Hylafax Mailing List Archives

[Date Prev][Date Next][Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: [hylafax-users] Mainpine Quatro+ not picking up in Hylafax



Many thanks for all of the input on the issue that John is trying to
overcome.

While I'm not a Linux or HylaFAX expert I thought some background
information on the Mainpine products might be useful.

As many of you are aware we recommend that if the kernel is recompiled then
multiport support is included.  This allows the serial driver to support
more than 4 ports on a single irq (CONFIG_SERIAL_MANY_PORTS) and to use the
interrupt status register that is a feature of our hardware
(CONFIG_SERIAL_EXTENDED).  That said,  CONFIG_SERIAL_MANY_PORTS is only
required on our 8 port OCTO+ board, and CONFIG_SERIAL_EXTENDED is only
recommended because we have identified a race condition which occasionally
results in the serial driver 'missing' an interrupt when under heavy load
(this was originally identified by a customer who is deploying 128 port
systems).  

Now, many customers systems work just fine without using the interrupt
status register (CONFIG_SERIAL_EXTENDED) which is to be expected because the
race condition is very rare.  It is also worth noting that multiport support
seems to have been extensively reworked in the 2.6 kernel including the
removal of the interrupt status register.  This would seem to leave us a
number of options.  We could try to fully debug the race condition (very
time consuming), lobby to have the interrupt status register supported once
again, or release our own version of the serial driver..... at this time the
jury is still out.....

It is interesting to note that the newer Kernels seem to be trying to
automatically configure the boards.... this would be great if it worked
every time but as John has reported it seems to remap the tty's only
occasionally.

Returning to John's specific problems it seems clear that the board has been
correctly installed and wired up because his system can send fax's without
problem.  So the question remains, why can't it receive ?  One thing I did
spot is that John reported Multitech 5634-series configuration;

>> # prototype config for MultiTech 5634-series modems using # the 
>> Lucent/Agere chipset supporting V.34-Fax.

Whereas the faq on our website seems to imply that HylaFAX used a
configuration specific to the RockForce DUO+;

http://www.mainpine.com/faqs/ans47.htm

As I said, I'm no expert and just wanted to raise this potential difference
while I wait for our engineering team to investigate John's problems.

All comments welcome !

Regards

Andrew Rinaldi
Support
Mainpine Ltd.
e: support@xxxxxxxxxxxx
w: http://www.mainpine.com
t: +44(0) 1373 830888
f: +44(0) 1373 830768

-----Original Message-----
From: hylafax-users-bounce@xxxxxxxxxxx
[mailto:hylafax-users-bounce@xxxxxxxxxxx] On Behalf Of Lee Howard
Sent: 14 July 2005 20:49
To: John Meuser
Cc: hylafax-users@xxxxxxxxxxx
Subject: Re: [hylafax-users] Mainpine Quatro+ not picking up in Hylafax

John Meuser wrote:

>Ah.. that would explain things.. it has assigned the interfaces to 
>completely new TTY's.... :
>
>meuserj@s-uhsfax1 ~$ dmesg | grep tty
>ttyS0 at I/O 0x3f8 (irq = 4) is a 16550A
>ttyS14 at I/O 0xec00 (irq = 185) is a 16550A
>ttyS15 at I/O 0xec08 (irq = 185) is a 16550A
>ttyS44 at I/O 0xec10 (irq = 185) is a 16550A
>ttyS45 at I/O 0xec18 (irq = 185) is a 16550A
>ttyS1 at I/O 0xec78 (irq = 185) is a 16450
>
>So, I reconfigured it using the new TTY's.. and it still doesn't work.
>It does exactly what it did with kernel 2.4.. I can send faxes, but it 
>won't pick up the call when I try to send to it.
>

I have a system with a Duo+ (and two external serial modems and two internal
ISA modems) running kernel 2.6.7.  I have the serial driver compiled as a
module.  During init (in /etc/rc.serial) I have:

  modprobe 8250
  setserial /dev/ttyS4 port 0xec00 irq 12 autoconfig ^fourport
  setserial /dev/ttyS5 port 0xec08 irq 12 autoconfig ^fourport

When I do 'dmesg | grep tty' I get:

  ttyS0 at I/O 0x3f8 (irq = 4) is a 16550A
  ttyS1 at I/O 0x2f8 (irq = 3) is a 16550A
  ttyS2 at I/O 0x3e8 (irq = 4) is a 16550A
  ttyS3 at I/O 0x2e8 (irq = 3) is a 16550A

These are not the ports from the Duo+.  However setserial on each device
reports:

  /dev/ttyS0, UART: 16550A, Port: 0x03f8, IRQ: 4
  /dev/ttyS1, UART: 16550A, Port: 0x02f8, IRQ: 3
  /dev/ttyS2, UART: 16550A, Port: 0x03e8, IRQ: 4
  /dev/ttyS3, UART: 16550A, Port: 0x02e8, IRQ: 3
  /dev/ttyS4, UART: 16550A, Port: 0xec00, IRQ: 12
  /dev/ttyS5, UART: 16550A, Port: 0xec08, IRQ: 12

And all of the modems run just fine.

I was seeing dmesg output like you are seeing and things didn't work right.
I don't know what the problem was, but in order to get rid of that messy
(using "high" tty numbers) dmesg output I decided to not compile-in the
serial driver into the kernel, but to use it as a module instead.  And after
doing that I didn't see the problems any more.

Lee.


____________________ HylaFAX(tm) Users Mailing List _______________________
  To subscribe/unsubscribe, click http://lists.hylafax.org/cgi-bin/lsg2.cgi
 On UNIX: mail -s unsubscribe hylafax-users-request@xxxxxxxxxxx < /dev/null
  *To learn about commercial HylaFAX(tm) support, mail sales@xxxxxxxxx*




____________________ HylaFAX(tm) Users Mailing List _______________________
  To subscribe/unsubscribe, click http://lists.hylafax.org/cgi-bin/lsg2.cgi
 On UNIX: mail -s unsubscribe hylafax-users-request@xxxxxxxxxxx < /dev/null
  *To learn about commercial HylaFAX(tm) support, mail sales@xxxxxxxxx*



Home
Report any problems to webmaster@hylafax.org

HylaFAX is a trademark of Silicon Graphics Corporation.
Internet connectivity for hylafax.org is provided by:
VirtuALL Private Host Services