[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20200911124258.GB3390477@lunn.ch>
Date: Fri, 11 Sep 2020 14:42:58 +0200
From: Andrew Lunn <andrew@...n.ch>
To: Matthias Schiffer <matthias.schiffer@...tq-group.com>
Cc: Marek Behún <marek.behun@....cz>,
Pavel Machek <pavel@....cz>, netdev@...r.kernel.org,
linux-leds@...r.kernel.org, Dan Murphy <dmurphy@...com>,
Ondřej Jirman <megous@...ous.com>,
Russell King <linux@...linux.org.uk>,
linux-kernel@...r.kernel.org,
"David S. Miller" <davem@...emloft.net>
Subject: Re: [PATCH net-next + leds v2 6/7] net: phy: marvell: add support
for LEDs controlled by Marvell PHYs
> - Do all PHYs support manual setting of the LED level, or are the PHYs
> that can only work with HW triggers?
There are PHYs with do not have simple on/off.
> - Is setting PHY registers always efficiently possible, or should SW
> triggers be avoided in certain cases? I'm thinking about setups like
> mdio-gpio. I guess this can only become an issue for triggers that
> blink.
There are uses cases where not using software frequently writing
registers would be good. PTP time stamping is one, where the extra
jitter can reduce the accuracy of the clock.
I also think activity blinking in software is unlikely to be
accepted. Nothing extra is allowed in the hot path, when you can be
dealing with a million or more packets per second.
So i would say limit software fallback to link and speed, and don't
assume that is even possible depending on the hardware.
Andrew
Powered by blists - more mailing lists