Documentation/networking/slicecom.txt - kernel/msm-4.9 - Gitiles


 SliceCOM adapter user's documentation - for the 0.51 driver version

 Written by Bartók István <bartoki@itc.hu>

 English translation: Lakatos György <gyuri@itc.hu>
 Mon Dec 11 15:28:42 CET 2000

 Last modified: Wed Aug 29 17:25:37 CEST 2001

 -----------------------------------------------------------------

 Usage:

 Compiling the kernel:

 Code maturity level options
 	[*] Prompt for development and/or incomplete code/drivers

 Network device support
 	Wan interfaces
 		<M> MultiGate (COMX) synchronous
 			<M> Support for MUNICH based boards: SliceCOM, PCICOM (NEW)
 			<M> Support for HDLC and syncPPP...


 Loading the modules:

 modprobe comx

 modprobe comx-proto-ppp		# module for  Cisco-HDLC and SyncPPP protocols

 modprobe comx-hw-munich		# the module logs information by the kernel
 				# about the detected boards


 Configuring the board:

 # This interface will use the Cisco-HDLC line protocol,
 # the timeslices assigned are 1,2 (128 KiBit line speed)
 # (the first data timeslice in the G.703 frame is no. 1)
 #
 mkdir /proc/comx/comx0.1/
 echo slicecom	>/proc/comx/comx0.1/boardtype
 echo hdlc	>/proc/comx/comx0.1/protocol
 echo 1 2	>/proc/comx/comx0.1/timeslots


 # This interface uses SyncPPP line protocol, the assigned
 # is no. 3 (64 KiBit line speed)
 #
 mkdir /proc/comx/comx0.2/
 echo slicecom	>/proc/comx/comx0.2/boardtype
 echo ppp	>/proc/comx/comx0.2/protocol
 echo 3		>/proc/comx/comx0.2/timeslots

 ...

 ifconfig comx0.1 up
 ifconfig comx0.2 up

 -----------------------------------------------------------------

 The COMX interfaces use a 10 packet transmit queue by default, however WAN
 networks sometimes use bigger values (20 to 100), to utilize the line better
 by large traffic (though the line delay increases because of more packets
 join the queue).

 # ifconfig comx0 txqueuelen 50

 This option is only supported by the ifconfig command of the later
 distributions, which came with 2.2 kernels, such as RedHat 6.1 or Debian 2.2.

 You can download a newer netbase packet from
 http://www.debian.org/~rcw/2.2/netbase/ for Debian 2.1, which has a new
 ifconfig. You can get further information about using 2.2 kernel with
 Debian 2.1 from http://www.debian.org/releases/stable/running-kernel-2.2

 -----------------------------------------------------------------

 The SliceCom LEDs:

 red	- on, if the interface is unconfigured, or it gets Remote Alarm-s
 green	- on, if the board finds frame-sync in the received signal

 A bit more detailed:

 red:	green:	meaning:

 -	-	no frame-sync, no signal received, or signal SNAFU.
 -	on	"Everything is OK"
 on	on	Reception is ok, but the remote end sends Remote Alarm
 on	-	The interface is unconfigured

 -----------------------------------------------------------------

 A more detailed description of the hardware setting options:

 The general and the protocol layer options described in the 'comx.txt' file
 apply to the SliceCom as well, I only summarize the SliceCom hardware specific
 settings below.

 The '/proc/comx' configuring interface:

 An interface directory should be created for every timeslot group with
 'mkdir', e,g: 'comx0', 'comx1' etc. The timeslots can be assigned here to the
 specific interface. The Cisco-like naming convention (serial3:1 - first
 timeslot group of the 3rd. board) can't be used here, because these mean IP
 aliasing in Linux.

 You can give any meaningful name to keep the configuration clear;
 e.g: 'comx0.1', 'comx0.2', 'comx1.1', comx1.2', if you have two boards
 with two interfaces each.

 Settings, which apply to the board:

 Neither 'io' nor 'irq' settings required, the driver uses the resources
 given by the PCI BIOS.

 comx0/boardnum	- board number of the SliceCom in the PC (using the 'natural'
 		PCI order) as listed in '/proc/pci' or the output of the
 	 	'lspci' command, generally the slots nearer to the motherboard
 		PCI driver chips have the lower numbers.

 		Default: 0 (the counting starts with 0)

 Though the options below are to be set on a single interface, they apply to the
 whole board. The restriction, to use them on 'UP' interfaces, is because the
 command sequence below could lead to unpredictable results.

 	# echo 0        >boardnum
 	# echo internal >clock_source
 	# echo 1        >boardnum

 The sequence would set the clock source of board 0.

 These settings will persist after all the interfaces are cleared, but are
 cleared when the driver module is unloaded and loaded again.

 comx0/clock_source - source of the transmit clock
 	Usage:

 	# echo line     >/proc/comx/comx0/clock_source
 	# echo internal >/proc/comx/comx0/clock_source

 	line	- The Tx clock is being decoded if the input data stream,
 		if no clock seen on the input, then the board will use it's
 		own clock generator.

 	internal - The Tx clock is supplied by the builtin clock generator.

 	Default: line

 	Normally, the telecommunication company's end device (the HDSL
 	modem) provides the Tx clock, that's why 'line' is the default.

 comx0/framing	- Switching CRC4 off/on

 	CRC4: 16 PCM frames (The 32 64Kibit channels are multiplexed into a
 	PCM frame, nothing to do with HDLC frames) are divided into 2x8
 	groups, each group has a 4 bit CRC.

 	# echo crc4	>/proc/comx/comx0/framing
 	# echo no-crc4	>/proc/comx/comx0/framing

 	Default is 'crc4', the Hungarian MATAV lines behave like this.
 	The traffic generally passes if this setting on both ends don't match.

 comx0/linecode	- Setting the line coding

 	# echo hdb3	>/proc/comx/comx0/linecode
 	# echo ami	>/proc/comx/comx0/linecode

 	Default a 'hdb3', MATAV lines use this.

 	(AMI coding is rarely used with E1 lines). Frame sync may occur, if
 	this setting doesn't match the other end's, but CRC4 and data errors
 	will come, which will result in CRC errors on HDLC/SyncPPP level.

 comx0/reg	- direct access to the board's MUNICH (reg) and FALC (lbireg)
 comx0/lbireg	circuit's registers

 	# echo >reg 0x04 0x0	- write 0 to register 4
 	# echo >reg 0x104	- write the contents of register 4 with
 				printk() to syslog

 WARNING! These are only for development purposes, messing with this will
 	result much trouble!

 comx0/loopback - Places a loop to the board's G.703 signals

 	# echo none   >/proc/comx/comx0/loopback
 	# echo local  >/proc/comx/comx0/loopback
 	# echo remote >/proc/comx/comx0/loopback

 	none   - normal operation, no loop
 	local  - the board receives it's own output
 	remote - the board sends the received data to the remote side

 	Default: none

 -----------------------------------------------------------------

 Interface (channel group in Cisco terms) settings:

 comx0/timeslots	- which timeslots belong to the given interface

 	Setting:

 	# echo '1 5 2 6 7 8' >/proc/comx/comx0/timeslots

 	# cat /proc/comx/comx0/timeslots
 	1 2 5 6 7 8
 	#

 	Finding a timeslot:

 	# grep ' 4' /proc/comx/comx*/timeslots
 	/proc/comx/comx0/timeslots:1 3 4 5 6
 	#

 	The timeslots can be in any order, '1 2 3' is the same as '1 3 2'.

 	The interface has to be DOWN during the setting ('ifconfig comx0
 	down'), but the other interfaces could operate normally.

 	The driver checks if the assigned timeslots are vacant, if not, then
 	the setting won't be applied.

 	The timeslot values are treated as decimal numbers, not to misunderstand
 	values of 08, 09 form.

 -----------------------------------------------------------------

 Checking the interface and board status:

 - Lines beginning with ' ' (space) belong to the original output, the lines
 which begin with '//' are the comments.

  papaya:~$ cat /proc/comx/comx1/status
  Interface administrative status is UP, modem status is UP, protocol is UP
  Modem status changes: 0, Transmitter status is IDLE, tbusy: 0
  Interface load (input): 978376 / 947808 / 951024 bits/s (5s/5m/15m)
                (output): 978376 / 947848 / 951024 bits/s (5s/5m/15m)
  Debug flags: none
  RX errors: len: 22, overrun: 1, crc: 0, aborts: 0
             buffer overrun: 0, pbuffer overrun: 0
  TX errors: underrun: 0
  Line keepalive (value: 10) status UP [0]

 // The hardware specific part starts here:
  Controller status:
          No alarms

 // Alarm:
 //
 // No alarms - Everything OK
 //
 // LOS  - Loss Of Signal - No signal sensed on the input
 // AIS  - Alarm Indication Signal - The remote side sends '11111111'-s,
 //	it tells, that there's an error condition, or it's not
 //	initialised.
 // AUXP - Auxiliary Pattern Indication - 01010101.. received.
 // LFA  - Loss of Frame Alignment - no frame sync received.
 // RRA  - Receive Remote Alarm - the remote end's OK, but signals error cond.
 // LMFA - Loss of CRC4 Multiframe Alignment - no CRC4 multiframe sync.
 // NMF  - No Multiframe alignment Found after 400 msec - no such alarm using
 //	no-crc4 or crc4 framing, see below.
 //
 // Other possible error messages:
 //
 // Transmit Line Short - the board felt, that it's output is short-circuited,
 // 	so it switched the transmission off. (The board can't definitely tell,
 //	that it's output is short-circuited.)

 // Chained list of the received packets, for debug purposes:

  Rx ring:
          rafutott: 0
          lastcheck: 50845731, jiffies: 51314281
          base: 017b1858
          rx_desc_ptr: 0
          rx_desc_ptr: 017b1858
          hw_curr_ptr: 017b1858
          06040000 017b1868 017b1898 c016ff00
          06040000 017b1878 017b1e9c c016ff00
          46040000 017b1888 017b24a0 c016ff00
          06040000 017b1858 017b2aa4 c016ff00

 // All the interfaces using the board: comx1, using the 1,2,...16 timeslots,
 // comx2, using timeslot 17, etc.

  Interfaces using this board: (channel-group, interface, timeslots)
           0 comx1: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
           1 comx2: 17
           2 comx3: 18
           3 comx4: 19
           4 comx5: 20
           5 comx6: 21
           6 comx7: 22
           7 comx8: 23
           8 comx9: 24
           9 comx10: 25
          10 comx11: 26
          11 comx12: 27
          12 comx13: 28
          13 comx14: 29
          14 comx15: 30
          15 comx16: 31

 // The number of events handled by the driver during an interrupt cycle:

  Interrupt work histogram:
  hist[ 0]:        0 hist[ 1]:        2 hist[ 2]:    18574 hist[ 3]:       79
  hist[ 4]:       14 hist[ 5]:        1 hist[ 6]:        0 hist[ 7]:        1
  hist[ 8]:        0 hist[ 9]:        7

 // The number of packets to send in the Tx ring, when a new one arrived:

  Tx ring histogram:
  hist[ 0]:     2329 hist[ 1]:        0 hist[ 2]:        0 hist[ 3]:        0

 // The error counters of the E1 interface, according to the RFC2495,
 // (similar to the Cisco "show controllers e1" command's output:
 // http://www.cisco.com/univercd/cc/td/doc/product/software/ios11/rbook/rinterfc.htm#xtocid25669126)

 Data in current interval (91 seconds elapsed):
    9516 Line Code Violations, 65 Path Code Violations, 2 E-Bit Errors
    0 Slip Secs, 2 Fr Loss Secs, 2 Line Err Secs, 0 Degraded Mins
    0 Errored Secs, 0 Bursty Err Secs, 0 Severely Err Secs, 11 Unavail Secs
 Data in Interval 1 (15 minutes):
    0 Line Code Violations, 0 Path Code Violations, 0 E-Bit Errors
    0 Slip Secs, 0 Fr Loss Secs, 0 Line Err Secs, 0 Degraded Mins
    0 Errored Secs, 0 Bursty Err Secs, 0 Severely Err Secs, 0 Unavail Secs
 Data in last 4 intervals (1 hour):
    0 Line Code Violations, 0 Path Code Violations, 0 E-Bit Errors
    0 Slip Secs, 0 Fr Loss Secs, 0 Line Err Secs, 0 Degraded Mins
    0 Errored Secs, 0 Bursty Err Secs, 0 Severely Err Secs, 0 Unavail Secs
 Data in last 96 intervals (24 hours):
    0 Line Code Violations, 0 Path Code Violations, 0 E-Bit Errors
    0 Slip Secs, 0 Fr Loss Secs, 0 Line Err Secs, 0 Degraded Mins
    0 Errored Secs, 0 Bursty Err Secs, 0 Severely Err Secs, 0 Unavail Secs

 -----------------------------------------------------------------

 Some unique options, (may get into the driver later):
 Treat them very carefully, these can cause much trouble!

 	modified CRC-4, for improved interworking of CRC-4 and non-CRC-4
 	devices: (see page 107 and g706 Annex B)
 		lbireg[ 0x1b ] |= 0x08
 		lbireg[ 0x1c ] |= 0xc0

 	- The NMF - 'No Multiframe alignment Found after 400 msec' alarm
 	comes into account.

 	FALC - the line driver chip.
 	local loop - I hear my transmission back.
 	remote loop - I echo the remote transmission back.

 	Something useful for finding errors:

 		- local loop for timeslot 1 in the FALC chip:

 	# echo >lbireg 0x1d 0x21

 		- Switching the loop off:

 	# echo >lbireg 0x1d 0x00

	SliceCOM adapter user's documentation - for the 0.51 driver version

	Written by Bartók István <bartoki@itc.hu>

	English translation: Lakatos György <gyuri@itc.hu>
	Mon Dec 11 15:28:42 CET 2000

	Last modified: Wed Aug 29 17:25:37 CEST 2001

	-----------------------------------------------------------------

	Usage:

	Compiling the kernel:

	Code maturity level options
	[*] Prompt for development and/or incomplete code/drivers

	Network device support
	Wan interfaces
	<M> MultiGate (COMX) synchronous
	<M> Support for MUNICH based boards: SliceCOM, PCICOM (NEW)
	<M> Support for HDLC and syncPPP...


	Loading the modules:

	modprobe comx

	modprobe comx-proto-ppp # module for Cisco-HDLC and SyncPPP protocols

	modprobe comx-hw-munich # the module logs information by the kernel
	# about the detected boards


	Configuring the board:

	# This interface will use the Cisco-HDLC line protocol,
	# the timeslices assigned are 1,2 (128 KiBit line speed)
	# (the first data timeslice in the G.703 frame is no. 1)
	#
	mkdir /proc/comx/comx0.1/
	echo slicecom >/proc/comx/comx0.1/boardtype
	echo hdlc >/proc/comx/comx0.1/protocol
	echo 1 2 >/proc/comx/comx0.1/timeslots


	# This interface uses SyncPPP line protocol, the assigned
	# is no. 3 (64 KiBit line speed)
	#
	mkdir /proc/comx/comx0.2/
	echo slicecom >/proc/comx/comx0.2/boardtype
	echo ppp >/proc/comx/comx0.2/protocol
	echo 3 >/proc/comx/comx0.2/timeslots

	...

	ifconfig comx0.1 up
	ifconfig comx0.2 up

	-----------------------------------------------------------------

	The COMX interfaces use a 10 packet transmit queue by default, however WAN
	networks sometimes use bigger values (20 to 100), to utilize the line better
	by large traffic (though the line delay increases because of more packets
	join the queue).

	# ifconfig comx0 txqueuelen 50

	This option is only supported by the ifconfig command of the later
	distributions, which came with 2.2 kernels, such as RedHat 6.1 or Debian 2.2.

	You can download a newer netbase packet from
	http://www.debian.org/~rcw/2.2/netbase/ for Debian 2.1, which has a new
	ifconfig. You can get further information about using 2.2 kernel with
	Debian 2.1 from http://www.debian.org/releases/stable/running-kernel-2.2

	-----------------------------------------------------------------

	The SliceCom LEDs:

	red - on, if the interface is unconfigured, or it gets Remote Alarm-s
	green - on, if the board finds frame-sync in the received signal

	A bit more detailed:

	red: green: meaning:

	- - no frame-sync, no signal received, or signal SNAFU.
	- on "Everything is OK"
	on on Reception is ok, but the remote end sends Remote Alarm
	on - The interface is unconfigured

	-----------------------------------------------------------------

	A more detailed description of the hardware setting options:

	The general and the protocol layer options described in the 'comx.txt' file
	apply to the SliceCom as well, I only summarize the SliceCom hardware specific
	settings below.

	The '/proc/comx' configuring interface:

	An interface directory should be created for every timeslot group with
	'mkdir', e,g: 'comx0', 'comx1' etc. The timeslots can be assigned here to the
	specific interface. The Cisco-like naming convention (serial3:1 - first
	timeslot group of the 3rd. board) can't be used here, because these mean IP
	aliasing in Linux.

	You can give any meaningful name to keep the configuration clear;
	e.g: 'comx0.1', 'comx0.2', 'comx1.1', comx1.2', if you have two boards
	with two interfaces each.

	Settings, which apply to the board:

	Neither 'io' nor 'irq' settings required, the driver uses the resources
	given by the PCI BIOS.

	comx0/boardnum - board number of the SliceCom in the PC (using the 'natural'
	PCI order) as listed in '/proc/pci' or the output of the
	'lspci' command, generally the slots nearer to the motherboard
	PCI driver chips have the lower numbers.

	Default: 0 (the counting starts with 0)

	Though the options below are to be set on a single interface, they apply to the
	whole board. The restriction, to use them on 'UP' interfaces, is because the
	command sequence below could lead to unpredictable results.

	# echo 0 >boardnum
	# echo internal >clock_source
	# echo 1 >boardnum

	The sequence would set the clock source of board 0.

	These settings will persist after all the interfaces are cleared, but are
	cleared when the driver module is unloaded and loaded again.

	comx0/clock_source - source of the transmit clock
	Usage:

	# echo line >/proc/comx/comx0/clock_source
	# echo internal >/proc/comx/comx0/clock_source

	line - The Tx clock is being decoded if the input data stream,
	if no clock seen on the input, then the board will use it's
	own clock generator.

	internal - The Tx clock is supplied by the builtin clock generator.

	Default: line

	Normally, the telecommunication company's end device (the HDSL
	modem) provides the Tx clock, that's why 'line' is the default.

	comx0/framing - Switching CRC4 off/on

	CRC4: 16 PCM frames (The 32 64Kibit channels are multiplexed into a
	PCM frame, nothing to do with HDLC frames) are divided into 2x8
	groups, each group has a 4 bit CRC.

	# echo crc4 >/proc/comx/comx0/framing
	# echo no-crc4 >/proc/comx/comx0/framing

	Default is 'crc4', the Hungarian MATAV lines behave like this.
	The traffic generally passes if this setting on both ends don't match.

	comx0/linecode - Setting the line coding

	# echo hdb3 >/proc/comx/comx0/linecode
	# echo ami >/proc/comx/comx0/linecode

	Default a 'hdb3', MATAV lines use this.

	(AMI coding is rarely used with E1 lines). Frame sync may occur, if
	this setting doesn't match the other end's, but CRC4 and data errors
	will come, which will result in CRC errors on HDLC/SyncPPP level.

	comx0/reg - direct access to the board's MUNICH (reg) and FALC (lbireg)
	comx0/lbireg circuit's registers

	# echo >reg 0x04 0x0 - write 0 to register 4
	# echo >reg 0x104 - write the contents of register 4 with
	printk() to syslog

	WARNING! These are only for development purposes, messing with this will
	result much trouble!

	comx0/loopback - Places a loop to the board's G.703 signals

	# echo none >/proc/comx/comx0/loopback
	# echo local >/proc/comx/comx0/loopback
	# echo remote >/proc/comx/comx0/loopback

	none - normal operation, no loop
	local - the board receives it's own output
	remote - the board sends the received data to the remote side

	Default: none

	-----------------------------------------------------------------

	Interface (channel group in Cisco terms) settings:

	comx0/timeslots - which timeslots belong to the given interface

	Setting:

	# echo '1 5 2 6 7 8' >/proc/comx/comx0/timeslots

	# cat /proc/comx/comx0/timeslots
	1 2 5 6 7 8
	#

	Finding a timeslot:

	# grep ' 4' /proc/comx/comx*/timeslots
	/proc/comx/comx0/timeslots:1 3 4 5 6
	#

	The timeslots can be in any order, '1 2 3' is the same as '1 3 2'.

	The interface has to be DOWN during the setting ('ifconfig comx0
	down'), but the other interfaces could operate normally.

	The driver checks if the assigned timeslots are vacant, if not, then
	the setting won't be applied.

	The timeslot values are treated as decimal numbers, not to misunderstand
	values of 08, 09 form.

	-----------------------------------------------------------------

	Checking the interface and board status:

	- Lines beginning with ' ' (space) belong to the original output, the lines
	which begin with '//' are the comments.

	papaya:~$ cat /proc/comx/comx1/status
	Interface administrative status is UP, modem status is UP, protocol is UP
	Modem status changes: 0, Transmitter status is IDLE, tbusy: 0
	Interface load (input): 978376 / 947808 / 951024 bits/s (5s/5m/15m)
	(output): 978376 / 947848 / 951024 bits/s (5s/5m/15m)
	Debug flags: none
	RX errors: len: 22, overrun: 1, crc: 0, aborts: 0
	buffer overrun: 0, pbuffer overrun: 0
	TX errors: underrun: 0
	Line keepalive (value: 10) status UP [0]

	// The hardware specific part starts here:
	Controller status:
	No alarms

	// Alarm:
	//
	// No alarms - Everything OK
	//
	// LOS - Loss Of Signal - No signal sensed on the input
	// AIS - Alarm Indication Signal - The remote side sends '11111111'-s,
	// it tells, that there's an error condition, or it's not
	// initialised.
	// AUXP - Auxiliary Pattern Indication - 01010101.. received.
	// LFA - Loss of Frame Alignment - no frame sync received.
	// RRA - Receive Remote Alarm - the remote end's OK, but signals error cond.
	// LMFA - Loss of CRC4 Multiframe Alignment - no CRC4 multiframe sync.
	// NMF - No Multiframe alignment Found after 400 msec - no such alarm using
	// no-crc4 or crc4 framing, see below.
	//
	// Other possible error messages:
	//
	// Transmit Line Short - the board felt, that it's output is short-circuited,
	// so it switched the transmission off. (The board can't definitely tell,
	// that it's output is short-circuited.)

	// Chained list of the received packets, for debug purposes:

	Rx ring:
	rafutott: 0
	lastcheck: 50845731, jiffies: 51314281
	base: 017b1858
	rx_desc_ptr: 0
	rx_desc_ptr: 017b1858
	hw_curr_ptr: 017b1858
	06040000 017b1868 017b1898 c016ff00
	06040000 017b1878 017b1e9c c016ff00
	46040000 017b1888 017b24a0 c016ff00
	06040000 017b1858 017b2aa4 c016ff00

	// All the interfaces using the board: comx1, using the 1,2,...16 timeslots,
	// comx2, using timeslot 17, etc.

	Interfaces using this board: (channel-group, interface, timeslots)
	0 comx1: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
	1 comx2: 17
	2 comx3: 18
	3 comx4: 19
	4 comx5: 20
	5 comx6: 21
	6 comx7: 22
	7 comx8: 23
	8 comx9: 24
	9 comx10: 25
	10 comx11: 26
	11 comx12: 27
	12 comx13: 28
	13 comx14: 29
	14 comx15: 30
	15 comx16: 31

	// The number of events handled by the driver during an interrupt cycle:

	Interrupt work histogram:
	hist[ 0]: 0 hist[ 1]: 2 hist[ 2]: 18574 hist[ 3]: 79
	hist[ 4]: 14 hist[ 5]: 1 hist[ 6]: 0 hist[ 7]: 1
	hist[ 8]: 0 hist[ 9]: 7

	// The number of packets to send in the Tx ring, when a new one arrived:

	Tx ring histogram:
	hist[ 0]: 2329 hist[ 1]: 0 hist[ 2]: 0 hist[ 3]: 0

	// The error counters of the E1 interface, according to the RFC2495,
	// (similar to the Cisco "show controllers e1" command's output:
	// http://www.cisco.com/univercd/cc/td/doc/product/software/ios11/rbook/rinterfc.htm#xtocid25669126)

	Data in current interval (91 seconds elapsed):
	9516 Line Code Violations, 65 Path Code Violations, 2 E-Bit Errors
	0 Slip Secs, 2 Fr Loss Secs, 2 Line Err Secs, 0 Degraded Mins
	0 Errored Secs, 0 Bursty Err Secs, 0 Severely Err Secs, 11 Unavail Secs
	Data in Interval 1 (15 minutes):
	0 Line Code Violations, 0 Path Code Violations, 0 E-Bit Errors
	0 Slip Secs, 0 Fr Loss Secs, 0 Line Err Secs, 0 Degraded Mins
	0 Errored Secs, 0 Bursty Err Secs, 0 Severely Err Secs, 0 Unavail Secs
	Data in last 4 intervals (1 hour):
	0 Line Code Violations, 0 Path Code Violations, 0 E-Bit Errors
	0 Slip Secs, 0 Fr Loss Secs, 0 Line Err Secs, 0 Degraded Mins
	0 Errored Secs, 0 Bursty Err Secs, 0 Severely Err Secs, 0 Unavail Secs
	Data in last 96 intervals (24 hours):
	0 Line Code Violations, 0 Path Code Violations, 0 E-Bit Errors
	0 Slip Secs, 0 Fr Loss Secs, 0 Line Err Secs, 0 Degraded Mins
	0 Errored Secs, 0 Bursty Err Secs, 0 Severely Err Secs, 0 Unavail Secs

	-----------------------------------------------------------------

	Some unique options, (may get into the driver later):
	Treat them very carefully, these can cause much trouble!

	modified CRC-4, for improved interworking of CRC-4 and non-CRC-4
	devices: (see page 107 and g706 Annex B)
	lbireg[ 0x1b ] \|= 0x08
	lbireg[ 0x1c ] \|= 0xc0

	- The NMF - 'No Multiframe alignment Found after 400 msec' alarm
	comes into account.

	FALC - the line driver chip.
	local loop - I hear my transmission back.
	remote loop - I echo the remote transmission back.

	Something useful for finding errors:

	- local loop for timeslot 1 in the FALC chip:

	# echo >lbireg 0x1d 0x21

	- Switching the loop off:

	# echo >lbireg 0x1d 0x00