[Asterisk-Users] te410p with serial console fails with error: TE410P: Double/missed interrupt detected

[Gavin Hollinger]

Thanks Mark, Sorry for all the questions, I am treading on un-familiar
ground here.  Can anyone think of a way to adjust how long the serial
console holds the interrupt when the kernel sends a message?  Anyone know
what the purpose of this behavior is?

Using the same port in /etc/inittab with:
co:2345:respawn:/sbin/agetty -h -t 60 ttyS0 9600 vt102
is not a problem.

When I add the kernel option, making it a console, the problem starts.
kernel /vmlinuz-2.4.20-20.9smp ro root=LABEL=/ hdc=ide-scsi console=tty0
console=ttyS0,9600n8

I truly appreciate your help, I really did not want a video card, monitor,
or keyboard associated with this machine.  (Rack mounted, equipment room
environment, trying to keep cost, power consumption and heat down.  In
addition, connecting a monitor is not practical.  BIOS settings and
everything are done via serial port.)  How are others dealing with this
issue?  Do certain hardware or kernels behave differently?  Or does
everyone attach keyboards and monitors?

>> The serial console likely is blocking interrupts for too long when it
>> engages.  Zaptel requires interrupts be serviced 1000 times per
>> second, for an interrupt latency of < 1ms max.  Generally this is not
>> a problem, but some systems in linux (e.g. IDE if DMA is enabled,
>> Framebuffer console, etc) block interrupts for exceptionally long
>> periods of time, thus causing this sort of problem.
>
> [Snip...]
>
> I guess in the case of IDE DMA, using hdparm -u1 would remove the
> interrupt blocking. DMA on IDE disks is a nice thing to have...
>
> Richard

Thanks Richard,
Interesting snip from man hdparm for others:
-u     Get/set  interrupt-unmask  flag  for  the drive.  A setting of 1
permits the driver to unmask other interrupts during  processing of  a
disk interrupt, which greatly improves Linux's responsiveness and
eliminates "serial port overrun" errors.  Use this feature  with  caution:
 some  drive/controller combinations do not tolerate the increased I/O
latencies possible when this  feature is enabled, resulting in massive
file system corruption.  In particular, CMD-640B and RZ1000 (E)IDE
interfaces can be unreliable (due  to  a  hardware flaw) when this option
is used with kernel versions earlier than 2.0.13.  Disabling the IDE 
prefetch  feature  of these interfaces (usually a BIOS/CMOS setting)
provides a safe fix for the problem for use with earlier kernels.