ohpaged

Sometimes Dev.ser generates +ohpaged and stops communication when cable
is removed from port and returned back. How to disable such behaviour ?
The communication is stopped until the flag is switched off. Furthermore
I don’t know how to switch the ohpaged flag off in C code (Watcom C
10.6, other attributes like DTR I can set up through qnx_ioctl(), but
not ohpaged). Now I am able to use stty utility only to resume
communication…

Thanks for any advice.

Andy

andy <andy@microstep-hdo.sk> wrote:

Sometimes Dev.ser generates +ohpaged and stops communication when cable
is removed from port and returned back. How to disable such behaviour ?
The communication is stopped until the flag is switched off. Furthermore
I don’t know how to switch the ohpaged flag off in C code (Watcom C
10.6, other attributes like DTR I can set up through qnx_ioctl(), but
not ohpaged). Now I am able to use stty utility only to resume
communication…

you can turn it off in a program … check out TCOOFFHW and TCOONHW in the
tflow() function.


Randy Martin email: randy@qnx.com www.qnx.com
QNX Software Systems Ltd. QUICS: randy (613) 591-0934 (data)
(613) 591-0931 x287 (voice) mail: 175 Terence Matthews Cr
(613) 591-3579 (fax) Kanata, Ontario, Canada K2M 1W8

I recommend starting Dev.ser with flow control disabled and in Raw mode for
communications ports:
Dev.ser -E -F … &
It seems more robust then disabling flow control with stty…
Don’t know why
-Paul

Example:

Dev.ser Settings:

-E raw mode

-F Disable Hardware Flow Control

-b baudrate

-H HUP Enable (default)

-I Input Buffer Size

-t 16550 Fifo Threshold (-t0 = no FIFO),

Threshold will hold irq for 4 idle char periods

-N Name Prefix

port,irq Port and Interrupt Pairs,

IRQ9 Replaces IRQ2 in PCAT Hardware, Use IRQ9 for IRQ2 Cards

PCs with PCI and EISA Slots may need IRQ selected in CMOS

Recommended Addresses for Extended Cards: 280-2Bf 100-1f0

Standard Com1-4: 3f8,4 2f8,3 3e8,9(2) 2e8,5

Dev.ser -E -F -t8 -I4096 -b115200 -N ser 3f8,4 2f8,3 &




andy <andy@microstep-hdo.sk> wrote in message
news:39C0A582.810C4365@microstep-hdo.sk

Sometimes Dev.ser generates +ohpaged and stops communication when cable
is removed from port and returned back. How to disable such behaviour ?
The communication is stopped until the flag is switched off. Furthermore
I don’t know how to switch the ohpaged flag off in C code (Watcom C
10.6, other attributes like DTR I can set up through qnx_ioctl(), but
not ohpaged). Now I am able to use stty utility only to resume
communication…

Thanks for any advice.

Andy

If another process has the port open when its configuration via stty,
then when the first process closes it may reset the port to
however it was when that process started, giving the appearance of
“less robust”.

Paul Russell wrote:

I recommend starting Dev.ser with flow control disabled and in Raw mode for
communications ports:
Dev.ser -E -F … &
It seems more robust then disabling flow control with stty…
Don’t know why
-Paul

Example:

Dev.ser Settings:

-E raw mode

-F Disable Hardware Flow Control

-b baudrate

-H HUP Enable (default)

-I Input Buffer Size

-t 16550 Fifo Threshold (-t0 = no FIFO),

Threshold will hold irq for 4 idle char periods

-N Name Prefix

port,irq Port and Interrupt Pairs,

IRQ9 Replaces IRQ2 in PCAT Hardware, Use IRQ9 for IRQ2 Cards

PCs with PCI and EISA Slots may need IRQ selected in CMOS

Recommended Addresses for Extended Cards: 280-2Bf 100-1f0

Standard Com1-4: 3f8,4 2f8,3 3e8,9(2) 2e8,5

Dev.ser -E -F -t8 -I4096 -b115200 -N ser 3f8,4 2f8,3 &

andy <> andy@microstep-hdo.sk> > wrote in message
news:> 39C0A582.810C4365@microstep-hdo.sk> …
Sometimes Dev.ser generates +ohpaged and stops communication when cable
is removed from port and returned back. How to disable such behaviour ?
The communication is stopped until the flag is switched off. Furthermore
I don’t know how to switch the ohpaged flag off in C code (Watcom C
10.6, other attributes like DTR I can set up through qnx_ioctl(), but
not ohpaged). Now I am able to use stty utility only to resume
communication…

Thanks for any advice.

Andy

Hi Andy,

I had a similar problem with software handshaking, you have to sense the
+ohpaged. If you want you can then force it off. I had a similar problem
when receiving a spurious XOFF. Here’s some sample code that worked fine for
me modifed to work with ohpaged:


int IsPaged( int handle)
{
struct termios tc;

if ( !tcgetattr( handle, &tc ) )
{
if( tc.c_status & TC_OHFLOW )
return( 1 );
}
return( 0 );
}

PagedOff( int handle)
{
tcflow( handle, TCOONHW );
}

I counted consecutive paged errors then forced it off, not really applicable
in you case I guess.

Regards


Glyn

“andy” <andy@microstep-hdo.sk> wrote in message
news:39C0A582.810C4365@microstep-hdo.sk

Sometimes Dev.ser generates +ohpaged and stops communication when cable
is removed from port and returned back. How to disable such behaviour ?
The communication is stopped until the flag is switched off. Furthermore
I don’t know how to switch the ohpaged flag off in C code (Watcom C
10.6, other attributes like DTR I can set up through qnx_ioctl(), but
not ohpaged). Now I am able to use stty utility only to resume
communication…

Thanks for any advice.

Andy

I have had success with using ‘Dev.ser -FEmSx’ in the sysinit.

Glyn Durban wrote:

Hi Andy,

I had a similar problem with software handshaking, you have to sense the
+ohpaged. If you want you can then force it off. I had a similar problem
when receiving a spurious XOFF. Here’s some sample code that worked fine for
me modifed to work with ohpaged:

int IsPaged( int handle)
{
struct termios tc;

if ( !tcgetattr( handle, &tc ) )
{
if( tc.c_status & TC_OHFLOW )
return( 1 );
}
return( 0 );
}

PagedOff( int handle)
{
tcflow( handle, TCOONHW );
}

I counted consecutive paged errors then forced it off, not really applicable
in you case I guess.

Regards

Glyn

“andy” <> andy@microstep-hdo.sk> > wrote in message
news:> 39C0A582.810C4365@microstep-hdo.sk> …
Sometimes Dev.ser generates +ohpaged and stops communication when cable
is removed from port and returned back. How to disable such behaviour ?
The communication is stopped until the flag is switched off. Furthermore
I don’t know how to switch the ohpaged flag off in C code (Watcom C
10.6, other attributes like DTR I can set up through qnx_ioctl(), but
not ohpaged). Now I am able to use stty utility only to resume
communication…

Thanks for any advice.

Andy

Yes, that’s solution for me. Thanks a lot.

Andy

Glyn Durban wrote:

Hi Andy,

I had a similar problem with software handshaking, you have to sense the
+ohpaged. If you want you can then force it off. I had a similar problem
when receiving a spurious XOFF. Here’s some sample code that worked fine for
me modifed to work with ohpaged:

int IsPaged( int handle)
{
struct termios tc;

if ( !tcgetattr( handle, &tc ) )
{
if( tc.c_status & TC_OHFLOW )
return( 1 );
}
return( 0 );
}

PagedOff( int handle)
{
tcflow( handle, TCOONHW );
}

I counted consecutive paged errors then forced it off, not really applicable
in you case I guess.

Regards

Glyn

“andy” <> andy@microstep-hdo.sk> > wrote in message
news:> 39C0A582.810C4365@microstep-hdo.sk> …
Sometimes Dev.ser generates +ohpaged and stops communication when cable
is removed from port and returned back. How to disable such behaviour ?
The communication is stopped until the flag is switched off. Furthermore
I don’t know how to switch the ohpaged flag off in C code (Watcom C
10.6, other attributes like DTR I can set up through qnx_ioctl(), but
not ohpaged). Now I am able to use stty utility only to resume
communication…

Thanks for any advice.

Andy