Andrey Chernov [Thu, 16 Jun 1994 11:13:44 +0000 (11:13 +0000)]
Choose secure single-user mode by default
Without this entry init.bsdi don't ask root password when it goes
to sigle-user. This entry must present here in any case,
subject of arguing can be only default mode, I mean
"secure" or "insecure" here. Please consider this entry
like template and change "insecure" to "secure" if you
are _shure_, but not back out whole line.
# This entry needed for asking password when init goes to single-user mode
console none unknown off insecure
David Greenman [Thu, 16 Jun 1994 10:09:47 +0000 (10:09 +0000)]
Correct minor foratting anomoly caused by last commit.
David Greenman [Thu, 16 Jun 1994 08:53:55 +0000 (08:53 +0000)]
Ran this through indent. John is going to hate me.
David Greenman [Thu, 16 Jun 1994 08:27:21 +0000 (08:27 +0000)]
Ran this puppy through indent with my own home brew set of options. The
code was really needing it after being tugged on a bunch over the past
three months.
Andrey Chernov [Thu, 16 Jun 1994 08:08:44 +0000 (08:08 +0000)]
Increase loopback delay for successfull probe of Boca
internal modem M2400I
Geoff Rehmet [Thu, 16 Jun 1994 08:01:19 +0000 (08:01 +0000)]
chpass was writing incorrect entries to /etc/passwd.
(This was caused by the fast update code being incorrect).
Cops reports:
Warning! Passwd file, line 48, does not have 7 fields:
g91v3505:*:107:123::0:0:Stuart van Onselen:/home/g91v3505:/usr/local/bin/tcsh
(The line above is in master.passwd format!)
To fix this: we provide 2 interfaces to pw_copy() , namely
pw_copy() (which behaves as before) and pw_copy_insecure(),
which does the stuff needed to update /etc/passwd. The original
pw_copy() routine now becomes a back end for these 2 routines.
(not the most elegant fix, but it works for 1.1.5 !!! grrrr)
Jordan K. Hubbard [Thu, 16 Jun 1994 05:32:10 +0000 (05:32 +0000)]
This is the IBM/National PCMCIA ethernet driver from Keith Moore,
based originally on work by David Greenman and adapted to FreeBSD
(and cleaned up a bit) by myself. It supports the IBM Credit Card
Adapter for the IBM Thinkpad, and I've had no trouble making it work
on my Toshiba T1910 with a National `InfoMover' NE4100 PCMCIA ethernet
card (I'm commiting this message through it right now :-).
This is actually sneaking it in after feature-freeze, but it's just
too useful to pass up! As always, necessity is a mother.
Adam David [Thu, 16 Jun 1994 03:53:29 +0000 (03:53 +0000)]
Changed delay mechanism to rely more on I/O spinning. If spinning on the
like this is bad news, it will have to be revised.
Shortened some verbose messages for when the kernel is loaded below 640k.
Updated version number.
David Greenman [Thu, 16 Jun 1994 02:55:28 +0000 (02:55 +0000)]
Disable pagescan daemon for now. It's an interesting idea, but consumes
lots of CPU time on slow machines. The downside of disabling this is
that compiles on 4MB machines will slow down by about 10% on fast CPU
machines.
Paul Traina [Thu, 16 Jun 1994 01:53:24 +0000 (01:53 +0000)]
Make loses argument information.
Fix courtesy of CMU's release notes for building AUIS (Andrew).
Adam David [Thu, 16 Jun 1994 00:45:02 +0000 (00:45 +0000)]
fix incorrect sense of defined(DEBUG).
only do '(skipping) strip' if NOSTRIP defined.
Normal DEBUG behaviour as per Rich and Rod
Jordan K. Hubbard [Wed, 15 Jun 1994 23:28:07 +0000 (23:28 +0000)]
Flags 0x80 now means enable the diags, not disable.
Jordan K. Hubbard [Wed, 15 Jun 1994 23:26:42 +0000 (23:26 +0000)]
The whole COM_QUIET thing is just too gross - our kernel is already way
too verbose as it is! I've reversed it and made it COM_VERBOSE, which
people can enable if _they really want to_. I'll change the doc as well
Jordan K. Hubbard [Wed, 15 Jun 1994 22:58:49 +0000 (22:58 +0000)]
Since we're using syscons now, make ^H the backspace character by
default (ack ack, evil evil, but easier than changing syscons at this
late hour).
Paul Richards [Wed, 15 Jun 1994 22:41:19 +0000 (22:41 +0000)]
Changes from Paul Kranenburg which bring us into sync with his sources:
handling of errors through the standard err() and warn()
more fixes for Geoff Rehmet's NULL pointer bug.
fixes NULL pointer bugs when linking mono and nested X servers.
supports a `-nostdlib' option.
accept object files without a symbol table
don't attempt dynamic linking when `-A' is given
a few variable names have chaged (desc -> fd), and the formatting has
changed which should make it much easier to track his sources.
I tested 'make world' for /usr/src and X twice with these changes.
Adam David [Wed, 15 Jun 1994 21:32:38 +0000 (21:32 +0000)]
Changed MDEC_OPT=BOOTWAIT= to BOOTWAIT=
Adam David [Wed, 15 Jun 1994 21:30:28 +0000 (21:30 +0000)]
Back out MDEC_OPT as it is totally unnecessary. I must have got confused over
the subtle differences between -DFOO=0 and FOO=0.
Is there a better way of reverting to a previous version if noone else has
touched the file? At the moment the version number has been incremented by
two, and the file not changed.
Jordan K. Hubbard [Wed, 15 Jun 1994 21:21:50 +0000 (21:21 +0000)]
1. Fill out the crunch target - this should do the trick (we'll see).
2. Remove the NOOBJ line - we want an object dir now.
Jordan K. Hubbard [Wed, 15 Jun 1994 21:19:59 +0000 (21:19 +0000)]
Whoops, missed the .cache files - those can be tossed too.
Jordan K. Hubbard [Wed, 15 Jun 1994 21:19:23 +0000 (21:19 +0000)]
Nuke all the intermediate files.
Jordan K. Hubbard [Wed, 15 Jun 1994 21:08:47 +0000 (21:08 +0000)]
Add crunch to list of things to build.
Jordan K. Hubbard [Wed, 15 Jun 1994 21:04:04 +0000 (21:04 +0000)]
Add a beep to the load_fd stuff for each floppy insert request so
that you can read a book and load the floppy distribution at the same
time.. :-)
Jordan K. Hubbard [Wed, 15 Jun 1994 21:02:36 +0000 (21:02 +0000)]
Remove the --exclude libc.so.1.1 line; with a static gzip and tar we
no longer need it, and it's just one more thing we would have to remember
to change with every shared lib version bump.
Jordan K. Hubbard [Wed, 15 Jun 1994 20:43:20 +0000 (20:43 +0000)]
Install manpath.config and mail.rc properly, and directly from the src
tree now (rather than having a local copy.
Adam David [Wed, 15 Jun 1994 20:24:16 +0000 (20:24 +0000)]
added comments for how to use CCFPU and MDEC_OPT
Adam David [Wed, 15 Jun 1994 20:21:00 +0000 (20:21 +0000)]
added MDEC_OPT for passing options to bootblock compilation.
It would be nice if someone who understands sys/Makefile to add similar
functionality there too :-)
Jordan K. Hubbard [Wed, 15 Jun 1994 20:15:12 +0000 (20:15 +0000)]
Commit Rod's magic EXTRACT scripts to the tree so that we can work on
them. These were al hand-copied over before.
Jordan K. Hubbard [Wed, 15 Jun 1994 20:14:29 +0000 (20:14 +0000)]
Lots of changes:
1. Properly use ${.CURDIR} now instead of hardcoded relative dirs.
2. Use ${BINOWN} and ${BINGRP} everywhere instead of root/wheel
3. Add target for copying over EXTRACT scripts (and add them here).
4. Start thinking about crunched floppy target (not in yet, next commit).
Jordan K. Hubbard [Wed, 15 Jun 1994 19:23:33 +0000 (19:23 +0000)]
1. Bump minimum root size to 8MB (we need that much now).
2. Bump version number to 1.1.5 for impending release.
Jordan K. Hubbard [Wed, 15 Jun 1994 19:16:58 +0000 (19:16 +0000)]
Make gzip static by default. This is needed because it's such a
frequent companion to tar.
Jordan K. Hubbard [Wed, 15 Jun 1994 19:16:04 +0000 (19:16 +0000)]
Make tar static by default. Don't scream, you'll thank me for this
someday.
Jordan K. Hubbard [Wed, 15 Jun 1994 19:09:14 +0000 (19:09 +0000)]
From Hellmuth Michaelis, reviewed by Bruce Evans: This enables
the keyboard clock rather than incorrectly disabling it.
Adam David [Wed, 15 Jun 1994 18:15:17 +0000 (18:15 +0000)]
make BOOTWAIT loopcount parameter available for tuning.
This changes nothing unless f.e. -DBOOTWAIT=0 or -DBOOTWAIT=640000 is used
Paul Richards [Wed, 15 Jun 1994 14:35:30 +0000 (14:35 +0000)]
CHanged MAXMEM to ISMAXMEM for safety
Jordan K. Hubbard [Wed, 15 Jun 1994 11:28:38 +0000 (11:28 +0000)]
Make DESTDIR work properly.
Jordan K. Hubbard [Wed, 15 Jun 1994 10:49:34 +0000 (10:49 +0000)]
1. Clean the intermediate files when building the default fixit floppy
(demonstration) target.
2. Add the kcopy and filesystem floppy configuration files from Rod.
Jordan K. Hubbard [Wed, 15 Jun 1994 10:33:49 +0000 (10:33 +0000)]
This commit was generated by cvs2svn to track changes on a CVS vendor
branch.
Jordan K. Hubbard [Wed, 15 Jun 1994 10:33:49 +0000 (10:33 +0000)]
This is the new crunch utility for making distribution floppies from
James da Silva. We need to integrate this before 1.1.5 so that we can
actually make the boot floppies.
Andrey Chernov [Wed, 15 Jun 1994 10:14:41 +0000 (10:14 +0000)]
Back out nostdinc change (discussion still in progress)
Adam David [Wed, 15 Jun 1994 08:24:25 +0000 (08:24 +0000)]
correct location of corefile data segment
Adam David [Wed, 15 Jun 1994 05:40:20 +0000 (05:40 +0000)]
stty `stty -g` was broken, corrected typos vmin --> min, vtime --> time
Jordan K. Hubbard [Wed, 15 Jun 1994 04:19:23 +0000 (04:19 +0000)]
Bring back bounce buffer support code from revision 1.14
Adam David [Wed, 15 Jun 1994 00:55:57 +0000 (00:55 +0000)]
(verbosely) skip the strip -x in the case of config -g
Jordan K. Hubbard [Tue, 14 Jun 1994 18:22:54 +0000 (18:22 +0000)]
Revert to previous version of this driver (1.19) until we figure
how how to make the 1.20 version work properly. I did not attempt to
bring any changes from the 1.20 version back in, so if there are any bugs
that reappear as a result of this, plesae let me know ASAP! Most people
had good results with this version, so I'm not too worried.
David Greenman [Tue, 14 Jun 1994 17:49:42 +0000 (17:49 +0000)]
slight improvement to last commit.
David Greenman [Tue, 14 Jun 1994 17:31:52 +0000 (17:31 +0000)]
Suggested by Bruce Evans - initialize eflags.
Jordan K. Hubbard [Tue, 14 Jun 1994 15:28:17 +0000 (15:28 +0000)]
Remove SYSCONS - no longer needed since syscons is the default in the
GENERIC* kernels, and LINT gives all the other relevant syscons options.
Add a new kernel configuration file suitable for building laptop kernels.
It's small and supports things like PS/2 mice out of the box.
Geoff Rehmet [Tue, 14 Jun 1994 12:45:41 +0000 (12:45 +0000)]
Bug reported by Jon-Dean Mountjoy <csjm@cs.ru.ac.za>
Paul Kranenburg's description:
ld is in error here, assuming that symbols with N_EXT set always have an
entry in the (global) symbol table: this is not the case for C++ generated
constructor/destructor symbols. I can reproduce your failure by fudging
a "multiply defined" constructor symbol by hand. Checking for `g == NULL'
seems to be a ok as a fence for now.
So:
for now, in do_file_warnings() we check if g == NULL, before trying to generate
any warning messages. This prevents a NULL pointer dereference.
Andrey Chernov [Tue, 14 Jun 1994 11:46:52 +0000 (11:46 +0000)]
Remove CXXFLAGS from here, handled in another place
Andrey Chernov [Tue, 14 Jun 1994 11:45:04 +0000 (11:45 +0000)]
Add -nostdinc, when DESTDIR
Andrey Chernov [Tue, 14 Jun 1994 11:05:19 +0000 (11:05 +0000)]
DPADD added
Andrey Chernov [Tue, 14 Jun 1994 11:02:17 +0000 (11:02 +0000)]
Correct name: libterm.a --> libtermcap.a
Andrey Chernov [Tue, 14 Jun 1994 11:00:35 +0000 (11:00 +0000)]
Add LIBREADLINE and LIBTERM to DPADD
Andrey Chernov [Tue, 14 Jun 1994 10:47:57 +0000 (10:47 +0000)]
Back out removing 'Escape ...' if VDISABLE,
breaks tradition
Geoff Rehmet [Tue, 14 Jun 1994 08:44:02 +0000 (08:44 +0000)]
Change /usr/lib/libskey.a in DPADD to ${LIBSKEY}
needed for compile with DESTDIR set to work.
Geoff Rehmet [Tue, 14 Jun 1994 08:41:58 +0000 (08:41 +0000)]
Add entries for LIBSKEY
Rod Grimes [Tue, 14 Jun 1994 07:31:42 +0000 (07:31 +0000)]
Touch the $Revision$ string to cause a cvs revision update of the boot
code (ie, should now print Revision 1.13).
David Greenman [Tue, 14 Jun 1994 03:41:10 +0000 (03:41 +0000)]
Fixes from John Dyson to fix a cache coherency problem with the VM
object cache when using NFS.
Rod Grimes [Tue, 14 Jun 1994 02:57:09 +0000 (02:57 +0000)]
Clean up small error in the ncvs code that was causing rcs co to be
called with -K-Keoptions -Kioptions. This should fix the problem with
$Id$ still getting changed.
I am also install both ncvs and ocvs on freefall with this change as
ocvs still had the bug with -I \! which I fixed but did not reinstall.
Garrett Wollman [Tue, 14 Jun 1994 02:05:38 +0000 (02:05 +0000)]
Fix flag punning bug which would cause IFF_ALTPHYS interfaces to generate
zero-entry RIP packets. Reported by Mark Taylor.
Garrett Wollman [Mon, 13 Jun 1994 22:41:04 +0000 (22:41 +0000)]
Added comment about registerd and kpasswdd not working in 1.x.
Deleted commented-out line which would start mountd; that's not
the right pplace to do it (don't confuse the users).
Should probablyhave uncommented rpc.rstatd, but didn't.
Andrey Chernov [Mon, 13 Jun 1994 22:03:57 +0000 (22:03 +0000)]
Add LIBREADLINE?= ...
Andrey Chernov [Mon, 13 Jun 1994 21:50:02 +0000 (21:50 +0000)]
Remove CXXFLAGS, unneeded here now
Andrey Chernov [Mon, 13 Jun 1994 21:40:22 +0000 (21:40 +0000)]
Print out "Escape character ..." only if it is not
equal _POSIX_VDISABLE
Geoff Rehmet [Mon, 13 Jun 1994 21:33:50 +0000 (21:33 +0000)]
Remove a conflicting prototype for malloc(),
which was causing the compile to fail when DESTDIR is set.
Geoff Rehmet [Mon, 13 Jun 1994 21:02:48 +0000 (21:02 +0000)]
Hack to allow libg++ to compile properly when DESTDIR is defined.
This involves adding -I${DESTDIR}/usr/include/${CXX} to the
compiler flags when C++ code is being compiled.
This needs to be revisited for 2.0.
Jordan K. Hubbard [Mon, 13 Jun 1994 20:19:35 +0000 (20:19 +0000)]
ISOFS fixed from Bruce Evans:
1. Make month array static so that it's not allocated all the time.
2. Correct application of timezone offset.
Adam David [Mon, 13 Jun 1994 19:39:43 +0000 (19:39 +0000)]
Make CCFPU available as a toplevel define to allow generation of FPU codes
Jordan K. Hubbard [Mon, 13 Jun 1994 19:27:52 +0000 (19:27 +0000)]
Be paranoid about setting of %dl register for braindead BIOS's,
initializing it to 0 in the floppy boot case. This allows my Toshiba
1910 laptop to finally boot, and may help out other machines cursed
with early Phoenix BIOS's as well.
Adam David [Mon, 13 Jun 1994 18:52:57 +0000 (18:52 +0000)]
Get groff to compile by adding to CXXFLAGS in the same way as CFLAGS.
There are probably better ways of getting it to work, but this does it.
Geoff Rehmet [Mon, 13 Jun 1994 18:49:29 +0000 (18:49 +0000)]
>From: stark!gene@newsserv.cs.sunysb.edu (Gene Stark)
The subject says it. The Section 1 man page for login is not getting
installed properly.
(This is probably due to the addition of login.access.5)
Add: MAN1= login.1
DPADD still needs fixing .
Andrey Chernov [Mon, 13 Jun 1994 08:03:32 +0000 (08:03 +0000)]
Remove PAGE_SIZE redifinition (from param.h)
Gary Clark II [Mon, 13 Jun 1994 05:38:52 +0000 (05:38 +0000)]
Changed entry for Seagate/FD driver to show as being in current.
Sean Vickery [Mon, 13 Jun 1994 03:31:30 +0000 (03:31 +0000)]
Admit being responsible for the Seagate ST01/02 & FD 950 SCSI driver
by Kent Palmkvist <kentp@isy.liu.se>
Sean Vickery [Mon, 13 Jun 1994 02:49:12 +0000 (02:49 +0000)]
Got Kent Palmkvist's ST01/02, FD 950 driver working for FreeBSD.
Things done:
- Added $Id$ lines to README.seagate and seagate.c
- Fixed typos in seagate.c isa_driver structure initialization
- Made entries in LINT and files.i386
Martin Renters [Mon, 13 Jun 1994 02:43:58 +0000 (02:43 +0000)]
Deleted wd80x3.c and added ether.c in its place. ether.c includes support
for NE1000/2000 and better WD/SMC support.
Sean Vickery [Mon, 13 Jun 1994 01:59:25 +0000 (01:59 +0000)]
Seagate ST01/02 & Future Domain 950 SCSI adaptor driver source and README.
From Kent Palmkvist <kentp@isy.liu.se>
Sean Vickery [Mon, 13 Jun 1994 01:59:25 +0000 (01:59 +0000)]
This commit was generated by cvs2svn to track changes on a CVS vendor
branch.
Geoff Rehmet [Sun, 12 Jun 1994 18:48:13 +0000 (18:48 +0000)]
>From: Ryutaroh Matsumoto <ryutaroh@fujiyama.iijnet.or.jp>
Description:
We lose one available file descriptor after calling gethostbyname().
Because init_services() in gethostnamadr.c does not close file.
- Close the file after we finish reading from it. (The patch submitted
was not quite correct.)
Geoff Rehmet [Sun, 12 Jun 1994 16:20:36 +0000 (16:20 +0000)]
Change default for directory permissions to be 0755.
Gary Clark II [Sun, 12 Jun 1994 15:13:31 +0000 (15:13 +0000)]
Added entry for Paul's comments on kernel debugging.
Gary Clark II [Sun, 12 Jun 1994 15:12:21 +0000 (15:12 +0000)]
Added Paul's comments on kernel dubugging of crash dumps.
Geoff Rehmet [Sun, 12 Jun 1994 11:05:04 +0000 (11:05 +0000)]
Modify messages about the resolver manual page as per Rod's suggestions.
The configure function now tells the user to type "man 5 resolver"
for more info on resolv.conf, but mentions that the bindist must
be fully installed before this can be done (actually a user won't
have reached this stage if he doesn't have an installed bindist ;-)
Andrey Chernov [Sun, 12 Jun 1994 10:51:01 +0000 (10:51 +0000)]
Add (#ifdef'ed by STARTUP_LOCALE) following line to crt0.c
(void) setlocale(LC_ALL, "");
It will be easiest way now to make national chars available
for all ctype-oriented programs at once by simple
setenv LANG Your_National_Charset
(currently only Russian.koi8-r available)
Default case (without "LANG" environment
variable) will be fully ANSI compatible (got "C" locale).
If "LANG" variable present, extention becomes active.
Effect of this extention is great: in one time all ctype
oriented programs can accept/print national characters
without any touching source/binary code, it is big win, IMHO.
This method is fully compatible with ISO8859-* and russian koi8-r
too (in general -- with all 8-bit character sets). I think
it is very useful.
I got this idea from Xenix locale implementation.
This extention is even never compiled in, unless you set
setenv STARTUP_LOCALE
before rebuilding crt0.c
Andrey Chernov [Sun, 12 Jun 1994 08:51:09 +0000 (08:51 +0000)]
Change includes in proper way, i.e.
"readline.h" --> <readline/readline.h>
Remove -I./readline from Makefile (again!)
David Greenman [Sun, 12 Jun 1994 04:13:21 +0000 (04:13 +0000)]
From John Dyson:
Fixed bug that could result in a page fault loop on page table pages.
Basically, bias the activation count a little and properly free the page
if it isn't held.
David Greenman [Sun, 12 Jun 1994 04:05:53 +0000 (04:05 +0000)]
Changed the policy of setting B_AGE to work properly for our relatively
large buffer caches. The old policy assumed that only read-ahead blocks
would be cached. This commit actually doesn't change much - since we
had already #if 0'd out most of the B_AGE set cases long ago.
Karl Lehenbauer [Sat, 11 Jun 1994 23:33:53 +0000 (23:33 +0000)]
Added an access check to the NFS server's create routine. Without it, clients
can zero out any file on the exported filesystem, just by creat(2)'ing it.
Problem noticed using NFS on SunOS4.1.1. This is fixed in BSD 4.4, and the fix
is the same.
Paul Richards [Sat, 11 Jun 1994 23:06:35 +0000 (23:06 +0000)]
Removed the following lines from m_compress:
if (in->m_flags & M_EXT) {
#ifdef DEBUG
ASSERT(in->m_len == 0);
#endif
}
since they do nothing unless DEBUG is set and if it is set the compile
falls over because ASSERT is only defined when the netiso code is
included. A more correct fix would probably be to wrap m_compress
with a ifdef NISO but m_compress is moved to the netiso directory in
4.4 anyway so simpler to just yank this bit of DEBUG code for 1.1.5
Gary Clark II [Sat, 11 Jun 1994 18:00:34 +0000 (18:00 +0000)]
Added warning about ST-?? driver.
Gary Clark II [Sat, 11 Jun 1994 17:59:19 +0000 (17:59 +0000)]
This is the systems FAQ. I'll be adding more entrys shortly.
Sorry for the delay.
Gary Clark II [Sat, 11 Jun 1994 17:58:17 +0000 (17:58 +0000)]
Changed entry for systems.FAQ (I have it now). and added mention of
ST-01/02 driver.
Paul Richards [Sat, 11 Jun 1994 17:09:39 +0000 (17:09 +0000)]
Two changes to DEBUG code:
Moved pads function before pmap_pvdump to avoid compiler warning
and ifdefed out the use of pv_flags as follows:
#ifdef used_to_be
printf(" -> pmap %x, va %x, flags %x",
pv->pv_pmap, pv->pv_va, pv->pv_flags);
#endif
printf(" -> pmap %x, va %x",
pv->pv_pmap, pv->pv_va);
Paul Richards [Sat, 11 Jun 1994 16:20:42 +0000 (16:20 +0000)]
Moved -DTRAD_CORE from gdb/Makefile to bfd/Makefile
Paul Richards [Sat, 11 Jun 1994 16:20:07 +0000 (16:20 +0000)]
Modev -DTRAD_CORE for gdb/Makefile to bfd/Makefile
Paul Richards [Sat, 11 Jun 1994 15:18:01 +0000 (15:18 +0000)]
Put back -I${.CURDIR}/readline in Makefile.
The sources are just broken and given that an extra -I flag is going
to be needed it may as well be as above to guarantee that the library
will build.
Geoff Rehmet [Sat, 11 Jun 1994 12:42:35 +0000 (12:42 +0000)]
Allow configuration of resolv.conf though "configure". The
configuration which is done is minimal, and the script refers the user
to the resolver(5) manual page for information on fine-tuning name
resolution.
David Greenman [Sat, 11 Jun 1994 07:58:08 +0000 (07:58 +0000)]
Added important missing functionality that is required in the
vnode_pager_setsize() routine. This additional code fixes the
"tail with NULLs bug".
David Greenman [Sat, 11 Jun 1994 02:36:32 +0000 (02:36 +0000)]
Fix from Bruce Evans:
Set npx_exists = 0 in the case of broken error reporting.
David Greenman [Sat, 11 Jun 1994 02:30:05 +0000 (02:30 +0000)]
Bruce found a bug in my changes to stop using the gs selector.
From Bruce Evans:
fu[i]byte() checked the wrong register. This caused interesting behaviour
in the GPL math emulator. The emulator does not check the values returned
by fu*() or su*() (:-() and it interpreted the address of -12(%ebp) as
-1(%ebp). The same probably occurs for all signed 8-bit offsets from
registers.
I cleaned up the new bzero() a bit.