linux-kernel - Re: [PATCH 2/2] clk: Add Fixed MMIO clock driver

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: Windows password security audit tool. GUI, reports in PDF.

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

Message-ID: <154356831048.88331.9290972200429107619@swboyd.mtv.corp.google.com>
Date:   Fri, 30 Nov 2018 00:58:30 -0800
From:   Stephen Boyd <sboyd@...nel.org>
To:     Janek Kotas <jank@...ence.com>
Cc:     "mark.rutland@....com" <mark.rutland@....com>,
        "mturquette@...libre.com" <mturquette@...libre.com>,
        "robh+dt@...nel.org" <robh+dt@...nel.org>,
        "linux-clk@...r.kernel.org" <linux-clk@...r.kernel.org>,
        "devicetree@...r.kernel.org" <devicetree@...r.kernel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 2/2] clk: Add Fixed MMIO clock driver

Quoting Janek Kotas (2018-11-30 00:52:16)
> 
>         We have a single register in a fixed location which contains 
>         the frequency of the main system clock.
>         This allows us to use the same image in emulation, FPGA
>         and simulation without any changes.
>         We usually don’t use a full bootloader, just a simple wrapper,
>         which initializes the necessary stuff and jumps to the kernel.
> 
> 
>     So the hardware team has decided to throw a frequency register in there?
>     Alright! Does that fixed rate clk generate its fixed rate from some
>     other clk? I’m curious if this fixed rate clk has a parent source.
> 
> 
> It depends, in simulation and emulation we can just generate 
> a clock source with a given frequency.
> On FPGA it’s usually an external oscillator, sometimes a fixed PLL.
> The driver is not intended to be used in full, complex SoCs.

Ok, sounds like it isn't worth modeling this then.

> 
> 
>     It would also be good to make sure that any clks registered from this
>     driver can't be populated from regions of memory like DDR. Can you
>     confirm? I think it will fail, but it would be worth checking
> 

Please try this.

>     Yes I'd prefer a platform driver unless there's some reason it can't be
>     done. We may need to have both in case this needs to be populated very
>     early, but if that isn’t the case then just a platform driver for now.
> 
> 
> I played with it a bit, and it and it looks like I need both.
> We use this clock source with some of our components, 
> and for some of them the platform driver was initialized too late.
> 

Ok. So do both then.