Linux Keystroke mini-HOWTO (part 1/1)

Linux Keystroke mini-HOWTO (part 1/1)

Post by Zenon Fortun » Wed, 03 Jan 1996 04:00:00

Archive-name: linux/howto/mini/keystroke
Last-modified: 9 May 95


*** The `Linux Keystroke mini-HOWTO' is posted automatically by the
*** Linux HOWTO coordinator, Greg Hankins <>.  Please
*** direct any comments or questions about this HOWTO to the author,
*** Zenon Fortuna <>.

- --- BEGIN Linux Keystroke mini-HOWTO part 1/1 ---

=================== Keystroke-HOWTO, version 2.0 ====================

                        Linux Keystroke HOWTO
                 Zenon Fortuna (

Version: 2.0                                            April 4th '95

        0.      What is "Keystroke-HOWTO" ?
        1.      History of changes
        2.      Short description
        3.      Tools for keyboard driver modification
        4.      Modifying keytable file
        4.0     Example of keytable file modification
        4.1     Temporary modification of the keyboard setup
        4.2     Permanent modification
        5.      Example of the key_macro script
        7.      Further ideas ?

- ----------------------------------------------------------------------------
0.      What is "Keystroke-HOWTO" ?

        This document is for users, who want to assign special action to some
        of keys of the keyboard. The suggested method is to use the loadkeys(1)
        or to modify the defkeymap.c file and relink the kernel.

        This text does NOT discuss remapping of keyboard keys, e.g. Backspace
        or Delete keys. For informations about remapping of keys read the
        Backspace Mini-HOWTO written by Stephen Lee.

        The method described below was tested on Linux 1.2.1 release, packaged
        in the Slackware 2.2.0 distribution.

1.      History of changes

        April 4th '95   Version: 2.0
           Adapted for the Linux 1.2.1:
           - simple changes in suggested modification of the ""
           - modified "key_macro" example with /dev/vcs* screen dumping

        May   7th '94   Version: 1.0
           The initial version of the "Keystroke-HOWTO", which worked for the
           Linux 1.0

2.      Short description

        The Linux virtual terminal and keyboard drivers assume default keyboard
        mapping as defined in the drivers/char/defkeymap.c file of the kernel
        source. The 12 PC keyboard function keys may get strings assigned to
        their action. After pressing any of those function keys, perhaps
        modified with the Alt or Ctrl keys, the current virtual terminal
        adds the specific string to its input and output buffers, in effect
        emulating entry of this string as typed in from the keyboard.

        Setting an appropriate string for chosen function key, we can simplify
        execution of selected command, for example calling a Shell-script
        "/usr/local/bin/key_macro", which we can create and modify as desired.

3.      Tools for keyboard driver modification

        We may use loadkeys(1), dumpkeys(1) and showkey(1):
          The loadkeys(1) utility helps to load new strings into the kernel
        keyboard buffers or prepares the new C-code to modify the kernel.
          The dumpkeys(1) should be used to get the current keyboard mapping
        table for inspection or modification.
          The showkey(1) may assist us to obtain the keycode of the selected
        function key.

        If your Linux system does not have these utilities, you may get them
        via anonymous ftp as kbd-0.89.tar.gz or kbd-0.90.tar.gz package from
      , or
        You should use the GNU tar to extract the needed files.

4.      Modifying keytable file

        Linux kernel includes compiled defkeymap.c code, which is generated
        with the loadkeys(1) utility from a file. Both files
        are included in the src/linux/drivers/char directory.

        We need to modify the file, so let's make a local
        copy of it either by
                # cp
                # dumpkeys >

        There is also a large collection of different keytable files in the
        /usr/lib/kbd/keytables directory, from which "" may be
        used as src/linux/drivers/char/ file on your system.

        The method which uses the dumpkeys(1) utility is recommended, because
        it may happen, that your kernel was already modified or generated
        for you with different file than the one you can find.

        Lets read the contents of the file: there are more
        than 300 lines of code, and we can find 3 groups of declarations:
        The first group consists of lines with the word "keycode", maybe
        prepended with additional words like "alt", "control", etc.
        The second group consists of lines with the word "string".
        The third group consists of lines with the word "compose".

        More about the keytables(5) syntax can be read with
                % man keytables

4.0     Example of keytable file modification

        As an example of assigning a macro-string to a function key stroke,
        let's make the "Ctrl-F1" to call our "/usr/local/bin/key_macro"

        First of all we should find out what is the keycode for the F1 function
        key.  We may use the showkey(1) utility to find the keycode with
        pressing F1.

        Instead we can search for the "F1" string in the "" file
        to find the following line:

        keycode  59 = F1

        This suggests, that the keycode for the F1 function key is 59.
        This line defines also, that after pressing the F1 key the keyboard
        driver would send out the string denoted by the string-code "F1".
        To see the contents of this string, one can search for the "string F1"
        pattern, to find

        string F1 = "\033[[A"

        This means, that after pressing the F1 key, the keyboard driver sends
        the "Esc [ [ A" (without blank spaces).

        We shouldn't change this string, because some applications depend on
        this string as default action of the F1 function key.

        However, we may define the new action for Ctrl-F1, provided it is not
        reserved by your kernel for other special actions.
        To see the mappings of the F1 key modified with the Ctrl-, Shift- or
        other modes, we may inspect the "" file with

                # grep 59

        In case when there is no line with "control keycode 59" we may use
        the Ctrl-F1 without problems. (when a line with the "shift control
        keycode 59" exists it is still OK)

        Let us add a following line to the "" file:

                control keycode 59 = key_symbol

        where the "key_symbol" would define the action of the Ctrl-F1 key.
        The Linux 1.2.* allows a dynamic allocation of strings, but a name
        of the "key_symbol" may be picked up only from a fixed set of names.
        Between other names the key-symbols F1-F246 are allowed.
        For "" on my system the F21 was unused, but you should
        inspect your "" and choose a proper key-symbol.
        So, we may end up with a line

                control keycode 59 = F21

        Now we have to define the contents of the F21, adding a line

                string F21 = "/usr/local/bin/key_macro\n"

        In the summary, we made two changes to the original ""
        file: we declared the new string F21, and we have declared
        that the Ctrl-F1 key will be calling the F21 contents.

4.1     Temporary modification of the keyboard setup

        Having properly modified "" we can copy the changes
        to the kernel keyboard driver, using the loadkeys(1) utility:

        % loadkeys

        The permission to modify the kernel keyboard driver is granted to
        everybody who has the read access to the "/dev/console" device.

        To verify that the intended changes were installed, we can use the
        dumpkeys(1) utility to check the F21 value, for example

        % dumpkeys | grep F21

        We may see:

        keycode 59 = F1         F11     Console_13      F21

        string F21 = "/usr/local/bin/key_macro\012"

        which is OK, because "\012", or LF, is equivalent to "\n".

        Now, pressing "Ctrl-F1" should call the "/usr/local/bin/key_macro"
        Shell-script, as intended.

4.2     Permanent modification

        The changes to the kernel keyboard driver imposed by the loadkeys(1)
        last until the next reboot (or the next call to loadkeys).

        We can modify the /etc/rc.d/rc.local to call the loadkeys with our file as an argument. Instead, we can modify the
        src/linux/drivers/char/defkeymap.c and re-link the kernel with
        new defaults.

        We should not modify the defkeymap.c manually, but rather generate
        it with the loadkeys(1) utility:

        # mv defkeymap.c defkeymap.c.ORIG
        # loadkeys --mktable > defkeymap.c

        Then we should generate the new kernel, essentially changing directory
        to the root of the linux kernel source, and using the make(1).

        Finally, we should use the lilo(1) to install and boot our new kernel.

5.      Example of the key_macro script

        A particularly useful script for simple-key-stroke operation may be
        a Shell-script preparing, or printing, a screen dump.

        This example has changed since the version 1.0 of Linux, because
        of the changes in the Linux kernel, which does not provide the
        ioctl(0,TIOCLINUX) system call anymore.

        To read the virtual console screen dumps one should prepare first
        some device files. As "root" user we may create the following files:

                # mknod /dev/vcs1  c 7 1
                # mknod /dev/vcs2  c 7 2
                # mknod /dev/vcs63 c 7 63

        Of course, it is sufficient to have only the /dev/vcs* files specific
        for the virtual consoles you are using.

        The code below should be regarded as an example of possible
        /usr/local/bin/key_macro file:

# This is an example of useful key_macro script

VT_NUMBER=`tty|cut -c9-`
cp /dev/vcs$VT_NUMBER $FILE
echo SCREEN DUMP saved in $FILE
# Uncomment the line below if you want to print the resulted dump-file
# lpr $FILE


        There is no practical limit on the sum the lengths of all strings
        which we would like to load to the keyboard driver. The previous fixed
        buffer of the length of FUNC_BUFSIZE (set to 512 bytes), has been
        replaced in the Linux 1.2.* by a strategy of dynamic buffer allocation
        ... in chunks of 512 bytes each.

        The most recent copy of the Keystroke-HOWTO can be found in


7.      Further ideas ?

        In case you find anything worth adding to this document, please send
        your comments to -- thanks (zf)

- --- END Linux Keystroke mini-HOWTO part 1/1 ---

Version: 2.6.2
Comment: finger for public key