[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <0c19d8dedda3e308c75a6c97358d2af3.sboyd@kernel.org>
Date: Thu, 12 Jan 2023 11:18:11 -0800
From: Stephen Boyd <sboyd@...nel.org>
To: Li Chen <me@...ux.beauty>
Cc: michael turquette <mturquette@...libre.com>,
li chen <lchen@...arella.com>,
linux-clk <linux-clk@...r.kernel.org>,
linux-kernel <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] clk: create write_enable file to control clk rate write and other dangerous ops permission
Quoting Li Chen (2023-01-11 19:05:34)
> Hi Stephen,
> ---- On Thu, 12 Jan 2023 02:37:29 +0800 Stephen Boyd wrote ---
> > Quoting Li Chen (2023-01-10 03:45:39)
> > > From: Li Chen lchen@...arella.com>
> > >
> > > It's common requirement for bsp debug/test to change clk rate from userspace.
> > >
> > > Currently, we must define CLKOCK_ALLOW_WRITE_DEBUGFS then re-compile kernel
> > > to allow this feature. Let's replace it with a "write_enable" file to
> > > allow enable it at runtime.
> >
> > Nak. This design is intentional.
>
> Thanks for your reply. Got it. But sometimes re-compile is somewhat low efficient. Is it acceptable to provide
> a write_enable to clk_core(not enable by default), and allow clk driver to enable it inside clk_ops->init
> via clk_hw->core->write_enable = 1?
>
No
Powered by blists - more mailing lists