Re: Machin dependent serial port patches

[David Woodhouse (dwmw2@infradead.org)]

On Tue, 2003-05-20 at 13:09, Alan Cox wrote:
> On Maw, 2003-05-20 at 01:56, J.C. Wren wrote:
> > 	One of the things I noticed in the port of 2.5.69 to the 386EX embedded 
> > system is that serial.h appears to not be a mach-xxx positionable file.  The 
> > 386EX board uses standard 8250 type serial ports, but at 3.6864Mhz instad of 
> > 1.8432Mhz.  There appears to be no way to build a patch set without modifying 
> > include/i386/serial.h.  Would this not be better places in mach-defaults?  
> > I'm trying very hard to modify as few files as possible when building these 
> > patch sets.
> 
> Making asm-i386/serial.h include a mach- file sounds the right thing to
> do. mach- for x86 is pretty new so a lot of stuff that maybe should be
> in it, hasnt migrated yet.

I disagree. I think it would be better to get rid of the hard-coded
table altogether and let something in the machine-specific code call
register_serial() during early boot.

Even on PeeCee hardware, that lets us do the superio chip probe and
potentially register high-speed serial ports, before the pnpbios probe
and finally falling back to the old standard I/O addresses.

-- 
dwmw2

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/