Re: R: How to avoid serial port buffer overruns?

[Russell King <rmk+lkml@arm.linux.org.uk>]

On Fri, Aug 18, 2006 at 08:01:06PM +0100, Russell King wrote:
> On Fri, Aug 18, 2006 at 02:52:53PM -0400, Lee Revell wrote:
> > On Fri, 2006-08-18 at 19:34 +0100, Russell King wrote:
> > > That "0000:00:0b.0" looks like a PCI device ID.  If it were a fourport
> > > board, it would be "serial8250.3" according to the current enumeration
> > > in linux/serial_8250.h.
> > > 
> > > Also, another give away is that IRQ185 is being setup as a PCI interrupt
> > > immediately prior to the devices being registered.
> > > 
> > > And I doubt that an ISA board (which is what fourport is) would ever get
> > > such a high IRQ number.
> > > 
> > 
> > So you're saying that the standard 8250 driver is being used?
> 
> Yes, which is also the case with 8250_fourport.  8250_fourport is just
> a probe module just like 8250_pnp or 8250_pci.
> 
> > http://www.moschip.com/html/MCS9845.html
> 
> That also clearly says its a PCI device. 8)
> 
> What problem are we talking about here again?  Sorry, I've completely lost
> track and this particular thread of 26 messages is soo convoluted and too
> much to re-read.
> 
> Since you only appear to be the messenger, wouldn't it be far better to get
> the person with the problem to report and respond rather than sitting in
> the middle?

BTW, let's have _one_ email which describes what problem it is, the
hardware which its being seen on, the kernel configuration, what
modules are loaded, the lspci output and so forth.  Let's not spread
all the information over 20 emails.

I just can't work with dribbled small bits of inforamtion sparsely
spread.  Neither can I work when it takes days to get that information -
I forget both the reported details and the thread of thought I was
following to maybe work towards a conclusion.

(Though at present I'm rather devoid of ideas in this area.)

-- 
Russell King
 Linux kernel    2.6 ARM Linux   - http://www.arm.linux.org.uk/
 maintainer of:  2.6 Serial core