provider *ctx,
>  		case branch_gate:
>  			flags |= CLK_SET_RATE_PARENT;
>  
> +			clk = clk_register_gate(NULL, list->name,
> +				list->parent_names[0], flags,
> +				ctx->reg_base + list->gate_offset,
> +				list->gate_shift, list->gate_flags, &ctx->lock);
> +			break;
> +		case branch_gate_no_set_rate:
> +			flags &= ~CLK_SET_RATE_PARENT;
> +
>  			clk = clk_register_gate(NULL, list->name,
>  				list->parent_names[0], flags,
>  				ctx->reg_base + list->gate_offset,
> diff --git a/drivers/clk/rockchip/clk.h b/drivers/clk/rockchip/clk.h
> index 1efc5c3a1e77..360d16402fe5 100644
> --- a/drivers/clk/rockchip/clk.h
> +++ b/drivers/clk/rockchip/clk.h
> @@ -519,6 +519,7 @@ enum rockchip_clk_branch_type {
>  	branch_divider,
>  	branch_fraction_divider,
>  	branch_gate,
> +	branch_gate_no_set_rate,
>  	branch_mmc,
>  	branch_inverter,
>  	branch_factor,
> @@ -844,6 +845,19 @@ struct rockchip_clk_branch {
>  		.gate_flags	= gf,				\
>  	}
>  
> +#define GATE_NO_SET_RATE(_id, cname, pname, f, o, b, gf)	\
> +	{							\
> +		.id		= _id,				\
> +		.branch_type	= branch_gate_no_set_rate,	\
> +		.name		= cname,			\
> +		.parent_names	= (const char *[]){ pname },	\
> +		.num_parents	= 1,				\
> +		.flags		= f,				\
> +		.gate_offset	= o,				\
> +		.gate_shift	= b,				\
> +		.gate_flags	= gf,				\
> +	}
> +
>  #define MMC(_id, cname, pname, offset, shift)			\
>  	{							\
>  		.id		= _id,				\
>