Re: File Access with VFS
Re: File Access with VFS
- Subject: Re: File Access with VFS
- From: Godfrey van der Linden <email@hidden>
- Date: Fri, 11 Apr 2003 10:46:17 -0700
G'day Scott,
Does your device change its matching parameters with the firmware
download? What sort of device is it, i.e. PCI, firewire etc.
In what way can we boot of your device. Can we net boot from your
network driver? Do you have the appropriate firmware to make it
accessible to OpenFirmware?
If you don't and you don't intend to write it in the future then your
best bet is not to be a boot driver at all. Then you can rely on a
single setProperties from your StartupItem script. If you are not a
driver REQUIRED for booting this is the easiest thing to do.
Godfrey van der Linden
At 10:14 -0700 03-4-11, Scott Lance wrote:
What I am currently doing is copying the firmware into big firmware arrays
in a C file, then loading all the firmware information when I initialize, I
guess it would make the driver a lot smaller in size if I created a 2nd
module, but for right now (in this development iteration) it is a lot easier
(and a lot less messy) to do it the way I currently am. Also, I don't
really have to worry about releasing any other module, but in answer to your
question Quinn and DTS seem to indicate that it is possible in Darwin.
----- Original Message -----
From: "Andrew Gallatin" <email@hidden>
To: "Scott Lance" <email@hidden>
Cc: "Darwin Kern Dev" <email@hidden>
Sent: Friday, April 11, 2003 10:04 AM
Subject: Re: File Access with VFS
In addition to statically compiling and having a user-client, would it
be possible to use a second kernel module?
Eg, we have the "real" device driver, and we have a module which
consists of nothing but firmware arrays. When the device is probed,
both modules are loaded (the driver moduke would "somehow" depend on
the firmware module). The driver then figures out what device(s) is
present, and copies the appropriate firmware array to malloced memory
in the driver. It then "somehow" releases the firmware module, and
that module is unloaded, saving memory. If a similar device was added
to the system, the firmware module would "somehow" get re-loaded, and
the new instance of the device would copy the firmware, as above..
As you can see by the "somehows", I'm a little fuzzy on the details.
> Is a scheme like this possible in Darwin?
_______________________________________________
darwin-kernel mailing list | email@hidden
Help/Unsubscribe/Archives:
http://www.lists.apple.com/mailman/listinfo/darwin-kernel
Do not post admin requests to the list. They will be ignored.