start

UNIX v6源代码 这几乎是最经典的unix版本 unix操作系统设计和莱昂氏unix源代码分析都是用的该版

c.c created above.
This is the major
device switch of each device class (block and character).
There is one line for each device configured in your system
and a null line for place holding for those devices
not configured.
The block special devices are put in first
by executing the following generic command
for each disk or tape drive.
(Note that some of these files already exist
in the directory /dev.
Examine each file with ls-I with \-l flag to
see if the file should be removed.)
.DS
	/etc/mknod /dev/NAME b MAJOR MINOR
.DE
The NAME is selected from the following list:
.DS
c.c	NAME	device
rf	rf0	RS fixed head disk
tc	tap0	TU56 DECtape
rk	rk0	RK03 RK05 moving head disk
tm	mt0	TU10 TU16 magtape
rp	rp0	RP moving head disk
hs	hs0	RS03 RS04 fixed head disk
hp	hp0	RP04 moving head disk
.DE
The major device number is selected by counting the
line number (from zero)
of the device's entry in the block configuration table.
Thus the first entry in the table bdevsw would be
major device zero.
.PP
The minor device is the drive number,
unit number or partition as described
under each device in section IV.
The last digit of the name
(all given as 0 in the table above)
should reflect the minor device number.
For tapes where the unit is dial selectable,
a special file may be made for each possible
selection.
.PP
The same goes for the character devices.
Here the names are arbitrary except that
devices meant to be used
for teletype access should be named
/dev/ttyX, where X is any character.
The files tty8 (console), mem, kmem, null are
already correctly configured.
.PP
The disk and magtape drivers provide a `raw' interface
to the device which provides direct transmission
between the user's core and the device and allows
reading or writing large records.
The raw device counts as a character device,
and should have the name of the corresponding
standard block special file with `r' prepended.
Thus the raw magtape
files would be called /dev/rmtX.
.PP
When all the special files have been created,
care should be taken to change
the access modes (chmod-I)
on these files to appropriate values.
.SH
The Source Disk
.PP
You should now extract the source disk.
This can be done as described above
or the UNIX command dd-I may be used.
The disk image begins at block 4100 on the tape,
so the command
.DS
dd if=/dev/mt0 of=/dev/rk1 count=4000 skip=4100
.DE
might be used to extract the disk
to RK drive 1.
.PP
This disk should be mounted (mount-VIII) on /usr/source; it contains
directories of source code.
In each directory is a Shell file run that
will recompile all the source in the directory.
These run files should be consulted whenever
you need to recompile.
.SH
Floating Point
.PP
UNIX only supports the 11/45 FP11-B
floating point unit.
For machines without this hardware,
there is a user subroutine
available that will catch illegal instruction
traps and interpret floating point operations.
(See fptrap-III.)
The system as delivered has this code included
in all commands that have floating point.
This code is never used if the FP hardware
is available and therefore does not
need to be changed.
The penalty is a little bit of disk space and
loading time for the few floating commands.
.PP
The C compiler in /usr/source/c probably
should be changed if floating point is available.
The fpp flag in c0t.s should be set and
C should be recompiled and reloaded and
installed.
This allows floating point C programs to be
compiled without the \-f flag and prevents
the floating point interpreter from getting
into new floating programs.
(See /usr/source/c/run.)
.SH
Time Conversion
.PP
If your machine is not in the Eastern time zone,
you must edit (ed-I) the subroutine
/usr/source/s4/ctime.c to reflect your local time.
The variable `timezone' should be changed
to reflect the time difference between local time and GMT.
For EST, this is 5*60*60; for PST it would be 8*60*60.
This routine also contains the names of the
standard and Daylight Savings time zone;
so `EST' and `EDT' might be changed to `PST' and `PDT'
respectively.
Notice that these two names are in upper case
and escapes may be needed (tty-IV).
Finally, there is a `daylight'
flag;
when it is 1 it causes the
time to shift to Daylight Savings automatically
between the last Sundays in April and October
(or other algorithms in 1974 and 1975).
Normally this will not have to be reset.
After ctime.c has been edited
it should be compiled and installed in
its library.
(See /usr/source/s4/run.)
Then you should
(at your leisure)
recompile and reinstall
all programs performing time conversion.
These include:
(in s1)
date, dump, ls, cron,
(in s2)
mail, pr, restor, who, sa and tp.
.SH
Disk Layout
.PP
If
there are to be more file systems mounted than just the root,
use mkfs-VIII to create the new file system and
put its mounting in the file /etc/rc (see init-VIII and mount-VIII).
(You might look at /etc/rc anyway to
see what has been provided for you.)
.PP
There are two considerations in deciding how to adjust the arrangement
of things on your disks:
the most important is making sure there is adequate space
for what is required;
secondarily, throughput should be maximized.
The RK disk (or its image)
as distributed has 4000 blocks for file storage,
and the remainder of the disk (872 blocks)
is set aside for swap space.
In our own system, which allows 14 simultaneous users,
this amount of swap space is not quite enough,
so we use 1872 blocks for this purpose;
it is large enough so running out of swap space never
occurs.
.PP
Many common system programs (C, the editor, the assembler etc.)
create intermediate files in the /tmp directory,
so the file system where this is stored also should be made
large enough to accommodate
most high-water marks.
In an idle state, we have about 900 free blocks
on the file system where /tmp resides,
and hit the bottom every few days or so.
(This causes a momentary disruption,
but not a crash, as swap-space runout does.)
All the programs that create files in /tmp try to take
care to delete them, but most are not immune to
events like being hung up upon, and can leave dregs.
The directory should be examined every so often and the old
files deleted.
.PP
Exhaustion of user-file space is certain to occur
now and then;
the only mechanisms for controlling this phenomenon
are occasional use of du-I and threatening
messages of the day and personal letters.
.PP
The efficiency with which UNIX is able to use the CPU
is largely dictated by the configuration of disk controllers.
For general time-sharing applications,
the best strategy is to try to split user files,
the root directory (including the /tmp directory)
and the swap area among three controllers.
In our own system, for example, we have user files on an RP,
the root on an RF fixed-head disk, and swap on an RK.
This is best for us since the RK has a faster transfer
rate than the rather slow RF, and in swapping
the transfer rate rather than access time
is the dominant influence on throughput.
.PP
Once you have decided how to make best use
of your hardware, the question is how to initialize it.
If you have the equipment,
the best way to move a file system
is to dump it (dump-VIII) to magtape,
use mkfs-VIII to create the new file system,
and restore the tape.
If you don't have magtape,
dump accepts an argument telling where to put the dump;
you might use another disk or DECtape.
Sometimes a file system has to be increased in logical size
without copying.
The super-block of the device has a word
giving the highest address which can be allocated.
For relatively small increases, this word can be patched
using the debugger (db-I)
and the free list reconstructed using icheck-VIII.
The size should not be increased very greatly
by this technique, however,
since although the allocatable space will increase
the maximum number of files will not (that is, the i-list
size can't be changed).
Read and understand the description given in file system-VI
before playing around in this way.
.PP
If you have only an RP disk,
see section rp-IV for some suggestions
on how to lay out the information on it.
The file systems distributed on tape, containing the binary, the source,
and the manuals,
are each only 4000 blocks long.
Perhaps the simplest way to integrate the latter two
into a large file system is to extract the
tape into the upper part of the RP,
dump it,
and restore it into an empty, non-overlapping
file system structure.
If you have to merge a file system into another, existing one,
the best bet is to
use ncheck-VIII
to get a list of names, then edit this list
into a sequence of mkdir and cp commands
which will serve as input to the Shell.
(But notice that owner information is
lost.)
.SH
New Users
.PP
Install new users by editing the password file
/etc/passwd (passwd-V).
You'll have to make current directories for the new users
and change their owners to the
newly installed name.
Login as each user to make sure the password
file is correctly edited.
For example:
.DS
ed /etc/passwd
$a
joe::10:1::/usr/joe:
.li
.
w
q
mkdir /usr/joe
chown joe /usr/joe
login joe
ls \-la
login root
.DE
This will make a new login entry for joe.
His default current directory is
/usr/joe
which has been created.
The delivered password file
has the user
.I
ken
.R
in it to be used as a prototype.
.SH
Multiple Users
.PP
If UNIX is to support simultaneous
access from more than just the console teletype,
the file /etc/ttys (ttys-V) has to be edited.
For some historical reason tty8 is the name of the console typewriter.
To add new typewriters be sure the device is configured
and the special file exists, then set
the first character of the appropriate line of /etc/ttys to 1
(or add a new line).
Note that init.c will have to be recompiled if there are to be
more than 20 typewriters.
Also note that if the special file is inaccessible when init tries to create a process
for it, the system will thrash trying and retrying to open it.
.SH
File System Health
.PP
Periodically (say every day or so) and always after a crash,
you should check all the file systems for consistency
(icheck, dcheck-VIII).
It is quite important to execute sync (VIII)
before rebooting or taking the machine down.
This is done automatically every 30 seconds by the update
program (VIII) when a multiple-user system is running,
but you should do it anyway to make sure.
.PP
Dumping of the file system should be done regularly,
since once the system is going it is very easy to
become complacent.
Just remember that our RP controller has failed three times,
each time in such a way that all information
on the disk was wiped out without any
error status from the controller.
Complete and incremental dumps are easily done with
the dump command (VIII) but restoration of individual
files is painful.
Dumping of files by name is best done by
tp (I) but the number of files is limited.
Finally if there are enough drives entire
disks can be copied using cp-I, or preferably with
dd-I using the raw special files and an appropriate
block size.
Note that there is no stand-alone program
with UNIX that will restore any of these formats.
Unless some action has been taken to
prevent destruction of a running version of
UNIX, you can find yourself stranded even though
you have backup.
.bp
.SH
Odds and Ends
.PP
The programs
dump,
icheck, dcheck, ncheck, and df
(source in /usr/source/s1 and /usr/source/s2)
should be changed to
reflect your default mounted file system devices.
Print the first few lines of these
programs and the changes will be obvious.
.PP
If you would like to share any UNIX compatible
software with others,
please let us know about it.
If you find bugs in the software or the
documentation,
again let us know.
.PP
Lastly,
there is a UNIX users' group forming.
To get on their mailing list,
send your name(s) and address to:
.DS
Prof. Melvin Ferentz
Physics Dept.
Brooklyn College of CUNY
Brooklyn, N.Y. 11210
.DE
.PP
.DS
		Good luck.
		Ken Thompson
		Dennis Ritchie
.DE