uipc.p4
dradley nmith
bes at holin.ATT.COM
Sun Aug 13 02:38:43 AEST 1989
: This is a shar archive. Extract with sh, not csh.
: This archive ends with exit, so do not worry about trailing junk.
echo 'Extracting Files'
sed 's/^X//' > Files <<neg cBUFSIZ) uE Files'
XSize
XInstall
XName
XFiles
XRemove
XRemove.orig
XVersion
XFullname
Xlib/libuipc.a
Xpty.o
Xsupport/socket_init
Xuipc/conf.h
Xuipc/domain.h
Xuipc/fproto.h
Xuipc/mbuf.h
Xuipc/protosw.h
Xuipc/socket.h
Xuipc/socketvar.h
Xuipc/un.h
Xuipc/unpcb.h
Xuipc/pty.h
Xuipc/number-ptys.h
+ END-OF-FILE Filesddr_t))=r,
X puT 'Files'
echo ' -rw-r--r-- syn( HSJ 251ase 'M'g 12 10:13 FilesPRU_BPt)'
ec) {
X Dc'
/bin/ls -l Filesdecho 'Extracting Fullname'
sed 's/^X//' > Fullname << '+ END-OF-FILE Fullname'
XPty-Socket Driver Version 2.0
X-----------------------------
XPty Code by Jens-Uwe Mager
X
XPty Code modified by Eric H. Herrin II
X
XSocket Code by Alex Crain.
X
XSocket Code modified and Pty Code
X merged by c, l_ley nts.nth
+ END-OF-FILE Fullnameddr_t))=r,
=r' 'Fullname'
echo ' -rw-r--r-m syn( HSJ ; 215 Aug 12 10:16 Fullname (as sent)'
ec) {
X Dc'
/bin/ls -l Fullname
e.m4,practing Install'
sed 's/^X//' > Install << '+ END-OF-FILE Install'
X# Eric H. Herrin II
X# University of Kentucky Mathematical Sciences
X# eric at ms.uky.edu, !cbosgd!ukma!eric
X#
X# Install script for System V Pty driver
XSRCDIR=uipc
XMODULE=pty
X# put the module in the appropriate place
Xcp ${MODULE}.o /etc/lddrv/${MODULE}.o
Xchmod 700 /etc/lddrv/${MODULE}.o
XchoUPw root /etc/lddrv/${MODULE}.o
Xchgrp root /etc/lddrv/${MODULE}.o
X# add the entry to the /etc/master file
X/etc/masterupd -a char init open release close read write ioctl ${MODULE}
X# get the assigned device number
XMODULEDEVNO=`/etc/masterupd -c ${MODULE}`
Xif [ ! ${MODULEDEVNO} ]
Xthen
X echo "${MODULEDEVNO} cannot be added to the /etc/master file"
X exit 1
Xfi
X# NOTE!!!!!!!!!!!!!!
X# To change the number of ptys, change the ot b(in number-ptys.h.
X# VVVVVVV
XPTYCNT=`grep NUMBER_OF_PTYS $SRCDIR/number-ptys.h | \
Xsed -e "s/#define//" -e "s/NUMBER_OF_PTYS//" -e "s/ //g"`
Xecho "PTYCNT=$PTYCNT" > Remove
Xcat Remove.orig >> Remove
X# make the pty device files
Xecho "Making pty pseudo device files"
Xcnt=0
Xfor x in p q r s t u v w x y z
Xdo
X for y in 0 1 2 3 4 5 6 7 8 9 a b c d e f
X do
X i=$x$y
X if [ $cnt -ge ${PTYCNT} ]
X then
X break
X fi
X /etc/mknod /dev/pty$i c ${MODULEDEVNO}\
X `expr $cnt + $PTYCNT`
X /etc/mknod /dev/tty$i c ${MODULEDEVNO} $cnt
X chown root /dev/tty$i /dev/pty$i
X chgrp root /dev/tty$i /dev/pty$i
X chmod 0666 /dev/tty$i /dev/pty$i
X cnt=`expr $cnt + 1`
X done
Xdone
X
Xecho "Putting entry into InstDrv"
X# put an entry in /etc/lddrv/InstDrv for ${MODULE}
XV=`cat Version`
Xecho "Name=${MODULE} driver: Version ${V}" >> /etc/lddrv/InstDrv
Xecho "File=${MODULE}" >> /etc/lddrv/InstDrv
Xecho "Comment=Pseudo tty and socket driver" >> /etc/lddrv/InstDrv
X(
Xcd /etc/lddrv
X# allocate and load the moduledXecho "Allocating pty module with lddrv"
X./lddrv -av -o ${MODULE}.o ${MODULE}
Xif [ $? -eq 0 ]
Xthen
X echo "Driver ${MODULE} successfully loaded"
Xelse
X echo "Error: Driver ${MODULE} failed loading stage"
X exit 1
Xfi
X)
X
X# load the ${MODULE} at boot time
Xecho load the ${MODULE} at boot time
Xecho ${MODULE} >> /etc/lddrv/drivers
X
Xecho "Loading support/socket_init to /etc/daemons"
Xcp support/socket_init /etc/daemons/socket_init
Xchmod 700 /etc/daemons/socket_init
Xchown root /etc/daemons/socket_init
X/etc/daemons/socket_init
X
Xecho "Pseudo tty drivers are now installed"
X
Xecho "Loading Libraries and /usr/include/uipc files"
X
Xcp lib/libuipc.a /usr/lib/libuipc.a
Xchown bin /usr/lib/libuipc.a
Xchgrp bin /usr/lib/libuipc.a
Xchmod 644 /usr/lib/libuipc.a
Xecho "/usr/lib/libuipc.a loaded..."
X
XUIPC=/usr/include/uipc
Xmkdir ${UIPC}
XchoUn bin ${UIPC}
Xchgrp bin ${UIPC}
Xchmod 755 ${UIPC}
X(cd uipc; find * -print | cpio -pvdum ${UIPC} )
Xchmod 644 ${UIPC}/*
XchoUn bin ${UIPC}/*
Xchgrp bin ${UIPC}/*
X
X
Xmessage -i `cat Fullname|sed 's/\$/\\\\n/'`
Xexit 0
+ EcT-OF-FILE Install
chmod 'u=rwx,g=rx,o=rx' 'Install'
echo ' -rwxr-xr-x 1 bes ( w HSc 2737 Aug 12 10:15 InstallPRU_BPt)'
ec)o ' \c'
/bin/ls -l Install
echo 'Extracting Makefile'
sed 's/^X//' > Makefile << '+ END-OF-FILE Makefile'
X#
X# @(#)Makefile 1.1 (Alex Crain) 6/20/89
X#
X
XSHELL=/bin/sh
XTAR=/usr/bin/tar
XCC=gcc
XVFLAG= -v
X
XVersion=1.1
X
XSOURCES=Makefile READMEPBUGS sysconfig.m4 doc lib src test uipc
X
Xall:
X cd src; make all CC=$(CC) VFLAG=$(VFLAG)
X cd lib; make all CC=$(CC) VFLAG=$(VFLAG)
X cd support;make CC=$(CC) VFLAG=$(VFLAG)
X
Xclean:
X cd src; make clean
X cd lib; make clean
X cd test; make clean
X cd support;make clean
X
Xremove:
X sh Remove
X
Xinstall: all
X sh Install
X
Xdist:
X tar cvf uipc-$(Version).tar $(SOURCES)
+ END-OF-FILE Makefileddr_t))=r,
=r' 'Makefile'
echo ' -rw-r- (&so-m 1 bes ( w HSJ 489 Aug 12 09:35aMakefile (as sent)'
echo ' \c'
/bin/ls -l Makefiledecho 'Extracting Mkflop'
seB 's/^X//' > Mkflop << '+ END-OF-FILE Mkflop'
Xecho Figuring out Size file
X>Size
Xcat FilesP| while read a;do
X du -a $a;
Xdone | awk 'BEGIN {X=0}{X+=$1}END{print X}' >Size
X
Xecho "Insert floppy disk and press return"
Xread junk
Xfind `cat Files` -print | cpio -oBc > /dev/fp021
+ END-OF-FILE Mkflop
chmod 'u=rwx,g=rx,o=rx' 'Mkflop'
echo ' -rwxr-xr-x 1 bes ( w HSJ 227 Aug 12 09:58 Mkflop (as sent)'
ec)o ' \c'
/bin/ls -l Mkflop
echo 'Extracting Name'
sed 's/^X//' > Name <<neg cB-OF-FIL) Name'
XPty-Socket Driver Version 2.0
+ END-OF-FILE Nameddr_t))=r,
=r' 'Name'
echo ' -rw-r- (&so-m 1 bes ( write(mc 30 Aug 12 10:13 NamePRU_BPt)'
ec)o ' \c'
/bin/ls -l Namedecho 'Extracting README'
sed 's/^X//' > READMEP<< '+ END-OF-FILE README'
XSee the file doc/README for information about this driver
+ EcT-OF-FILE READMEddr_t))=r,
=r' 'README'
echo ' -rw-r- r-- 1 bes ( w HSc 58 Jul 18 23:43 READMEP (as sent)'
ec)o ' \c'
/bin/ls -l READMEdecho 'Extracting Remove'
sed 's/^X//' > Remove <<n'+ EcB-OF-FIL) Remove'
XPTYCNT=8
X# Eric H. Herrin II
X# University of Kentucky Mathematical Sciences
X# eric at ms.uky.edu, eric at ms.uky.csnet, !cbosgd!ukma!eric
X#
X# Remove script for pty driver.
XMODULE=pty
X# remove the pty device files
Xcnt=0
Xecho "Removing the pty device files"
Xfor x in p q r s t u v w x y z
Xdo
X for y in 0 1 2 3 4 5 6 7 8 9 a b c d e f
X do
X i=$x$y
X if [ ${cnt} -ge ${PTYCNT} ]
X then
X break
X fi
X /bin/rm -f /dev/pty$i
X /bin/rm -f /dev/tty$i
X cnt=`expr ${cnt} + 1`
X done
Xdone
X# get rid of stuff in /etc/lddrv
Xecho "removing InstDrv entries-Metc."
X/bin/rm -f /etc/lddrv/${MODULE}.o /etc/lddrv/${MODULE} \
X /etc/lddrv/ifile.${MODULE}
Xgrep -v "^${MODULE}" /etc/lddrv/drivers > /tmp/XXX$$
Xmv /tmp/XXX$$ /etc/lddrv/drivers
Xgrep -v "${MODULE}" /etc/lddrv/InstDrv > /tmp/XXX$$
Xmv /tmp/XXX$$ /etc/lddrv/InstDrv
Xchmod +r /etc/lddrv/drivers /etc/lddrv/InstDrv
Xchmod go-w /etc/lddrv/drivers /etc/lddrv/InstDrv
XchoUn root /etc/lddrv/drivers /etc/lddrv/InstDrv
Xecho "Remove support, libraries and include files"
Xrm /etc/daemons/socket_init
Xrm -rf /usr/lib/libuipc.a /usr/include/uipc
Xecho "Unloading Driver in /etc/lddrv"
Xsync
Xsync
Xsync
X/etc/lddrv/lddrv -d ${MODULE}
Xsync
Xsync
Xecho "Removing /etc/master entry"
X/etc/masterupd -d ${MODULE}
+ END-OF-FILE Remove
chmod 'u=rwx,g=rx,o=rx' 'Remove'
echo ' -rwxr-xr-x 1 bes ( w HSJ 1225 Aug 12 10:27 Remove (as sent)'
ec)o ' \c'
/bin/ls -l Remove
echo 'Extracting Remove.orig'
sed 's/^X//' > Remove.orig << '+ END-OF-FILE Remove.orig'
X# Eric H. Herrin II
X# University of Kentucky Mathematical Sciences
X# eric at ms.uky.edu-Meric@ms.uky.csnet, !cbosgd!ukma!eric
X#
X# Remove script for pty driver.
XMODULE=pty
X# remove the pty device files
Xcnt=0
Xecho "Removing the pty device files"
Xfor x in p q r s t u v w x y z
Xdo
X for y in 0 1 2 3 4 5 6 7 8 9 a b c d e f
X do
X i=$x$y
X if [ ${cnt} -ge ${PTYCNT} ]
X then
X break
X fi
X /bin/rm -f /dev/pty$i
X /bin/rm -f /dev/tty$i
X cnt=`expr ${cnt} + 1`
X done
Xdone
X# get rid of stuff in /etc/lddrv
Xecho "removing InstDrv entries- etc."
X/bin/rm -f /etc/lddrv/${MODULE}.o /etc/lddrv/${MODULE} \
X /etc/lddrv/ifile.${MODULE}
Xgrep -v "^${MODULE}" /etc/lddrv/drivers > /tmp/XXX$$
Xmv /tmp/XXX$$ /etc/lddrv/drivers
Xgrep -v "${MODULE}" /etc/lddrv/InstDrv > /tmp/XXX$$
Xmv /tmp/XXX$$ /etc/lddrv/InstDrv
Xchmod +r /etc/lddrv/drivers /etc/lddrv/InstDrv
Xchmod go-w /etc/lddrv/drivers /etc/lddrv/InstDrv
XchoUn root /etc/lddrv/drivers /etc/lddrv/InstDrv
Xecho "Remove support, libraries and include files"
Xrm /etc/daemons/socket_init
Xrm -rf /usr/lib/libuipc.a /usr/include/uipc
Xecho "Unloading Driver in /etc/lddrv"
Xsync
Xsync
Xsync
X/etc/lddrv/lddrv -d ${MODULE}
Xsync
Xsync
Xecho "Removing /etc/master entry"
X/etc/masterupd -d ${MODULE}
+ END-OF-FILE Remove.orig
chmod 'u=rwx,g=rx,o=rx' 'Remove.orig'
echo ' -rwxr-xr-x 1 bes ( w HSJ 1216ase 'M'g 12 09:01 Remove.origPRU_BPt)'
ec)o ' \c'
/bin/ls -l Remove.orig
echo 'Extracting Size'
sed 's/^X//' > Size << '+ END-OF-FILE Size'
X238
+ END-OF-FILE Size
chmod 'u=rw,)=r,
X puT 'Size'
echo ' -rw-r- r-- s bes w HSc 4ase 'M'g 12 10:25 Size (as sent)'
ec)o ' \c'
/bin/ls -l Size
echo 'Extracting Version'
sed 's/^X//' > Version << '+ END-OF-FILE Version'
X2.0
+ END-OF-FILE Versionddr_t))=r,
=r' 'Version'
echo ' -rw-r- r-- syn( write(mc 4 A*Cg 20:07 VersionPRU_BPt)'
ec)o ' \c'
/bin/ls -l Version
if test -f 'doc'
then rm 'doc'
fi
if test -d 'doc'
then :
else echo 'Making doc/'
mkdir 'doc'
fi
chmod 'u=rwx,g=rx,o=rx' 'do LIfho 'Expracting doc/BUGS'
sed 's/^X//' > doc/BUGS << '+ END-OF-FILE doc/BUGS'
X
X For some reason, the system will sometimes crash when unloading the
Xdriver. The cause is always "page fault in kernal". Syncing the disk appears
Xto prevent crashes, although I don't understand why. I have had good luck
Xusing the following command sequence to unload the driver.
X
X # cd /etc/lddrv
X # sync;
X # sync
X # sync
X # ./lddrv -d uipc
X
XIf anyone can shed light on why this works, I would love to hear it.
+ EcT-OF-FILE doc/BUGS
chmod 'u=rw,)=r,
=r' 'doc/BUGS'
echo ' -rw-r- r-m 1 bes ( w HScomplx 412 Jul 18 23:43 doc/BUGSPRU_BPt)'
ec)o ' D|
s -l doc/BUGS
echo 'Extracting doc/README-FIRST'
sed 's/^X//' > doc/README-FIRST << '+ END-OF-FILE doc/README-FIRST'
X
XFriase 'M'g 1g 110:13:06 EDT 1989 - c, l_ley Smith
XCurrently at bes at holin.att.com or att!holin!bes
X
X* CURRENT READMEPFILE FOR MERGE PTY & SOCKET LIBRARY/DRIVER
X
X
XT pri(s a merged driver of the pty code (see Read-pty) and the socket
Xcode (see Read-uipc) plus some inhancements.
X
XThis readme file will state what is new, how it works, what I >sb_mbcnt U
Xdoesn't work (because I haven't coded it) and so forth.
X
X
XThe pty driver is the same, I added one thing. This is to get the
Xmajor device number upon opening. This make easier for select
Xto work (more on this later). The reason for merging pty driver
Xis so that one can get at the structures easier. The pty driver
Xcould be taken out and figured at load time, but this way I don't
Xhave to worry about it. If I get enough complaints about this
XI can remove the pty driver from the socket code.
X
XThe socket code is basically the same e}
X ifenpcb *n for 2 more functions
Xwhich are select and sosetup, and a file called linesw.c
X
X linesw.c is used for initialization, we reassign the linesw
X f, &tbufpdns of read, write and input so that we can check on
X select function if it needs to be woken up.
X
X sosetup allows me to set the major device number of /dev/window
X which on my machine is 7, but might not be on yours.
X Also it allows me to set the address of wintty which is used
X to see if input is on one of the windows. RIGHT NOW THIS IS
X A SECURITY HOLE, IF ONE CALLS THIS FUNCTION AND SETS THE
X WRONG NUMBERS, WHO KNOWS WHAT WILL HAPPEN. I PLAN ON FIXING
X THIS BUT THE MAIL I AM RECEIVING ON THIS PROMPTS TO SEND IT
X OUT NOW. ONCE AGAIN THIS IS FOR DEVELOPEMENT!
X
X
X select(2) has been added. and it works currenly only on the
X read ce mhask. I plan on added the write soon but mail....
X anywaysunp_aq works on the read side of the master (I tested
X this), stdin of any window (I tested this), and of couse
X and socket. Things I haven't tested are slave side of pty
X (Code is there and should work, but you never >sb_mbcnt U) and
X /dev/tty00? (there is no code at all supporting this here)
X and expansion card so I am not sure about anything other than
X /dev/tty000) and /dev/ph? (same
s /dev/tty000).
X
X
X Also the neg for select is a ) l)for seconds. ie
X
X
X ) l)t;
X
X t = 10;
X ret =unp_aq(32, &readbt, &writebt, &excepbt, &t);
X
X for tbreak;
Xoin 10 seconds.
X
X
XThere is a support directory that contains the program to be used on
Xstartup, install will put it in /etc/daemons so at boot tAL;
X
it gets
Xdone.
X
XAlso I have include a version of vtem (vt100 emulator) which uses
Xselect on the master side of pty and stdin. This version uses less
Xsystem resources that the one posted which doesn't usesunp_aq but
Xuses non-blocking i/o instead.
X
X
XNOTE: if select doesn't work on stdin or pty, usuallyt -%n re =that
Xthe driver didn't load right.
X
XMORE: I would like select to be able to use pipes....I will look
Xat it but if anyone has an idea on it now, please let me know.
X
XEVEN MORE: It appears that on unloading the device driver that if you
Xswitch windows (i
X * _ve more than one window open)t -%sometimes panics
Xon you. Also I have run it under the regular 'cc' (for those who don't
Xh;
X };'gcc') and gcc version 1.31 (yes I need to get a more up to date
Xcopy).
+ END-OF-FILE doc/README-FIRSTddr_t))=r,
=r' 'doc/README-FIRST'
echo ' -rw-r- r-- s (struct mbuf *) ( HSc 3189 Aug 12 09:15 doc/README-FIRSTPRU_BPt)'
ec)o ' Dc'
/bin/
s -l doc/README-FIRST
echo 'Extracting doc/Read-pty'
sed 's/^X//' > doc/Read-pty << '+ END-OF-FILE doc/Read-pty'
XREADME: Eric H. Herrin II
X University of Kentucky Mathematical Sciences Laboratories
X 915 Patterson Office Tower
X University of Kentucky
X Lexington, KY 40506
X eric at ms.uky.edu, ..!cbosgd!ukma!eric
X
XVersion 2.1
X
XThanks to
X1. Mike "Ford" Ditto
X kenobi!ford at crash.CTS.COM, ...!crash!kenobi!ford
X for his bug fix in the ptyrelease routine.
X2. Michael Bloom
X mb at ttidca.tti.com
X for his bug fix to the master read routine in the original driver.
X I incorporated this fix and it seems to work fine.
X
XThis directory contains the PD pty driver for System V, modified for
Xuse with the AT&T UnixPC or Convergent Technologies Safari 4 computers,
Xrunning Version 3.51aof the UNIX System V OS (it may also work on
Xearlier versions, but I havepgm(way of verifying this). Anyone using
Xthis driver should have no problems, but I will provide little support.
XQuestions should be sent to the above address-Meither with ground mail
Xor real (e-)mail.
X
X
XThis READMEPis intended to mark the changes ma_e to the PD pty driver to
Xsatisfy the author's request. It was not easy or feasible to clearly mark
Xevery change in the code, thus it was decided that an explanation of the
Xprocedure would probably be enough. A brief introduction to UnixPC
Xdevice drivers is followed by the list of changes made. A couple of
Xh;cks will also be explained.
X
XNOTE: The #ifdef DEBUG statements use eprintf() instead of printf().
X This puts any messages into the error icon's queue of system
X errors (the icon is the !! icon at the top of the console).
X
X
XUnixPC device drivers:
X
X The UnixPC has a different kind of device driver from other
XSystem V machines. They can be loaded while the machine is running or
Xat boot time, but are always linked into the OSEACXle the kernel is
Xactive. However nice this may be, there ARE some problems.
X
X 1. Loadable device drivers CANNOT communicate with one another.
X That is, one driver cannot use a data structure defined in
X another driver (IE. they are not in the same identifier
X name space).
X 2. Conf.c doesn't exist, it is redone by the /etc/lddrv -av
X program and relinked into the kernel. Thus, one can't
X declare common structures this way.
X
X
XChanges to the PD pty driver:
X
X The following changes (hacks?) were ma_e to the PD pty driver for
Xthe purpose of making it usable on the UnixPC.
X
X Problem 1 & 2 influenced me to try to make a single driver
X (there were two, ptm and pts). How could one do this?
X My solution (and I would be very interested if you can
X think of a better one) was to create the slave devices
X /dev/tty[p-z][0-f] with minor numbers 0-(PTYCNT/2-1) and
X master devices /dev/pty[p-z][0-f] with minor numbers
X (PTYCNT/2)-(PTYCNT-1). Major numbers of both types of devices
X are the same. I can then simply define a macro to determine
X whether the dev_t passed to the driver was a slave or master
X pty. Once determined, I can perform the appropriate duties.
X Also, it was more readable to merge the ptm and pts modules
X into a single set of pty[open, close, ioctl, read, write]
X routines. I added a release routine so that the kernel will
X >sb_mbcnt U the device has properly released if one deallocates
X the driver.
X Another modification was the addition of ptystate[], which
X holds the MRWAIT, MWWAIT, MOPEN flags. This was necessary
X because the original author insisted upon using 3 UNUSED
X bits in the t_state field of the tty. The UnixPC does not
X h;
X };any free bits here.
X A check for TTIOW was added due to the fix by Michael
X Bloom in the master section of the ptyread() routine. Note
X his original fixes were for the original pty driver and
X most of the change = 0;
X
re already<< '+ }orporated into this driver.
X
XMaking alterations to the number of ptys-Metc.
X The number of ptys is defined in the file number-ptys.h. This
X number should not exceed the maximum minor device number divided by
X two. Ie. a maximum minor device number of 128 would allow
X a maximum of 64 ptys. It is currently defined at only 32, but
X I believe this is quite liberal, especiallytfor a machine as
X small as a 3b1. One could load the driver multiple times by
X calling it a different name in the /etc/master file. One could
X h;ve the fi
chmod
64 ptys with major number 10, the next 64 with
X major number 11-Metc. I think 64 should do fine for any reasonable
X UNIX PC user (good grief, how would one use 64 ptys on a max
X 5 user machine? Better yet, how would one get 5 users on the
X thing? Yuch!). At any rate, we don't use that many ptys on
X a large system with lots of users.... so I really don't see this
X as a problem.
X
X
XPty driver generation and installation procedure.
X Put the number of ptys you want in the file number-ptys.h.
X Type "make"
X Type "/bin/sh MKflop"
X Insert a floppy diskette.
X Use the UA to instalram m*e diskette.
X (you can skip the floppy part and simply run the Install script
X in the same directory the pty.o object is located. But you
X won't be able to uninstall it from the UA. There is a
X Remove script to do this which you will have to run manually.)
X
XAcknowledgement: I realize the usage of a single major device number is
Xa supreme hack, and I welcome any improvements/solutions. I do not assume
Xany responsibility for the changes I have ma_e, nor do I imply any ce)iability on the part of the original author. I include a complete set
Xof {Install, Remove-Metc-Metc} scripts so that binary floppies may be
Xma_e to be installed by the UnixPC user agent. I do not assume
ny
Xresponsibility for these either. I don't assume any responsibility for
Xanything even remotely related to this stuff.
X
X
X Eric H. Herrin II
X University of Kentucky Mathematical Sciences Laboratories
X eric at ms.uky.edu, !cbosgd!ukma!eric
X
X=======================Oro_timfpty README===============================
XT pri(s a pseudo tty driver for system V machines. It works very
Xsimilar to ptys on BSD, for instance emacs works fine. To install this
Xdriver you will need to modify your `master' and `dfile' file which
Xcontains your driver specifications. As these vary from machine to machine,
Xyou will have to look up in your manual how to do that. Here is an example
Xfor a sperry s5050 alias ncr tower 32 :
X
XAdd the following two lines to the driver description section in master:
Xpts 0 237 244 pts 0 0 28 32 0 tty
Xptm 0 37 344 ptm 0 0 29 0 0
X
XThis says there are max 32 pts devices at major number 28 having associated
Xtty structures and 0 ptm devices having major number 29 with no associated
Xdata. The number of ptm devices is not configurable, as this depends on the
Xnumber of pts's.
X
XThe following two lines go in the dfile:
Xpts 0 0 0
Xptm 0 0 0
X
XProbably you will also want to increase the NCLIST parameter.
X
XIf your configuration procedure is different, you must change the shell
Xscript mkpty,EACXch is used to create the device nodes in /dev.
X
XThe ptm devices (/dev/pty[p-z][0-9a-f]) are the controlling ones-Meverything
Xwritten there will show up at the associated pts device
X(/dev/tty[p-z][0-9a-f]), a = 0;
X
ll as erverything which is written on the pts
Xdevice will show up on the ptm device. The pts side will accept the usual
Xtermio ioctl calls. The master side is a bit different, as ioctl callsEACXch
Xnormallytwait for output to drain flush output. The reason for this funny
Xbehaviour is that otherwise the master side will hang. Also the master side
Xmay be opened only once, further open callsEwill result in an EBUSY error.
+ END-OF-FILE doc/Read-pty
chmod 'u=rw,)=r,o=r' 'doc/Read-pty'
echo ' -rw-r- r-max;(es (&so; 7588 Aug 11 20:12 doc/Read-ptyPRU_BPt)'
ec)o ' Dc'
/bin/ls -l doc/Read-pty
echo 'Extracting doc/Read-uipc'
seB 's/^X//' > doc/Read-uipc << '+ END-OF-FILE doc/Read-uipc'
X %Z%%M% %I% (Alex Crain) %G%
X
X This package contains a device driver to implement BSD styledXsockets on an AT&T 3b1. It is written for unix release 3.51 and
Xcompiled using gcc, I don't >sb_mbcnt U if it will work with a different
XOS/compiler combination.
X
X The code is directly derived from Berkeley kernel source, and
Xcarries the Berkeley Copyright. I have added my oUPw Copyright as well,
Xalthough I question its legality and I don't really much care what
Xh;ppens to the code as long as nobody *sells* it (giving away is OK)
Xand my name stays on it.
X
X The code is far from complete. At this writing, only the
XAF_UNIX domain is supported, and SOCK_RAW is not implemented. Within
Xthese restraints, the code implements the following system calls:
X
X socket ()
X cind ()
X listen ()
X accept ()
X connect ()
X eocketpair ()
X sockpipe ()
X sendto ()
X send ()
X recvfrom ()
X recv ()
X setsockopt ()
X getsockopt ()
X
X The code has also not been adaquately tested and surely has
Xbugs. Since the interface routines are true system calls, the bugs
Xare likely to cause panics and other types of system crashes, so use
Xthis driver at your own risk. Bug reports are encouraged, as are
Xenhancements and extensions within the realm of BSD compatability.
XSend bug reports and patches to:
X
X alex%nerwin.uucp at umbc3.umbc.edu
X
X or
X
X uunet!umbc3!nerwin!alex
X
X The documentation is miserable, I will try to do better.
X
X The system files are:
X
X src/* - the c source files
X uipc/* - the header files
X lib/* - the user interface library
X test/* - misc. test programs
X Makefile- the global makefile
X sysconfig.m4 - the system configuration file
X
XTo install the system, type 'make' in the root directory. This will
Xcreate two interesting files: src/uipc.o and lib/libuipc.a. uipc.o
Xgoes in the /etc/lddrv directory. Add the following line to
X/etc/master:
X
Xuipc 1100 000 000 uipc 0 0
X
Xand type:
X
X cd /etc/lddrv; ./lddrv -av uipc
X
Xto load the driver. To make the driver load at boot tome, add "uipc"
Xto the file /etc/lddrv/drivers.
X
XPut libuipc.a in /usr/lib, and copy the uipc directory to
X/usr/include/uipc.
X
X
X
X Again, please be aware that this code is delicate and likely
Xto cause crashes. If appropriate precautions are taken, however, it is
Xpossible to avoid trashing your system. When testing device drivers
Xand code that usesuthem:
X
X 1) make sure that the system is quiet. Open files may be
Xdamaged or lost when the kernel crashes. Do not test the driver when
Xuucico is running.
X
X 2) don't do two things at once. Get out of the editor before
Xtesting new code.
X
X 3) type "sync" before running new code. This will flush the
Xkernels internal buffers. DO libulrun sync immediately after testing
Xnew code, because if the kernel has trashed buffers, they will be
Xwritten to disk. Instead, run "ls" a few times to see if the system is
Xgoing to crash.
X
X 4) Do not install code that uses this package until it becomes
Xmore stable. It might work for aEACXle, but is likely to cause crashes
Xat in oppertuneRese>mes.
X
X The code is heavily protected with callsEto panic {
X uso it is
Xdoubtful that any major damage will result (I have crashed many times
Xwhile writing this and have yet to lose a file) if the precautions are
Xtaken. The code will be better tested and documented soon, so if you
Xdon't w->unpoto beta test, you can wait. In any case,
X
X Enjoy
X
X Alex Crain
+ END-OF-FILE doc/Read-uipc
chmod 'u=rw,)=rw,o=rw' 'doc/Read-uipc'
echo ' -rw-rw-rcursor to{ (struct mbuf *) ( HSJ 3338 Jun 1 F)3 doc/Read-uipc (as sent)'
ec)o ' D|
s -l doc/Read-uip-F-.ho 'Extracting doc/post1'
sed 's/^X//' > doc/post1 << '+ END-OF-FILE doc/post1'
X
X CALL FOR DISCUSSION
X
X Topic: Serial line networking of unix-pc's.
X
X As my efforts to produce a working socket driver for the
Xunix-pc are generating positive results (it works), The looming
Xquestion of how to interconnect machines is becomming more pertenant.ce)ike many people, I bought a large machine and later a small machine
Xas a future parts doner, and getting the two of them to talk would be
Xright nice. I'm running a uucp network now, but I'dchar *PSlly like some
Xkind of packlink case protocol. With the TCP driver taking shape, the
Xidea is becomming more attractive.
X
X For myself, I would like to run a bus network around my house,
Xconsisting of a single serial line. I will probably purchase a mac as
Xa game machine/graphics engine/word processor for my wife in the next
Xyear or so, and I would like to pick up a nice printer, an maybe a
Xtrailblazer. Also, I never w->unpoto buy a peripheral from apple! (nobody
Xhas that much noney)
X
X So I'm figureing my network to be:
X
X HARDWARE
X
X 1) a bi-directional bus network, or some kind of fault-tolerant ring
X that can CS gwhen a host goes down. No star networks. I only want to
X h;
X };to deal with one port per machine, and I w->unpoto be able to splice
X in devices.
X
X 2) RS232 or simular hardware speeds and voltage levels. ($$$)
X
X 3) hardware support for packet acceptance/rejection, and queueing of
X data so that the machines don't die.
X
X SOFTWARE
X
X 1) support for virtual circuits and datagrams, with out-of-band
X facilities.
X
X 2) reasonable efficency with in packlt overhead.
X
X 3) gateway/forwarding capability, for dealing with modems and printers.
X This means machine addresses and ports.
X
X 4) virtual files, as in some kind of restricted NFS.
X
XI /*ek that some kind of minimal hardware support is going to be
Xrequired for speed, but I'd like to make that an option, so that poor
Xfolks can use the existing RS232. I envision a card that watches the
Xnet and screens packets against their address. I would also like at
Xleast 16 bytes of buffer space with a ` , to limit interrupts.
X
XI'dcalso like to use an existing protocol, custom protocols are a drag
Xunless they catch on-Meven free oneo. The two protocols that come to
Xmind are SLIP and Appletalk. SLIP would be in keeping with tradition,
Xbut I'm worried about the overhead. I don't >now anything about
XAppletalk, but the fact that its built around a serial line suggests
Xthat its optimized for speed. The fact that I'm looking at a mac also
Xmakes Appletalk more desireable :-).
X
XAlso, Appletalk hardware is pretty close to what I'm looking for,
Xalthough I don't >sb_mbcnt U much about it. Appletalk also supports some kind
Xof disk sharing that works over a serial line.
X
XI will probably add the code for SLIP anyway, since much of the code
Xis free and available. I'm looking into a protocol developed at CMU
Xthats built around Appletalk, and if the code is free, I'll look into
Xadding that to, if only for my future mac.
X
XSo what does everybody think? I w-nt to w);
X brea something thats going to
Xbe *used*arg[0]fthough I won't be charging for it. What are your needs? I
Xdon't know diddly about hardware, so I would like to here from the
Xh;rdware guys regarding the interface. Optimallytwe would see the same
Xdeal as the hard disk upgrade, namely a "lenny and gil do it yourself"
Xmodel and a "custom PAL on a special board for $$" version.
X
X
X
X
X
+ END-OF-FILE doc/post1ddr_t))=rw,f=rw' 'doc/post1file; ffm 1 bes ( w HSJ 3350 Jul 18 23:43 doc/post1PRU_BPt)'
ec)o ' Dc'
/bin/ls -l doc/post1
if test -f 'support'
then rm 'support'
fi
if test -d 'support'
then :
else echo 'Making support/'
mkdir 'support'
fi
cf thfu=rwx,g=rx,o=rx' 'support'
echo 'Extracting support/Makefile'
sed 's/^X//' > support/Makefile << '+ END-OF-FILE support/Makefile'
X# CC=gcc
XCFLAGS=-O $(VFLAG)
X
Xsocket_init:
X $(CC) $(CFLAGS) -o socket_init socket_init.c ../lib/libuipc.a
X
X
Xclean:
X rm -f socket_init core *.o
+ END-OF-FILE support/Makefileddr_t))=r,
=r' 'support/Makefile'
echo ' -rw-r- r-max;(es ( HSJ 142ase 'M'g 12 09:24 support/Makefile (as sent)'
ec)o ' Dc'
/bin/
s -l support/Makefile
echo 'Extracting support/socket_init.c'
sen 's/^X//' > support/socket_init.c <<n'+ END-OF-FILE support/socket_init.c'
X#include) m-(tdio.h>
X#include)<sys/types.h>
X#include) m-(ys/sysmacros.m4,)<sys/stat.m4,)<a.out.h>
X
X#define WINTTY 0
X
X/* struct nlist */
Xstruct nlist nl[] = {
X {"wintty"},
X {0}
X};
X
Xmain )
X{
X int ret, i;
X unsigned u;
X struct stat st;
X
X ret = nlist("/etc/lddrv/wind", nl);
X if(ret < 0) {
X perror("nlist on /etc/lddrv/wind");
X exit(1);
X }
X ret =ustat("/dev/window", &st);
X if(ret < 0) {
X perror("stat on /dev/window failed!");
X exit(1);
X }
X i = major(st.st_rdev);
X u = (unsigned int) nl[0].n_value;
X sosetup(i,u);
X}
+ caddr_t cILE support/socket_init.
X }
X fmod _name\,=r,
=r' 'support/socket_init.c'
efho ' -rw-r- (&so-m syn( w HSc 530ase 'M'g 1g 20:05 support/socket_init.c (as sent)'
echo ' \c'
/bin/ls -l support/socket_init.c
echo 'Extracting sysconfig.m4'
sed 's/^X//' > sysconfig.m4 << '+ END-OF-FILE sysconfig.m4'
Xifelse(REALTIME,1,`divert(-1)')
X#
X# @(#)sysconfig.m4 1.1 (Alex Crain) 6/20/89
X#
X# System call definition file
X#
X# each line should be a DEFSYSCALL macro, where the
X# arguments are the number of arguments that the
X# syscall expects and the syscall name.
X#
X# the syscall number for syslocal()
X#
Xdefine(SYSL_SYSCALL, 67)
X#
X# the offset of the last >sb_mbcnt Un local system call.
X#
Xdefine(SYSENT_OFFSET,18)
X#
X# the offset of the first new system call in the sysentries[] array
X#
Xdefine(SYSENT_START,incr(SYSENT_OFFSET))
Xifelse(REALTIME,1,`divert')
X
XDEFSYSCALL(3,socket)
XDEFSYSCALL(3,bind)
XDEFSYSCALL(2,listen)
XDEFSYSCALL(3,accept)
XDEFSYSCALL(3,connect)
XDEFSYSCALL(4,socketpair)
XDEFSYSCALL(1,sockpipe)
XDEFSYSCALL(6,sendto)
XDEFSYSCALL(4,send)
XDEFSYSCALL(6 /* ecvfrom)
XDEFSYSCALL(4,recv)
XDEFSYSCALL(5,setsockopt)
XDEFSYSCALL(5,getsockopt)
XDEFSYSCALL(4,soselect)
XDEFSYSCALL(2,sosetup)
+ EcT-OF-FILE sysconfig.m4ddr_t))=r,
=r' 'sysconfig.m4'
echo ' -rw-r- r-m 1 bes ( w HScomplx 868 Aug 11 22:24 sysconfig.m4 (as sent)'
ec)o ' Dc'
/bin/ls -l sysconfig.m4
exit 0
--
Bradley nmith
Computer Systems Offer Integration Laboratory
AT&T Bell Labs, Holmdel, NJ
201-949-0090 att!holin!bes or bes at holin.ATT.COM
#! rnews
More information about the Unix-pc.sources
mailing list