[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <Y/NBJ5xzUoNHEOm2@kroah.com>
Date: Mon, 20 Feb 2023 10:45:11 +0100
From: Greg KH <gregkh@...uxfoundation.org>
To: Kumaravel.Thiagarajan@...rochip.com
Cc: arnd@...db.de, linux-gpio@...r.kernel.org, michael@...le.cc,
srinivas.kandagatla@...aro.org, linux-kernel@...r.kernel.org,
UNGLinuxDriver@...rochip.com, Tharunkumar.Pasumarthi@...rochip.com
Subject: Re: [PATCH v5 char-misc-next] misc: microchip: pci1xxxx: Add
OTP/EEPROM driver for the pci1xxxx switch
On Mon, Feb 20, 2023 at 09:31:45AM +0000, Kumaravel.Thiagarajan@...rochip.com wrote:
> > I do not recommend using a block interface for this at all. Why not
> > the
> > "normal" EEPROM interface that the kernel has today (i.e. a binary
> > sysfs
> > file)? That way you can mmap it and edit locations how ever you
> > want.
> Greg, I have one question about the sysfs interface. If OTP and EEPROM
> are enumerated as two sysfs files of 8KB each, will that start
> occupying the RAM permamently or is the swapping in and out of the RAM
> handled automatically by the kernel based on the user trying to access
> it.
It depends on exactly what you are doing here with the binary sysfs
file. Try it yourself and see how it works.
thanks,
greg k-h
Powered by blists - more mailing lists