This document contains instructions for the installation and operation of the 2.11BSD PDP-11\(ua UNIX*** system.
It discusses procedures for installing 2.11BSD UNIX on a PDP-11, including explanations of how to lay out file systems on available disks, how to set up terminal lines and user accounts, how to do system-specific tailoring, and how to install and configure the networking facilities. Finally, the document details system operation procedures: shutdown and startup, hardware error reporting and diagnosis, file system backup procedures, resource control, performance monitoring, and procedures for recompiling and reinstalling system software.
The ``bugs'' address supplied with this release will work for some unknown period of time; make sure the ``Index:'' line of the bug report indicates that the release is ``2.11BSD''. See the sendbug(8) program for more details. All fixes that I make, or that are sent to me, will be posted on USENET, in the news group ``comp.bugs.2bsd''.
This document explains how to install 2.11BSD UNIX for the PDP-11 on your system. This document has been revised several times since the first release of 2.11BSD, most recently in July 1995 to reflect the addition of disk labels to the system. The format of the bootable tape has changed. There is now a standalone disklabel program present. While the system call interface is the same as that of 2.10.1BSD, a full bootstrap from the distribution tape is required because the filesystem has changed to allow file names longer than 14 characters. Also, the 3 byte block number packing scheme used by earlier versions of UNIX for the PDP-11 has been eliminated. Block numbers are always 4 byte longs now.
The procedure for performing a full bootstrap is outlined in chapter 2. The process includes copying a root file system from the distribution tape into a new file system, booting that root filesystem, and then reading the remainder of the system binaries and sources from the archives on the tapes.
As 2.11BSD is not compatible at the filesystem level with previous versions of UNIX on the PDP-11, any upgrade procedure is essentially a full bootstrap. There is a limited ability to access old filesystems which may be used after the system partitions have been loaded from a full bootstrap. It is desirable to recompile most local software after the conversion, as there are changes and performance improvements in the standard libraries.
Binaries from 2.10.1BSD which do not read directories or inode structures may be used but should be recompiled to pick up changes in the standard libraries. Note too, that the portable ASCII format of ar(1) archives is now in place - any local archive files will have to be converted using /usr/old/arcv.
This distribution can be booted on a PDP-11 with 1Mb of memory or more\(ua, separate I&D, and with any of the following disks:
Other disks are supported (RX23, RX33, RX50, RD51) but are not large enough to hold a root filesystem plus a swap partition. The old restriction of using RL02 drives in pairs has been lifted. It is now possible to define a root ('a') partition and a swap partition ('b') and load at least the root filesystem to a single RL02. Discs which are too small to hold even a root filesystem (floppies for example) may be used as data disks or as standalone boot media, but are not useable for loading the distribution. Others, while listed above, are not very well suited to loading the distribution. The RK06/07 drives are hard pressed to even hold the system binaries, much less the sources.
The tape drives supported by this distribution are:
The basic distribution contains the following items:
The distribution does not fit on several standard PDP-11 configurations that contain only small disks. If your hardware configuration does not provide at least 75 Megabytes of disk space you can still install the distribution, but you will probably have to operate without source for the user level commands and, possibly, the source for the operating system.
The root file system now occupies a minimum of 4Mb. If at all possible a larger, 6 or 7Mb, root partition should be defined when using the standalone disklabel program.
If you have the facilities, it is a good idea to copy the magnetic tape(s) in the distribution kit to guard against disaster. The tapes are 9-track 1600 BPI, TK50 or TK25 cartridges and contain some 512-byte records, followed by some 1024-byte records, followed by many 10240-byte records. There are interspersed tape marks; end-of-tape is signaled by a double end-of-file.
The basic bootstrap material is present in six short files at the beginning of the first tape. The first file on the tape contains preliminary bootstrapping programs. This is followed by several standalone utilities (disklabel, mkfs(8), restor(8), and icheck(8)\(ua) followed by a full dump of a root file system (see dump(8)). Following the root file system dump is a tape archive image of /usr except for /usr/src (see tar(1)). Finally, a tape archive of source for include files and kernel source ends the first tape. The second tape contains a tape archive image, also in tar format, of all the remaining source that comes with the system.
The entire distribution (barely) fits on a single TK50 cartridge, references to the second tape should be treated as being the 9th file on the TK50. Many of the programs in /usr/src/new have been tar+compress'd in order to keep the distribution to a single tape.
TAPE 1:
Tape file Record size Records^ Contents
---------------------------------------------------------------------------------------
0 512 1 primary tape boot block
512 1 boot block (some tape boot ROMs go for this copy)
512 69 standalone boot program
1 1024 37 standalone disklabel
2 1024 33 standalone mkfs(8)
3 1024 35 standalone restor(8)
4 1024 32 standalone icheck(8)
5 10240 285 dump of ``root'' file system
6 10240 3368 tar dump of /usr, excepting /usr/src
7 10240 519 tar dump of /usr/src/include and /usr/src/sys
TAPE 2:
Tape file Record size Records^ Contents
-------------------------------------------------------------------------------------
0 10240 4092 tar dump of /usr/src, excepting include and sys
UNIX has a set of names for devices which are different from the DEC names for the devices. The disk and tape names used by the bootstrap and the system are:
MSCP disks and TMSCP tapes include SCSI drives attached to the RQZX1 controller on the PDP-11/93. MSCP disks and TMSCP tapes also include SCSI drives attached to the Emulex UC07 or UC08 Q-BUS controllers on Q-bus systems as well as the UC17 and UC18 controllers on UNIBUS systems.
The standalone system used to bootstrap the full UNIX system uses device names of the form:
Example: if there are two MSCP controllers in the system addressed as 0172150 and 0172154 respectively booting from the controller at 172154 requires that c be given as 1. Booting from the standard CSR of 0172150 would be done by specifying c as 0 or omitting the c parameter. boot automatically detects if the first (standard) CSR is being used. All future references will ignore the c parameter by assuming the default value.
The value y specifies the device or drive unit to use. The z value is interpreted differently for tapes and disks: for disks it is a partition number (0 thru 7) corresponding to partitions 'a' thru 'h' respectively. This should always be zero unless you really know what you are doing. The ability to load a kernel from the swap area is planned for the future but does not presently exist. For tapes z is a file number on the tape.\(ua
In all simple cases, a drive with unit number 0 (determined either by a unit plug on the front of the drive, or jumper settings on the drive or controller) will be called unit 0 in its UNIX file name. file name. If there are multiple controllers, the drive unit numbers will normally be counted within each controller. Thus drives on the the first controller are numbered 0 thru 7 and drives on the second controller are numbered 0 thru 7 on controller 1. Returning to the discussion of the standalone system, recall that tapes also took two integer parameters. In the case of a TE16/TU tape formatter on drive 0, the files on the tape have names ``ht(0,0)'', ``ht(0,1)'', etc. Here ``file'' means a tape file containing a single data stream separated by a single tape mark. The distribution tapes have data structures in the tape files and though the first tape contains only 7 tape files, it contains several thousand UNIX files.
Each UNIX physical disk is divided into 8 logical disk partitions, each of which may occupy any consecutive cylinder range on the physical device. While overlapping partitions are allowed they are not a good idea, being an accident waiting to happen (making one filesystem will destroy the other overlapping filesystems). The cylinders occupied by the 8 partitions for each drive type are specified by the disk label read from the disk.
If no label exists the disk will not be bootable and while the kernel attempts not to damage unlabeled disks (by swapping to or doing a crash dump on a live filesystem) there is a chance that filesystem damage will result if a kernel is loaded from an unlabeled disk.
The standalone disklabel program is used to define the partition tables. Each partition may be used either as a raw data area (such as a swapping area) or to store a UNIX file system. It is mandatory for the first partition on a disk to start at sector offset 0 because the 'a' partition is used to read and write the label (which is at the beginning of the disk). If the drive is to be used to bootstrap a UNIX system then the 'a' partition must be of type 2.11BSD (FS_V71K in disklabel.h) and at least 4Mb is size. A 'b' partition of at least 2-3Mb (4Mb is a good choice if space is available) for swapping is also needed. If a drive is being used solely for data then that drive need not have a 'b' (swap) partition but partition 'a' must still span the first part of the disk. The second partition is used as a swapping area, and the rest of the disk is divided into spaces for additional ``mounted file systems'' by use of one or more additional partitions.
The third ('c') logical partition of each physical disk also has a conventional usage: it allows access to the entire physical device, including the bad sector forwarding information recorded at the end of the disk (one track plus 126 sectors). It is occasionally used to store a single large file system or to access the entire pack when making a copy of it on another. Care must be taken when using this partition not to overwrite the last few tracks and thereby destroying the bad sector information.
Unfortunately while the drivers can follow the rules above the entries in /etc/disktab (disktab(5)) do not. The entries in /etc/disktab are translations of the old partition tables which used to be embedded in the device drivers and are thus probably not suitable for use without editing. In some cases it may be that the 8th ('h') partition is used for access to the entire disk rather than the third ('c') partition. Caution should be observed when using the newfs(8) and disklabel(8) commands.
UNIX makes a distinction between ``block'' and ``raw'' (character) devices. Each disk has a block device interface where the system makes the device byte addressable and you can write a single byte in the middle of the disk. The system will read out the data from the disk sector, insert the byte you gave it and put the modified data back. The disks with the names ``/dev/xx0a'', etc are block devices. There are also raw devices available. These have names like ``/dev/rxx0a'', the ``r'' here standing for ``raw''. Raw devices bypass the buffer cache and use DMA directly to/from the program's I/O buffers; they are normally restricted to full-sector transfers. In the bootstrap procedures we will often suggest using the raw devices, because these tend to work faster. Raw devices are used when making new filesystems, when checking unmounted filesystems, or for copying quiescent filesystems. The block devices are used to mount file systems, or when operating on a mounted filesystem such as the root.
You should be aware that it is sometimes important whether to use the character device (for efficiency) or not (because it wouldn't work, e.g. to write a single byte in the middle of a sector). Don't change the instructions by using the wrong type of device indiscriminately.
The standalone disklabel program must be used to alter the 'a' and 'b'
partitions of a drive being used for a bootable system. This is because
the kernel will not permit an open partition to change size or offset.
The root and and swap partitions are always open when the kernel is
running.
This section explains the bootstrap procedure that can be used to get the kernel supplied with this distribution running on your machine. It is mandatory to do a full bootstrap since the filesystem has changed from 2.10.1BSD to 2.11BSD.
The safest route is to use tar(1) to dump all of your current file systems, do a full bootstrap of 2.11BSD and then restore user files from the backups. There is also an untested version of 512restor(8) available for V7 sites that need to read old dump tapes.
It is also desirable to make a convenient copy of system configuration files for use as guides when setting up the new system; the list of files to save from earlier PDP-11 UNIX systems, found in chapter 3, may be used as a guideline.
2.11BSD restor(8) is able to read and automatically convert to the new on disk directory format dump(8) tapes made under 2.9BSD, 2.10BSD and 2.10.1BSD.
The tape bootstrap procedure used to create a working system involves the following major steps:
Certain of these steps are dependent on your hardware configuration. If your disks require formatting, standard DEC diagnostic utilities will have to be used, they are not supplied on the 2.11BSD distribution tape.
To load the tape bootstrap monitor, first mount the magnetic tape on drive 0 at load point, making sure that the write ring is not inserted. Then use the normal bootstrap ROM, console monitor or other bootstrap to boot from the tape.
NOTE: The boot blocks expect the CSR of the booting controller in r0 and the unit number in r1. boot may be booted from any controller or unit, the earlier restrictions of controller 0 and unit 0 have been lifted.
If no other means are available, the following code can be keyed in and executed at (say) 0100000 to boot from a TM tape drive (the magic number 172526 is the address of the TM-11 current memory address register; an adjustment may be necessary if your controller is at a nonstandard address):
The console should type
ctlr is the controller number that Boot was loaded from. It is 0 unless booting from a non-standard CSR.
drive is the drive unit number.
The part number is disk partition or tapefile number booted from. This will always be 0 for the tape Boot program.
csr is an octal number telling the CSR of the controller from which Boot was loaded.
You are now talking to the tape bootstrap monitor. At any point in the following procedure you can return to this section, reload the tape bootstrap, and restart. Through the rest of this section, substitute the correct disk type for dk and the tape type for tp.
The standalone disklabel program is then run:
If there is an existing label present on dk(0,0) it will be used as the default. To have disklabel create a new default based on its idea of what the drive is select D. Then enter m to modify/edit the label.
The MSCP driver is quite good at identifying drives because it can query the controller. Other drivers (notably the SMD (xp) driver) have to deal with a much wider range of controllers which do not all have the same capabilities for drive identification. When dealing with SMD drives you must know the geometry of the drive so you can verify and correct disklabel's choices.
You can however, if using non-DEC SMD controllers, make things easy for disklabel to determine what type of drive is being used. If your controller offers the choice of RM02 emulation you should select that choice. The standalone xp driver uses RM02 as the indication that drive identification capabilities not offered by DEC controllers are present. The driver will be able to determine the geometry of the drive in this case. This is optional because you can explicitly specify all of the parameters to the standalone disklabel program.
A full description of the standalone disklabel program is in Appendix B of this document.
Now create the root file system using the following procedure.\(ua
The size of the root ('a') filesystem was assigned in step 2 (creating the disk label). mkfs will not allow a filesystem to be created if there is not a label present or if the partition size is 0. mkfs looks at partition 0 ('a') in the disklabel for the root file system size.
Finally, determine the proper interleaving factors m and n for your disk. Extensive testing has demonstrated that the choice of m is non critical (performance of a file system varying only by 3 to 4% for a wide range of m values). Values for m within the range from 2 to 5 give almost identical performance. Increasing m too much actually causes degraded performance because the free blocks are too far apart. Slower processors (such as the 73 and 44) may want to start with a m of 3 or 4, faster processors (such as the 70 and 84) may start with a m of 2 or 3. On the other hand, the n value is moderately important. It should be the number of filesystem blocks contained by one cylinder of the disk, calculated by dividing the number of sectors per cylinder by 2, rounding down if needed. (This is what mkfs does by default, based on the geometry information in the disk label.) These numbers determine the layout of the free list that will be constructed; the proper interleaving will help increase the speed of the file system.
The number of bytes per inode determines how many inodes will be allocated in the filesystem. The default of 4096 bytes per inode is normally enough (resulting in about 2000 inodes for a 8mb root filesystem and 1000 inodes for the 4mb distribution ``generic'' root filesystem). If more inodes are desired then a lower value (perhaps 3072) should be specified when prompted for the number of bytes per inode.
Then run the standalone version of the mkfs (8) program. The values in square brackets at the size prompt is the default from the disklabel. Simply hit a return to accept the default. mkfs will allow you to create a smaller filesystem but you can not enter a larger number than the one in brackets. In the following procedure, substitute the correct types for tp and dk and the size determined above for size:
The number nnnn is the device number of the device (high byte is the major device number and the low byte is the unit number). The mmmmmm number is the CSR of the device. This information is mainly used as a reminder and diagnostic/testing purposes.
You now have an empty UNIX root file system.
To restore the root file system onto it, type
This takes about 8 minutes with a TZ30 on a 11/93 and about 15 minutes using a TK50 on a 11/73.
You are now ready to boot from disk. Type:
UNIX itself then runs for the first time and begins by printing out a banner identifying the release and version of the system that is in use and the date that it was compiled.
Next the mem messages give the amount of real (physical) memory, the amount of memory left over after the system has allocated various data structures, and the amount of memory available to user programs in bytes.
The information about different devices being attached or not being found is produced by the autoconfig(8) program. Most of this is not important for the moment, but later the device table, /etc/dtab, can be edited to correspond to your hardware. However, the tape drive of the correct type should have been detected and attached.
The ``erase ...'' message is part of /.profile that was executed by the root shell when it started. This message is present to remind you that the character erase, line erase, and interrupt characters are set to what is standard for DEC systems; this insures that things are consistent with the DEC console interface characters.
UNIX is now running single user on the installed root file system, and the `UNIX Programmer's Manual' applies. The next section tells how to complete the installation of distributed software on the /usr file system. The `#' is the prompt from the shell, and lets you know that you are the super-user, whose login name is ``root''.
The disk with the new root file system on it will not be bootable directly until the block 0 bootstrap program for your disk has been installed. There are copies of the bootstraps in /mdec. Use dd(1) to copy the right boot block onto block 0 of the disk.
Once this is done, booting from this disk will load and execute the block 0 bootstrap, which will in turn load /boot. /boot will print on the console:
As distributed /boot will load dk(0,0)unix by default if a carriage return is typed at the : prompt.
NOTE: NONE the primary bootstraps have a prompt or alternate program name capability because of space considerations. No diagnostic message results if the file cannot be found.
First set a shell variable to the name of your disk, so the commands used later will work regardless of the disk you have; do one of the following:
The next thing to do is to extract the rest of the data from the tape. You might wish to review the disk configuration information in section 4.3 before continuing; you will have to select a partition to restore the /usr file system into which is at least 25 Megabytes in size (this is just barely enough for the system binaries and such and leaves no room for the system source.)\(ua
In the command below part is the partition name (a-h) for the partition which will hold /usr.
If you have an existing/old password file to be merged back into 2.11BSD, special steps are necessary to convert the old password file to the shadow password file format (shadow password file and password aging were ported from 4.3BSD and are standard in 2.11BSD ).
The first tape has been been completely loaded. You can check the consistency of the /usr file system by doing
If there are inconsistencies in the file system, you may be prompted to apply corrective action; see the document describing fsck for information.
To use the /usr file system, you should now remount it by saying
You can then extract the source code for the commands from the second distribution tape\(ua (with the exception of RK07's, RM03's, and RD52's and other small disks this will fit in the /usr file system):
After setting up the new 2.11BSD filesystems, you may restore the user files that were saved on tape before beginning the conversion. Note that the 2.11BSD restor program does its work by accessing the raw file system device and depositing inodes in the appropriate locations on disk. This means that file system dumps might not restore correctly if the characteristics of the file system have changed (eg. if you're restoring a dump of a file system into a file system smaller than the original.) To restore a dump tape for, say, the /u file system something like the following would be used:
If tar images were written instead of doing a dump, you should be sure to use the `p' option when reading the files back. No matter how you restore a file system, be sure and check its integrity with fsck when the job is complete.
tar tapes are preferred (when possible) because the inode allocation is performed by the kernel rather than the restor(8) program. This has the benefit of allocating inodes sequentially starting from the beginning of the inode portion of the filesystem rather than preserving the fragmented/randomized order of the old filesystem.
Begin by reading the document ``Changes to the System in 2.11BSD'' to get an idea of how the system changes will affect your local modifications. If you have local device drivers, see the file /sys/OTHERS/README for hints on how to integrate your drivers into 2.11BSD.
The only upgrade path to 2.11BSD is to do a full bootstrap as described in Chapter 2. As always, full backups of the existing system should be made to guard against errors or failures. NOTE: The old filesystems can not be mounted by the new kernel. If you must access old discs or filesystems, there is a version of dump(8) in /usr/src/old/dump which can be used with the raw disc to dump old filesystems.
The archive file format has changed, the 4.3BSD ar(5) format is now used. Local archives will have to be converted by the /usr/old/arcv program.
The following list enumerates the standard set of files you will want to save and suggests directories in which site specific files should be present. Note that because 2.10BSD changed so radically from previous versions of UNIX on the PDP-11, many of these files may not exist on your system, and will almost certainly require extensive changes for 2.11BSD, but it's still handy to have them around as you're configuring 2.11BSD. This list will likely be augmented with non-standard files you have added to your system.
You should create a tar image of (at a minimum) the following files before the new file systems are created. In addition, you should do a full dump before rebuilding the file system to guard against missing something the first time around. The 2.11BSD restor(8) program can read and convert old dump(8) tapes.
\(uaFiles that can be used from 2.10BSD without change. ***Files that need local modifications merged into 2.11BSD files. *Files that require special work to merge and are discussed below.
The next step is to build a working 2.11BSD system. This can be done by following the steps in section 2 of this document for extracting the root and /usr file systems from the distribution tape onto unused disk partitions.
Once you have extracted the 2.11BSD system and booted from it, you will have to build a kernel customized for your configuration. If you have any local device drivers, they will have to be incorporated into the new kernel. See section 4.2.3 and ``Building 2.11BSD UNIX Systems.''
With the introduction of disklabels the disk partitions in 2.11BSD the /etc/disktab file has changed dramatically. There is a detailed description later in this chapter about the changes. If you have modified the partition tables in previous versions of 2.11BSD you will need to create a new disktab entry or modify an existing one.
When your system is booting reliably and you have the 2.11BSD root and /usr file systems fully installed you will be ready to continue with the next step in the conversion process, merging your old files into the new system.
If you saved the files on a tar tape, extract them into a scratch directory, say /usr/convert:
For sites running 2.10.1BSD, converting local configuration files should be very simple. In general very little has changed between 2.10.1BSD and 2.11BSD with regard to these files.
For sites running a pre-2.10BSD UNIX, there is very little that can be said here as the variety of previous versions of PDP-11 UNIX systems and how they were administered is large. As an example, most previous versions of PDP-11 UNIX systems used the files /etc/ttys and /etc/ttytype to administer which terminals should have login processes attached to them and what the types of terminals those were. Under 2.11BSD /etc/ttytype has disappeared entirely, its functions subsumed by /etc/ttys along with several new functions. In general you will simply have to use your previous configuration files as references as you configure 2.11BSD to your site needs. Familiarity with 4.3BSD configuration is very helpful at this point since 2.11BSD is nearly identical in most of the files listed in the previous section.
If you have any home grown device drivers that use major device numbers reserved by the system you will have to modify the commands used to create the devices or alter the system device configuration tables in /sys/pdp/conf.c. Note that almost all 2.11BSD major device numbers are different from those in previous PDP-11 UNIX systems except 2.10.1BSD. A couple more device numbers were added since the release of 2.10.1BSD for the kernel logging facility (/dev/klog) and a (new) TK50/TU81 driver.
System security changes require adding several new ``well-known'' groups to /etc/group. The groups that are needed by the system as distributed are:
Several new users have also been added to the group of ``well-known'' users in /etc/passwd. The current list is:
After restoring your old password file from tape/backups, a conversion is required to create the shadow password file. Only the steps to convert /etc/passwd are given here, see the various man pages for chpass(1), vipw(8), mkpasswd(8), etc.
The format of the cron table, /etc/crontab, is the same as that of 2.10.1BSD.
Some of the commands previously in /etc/rc.local have been moved to /etc/rc; several new functions are now handled by /etc/rc.local. You should look closely at the prototype version of /etc/rc.local and read the manual pages for the commands contained in it before trying to merge your local copy. Note in particular that ifconfig has had many changes, and that host names are now fully specified as domain-style names (e.g, boris.Oswego.EDU).
The C library and system binaries on the distribution tape are compiled with versions of gethostbyname and gethostbyaddr which use ndbm host table lookup routines instead of the name server. You must run mkhosts(8) to create the ndbm host table database from /etc/hosts. For 2.11BSD the mkhosts program has been enhanced to support multiple addresses per host with order being preserved (the order in which the multiple addresses appear in /etc/hosts for the same host is the same order the addresses will be returned to the caller of gethostbyname).
There is a version of the nameserver which runs under 2.11BSD. However in addition to having a voracious appetite for memory there are memory leaks which cause named(8) to crash after running for an extended period. Restarting named(8) nightly from cron is the only work around solution at present.
If you want to compile your system to use the name server resolver routines instead of the ndbm host table, you will need to modify /usr/src/lib/libc/Makefile according to the instructions there and then recompile all of the system and local programs (see section 6.5).\(ua
The format of /etc/ttys is the same as it was under 2.10BSD. It includes the terminal type and security options that were previously in /etc/ttytype and /etc/securettys.
syslog is the 4.4BSD-Lite version now. See syslog(3) and syslogd(8) for details. They are used by many of the system daemons to monitor system problems more closely, for example network routing changes.
Again, it must be emphasized that the nameserver is not robust under 2.11BSD, and if the hosts files are not desired then the best alternative is to use the resolver(5) routines and use the nameserver on a remote larger machine. The resolver(5) routines are known to work.
The spooling directories saved on tape may be restored in their eventual resting places without too much concern. Be sure to use the ``p'' option to tar so that files are recreated with the same file modes:
The ownership and modes of two of these directories needs to be changed, because at now runs set-user-id ``daemon'' instead of root. Also, the uucp directory no longer needs to be publicly writable, as tip reverts to privileged status to remove its lock files. After copying your version of /usr/spool, you should do the following:
Whatever else is left is likely to be site specific or require careful scrutiny before placing in its eventual resting place. Refer to the documentation and source code before arbitrarily overwriting a file.
This section summarizes some of the significant changes in 2.11BSD from 2.10.1BSD. The installation guide for 2.10.1BSD is included in the distribution as /usr/doc/2.10/setup.2.10 and should be read if you are not presently running 2.10BSD or 2.10.1BSD. It does not include changes in the network; see chapter 5 for information on setting up the network.
Old core files will not be intelligible by the current debuggers because of numerous changes to the user structure. Also removed from the user structure are the members u_offset, u_count, u_base, u_segflg, the 4.3BSD uio/iovec/rdwri kernel i/o model having been put in place. The 4.3BSD namei argument encapsulation technique has been ported, which adds the u_nd member to the user structure.
Note, once your system is installed and running, you should make sure that you recompile and reinstall the directory /usr/src/share/zoneinfo. Read through the Makefile first, if you're not located on the West Coast you will have to change it. This directory is an addition since 4.3BSD, and is intended to solve the Daylight Savings Time problems once and for all.
The incore inode structure has had the i_id member added as part of implementing the 4.3BSD namei cache. The di_addr member of the on disk inode structure is now an array of type daddr_t instead of char. The old 3 byte packed block number is obsolete at last.
The on disk directory structure is that of 4.3BSD with the difference that the inode number is an unsigned short instead of a long. This was done to reduce the amount of long arithmetic in the kernel and to maintain compatibility with earlier versions with regard to the maximum number of inodes per filesystem. Given the typical size of discs used with 2.11BSD the limit on the number of inodes per filesystem will not be a problem.
And again, 2.11BSD is not filesystem compatible with any previous PDP-11 UNIX system.
If you want to use ps after booting a new kernel, and before going multiuser, you must initialize its name list database by running ps -U.
UT_NAMESIZE in <utmp.h was changed from 8 to 15. This won't affect correctly written programs (those which do not hard code the constant 8) at the source level but does cause changes in various databases. This means that old binaries won't be able to cope with new databases (passwd, aliases, etc) and vice versa.
This change was necessary since the systems available for 2.11BSD development had to be shared with systems in which UT_NAMESIZE was set at 15. If this change/incompatibility is not desired, then utmp.h and wtmp.h will have to be modified and the system libraries and applications rebuilt before proceeding to load local software.
The simplest way to deal with this incompatibility is simply to rebuild all your databases from the source data. In particular, you should be sure you rebuild /etc/passwd, /etc/hosts, and /etc/aliases databases via the commands: mkpasswd /etc/passwd, mkhosts /etc/hosts, and /usr/ucb/newaliases.
The manual system continues to track the changes going on in 4BSD. I'm not convinced the new setup is better, but it does seem to be the method of the moment. The setup is essentially the same as that in the 4.3BSD-TAHOE distribution with the manual source in /usr/src/man.
The value of NMOUNT in /sys/h/param.h is set to 5 in the distribution system. This will be too small for many sites. Since each mount table entry costs about 440 bytes of valuable kernel dataspace this number should be chosen with care. See Appendix A for an explanation of how to reconfigure NMOUNT.
The May 1989 release of the 4.3BSD shadow password file has been ported to 2.11BSD. Password aging is also implemented.
/etc/rc and /etc/rc.local have changed fairly significantly, and
/etc/netstart has been added to configure site specific network features (much of this was pulled from the old rc.local). /etc/netstart uses the tiny program testnet which attempts to create a socket and prints NO on stdout if an error is returned by the kernel, YES if no error was returned.
mkfs(8) no longer can populate a filesystem with files. The 4.3BSD versions of mkfs(8) and mkproto(8) were ported to 2.11BSD. There is a limit on the size of the file which mkproto(8) can place on a newly created filesystem. Only files up to single indirect (about 260kb) may be copied at this time.
mklost+found(8) is a ported version from 4.3BSD, the only change being to use 63 character file names (MAXNAMLEN is 63 at this time in 2.11BSD) instead of 255. mklost+found(8) is really not needed, fsck(8) is now capable of automatically extending lost+found by up to the number of direct blocks in an inode.
The format of /etc/disktab is now the same as 4.3BSD-Reno and 4.4BSD. Previously to describe a drive (an RM03 for example) the /etc/disktab file had entries of the form:
:ty=removable:ns#32:nt#5:nc#823:sf:
:b0=/mdec/rm03uboot:
:pa#9600:ba#1024:fa#1024:
:pb#9600:bb#1024:fb#1024:
:pc#131520:bc#1024:fc#1024:
:pf#121920:bf#1024:ff#1024:
:pg#112320:bg#1024:fg#1024:
:ph#131520:bh#1024:fh#1024:
Note that there is no information at all about which cylinder a partition starts at or which partitions overlap and may not be used simultaneously. That information was kept in tables in the driver. If you modified /etc/disktab it would have no effect without also changing the driver and recompiling the kernel.
The new /etc/disktab file looks like this:
:ty=removable:ns#32:nt#5:nc#823:sf:
:b0=/mdec/rm03uboot:
:pa#9600:oa#0:ba#1024:fa#1024:ta=2.11BSD:
:pb#9600:ob#9600:bb#1024:fb#1024:tb=swap:
:pc#131520:oc#0:bc#1024:fc#1024:
:pf#121920:of#9600:bf#1024:ff#1024:tf=2.11BSD:
:pg#112320:og#19200:bg#1024:fg#1024:tg=2.11BSD:
:ph#131520:oh#0:bh#1024:fh#1024:th=2.11BSD
There are two new fields per partition, the 'o' (oa, ob, usw.) field specifies the offset in sectors that the partition begins at. The 't' field specifies the partition type. Only those partitions which are 2.11BSD will be recognized by newfs(8) and the kernel as filesystems. The kernel also will not swap or place a crash dump on a partition that is not of type swap.
The two examples above are equivalent and provide an example of a translating an old style disktab entry into a new style entry. To translate a customized disktab entries you will need: 1) a copy of your current partition tables from the device driver, 2) a copy of the old disktab entry, 3) your current /etc/fstab file. In new disktab entries you should only place those partitions you actually use. There is no need to declare (as was done in the examples above) all of the possible partitions.
If you
have changed the disk partition sizes, be sure to make the necessary
/etc/disktab changes and label your disks
BEFORE trying to access any of
your old file systems!
There are two ways to label your disks. The standalone disklabel program
is one way. It is also possible to label disks using disklabel(8)
with the -r option - this works even when running on a kernel which does
not support labels (-r reads and writes the raw disk, thus it is possible
to label disks on an older kernel as long as the disklabel(8)
program is present).
This section describes procedures used to set up a PDP-11 UNIX system. These procedures are used when a system is first installed or when the system configuration changes. Procedures for normal system operation are described in the next section.
/boot uses the device information passed to it from the bootstrap in determining the device, unit and file to load. If an autoreboot is being done the kernel will have passed the device information to the bootstrap as well as setting the autoreboot flag.
/boot does not require recompilation to adapt to a new autoreboot device.
This section briefly describes the layout of the kernel code and how files for devices are made.
As distributed, the kernel source is in a separate tar image. The source may be physically located anywhere within any file system so long as a symbolic link to the location is created for the file /sys (many files in /usr/include are normally symbolic links relative to /sys). In further discussions of the system source all path names will be given relative to /sys.
The directory /sys/sys contains the mainline machine independent operating system code. Files within this directory are conventionally named with the following prefixes:
The remaining directories are organized as follows:
Many of these directories are referenced through /usr/include with symbolic links. For example, /usr/include/sys is a symbolic link to /sys/h. The system code, as distributed, is mostly independent of the include files in /usr/include. Unfortunately not all references to /usr/include have been eradicated, so compiling the system requires the /usr file system to be mounted.
Devices supported by UNIX are implemented in the kernel by drivers whose source is kept in /sys/pdp, /sys/pdpuba, or /sys/pdpmba. These drivers are loaded into the system when included in a cpu specific configuration file kept in the conf directory. Devices are accessed through special files in the file system, made by the mknod(8) program and normally kept in the /dev directory. For all the devices supported by the distribution system, the files in /dev are created by the /dev/MAKEDEV shell script.
Determine the set of devices that you have and create a new /dev directory by running the MAKEDEV script. First create a new directory /newdev, copy MAKEDEV into it, edit the file MAKEDEV.local to provide an entry for local needs, and run it to generate a /newdev directory. For instance, if your machine has a single DZ11, a single DH11, an RM03 disk, an EMULEX UNIBUS SMD disk controller, an AMPEX 9300 disk, and a TE16 tape drive you would do:
You can then do
The kernel configuration of each UNIX system is described by a single configuration file, stored in the /sys/conf directory. The format of this file is very simple consisting of lines starting with an identifier followed by a value. Blank lines and anything past a ``#'' (including the #) are comments. This file is processed by the shell script config in the same directory. The manual pages in section 4 of the UNIX manual specify the configuration lines necessary for various devices. A comprehensive list of system options with descriptions of their meanings and effects can be found in appendix A.
The configuration file GENERIC in the conf directory was used to build the generic distribution kernel. To build a local configuration file, copy GENERIC to a new file SYSTEM, edit SYSTEM for your local system configuration, and then type "./config SYSTEM". This will create the directory ../SYSTEM and copy specially edited files into based on the definitions in SYSTEM. Change directory into the new system directory and type "make all".\(ua,
Note that the overlay scheme in the Makefile copied into the new system directory may fail because either the base segment is too small, too large or one or more overlay segments are too large. If this happens the system objects will have to be re-arranged in the base and overlay segments. The comments in the Makefile should make it fairly clear what the restrictions on object placement are in the system.
The configured system image ``unix''*** should be copied to the root, and then booted to try it out. It is best to save the old kernel to a known name so as not to destroy the working system until you're sure the new one does work. It is an better idea to have a non network kernel (/emergencyunix) always kept on the system:
This section describes how to layout file systems to make use of the available space and to balance disk load for better system performance.
Each physical disk drive can be divided into up to 8 partitions; UNIX typically uses only 3 or 4 partitions. For instance, on an RP06 the first partition, xp0a, is used for a root file system, a backup thereof, or a small file system like, /tmp; the second partition, xp0b, is used for swapping or a small file system; and a combination of the remaining partitions (xp0d, xp0e, xp0f, xp0g, xp0h) would hold user file systems.
Warning: for disks on which DEC standard 144 bad sector forwarding is supported, the last track and up to 126 preceding sectors contain replacement sectors and bad sector lists. Disk-to-disk copies should be careful to avoid overwriting this information. See bad144(8). Bad sector forwarding is optional in the hk and xp drivers. The partition sizes listed in /etc/disktab that newfs(8) uses automatically reserve the maximum amount of room that may be used by bad block forwarding on a disk.
Note also that bad144 style bad block forwarding can not be used with SI controllers on the xp driver as the controllers use their own internal scheme for bad block forwarding, and you can in fact make your disks unusable on the SI controllers if you write anything in the last five cylinders. The partition sizes in /etc/disktab also handle this constraint automatically.
The generic distribution kernel does not do bad block forwarding. There is unfortunately no way to run bad144 style bad block forwarding on some of your disks, but not others. As a final bug, the hk and xp drivers do not reread the bad sector forwarding information when disk packs are changed and so will erroneously use bad block forwarding information from the wrong packs!
The space available on a disk varies, not surprisingly, per device. Disklabels make a table giving sizes meaningless since there are no predefined partition sizes embedded in the kernel any longer. The root filesystem (a) must be at least 4Mb, preferably 6 to 7Mb if possible. The swap area (almost always the b partition) should be about 3Mb or so. If your system has a small amount (less than 2Mb) of memory you will need more swap space, perhaps 4 or 5Mb. It is a rare case where more than 5 or 6Mb of swap space is required. The system will run out of other resources by the time enough activity is generated to need that much swap space.
The system (boot) disk has a swapping area and a root file system. Other drives may use those partitions for data. Remember: the a partition must start at sector 0 or disklabel(8) or else the kernel will not be able to read/write the label.
The distributed system binaries occupy about 34 Megabytes while the major sources occupy another 36 Megabytes. Adding in the miscellaneous sources, a few locate works of art bring the total for a complete system to about 90 Megabytes. This overflows RK07, RL02 and RM03 systems, but fits easily on most other hardware configurations. 2.11BSD is quite happy on RD54 or larger. Simply fitting the distribution isn't enough, there must still be space left for user files, objects when compiling programs, spooling directories, usw.
Be aware that the disks have their sizes measured in disk sectors (512 bytes), while the UNIX file system blocks are 1024 bytes each. Thus if a disk partition has 10000 sectors (disk blocks), it will have only 5000 UNIX file system blocks, and you must divide by 2 to use 5000 when specifying the size to the mkfs command for instance. The newfs(8) program performs this calculation automatically. You should never need to run mkfs manually. All user programs report disk space in kilobytes and, where needed, disk sizes are always specified in units of sectors. The /etc/disktab file used in making file systems specifies disk partition sizes in sectors; the default sector size may be overridden with the ``se'' attribute. Note that the only sector size currently supported is NBPG as defined in /sys/pdp/machparam.h. This restriction is enforced in several places in the disklabeling process as a safeguard against specifying a sector size other than NBPG (512). Any other sector size would produce strange results and almost certainly curdled filesystems.
There are several considerations in deciding how to adjust the arrangement of things on your disks. The most important is making sure that there is adequate space for what is required; secondarily, throughput should be maximized. Swap space is an important parameter since it defines the maximum process image load that may be run. If, for instance, your swap area were smaller than the amount of main memory available after the kernel took its share, some of your memory would never be used.
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; if you have several disks, it makes sense to mount this in a ``root'' (i.e. first partition) file system on another disk. All the programs that create files in /tmp take care to delete them, but are not immune to rare events and can leave dregs. The directory should be examined every so often and the old files deleted.
The efficiency with which UNIX is able to use the CPU is often strongly affected by the configuration of disk controllers. For general time-sharing applications, the best strategy is to try to split the most actively-used sections among several disk arms.
It is critical for good performance to balance disk load. There are at least five components of the disk load that you can divide between the available disks:
+--------------------------+ +----------+---------------+ | | | disks | |what | 2 | 3 | 4 | +----------+---+-----+-----+ |/ | 0 | 0 | 0 | |tmp | 1 | 2 | 3 | |usr | 1 | 1 | 1 | |swapping^ | 0 | 2 | 2 | |users | 0 | 0+2 | 0+2 | |archive | x | x | 3 | +----------+---+-----+-----+ +--------------------------+
The most important things to consider are to even out the disk load as much as possible, and to do this by decoupling file systems (on separate arms) between which heavy copying occurs. Note that a long term average balanced load is not important; it is much more important to have an instantaneously balanced load when the system is busy. When placing several busy file systems on the same disk, it is helpful to group them together to minimize arm movement, with less active file systems off to the side.
Intelligent experimentation with a few file system arrangements can pay off in much improved performance. It is particularly easy to move the root, the /tmp file system and the swapping area. Note, though, that the disks containing the root and swapping area can never be removed while UNIX is running. Place the user files and the /usr directory as space needs dictate and experiment with the other, more easily moved file systems.
To put a chosen disk layout into effect, you should use the newfs(8) command to create each new file system. Each file system must also be added to the file /etc/fstab so that it will be checked and mounted when the system is bootstrapped.
As an example, consider a system with RA80's. On the first RA80, ra0, we will put the root file system in ra0a, and the /usr file system in ra0c, which has enough space to hold it and then some. The /tmp directory will be part of the root file system, as no file system will be mounted on /tmp. If we had only one RA80, we would put user files in the ra0c partition with the system source and binaries.
If we had a second RA80, we would place /usr in ra1c. We would put user files in ra0c, calling the file system /mnt. We would put swap on ra0b. We would keep a backup copy of the root file system in the ra1a disk partition and put /tmp on ra1b. /etc/fstab would then contain
To make the /mnt file system we would do:
If UNIX is to support simultaneous access from directly-connected terminals other than the console, the file /etc/ttys (ttys(5)) must be edited.
Terminals connected via DZ11 interfaces are conventionally named ttyDD where DD is a decimal number, the ``minor device'' number. The lines on dz0 are named /dev/tty00, /dev/tty01, ... /dev/tty07. By convention, all other terminal names are of the form ttyCX, where C is an alphabetic character according to the type of terminal multiplexor and its unit number, and X is a digit for the first ten lines on the interface and an increasing lower case letter for the rest of the lines. C is defined for the number of interfaces of each type listed below. Since tty structures are approximately 78 bytes each, it is highly doubtful that more than 3 or 4 terminal interface boards will ever be attached to a PDP-11 (especially in a BA23 cabinet).
To add a new terminal device, be sure the device is configured into the system and that the special files for the device have been made by /dev/MAKEDEV. Then, enable the appropriate lines of /etc/ttys by setting the ``status'' field to on (or add new lines). Note that lines in /etc/ttys are one-for-one with entries in the file of current users (/var/run/utmp), and therefore it is best to make changes while running in single-user mode and to add all of the entries for a new device at once.
The format of the /etc/ttys file is the same in 2.11BSD as in 2.10BSD and 4.3BSD. Each line in the file is broken into four tab separated fields (comments are shown by a `#' character and extend to the end of the line). For each terminal line the four fields are: the device (without a leading /dev), the program /etc/init should startup to service the line (or none if the line is to be left alone), the terminal type (found in /etc/termcap), and optional status information describing if the terminal is enabled or not and if it is ``secure'' (i.e. the super user should be allowed to login on the line). All fields are character strings with entries requiring embedded white space enclosed in double quotes. Thus a newly added terminal /dev/tty00 could be added as
Dialup terminals should be wired so that carrier is asserted only when the phone line is dialed up. For non-dialup terminals from which modem control is not available, you must either wire back the signals so that the carrier appears to always be present, or show in the minor device number that carrier is to be assumed to be present by adding 128 decimal to the minor device number when creating the device node. This differs from 4.3BSD where the softcarrier state is specified at kernel configuration time.
For network terminals (i.e. pseudo terminals), no program should be started up on the lines. Thus, the normal entry in /etc/ttys would look like
When the system is running multi-user, all terminals that are listed in /etc/ttys as on have their line are enabled. If, during normal operations, it is desired to disable a terminal line, you can edit the file /etc/ttys to change the terminal's status to off and then send a hangup signal to the init process, by doing
Note that if a special file is inaccessible when init tries to create a process for it, init will log a message to the system error logging process (/usr/sbin/syslogd) and try to reopen the terminal every minute, reprinting the warning message every 10 minutes. Messages of this sort are normally printed on the console, though other actions may occur depending on the configuration information found in /etc/syslog.conf.
Finally note that you should change the names of any dialup terminals to ttyd? where ? is in [0-9a-zA-Z], as some programs use this property of the names to determine if a terminal is a dialup. Shell commands to do this should be put in the /dev/MAKEDEV.local script.
New users can be added to the system by adding a line to the password file /etc/passwd. The procedure for adding a new user is described in adduser(8).
You should add accounts for the initial user community, giving each a directory and a password, and putting users who will wish to share software in the same groups.
Several guest accounts have been provided on the distribution system; these accounts are for people at Berkeley, Bell Laboratories, and others who have done major work on UNIX in the past. You can delete these accounts, or leave them on the system if you expect that these people would have occasion to login as guests on your system.
All programs that require the site's name, or some similar characteristic, obtain the information through system calls or from files located in /etc. Aside from parts of the system related to the network, to tailor the system to your site you must simply select a site name, then edit the file
The mail system consists of the following commands:
To set up the mail facility you should read the instructions in the file READ_ME in the directory /usr/src/usr.lib/sendmail and then adjust the necessary configuration files. You should also set up the file /etc/aliases for your installation, creating mail groups as appropriate. Documents describing sendmail's operation and installation are also included on the distribution tape.
The version of uucp included in 2.11BSD is an enhanced version of the one originally distributed with 32/V\(ua. The enhancements include:
This section gives a brief overview of uucp and points out the most important steps in its installation.
To connect two UNIX machines with a uucp network link using modems, one site must have an automatic call unit and the other must have a dialup port. It is better if both sites have both.
You should first read the paper in the UNIX System Manager's Manual: ``Uucp Implementation Description''. It describes in detail the file formats and conventions, and will give you a little context. In addition, the document ``setup.tblms'', located in the directory /usr/src/usr.bin/uucp/UUAIDS, may be of use in tailoring the software to your needs.
The uucp support is located in three major directories: /usr/bin, /etc/uucp, and /usr/spool/uucp. User commands are kept in /usr/bin, operational commands in /etc/uucp, and /usr/spool/uucp is used as a spooling area. The commands in /usr/bin are:
To install uucp on your system, start by selecting a site name. A uucp account must be created in the password file and a password set up. Then, create the appropriate spooling directories with mode 755 and owned by user uucp, group daemon.
If you have an auto-call unit, the L.sys, L-dialcodes, and L-devices files should be created. The L.sys file should contain the phone numbers and login sequences required to establish a connection with a uucp daemon on another machine. For example, my L.sys file looks something like:
The first field is the name of a site, the second shows when the machine may be called, the third field specifies how the host is connected (through an ACU, a hard-wired line, etc.), then comes the phone number to use in connecting through an auto-call unit, and finally a login sequence. The phone number may contain common abbreviations that are defined in the L-dialcodes file. The device specification should refer to devices specified in the L-devices file. Listing only ACU causes the uucp daemon, uucico, to search for any available auto-call unit in L-devices. Our L-dialcodes file is of the form:
As uucp operates it creates (and removes) many small files in the directories underneath /usr/spool/uucp. Sometimes files are left undeleted; these are most easily purged with the uuclean program. The log files can grow without bound unless trimmed back; uulog maintains these files. Many useful aids in maintaining your uucp installation are included in a subdirectory UUAIDS beneath /usr/src/usr.bin/uucp. Peruse this directory and read the ``setup'' instructions also located there.
The following section has been lightly edited to correspond to the current 2.11BSD networking. Several parts of it do not really apply to 2.11BSD, for example, it is unlikely that anyone will connect a PDP-11 to an IMP but it is possible as the LH/DH-11 networking interface and the IMP modules have been ported and lightly tested, or that anyone will run the nameserver. The ``correct'' use of the networking in 2.11BSD is probably with a list of the local net addresses in the /etc/hosts file and with one default gateway for all network traffic. In particular, do not run routed(8) unless you're extremely sure that you know what you're doing. This is doubly true if SL/IP is being used as the primary connection to the outside world. The IMP and PRONET drivers are known to work, but long term robustness is unknown. Sites that wish to hook 2.11BSD into more than a simple local ethernet may have some work ahead of them. If any additional drivers are ported, I would really like a copy.
The networking in 2.11BSD, runs in supervisor mode, separate from the mainstream kernel. There is room without overlaying to hold both a SL/IP and ethernet driver. This is a major win, as it allows the networking to maintain its mbufs in normal data space, among other things. The networking portion of the kernel resides in ``/netnix'', and is loaded after the kernel is running. Since the kernel only looks for the file ``/netnix'', it will not run if it is unable to load ``/netnix'' , sites should build and keep a non-networking kernel in ``/'' at all times, as a backup. NOTE: The ``/unix'' and ``/netnix'' imagines must have been created at the same time, do not attempt to use mismatched images. The ability to have boot tell the kernel which network image to load is on the wish list (had to have something take the place of wishing for disklabels ;-)).
2.11BSD provides support for the DARPA standard Internet protocols IP, ICMP, TCP, and UDP. These protocols may be used on top of a variety of hardware devices ranging from the IMP's (PSN's) used in the Internet to local area network controllers for the Ethernet. Network services are split between the kernel (communication protocols) and user programs (user services such as TELNET and FTP). This section describes how to configure your system to use the Internet networking support. 2.11BSD also includes code to support the Xerox Network Systems (NS) protocols; the basic porting work has been done, but it is completely untested.
To configure the kernel to include the Internet communication protocols, define the INET option. This automatically defines the NLOOP option. TCP_COMPAT_42 is always defined. Xerox NS support is enabled with the NS option. In either case, include the pseudo-device ``pty'' in your machine's configuration file, using the NPTY options. The ``pty'' pseudo-device forces the pseudo terminal device driver to be configured into the system, see pty(4). The NLOOP option forces inclusion of the software loopback interface driver. The loop driver is used in network testing as well as for the system talking to itself rather than transmitting the data over the wire.
If you are planning to use the Internet network facilities on a 10Mb/s Ethernet, the pseudo-device ``ether'' should also be included in the configuration using the NETHER option; this forces inclusion of the Address Resolution Protocol module used in mapping between 48-bit Ethernet and 32-bit Internet addresses. Also, if you have an IMP connection, you will need to include the pseudo-device ``imp'', using the option NIMP. The IMP software is ported and is in use at at least one site.
Before configuring the appropriate networking hardware, you should consult the manual pages in section 4 of the Programmer's Manual. The following table lists the devices for which software support exists. Again, much of this software is unported and untested; only the basic networking has been stressed at all. Many other devices are available, but unported. Porting should simply be a matter of making the hardware device work. The directories ``/sys/pdpif'' and ``/sys/vaxif'' contain many drivers. The ones in ``pdpif'' are either the current, working drivers, or drivers that, at some time, worked on PDP-11's. The ones in ``vaxif'' are the current VAX drivers, and, as such, will have to have their memory usage changed, but serve as an excellent example of how the hardware works.
SL/IP is also available. It is surprisingly efficient. Over a 9600 baud line it is not unusual to see ftp rates in the 800 bytes per second range (depending how busy the system is).
All network interface drivers including the loopback interface, require that their host address(es) be defined at boot time. This is done with ifconfig(8) commands included in the /etc/rc.local file. Interfaces that are able to dynamically deduce the host part of an address may check that the host part of the address is correct. The manual page for each network interface describes the method used to establish a host's address. Ifconfig(8) can also be used to set options for the interface at boot time. Options are set independently for each interface, and apply to all packets sent using that interface. These options include disabling the use of the Address Resolution Protocol; this may be useful if a network is shared with hosts running software that does not yet provide this function. Alternatively, translations for such hosts may be set in advance or ``published'' by a 2.11BSD host by use of the arp(8) command. Note that the use of trailer link-level is now negotiated between 2.11BSD hosts using ARP, and it is thus no longer necessary to disable the use of trailers with ifconfig. It is STRONGLY recommended, however, that 2.11BSD networking be run without trailers, as the trailer code in most of the drivers has either been removed, commented out, is untested or is known not to work. This is a problem with certain releases of Ultrix, which has to be explicitly configured not to send trailers if it and 2.11BSD are to coexist.
To use the pseudo terminals just configured, device entries must be created in the ``/dev'' directory. To create 32 pseudo terminals (plenty, you can probably get by with many less) execute the following commands.
In 2.11BSD the DARPA Internet support includes the notion of ``subnetworks''. This is a mechanism by which multiple local networks may appears as a single Internet network to off-site hosts. Subnetworks are useful because they allow a site to hide their local topology, requiring only a single route in external gateways; it also means that local network numbers may be locally administered. The standard describing this change in Internet addressing is RFC-950.
To set up local subnetworks one must first decide how the available address space (the Internet ``host part'' of the 32-bit address) is to be partitioned. Sites with a class A network number have a 24-bit address space with which to work, sites with a class B network number have a 16-bit address space, while sites with a class C network number have an 8-bit address space\(ua. To define local subnets you must steal some bits from the local host address space for use in extending the network portion of the Internet address. This reinterpretation of Internet addresses is done only for local networks; i.e. it is not visible to hosts off-site. For example, if your site has a class B network number, hosts on this network have an Internet address that contains the network number, 16 bits, and the host number, another 16 bits. To define 254 local subnets, each possessing at most 255 hosts, 8 bits may be taken from the local part. (The use of subnets 0 and all-1's, 255 in this example, is discouraged to avoid confusion about broadcast addresses.) These new network numbers are then constructed by concatenating the original 16-bit network number with the extra 8 bits containing the local subnetwork number.
The existence of local subnetworks is communicated to the system at the time a network interface is configured with the netmask option to the ifconfig program. A ``network mask'' is specified to define the portion of the Internet address that is to be considered the network part for that network. This mask normally contains the bits corresponding to the standard network part as well as the portion of the local part that has been assigned to subnets. If no mask is specified when the address is set, it will be set according to the class of the network. For example, at Berkeley (class B network 128.32) 8 bits of the local part have been reserved for defining subnetworks; consequently the /etc/rc.local file contains lines of the form
The address defined as the broadcast address for Internet networks according to RFC-919 is the address with a host part of all 1's. The address used by 4.2BSD was the address with a host part of 0. 2.11BSD uses the standard broadcast address (all 1's) by default, but allows the broadcast address to be set (with ifconfig) for each interface. This allows networks consisting of both 4.2BSD and 2.11BSD hosts to coexist. In the presence of subnets, the broadcast address uses the subnet field as for normal host addresses, with the remaining host part set to 1's (or 0's, on a network that has not yet been converted). 2.11BSD hosts recognize and accept packets sent to the logical-network broadcast address as well as those sent to the subnet broadcast address, and when using an all-1's broadcast, also recognize and receive packets sent to host 0 as a broadcast.
If your environment allows access to networks not directly attached to your host you will need to set up routing information to allow packets to be properly routed. Two schemes are supported by the system. The first scheme employs the routing table management daemon routed to maintain the system routing tables. The routing daemon uses a variant of the Xerox Routing Information Protocol to maintain up to date routing tables in a cluster of local area networks. By using the /etc/gateways file created by htable(8), the routing daemon can also be used to initialize static routes to distant networks (see the next section for further discussion). When the routing daemon is started up (usually from /etc/rc.local) it reads /etc/gateways if it exists and installs those routes defined there, then broadcasts on each local network to which the host is attached to find other instances of the routing daemon. If any responses are received, the routing daemons cooperate in maintaining a globally consistent view of routing in the local environment. This view can be extended to include remote sites also running the routing daemon by setting up suitable entries in /etc/gateways; consult routed(8) for a more thorough discussion.
The second approach is to define a default or wildcard route to a smart gateway and depend on the gateway to provide ICMP routing redirect information to dynamically create a routing data base. This is done by adding an entry of the form
The system always listens, and processes, routing redirect information, so it is possible to combine both of the above facilities. For example, the routing table management process might be used to maintain up to date information about routes to geographically local networks, while employing the wildcard routing techniques for ``distant'' networks. The netstat(1) program may be used to display routing table contents as well as various routing oriented statistics. For example,
Only sheer insanity could prompt the use of 2.11BSD machines as gateways. If you really want to do this then the best recourse is to prowl the sources and see what has to be done. The code is all there, and the "ipforwarding" variable is present.
Local area routing within a group of interconnected Ethernets and other such networks may be handled by routed(8). Gateways between the Internet and one or more local networks require an additional routing protocol, the Exterior Gateway Protocol (EGP), to inform the core gateways of their presence and to acquire routing information from the core.
In 2.11BSD most of the server programs are started up by a ``super server'', the Internet daemon. The Internet daemon, inetd, acts as a master server for programs specified in its configuration file, /etc/inetd.conf, listening for service requests for these servers, and starting up the appropriate program whenever a request is received. The configuration file contains lines containing a service name (as found in /etc/services), the type of socket the server expects (e.g. stream or dgram), the protocol to be used with the socket (as found in /etc/protocols), whether to wait for each server to complete before starting up another, the user name as which the server should run, the server program's name, and at most five arguments to pass to the server program. Some trivial services are implemented internally in inetd, and their servers are listed as ``internal.'' For example, an entry for the file transfer protocol server would appear as
Several data files are used by the network library routines and server programs. Most of these files are host independent and updated only rarely.
When using the host address routines that use the Internet name server, the file /etc/hosts is only used for setting interface addresses and at other times that the server is not running, and therefore it need only contain addresses for local hosts. There is no equivalent service for network names yet. The days of retrieving a host file containing all systems on the Internet are over. Besides, you would grow very old and run out of disk space while waiting for mkhosts(8) to process a hosts file containing the several million entries. Therefore the details of retrieving a master hosts file using htable(8) and gettable(8) have been removed from this document. However if you do use local hosts files you will still need to run mkhosts(8) and this is described below.
If you are using the host table for host name and address mapping, you should run mkhosts(8) after installing /etc/hosts. The mkhosts(8) program has been enhanced for 2.11BSD to allow multiple addresses per host. The order in which the addresses are given in /etc/hosts is preserved, so the entries for a given host should be in order of importance. If you are using the name server for the host name and address mapping, you only need to install networks and a small copy of hosts describing your local machines. The full host table in this case might be placed somewhere else for reference by users. The gateways file may be installed in /etc/gateways if you use routed(8) for local routing and wish to have static external routes installed when routed is started. This procedure is essentially obsolete, however, except for individual hosts that are on the Milnet and do not forward packets from a local network. Other situations require the use of gated. That program can never be made to run on a PDP-11 due to address space considerations. Also, the networking code could not even begin to handle the number of routes which would be received.
If you are connected to the Internet, it is highly recommended that you use the name server resolver routines for your host name and address mapping, as this provides access to a much larger set of hosts than are provided in the host table. Many large organization on the network, currently have only a small percentage of their hosts listed in the host table retrieved from NIC.
The remote login and shell servers use an authentication scheme based on trusted hosts. The hosts.equiv file contains a list of hosts that are considered trusted and, under a single administrative control. When a user contacts a remote login or shell server requesting service, the client process passes the user's name and the official name of the host on which the client is located. In the simple case, if the host's name is located in hosts.equiv and the user has an account on the server's machine, then service is rendered (i.e. the user is allowed to log in, or the command is executed). Users may expand this ``equivalence'' of machines by installing a .rhosts file in their login directory. The root login is handled specially, bypassing the hosts.equiv file, and using only the /.rhosts file.
Thus, to create a class of equivalent machines, the hosts.equiv file should contain the official names for those machines. If you are running the name server, you may omit the domain part of the host name for machines in your local domain. For example, several machines on my local network are considered trusted, so the hosts.equiv file is of the form:
Most network servers are automatically started up at boot time by the command file /etc/rc (if they are installed in their presumed locations) or by the Internet daemon (see above). These include the following:
To have other network servers started up as well, the appropriate line should be added to the Internet daemon's configuration file /etc/inetd.conf, or commands similar to the following should be placed in the site dependent file /etc/rc.local.
The FTP server included in the system provides support for an anonymous FTP account. Because of the inherent security problems with such a facility you should read this section carefully if you consider providing such a service.
An anonymous account is enabled by creating a user ftp. When a client uses the anonymous account a chroot(2) system call is performed by the server to restrict the client from moving outside that part of the file system where the user ftp home directory is located. Because a chroot call is used, certain programs and files used by the server process must be placed in the ftp home directory. Further, one must be sure that all directories and executable images are unwritable. The following directory setup is recommended.
NOTE: Mode 777 on the 'pub' directory can and has been abused! Changing the mode to 555 is a good choice but would require administrative assistance for placing files in the 'pub' directory. Probably not a bad idea though.
Another issue to consider is the copy of /etc/passwd placed here. It may be copied by users who use the anonymous account. They may then try to break the passwords of users on your machine for further access. A good choice of users to include in this copy might be root, daemon, uucp, and the ftp user.
Aside from the problems of directory modes and such, the ftp server may provide a loophole for interlopers if certain user accounts are allowed. The file /etc/ftpusers is checked on each connection. If the requested user name is located in the file, the request for service is denied. This file normally has the following names on our systems.
This section describes procedures used to operate a PDP-11 UNIX system. Procedures described here are used periodically, to reboot the system, analyze error messages from devices, do disk backups, monitor system performance, recompile system software and control local changes.
In a normal reboot, the system checks the disks and comes up multi-user without intervention at the console. Such a reboot can be stopped (after it prints the date) with a ^C (interrupt). This will leave the system in single-user mode, with only the console terminal active. It is also possible to allow the filesystem checks to complete and then to return to single-user mode by signaling fsck with a QUIT signal (^\).
If booting from the console command level is needed, then the command
To bring the system up to a multi-user configuration from the single-user all you have to do is hit ^D on the console. The system will then execute /etc/rc, a multi-user restart script (and /etc/rc.local), and come up on the terminals listed as active in the file /etc/ttys. See init(8) and ttys(5). Note, however, that this does not cause a file system check to be performed. Unless the system was taken down cleanly, you should run ``fsck'' or force a reboot with reboot(8) to have the disks checked.
To take the system down to a single user state you can use
Each system shutdown, crash, processor halt and reboot is recorded in the file /usr/adm/shutdownlog with the cause.
When serious errors occur on peripherals or in the system, the system prints a warning diagnostic on the console. These messages are written to the kernel logger where they are retrieved by syslogd(8) via /dev/klog - dmesg(8) is now obsolete. dmesg(8) is present in the distribution but no longer used. The message buffer is now 4kb in size and external to the kernel.
Error messages printed by the devices in the system are described with the drivers for the devices in section 4 of the programmer's manual. Some drivers have been modified to use the kernel logger, others still simply do printf statements. If errors occur suggesting hardware problems, you should contact your hardware support group or field service. It is a good idea to examine the error log files regularly (e.g. with ``tail -r /usr/adm/messages'').
Periodically (say every week or so in the absence of any problems) and always (usually automatically) after a crash, all the file systems should be checked for consistency by fsck(8). The procedures of reboot(8) should be used to get the system to a state where a file system check can be performed manually or automatically.
Dumping of the file systems should be done on a regular schedule, since once the system is going it is easy to become complacent. Complete and incremental dumps are easily done with dump(8). You should arrange to do a towers-of-hanoi dump sequence; we tune ours so that almost all files are dumped on two tapes and kept for at least a week in most every case. We take full dumps every month (and keep these indefinitely).
More precisely, we have three sets of dump tapes: 10 daily tapes, 5 weekly sets of 2 tapes, and fresh sets of three tapes monthly. We do daily dumps circularly on the daily tapes with sequence `3 2 5 4 7 6 9 8 9 9 9 ...'. Each weekly is a level 1 and the daily dump sequence level restarts after each weekly dump. Full dumps are level 0 and the daily sequence restarts after each full dump also.
Thus a typical dump sequence would be:
tape name level number date opr size
------------------------------------------------------
FULL 0 Nov 24, 1979 jkf 137MB
D1 3 Nov 28, 1979 jkf 29MB
D2 2 Nov 29, 1979 rrh 34MB
D3 5 Nov 30, 1979 rrh 19MB
D4 4 Dec 1, 1979 rrh 22MB
W1 1 Dec 2, 1979 etc 40MB
D5 3 Dec 4, 1979 rrh 15MB
D6 2 Dec 5, 1979 jkf 25MB
D7 5 Dec 6, 1979 jkf 15MB
D8 4 Dec 7, 1979 rrh 19MB
W2 1 Dec 9, 1979 etc 118MB
D9 3 Dec 11, 1979 rrh 15MB
D10 2 Dec 12, 1979 rrh 26MB
D1 5 Dec 15, 1979 rrh 14MB
W3 1 Dec 17, 1979 etc 71MB
D2 3 Dec 18, 1979 etc 13MB
FULL 0 Dec 22, 1979 etc 135MB
Weekly dumps are done often enough that daily dumps always fit on one tape.
Dumping of files by name is best done by tar(1) but the amount of data that can be moved in this way is limited to a single tape. Finally if there are enough drives entire disks can be copied with dd(1) using the raw special files and an appropriate blocking factor; the number of sectors per track is usually a good value to use, consult /etc/disktab.
It is desirable that full dumps of the root file system be made regularly. These dumps should be made in ``bootable`` format, including the standalone programs mentioned back in chapter 2 (boot, mkfs, restor and icheck). This can easily be done by going to /sys/pdpstand and doing:
Exhaustion of user-file space is certain to occur now and then and may be managed with a combination of disc quotas (the 4.3BSD disc quota system is available as a kernel configuration option), threatening messages of the day, and personal letters.
If you have the equipment, the best way to move a file system is to dump it to magtape using dump(8), use newfs(8) to create the new file system, and restore the tape, using restor(8). If for some reason you don't want to use magtape, dump accepts an argument telling where to put the dump; you might use another disk. Filesystems may also be moved by piping the output of a tar(1) to another tar. The restor program accesses the raw device, laying down inodes and blocks in the same place they came from as recorded by dump. Care must therefore be taken when restoring a dump into a file system smaller than the original file system.
If you have to shrink a file system or merge a file system into another, existing one, the best bet is to use tar(1). If you are playing with the root file system and only have one drive, the procedure is more complicated. If the only drive is a Winchester disk, this procedure may not be used without overwriting the existing root or another partition. What you do is the following:
Note that if you add new disk drivers they should also be added to the standalone system in /sys/pdpstand. If you change the disk partition tables the default disk partition tables in /etc/disktab should be modified.
It is easy to regenerate the system, and it is a good idea to try rebuilding pieces of the system to build confidence in the procedures. The system consists of two major parts: the kernel itself (/sys) and the user programs (/usr/src and subdirectories). The major part of this is /usr/src.
The three major libraries are the C library in /usr/src/lib/libc and the FORTRAN libraries /usr/src/usr.lib/libI77 and /usr/src/usr.lib/libF77. In each case the library is remade by changing into the corresponding directory and doing
The source for all other libraries is kept in subdirectories of /usr/src/usr.lib; each has a makefile and can be recompiled by the above recipe.
If you look at /usr/src/Makefile, you will see that you can recompile the entire system source with one command. To recompile a specific program, find out where the source resides with the whereis(1) command, then change to that directory and remake it with the makefile present in the directory. For instance, to recompile ``date'', all one has to do is
If you wish to recompile and install all programs in a particular target area you can override the default target by doing:
To regenerate all the system source you can do
If you modify the C library, say to change a system call, and want to rebuild and install everything from scratch you have to be a little careful. You must insure that the libraries are installed before the remainder of the source, otherwise the loaded images will not contain the new routine from the library. The following sequence will accomplish this.
/usr/new is used by default for the programs that constitute the contributed software portion of the distribution but which may not have man pages installed. Locally written commands that aren't distributed (or whose man pages are not up to date) are kept in /usr/src/local and their binaries are kept in /usr/local. This allows /usr/bin, /usr/ucb, and /bin to correspond to the distribution tape People using /usr/local commands are made aware that the programs may not be in the base system yet.
UNIX optionally records two kinds of accounting information: connect time accounting and process resource accounting. The connect time accounting information is stored in the file /usr/adm/wtmp, which is summarized by the program ac(8). The process time accounting information is stored in the file /usr/adm/acct after it is enabled by accton(8), and is analyzed and summarized by the program sa(8).
If you need to recharge for computing time, you can develop procedures based on the information provided by these commands. A convenient way to do this is to give commands to the clock daemon cron to be executed every day at a specified time. This is done by adding lines to /usr/adm/crontab; see cron(8) for details.
Resource control in 2.11BSD is more elaborate than in previous PDP-11 UNIX systems. The resources consumed by any single process can be limited by the mechanisms of setrlimit(2). As distributed, the mechanism is voluntary, though sites may choose to modify the login mechanism to impose limits. Csh now has the limits builtin command enabled. Another available option is the 4.3BSD disc quota system.
The discussion of system operations is concluded by listing the files that require periodic attention or are system specific
/etc/fstab how disk partitions are used /etc/disktab disk partition sizes /etc/printcap printer data base /etc/gettytab terminal type definitions /etc/remote names and phone numbers of remote machines for tip(1) /etc/group group memberships /etc/motd message of the day /etc/master.passwd password file; each account has a line /etc/rc.local local system restart script; runs reboot; starts daemons /etc/inetd.conf local internet servers /etc/hosts host name data base /etc/networks network name data base /etc/netstart Startup file to configure network /etc/services network services data base /etc/hosts.equiv hosts under same administrative control /etc/syslog.conf error log configuration for syslogd(8) /etc/ttys enables/disables ports /etc/crontab commands that are run periodically /etc/aliases mail forwarding and distribution groups /usr/adm/acct raw process account data /usr/adm/messages system error log /usr/adm/shutdownlog log of system reboots /usr/adm/wtmp login session accounting
The 2.11BSD kernel has a number of parameters and options that can be used to tailor the kernel to site specific needs. This appendix lists the parameters and options used in the kernel. The parameters have numeric values, usually table sizes. The options flags are either defined or undefined (via the values YES or NO respectively.)
Prototypes for all the following options can be found in the generic kernel configuration file /sys/conf/GENERIC. The process of configuring a new kernel consists simply of copying the generic configuration file to a new file, SYSTEM and then editing the options in SYSTEM to reflect your needs. You can treat the items copied from GENERIC as a ``grocery list'', checking off those options you want, crossing out those you don't and setting numeric parameters to reasonable values.
Because of time constraints the NMOUNT constant was not moved into the kernel configuration file where it belongs. NMOUNT is used to configure the number of mountable file systems in 2.11BSD. Since each slot in the kernel mount table takes up close to a half Kb of valuable kernel data space, the distribution kernel comes configured with NMOUNT set to 5. This is almost certainly too small for most sites and should be increased to the number of file systems you expect to mount.
NMOUNT is defined in /sys/h/param.h. If you change its value, you must recompile the kernel (obviously) and the following applications: mount, quotaon, edquota, umount, and df.
All of the generic kernels support the following devices:
Device Number - RK06/07 2 MSCP (RA) Controllers 2 MSCP (RA) Disks 3 RL01/02 Drives 2 SMD (XP) Controllers 1 SMD (XP) Disks 2 TE16, TU45, TU77 (HT) Tape drives 2 TM11 (TM) Tape drives 2 TS11 (TS) Tape drives 2 TK50 (TMSCP) Tape drives 2
The generic kernel adapts automatically to the booted device. The 'a' partition on the booted device is automatically made the root filesystem and the 'b' partition the swap area (except for the RL02 which uses the second drive). The size of the swap partition is determined at run time, the kernel queries the driver for the number of block in the 'b' partition. NOTE: If the swap partition is not labeled as being of type swap the kernel will panic.
# Machine configuration file for 2.11BSD distributed kernel.
#
# Format:
# name value comments
# An item's value may be either numerical, boolean or a string; if it's
# boolean, use "YES" or "NO" to set it or unset it, respectively. Use
# the default value and the comments field as indicators of the type of
# field it is.
#########################################
# MACHINE DEPENDENT PARAMETERS #
#########################################
# Machine type
# Split I/D and hardware floating point are required.
#
# Including UNIBUS map support for machines without a UNIBUS will not cause
# a kernel to die. It simply includes code to support UNIBUS mapping if
# present.
#
# The define UNIBUS_MAP implements kernel support for UNIBUS mapped
# machines. However, a kernel compiled with UNIBUS_MAP does *not* have to
# be run on a UNIBUS machine. The define simply includes support for UNIBUS
# mapping if the kernel finds itself on a machine with UNIBUS mapping.
UNIBUS_MAP YES # include support for UNIBUS mapping
# The define Q22 has been removed. The references to it were incorrect
# (i.e. using it to distinguish between an Emulex CS02 and a DH11) or
# inappropriate (the if_il.c driver should have been checking if a Unibus
# Map was present at runtime).
#LINEHZ 50 # clock frequency European
LINEHZ 60 # clock frequency USA
# PDP-11 machine type; allowable values are GENERIC, 44, 70, 73. GENERIC
# should only be used to build a distribution kernel. The only use of this
# option is to select the proper in-line PS instructions (references to the
# PSW use 'spl', 'mfps/mtps' or 'movb' instructions depending on the cpu type).
PDP11 GENERIC # distribution kernel
#PDP11 44 # PDP-11/44
#PDP11 70 # PDP-11/70,45,50,55
#PDP11 73 # PDP-11/73,53,83,93,84,94
#########################################
# GENERAL SYSTEM PARAMETERS #
#########################################
IDENT GENERIC # machine name
MAXUSERS 4 # maxusers on machine
# BOOTDEV is the letter combination denoting the autoboot device,
# or NONE if not using the autoboot feature.
BOOTDEV NONE # don't autoboot
#BOOTDEV dvhp # DIVA Comp/V boot device
#BOOTDEV hk6 # rk06 boot device
#BOOTDEV hk7 # rk07 boot device
#BOOTDEV ra # MSCP boot device
#BOOTDEV rl # rl01/02 boot device
#BOOTDEV rm # rm02/03/05 boot device
#BOOTDEV br # Eaton BR1537/BR1711 boot device
#BOOTDEV sc11 # Emulex SC11/B boot device
#BOOTDEV sc21 # Emulex SC21 boot device
#BOOTDEV si # si 9500 boot device
# Timezone, in minutes west of GMT
#TIMEZONE 300 # EST
#TIMEZONE 360 # CST
#TIMEZONE 420 # WST
TIMEZONE 480 # PST
DST 1 # Daylight Savings Time (1 or 0)
# Filesystem configuration
# Rootdev, swapdev etc. should be in terms of makedev. For example,
# if you have an SMD drive using the xp driver, rootdev would be xp0a,
# or "makedev(10,0)". Swapdev would be the b partition, xp0b, or
# "makedev(10,1)". The 10 is the major number of the device (the offset
# in the bdevsw table in conf.c) and the 0 and 1 are the minor numbers
# which correspond to the partitions as described in the section 4 manual
# pages. You can also get the major numbers from the MAKEDEV script in
# /dev.
PIPEDEV makedev(10,0) # makedev(10,0) xp0a
ROOTDEV makedev(10,0) # makedev(10,0) xp0a
SWAPDEV makedev(10,1) # makedev(10,1) xp0b
# DUMPROUTINE indicates which dump routine should be used. DUMPDEV
# should be in terms of makedev. If DUMPDEV is NODEV no automatic
# dumps will be taken, and DUMPROUTINE needs to be set to "nulldev" to
# resolve the reference. See param.h and ioconf.c for more information.
# DUMPLO should leave room for the kernel to start swapping without
# overwriting the dump.
DUMPLO 512 # dump start address
DUMPDEV NODEV # makedev(10,1) xp0b
DUMPROUTINE nulldev # no dump routine.
#DUMPROUTINE hkdump # hk driver dump routine
#DUMPROUTINE hpdump # hp driver dump routine
#DUMPROUTINE radump # ra driver dump routine
#DUMPROUTINE rldump # rl driver dump routine
#DUMPROUTINE rmdump # rm driver dump routine
#DUMPROUTINE brdump # br driver dump routine
#DUMPROUTINE sidump # si driver dump routine
#DUMPROUTINE xpdump # xp driver dump routine
#DUMPROUTINE tmsdump # tms driver dump routine
#########################################
# KERNEL CONFIGURATION #
#########################################
BADSECT NO # bad-sector forwarding
EXTERNALITIMES YES # map out inode time values
UCB_CLIST NO # clists moved from kernel data space
NOKA5 NO # KA5 not used except for buffers
# and clists (_end < 0120000);
QUOTA NO # dynamic file system quotas
# NOTE -- *very* expensive
# UCB_METER is fairly expensive, but various programs (iostat, vmstat, etc)
# use it.
UCB_METER NO # vmstat performance metering
# NBUF is the size of the buffer cache, and is directly related to the UNIBUS
# mapping registers. There are 32 total mapping registers, of which 30 are
# available. The 0'th is used for CLISTS, and the 31st is used for the I/O
# page on some PDP's. It's suggested that you allow 7 mapping registers
# per UNIBUS character device so that you can move 56K of data on each device
# simultaneously. The rest should be assigned to the block buffer pool. So,
# if you have a DR-11 and a TM-11, you would leave 14 unassigned for them and
# allocate 16 to the buffer pool. Since each mapping register addresses 8
# buffers for a 1K file system, NBUF would be 128. A possible exception would
# be to reduce the buffers to save on data space, as they were 24 bytes each
# Should be 'small' for GENERIC, so room for kernel + large program to run.
NBUF 32 # buffer cache, *must* be <= 240
# DIAGNOSTIC does various run-time checks, some of which are pretty
# expensive and at a high priority. Suggested use is when the kernel
# is crashing and you don't know why, otherwise run with it off.
DIAGNOSTIC NO # misc. diagnostic loops and checks
#########################################
# PERIPHERALS: DISK DRIVES #
#########################################
NBR 0 # EATON BR1537/BR1711, BR1538A, B, C, D
NHK 2 # RK611, RK06/07
NRAC 1 # NRAD controllers
NRAD 2 # RX50, RC25, RD51/52/53, RA60/80/81
NRK 0 # RK05
NRL 2 # RL01/02
NRX 0 # RX02
NSI 0 # SI 9500 driver for CDC 9766 disks
# Because the disk drive type registers conflict with other DEC
# controllers, you cannot use XP_PROBE for the Ampex 9300 and
# Diva drives. Read through /sys/pdpuba/hpreg.h and /sys/pdpuba/xp.c
# for information on how to initialize for these drives.
NXPC 1 # NXPD controllers (RH70/RH11 style)
NXPD 2 # RM02/03/05, RP04/05/06, CDC 9766,
# Ampex 9300, Diva, Fuji 160, SI Eagle.
XP_PROBE YES # check drive types at boot
NRAM 0 # RAM disk size (512-byte blocks)
#########################################
# PERIPHERALS: TAPE DRIVES #
#########################################
NHT 2 # TE16, TU45, TU77
# Setting AVIVTM configures the TM driver for the AVIV 800/1600/6250
# controller (the standard DEC TM only supports 800BPI). For more details,
# see /sys/pdpuba/tm.c.
NTM 2 # TM11
AVIVTM YES # AVIV 800/1600/6250 controller
NTS 2 # TS11
NTMSCP 2 # TMSCP controllers
NTMS 2 # TMSCP drives
TMSCP_DEBUG NO # debugging code in TMSCP drive (EXPENSIVE)
#########################################
# PERIPHERALS: TERMINALS #
#########################################
# NKL includes both KL11's and DL11's.
# It should always be at least 1, for the console.
NKL 1 # KL11, DL11
NDH 0 # DH11; NDH is in units of boards (16 each)
CS02 NO # DH units above are really Emulex CS02
# boards on a 22bit Qbus.
NDM 0 # DM11; NDM is in units of boards (16 each)
NDHU 0 # DHU11
NDHV 0 # DHV11
NDZ 0 # DZ11; NDZ is in units of boards (8 each)
#########################################
# PERIPHERALS: OTHER #
#########################################
NDN 0 # DN11 dialer
NLP 0 # Line Printer
LP_MAXCOL 132 # Maximum number of columns on line printers
NDR 0 # DR11-W
#########################################
# PSEUDO DEVICES, PROTOCOLS, NETWORKING #
#########################################
# Networking only works with split I/D and SUPERVISOR space, i.e. with the
# 11/44/45/50/53/55/70/73/83/84. NETHER should be non-zero for networking
# systems using any ethernet. CHECKSTACK makes sure the networking stack
# pointer and the kernel stack pointer don't collide; it's fairly expensive
# at 4 extra instructions for EVERY function call AND return, always left
# NO unless doing serious debugging.
INET NO # TCP/IP
CHECKSTACK NO # Kernel & Supervisor stack pointer checking
NETHER 0 # ether pseudo-device
# Note, PTY's and the select(2) system call do not require the kernel to
# be configured for networking (INET). Note that you can allocate PTY's
# in any number (multiples of 8, of 16, even, odd, prime, whatever). Nothing
# in the kernel cares. PTY's cost 78 bytes apiece in kernel data space. You
# should probably have at least 8-10 since several applications use them:
# script, jove, window, rlogin, ...
NPTY 0 # pseudo-terminals - GENERIC sys needs NONE
# To make the 3Com Ethernet board work correctly, splimp has to be promoted
# to spl6; splfix files that do this are in conf/3Com; the config script
# does the right thing.
NEC 0 # 3Com Ethernet
NDE 0 # DEUNA/DELUA
NIL 0 # Interlan Ethernet
NSL 0 # Serial Line IP
NQE 0 # DEQNA
NQT 0 # DEQTA (DELQA-YM, DELQA-PLUS)
NVV 0 # V2LNI (Pronet)
NACC 0 # ACC LH/DH ARPAnet IMP interface
PLI NO # LH/DH is connected to a PLI
NIMP 0 # ARPAnet IMP 1822 interface
# The following are untested in 2.11BSD; some are untested since before 2.9BSD
# Some won't even compile. Most will require modification. Good luck.
ENABLE34 NO # if have the ENABLE34 board
NCSS 0 # DEC/CSS IMP11-A ARPAnet IMP interface
NDMC 0 # DMC11
NEN 0 # Xerox prototype (3 Mb) Ethernet
NHY 0 # Hyperchannel
NS 0 # Xerox NS (XNS)
NSRI 0 # SRI DR11c ARPAnet IMP
NTB 0 # RS232 interface for Genisco/Hitachi tablets
# Defining FPSIM to YES compiles a floating point simulator into the kernel
# which will catch floating point instruction traps from user space. This
# doesn't work at present.
FPSIM NO # floating point simulator
# To enable profiling, the :splfix script must be changed to use spl6 instead
# of spl7 (see conf/:splfix.profile), also, you have to have a machine with a
# supervisor PAR/PDR pair, i.e. an 11/44/45/50/53/55/70/73/83/84, as well
# as both a KW11-L and a KW11-P.
#
# Note that profiling is not currently working. We don't have any plans on
# fixing it, so this is essentially a non-supported feature.
PROFILE NO # system profiling with KW11P clock
INGRES NO # include the Ingres lock driver
This is a real example of using the disklabel program to place a label on a disk. User input is in bold type. The disklabel program was loaded from a bootable TK50. The disk being labeled in a RD54. The BOOT> prompt is from the 11/73 console ODT, if you are using an 11/44 the prompt will be >>>.
The first thing that is done is request disklabel to create a default partition ('a') which spans the entire disk. Some disk types have fixed sizes and geometries, for example RK05 (rk), RK06/7 (hk) and RL02 (rl) drives. With this type of disk the standalone disklabel program will generate a label with the correct geometry and 'a' partition size. With MSCP ('ra') disks disklabel will query the controller for the information it needs. The last type of disk, SMD (xp), presents many problems, disklabel will attempt to determine the drive type and geometry but you will have to verify the information.
Indented paragraphs like this one are explanatory comments and
are not part of the output from the disklabel program. In the case of MSCP
drives the number of cylinders may be 1 too low. This is discussed in the
example below.
BOOT> MU 0
73Boot from tms(0,0,0) at 0174500
: tms(0,1)
Boot: bootdev=06001 bootcsr=0174500
disklabel
Disk? ra(0,0)
d(isplay) D(efault) m(odify) w(rite) q(uit)? D
d(isplay) D(efault) m(odify) w(rite) q(uit)? d
type: MSCP
disk: RD54
flags:
bytes/sector: 512
sectors/track: 17
tracks/cylinder: 15
sectors/cylinder: 255
cylinders: 1220
rpm: 3600
drivedata: 0 0 0 0 0
1 partitions:
# size offset fstype [fsize bsize]
a: 311200 0 2.11BSD 1024 1024 # (Cyl. 0 - 1220*)
The columns do not line up nicely under the headings due to limitations
of the sprintf() routine in the standalone I/O package. There is no
capability to justify the output. It should be obvious which column
belongs under which heading. The '*' says that the partition does not
end on a cylinder boundary. This is due to the peculiar way in which
MSCP returns the geometry information: sectors/track * tracks/cylinder *
cylinders != sectors per volume.
d(isplay) D(efault) m(odify) w(rite) q(uit)? m
modify
d(isplay) g(eometry) m(isc) p(artitions) q(uit)? m
It is normally not necessary to change the geometry of an MSCP disk. On
the other hand it will almost always be necessary to specify the geometry
of an SMD drive (one which uses the XP driver). Since the drive being
labeled is an MSCP drive the next step is to set the pack label to something
other than DEFAULT.
modify misc
d(isplay) t(ype) n(ame) l(able) f(lags) r(pm) D(rivedata) q(uit)? l
label [DEFAULT]: TESTING
modify misc
d(isplay) t(ype) n(ame) l(able) f(lags) r(pm) D(rivedata) q(uit)? q
modify
d(isplay) g(eometry) m(isc) p(artitions) q(uit)? p
modify partitions
d(isplay) n(umber) s(elect) q(uit)? d
type: MSCP
disk: RD54
flags:
bytes/sector: 512
sectors/track: 17
tracks/cylinder: 15
sectors/cylinder: 255
cylinders: 1220
rpm: 3600
drivedata: 0 0 0 0 0
1 partitions:
# size offset fstype [fsize bsize]
a: 311200 0 2.11BSD 1024 1024 # (Cyl. 0 - 1220*)
modify partitions
d(isplay) n(umber) s(elect) q(uit)? s
a b c d e f g h q(uit)? a
sizes and offsets may be given as sectors, cylinders
or cylinders plus sectors: 6200, 32c, 19c10s respectively
modify partition 'a'
d(isplay) z(ero) t(ype) o(ffset) s(ize) f(rag) F(size) q(uit)? s
'a' size [311200]: 15884
d(isplay) z(ero) t(ype) o(ffset) s(ize) f(rag) F(size) q(uit)? q
modify partitions
d(isplay) n(umber) s(elect) q(uit)? s
a b c d e f g h q(uit)? b
sizes and offsets may be given as sectors, cylinders
or cylinders plus sectors: 6200, 32c, 19c10s respectively
modify partition 'b'
d(isplay) z(ero) t(ype) o(ffset) s(ize) f(rag) F(size) q(uit)? o
'b' offset [0]: 15884
modify partition 'b'
d(isplay) z(ero) t(ype) o(ffset) s(ize) f(rag) F(size) q(uit)? s
'b' size [0]: 16720
modify partition 'b'
d(isplay) z(ero) t(ype) o(ffset) s(ize) f(rag) F(size) q(uit)? t
'b' fstype [unused]: swap
modify partition 'b'
d(isplay) z(ero) t(ype) o(ffset) s(ize) f(rag) F(size) q(uit)? q
modify partitions
d(isplay) n(umber) s(elect) q(uit)? s
a b c d e f g h q(uit)? c
sizes and offsets may be given as sectors, cylinders
or cylinders plus sectors: 6200, 32c, 19c10s respectively
modify partition 'c'
d(isplay) z(ero) t(ype) o(ffset) s(ize) f(rag) F(size) q(uit)? o
'c' offset [0]: 0
modify partitions 'c'
d(isplay) z(ero) t(ype) o(ffset) s(ize) f(rag) F(size) q(uit)? s
'c' size [0]: 311200
modify partitions 'c'
d(isplay) z(ero) t(ype) o(ffset) s(ize) f(rag) F(size) q(uit)? t
'c' fstype [unused]: unused
modify partitions 'c'
d(isplay) z(ero) t(ype) o(ffset) s(ize) f(rag) F(size) q(uit)? q
modify partitions
d(isplay) n(umber) s(elect) q(uit)? s
a b c d e f g h q(uit)? g
sizes and offsets may be given as sectors, cylinders
or cylinders plus sectors: 6200, 32c, 19c10s respectively
modify partition 'g'
d(isplay) z(ero) t(ype) o(ffset) s(ize) f(rag) F(size) q(uit)? o
'g' offset [0]: 32604
modify partition 'g'
d(isplay) z(ero) t(ype) o(ffset) s(ize) f(rag) F(size) q(uit)? s
'g' size [0]: 278596
modify partition 'g'
d(isplay) z(ero) t(ype) o(ffset) s(ize) f(rag) F(size) q(uit)? t
'g' fstype [unused]: 2.11BSD
modify partition 'g'
d(isplay) z(ero) t(ype) o(ffset) s(ize) f(rag) F(size) q(uit)? q
modify partitions
d(isplay) n(umber) s(elect) q(uit)? n
Number of partitions (8 max) [7]? 7
modify partitions
d(isplay) n(umber) s(elect) q(uit)? q
modify
d(isplay) g(eometry) m(isc) p(artitions) q(uit)? d
type: MSCP
disk: RD54
label: TESTING
flags:
bytes/sector: 512
sectors/track: 17
tracks/cylinder: 15
sectors/cylinder: 255
cylinders: 1220
rpm: 3600
drivedata: 0 0 0 0 0
7 partitions:
# size offset fstype [fsize bsize]
a: 15884 0 2.11BSD 1024 1024 # (Cyl. 0 - 62*)
b: 16720 15884 swap # (Cyl. 62*- 127*)
c: 311200 0 unused 1024 1024 # (Cyl. 0 - 1220*)
g: 278596 32604 2.11BSD 1024 1024 # (Cyl. 127- 1220*)
modify
d(isplay) g(eometry) m(isc) p(artitions) q(uit)? q
On MSCP disks it is possible you will see a warning error like this:
partition c: extends past end of unit 0 311200 311100
partition g: extends past end of unit 32604 278596 311100
This is not cause for panic. What this is saying is that the number of cylinders is one too low. MSCP devices do not necessarily use all of the last cylinder. The total number of blocks is precisely known for MSCP devices (it is returned in the act of bringing the drive online). However the number of sectors on the volume is not necessarily evenly divisible by the number of sectors per track (311200 divided by 17*15 gives 1220.392). Basically the last cylinder is not fully used. What must be done is raise the number of cylinders by 1.
NOTE: For any other disk type it is cause for concern if the warning above is issued - it means that incorrect partition or geometry information was entered by the user and needs to be corrected.
d(isplay) D(efault) m(odify) w(rite) q(uit)? m
modify
d(isplay) g(eometry) m(isc) p(artitions) q(uit)? g
modify geometry
d(isplay) s(ector/trk) t(rk/cyl) c(yl) S(ector/cyl) q(uit)? c
cylinders [1220]: 1221
modify geometry
d(isplay) s(ector/trk) t(rk/cyl) c(yl) S(ector/cyl) q(uit)? q
modify
d(isplay) g(eometry) m(isc) p(artitions) q(uit)? q
d(isplay) D(efault) m(odify) w(rite) q(uit)? w
d(isplay) D(efault) m(odify) w(rite) q(uit)? q
73Boot from tms(0,0,1) at 0174500
: ra(1,0,0)unix
ra1 csr[00]: 0172154
The last string entered shows how I boot from an alternate controller. In
normal use, i.e. with a single MSCP controller, the string would simply
be ra(0,0)unix.
The standalone disklabel program is the second file on a boot tape (after the bootblocks and boot program). It is used to place an initial label on a disk describing the disk and its partitions. The program is also used when the root ('a') or swap ('b') partitions of a previously labeled system disk must be modified. The second use is mandated because the root and swap partitions can not be modified while the kernel has them open.
disklabel effectively runs in CBREAK mode - you do not need to hit the RETURN key except when prompted for a multicharacter response such as a string (the pack label) or a number (partition size). Defaults are placed inside square brackets ([default]). Entering RETURN accepts the default.
The program is organized into several levels. disklabel prints the current level out before prompting. At each level there is always the choice of d(isplaying) the current label and q(uit)ing the current level and returning to the previous level. If you are at the top level and enter q the program will exit back to Boot unless you have made changes to the disklabel. In that case you will be asked if you wish to discard the changes, if you answer y the changes will be discarded. If the answer is n the q is ignored and disklabel does not exit.
In the following paragraphs the convention is to bold the user input while leaving the output from disklabel in normal type. The devices used were a TK50 and an RD54, thus the tape device is tms and the disk device is ra.
The TK50 was booted resulting in the usual message from Boot:
73Boot from tms(0,0,0) at 0174500
: tms(0,1)
Boot: bootdev=06001 bootcsr=0174500
disklabel
Disk? ra(0,0)
d(isplay) D(efault) m(odify) w(write) q(uit)? m
The 'D' option will request disklabel to create a default label based
on what the program can determine about the drive. For some devices, such
as RL01/02, RK06/07, MSCP (RD54, RA81, usw.), disklabel can
determine what the drive type is and how many sectors it has. For other
devices, such as SMD drives supported by the xp driver, the task is
complicated by the number of different controllers and emulations supported.
Some 3rd party controllers have capabilities that DEC controllers do not and
the xp has no way of knowing exactly which type of controller is
present. In this case disklabel will guess and then depend on
you to enter the correct data.
modify
d(isplay) g(eometry) m(isc) p(artitions) q(uit)? g
modify geometry
d(isplay) s(ector/trk) t(rk/cyl) c(yl) S(ector/cyl) q(uit)? q
The Sector/cyl entry is rarely used. disklabel will calculate
this quantity for you from the sector/trk and trk/cyl quantities.
modify
d(isplay) g(eometry) m(isc) p(artitions) q(uit)? m
d(isplay) t(ype) n(ame) l(abel) f(lags) r(pm) D(rivedata) q(quit)? f
Type is one of: SMD, MSCP, old DEC, SCSI, ESDI, ST506, floppy.
Name is a string up to 16 characters in length. It is typically something like rd54 or rm03 but may be any meaningful string.
Label is an arbitrary string up to 16 characters in length - nothing in the system checks for or depends on the contents of the pack label string.
Rpm is the rotational speed of the drive. Nothing in the system uses or depends on this at the present time. Default is 3600.
Drivedata consists of 5 longwords of arbitrary data. Reserved for future use.
modify misc flags
d(isplay) c(lear) e(cc) b(adsect) r(emovable) q(uit)? q
Ecc says that the controller/driver can correct errors.
Badsect indicates that the controller/driver supports bad sector replacement.
Removable indicates that the drive uses removable media (floppy, RL02,
RA60 for example).
modify misc
d(isplay) t(type) n(ame) l(abel) f(lags) r(pm) D(rivedata) q(uit)? q
modify
d(isplay) g(eometry) m(isc) p(artitions) q(uit)? p
modify partitions
d(isplay) n(umber) s(elect) q(uit)? n
Number of partitions (8 max) [7]? 7
This is the highest partition number considered to be valid.
disklabel will adjust this parameter semi-automatically at the
p level but it may be necessary to use n in cases where
some partitions are not to be used or contain invalid information.
modify partitions
d(isplay) n(umber) s(elect) q(uit)? s
a b c d e f g h q(uit)? a
sizes and offsets may be given as sectors, cylinders
or cylinders plus sectors: 6200, 32c, 19c10 respectively
modify partition 'a'
d(isplay) z(ero) t(ype) o(ffset) s(ize) f(rag) F(size) q(uit)? q
Zero clears the size and offset fields of a partition entry and sets the filesystem type to unused.
Type is the filesystem type and of the possible choices only 2.11BSD, swap and unused make any sense to specify.
Offset is the number of sectors from the beginning of the disk at which the partition starts.
Size is the number of sectors which the partition occupies.
Frag is the number of fragments a filesystem block can be broken into. It is not presently used and should be left at the default of 1.
Fsize is the filesystem blocksize and should be left at the default of 1024.
modify partitions
d(isplay) n(umber) s(elect) q(uit)? q
modify
d(isplay) g(eometry) m(isc) p(artitions) q(uit)? q
d(isplay) D(efault) m(odify) w(write) q(uit)? q
Label changed. Discard changes [y/n]? y
73Boot from tms(0,0,1) at 0174500
Define only those partitions you actually will use. There is no need to set up all 8 partitions. Drives less than 200Mb probably will only have 3 partitions defined, 'a', 'b' and 'd' for /, swap and /usr respectively. Remember to set the number of partitions. Disklabel will attempt to do this for you by keeping track of the highest partition you modify but this is not foolproof.
Do not define overlapping partitions unless you are sure what you are doing. disklabel will warn you of overlapping partitions but will not prohibit you from writing such a label to disk.
Remember that the prompt levels nest in disklabel. It will be necessary in several cases to enter multiple q commands to get back to the top level.
IMPORTANT: Keep at least 1, preferably more, bootable tape or floppy with disklabel on it present at all times. If the label on a disk ever becomes corrupted the kernel will be very unhappy and probably won't boot. If this happens you will need to boot the standalone disklabel program and relabel the disk. At least 2.11BSD provides a standalone disklabel - previous 4BSD systems which implemented disklabels did not and the cold-start of those systems was painful indeed.
IMPORTANT: Write down in at least one place, and keep with the tape/floppy mentioned above, the geometry and partition layout you assign to the disk. The disklabel(8) program should be used to produce a hardcopy of the disklabel.
Note: The standalone tape drive unit number is specially encoded to specify both unit number and tape density (BPI). Most tape subsystems either automatically adjust to tape density or have switches on the drives to force the density to a particular setting, but for those which don't the following density select mechanisms may be necessary. The ts only operates at 1600BPI, so there is no special unit density encoding. The ht will operate at either 800BPI or 1600BPI. Units 0 through 3 corresponding to 800BPI, and Units 4 through 7 corresponding to 1600BPI on drives 0 through 3 respectively. The standard DEC tm only supports 800BPI (and hence can't be used with the 2.11BSD distribution tape), but several widely used tm emulators support 1600BPI and even 6250BPI. Units 0 through 3 corresponding to 800BPI, Units 4 through 7 corresponding to 1600BPI, and Units 8 through 11 corresponding to 6250BPI on drives 0 through 3 respectively.