Re: [REVIEW] NVM Express driver

[Greg KH <greg@kroah.com>]

On Fri, Mar 04, 2011 at 12:43:48PM +0000, Alan Cox wrote:
> On Thu, 3 Mar 2011 14:22:26 -0800
> > > How would the driver know that it should call request_firmware()?
> > > Do it every 60 seconds in case somebody's downloaded some new firmware?
> > 
> > Ick, no, just use the function provided that lets you create a firmware
> > request and be notified when it is written to,
> > request_firmware_nowait().  That is what it is there for.
> 
> Bunkum
> 
> It's there for automated loading of needed firmware

And non-automated loading of firmware as well.  Dell uses this for
updating their BIOSes just fine, with a userspace tool that initiates
the loading of the firmware.

> > Anyway, just use request_firmware_nowait(), you will be fine.
> 
> Not really. You've got no way with the request_firmware interface for the
> user space and kernel space to reliably time objects together eg to do
> user side work before a firmware load.

How does Dell do it?

> Almost all the other stuff in the kernel which is for firmware
> programming as opposed to automatic loading is not using request_firmware
> - and for good reasons like needing to be able to specify the path to the
> object reliably - which cannot be done with namespaces.
> 
> From a security perspective, a correctness perspective, for reliability
> and managability request_firmware is the wrong interface for flashing new
> firmware.

Tell me how you really feel now :)

So, what could be changed in the current firmware interface to fix this
problem in a manner which would solve these perceived issues?

thanks,

greg k-h