On Tue, Feb 18, 2003 at 10:45:46AM +0100, Thomas Jarosch wrote:
> > > I found Documentation/ioctl-numbers.txt for new-style ioctls.
> > > Currently I'm using 'N' with a range of 20 to 3F.
> > > Before a final release, I'll check if they're still unused.
> >
> > Heh, good luck, I don't think I'll take a patch adding new ioctls
> > without a _very_ good reason.
>
> What about the ioctls to control the latency of the on-chip buffers?
> They are very useful for performance tuning. IMHO there are enough
> useful uses for the new ioctls.
For "normal" serial devices, or this odd bit controlled device?
> > Also, you'll need to make the necessary
> > 64 bit thunking layer if somehow you convince me :)
>
> I thought using the _IO macros should do it
> or did I miss something?
No, those macros don't do it. If you have a strong stomach, look at:
arch/sparc64/kernel/ioctl32.c
for just one of the arch specific files that you will have to modify.
Make sure you get them all, or you will have some angry maintainers to
deal with...
> > > [FTDI Bit Bang mode]
> > > > It looks like this mode does not even look like a serial device anymore.
> > > > I'd recommend just using libusb to talk to the device to use this new
> > > > functionality (and to send new firmware to the device), that way you
> > > > don't have to mess with kernel programming at all.
> > >
> > > Hehe, it's still serial, because you have to write out bytes
> > > to change the status of the pins. It's more like synchronous serial.
> > > If you switch one line from zero to one and back constantly,
> > > you even have a clock line to talk to other chips.
> >
> > But it isn't serial data, so you don't need the tty layer. Why not just
> > do it all though libusb/usbfs in userspace?
>
> But you have to send the data using the serial way.
> A userspace program using libusb would have to copy all the init
> and data writing code of the kernel.
What init and data writing code?
> Also it would be better
> if people could use a stock kernel instead of having to download
> libusb and special software if they want to use the bitbang mode.
Heh, it's _much_ easier to use a userspace program using libusb than to
tell someone to upgrade their kernel version. Trust me on this one.
> The bitbang mode just changes the final behavior of the chip,
> all data transfer is done the same way.
But the end use of the device is then different, right?
> [eeprom checkusm generation]
> > No, that should be done in userspace, not kernelspace.
>
> Already removed the code some hours later
> after my first post about it :-)
I'll slowly convince you yet... :)
thanks,
greg k-h
-------------------------------------------------------
This SF.net email is sponsored by: SlickEdit Inc. Develop an edge.
The most comprehensive and flexible code editor you can use.
Code faster. C/C++, C#, Java, HTML, XML, many more. FREE 30-Day Trial.
www.slickedit.com/sourceforge
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
