PC World 1999 August/PCWorld_1999-08_cd.bin/doc/HOWTO/Modem-HOWTO

Text File | 1999-05-24 | 153.4 KB | 3,565 lines

  Modem-HOWTO
  David S.Lawyer    <mailto:[e-mail]> Contains much info
  from Serial-HOWTO by Greg Hankins
  v0.03, May 1999

  Help with selecting, connecting, configuring, trouble-shooting, and
  understanding modems for a PC.  See Serial-HOWTO for multiport serial
  boards.
  ______________________________________________________________________

  Table of Contents























































  1. Introduction

     1.1 Copyright, Trademarks, Disclaimer, & Credits
        1.1.1 Copyright
        1.1.2 Trademarks
        1.1.3 Disclaimer
        1.1.4 Credits
     1.2 Future Plans, You Can Help
     1.3 New Versions of this HOWTO
     1.4 What is a Modem ?
     1.5 Quick Install
        1.5.1 External Modem Install
        1.5.2 Internal Modems (on ISA bus)
        1.5.3 All Modems

  2. Modem & Serial Port Basics

     2.1 Modem Converts Digital to Analog (and conversely)
     2.2 What is a Serial Port ?
        2.2.1 Intro to Serial
        2.2.2 Pins and Wires
        2.2.3 Internal Modem Contains Serial Port
     2.3 IO Address & IRQ
     2.4 Interrupts
     2.5 Data Compression (by the Modem)
     2.6 Error Correction
     2.7 Data Flow (Speeds)
     2.8 Flow Control
        2.8.1 Flow Control Explained by an Example
        2.8.2 Symptoms of No Flow Control
        2.8.3 Hardware vs. Software Flow Control
        2.8.4 Modem-to-Modem Flow Control
     2.9 Data Flow Path; Buffers
     2.10 Complex Flow Control Example
     2.11 Modem Commands

  3. Configuring Modems (including the serial port)

     3.1 Configuring Overview
     3.2 Configuring the IRQ and IO-Address of the Serial Port
        3.2.1 Plug-and-Play
        3.2.2 Using a PnP BIOS to I0-IRQ Configure
        3.2.3 External Modem Configure
        3.2.4 If More Than 2 Serial Ports
        3.2.5 Methods for Setting IO Addresses and IRQs
        3.2.6 Avoiding IO Address Conflicts of IBM 8514 Video Board
     3.3 What is the current IO address and IRQ of my Serial Port ?
        3.3.1 What does the device driver think?
        3.3.2 What is set in my serial port hardware?
     3.4 Other Configuring
        3.4.1 Configuring Hardware Flow Control (RTS/CTS)
     3.5 Modem Configuration (excluding serial port)
        3.5.1 AT Commands
        3.5.2 Init Strings: Saving and Recalling
        3.5.3 Other Modem Commands

  4. Modems for a Linux PC

     4.1 External vs. Internal
     4.2 External Modems
        4.2.1 PnP External Modems
        4.2.2 Cabling & Installation
        4.2.3 What the Lights (LED's) Mean
     4.3 Internal Modems
     4.4 Internal Modems You Must Avoid (winmodems, etc.)
     4.5 Which Internal Modems are Best to Avoid ?
        4.5.1 MWave and DSP Modems
        4.5.2 Rockwell (RPI) Drivers
        4.5.3 PCI Modems

  5. Serial Port Devices

     5.1 Serial Port Device Names & Numbers
     5.2 Creating Devices In the /dev directory
     5.3 Link ttySN to /dev/modem ?
     5.4 The cua Device

  6. Interesting Programs You Should Know About

     6.1 What is getty?
        6.1.1 About mgetty
        6.1.2 About getty_ps
        6.1.3 About agetty and mingetty
     6.2 What is Setserial ?
        6.2.1 Intro to Setserial
        6.2.2 Probing
        6.2.3 Boot-time Configuration
        6.2.4 IRQs
     6.3 What is isapnp ?
     6.4 What is wvdialconf ?
     6.5 What is stty ?

  7. Trying Out Your Modem (Dialing Out)

     7.1 Are You Ready to Dial Out ?
     7.2 Dialing Out with Minicom
     7.3 Dialing Out with Kermit

  8. Uugetty for Dial-In (from the old Serial-HOWTO)

     8.1 Installing getty_ps
     8.2 Setting up uugetty
        8.2.1 Modern Modems
        8.2.2 Old slow modems
        8.2.3 Login Banner
     8.3 Customizing uugetty

  9. What Speed Should I Use with My Modem?

     9.1 Speed and Data Compression
     9.2 Where do I Set Speed ?
     9.3 Can't Set a High Enough Speed
     9.4 Speed Table

  10. Communications Programs And Utilities

     10.1 Minicom vs. Kermit
     10.2 Lists of Programs
        10.2.1 Least Popular Dialout
        10.2.2 Most Popular Dialout
        10.2.3 Fax
        10.2.4 Voicemail
        10.2.5 Dial-in (uses getty)
        10.2.6 Other
     10.3 (TT

  11. What Are UARTs?  How Do They Affect Performance?

  12. Troubleshooting

     12.1 Software
     12.2 My Modem is Physically There but Can't be Found
     12.3 "Operation not supported by device" (error message) for ttySx
     12.4 Slow. Text appears on the screen slowly after long delays
     12.5 Uploading (downloading) files is broken/slow
     12.6 For Dial-in I Keep Getting "line NNN of inittab invalid"
     12.7 When I Try To Dial Out, It Says "/dev/ttySN: Device or resource busy"
     12.8 I Keep Getting "Getty respawning too fast: disabled for 5 minutes"
     12.9 My Modem is Hosed after Someone Hangs Up, or uugetty doesn't respawn
     12.10 uugetty Still Doesn't Work

  13. Flash Upgrades

  14. Problems Explained

     14.1 Interrupt Mis-set
     14.2 Interrupt Conflicts

  15. Other Sources of Information

     15.1 Misc
     15.2 Books
     15.3 HOWTOs
     15.4 Usenet newsgroups
     15.5 Web Sites

  16. Appendix A:  How Modems Work (technical) (unfinished)

     16.1 Modulation Details
        16.1.1 Intro to Modulation
        16.1.2 Frequency Modulation
        16.1.3 Amplitude Modulation
        16.1.4 Phase Modulation
        16.1.5 Combination Modulation
        16.1.6 56k Modems (v.90)

  17. Appendix B:  "baud" vs. "bps"

     17.1 A simple example
     17.2 Real examples

  18. Appendix C:  Terminal Server Connection

  19. Appendix D:  Other Types of Modems

     19.1 Digital-to-Digital "Modems"
     19.2 ISDN "Modems"
     19.3 Digital Subscriber Line (DSL)


  ______________________________________________________________________

  1.  Introduction

  This covers conventional modems for PC's, mainly modems on the ISA bus
  (although much of this should also apply to the PCI bus).  For modems
  on the PCMCIA bus see the PCMCIA-HOWTO: PCMCIA serial and modem
  devices.


  1.1.  Copyright, Trademarks, Disclaimer, & Credits

  1.1.1.  Copyright

  Copyright (c) 1998-9 by David S. Lawyer and Greg Hankins.  Please
  freely copy and distribute (sell or give away) this document.  For
  corrections and minor changes contact the maintainer.  Otherwise you
  may create derivative works and distribute them provided you:
  1. Discuss it with the maintainer (if there is one).  2. Put the
  derivative work at the mirrored LDP Internet site (or the like) for
  free downloading.  3. License the work in the spirit of this license
  or use GPL.  4. Give due credit to previous authors and major
  contributors.


  1.1.2.  Trademarks

  If certain words are trademarks, the context should make it clear to
  whom they belong.  For example "MS Windows" (or just "Windows")
  implies that "Windows" belongs to Microsoft (MS).  "Hayes" is a
  trademark of Microcomputer Products Inc.  I use "winmodem" to mean any
  modem which requires MS-Windows and not in the trademark sense.


  1.1.3.  Disclaimer

  Much of the info in this HOWTO was obtained from the previous Serial-
  HOWTO, the Internet, etc. and may be unreliable.  While I haven't
  intentionally tried to mislead you, there are likely a number of
  errors in this document.  Please let me know about them.  Since this
  is free documentation, it should be obvious that neither I nor
  previous authors can be held legally responsible for any errors.


  1.1.4.  Credits

  The following is only a rough approximation of how version 0.0 of this
  document was created:  About 1/3 of the material here was lifted
  directly from Serial-HOWTO v. 1.11 by Greg Hankins.
  <mailto:[e-mail]> (with his permission).  About another 1/3
  was taken from that Serial-HOWTO and revised.  The remaining 1/3 is
  newly created by the author: David S. Lawyer  <mailto:[e-mail]>.


  1.2.  Future Plans, You Can Help

  An explanation of how to set up modems for dial-in is lacking in this
  version but should be included in the next.  Please let me know of any
  errors in facts, opinions, logic, spelling, grammar, clarity, links,
  etc.  But first, if the date is over a months old, check to see that
  you have the latest version.  Please send me any other info that you
  think belongs in this document.  The French "Modems-HOWTO" needs to be
  somehow merged with this document (but I don't know French).


  1.3.  New Versions of this HOWTO

  New versions of this Modem-HOWTO come out every month or so since
  modem situation is rapidly changing (and since I'm still learning).
  Your problem might be solved in the latest version.  It will be
  available to browse and/or download at LDP mirror sites.  For a list
  of such sites see:  <http://metalab.unc.edu/LDP/mirrors.html> If you
  only want to quickly check the date of the latest version you may not
  want to use a mirror site so check out:
  <http://metalab.unc.edu/LDP/HOWTO/Modem-HOWTO.html>.


  1.4.  What is a Modem ?

  A modem is a device that lets one send digital signals over ordinary
  telephone lines not designed for digital signals.  If telephone lines
  were all digital then you wouldn't need a modem.  It permits your
  computer to connect to and communicate with the rest of the world.
  When you use a modem, you normally use a communication program or web
  browser (which includes such a program) to utilize the modem and dial-
  out on a telephone line.  Advanced modem users can set things up so
  that others may phone in to them and use their computer.  This is
  called "dial-in".

  There are two basic types of modems for a PC: external and internal.
  The external sets on your desk outside the PC while the internal is
  not visible since it's inside the PC.  The external modem plugs into a
  connector on the back of the PC known as a "serial port".  The
  internal modem is a card that is inserted inside the computer and has
  an (invisible) serial port built into it.  For a more detailed
  comparison see ``External vs. Internal''.  Thus when you get an
  internal modem, you also get a dedicated serial port (which can only
  be used with the modem and not with anything else such as another
  modem or a printer).  In Linux, the serial ports are named ttyS0,
  ttyS1, etc. (corresponding respectively to COM1, COM2, etc. in
  Dos/Windows).

  The serial port is not to be confused with the "Universal Serial Bus"
  (USB) which uses a special modular connector and may be used with
  modems in the future.  See ``Modem & Serial Port Basics'' for more
  details on modems and serial ports.


  1.5.  Quick Install

  1.5.1.  External Modem Install

  With a straight-thru or modem cable, connect the modem to an unused
  serial port on the PC.  Make sure you know the name of the serial
  port: COM1 is ttyS0, COM2 is ttyS1.  You may need to check the BIOS
  setup menu to determine this.  Plug in the power cord to provide power
  to the modem.  See ``All Modems'' for further instructions.


  1.5.2.  Internal Modems (on ISA bus)

  (See ``PCI Modems'' for the PCI bus) If the modem says it will only
  work under MS Windows, you are out of luck.  If you already have 2
  serial ports, make this the 3rd serial port (ttyS2 = COM3).  Find an
  unused IRQ number to use.  An unused one is often IRQ5 for the 2nd
  parallel port or a sound card (which you may not have).  Then set the
  jumpers (or the like) on the internal modem to the unused IRQ and IO
  address 3E8 (ttyS2) .

  "Or the like" (in the previous sentence) may be a bit tricky.  If the
  modem is a Plug and Play (PnP) for the ISA bus, the equivalent
  probably can be done using the "isapnp" program which comes with
  "isapnptools".  See "man isapnp" or the FAQ for it.  See also "Plug-
  and-Play-HOWTO.  With a PnP-BIOS you may be able to tell the CMOS
  setup menu that you don't have a PnP OS and then the BIOS may set a
  suitable IRQ and IO address in the modem card.  I once thought that if
  you configured the modem under Windows 9x then Windows would be nice
  and put that configuration info (IRQ and IO address) into the BIOS's
  flash memory so that it could be used by the BIOS each time you start
  Linux.  But it may not do this so try it only as a last resort.  You
  might try using ICU utility under Windows..  There may even be a way
  to disable PnP using software (under Windows) that came with the
  modem.

  Finally you must also find the file where "setserial" is run and add a
  line something like: "setserial /dev/ttyS2 irq5".  See ``All Modems''
  for further instructions.



  1.5.3.  All Modems

  Plug the modem into a telephone line.  Then start up a communication
  program such as minicom and go to the configuration menu for the
  serial port.  Assign it a high baud rate a few times higher than the
  bit rate of your modem.  See ``Speed Table'' for the "best" speeds to
  use.  Tell it the full name of your serial port such as /dev/ttyS1.
  Set hardware flow control (RTS/CTS).  Now you need to save these
  settings and exit minicom.  Then start minicom again, type AT to see
  if your modem is there and responds with OK.  Then go to the dial
  directory (or menu) and dial a number.


  2.  Modem & Serial Port Basics

  You don't have to understand the basics to use and install a modem.
  But understanding it may help to determine what is wrong if you run
  into problems.  After reading this section, if you want to understand
  it even better you may want to see ``How Modems Work'' in this
  document (not yet complete).  A future version of Serial-HOWTO
  (expected by Feb. 1999).  should cover more on the serial port itself.


  2.1.  Modem Converts Digital to Analog (and conversely)

  Most all telephone main lines are digital already but the lines
  leading to your house (or business) are usually analog which means
  that they were designed to transmit a voltage wave which is an exact
  replica of the sound wave coming out of your mouth.  Such a voltage
  wave is called "analog".  If viewed on an oscilloscope it looks like a
  sine wave of varying frequency and amplitude.  A digital signal is
  like a square wave.  For example 3 v (volts) might be a 1-bit and 0 v
  could be a 0-bit.  For most serial ports (used by external modems) +12
  v is a 0-bit and -12 v is a 1-bit (some are + or - 5 v).

  To send data from your computer over the phone line, the modem takes
  the digital signal from your computer and converts it to "analog".  It
  does this by both creating an analog sine wave and then "MODulating"
  it.  Since the result still represents digital data, it could also be
  called a digital signal instead of analog.  But it looks something
  like an analog signal and almost everyone calls it analog.  At the
  other end of the phone line another modem "DEModulates" this signal
  and the pure digital signal is recovered.  Put together the "mod" and
  "dem" parts of the two words above and you get "modem" (if you drop
  one of the two d's).  A "modem" is thus a MODulator-DEModulator.  Just
  what modulation is may be found in the section ``Modulation Details''.


  2.2.  What is a Serial Port ?

  2.2.1.  Intro to Serial

  Since modems have a serial port between them and the computer, it's
  necessary to understand the serial port as well as the modem.  The
  serial port is an IO (Input/Output) device.  Most PC's have one or two
  serial ports.  Each has a 9-pin connector (sometimes 25-pin) on the
  back of the computer.  Computer programs can send data (bytes) to the
  transmit pin (output) and receive bytes from the receive pin (input).
  The other pins are for control purposes and ground.

  The serial port is much more than just a connector.  It converts the
  data from parallel to serial and changes the electrical representation
  of the data.  Inside the computer, data bits flow in parallel (using
  many wires at the same time).  Serial flow is a stream of bits over a
  single wire (such as on the transmit or receive pin of the serial
  connector).  For the serial port to create such a flow, it must
  convert data from parallel (inside the computer) to serial on the
  transmit pin (and conversely).

  Most of the electronics of the serial port is found in a computer chip
  (or a section of a chip) known as a UART.  For more details on UARTs
  see the section ``What Are UARTs?  How Do They Affect Performance?''.
  But you may want to finish this section first so that you will
  hopefully understand how the UART fits into the overall scheme of
  things.


  2.2.2.  Pins and Wires

  Old PC's used 25 pin connectors but only about 9 pins were actually
  used, so today most connectors are only 9-pin.  Each of the 9 pins
  connects to a wire.  Besides the two wires used for transmitting and
  receiving data, another pin (wire) is signal ground.  The voltage on
  any wire is measured with respect to this ground.  There are still
  more wires which are for control purposes (signalling) only and not
  for sending bytes.  All of these signals could have been sent on a
  single wire, but instead, there is a separate dedicated wire for every
  type of signal.  Some (or all) these control wires are called "modem
  control lines".  These control wires are either in the asserted state
  (on) of +12 volts or in the negated state (off) of -12 volts.  There
  is a wire to signal the computer to stop sending bytes to the modem.
  Conversely, another wire signals the modem to stop sending bytes to
  the computer.  Other wires may tell the modem to hang up the telephone
  line or tell the computer that a connection has been made or that the
  telephone line is ringing (someone is attempting to call in).


  2.2.3.  Internal Modem Contains Serial Port

  For an internal modem there is no 9-pin connector but the behavior is
  exactly as if the above mentioned cable wires existed.  Instead of a a
  12 volt signal in a wire giving the state of a modem control line, the
  internal modem may just use a status bit in its own memory (a
  register) to determine the state of this non-existent "wire".  The
  internal modem's serial port looks just like a real serial port to the
  computer.  It even includes the speed limits that one may set at
  ordinary serial ports such as 115200 bits/sec.  Unfortunately today,
  many internal modems don't work exactly this way but instead use MS
  Windows software to help do their job and will not work under Linux.
  See ``Avoid: winmodems''.


  2.3.  IO Address & IRQ

  Since the computer needs to communicate with each serial port, the
  operating system (OS) must know that the serial port exists, where it
  is (its IO address) and what wire (IRQ number) the serial port is to
  use to request service from the computer's CPU.  Thus every serial
  port device must store in its non-volatile memory both its IO address
  and its Interrupt ReQuest number: IRQ.  The IRQ determines what wire
  is used to request service using interrupt signals.  See
  ``Interrupts''.

  The serial ports are labeled ttyS0, ttyS1, etc. (corresponding to
  COM1, COM2, etc. in DOS).  There is also an obsolete set of names:
  cua0, cua1, etc which are almost the same as ttyS0, ttyS1, etc.  See
  ``The cua Device''.  Which one of these names refers to certain
  physical serial port is determined (in part) by the IO address stored
  inside the hardware chip of the physical port.  This mapping of names
  (such as ttyS1) to IO addresses and IRQ's may be set by the
  "setserial" command.  ``What is Setserial''.  This does not set the IO
  address and IRQ on the hardware itself (which is set by jumpers or by
  plug-and-play).


  2.4.  Interrupts

  Bytes come in over the phone line to the modem, are converted from
  analog to digital by the modem and passed along to the serial port on
  their way to their destination inside your computer.  When the serial
  port gets say 8 bytes from the modem (may be set to 1, 4, 8, or 14) it
  signals the CPU to fetch them by sending an electrical signal known as
  an interrupt on a dedicated conductor.  Old serial ports would always
  send an interrupt for every byte received.

  Each interrupt conductor (inside the computer) has a number (IRQ) and
  the serial port must know which conductor to use to signal on.  For
  example, ttyS0 normally uses IRQ number 4 known as IRQ4 (or IRQ 4).  A
  list of them and more will be found in "man setserial" (search for
  "Configuring Serial Ports").  Interrupts are issued whenever the
  serial port needs to get the CPU's attention.  It's important to do
  this in a timely manner since the buffer inside the serial port can
  hold only 16 (1 in old modems) incoming bytes.  If the CPU fails to
  remove such received bytes promptly, then there will not be any space
  left for any more incoming bytes and the small buffer may overflow
  (overrun).  Bytes will be lost.  For an external modem, there is no
  way to stop the flow rapidly enough to prevent this.  For an internal
  modem the 16-byte buffer is on the same card and a good modem will not
  write to it if it's full.  Thus a good internal modem will not overrun
  the 16-byte buffers and this is one advantage of an internal modem.

  Interrupts are also issued when the serial port has just sent out all
  16 of it bytes from its small transmit buffer to the modem.  It then
  has space for 16 more outgoing bytes.  The interrupt is to notify the
  CPU of that fact so that it may put more bytes in the small transmit
  buffer to be transmitted.  Also, when a modem control line changes
  state an interrupt is issued.

  The buffers mentioned above are all hardware buffers.  The serial port
  also has large buffers in main memory.  This will be explained later

  Interrupts convey a lot of information but only indirectly.  The
  interrupt itself just tells a chip called the interrupt controller
  that a certain serial port needs attention.  The interrupt controller
  then signals the CPU.  The CPU runs a special program to service the
  serial port.  That program is called an interrupt service routine
  (part of the serial driver software).  It tries to find out what has
  happened at the serial port and then deals with the problem such a
  transferring bytes from (or to) the serial port's hardware buffer.
  This program can easily find out what has happened since the serial
  port has registers at IO addresses known to the the serial driver
  software.  These registers contain status information about the serial
  port.  The software reads these registers and by inspecting the
  contents, finds out what has happened and takes appropriate action.


  2.5.  Data Compression (by the Modem)

  Before continuing with the basics of the serial port, one needs to
  understand about something done by the modem: data compression.  In
  some cases this task is actually done by software run on the
  computer's CPU but unfortunately at present, such software only works
  for MS Windows.  The discussion here will be for the case where the
  modem itself does the compression since this is what must happen in
  order for the modem to work under Linux.

  In order to send data faster over the phone line, one may compress
  (encode it) using a custom encoding scheme which itself depends on the
  data.  The encoded data is smaller than the original (less bytes) and
  can be sent over the Internet in less time.  This process is called
  "data compression".

  If you download files from the Internet, they are likely already
  compressed and it is not feasible for the modem to try to compress
  them further.  Your modem may sense that what is passing thru has
  already been compressed and refrain from trying a compress it any
  more.  If you are receiving data which has been compressed by the
  other modem, your modem will decompress it and create many more bytes
  than were sent over the phone line.  Thus the flow of data from your
  modem into your computer will be higher than the flow over the phone
  line to you.  The ratio of this flow is called the compression ratio.
  Compression ratios as high as 4 are possible, but not very likely.


  2.6.  Error Correction

  Similar to data compression, modems may be set to do error correction.
  While there is some overhead cost involved which slows down the
  byte/sec flow rate, the fact that error correction strips off start
  and stop bits actually increases the data byte/sec flow rate.

  For the serial port's interface with the external world, each 8-bit
  byte has 2 extra bits added to it: a start-bit and a stop-bit.
  Without error correction, these extra stop and stop bits usually go
  right thru the modem and out over the phone lines.  But when error
  correction is enabled, these extra bits are stripped off and the 8-bit
  bytes are put into packets.  This is more efficient and results in
  higher byte/sec flow in spite of the fact that there are a few more
  bytes added for packet headers and error correction purposes.


  2.7.  Data Flow (Speeds)

  Data (bytes representing letters, pictures, etc.) flows from your
  computer to your modem and then out on the telephone line (and
  conversely).  Flow rates (such as 56k (56000) bits/sec) are
  (incorrectly) called "speed".  But almost everyone says "speed"
  instead of "flow rate".  If there were no data compression the flow
  rate from the computer to the modem would be about the same as the
  flow rate over the telephone line.

  Actually there are two different speeds to consider at your end of the
  phone line:


  ╖  The speed on the phone line itself (DCE speed) modem-to-modem

  ╖  The speed from your computer's serial port to your modem (DTE
     speed)

  When you dial out and connect to another modem on the other end of the
  phone line, your modem often sends you a message like "CONNECT 28800"
  or "CONNECT 115200".  What do these mean?  Well, its either the DCE
  speed or the DTE speed. If it's higher than the advertised modem speed
  it must be the DTE modem-to-computer speed.  This is the case for the
  115200 speed shown above.  The 28800 must be a DCE (modem-to-modem)
  speed since the serial port has no such speed.  One may configure the
  modem to report either speed.  Some modems report both speeds and
  report the modem-to-modem speed as (for example): CARRIER 28800.

  If you have an internal modem you would not expect that there would be
  any speed limit on the DTE speed from your modem to your computer
  since you modem is inside your computer and is almost part of your
  computer.  But there is since the modem contains a dedicated serial
  port within it.

  It's important to understand that the average speed is often less than
  the specified speed, especially on the short DTE computer-to-modem
  line.  Waits (or idle time) result in a lower average speed.  These
  waits may include long waits of perhaps a second due to ``Flow
  Control''.  At the other extreme there may be very short waits (idle
  time) of several micro-seconds separating the end of one byte and the
  start of the next byte.  In addition, modems will fallback to lower
  speeds if the telephone line conditions are less than pristine.

  For a discussion of what DTE speed is best to use see section ``What
  Speed Should I Use''.


  2.8.  Flow Control

  Flow control means the ability to stop the flow of bytes in a wire.
  It also includes provisions to restart the flow without any loss of
  bytes.  Flow control is needed for modems to allow a jump in flow
  rates.


  2.8.1.  Flow Control Explained by an Example

  For example, consider the case where your 33.6k modem is not doing any
  data compression or error correction.  You have set the serial port
  speed to 115,200 bits/sec (bps).  You are sending data from your
  computer to the phone line.  Then the flow from the your computer to
  your modem is at 115.2k bps.  However the flow from your modem out the
  phone line is at best only 33.6k bps.  Since a faster flow (115.2k) is
  going into your modem than is coming out of it, the modem is storing
  the excess flow (115.2k -33.6k = 81.6k) in one of its buffers.  This
  buffer would eventually overrun (run out of storage space) unless the
  115.2k flow is stopped.

  But now flow control comes to the rescue.  When the modem's buffer is
  almost full, the modem sends a stop signal to the serial port.  The
  serial port passes on the stop signal to the device driver and the
  115.2k bps flow is halted.  Then the modem continues to send out data
  at 33.6k bps drawing on the data it previous accumulated in its
  buffer.  Since nothing is coming into the buffer, the level of bytes
  in it starts to drop.  When almost no bytes are left in the buffer,
  the modem sends a start signal to the serial port and the 115.2k flow
  from the computer to the modem resumes.  In effect, flow control
  creates an average flow rate (in this case 33.6k) which is
  significantly less than the "on" flow rate of 115.2k bps.  This is
  "start-stop" flow control.

  The above is an example of flow control for flow from the computer to
  the modem , but there is also flow control which is used for the
  opposite direction of flow: from a modem to a computer.  You don't
  often need it in this direction (for flow from a modem to PC), but a
  for complex example of a case where it's needed see ``Complex Flow
  Control Example''.  But if you don't have a high enough speed set
  between the modem and the computer (serial port speed) then you do
  need to slow down the flow from the modem to the PC.  To do this you
  must stop the incoming flow of bytes over the telephone line.   Your
  modem must tell the other modem to stop sending.  See ``Modem-to-Modem
  Flow Control''   More details on flow control may eventually be put
  into the Serial-HOWTO.





  2.8.2.  Symptoms of No Flow Control

  Understanding flow-control theory can be of practical use.  For
  example I used my modem to access the Internet and it seemed to work
  fine.  But after a few months I tried to send long files from my PC to
  an ISP and a huge amount of retries and errors resulted (but
  eventually Kermit could send a long file after many retries).
  Receiving in the other direction (from my ISP to me) worked fine.  The
  problem turned out to be a hardware defect in my modem that had
  resulted in disabling flow control.  My modem's buffer was overflowing
  (overrunning) on long outgoing files since no "stop" signal was ever
  sent to the computer to halt sending to the modem.  There was no
  problem in the direction from the modem to my computer since the
  capacity (say 115.2k) was always higher than the flow over the
  telephone line.  The fix was to enable flow control by putting an
  enable-flow-control command for the modem last in the init string.


  2.8.3.  Hardware vs. Software Flow Control

  For modems, it's best to use "hardware" flow control that uses two
  dedicated "modem control" wires to send the "stop" and "start"
  signals.  Software flow control uses the main receive and transmit
  wires to send the start and stop signals.  It uses the ASCII control
  characters DC1 (start) and DC3 (stop) for this purpose.  They are just
  inserted into the regular stream of data.  Software flow control is
  not only slower in reacting but also does not allow the sending of
  binary data thru the modem unless special precautions are taken.
  Since binary data will likely contain DC1 and DC3, special means must
  be taken to distinguish between a DC3 that means a flow control stop
  and a DC3 that is part of the binary code.  Likewise for DC1.  To get
  software flow control to work for binary data requires both modem
  (hardware) and software support.


  2.8.4.  Modem-to-Modem Flow Control

  This is the flow control of the data sent over the telephone lines
  between two modems.  Practically speaking, it only exists when you
  have error correction set.  Actually, even without error correction
  it's possible to enable software flow control between modems but it
  may interfere with sending binary data so it's not often used.


  2.9.  Data Flow Path; Buffers

  Although much has been explained about this including flow control, a
  pair of 16-byte serial port buffers (in the hardware), and a pair of
  larger buffers inside the modem, there is still another pair of
  buffers.  These are large buffers (perhaps 8k) in main memory also
  known as serial port buffers.  When an application program sends bytes
  to the serial port (and modem), they first get stashed in the the
  transmit serial port buffer in main memory.  The pair consists of both
  this transmit buffer and a receive buffer for the opposite direction
  of byte-flow.

  The serial device driver takes out say 16 bytes from this transmit
  buffer, one byte at a time and puts them into the 16-byte transmit
  buffer in the serial hardware for transmission.  Once in that transmit
  buffer, there is no way to stop them from being transmitted.  They are
  then transmitted to the modem which also has a fair sized (say 1k)
  buffer.  When the device driver (on orders from flow control) stops
  the flow of outgoing bytes from the computer, what it actually stops
  is the flow of outgoing bytes from the large transmit buffer in main
  memory.  Even after this has happened and the flow to the modem has
  stopped, an application program may keep sending bytes to the 8k
  transmit buffer until it becomes fill.  When it gets fill, the
  application program can't send any more bytes to it (a "write"
  statement in a C_program blocks) and the application program
  temporarily stops running and waits until some buffer space becomes
  available.  Thus a flow control "stop" is ultimately able to stop the
  program that is sending the bytes.  Even though this program stops,
  the computer does not necessarily stop computing.  It may switch to
  running other processes while it's waiting at a flow control stop.
  The above was a little oversimplified since there is another
  alternative of having the application program itself do something else
  while it is waiting to "write".


  2.10.  Complex Flow Control Example

  For many situations, there is a transmit path involving several links,
  each with its own flow control.  For example, I type at a text-
  terminal connected to a PC.  Inside the PC is a modem to access a BBS.
  For this I use the application program "minicom" which deals with 2
  serial ports: one connected to the modem and another connected to the
  text-terminal.  What I type at the text terminal goes into the first
  serial port to minicom, then from minicom out the second serial port
  to the modem, and then onto the telephone line to the BBS.  The text-
  terminal has a limit to the speed at which bytes can be displayed on
  its screen and issues a flow control "stop" from time to time to slow
  down the flow.  What happens when such a "stop" is issued?  Let's
  consider a case where the "stop" is long enough to get thru to the BBS
  and stop the program at the BBS which is sending out the bytes.

  Let's trace out the flow of this "stop" (which may be "hardware" on
  some links and "software" on others).  First, suppose I'm "capturing"
  a long file from the BBS which is being sent simultaneously to both my
  text-terminal and a to file on my hard-disk.  The bytes are coming in
  faster than the terminal can handle them so it sends a "stop" out its
  serial port to the first serial port on my PC.  The device driver
  detects it and stops sending bytes from the 8k serial buffer (in main
  memory) to the terminal.  Now minicom still keeps sending out bytes
  for the terminal into this 8k buffer.

  When this 8k transmit buffer (on the first serial port) is full,
  minicom must stop writing to it.  Minicom stops and waits.  But this
  also causes minicom to stop reading from the 8k receive buffer on the
  2nd serial port connected to the modem.  Flow from the modem continues
  until this 8k buffer too fills up and sends a different "stop" to the
  modem.  Now the modem's buffer ceases to send to the serial port and
  also fills up.  The modem (assuming error correction is enabled) sends
  a "stop signal" to the other modem at the BBS.  This modem stops
  sending bytes out of its buffer and when its buffer gets fill, another
  stop signal is sent to the serial port of the BBS.  At the BBS, the
  8-k (or whatever) buffer fills up and the program at the BBS can't
  write to it anymore and thus temporarily halts.

  Thus a stop signal from a text terminal has halted a programs on a BBS
  computer.  What a Rube Goldberg (complex) sequence of events!  Note
  that the stop signal passed thru 4 serial ports, 2 modems, and one
  application program (minicom).  This counts the serial port attached
  to the terminal.  Each serial port has 2 buffers (in one direction of
  flow): the 8k one and the hardware 16-byte one.  The application
  program may have a buffer in main memory as specified in its C-code.
  This adds up to 11 different buffers the data is passing thru.  Note
  that the small serial hardware buffers do not participate directly in
  flow control.

  If the terminal speed limitation is the bottleneck in the flow from
  the BBS to the terminal, then its flow control "stop" is actually
  stopping the program that is sending from the BBS as explained above.
  But you may ask, how can a "stop" last so long that 11 buffers (some
  of them large) all get filled up?   It can actually happen this way if
  all the buffers were near their upper limits when the terminal sent
  out the "stop".

  But if you were to run a simulation on it you would discover that it's
  usually more complicated than this.  At an instant of time some links
  are flowing and others are stopped (due to flow control).  A "stop"
  from the terminal seldom propagates back to the BBS neatly as
  described above.  It may take a few "stops" from the terminal to
  result in a "stop" at the BBS.  To understand what is going on you
  really need to observe a simulation which can be done for a simple
  case with coins on a table.  Use only a few buffers and set the upper
  level for each buffer at only a few coins.

  Does one really need to understand all this?  Well, understanding this
  explained to me why capturing text from a BBS was loosing text.  The
  situation was exactly the above example but modem-to-modem flow
  control was disabled.  Chunks of captured text that were supposed to
  get to the hard-disk never got there because of an overflow (overrun)
  at the modem buffer due to flow control "stops" from the terminal.
  Even though the BBS had a flow path to the hard-disk without
  bottlenecks, the overflow due to the terminal happened on this path
  and chunks of text were lost.


  2.11.  Modem Commands

  Commands to the modem are sent to it from the communication software
  over the same conductor as used to send data.  The commands are short
  ASCII strings.  Examples are "AT&K3" for enabling hardware flow
  control (RTS/CTS) between your computer and modem; and "ATDT5393401
  for Dialing the number 5393401.  Note all commands are prefaced by
  "AT".   Some commands such as enabling flow control help configure the
  modem.  Other commands such as dialing a number actually do something.
  There are about a hundred or so different possible commands.  When
  your communication software starts running, it first sends an "init"
  string of commands to the modem to configure it.  All commands are
  sent on the ordinary data line before the modem dials (or receives a
  call).

  Once the modem is connected to another modem (on-line mode),
  everything that is sent from your computer to your modem goes directly
  to the other modem and is not interpreted by the modem as a command.
  There is a way to "escape" from this mode of operation and go back to
  command mode where everything sent to the modem will be interpreted as
  a command.  The computer just sends "+++" with a specified time
  spacing before and after it.  If this time spacing is correct, the
  modem reverts to command mode.  Another way to do this is by a signal
  on a certain modem control line.

  There are a number of lists of modem commands on the Internet.  The
  section ``Web Sites'' has links to a couple of such web-sites.
  Different models and brands of modems do not use exactly the same set
  of such commands.  So what works for one modem might not work for
  another.  Some common command (not guaranteed to work on all modems)
  are listed in this HOWTO in the section ``Modem Configuration''


  3.  Configuring Modems (including the serial port)

  If you want to use a modem only for MS Windows/Dos, then you can just
  install almost any modem and it will work OK.  With a Linux PC it's
  not usually this easy unless you use an external modem.  All external
  modems should work OK (even if they are labeled "Plug and Play")  But
  most new internal modems are Plug-and-Play (PnP) and have PnP serial
  ports.  You may need to use the Linux "isapnp" program to configure
  these PnP serial ports.  See the Plug-and-Play-HOWTO for more
  information.


  3.1.  Configuring Overview

  Since each modem has an associated serial port there are two parts to
  configuring a modem:


  ╖  Configuring the modem itself: Done by the communication program

  ╖  Configuring the modem's serial port: Done only partly by the
     communication program

  Most of the above configuring (but not necessarily most of the effort)
  is done by the communication program that you use with the modem such
  as minicom or seyon, or by the PPP part of your Web Browser.  If you
  use the modem for dial-in, then the getty program which you use to
  present outsiders with a login-prompt, will help configure.  Thus to
  configure the modem (and much of the serial port) you need to
  configure the communication program (or PPP or getty).  The
  documentation for these programs and/or the PPP-HOWTO should be
  helpful.

  But note that not all of the configuration of the serial port is done
  by the communication program (or getty).  The remaining configuring is
  simple to state (but sometimes difficult to do).  It mainly consists
  of setting the IO address of the port and its IRQ number.  In fact,
  plug-and-play could set these without you doing a thing.  But there's
  a serious problem: Linux (as of early 1999) doesn't support plug-and-
  play very well.  This may create a difficult problem for you.


  3.2.  Configuring the IRQ and IO-Address of the Serial Port

  Prior to firing up (and configuring) your communication program, you
  must do the configuring that your communication program can't do.
  Oversimplified, this consists only of identifying the serial port on
  which the modem resides:  Is it ttyS2 (=COM3) or ttyS1 (=COM2) etc.?
  If you know the answer for sure (and there are no IRQ conflicts), then
  there is nothing to do and you may start your communication program .

  Otherwise, you must establish the serial port identification and
  assign it an IRQ number.  This is done by putting two values (an IRQ
  number and IO address) into two places:

  1. A memory register of the serial port hardware itself

  2. the device driver (often by running "setserial" at boot-time)

  Both of the above are supposed to by done by a plug-and-play operating
  system (OS).  For item 1. setting these numbers in a modem card (or
  for the serial port in the case of an external modem) was formerly
  done by jumpers.  Today it's supposed to be done at boot-time by plug-
  and-play (PnP).  A major difference between jumpers and PnP is that
  the jumper setting remain the same after you turn off the PC.  For PnP
  the equivalent settings are not remembered by the card and the
  settings must be supplied anew (via PnP) to the card each time the PC
  is powered up.

  For item 2. if you accept the default settings there is no need to use
  setserial.  But if you use "setserial" the IRQ and IO address you tell
  it must be exactly the same as what is set inside the serial port
  hardware (or will be set via PnP).  We might call all of this "io-irq"
  configuring for short.  In the Wintel world, the IO address and IRQ
  are called "resources" and we are thus configuring certain resources.


  3.2.1.  Plug-and-Play

  Plug-and-Play was designed to automate this io-irq configuring, but
  for Linux at present, it has made life more complicated.  The standard
  kernels for Linux don't support plug-and-play very well.  If you use a
  patch to the Linux kernel to covert it to a plug-and-play operating
  system, then all of the above should be handled automatically by the
  OS.  But when you want to use this to automate configuring devices
  other that the serial port, you may find that you'll still have to
  configure the drivers manually since many Linux drivers are not
  written to support a Linux PnP OS.  If you use isapnptools or the BIOS
  for configuring plug-and-play this will only put the two values into
  the registers of the serial port section of the modem card and you
  will likely still need to set up setserial.  None of this is easy or
  very well documented as of early 1999.  See Plug-and-Play-HOWTO and
  the isapnptools FAQ.


  3.2.2.  Using a PnP BIOS to I0-IRQ Configure

  While the explanation of how to use a PnP OS or isapnp for io-irq
  configuring should come with such software, this is not the case if
  you want to let a PnP BIOS do such configuring.  Not all PnP BIOS can
  do this.  The BIOS usually has a CMOS menu for setting up the first
  two serial ports.  There is often little to choose from.  Unless
  otherwise indicated in menus, these first two ports get set at the
  standard IO addresses and IRQs.  See ``Serial Port Device Names &
  Numbers''

  Whether you like it or not, when you start up a PC a PnP BIOS starts
  to do PnP (io-irq) configuring of hardware devices.  It may do the job
  partially and turn the rest over to a PnP OS (which you probably don't
  have) or if thinks you don't have a PnP OS it may fully configure all
  the PnP devices but not configure the device drivers.  This is what
  you want but it's not always easy to figure out exactly what the PnP
  BIOS has done.

  If you tell the BIOS that you don't have a PnP OS, then the PnP BIOS
  should do the configuring of all PnP serial ports --not just the first
  two.  If you have MS Windows9x on the same PC, the BIOS might ?? have
  saved the io-irq configuration used for MS Windows in its non-volatile
  memory.  If you're now reading the latest version of this HOWTO let me
  know if Windows puts it there.  If so the BIOS will use this same
  configuration for Linux.  In this case, if you can find out how MS
  Windows has set up io-irq then it should be the same under Linux.

  If you add a new PnP device, the BIOS should change its PnP
  configuration to accommodate it.  It could even change the io-irq of
  existing devices if required to avoid any conflicts.  For this
  purpose, it keeps a list of non-PnP devices provided that you have
  told the BIOS how these non-PnP devices are io-irq configured.  One
  way to tell the BIOS this is by running a program called ICU under
  DOS/Windows.

  But how do you find out what the BIOS has done so that you set up the
  device drivers with this info?  The BIOS itself may provide some info,
  either in it's setup menus of via messages on the screen when you turn
  on your computer.  See ``What is set in my serial port hardware?''




  3.2.3.  External Modem Configure

  If you use an external modem and plug it into say the ttyS1 connector,
  then there's (usually) no io-irq configuring to do since ttyS1 has
  likely already been io-irq configured.  Your CMOS BIOS may have a menu
  to do this for COM1 and COM2.


  3.2.4.  If More Than 2 Serial Ports

  Normally you don't need to configure the first two serial ports that
  your computer comes with since the default configuration works fine.
  The Linux distribution you get probably has these defaults built into
  it so you have nothing to do.  Everything is different when you want
  to add a third serial port.  The io-irq configuring of it uses the
  same principles as for the first two serial ports but this time
  (unless you use a plug-and-play operating system) you have to do it
  yourself.  To do this (as already mentioned) you'll need to run
  setserial: ``What is Setserial'' and also set the io-irq data into the
  modem card by PnP methods (or by physical jumpers or switches on old
  cards).


  3.2.5.  Methods for Setting IO Addresses and IRQs

  Here's a summary of what was just discussed (and more).  The modem is
  entirely configured by sending commands to it from the computer (via
  the communication program) as is much of the serial port (such as the
  baud rate and hardware flow control).  For more details on the PnP
  options see Plug-and-Play-HOWTO.  You configure the io-irq of the
  serial port by:


  1. Setting the IRQ number and the IO address in the port hardware (io-
     irq) by doing one of the following:

  ╖  Doing nothing for an external modem where the existing io-irq is OK

  ╖  Using a PnP BIOS CMOS setup menu (usually only for external modems
     on ttyS0 (Com1) and ttyS1 (Com2))

  ╖  Doing nothing if you have both a PnP internal modem and a PnP Linux
     operating system.

  ╖  Setting jumpers (if they exist) on an internal modem, old serial
     card, or old motherboard

  ╖  Using isapnp for a PnP internal modem (non-PCI)

  ╖  Letting a PnP BIOS automatically configure a PnP internal modem See
     ``Using a PnP BIOS to I0-IRQ Configure''

  ╖  Using a setup program which comes with the modem (run under MS
     Windows) that disables PnP

  2. Run "setserial": Assigns say ttyS2 to an IO address and IRQ number.
     Except that you probably don't need to run it if you have only 1 or
     2 serial ports total or if you use a PnP operating system.


  3.2.6.  Avoiding IO Address Conflicts of IBM 8514 Video Board

  The IO address of the IBM 8514 video board (and it's clones) is
  allegedly 0x2e8, the same as the IO address of ttyS3.  That is bad
  news if you try to use ttyS3 at this IO address.

  3.3.  What is the current IO address and IRQ of my Serial Port ?

  There are two answers to this question: 1. What the device driver
  thinks has been set (This is what setserial "sets").  2. What is
  actually set in the hardware.  They both should be the same.  If
  you're having trouble (including communication programs that can't
  communicate) it may mean that these two items are not set the same.
  In other words, this means that the driver has incorrect info on the
  serial port.  If the driver has the wrong IO address it will try to
  send data to a non-existing serial port --or even worse, to an actual
  device that is not a serial port.  If it has the wrong IRQ it will not
  get interrupt service requests from the serial port, resulting in the
  possible overflow of the serial port's buffer and in very slow
  response (due to fallback to very slow "polling" methods instead of
  interrupts).  If it has the wrong model of UART there is also apt to
  be trouble.


  3.3.1.  What does the device driver think?

  How do you insure that the device driver has the correct info?  Well,
  if everything seems OK then there's no need to look into this.  But
  otherwise, it's easy to see what the device driver thinks.  One way is
  to just type "setserial -g /dev/ttyS*".  At boot-time, a message on
  the console should show this.  You may look at /proc/ioports but it
  only shows the same IO address which setserial has "set" but are not
  necessarily the way its actually set in the hardware.  To see the IRQs
  used by currently running processes (that have devices open) look at
  /proc/interrupts.  It also shows how many actual interrupts have been
  issued (often thousands).  None of the above tests show what is
  actually set in the device.  But if everything works fine, the devices
  are likely actually set up that way in the hardware.


  3.3.2.  What is set in my serial port hardware?

  How do you find out what IO address and IRQ are actually set in the
  device hardware?  Perhaps the BIOS messages will tell you some info
  before Linux starts booting.  Use the shift-PageUp key to step back
  thru the boot-time messages and look at the very first ones which are
  from the BIOS.  This is how it was before Linux started.  Setserial
  can't change it but isapnp can.

  For Plug-and-Play (PnP) modems (or PnP serial ports) on the ISA bus
  one may try the pnpdump program (part of isapnptools).  If you use the
  --dumpregs option then it should tell you the actual IO address and
  IRQ set in the modem card.  The address it "trys" is not the device's
  IO address, but a special read-port used only for PnP purposes.

  For an older card, the jumper setting may tell you how its set.  If
  the modem is PCI look at /proc/pci or /proc/bus/pci/devices.  One
  crude method is try probing with setserial using the "autoconfig"
  option.  You'll need to guess the addresses to probe at.  See ``What
  is Setserial''.

  Checking on how it's configured under MS-Windows may not be of much
  help.  Windows stores its configuration info in its Registry which is
  not used by Linux.  It may supply the BIOS's non-volatile memory with
  some info but it may not be kept in sync with the current Window
  configuration in the Registry ??  If you let a PnP BIOS automatically
  do the configuring when you start Linux (and have told the BIOS that
  you don't have a PnP operating system when running Linux) then it
  should use whatever configuration is in the BIOS's non-volatile
  memory.


  3.4.  Other Configuring

  3.4.1.  Configuring Hardware Flow Control (RTS/CTS)

  See ``Flow Control'' for an explanation of it.  You should always use
  hardware flow control if possible.  Your communication program or
  "getty" should have an option for setting it (and if you're in luck it
  might be enabled by default).  It needs to be set both inside your
  modem (by an init string or default) and in the device driver.  Your
  communication program should set both of these (if you configure it
  right).

  If none of the above will fully enable hardware flow control.  Then
  you must do it yourself.  For the modem, make sure that it's either
  done by the init string or is on by default.  If you need to tell the
  device driver to do it is best done on startup by putting a file that
  runs at boot-time.  See the subsection ``Boot-time Configuration'' You
  need to add the following to such a file for each serial port (example
  is ttyS2) you want to enable hardware flow control on:




       stty crtscts < /dev/ttyS2





  If you want to see if flow control is enabled do the following:  In
  minicom (or the like) type AT&V to see how the modem is configured and
  look for &K3 which means hardware flow control.  Then see if the
  device driver knows about it by typing: stty -a < /dev/ttyS2  Look for
  "crtscts" (without a disabling minus sign).


  3.5.  Modem Configuration (excluding serial port)

  3.5.1.  AT Commands

  While the serial port on which a modem resides requires configuring,
  so does the modem itself.  The modem is configured by sending AT
  commands (or the like) to it on the same serial line that is used to
  send data.

  Most modems use an AT command set.  These are cryptic and short ASCII
  commands where all command strings are prefaced by the letters AT.
  For example: ATZ&K3  There are two commands here Z and &K3.
  Unfortunately there are many different variations of the AT command
  set so that what works for one modem may or may not work for another
  modem.  Thus there is no guarantee that the AT commands given in this
  section will work on your modem.  Another point is that to get the
  modem to act on the AT command string, a return character must be sent
  at the end of the string.  Sometimes the AT is prefaced by a return
  character and sometimes there are symbols added to the string (such as
   ) which only tell the communication program to pause for a tiny
  interval of time at that point.

  If you have a manual for your modem you can likely look up the AT
  command set in it.  Otherwise, you may try to find it on the Internet.
  One may use a search engine and include some actual commands in the
  search terms to avoid finding sites that just talk about such commands
  but fail to list them.  You might also try a few of the sites listed
  in the subsection ``Web Sites''


  3.5.2.  Init Strings: Saving and Recalling

  The examples given in this subsection are from the Hayes AT modem
  command set.  All command strings must be prefaced by the two letters
  AT (for example: AT&C1&D3 ).  When a modem is powered on, it
  automatically configures itself with one of the configurations it has
  stored in its non-volatile memory.  If this configuration is
  satisfactory there is nothing further to do.

  If it's not satisfactory, then one may either alter it or configure
  the modem each time you use it by sending it a string of commands
  known as an "init string" (= initialization string).  Normally a a
  communication program does this.  What it sends will depend on how you
  configured the communications program or what script you wrote for it
  if you use Kermit.  You can usually edit the init string and change it
  to whatever you want.  Sometimes the communications program will let
  you select the model of your modem and then it will use an init string
  that it thinks is best for that modem.

  So there is both a default "string" (called a profile) stored inside
  the modem and another (the init string) that the communications
  program sends it.  The modem will wind up configured like the default
  set it except that it will be modified by the commands included in the
  init string.  If the init string is empty then it will of course use
  the default configuration.

  Actually there is more than one "default" configuration (or profile)
  stored in the modem's non-volatile memory (it's still there when you
  turn it off).  In my modem there are two factory profiles (0 and 1,
  neither of which can be changed) and two user defined profiles (0 and
  1) that the user may set and store.  Your modem may have more.  Which
  one of these user-defined profiles is used at power-up depends on
  another item stored in the profile.  If the command &Y0 is given then
  in the future profile 0 will be used at power-on.  If it's a 1 instead
  of a 0 then profile 1 will be used at power-on.

  There are also commands to recall (use it now) any of the 4 stored
  profiles.  One may put such a command in an init string.  Of course if
  it recalls the same profile as was automatically loaded at power-up,
  nothing is changed unless the active profile has been modified since
  power-up.  Since it could have been modified It's a good idea to use
  some kind of an init string even if it does nothing more than
  recalling a stored profile.

  Recalling a saved profile (use 1 instead of 0 for profile 1):
  Z0 recalls user-defined profile 0 and resets (hangs up, etc.)
  &F0 recalls factory profile 0

  Once you have sent commands to the modem to configure it the way you
  want (including recalling a factory profile and modifying it a little)
  you may save this as a user-defined profile:
  &W0 saves the current configuration to user-profile 0

  Many people don't bother saving a good configuration in their modem,
  but instead, send the modem a longer init string each time the modem
  is used.  Another method is to restore the factory default at the
  start of the init string and then modify it a little by adding a few
  other commands to the end of the init string.  By doing it this way no
  one can cause problems by modifying the user-defined profile which is
  loaded at power-on.

  You may pick an init string supplied by someone else that they think
  is right for your modem.  Some communication programs have a library
  of init strings to select from.  The most difficult method (and one
  which will teach you the most about modems) is to study the modem
  manual and write one yourself.  You could save this configuration
  inside the modem so that you don't need an init string.  A third
  alternative is to start with an init string someone else wrote, but
  modify it to suit your purposes.


  3.5.3.  Other Modem Commands

  Future editions of Modem-HOWTO may contain more AT commands but the
  rest of this section is what was in the old Serial-HOWTO.  All strings
  must start with AT.  Here's a few Hayes AT codes that should be in the
  string (if they are not set by using a factory default or by a saved
  configuration).



       E1       command echo ON
       Q0       result codes are reported
       V1       verbose ON
       S0=0     never answer (uugetty does this with the WAITFOR option)




  Here's some more codes concerning modem control lines DCD and DSR:


       &C1     DCD is on after connect only
       &S0     DSR is always on




  These affect what your modem does when calls start and end.  What DTR
  does may also be set up but it's more complicated.


  If your modem does not support a stored profile, you can set these
  through the INIT string in a config file (or the like).  Some older
  modems come with DIP switches that affect register settings.  Be sure
  these are set correctly, too.


  Greg Hankins has a collection of modem setups for different types of
  modems.  If you would like to send him your working configuration,
  please do so:  <mailto:[e-mail]> You can get these setups
  at ftp://ftp.cc.gatech.edu/pub/people/gregh/modem-configs.


  Note: to get his USR Courier V.34 modem to reset correctly when DTR
  drops, Greg Hankins had to set &D2 and S13=1 (this sets bit 0 of
  register S13).  This has been confirmed to work on USR Sportster V.34
  modems as well.


  Note: some Supra modems treat DCD differently than other modems.  If
  you are using a Supra, try setting &C0 and not &C1.  You must also set
  &D2 to handle DTR correctly.


  4.  Modems for a Linux PC

  4.1.  External vs. Internal

  A modem for a PC may be either internal or external.  The internal one
  is installed inside of your PC (you must remove screws, etc. to
  install it) and the external one just plugs into a serial port
  connector on a PC.  Internal modems are less expensive, are less
  likely to overrun, usually use less electricity, and use up no space
  on your desk.  External modems are much easier to install, require
  less configuration, and have lights which may give you a clue as to
  what is happening.  External modems are easy to move to another
  computer.  Most external modems have no switch to turn off the power
  supply when not in use and thus are likely to consume a little
  electricity even when turned off (unless you unplug the power supply
  from the wall).  Each watt they draw costs you about $1/yr.  Another
  possible disadvantage of an external is that you will be forced to use
  an existing serial port which may not support a speed of over 115,200
  k (although as of late 1998 most new internal modems don't either
  --but some do).  If a new internal modem had a 16650 UART it would put
  less load on the CPU (but almost none do as of late 1998).

  Internal modems present a special problem for Linux, but will work
  just as well as external modems provided you avoid the high percentage
  of them that will work only for MS Windows, and also provided that you
  spend time (sometimes a lot of time) to configure them correctly.
  Some of the modems which will work only under MS Windows are,
  unfortunately, not labeled as such.  If you buy a new one, make sure
  that you can return it for a refund if it will not work under Linux.

  While most new modems are plug-and-play you have various ways to deal
  with them:

  ╖  Use the "isapnp" program

  ╖  Have a PnP BIOS do the configuring

  ╖  Patch the kernel to create a PnP Linux

     Each of the above has shortcomings.  Isapnp documentation is
     difficult to understand although reading the Plug-and-Play-HOWTO
     (at present incomplete) will aid in understanding it.  If you want
     the PnP BIOS to do the configuring, all you need to do is to make
     sure that it knows you don't have a PnP operating system.  But it
     may not do it correctly.  To find out what it's done see ``What is
     set in my serial port hardware?''.   Patching the kernel can be
     complicated too and there might not be a patch against recent
     kernels.

  There are many Linux users that say that it's a lot simpler just to
  get an external modem and plug it in.  But since new peripherals are
  mostly PnP today, you may eventually need to deal with it, so why
  delay it?  Still, the most expedient (and expensive) solution is an
  external modem (if you have a free serial port).


  4.2.  External Modems

  4.2.1.  PnP External Modems

  Many external modems are labeled "Plug and Play" (PnP) but they should
  all work fine as non-PnP modems.  Since you usually plug the modem
  into a serial port which has its own IRQ number and IO address, the
  modem needs no PnP features to set these up.  However, the serial port
  itself may need to be configured (IRQ number and IO address) unless
  the default configuration is OK.

  How can an external modem be called PnP since it can't be configured
  by PnP?   Well, it has a special PnP identification built into it that
  can be read (thru the serial port) by a PnP operating system.  Such an
  operating system would then know that you have a modem on a certain
  port and would also know the model number.  Then you might not need to
  configure application programs by telling them what port the modem is
  on (such as /dev/ttyS2 or COM3).  But if you don't have such a PnP
  operating system you will need to configure your application program
  manually by giving it the /dev id (such as /dev/ttyS2).


  4.2.2.  Cabling & Installation

  Connecting an external modem is simple compared to connecting most
  other devices to a serial port that require various types of "null
  modem" cables.  Modems use straight through cable, with no pins
  crossed over.  Most computer stores should have these.  Make sure you
  get the correct gender.  If you are using the DB9 or DB25 serial port
  at your computer, it will always be male which means that the
  connector on the cable should be female.  Hook up your modem to one of
  your serial ports.  If you are willing to accept the default IRQ and
  IO address of the port you connect it to, then you are ready to start
  your communication program and configure the modem itself.


  4.2.3.  What the Lights (LED's) Mean


  ╖  TM Test Modem

  ╖  AA Auto Answer (If on, your modem will answer an incoming call)

  ╖  RD Receive Data line = RxD

  ╖  SD Send Data line = TxD

  ╖  TR data Terminal Ready = DTR (set by your PC)

  ╖  RI Ring Indicator (If on, someone is "ringing" your modem)

  ╖  OH Off Hook (If off, your modem has hung up the phone line)

  ╖  MR Modem Ready = DSR ??

  ╖  EC Error Correction

  ╖  DC Data Compression

  ╖  HS High Speed (for this modem)


  4.3.  Internal Modems

  An internal modem is installed in a PC by taking off the cover of the
  PC and inserting the modem card into a vacant slot on the motherboard.
  There are modems for the ISA slots and others for the PCI slots.
  While external modems plug into the serial port (via a short cable)
  the internal modems have the serial port built into the modem.  In
  other words, the modem card is both a serial port and a modem.

  Setting the IO address and IRQ for a serial port was formerly done by
  jumpers on the card.  These are little black rectangular "cubes" about
  5x4x2 mm in size which push in over pins on the card.  Plug-and-Play
  modems (actually the serial port part of the modems) don't use jumpers
  for setting these but instead are configured by sending configuration
  commands to them (via IO address space on the ISA bus inside the
  computer).  Such configuration commands can be sent by a PnP BIOS, the
  isapnp program (for the ISA bus only) or by a PnP operating system.
  The configuring of them is built into Windows 95/98 OSs.  Under Linux
  you have a choice of ways (none of which is always easy) to io-irq
  configure them:

  1. Use "isapnp" which may be run automatically at every boot-time

  2. Use a PnP BIOS alone (which runs at every boot-time)

  3. Patch Linux to make it a PnP operating system


  4.4.  Internal Modems You Must Avoid (winmodems, etc.)

  A majority of internal modems made after about mid-1998 don't work
  with Linux since they are "winmodems" or the like.   Names used
  include: HSP, HCF, and soft-... modem.  Such modems turn over much (or
  even almost all) of the work of the modem to the main processor (CPU)
  chip of your computer (such as a Pentium chip).  Since only Windows
  (and not Linux) software is provided to do this, it will not work
  under Linux.  A list of modems which do and don't work under Linux is
  at Linux modem list <http://www.o2.net/~gromitkc/winmodem.html>

  A better term for "winmodem" might be "software modem" or "soft-
  modem".  But since this software is only for MS Windows, the term
  "winmodem" is now a good one since it also implies it's "Windows-
  only".   The term "Winmodem" is a trademark for a certain type of
  "winmodem".  Here is some more precise terminology regarding
  "winmodems":

  HSP (Host Signal Processor) means that the host processor (your CPU
  chip) creates the code needed to produce the electrical signal on the
  phone line.  The modem itself just creates whatever electrical
  waveshape the CPU tells it to.  In contrast to this, a
  "controllerless" modem can create the waveshapes on its own (but can't
  control the modem).  It contains no facilities to deal with bytes
  being sent and received.  It can't compress strings of bytes; it can't
  check for errors; it can't put them into packets.  In other words it
  can't control the modem but instead has the CPU do all this work using
  a program for Windows.  The Rockwell HCF (Host Controlled Family) does
  this.  If the software that does all this could be ported to Linux and
  then there wouldn't be this problem.  Besides the above, a modem which
  doesn't simulate a serial port will not work under Linux.

  How do you determine if an internal modem will work under Linux?  If
  you don't know the model of the modem and you also have Windows on
  your Linux PC, click on the "Modem" icon in the "Control Panel".
  First check out the modem list on the Web mentioned 3 paragraphs
  above.  If that doesn't work (or isn't feasible),  you can look at the
  package it came in (or a manual) find the section on the package that
  says something like "Minimum System Requirements" or just "System
  Requirements".  It may be in fine print.  Read it closely.  If Windows
  is listed as one of the requirements then it will likely not work
  under Linux.

  Otherwise, it may work under Linux if it fails to state explicitly
  that you must have Windows.  By saying it's "designed for Windows" it
  may only mean that it fully supports Microsoft's plug-and-play which
  is OK since Linux uses the same plug-and-play specs (but doesn't
  support them very well as yet).  Being "designed for Windows" thus
  gives no clue as to whether or not it will work under Linux.  You
  might check the Website of the manufacturer or inquire via email.  I
  once saw a web-page that specifically stated that one model worked
  under Linux while implying that another model didn't.

  As far as the author knows, there is no effort currently underway to
  support winmodems in Linux.  You might request that modem
  manufacturers port their code to Linux (or the like).



  4.5.  Which Internal Modems are Best to Avoid ?



  ╖  ``Avoid: winmodems'' or the like.  They will NOT work at all

  ╖  ``PCI Modems'' seldom work under Linux

  ╖  ``MWave and DSP Modems'' might work, but only if you first start
     Windows/Dos each time you power on your PC

  ╖  Modems with ``RPI (Rockwell)'' drivers work but with reduced
     performance


  4.5.1.  MWave and DSP Modems

  Such modems use DSP's (Digital Signal Processors) which are programmed
  by algorithms which must be downloaded from the hard disk to the DSP's
  memory just before using the modem.  Unfortunately, the downloading is
  done by Dos/Windows programs so one can't do it from Linux.  Ordinary
  modems that work with Linux often have a DSP too (and may mention this
  on the packaging), but the program that runs it is stored inside the
  modem.  This is not a "DSP modem" in the sense of this section and
  should work OK under Linux.

  If a DSP modem modem simulates a serial port, then it is usable with
  Linux which communicates with modems via the serial port.  If you also
  have Dos/Windows on the same PC you may be able to use the modem:
  First start Dos/Windows (make sure the modem gets initialized) and
  then without turning off the computer, go into Linux.  One way to do
  this may be to press CTRL-ALT-DEL.  An example of a DSP modem is IBM's
  Aptiva MWAVE.


  4.5.2.  Rockwell (RPI) Drivers

  Modems that require Rockwell RPI drivers are not fully usable since
  the driver software doesn't work on Linux.  The RPI does compression
  and error correction using MS Windows software on your computer's CPU.
  If you are willing to operate the modem without using the RPI (and
  have no compression nor error correction) then you may easily disable
  RPI by sending the modem (via the initialization string) a command to
  do so each time you power on your modem.  On my modem this command is
  +H0.  Not having data compression available may not be much of a
  handicap since most long files which you download from the Internet
  are already compressed and attempts at further compression may only
  slow things down a bit.


  4.5.3.  PCI Modems

  A PCI modem card is one which inserts into a PCI-bus slot on the
  motherboard of a PC.  Unfortunately, it seems that almost all PCI
  modems will not work under Linux.  But since a few people have gotten
  PCI modems to work under Linux it seems that a tiny minority of such
  modems may work under Linux.  Some claim that no PCI modem will work
  under Linux but I've seen a few messages claiming it was done (in one
  case by using jumpers).

  After you install a PCI modem, look at /proc/pci.   If you see a high
  main memory address (like 0xfebfff00), it probably works by shared
  memory which is not supported by Linux.  If there is no such high
  memory address and the only address you see is an IO address under
  0xffff ??, then it might work OK.  I'm not really sure of the
  /proc/pci signature to look for to determine whether or not the modem
  will work under Linux.  If this request is still in the latest version
  of this HOWTO, please send me info on what /proc/pci looks like for
  PCI modems that work under Linux or let me know what signatures to
  look for in /proc/pci to determine whether or not the modem will work
  with Linux.


  5.  Serial Port Devices /dev/ttySN

  5.1.  Serial Port Device Names & Numbers

  Devices in Linux have major and minor numbers.  Each serial port may
  have 4 possible names, only 2 of which are official and found in the
  /dev directory: ttyS and cua.   The cua name is deprecated and may not
  be used in the future.  See ``The cua Device''.  Dos/Windows use the
  COM name while the setserial program uses tty00, tty01, etc.  Don't
  confuse dev/tty0, dev/tty1, etc. which are used for the console (your
  PC monitor) but are not serial ports.



              set-                                              IO
       dos   serial          major minor         major minor  address
       COM1  tty00  /dev/ttyS0  4,  64;  /dev/cua0  5,  64      3F8
       COM2  tty01  /dev/ttyS1  4,  65;  /dev/cua1  5,  65      2F8
       COM3  tty02  /dev/ttyS2  4,  66;  /dev/cua2  5,  66      3E8
       COM4  tty03  /dev/ttyS3  4,  67;  /dev/cua3  5,  67      2E8




  Note that all distributions should come with ttyS devices (and cua
  devices until cua is finally abolished) already made correctly.  You
  can verify this by typing:



       linux% ls -l /dev/cua*
       linux% ls -l /dev/ttyS*





  5.2.  Creating Devices In the /dev directory

  If you don't have a device, you will have to create it with the mknod
  command.  Example, suppose you needed to create devices for ttyS0:


       linux# mknod -m 666 /dev/cua0 c 5 64
       linux# mknod -m 666 /dev/ttyS0 c 5 64




  You can use the MAKEDEV script, which lives in /dev.  See the man page
  for it.  This simplifies the making of devices.  For example, if you
  needed to make the devices for ttyS0 you would type:


       linux# cd /dev
       linux# ./MAKEDEV ttyS0



  This handles the devices creation and should set the correct permis¡
  sions.


  5.3.  Link ttySN to /dev/modem ?

  On some installations, two extra devices will be created, /dev/modem
  for your modem and /dev/mouse for your mouse.  Both of these are
  symbolic links to the appropriate device in /dev which you specified
  during the installation (unless you have a bus mouse, then /dev/mouse
  will point to the bus mouse device).

  There has been some discussion on the merits of /dev/mouse and
  /dev/modem.  The use of these links is discouraged.  In particular, if
  you are planning on using your modem for dialin you may run into
  problems because the lock files may not work correctly if you use
  /dev/modem. Use them if you like, but be sure they point to the right
  device.  However, if you change or remove this link, some applications
  might need reconfiguration.


  5.4.  The cua Device

  Each ttyS device has a corresponding cua device.  It is planned to
  eventually abolish cua so it's best to use ttyS (unless cua is
  required).  There is a difference between cua and ttyS but a savvy
  programmer can make a ttyS port behave just like a cua port so there
  is no real need for the cua anymore.  Except some older programs may
  need to use the cua.

  What's the difference?  The  main difference between cua and ttyS has
  to do with what happens in a C-program when an ordinary "open" command
  tries to open the port. If a cua port has been set to check modem
  control signals, the port can be opened even if the DCD modem control
  signal says not to.  Astute programming (by adding additional lines to
  the program) can force a ttyS port to behave this way also.  But a cua
  port can be more easily programmed to open for dialing out on a modem
  even when the modem fails to assert DCD (since no one has called into
  it and there's no carrier).  That's why cua was once used for dial-out
  and ttyS used for dial-in.

  Starting with Linux kernel 2.2, a warning message will be put in the
  kernel log when one uses cua.  This is an omen that cua will sometime
  disappear.


  6.  Interesting Programs You Should Know About

  6.1.  What is getty?

  "getty" is a program that handles the login process when you log into
  a Linux box.  You will need to use some type of "getty" if you want
  others to be able to dial in to your Linux PC with a modem.  You do
  not need to use getty if you only want to dial out with your modem.
  There are three getty versions that modems may use with Linux: mgetty,
  getty_ps, and agetty.  agetty is the simplest (and weakest) of the
  three and some consider it mainly for text-terminals.  The syntax for
  these programs differs, so be sure to check that you are using the
  correct syntax for whatever getty you use.  If this sentence appears
  in the latest version of this howto, then if you can write a few
  paragraphs comparing mgetty with getty_ps, etc. please submit it so
  that it can be included here.  Mgetty is only a few years old and
  seems to be more popular for new installations than the older
  getty_ps.


  6.1.1.  About mgetty

  mgetty is a version of getty mainly for use with modems.  It may be
  used for hard-wired terminals but the documentation is about 99%
  related to modems.  In addition to allowing dialup logins, mgetty also
  provides FAX support and auto PPP detection.  There is a supplemental
  program called vgetty which handles voicemail for some modems.  mgetty
  documentation (supplied in texinfo format) is good, and does not need
  supplementing.  Please refer to it for installation instructions.  You
  can find the latest information on mgetty at
  http://www.leo.org/~doering/mgetty/ and
  <http://alpha.greenie.net/mgetty>


  6.1.2.  About getty_ps


  getty_ps  contains two programs: getty is used for console and
  terminal devices, and uugetty for modems.  Greg Hankins (former author
  of Serial-HOWTO) used uugetty so his writings about it are included
  here.  See ``Uugetty''.  The other gettys are well covered by the
  documentation that comes with them.


  6.1.3.  About agetty and mingetty

  agetty is the third variation of getty. It's a simple, completely
  functional implementation of getty which is best suited for virtual
  consoles or terminals rather than modems.  But it works fine with
  modems under favorable conditions.

  mingetty is a small getty that will work only for consoles (monitors).
  While the previous 3 variations of getty will work for both real
  terminals and dial-in modems, mingetty will not do this.


  6.2.  What is Setserial ?

  6.2.1.  Intro to Setserial

  setserial is a program which allows you to tell the device driver
  software the IO address of the serial port, which IRQ is set in the
  port's hardware, etc.  With appropriate options, it can also probe (at
  a given IO address) for a serial port but you must guess the IO
  address (or it may use whatever address the driver thinks your
  /dev/ttySx is at).  Setserial does not set either IRQ's nor IO
  addresses in the serial port hardware itself.  You must tell setserial
  the identical values that have been set in the hardware.  It's set in
  the hardware either by jumpers or by plug-and-play.  Do not just
  invent some values that you think would be nice to use.  However, if
  you know the IO address but don't know the IRQ you may command
  setserial to attempt to determine it.

  You can see a list of possible commands to use (but not the one-letter
  options such as -v for verbose --which you should normally use when
  troubleshooting) by typing setserial with no arguments.  Note that
  setserial calls an IO address a "port".  If the argument to setserial
  is for example just /dev/ttyS1, then you'll see some info about how
  that device driver is configured for that port.  But this doesn't tell
  you if the hardware actually has these values set in it.  If fact, you
  can run setserial and assign a purely fictitious IO address, any IRQ,
  and whatever uart type you would like to have.  Then the next time you
  type "setserial ..." it will display these bogus values without
  complaint.  Note that assignments made by setserial are lost when the
  PC is powered down so it is usually run automatically somewhere each
  time that Linux is booted.
  6.2.2.  Probing

  In order to try to find out if you have a certain piece of serial
  hardware you must first know its IO address (or the device driver must
  have an IO address for it, likely previously set by setserial).  To
  try to detect the physical hardware use the -v (verbose) and
  autoconfig command to setserial.  If the resulting message shows a
  uart type such as 16550A, then you're OK.  If instead it shows
  "unknown" for the uart type, then there is likely no serial port at
  all at that IO address.  Some cheap serial ports don't identify
  themselves correctly so if you see "unknown" you still might have
  something there.  See the file in which "setserial" is run at boot-
  time.  Besides auto-probing for uart type, setserial can auto-probe
  for IRQ's but this doesn't always work right either.


  6.2.3.  Boot-time Configuration

  There is usually a file somewhere that runs setserial at boot-time.
  If it's not run at boot-time then your Linux system will automatically
  configure only ttyS{0-3} using the default IRQs of 4 and 3 (with the
  default IRQ conflicts).  In 1998 it was (temporarily ?) changed to
  only ttyS{0-1}.  So if you have more than 2 serial ports, or want to
  have control over how the ports are configured you should configure
  using setserial.  In fact, your distribution may have set things up so
  that the setserial program runs automatically at boot-time.

  The file that runs setserial at boot-time is likely somewhere in the
  /etc directory-tree.  You might use "locate" to find a file named:
  rc.serial, or 0setserial (Debian), etc.  This supplied file which runs
  setserial at start-up may contain a number of commented-out examples.
  By uncommenting some of these and/or modifying them, you may be able
  to set things up correctly or run some tests.  You could copy a few of
  them to another file and then execute it as a shell script but don't
  forget to also copy any capitalized definitions needed such as
  SETSERIAL=/bin/setserial.

  If you use setserial you could test it on the command line first, and
  then when you have it working, put it into the file which runs it at
  boot-time: /etc/rc.d/rc.serial or /etc/rc.boot/0setserial so that it
  is run at startup.  If those files don't exist try /etc/rc.d/rc.local
  (someone reported that with one kernel, rc.local was executed too late
  after the serial port had already been opened).  Or you could just
  edit one of the above files and cross your fingers.  Make sure that
  you are using a valid path for setserial, and a valid device name.


  6.2.4.  IRQs

  By default, both ttyS0 and ttyS2 share IRQ 4, while ttyS0 and ttyS3
  share IRQ 3.  But sharing serial interrupts is not permitted unless
  you have kernel 2.2 or better.  If you don't have this modern kernel
  but only have two serial ports ttyS0 and ttyS1 you're still OK since
  IRQ sharing conflicts don't exist for non-existent devices.

  But if you do have more than 2 serial ports, then for kernels < 2.2
  such sharing may be dangerous if the two devices with the same IRQ are
  being used at the same time.  If you add an internal modem and retain
  ttyS0 and ttyS1, then you should attempt to find an unused IRQ and set
  it both on your modem card (or serial port) and then use setserial to
  assign it to your device driver.  If IRQ 5 is not being used for a
  sound card, this may be one you can use for a modem.  To set the IRQ
  in hardware you may need to use isapnp, a PnP BIOS (See ``Using a PnP
  BIOS to I0-IRQ Configure'') or patch Linux to make it PnP.  To help
  you determine which spare IRQ's you might have, type "man setserial"
  and search for say: "IRQ 11".
  6.3.  What is isapnp ?

  isapnp is a program to configure Plug-and-Play (PnP) devices on the
  ISA bus including internal modems.  It comes in a package called
  "isapnptools" and includes another program, "pnpdump" which finds all
  your ISA PnP devices and shows you options for configuring them in a
  format which may be added to the PnP configuration file:
  /etc/isapnp.conf.  It may also be used with the --dumpregs option to
  show the current IO address and IRQ of the modem's serial port.  The
  isapnp command may be put into a startup file so that it runs each
  time you start the computer and thus will configure ISA PnP devices.
  It is able to do this even if your BIOS doesn't support PnP.  See
  Plug-and-Play-HOWTO.


  6.4.  What is wvdialconf ?

  wvdialconf will try to find which serial port has a modem on it.  It
  also creates a configuration program for the wvdial program.  wvdial
  is used for simplified dialing out using the PPP protocol to an ISP.
  But you don't need to install PPP in order to use wvdialconf.  It will
  not find a modem if it's in use.  It will also automatically devise a
  "suitable" init strings but sometimes gets it wrong.  Since this
  command has no options, it's simple to use but you must give it the
  name of a file to put the init string (and other data) into.  For
  example type: wvdialconf my_file_name.


  6.5.  What is stty ?

  stty is like setserial but it sets the baud rate and other parameters
  of a serial port.  Typing "stty -a < /dev/ttyS2" should show you how
  ttyS2 is configured.  Most of the settings are for things that you
  never need to use with modems (such as some used only for old
  terminals of the 1970s).  Your communication package should
  automatically set up all the setting correctly for modems.  But stty
  is sometimes useful for trouble-shooting.

  Two items set by stty are: 1. Hardware flow control by "crtscts" and
  2. Ignore the DCD signal from the modem: "clocal".  If the modem is
  not sending a DCD signal and clocal is disabled (stty shows -clocal)
  then a program may not be able to open the serial port.  If the port
  can't open, the program may just hang, waiting (often in vain) for a
  DCD signal from the modem.

  Minicom sets clocal automatically when it starts up so there is no
  problem.  But version 6.0.192 of Kermit hung when I set -clocal and
  tried to "set line ..." If -clocal is set and there is no DCD signal
  then even the "stty" command will hang and there is seemingly no way
  to set clocal (except by running minicom).  But minicom will restore
  -clocal when it exits.  One way to get out of this is to use minicom
  to send the "AT&C" to the modem (to get the DCD signal) and then exit
  minicom with no reset so that the DCD signal remains on.  Then you may
  use stty again.


  7.  Trying Out Your Modem (Dialing Out).

  7.1.  Are You Ready to Dial Out ?

  Once you've plugged in your modem and know which serial port it's on
  you're ready to try using it.  Before you try to get the Internet on
  it or have people call in to you, first try something simpler like
  dialing out to some number to see if your modem is working OK.  Find a
  phone number that is connected to a modem.  It you don't know what
  number to call, ask at computer stores for such phone numbers of
  bulletin boards, etc. or see if a local library has a phone number for
  their on-line catalog.

  Then make sure you are ready to phone.  Do you know what serial port
  (such as ttyS2) your modem is on?  You should have found this out when
  you io-irq configured your serial ports.  Have you decided what speed
  you are going to use for this port?  See ``Speed Table'' for a quick
  selection or ``What Speed Should I Use'' for more details.  If you
  have no idea what speed to set, just set it a few times faster than
  the advertised speed of your modem.  Also remember that if you see a
  menu where an option is "hardware flow control" and/or "RTS/CTS" or
  the like, select it.  Is a live telephone cable plugged in to your
  modem?  You may want to connect the cable to a real telephone to make
  sure that it can produce a dial tone.

  Now you need to select a communication (dialing) program to use to
  dial out.  Dialing programs include: minicom, seyon (X-windows), and
  kermit.  See section ``Communications Programs'' about some
  communications programs.  Two examples are presented next: ``Dialing
  Out with Minicom'' and ``Dialing Out with Kermit''



  7.2.  Dialing Out with Minicom

  Minicom comes with most Linux distributions.  To configure it you
  should be the root user.  Type "minicom -s" to configure.  This will
  take you directly to the configuration (set-up) menus.  Alternatively
  you could just run "minicom" and then type ^A to see the bottom status
  line.  This shows to type ^A Z for help (you've already typed the ^A
  so just type z).  From the help menu go to the Configuration menu.

  Most of the options don't need to be set for just simply dialing out.
  To configure you have to supply a few basic items: the name of the
  serial port your modem is on such as /dev/ttyS2 and the speed such as
  115200.  These are set at the serial port menu.  Go to it and set
  them.  Also (if possible) set hardware flow control (RTS/CTS).  Then
  save them.  When typing in the speed, you should also see something
  like "8N1" which you should leave alone.  It means: 8-bit bytes, No
  parity, 1 stop-bit appended to each byte.  If you can't find the speed
  you want, a lower speed will always work for a test.   Exit (hit
  return) when done and save the configuration as default (dfl) using
  the menu.  You may want to exit minicom and start it again so it can
  now find the serial port and initialize the modem, or you could go to
  help and tell minicom to initialize the modem.

  Now you are ready to dial.  But first at the main screen you get after
  you first type "minicom" make sure there's a modem there by typing AT
  and then hit the "enter/return" key.  It should display OK.  If it
  doesn't something is wrong and there is no point of trying to dial.

  If you got the "OK" go back to help and select the dialing directory.
  You may edit it and type in a phone number, etc. into the directory
  and then select "dial" to dial it.  Alternatively, you may just dial
  manually (by selecting "manual" and then type the number at the
  keyboard).  If it doesn't work, carefully note any error messages and
  try to figure out what went wrong.


  7.3.  Dialing Out with Kermit

  You can find the latest version of kermit at
  http://www.columbia.edu/kermit/.  For example, say your modem was on
  ttyS3, and it's speed was 115200 bps.  You would do the following:


  linux# kermit
  C-Kermit 6.0.192, 6 Sep 96, for Linux
   Copyright (C) 1985, 1996,
    Trustees of Columbia University in the City of New York.
  Default file-transfer mode is BINARY
  Type ? or HELP for help.
  C-Kermit>set line /dev/ttyS3
  C-Kermit>set carrier-watch off
  C-Kermit>set speed 115200
  /dev/ttyS3, 115200 bps
  C-Kermit>c
  Connecting to /dev/ttyS3, speed 115200.
  The escape character is Ctrl-\ (ASCII 28, FS)
  Type the escape character followed by C to get back,
  or followed by ? to see other options.
  ATE1Q0V1                           ; you type this and then the Enter key
  OK                                 ; modem should respond with this




  If your modem responds to AT commands, you can assume your modem is
  working correctly on the Linux side.  Now try calling another modem by
  typing:


       ATDT7654321




  where 7654321 is a phone number.  Use ATDP instead of ATDT if you have
  a pulse line.  If the call goes through, your modem is working.

  To get back to the kermit prompt, hold down the Ctrl key, press the
  backslash key, then let go of the Ctrl key, then press the C key:


       Ctrl-\-C
       (Back at linux)
       C-Kermit>quit
       linux#




  This was just a test using the primitive "by-hand" dialing method.
  The normal method is to let kermit do the dialing for you with its
  built-in modem database and automatic dialing features, for example
  using a US Robotics (USR) modem:
















  linux# kermit
  C-Kermit 6.0.192, 6 Sep 1997, for Linux
   Copyright (C) 1985, 1996,
    Trustees of Columbia University in the City of New York.
  Default file-transfer mode is BINARY
  Type ? or HELP for help
  C-Kermit>set modem type usr        ; Select modem type
  C-Kermit>set line /dev/ttyS3       ; Select communication device
  C-Kermit>set speed 115200          ; Set the dialing speed
  C-Kermit>dial 7654321              ; Dial
   Number: 7654321
   Device=/dev/ttyS3, modem=usr, speed=115200
   Call completed.<BEEP>
  Connecting to /dev/ttyS3, speed 115200
  The escape character is Ctrl-\ (ASCII 28, FS).
  Type the escape character followed by C to get back,
  or followed by ? to see other options.

  Welcome to ...

  login:





  8.  Uugetty for Dial-In (from the old Serial-HOWTO)

  8.1.  Installing getty_ps

  Since uugetty is part of getty_ps you'll first have to install
  getty_ps.  If you don't have it, get the latest version from
  metalab.unc.edu:/pub/Linux/system/serial.  In particular, if you want
  to use high speeds (57600 and 115200 bps), you must get version 2.0.7j
  or later.  You must also have libc 5.x or greater.


  By default, getty_ps will be configured to be Linux FSSTND (File
  System Standard) compliant, which means that the binaries will be in
  /sbin, and the config files will be named /etc/conf.{uu}getty.ttySN.
  This is not apparent from the documentation!  It will also expect lock
  files to go in /var/lock.  Make sure you have the /var/lock directory.

  If you don't want FSSTND compliance, binaries will go in /etc, config
  files will go in /etc/default/{uu}getty.ttySN, and lock files will go
  in /usr/spool/uucp.  I recommend doing things this way if you are
  using UUCP, because UUCP will have problems if you move the lock files
  to where it isn't looking for them.


  getty_ps can also use syslogd to log messages.  See the man pages for
  syslogd(1) and syslog.conf(5) for setting up syslogd, if you don't
  have it running already.  Messages are logged with priority LOG_AUTH,
  errors use LOG_ERR, and debugging uses LOG_DEBUG.  If you don't want
  to use syslogd you can edit tune.h in the getty_ps source files to use
  a log file for messages instead, namely /var/adm/getty.log by default.


  Decide on if you want FSSTND compliance and syslog capability.  You
  can also choose a combination of the two.  Edit the Makefile, tune.h
  and config.h to reflect your decisions.  Then compile and install
  according to the instructions included with the package.

  >From this point on, all references to getty will refer to getty_ps.
  References to uugetty will refer to the uugetty that comes with the
  getty_ps package.  These instructions will not work for mgetty or
  agetty.


  8.2.  Setting up uugetty

  Make sure that you have an outgoing and incoming device for the serial
  port your modem is on.  If you have your modem on ttyS3 you
  will need the /dev/cua3, and /dev/ttyS3 devices.  If you don't have
  the correct devices, see section ``Creating Devices In <tt>/dev</tt>''
  on how to create devices, and create the devices.  If you want to be
  able to dial out with your modem while uugetty is watching the port
  for logins, use the /dev/cuaN device instead of the /dev/ttySN device
  [One wouldn't need cua if the software was written to avoid it.]


  uugetty does important lock file checking.  Update /etc/gettydefs to
  include an entry for your modem.  When you are done editing
  /etc/gettydefs, you can verify that the syntax is correct by doing:


       linux# getty -c /etc/gettydefs





  8.2.1.  Modern Modems

  If you have a 9600 bps or faster modem with data compression, you can
  lock your serial port to one speed.  For example:


       # 115200 fixed speed
       F115200# B115200 CS8 # B115200 SANE -ISTRIP HUPCL #@S @L @B login: #F115200





  If you have your modem set up to do RTS/CTS hardware flow control, you
  can add CRTSCTS to the entries:


       # 115200 fixed speed with hardware flow control
       F115200# B115200 CS8 CRTSCTS # B115200 SANE -ISTRIP HUPCL CRTSCTS #@S @L @B login: #F115200





  8.2.2.  Old slow modems

  If you have a slow modem (under 9600 bps) Then, instead of one line
  for a single speed, your need several lines to try a number of speeds.
  Note the these lines are linked to each other by the last "word" in
  the line such as #38400.  Blank lines are needed between each entry.










  # Modem entries
  115200# B115200 CS8 # B115200 SANE -ISTRIP HUPCL #@S @L @B login: #57600

  57600# B57600 CS8 # B57600 SANE -ISTRIP HUPCL #@S @L @B login: #38400

  38400# B38400 CS8 # B38400 SANE -ISTRIP HUPCL #@S @L @B login: #19200

  19200# B19200 CS8 # B19200 SANE -ISTRIP HUPCL #@S @L @B login: #9600

  9600# B9600 CS8 # B9600 SANE -ISTRIP HUPCL #@S @L @B login: #2400

  2400# B2400 CS8 # B2400 SANE -ISTRIP HUPCL #@S @L @B login: #115200





  8.2.3.  Login Banner

  If you want, you can make uugetty print interesting things in the
  login banner.  In Greg's  examples, he has the system name, the serial
  line, and the current bps rate.  You can add other things:



              @B    The current (evaluated at the time the @B is seen) bps rate.
              @D    The current date, in MM/DD/YY.
              @L    The serial line to which getty is attached.
              @S    The system name.
              @T    The current time, in HH:MM:SS (24-hour).
              @U    The number of currently signed-on users.  This is  a
                    count of the number of entries in the /etc/utmp file
                    that have a non-null ut_name field.
              @V    The value of VERSION, as given in the defaults file.
              To display a single '@' character, use either '\@' or '@@'.





  8.3.  Customizing uugetty

  There are lots of parameters you can tweak for each port you have.
  These are implemented in separate config files for each port.  The
  file /etc/conf.uugetty will be used by all instances of uugetty, and
  /etc/conf.uugetty.ttySN will only be used by that one port.  Sample
  default config files can be found with the getty_ps source files,
  which come with most Linux distributions.  Due to space concerns, they
  are not listed here.  Note that if you are using older versions of
  getty (older than 2.0.7e), or aren't using FSSTND, then the default
  file will be /etc/default/uugetty.ttySN.  Greg's
  /etc/conf.uugetty.ttyS3 looked like this:














  # sample uugetty configuration file for a Hayes compatible modem to allow
  # incoming modem connections
  #
  # alternate lock file to check... if this lock file exists, then uugetty is
  # restarted so that the modem is re-initialized
  ALTLOCK=cua3
  ALTLINE=cua3
  # line to initialize
  INITLINE=cua3
  # timeout to disconnect if idle...
  TIMEOUT=60
  # modem initialization string...
  # format: <expect> <send> ... (chat sequence)
  INIT="" AT\r OK\r\n
  WAITFOR=RING
  CONNECT="" ATA\r CONNECT\s\A
  # this line sets the time to delay before sending the login banner
  DELAY=1
  #DEBUG=010




  Add the following line to your /etc/inittab, so that uugetty is run on
  your serial port, substituting in the correct information for your
  environment - run-levels (2345 or 345, etc.)  config file location,
  port, speed, and default terminal type:



       S3:2345:respawn:/sbin/uugetty -d /etc/default/uugetty.ttyS3 ttyS3 F115200 vt100




  Restart init:


       linux# init q




  For the speed parameter in your /etc/inittab, you want to use the
  highest bps rate that your modem supports.

  Now Linux will be watching your serial port for connections.  Dial in
  from another machine and login to you Linux system.

  uugetty has a lot more options, see the man page for getty(1m) for a
  full description.  Among other things there is a scheduling feature,
  and a ringback feature.


  9.  What Speed Should I Use with My Modem?

  By "speed" we really mean the "data flow rate" but almost everybody
  incorrectly calls it speed.  For all modern modems you have no choice
  of the speed that the modem uses on the telephone line since it will
  automatically choose the highest possible speed that is possible under
  the circumstances.  But you do have a choice as to what speed will be
  used between your modem and your computer.  This is sometimes called
  "DTE speed" where "DTE" stands for Data Terminal Equipment (Your
  computer is a DTE.)  You need to set this speed high enough so this
  part of the signal path will not be a bottleneck.  The setting for the
  DTE speed is the maximum speed of this link.  Most of the time it will
  likely operate at lower speeds.

  For an external modem, DTE speed is the speed (in bits/sec) of the
  flow over the cable between you modem and PC.  For an internal modem,
  it's the same idea since the modem also emulates a serial port.  It
  may seem ridiculous having a speed limit on communication between a
  computer and a modem card that is directly connected inside the
  computer to a much higher speed bus.  But it's that way since the
  modem card probably includes a dedicated serial port which does have
  speed limits (and settable speeds).


  9.1.  Speed and Data Compression

  What speed do you choose?  If it were not for "data compression" one
  might try to choose a DTE speed exactly the same as the modem speed.
  Data compression takes the bytes sent to the modem from your computer
  and encodes them into a fewer number of bytes.  For example, if the
  flow (speed) from the PC to the modem was 20,000 bytes/sec (bps) and
  the compression ratio was 2 to 1, then only 10,000 bytes/sec would
  flow over the telephone line.  Thus for a 2:1 compression ratio you
  need to set the speed double the maximum modem speed on the phone
  line.  If the compression ratio is 3 to 1 you need to set it 3 times
  faster.


  9.2.  Where do I Set Speed ?

  This DTE speed is normally set by a menu in your communications
  program or by an option given to the getty command if someone is
  dialing in.  You can't set the DCE modem-to-modem speed.


  9.3.  Can't Set a High Enough Speed

  You need to find out the highest speed supported by your hardware.  As
  of late 1998 most hardware only supported speeds up to 115.2K bps.  A
  few 56K internal modems support 230.4K bps.  If you  have a
  communications programs that doesn't show high enough speeds in its
  menu, then there are some options you can give to the setserial
  command so that a low speed command from the communication program
  will actually result in a higher speed.  With these options, when you
  set the speed for 38400 the actual speed will be much higher.  See the
  man page for "setserial" and search for spd_hi, spd_vhi, spd_cust,
  baud_base, and divisor.  Note that you must set baud_base to the
  actual maximum speed of the hardware.  This speed is usually lower
  than the frequency of the crystal oscillator in the hardware since the
  crystal frequency is often divided by 16 in the hardware to get the
  actual top speed.  The reason the crystal frequency needs to be higher
  is so that this high crystal speed can be used to take a number of
  samples of each  bit to determine if it's a 1 or a 0.  To get a speed
  of 230400 (if this is what your serial port hardware supports --few
  do) on ttyS2 you could use:

  setserial /dev/ttyS2 spd_cust baud_base 230400 divisor 1
  In some cases this works when "stty 230400" doesn't.  Since you've
  used spd_cust you'll have to claim the speed is 38400 somewhere else
  to obtain the actual speed of 230400.  This method of setting speed is
  sort of a hack and when applications catch up to higher speeds it
  might not be needed anymore.


  9.4.  Speed Table

  It's best to have at least a 16650 UART for a 56K modem but few modems
  support it.  Second best is a 16550 that has been tweaked to give
  230,400 bps.  Here are some suggested speeds to set your serial line
  if your modem speed is:

  ╖  56K (V.90) use 115200 bps or 230400 bps (a few % faster ?)

  ╖  28.8K (V.34), 33.6K (V.34) use 115200 bps

  ╖  14400 bps (V.32bis), with V.42bis data compression, use 57600 bps

  ╖  9600 bps (V.32), with V.42bis data compression, use 38400 bps

  ╖  slower than a 9600 bps (V.32) modem, set your speed to the highest
     speed your modem supports.


  10.  Communications Programs And Utilities

  PPP is by far the most widely used.  It's used for Internet access.
  For dialing out to public libraries, bulletin boards, etc.  minicom is
  the most popular followed by Seyon (X-Windows only) and Kermit.


  10.1.  Minicom vs. Kermit

  Minicom is only a communications program while Kermit is both a
  communications program and a file transfer protocol.  But one may use
  the Kermit protocol from within Minicom (provided one has Kermit
  installed on one's PC) .  Minicom is menu based while Kermit is
  command line based (interactive at the special Kermit prompt).  While
  the Kermit program is free software, the documentation is not all
  free.  There is no detailed manual supplied and it is suggested that
  you purchase a book as the manual.  However Kermit has interactive
  online help which tells all but lacks tutorial explanations for the
  beginner.  Commands may be put in a script file so you don't have to
  type them over again each time.  Kermit (as a communications program)
  is more powerful than Minicom.

  Although all Minicom documentation is free, it's not as extensive as
  Kermit's.  Since permission is required to include Kermit in a
  commercial distribution, and since the documentation is not entirely
  free, some distributions don't include Kermit.  In my opinion it's
  easier to set up Minicom and there is less to learn.


  10.2.  Lists of Programs

  Here is a list of some communication software you can choose from,
  available via FTP, if they didn't come with your distribution.  I
  would like comparative comments on the dialout programs.  Are the
  least popular ones obsolete?


  10.2.1.  Least Popular Dialout



  ╖  ecu - a communications program

  ╖  pcomm - procomm-like communications program with zmodem

  ╖  xc - xcomm communication package





  10.2.2.  Most Popular Dialout



  ╖  minicom - telix-like communications program.  Supports scripts,
     zmodem, kermit

  ╖  C-Kermit <http://www.columbia.edu/kermit/> - portable, scriptable,
     serial and TCP/IP communications including file transfer,
     character-set translation, and zmodem support

  ╖  seyon - X based communication program


  10.2.3.  Fax



  ╖  efax a small fax program

  ╖  hylafax a large fax program based on the client-server model.

  ╖  mgetty+fax handles fax stuff and login for dial-ins


  10.2.4.  Voicemail



  ╖  mvm <http://www-internal.alphanet.ch/~schafer/mvm> is a Minimal
     VoiceMail for Linux

  ╖  vgetty is an extension to mgetty that handles voicemail for some
     modems.  It should come with recent releases of mgetty.


  10.2.5.  Dial-in (uses getty)



  ╖  mgetty+fax is for modems and is well documented.  It also handles
     fax stuff and provides an alternative to uugetty.  It's
     incorporating callback and voicemail (using vgetty) features.  See
     ``About mgetty''

  ╖  uugetty is for modems.  It comes as a part of the ps_getty package.
     See ``About getty_ps''


  10.2.6.  Other



  ╖  callback is where you dial out to a remote modem and then that
     modem hangs up and calls you back (to save on phone bills).

  ╖  SLiRP and term provide a PPP-like service that you can run in user
     space on a remote computer with a shell account.  See ``term and
     SLiRP'' for more details

  ╖  ZyXEL is a control program for ZyXEL U-1496 modems.  It handles
     dialin, dialout, dial back security, FAXing, and voice mailbox
     functions.

  ╖  SLIP and PPP software can be found at
      ftp://metalab.unc.edu/pub/Linux/system/network/serial.
  ╖  Other things can be found on
     ftp://metalab.unc.edu/pub/Linux/system/serial and
     ftp://metalab.unc.edu/pub/Linux/apps/serialcomm or one of the many
     mirrors.  These are the directories where serial programs are kept.


  10.3.  SLiRP  and term

  SLiRP and term are programs which are of use if you only have a dial-
  up shell account on a Unix-like machine and want to get the equivalent
  of a PPP account (or the like) without being authorized to have it
  (possibly because you don't want to pay extra for it, etc.).  SLiRP is
  more popular than term which is almost obsolete.

  To use SLiRP you install it in your shell account on the remote
  computer.  Then you dial up the account and run SLiRP on the remote
  and PPP on your local PC.  You now have a PPP connection over which
  you may run a web browser on your local PC such as Netscape, etc.
  There may be some problems as SLiRP is not as good as a real PPP
  account.  Some accounts may provide SLiRP since it saves on IP
  addresses (You have no IP address while using SLiRP).

  term is something like SLiRP only you need to run term on both the
  local and remote computer.  There is no PPP on the phone line since
  term uses its own protocol.  To use term from your PC you need to use
  a term-aware version of ftp to do ftp, etc.  Thus it's easier to use
  SLiRP since the ordinary version of ftp works fine with SLiRP.  There
  is an unmaintained Term HOWTO.


  11.  What Are UARTs?  How Do They Affect Performance?

  UARTs (Universal Asynchronous Receiver Transmitter) are serial chips
  on your PC motherboard (or on an internal modem card).   The UART
  function may also be done on a chip that does other things as well.
  On older computers like many 486's, the chips were on the disk IO
  controller card.  Still older computer have dedicated serial boards.

  The UART's purpose is to convert bytes from the PC's parallel bus to a
  serial bit-stream.  The cable going out of the serial port is serial
  and has only one wire for each direction of flow.  The serial port
  sends out a stream of bits, one bit at a time.  Conversely, the bit
  stream that enters the serial port via the external cable is converted
  to parallel bytes that the computer can understand.  UARTs deal with
  data in byte sized pieces, which is conveniently also the size of
  ASCII characters.

  Say you have a terminal hooked up to your PC.  When you type a
  character, the terminal gives that character to it's transmitter (also
  a UART).  The transmitter sends that byte out onto the serial line,
  one bit at a time, at a specific rate.  On the PC end, the receiving
  UART takes all the bits and rebuilds the (parallel) byte and puts it
  in a buffer.

  There are two basic types of UARTs: dumb UARTS and FIFO UARTS.  Dumb
  UARTs are the 8250, 16450, early 16550, and early 16650.  They are
  obsolete but if you understand how they work it's easy to understand
  how the modern ones work with FIFO UARTS ( late 16550, 16550A, 16c552,
  late 16650, 16750, and 16C950).

  There is some confusion regarding 16550.  Early models had a bug and
  worked properly only as 16450's.  Later models with the bug fixed were
  named 16550A but many manufacturers did not accept the name change and
  continued calling it a 16550.  Most all 16550's in use today are like
  16550A's.  Linux will report it as being a 16550A even though your
  hardware manual (or a label note) says it's a 16550.  A similar
  situation exists for the 16650 (only it's worse since the manufacturer
  allegedly didn't admit anything was wrong).  Linux will report a late
  16650 as being a 16650V2.  If it reports it as 16650 it is bad news
  and only is used as if it had a one-byte buffer.

  To understand the differences between dumb and FIFO (First In, First
  Out queue discipline) first let's examine what happens when a UART has
  sent or received a byte.  The UART itself can't do anything with the
  data passing thru it, it just receives and sends it.  For the original
  dumb UARTS, the CPU gets an interrupt from the serial device every
  time a byte has been sent or received.  The CPU then moves the
  received byte out of the UART's buffer and into memory somewhere, or
  gives the UART another byte to send.  The 8250 and 16450 UARTs only
  have a 1 byte buffer.  That means, that every time 1 byte is sent or
  received, the CPU is interrupted.  At low transfer rates, this is OK.
  But, at high transfer rates, the CPU gets so busy dealing with the
  UART, that is doesn't have time to adequately tend to other tasks.  In
  some cases, the CPU does not get around to servicing the interrupt in
  time, and the byte is overwritten, because they are coming in so fast.
  This is called an "overrun" or "overflow".

  That's where the FIFO UARTs are useful.  The 16550A (or 16550) FIFO
  chip comes with 16 byte FIFO buffers.  This means that it can receive
  up to 14 bytes (or send 16 bytes) before it has to interrupt the CPU.
  Not only can it wait for more bytes, but the CPU then can transfer all
  14 (or more) bytes at a time.  Although the interrupt threshold
  (trigger level) may be set at 8 instead of 14, this is still a
  significant advantage over the other UARTs, which only have 1 byte
  buffers.  The CPU receives less interrupts, and is free to do other
  things.  Data is not lost, and everyone is happy.

  While most PC's only have a 16550 with 16-byte buffers, better UARTS
  have even larger buffers.  Note that the interrupt is issued slightly
  before the buffer get full (at say a "trigger level" of 14 bytes for a
  16-byte buffer).  This allows room for a few more bytes to be received
  during the time that the interrupt is being serviced.  The trigger
  level may be set to various permitted values by kernel software.  A
  trigger level of 1 will be almost like a dumb UART (except that it
  still has room for 15 more bytes after it issues the interrupt).

  If you type something while visiting a BBS, the characters you type go
  out thru the serial port.  Your typed characters that you see on the
  screen are what was echoed back thru the telephone line thru your
  modem and then thru your serial port to the screen.  If you had a
  16-byte buffer on the serial port which held back characters until it
  had 14 of them, you would need to type many characters before you
  could see what you typed (before they appeared on the screen).  This
  would be very confusing but there is a "timeout" to prevent this.
  Thus you normally see a character on the screen just as soon as you
  type it.

  The "timeout" works like this for the receive UART buffer: If
  characters arrive one after another, then an interrupt is issued only
  when say the 14th character reaches the buffer.  But if a character
  arrives and the next character doesn't arrive soon thereafter, then an
  interrupt is issued.  This happens even though there are not 14
  characters in the buffer (there may only be one character in it).
  Thus when what you type goes thru this buffer, it acts almost like a
  1-byte buffer even though it is actually a 16-byte buffer (unless your
  typing speed is a hundred times faster than normal).  There is also
  "timeout" for the transmit buffer as well.

  Here's a list of UARTs.  TL is Trigger Level

  ╖  8250, 16450, early 16550: Obsolete with 1-byte buffers

  ╖  16550, 16550A, 16c552: 16-byte buffers, TL=1,4,8,14

  ╖  16650:  32-byte buffers. Speed up to 460.8 Kbps

  ╖  16750:  64-byte buffer for send, 56-byte for receive.  Speed up to
     921.6 Kbps

  ╖  Hayes ESP: 1K-byte buffers.

  The obsolete ones are only good for modems no higher than 14.4k (DTE
  speeds up to 38400 bps).  For modern modems you need at least a 16550
  (and not an early 16550).  For V.90 56k modems, it may be a several
  percent faster with a 16650 (especially if you are downloading
  uncompressed files).  The main advantage of the 16650 is its larger
  buffer size as the extra speed isn't needed unless the modem
  compression ratio is high.  Some 56k internal modems may come with a
  16650 ??

  Non-UART, and intelligent multiport boards use DSP chips to do
  additional buffering and control, thus relieving the CPU even more.
  For example, the Cyclades Cyclom, and Stallion EasyIO boards use a
  Cirrus Logic CD1400 RISC UART, and many boards use 80186 CPUs or even
  special RISC CPUs, to handle the serial IO.

  Most newer PC's (486's, Pentiums, or better) come with 16550A's
  (usually called just 16550's).  If you have something really old the
  chip may unplug so that you may be able to upgrade by buying a 16550A
  chip and replacing your existing 16450 UART.  If the functionality has
  been put on another type of chip, you are out of luck.  If the UART is
  socketed, then upgrading is easy (if you can find a replacement).  The
  new and old are pin-to-pin compatible.  It may be more feasible to
  just buy a new serial board on the Internet (few retail stores stock
  them today).


  12.  Troubleshooting

  12.1.  Software



  ╖  modemstat and statserial show the current state of various modem
     signal lines (such as DTR, CTS, etc.)

  ╖  irqtune will give serial port interrupts higher priority to improve
     performance.

  ╖  hdparm for hard-disk tuning may help some more.


  12.2.  My Modem is Physically There but Can't be Found

  For the PCI bus look at /proc/pci.  Otherwise see ``What is the
  current IO address and IRQ of my Serial Port ?''

  The "wvdialconf" program will try to find a modem on your serial
  ports.  See ``What is wvdialconf ?''  Your problem could be due to a
  winmodem (or the like) which can't be used with Linux.  See ``Avoid:
  winmodems''.  The "setserial program may be used to detect serial
  ports but will  not detect modems on them.  Thus "wvdialconf" is best
  to try first.

  You also must type a command for each IO address you probe.  Thus it's
  harder to use but you can try any IO address while wvdialconf likely
  trys only the most common IO addresses.  With "setserial" you must
  give the "autoconfig" argument at the IO address you think the modem
  is at.  If it shows "unknown" for UART type there may be nothing
  there.  See ``What is setserial''.

  Here are some common mistakes people make:

  ╖  setserial: They run it (without the "autoconfig" option) or see it
     displayed on the screen at boot-time, and erroneously think that
     the result shows how their hardware is actually configured.

  ╖  /proc/interrupts: When their modem isn't in use they don't see
     their modem's interrupt there, and erroneously conclude that their
     modem can't be found (or doesn't have an interrupt set).

  ╖  /proc/ioports: People think this shows the hardware configuration
     when it only shows about the same data (possibly erroneous) as
     setserial.



  12.3.  "Operation not supported by device" (error message) for ttySx

  This means that an operation requested by setserial, stty, etc.
  couldn't be done because the kernel doesn't support doing it.
  Formerly this was often due to the "serial" module not being loaded.
  But with the advent of PnP, it may likely mean that there is no modem
  at the address where the driver (and setserial) thinks it is.  If
  there is no modem there, commands (for operations) sent to that
  address obviously don't get done.  See ``What is set in my serial port
  hardware?''

  Another reason is that the "serial" module wasn't loaded at the time.
  "lsmod" will show you if it's now loaded but it sometimes is
  automatically loaded when needed so it may be loaded now but wasn't
  loaded when you got the error message.  The "serial" module should be
  listed in the file: /etc/modules.conf.  The actual module should
  reside in: /lib/modules/.../misc/serial.o.


  12.4.  Slow. Text appears on the screen slowly after long delays

  This will happen from the very start of using the modem.  One symptom
  happens when you are manually typing to your modem: You type but the
  screen remains blank (until after several seconds when you finally
  might see what you typed).   Another symptom is that only a few words
  at a time appear on the screen (possibly with missing text).

  This may be due to a mis-set IRQ.  This means that the IRQ used by the
  device driver does not correspond to the IRQ set in hardware (IRQ mis-
  set).  With a mis-set IRQ you may loose received data and get "input
  overrun" error messages (or find them in logs).  See ``Interrupt Mis-
  set'' for more details.

  It could also be an interrupt conflict. See ``Interrupt Conflicts''
  Make sure there are no IRQs being shared.  Check all your boards
  (serial, ethernet, SCSI, etc...).  Make sure the jumper (or PnP)
  settings, and the setserial parameters are correct for all your serial
  devices.  Also check /proc/ioports and /proc/interrupts and /proc/pci
  for conflicts.


  12.5.  Uploading (downloading) files is broken/slow

  Flow control (both at your PC and/or modem) may not be enabled.  If
  you have set a high DTE speed (like 115.2K) then flow from your modem
  to your PC may work OK but a lot of flow in the other direction will
  not all get thru due to the telephone line bottleneck.  This will
  result in many errors and the resending of packets.  It may thus take
  far too long to send a file.  In some cases, files don't make it thru
  at all.  If you're downloading long uncompressed files or web pages
  (and your modem uses data compression) or you've set a low DTE speed,
  then downloading may also be broken due to no flow control.


  12.6.  For Dial-in I Keep Getting "line NNN of inittab invalid"

  Make sure you are using the correct syntax for your version of init.
  The different init's that are out there use different syntax in the
  /etc/inittab file.  Make sure you are using the correct syntax for
  your version of getty.


  12.7.  When I Try To Dial Out, It Says "/dev/ttySN: Device or resource
  busy"

  This problem can arise when DCD or DTR are not implemented correctly.
  DCD should only be on (asserted) when there is an actual connection
  (ie someone has dialed in), not when getty is watching the port.
  Check to make sure that your modem is configured to only assert DCD
  when there is a connection.  DTR should be on (asserted)  whenever
  something is using, or watching the line, like getty, kermit, or some
  other comm program.

  Another common cause of ``device busy'' errors, is that you set up
  your serial port with an interrupt already taken by something else.
  As each device initializes, it asks Linux for permission to use its
  hardware interrupt.  Linux keeps track of which interrupt is assigned
  to whom, and if your interrupt is already taken, your device won't be
  able to initialize properly.  The device really doesn't have much of
  any way to tell you that this happened, except that when you try to
  use it, it will return a ``device-busy'' error.  Check the interrupts
  on all of your boards (serial, ethernet, SCSI, etc.).  Look for IRQ
  conflicts.


  12.8.  I Keep Getting "Getty respawning too fast: disabled for 5 min¡
  utes"

  Make sure your modem is configured correctly.  Look at registers E and
  Q.  This can occur when your modem is chatting with getty.

  Make sure you are calling getty correctly from your /etc/inittab.
  Using the wrong syntax or device names will cause serious problems.

  For uugetty, verify that your /etc/gettydefs syntax is correct by
  doing the following:


       linux# getty -c /etc/gettydefs





  This can also happen when the uugetty initialization is failing.  See
  section ``uugetty Still Doesn't Work''.



  12.9.  My Modem is Hosed after Someone Hangs Up, or uugetty doesn't
  respawn


  This can happen when your modem doesn't reset when DTR is dropped.
  Greg Hankins saw his RD and SD LEDs go crazy when this happened.  You
  need to have your modem reset.  Most Hayes compatible modems do this
  with &D3, but on his USR Courier, he had to set &D2 and S13=1.  Check
  your modem manual (if you have one).


  12.10.  uugetty Still Doesn't Work

  There is a DEBUG option that comes with getty_ps.  Edit your config
  file /etc/conf.{uu}getty.ttySN and add DEBUG=NNN.  Where NNN is one of
  the following combination of numbers according to what you are trying
  to debug:


       D_OPT   001            option settings
       D_DEF   002            defaults file processing
       D_UTMP  004            utmp/wtmp processing
       D_INIT  010            line initialization (INIT)
       D_GTAB  020            gettytab file processing
       D_RUN   040            other runtime diagnostics
       D_RB    100            ringback debugging
       D_LOCK  200            uugetty lockfile processing
       D_SCH   400            schedule processing
       D_ALL   777            everything




  Setting DEBUG=010 is a good place to start.

  If you are running syslogd, debugging info will appear in your log
  files.  If you aren't running syslogd info will appear in
  /tmp/getty:ttySN for debugging getty and /tmp/uugetty:ttySN for
  uugetty, and in /var/adm/getty.log.  Look at the debugging info and
  see what is going on.  Most likely, you will need to tune some of the
  parameters in your config file, and reconfigure your modem.

  You could also try mgetty.  Some people have better luck with it.


  13.  Flash Upgrades

  Many modems can be upgraded by reprogramming their flash memories with
  an upgrade program which you get from the Internet.  By sending this
  "program" via the serial port to the modem, the modem will store this
  program in its non-volatile memory (it's still there when the power is
  turned off).  The instructions on installing it are usually on how to
  do in under Windows so you'll need to figure out how to do the
  equivalent under Linux (unless you want to install the upgrade under
  Windows).

  If the latest version of this HOWTO still contains this request (see
  ``New Versions of this HOWTO'') please send me your experiences with
  installing such upgrades that will be helpful to others.

  If you need to send a file (program) to your modem, how do you do it?
  First, there may be a command that you need to send your modem to tell
  it that what follows is a flash ROM upgrade.  In one case this was
  AT**  Next, you need to send the file directly to the modem.
  Communication programs often use zmodem or kermit to send files to the
  modem (and beyond) but these put the file into packets which append
  headers and you want the exact file, not a modified one.  But the
  kermit program has a "transmit" command that will send the file
  directly (without using the kermit packets) so this is one way to send
  a file directly.  Another way would be to escape from the
  communication program to the shell (in minicom this is ^aj) and then:
  cat upgrade_file_name > /dev/ttyS2 .  Then go back to the
  communication program (type fg at the command line prompt in minicom)
  to see what happened.


  14.  Problems Explained

  While the section ``Troubleshooting'' lists problems by symptom, this
  section explains what will happen if something is set incorrectly.
  This section helps you understand what caused the symptom and what
  other symptoms might be due to the same problem.


  14.1.  Interrupt Mis-set

  If you don't understand what an interrupt does see ``Interrupts''.  If
  a serial port has one IRQ set in the hardware but a different one set
  in the device driver, the device driver will not receive any
  interrupts sent by the serial port.  Since the serial port uses
  interrupts to tell its driver when it needs service (fetching bytes
  from it's 16-byte receive buffer or putting another 16-bytes in its
  transmit buffer) one might expect that the serial port would not work
  at all.

  But it still may work anyway --sort of.  Why?  Well, besides the
  interrupt method of servicing the port there's a polling method that
  doesn't need interrupts.  The way it works is that every so often the
  device driver checks the serial port to see if it needs anything such
  as if it has some bytes that need fetching from its receive buffer.
  If interrupts don't work, the serial driver falls back to this polling
  method.  But this polling method was not intended to be used a
  substitute for interrupts.   It's so slow that it's not practical to
  use and may cause buffer overruns.  Its purpose may have been to get
  things going again if just one interrupt is lost or fails to do the
  right thing.  It's also useful in showing you that interrupts have
  failed.

  For the 16-byte transmit buffer, 16 bytes will be transmitted and then
  it will wait until the next polling takes place (several seconds
  later) before the next 16 bytes is sent out.  Thus transmission is
  very slow and in small chunks.  Receiving is slow too since bytes that
  are received by the receive buffer are likely to remain there for
  several seconds until it is polled.

  This explains why it takes so long before you see what you typed.
  When you type say AT to the modem, the AT goes out the serial port to
  the modem.  The modem then echos the AT back thru the serial port to
  the screen.  Thus the AT characters had to pass twice thru the serial
  port.  Normally this happens so fast that AT seems to appear on the
  screen at the same time that you hit the keys on the keyboard.  With
  polling delays thru the serial port, you don't see what you typed
  until many seconds later.

  What about overruns of the 16-byte receive buffer?  This will happen
  with an external modem since the modem just sends to the serial port
  at high speed which is likely to overrun the 16-byte buffer.  But for
  an internal modem, the serial port is on the same card and it's likely
  to check that this receive buffer has room for more bytes before
  putting received bytes into it.  In this case there will be no overrun
  of this receive buffer, but text will just appear on your screen in
  16-byte chunks at intervals of several seconds.

  Even with an external modem you might not get overruns.  If just a few
  characters (under 16) are sent you don't get overruns since the buffer
  likely has room for them.  But attempts to send a larger number of
  bytes from your modem to your screen may result in overruns.  However,
  more than 16 (with no gaps) can get thru OK if the timing is right.
  For example, if 32 bytes were received (and no more bytes followed),
  the polling might just happen after the first 16 bytes had been
  received.  Then there would be space for the next 16 bytes so that 32
  bytes gets thru OK.  Similar conditions might pass between 16 to 31
  bytes thru OK.  But it's also likely that only an occasional 16-byte
  chunk will get thru and huge gaps of missing data will be lost.

  If you have an obsolete serial port with only a 1-byte buffer (or it's
  been incorrectly set to work like a 1-byte buffer) then the situation
  will be much worse than described above and only one character will
  occasionally make it thru the port.  This character is likely to be
  just a line-feed since this is often the last character to be
  transmitted in a burst of characters sent to your screen.  Thus you
  may type AT to the modem but never see AT on the screen.  All you see
  several seconds later is that the cursor drops down one line.  This
  has happened to me even with a 16-byte buffer that was somehow
  behaving like a 1-byte buffer.

  When a communication program starts up, it expects interrupts to be
  working.  It's not geared to using this slow polling-like mode of
  operation.  Thus all sorts of mistakes may be made such as setting up
  the serial port and/or modem incorrectly.  It may fail to realize when
  a connection has been made.  If a script is being used for login, it
  may fail (caused by timeout) due to the polling delays.


  14.2.  Interrupt Conflicts

  When two devices have the same IRQ number it's called sharing
  interrupts.  Under some conditions this sharing works out OK.
  Starting with kernel version 2.2, serial ports may share interrupts
  with other serial ports.  Devices on the PCI bus may share the same
  IRQ interrupt with other devices on the PCI bus.  In other cases where
  there is potential for conflict, there should be no problem if no two
  devices with the same IRQ are ever "in use" at the same time.   More
  precisely, "in use" really means "open" (in programmer jargon).  In
  cases other than the exceptions mentioned above (unless special
  software permits sharing), sharing is not allowed and conflicts arise
  if sharing is attempted.

  Even if two processes with conflicting IRQs run at the same time, one
  of the devices will likely have its interrupts sent to its device
  driver and will work OK.  The other device will not have its
  interrupts sent to the correct driver and will likely behave just like
  a process with mis-set interrupts.  See ``Interrupt Mis-set'' for more
  details.


  15.  Other Sources of Information

  15.1.  Misc


  ╖  man pages for: agetty(8), getty(1m), gettydefs(5), init(1),
     isapnp(8), login(1), mgetty(8), setserial(8)

  ╖  Your modem manual (if it exists).  Some modems come without
     manuals.

  ╖  The Linux serial mailing list.  To join, send email to
     [e-mail], with ``subscribe linux-serial'' in the
     message body.  If you send ``help'' in the message body, you get a
     help message.  The server also serves many other Linux lists.  Send
     the ``lists'' command for a list of mailing lists.
  15.2.  Books

  I've been unable to find a good up-to-date book on modems.

  ╖  The Complete Modem Reference by Gilbert Held, 1997.  Contains too
     much info about obsolete topics.  More up-to-date info may be found
     on the Internet.

  ╖  Modems For Dummies by Tina Rathbone, 1996.  (Have never seen it.)


  15.3.  HOWTOs



  ╖  Cable-Modem mini-howto

  ╖  ISDN Howto (not a LDP Howto)
     <http://www.suse.de/Support/sdb_e/isdn.html>: drivers for ISDN
     "Modems"; Much related info on this is in German

  ╖  Modems-HOWTO: In French (Not used in creating this Modem-HOWTO)

  ╖  NET-3-HOWTO: all about networking, including SLIP, CSLIP, and PPP

  ╖  PPP-HOWTO: help with PPP including modem set-up

  ╖  Serial-HOWTO has info on Multiport Serial Cards used for both
     terminals and banks of modems.  Technical info on the serial port
     will appear in the next revision of it.

  ╖  Serial-Programming-HOWTO: for some aspects of serial-port
     programming

  ╖  Text-Terminal-HOWTO: (including connecting up with modems)

  ╖  UUCP-HOWTO: for information on setting up UUCP


  15.4.  Usenet newsgroups



  ╖  comp.os.linux.answers FAQs, How-To's, READMEs, etc. about Linux.

  ╖  comp.os.linux.hardware Hardware compatibility with the Linux
     operating system.

  ╖  comp.os.linux.setup Linux installation and system administration.

  ╖  comp.dcom.modems Modems for all OS's


  15.5.  Web Sites



  ╖  Hayes AT modem commands Technical Reference for Hayes (tm) Modem
     Users <http://www.hayes.com/TechSupport/techref/>

  ╖  Rockwell-based modem commands
     <http://www.rss.rockwell.com/techinfo/>

  ╖  A white paper discussing serial communications and multiport serial
     boards is available from Cyclades at http://www.cyclades.com.

  ╖  Modem FAQs:
     Navas 28800 Modem FAQ
     <http://web.aimnet.com/~jnavas/modem/faq.html>

  ╖  Curt's High Speed Modem Page
     <http://www.teleport.com/~curt/modems.html>

  ╖  Much info on 56k modems 56k Modem = v.Unreliable
     <http://808hi.com/56k/>

  ╖  Links to modem manufacturers
     <http://www.56k.com/links/Modem_Manufacturers/>

  ╖  Identifying modems by FCC ID
     <http://www.sbsdirect.com/fccenter.html>

  ╖  Partial list of  modems which work/don't_work under Linux modem
     list <http://www.o2.net/~gromitkc/winmodem.html>


  16.  Appendix A:  How Modems Work (technical) (unfinished)

  16.1.  Modulation Details

  16.1.1.  Intro to Modulation

  Modulation is the conversion of a digital signal represented by binary
  bits (0 or 1) into an analog signal something like a sine wave.  The
  modulated signal consists pure sine wave "carrier" signal which is
  modified to convey information.  A pure carrier sine wave, unchanging
  in frequency and voltage, provides no flow of information at all
  (except that a carrier is present).  To make it convey information we
  modify (or modulate) this carrier.  There are 3 basic types of
  modulation: frequency, amplitude, and phase.  They will be explained
  next.


  16.1.2.  Frequency Modulation

  The simplest modulation method is frequency modulation.  Frequency is
  measured in cycles per second (of a sine wave).  It's the count of the
  number of times the sine wave shape repeats itself in a second.  This
  is the same as the number of times it reaches it peak value in a
  second.  The word "Hertz" (abbreviated Hz) is used to mean "cycles per
  second".

  A simple example of frequency modulation is where one frequency means
  a 0 and another means a 1.  For example, for some obsolete 300 baud
  modems 1070 Hz meant a binary 0 while 1270 Hz meant a binary 1.  This
  was called "frequency shift keying".  Instead of just two possible
  frequencies, more could be used to allow more information to be
  transmitted.  If we had 4 different frequencies (call them A, B, C,
  and D) then each frequency could stand for a pair of bits.  For
  example, to send 00 one would use frequency A.  To send 01, use
  frequency B; for 10 use C; for 11 use D.  In like manner, by using 8
  different frequencies we could send 3 bits with each shift in
  frequency.  Each time we double the number of possible frequencies we
  increase the number of bits it can represent by 1.


  16.1.3.  Amplitude Modulation

  Once one understands frequency modulation example above including the
  possibilities of representing a few bits by a single shift in
  frequency, it's easier to understand both amplitude modulation and
  phase modulation.  For amplitude modulation, one just changes the
  height (voltage) of the sine wave analogous to changing the frequency
  of the sine wave.  For a simple case there could only be 2 allowed
  amplitude levels, one representing a 0-bit and another representing a
  1-bit.  As explained for the case of frequency modulation, having more
  possible amplitudes will result in more information being transmitted.


  16.1.4.  Phase Modulation

  To change the phase of a sine wave at a certain instant of time, we
  stop sending this old sine wave and immediately begin sending a new
  sine wave of the same frequency and amplitude.  If we started sending
  the new sine wave at the same voltage level (and slope) as existed
  when we stopped sending the old sine wave, there would be no change in
  phase (and no detectable change at all).  But suppose that we started
  up the new sine wave at a different point on the sine wave curve.
  Then there would likely be a sudden voltage jump at the point in time
  where the old sine wave stopped and the new sine wave began.  This is
  a phase shift and it's measured in degrees (deg.)  A 0 deg. (or a 360
  deg.) phase shift means no change at all while a 180 deg. phase shift
  just reverses the voltage (and slope) of the sine wave.  Put another
  way, a 180 deg. phase shift just skips over a half-period (180 deg.)
  at the point of transition.  Of course we could just skip over say 90
  deg. or 135 deg. etc.  As in the example for frequency modulation, the
  more possible phase shifts, the more bits a single shift in phase can
  represent.


  16.1.5.  Combination Modulation

  Instead of just selecting either frequency, amplitude, or phase
  modulation, we may chose to combine modulation methods.  Suppose that
  we have 256 possible frequencies and thus can send a byte (8 bits) for
  each shift in frequency (since 2 to the 8 power is 256).  Suppose also
  that we have another 256 different amplitudes so that each shift in
  amplitude represents a byte.  Also suppose there are 256 possible
  phase shifts.  Then a certain points in time we may make a shift in
  all 3 things: frequency, amplitude and phase.  This would send out 3
  bytes for each such transition.

  No modulation method in use today actually does this.  It's not
  practical due to the relatively long time it would take to detect all
  3 types of changes.  But what is quite common is the simultaneous
  change in both phase and amplitude.  This is called phase-amplitude
  modulation (sometimes also called quadrature amplitude modulation =
  QAM).  This method is used for the common modem speeds of 14.4k,
  28.8k, and 33.6k.  The only significant case where this modulation
  method is not used today is for 56k modems.  But even 56k modems
  exclusively use QAM (phase-amplitude modulation) in the direction from
  your PC out the telephone line.  Sometimes even the other direction
  will also fall back to QAM when line conditions are not good enough.
  Thus QAM (phase-amplitude modulation) still remains the most widely
  used method on ordinary telephone lines.


  16.1.6.  56k Modems (v.90)

  The modulation method used above 33.6k is entirely different than the
  common phase-amplitude modulation.  The details of exactly how it
  works seem to be obscure and I couldn't find them on the Internet as
  of late 1998.  But the basic idea behind it is easy to understand.
  Since ordinary telephone calls are converted to digital signals at the
  local offices of the telephone company, the fastest speed that you can
  send digital data by an ordinary telephone call is the same speed that
  the telephone company uses over its digital portion of the phone call
  transmission.  What is this speed?  Well, in the USA it's exactly 56k!
  In other countries it may be slightly higher.

  Thus in the USA 56k is the absolute top speed possible for an ordinary
  telephone call using the digital portion of the circuit that was
  designed to send digital encodings of the human voice.  In order to
  use 56k, the modem must know exactly how the telephone company is
  doing its digital encoding of the analog signals.  This task is far
  too complicated if both sides of a telephone call have only an analog
  interface to the telephone company.  But if one side has a digital
  interface, then it's possible (at least in one direction).  Thus if
  your ISP has a digital interface to the phone company, the ISP may
  send out a certain digital signal over the phone lines toward your PC.
  The digital signal from the ISP gets converted to analog at the local
  telephone office near your PC's location (perhaps near your home).
  Then it's your modem's task to try to figure out exactly what that
  digital signal was.  If it can do this then transmission at 56k (the
  speed of the telephone company's digital signal) is possible in this
  direction.

  What method does the telephone company use to digitally encode analog
  signals?  It uses a method of sampling the amplitude of the analog
  signal at a rate of 8000 samples per second.  Each sample amplitude is
  encoded as a 7-bit (ASCII-like) byte.  (Note: 7 x 8000 = 56k)  This is
  called "Pulse Code Modulation" = PCM.  These bytes are then sent
  digitally on the telephone company's digital circuits where many calls
  share a single circuit using a time-sharing scheme known as "time
  division multiplexing".  Then finally at the local telephone office
  near your home, the digital signal is de-multiplexed resulting in the
  same digital signal as was originally created by PCM.  This signal is
  then converted back to analog and sent to your home.  Each 7-bit byte
  creates a certain amplitude of the analog signal.  Your modem's task
  is to determine just what that PCM 7-bit byte was based on the
  amplitude it detects.

  This is (sort of) "amplitude demodulation" but not really.  It's not
  amplitude demodulation because there is no carrier.  Actually, it's
  called "modulus conversion" which is the inverse of PCM.  In order to
  determine the digital codes the telephone Co. used to create the
  analog signal, the modem must sample this analog signal amplitude at
  exactly the same points in time the phone Co. used when it created the
  analog signal.  In order to get the modem to do this correctly the
  modem must go thru "training" periods where the ISP's modem sends out
  known digital signals and the modem trains itself to recognize those
  signals.  (At least that's the way I think it works ??)

  Note that the digital part of the telephone network is bi-directional.
  Two such circuits are used for a phone call, one in each direction.
  Also, while 7-bit bytes are used to encode the amplitude, the bytes
  sent are 8-bit ones with the extra bit used by the telephone company
  for its signalling purposes.  The telephone users have no control over
  this extra bit.  This means that while the digital signal is actually
  64k bits/sec, only 56k can be controlled by the user.


  17.  Appendix B:  "baud" vs. "bps"

  17.1.  A simple example

  ``baud'' and ``bps'' are perhaps one of the most misused terms in the
  computing and telecommunications field.  Many people use these terms
  interchangeably, when in fact they are not!  bps is simply the number
  of bits transmitted per second.  The baud rate is a measure of how
  many times per second a signal changes (or could change).  For a
  typical serial port a 1-bit is -12 volts and a 0-bit is +12 v (volts).
  If the bps is 38,400 a sequence of 010101... would also be 38,400 baud
  since the voltage shifts back and forth from positive to negative to
  positive ... and there are 38,400 shifts per second.  For another
  sequence say 111000111... there will be fewer shifts of voltage since
  for three 1's in sequence the voltage just stays at -12 volts yet we
  say that its still 38,400 baud since there is a possibility that the
  number of changes per second will be that high.

  Looked at another way, put an imaginary tic mark separating each bit
  (even thought the voltage may not change).  38,400 baud then means
  38,400 tic marks per second.  The tic marks at at the instants of
  permitted change and are actually marked by a synchronized clock
  signal generated in the hardware but not sent over the external cable.

  Suppose that a "change" may have more than the two possible outcomes
  of the previous example (of +- 12 v).  Suppose it has 4 possible
  outcomes, each represented by a unique voltage level.  Each level may
  represent a pair of bits (such as 01).  For example, -12v could be 00,
  -6v 01, +6v 10 and +12v 11.  Here the bit rate is double the baud
  rate.  For example, 3000 changes per second will generate 2 bits for
  each change resulting in 6000 bits per second (bps).  In other words
  3000 baud results in 6000 bps.


  17.2.  Real examples

  The above example is overly simple.  Real examples are more
  complicated but based on the same idea.  This explains how a modem
  running at 2400 baud, can send 14400 bps (or higher).  The modem
  achieves a bps rate greater than baud rate by encoding many bits in
  each signal change (or transition).  Thus, when 2 or more bits are
  encoded per baud, the bps rate exceeds the baud rate. If your modem-
  to-modem connection is at 14400 bps, it's going to be sending 6 bits
  per signal transition at 2400 baud.  A speed of 28800 bps is obtained
  by 3200 baud at 9 bits/baud.  When people misuse the word baud, they
  may mean the modem speed (such as 33.6K).

  Common modem bps rates were formerly  50, 75, 110, 300, 1200, 2400,
  9600.  These were also the bps rates over the serial_port-to-modem
  cables.  Today the bps modem-to-modem rates are 14.4K, 28.8K, 33.6K,
  and 56K, but the rates over the serialPort-to-modem cables are not the
  same but are: 19.2K, 38.4K, 57.6K and 115.2K).  Using modems with
  V.42bis compression (max 4:1 compression), rates up to 115.2K bps are
  possible for 33.6K modems (230.4K is possible for 56K modems).

  Except for the 56k modems, most modems run at 2400, 3000, or 3200
  baud.  Because of the bandwidth limitations on voice-grade phone
  lines, baud rates greater than 2400 are harder to achieve, and only
  work under conditions of pristine phone line quality.

  How did this confusion between bps and baud start?  Well, back when
  antique low speed modems were high speed modems, the bps rate actually
  did equal the baud rate.  One bit would be encoded per phase change.
  People would use bps and baud interchangeably, because they were the
  same number.  For example, a 300 bps modem also had a baud rate of
  300. This all changed when faster modems came around, and the bit rate
  exceeded the baud rate.  ``baud'' is named after Emile Baudot, the
  inventor of the asynchronous telegraph printer.


  18.  Appendix C:  Terminal Server Connection

  This section was adapted from Text-Terminal-HOWTO.

  A terminal server is something like an intelligent switch that can
  connect many modems (or terminals) to one or more computers.  It's not
  a mechanical switch so it may change the speeds and protocols of the
  streams of data that go thru it.  A number of companies make terminal
  servers: Xyplex, Cisco, 3Com, Computone, Livingston, etc.  There are
  many different types and capabilities.  Another HOWTO is needed to
  compare and describe them (including the possibility of creating your
  own terminal server with a Linux PC).  Most are used for modem
  connections rather than directly connected terminals.

  One use for them is to connect many modems (or terminals) to a high
  speed network which connects to host computers.  Of course the
  terminal server must have the computing power and software to run
  network protocols so it is in some ways like a computer.  The terminal
  server may interact with the user and ask what computer to connect to,
  etc. or it may connect without asking.  One may sometimes send jobs to
  a printer thru a terminal server.

  A PC today has enough computing power to act like a terminal server
  except that each serial port should have its own hardware interrupt.
  PC's only have a few spare interrupts for this purpose and since they
  are hard-wired you can't create more by software.  A solution is to
  use an advanced multiport serial card which has its own system of
  interrupts (or on lower cost models, shares one of the PC's interrupts
  between a number of ports).  See Serial-HOWTO for more info.  If such
  a PC runs Linux with getty running on many serial ports it might be
  thought of as a terminal server.  It is in effect a terminal server if
  it's linked to other PC's over a network and if its job is mainly to
  pass thru data and handle the serial port interrupts every 14 (or so)
  bytes.  Software called "radius" is sometimes used.

  Today real terminal servers serve more than just terminals.  They also
  serve PC's which emulate terminals, and are sometimes connected to a
  bank of modems connected to phone lines.  Some even include built-in
  modems.  If a terminal (or PC emulating one) is connected directly to
  a modem, the modem at the other end of the line could be connected to
  a terminal server.  In some cases the terminal server by default
  expects the callers to use PPP packets, something that real text
  terminals don't generate.


  19.  Appendix D:  Other Types of Modems

  This HOWTO currently only deals with the common type of modem used to
  connect PC's to ordinary analog telephone lines.  There are various
  other types of modems, including devices called modems that are not
  really modems.


  19.1.  Digital-to-Digital "Modems"

  The standard definition of a modem is sometimes broadened to include
  "digital" modems.  Today direct digital service is now being provided
  to many homes and offices so a computer there sends out digital
  signals directly (well almost) into the telephone lines.  But a device
  is still needed to convert the computer digital signal into type
  allowed on telephone circuits and this device is sometimes called a
  modem.  The next 2 sections: ISDN and DSL concern digital-to-digital
  "modems".


  19.2.  ISDN "Modems"

  The "modem" is  really a Terminal Adapter (TA).  A Debian package
  "isdnutils" is available.  There is a ISDN Howto in German with an
  English translation:  <http://www.suse.de/Support/sdb_e/isdn.html>.
  It's put out by the SuSE distribution of Linux and likely is about
  drivers available in that distribution.  There is an isdn4linux
  package and a newsgroup: de.alt.comm.isdn4linux.  Many of the postings
  are in German.  You might try using a search engine (such as DejaNews)
  to find "isdn4linux".


  19.3.  Digital Subscriber Line (DSL)

  DSL uses the existing twisted pair line from your home (etc.) to the
  local telephone office.  This can be used if your telephone line can
  accept higher speeds than an ordinary modem (say 56k) sends over it.
  It replaces the analog-to-digital converter at the local telephone
  office with a converter which can accept a much faster flow of data
  (in a different format of course).  The device which converts the
  digital signals from your computer to the signal used to represent
  digital data on the local telephone line is also called a modem.  This
  document presently does not cover the special aspects of these modems.

  END OF Modem-HOWTO