]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
30 years agoConvert to our man installation style. Also fixed long-standing bug
wollman [Fri, 5 Aug 1994 02:42:42 +0000 (02:42 +0000)]
Convert to our man installation style.  Also fixed long-standing bug
in `fastboot'/`fasthalt' in which the interpreter would hang around
after `reboot' or `halt' is run, causing an irritating ``Killed'' message.

30 years agoAllow libtelnet to compile, by removing references to Kerberos and DES in
wollman [Fri, 5 Aug 1994 02:02:36 +0000 (02:02 +0000)]
Allow libtelnet to compile, by removing references to Kerberos and DES in
the Makefile.  We still need to determine the appropriate source organization
for this.

30 years agoMake up for the fact that rpc headers are now installed from include.
wollman [Fri, 5 Aug 1994 01:59:40 +0000 (01:59 +0000)]
Make up for the fact that rpc headers are now installed from include.

30 years agoTemporarily force creation of librpc.so.2.0 (by including lib-directory
wollman [Fri, 5 Aug 1994 01:58:43 +0000 (01:58 +0000)]
Temporarily force creation of librpc.so.2.0 (by including lib-directory
Makefile.inc).  Eventually this will be moved back into libc as in 1.1.5.

30 years agoMake it work with our make macros.
wollman [Fri, 5 Aug 1994 01:52:05 +0000 (01:52 +0000)]
Make it work with our make macros.

30 years agoMake it work with our make macros.
wollman [Fri, 5 Aug 1994 01:42:04 +0000 (01:42 +0000)]
Make it work with our make macros.

30 years agoFirst crack at making libc work with the new make macros. It compiles on
wollman [Fri, 5 Aug 1994 01:19:12 +0000 (01:19 +0000)]
First crack at making libc work with the new make macros.  It compiles on
my machine, and a simple static (genassym) and shared (sysctl) executable
both work.  Still to be done: RPCand YP merge.

30 years agoFrom 1.1.5: define default shared library versions. For 2.0, use version
wollman [Thu, 4 Aug 1994 22:44:00 +0000 (22:44 +0000)]
From 1.1.5: define default shared library versions.  For 2.0, use version
2.0.

30 years agoFor the moment, don't build kpasswdd until we figure out what
wollman [Thu, 4 Aug 1994 21:54:38 +0000 (21:54 +0000)]
For the moment, don't build kpasswdd until we figure out what
to do with it.

30 years agoMake obj directory work right. md?hl.c should probably be
wollman [Thu, 4 Aug 1994 21:20:03 +0000 (21:20 +0000)]
Make obj directory work right.  md?hl.c should probably be
deleted, but I'll leave that up to PHK.

30 years agoAdd a couple of files we created in 1.x.
wollman [Thu, 4 Aug 1994 21:10:08 +0000 (21:10 +0000)]
Add a couple of files we created in 1.x.

30 years agoSpplat our 1.1.5 `mk' ifiles over the top of the 4.4 ones. So far
wollman [Thu, 4 Aug 1994 21:09:27 +0000 (21:09 +0000)]
Spplat our 1.1.5 `mk' ifiles over the top of the 4.4 ones.  So far
as I can tell, this is ts the right thing to do.

30 years agoUse __GNU_P not __P.
wollman [Thu, 4 Aug 1994 20:57:07 +0000 (20:57 +0000)]
Use __GNU_P not __P.

30 years agoUse __GNU_P ratherthan __P.
wollman [Thu, 4 Aug 1994 20:53:14 +0000 (20:53 +0000)]
Use __GNU_P ratherthan __P.

30 years agohistedit.h is now installed from here, not from libedit.
wollman [Thu, 4 Aug 1994 20:49:28 +0000 (20:49 +0000)]
histedit.h is now installed from here, not from libedit.

30 years agohistedit.h is now installed from include.
wollman [Thu, 4 Aug 1994 20:47:46 +0000 (20:47 +0000)]
histedit.h is now installed from include.

30 years agoInstall RPC headers from include, like they always should have been.
wollman [Thu, 4 Aug 1994 20:39:34 +0000 (20:39 +0000)]
Install RPC headers from include, like they always should have been.

30 years agoAdded assembly versions of ffs() and bcmp().
dg [Thu, 4 Aug 1994 19:51:01 +0000 (19:51 +0000)]
Added assembly versions of ffs() and bcmp().

30 years agoInlined insque and remque.
dg [Thu, 4 Aug 1994 19:46:57 +0000 (19:46 +0000)]
Inlined insque and remque.

30 years agoFixed bug that would cause free memory reserves to be depleted and cause a
dg [Thu, 4 Aug 1994 19:43:13 +0000 (19:43 +0000)]
Fixed bug that would cause free memory reserves to be depleted and cause a
panic in some cases.
Submitted by: John Dyson

30 years agoAdded some code that was accidently left out early in the 1.x -> 2.0 VM
dg [Thu, 4 Aug 1994 19:40:47 +0000 (19:40 +0000)]
Added some code that was accidently left out early in the 1.x -> 2.0 VM
system conversion.
Submitted by: John Dyson

30 years agoMove ieeefp.h over, and put it in the correct subdirectory this time.
wollman [Thu, 4 Aug 1994 19:16:37 +0000 (19:16 +0000)]
Move ieeefp.h over, and put it in the correct subdirectory this time.

Submitted by: Andrew Moore

30 years agoAdded glueo build rpcsvc stuff, and install floatingpoint.h.
wollman [Thu, 4 Aug 1994 19:04:38 +0000 (19:04 +0000)]
Added glueo build rpcsvc stuff, and install floatingpoint.h.

30 years agoMoved 1.1.5 RPC service files from 1.1.5. Tese are the correct ones;
wollman [Thu, 4 Aug 1994 19:01:57 +0000 (19:01 +0000)]
Moved 1.1.5 RPC service files from 1.1.5.  Tese are the correct ones;
the ones in /usr/src/lib/librpc/rpcsvc are somewhat bogus and will
be deleted.

Submitted by: Original work in 1.1 by J.T. Conklin.

30 years agoUpdated to include improvements from FreeBSD 1.1.5. Fixed brokeness
dg [Thu, 4 Aug 1994 17:42:35 +0000 (17:42 +0000)]
Updated to include improvements from FreeBSD 1.1.5. Fixed brokeness
with multicast support and BPF.

30 years agoNuke redefinition of insque and remque.
dg [Thu, 4 Aug 1994 06:15:10 +0000 (06:15 +0000)]
Nuke redefinition of insque and remque.

30 years agoNuked #if 0'd _insque and _remque routines - they are now inlined in
dg [Thu, 4 Aug 1994 06:10:27 +0000 (06:10 +0000)]
Nuked #if 0'd _insque and _remque routines - they are now inlined in
cpufunc.h.

30 years ago Made NFS attribute cache timeouts kernel config file tunable via
dg [Thu, 4 Aug 1994 06:03:46 +0000 (06:03 +0000)]
   Made NFS attribute cache timeouts kernel config file tunable via
NFS_MINATTRTIMO and NFS_MAXATTRTIMO.

30 years agoIntegrated VM system improvements/fixes from FreeBSD-1.1.5.
dg [Thu, 4 Aug 1994 03:06:48 +0000 (03:06 +0000)]
Integrated VM system improvements/fixes from FreeBSD-1.1.5.

30 years agoChanged occurrances of "itrunc" to "ffs_truncate" to make Bruce happy.
dg [Wed, 3 Aug 1994 08:19:35 +0000 (08:19 +0000)]
Changed occurrances of "itrunc" to "ffs_truncate" to make Bruce happy.

30 years ago Merged in post-1.1.5 work done by myself and John Dyson. This includes:
dg [Wed, 3 Aug 1994 02:45:30 +0000 (02:45 +0000)]
   Merged in post-1.1.5 work done by myself and John Dyson. This includes:

me:
1) TLB flush optimization that effectively eliminates half of all of the
   TLB flushes. This works by only flushing the TLB when a page is "present"
   in memory (i.e. the valid bit is set in the page table entry). See section
   5.3.5 of the Intel 386 Programmer's Reference Manual.
2) The handling of "CMAP" has been improved to catch attempts at multiple
   simultaneous use.

John:
1) Added pmap_qenter/pmap_qremove functions for fast mapping of pages into
   the kernel. This is for future optimizations and support for the upcoming
   merged VM/buffer cache.

Reviewed by: John Dyson

30 years agoThis commit was generated by cvs2svn to compensate for changes in r1823,
phk [Tue, 2 Aug 1994 20:15:59 +0000 (20:15 +0000)]
This commit was generated by cvs2svn to compensate for changes in r1823,
which included commits to RCS files with non-trunk default branches.

30 years agoHere comes the right import of gcc-2.6.0.
phk [Tue, 2 Aug 1994 20:15:59 +0000 (20:15 +0000)]
Here comes the right import of gcc-2.6.0.

30 years agoHere comes the right import of gcc-2.6.0.
phk [Tue, 2 Aug 1994 20:15:59 +0000 (20:15 +0000)]
Here comes the right import of gcc-2.6.0.

30 years agoAdded MAP_FILE definition that does nothing - for backward source
dg [Tue, 2 Aug 1994 15:06:58 +0000 (15:06 +0000)]
Added MAP_FILE definition that does nothing - for backward source
compatibility.

30 years agoCompleted (hopefully) the kernel support for old style "fastlinks".
dg [Tue, 2 Aug 1994 13:51:05 +0000 (13:51 +0000)]
Completed (hopefully) the kernel support for old style "fastlinks".

30 years agoFixed so that it can grok old style "fastlinks".
dg [Tue, 2 Aug 1994 12:54:27 +0000 (12:54 +0000)]
Fixed so that it can grok old style "fastlinks".

30 years agoDon't call getmntopts with a null pointer as an option string; it'll
dg [Tue, 2 Aug 1994 11:45:43 +0000 (11:45 +0000)]
Don't call getmntopts with a null pointer as an option string; it'll
dereference it. Pass optarg from getopts as it should have been done.
This bug caused mount_cd9660 to segfault whenever -o was used.

30 years agoFixed NULL pointer dereference that occured when any options were
dg [Tue, 2 Aug 1994 11:42:10 +0000 (11:42 +0000)]
Fixed NULL pointer dereference that occured when any options were
specified.

30 years agoAdded $Id$
dg [Tue, 2 Aug 1994 07:55:43 +0000 (07:55 +0000)]
Added $Id$

30 years agoThis commit was generated by cvs2svn to compensate for changes in r1814,
phk [Tue, 2 Aug 1994 05:22:24 +0000 (05:22 +0000)]
This commit was generated by cvs2svn to compensate for changes in r1814,
which included commits to RCS files with non-trunk default branches.

30 years agoThis is the tcl-scripts which will convert gcc-2.6.0 to bmake form,
phk [Tue, 2 Aug 1994 05:22:24 +0000 (05:22 +0000)]
This is the tcl-scripts which will convert gcc-2.6.0 to bmake form,
a patch to gcc from Bruce, and to libg++ by me.  I will check in the
actual compiler when Jordan has put in the rest of src/gnu.

30 years agofixed bug where large amounts of unidirectional UDP traffic would fill
dg [Mon, 1 Aug 1994 12:01:45 +0000 (12:01 +0000)]
fixed bug where large amounts of unidirectional UDP traffic would fill
the interface output queue and further udp packets would be fragmented
and only partially sent - keeping the output queue full and jamming the
network, but not actually getting any real work done (because you can't
send just 'part' of a udp packet - if you fragment it, you must send
the whole thing). The fix involves adding a check to make sure that the
output queue has sufficient space for all of the fragments.

30 years agoFixed bug with Nagel Congestion Avoidance where a tcp connection would
dg [Mon, 1 Aug 1994 12:00:25 +0000 (12:00 +0000)]
Fixed bug with Nagel Congestion Avoidance where a tcp connection would
stall unnecessarily - always send an ACK when a packet len of < mss is
received.

30 years agoReduced loopback MTU from 65535 to 65532 because some things like NFS
dg [Mon, 1 Aug 1994 11:39:43 +0000 (11:39 +0000)]
Reduced loopback MTU from 65535 to 65532 because some things like NFS
really like it to be rounded to a longword.

30 years agoRemoved all code related to the pagescan daemon, and changed 'act_count'
dg [Mon, 1 Aug 1994 11:25:45 +0000 (11:25 +0000)]
Removed all code related to the pagescan daemon, and changed 'act_count'
adjustments to compensate for a world without the pagescan daemon.

30 years agoAllow for output processing routine to handle entire character buffer
dg [Mon, 1 Aug 1994 10:42:26 +0000 (10:42 +0000)]
Allow for output processing routine to handle entire character buffer
and thus not require a sleep/wakeup.
Reviewed by:
Submitted by:

30 years agoFixed a bug that was introduced in the conversion from 1.1.5 to 2.0.
dg [Mon, 1 Aug 1994 10:38:19 +0000 (10:38 +0000)]
Fixed a bug that was introduced in the conversion from 1.1.5 to 2.0.

30 years agoReviewed by:
ats [Sat, 30 Jul 1994 17:16:22 +0000 (17:16 +0000)]
Reviewed by:
Submitted by:
Added a accton man page. Changed the Makefile to install the man page
as accton.8.

30 years agoRemoved pager declarations. They are NOT optional, and don't need to be
dg [Thu, 28 Jul 1994 05:53:44 +0000 (05:53 +0000)]
Removed pager declarations. They are NOT optional, and don't need to be
declared.

30 years agoDocumented the new CVS template. This really isn't the right place for it,
wollman [Sun, 24 Jul 1994 22:27:27 +0000 (22:27 +0000)]
Documented the new CVS template.  This really isn't the right place for it,
but I wanted to get this out for review without an interminable debate about
where it really should go.

30 years agoReviewed by: phk
phk [Sun, 24 Jul 1994 03:30:47 +0000 (03:30 +0000)]
Reviewed by: phk
Added libmd.

30 years agoReviewed by: phk
phk [Sun, 24 Jul 1994 03:29:56 +0000 (03:29 +0000)]
Reviewed by: phk
Imported libmd.  This library contains MD2, MD4 and MD5.
These three boggers pop up all over the place all of the time, so I
decided we needed a library with them.  In general they are used for
security checks, so if you use them you want to link them static.

30 years agoThis commit was generated by cvs2svn to compensate for changes in r1802,
phk [Sun, 24 Jul 1994 03:29:56 +0000 (03:29 +0000)]
This commit was generated by cvs2svn to compensate for changes in r1802,
which included commits to RCS files with non-trunk default branches.

30 years agoReviewed by: phk
phk [Sun, 24 Jul 1994 02:44:28 +0000 (02:44 +0000)]
Reviewed by: phk
changed ^G to \007

30 years agoReviewed by: phk
phk [Sun, 24 Jul 1994 02:04:51 +0000 (02:04 +0000)]
Reviewed by: phk
Killed a bunch of control chars.

30 years agoReviewed by: phk
phk [Sun, 24 Jul 1994 01:45:20 +0000 (01:45 +0000)]
Reviewed by: phk
Added a clean target.

30 years agoReviewed by: phk
phk [Sun, 24 Jul 1994 01:42:51 +0000 (01:42 +0000)]
Reviewed by: phk
changed a ctrl-A to \001

30 years agoReviewed by: phk
phk [Sun, 24 Jul 1994 01:39:14 +0000 (01:39 +0000)]
Reviewed by: phk
More binary files killed using uuencode.

30 years agoSubmitted by: phk
phk [Sat, 23 Jul 1994 18:14:06 +0000 (18:14 +0000)]
Submitted by: phk
Got rid of a couple of binary files by uuencoding.  49 more to go.
BTW: ed(1) seems to have an exit(2) problem, 5 tests fail.

30 years agoThe FreeBSD bootloader loads the kernel at the a_entry address in the
csgr [Fri, 22 Jul 1994 13:17:07 +0000 (13:17 +0000)]
The FreeBSD bootloader loads the kernel at the a_entry address in the
kernel a.out file, meaning that this is where the kernel starts.
(not at KERNBASE) - KERNBASE is 0xf0000000, while the kernel loads at
0xf0100000
Reviewed by:

30 years agoRewrite nlist to mmap the whole a.out file (at Davidg's suggestion).
csgr [Fri, 22 Jul 1994 12:22:51 +0000 (12:22 +0000)]
Rewrite nlist to mmap the whole a.out file (at Davidg's suggestion).
This means that we don't have to do rounding calculations for page
boundaries.  (We do all our accesses via the mmapped area now.)
Reviewed by:

30 years agoArgh! What a boneheaded bug! Finally fix the blank-line-in-packing list
jkh [Mon, 11 Jul 1994 01:11:14 +0000 (01:11 +0000)]
Argh!  What a boneheaded bug!  Finally fix the blank-line-in-packing list
problem!  I am pond scum!

30 years agoAdded code to allocate and deallocate a number of cblocks on open/close of
dg [Wed, 6 Jul 1994 06:42:34 +0000 (06:42 +0000)]
Added code to allocate and deallocate a number of cblocks on open/close of
a tty.
Note that this might  conflict with the collateral use of TS_WOPEN, but
for the moment I can find no problems associated with this. (TS_WOPEN
will likely go away in the future anyway). This should be looked at
again in the future (the potential problem is that the cblock pool
may either run out or accumulate too many cblocks).

30 years agoThe file tmac.r was not shipped on the 4.4 tape, mark this fact.
rgrimes [Wed, 6 Jul 1994 02:17:29 +0000 (02:17 +0000)]
The file tmac.r was not shipped on the 4.4 tape, mark this fact.

30 years agoall: does not need to depend on ${MAN5}, remove special case clean:
rgrimes [Wed, 6 Jul 1994 02:16:59 +0000 (02:16 +0000)]
all: does not need to depend on ${MAN5}, remove special case clean:
and cleandir: targets, simple use a CLEANFILES+= to handle this very
simple special case.
Add ${COPY} knob to install commands so that files don't disappear out
of the obj tree after a make install.

30 years agoMark the fact that we have not imported man8.{hp300,tahoe,vax}. (There
rgrimes [Wed, 6 Jul 1994 02:14:36 +0000 (02:14 +0000)]
Mark the fact that we have not imported man8.{hp300,tahoe,vax}. (There
was not man8.sparc :-).)

30 years agoMark the fact that we have not imported man4.{hp300,sparc,tahoe,vax}.
rgrimes [Wed, 6 Jul 1994 02:13:54 +0000 (02:13 +0000)]
Mark the fact that we have not imported man4.{hp300,sparc,tahoe,vax}.

30 years agoThe directory man3f did not come on the tape, mark it XXX MISSING.
rgrimes [Wed, 6 Jul 1994 02:13:00 +0000 (02:13 +0000)]
The directory man3f did not come on the tape, mark it XXX MISSING.

30 years agoThe files connectives and words did not come on the tape, mark them
rgrimes [Wed, 6 Jul 1994 02:12:25 +0000 (02:12 +0000)]
The files connectives and words did not come on the tape, mark them
XXX MISSING.

30 years agoDisable doc for now since it had a bunch of symbolic links I am not
rgrimes [Wed, 6 Jul 1994 02:11:36 +0000 (02:11 +0000)]
Disable doc for now since it had a bunch of symbolic links I am not
ready to go deal with just yet.
Disable man for now it will be fixed shortly, just wanted all the man
page converion stuff to be done togeather since that is a major functional
change and really belongs in a seperate commit.

30 years agoOh GACK! src-clean doesn't quite work that easily since cleandist rebuilds the
jkh [Wed, 29 Jun 1994 21:19:38 +0000 (21:19 +0000)]
Oh GACK!  src-clean doesn't quite work that easily since cleandist rebuilds the
obj links again and they need to be nuked out one more time after it's finished.
Sigh..  Oh well, this code will probably never be run again anyway.

30 years agoPut some of Rod's ugly hacks back - a couple of them prevent Yet Another Reboot
jkh [Wed, 29 Jun 1994 20:29:12 +0000 (20:29 +0000)]
Put some of Rod's ugly hacks back - a couple of them prevent Yet Another Reboot
after bin/des distribution is extracted, so I figure it's worth it.

30 years agoClean this puppy up quite a bit. Most notably:
jkh [Wed, 29 Jun 1994 11:06:30 +0000 (11:06 +0000)]
Clean this puppy up quite a bit.  Most notably:

1. Use ${MAKE} everywhere again.  Whoops.
2. Replace multiple invocations of gzip ... split ... with one variable.
3. Add src-clean target for making the src tree presentable before
   making a src tarball out of it.

30 years ago$rotdelay is now initialized to "-d 0 -n 1" and no longer changed based
rgrimes [Wed, 29 Jun 1994 06:46:02 +0000 (06:46 +0000)]
$rotdelay is now initialized to "-d 0 -n 1" and no longer changed based
upon disk type.  In far more cases than not this is the optimal setting
for any disk drive made after 1990.

This now means all installs will have the disks newfs'ed with either:
newfs -b 8192 -f 1024 -d 0 -n 1
or
newfs -n 4096 -f 512 -d 0 -n 1
depending on what the user chooses for the blocking factor.

30 years agoClean up the installation docs a bit (geeze Loise were they EVER out of
jkh [Tue, 28 Jun 1994 09:01:55 +0000 (09:01 +0000)]
Clean up the installation docs a bit (geeze Loise were they EVER out of
date!!) and rename them to something more eye-catching so people will read them
again (considering the previous state of affairs, I'm actually rather glad they didn't!).

30 years agoAdd --unlink flag to all tar commands, but each for different reasons.
jkh [Tue, 28 Jun 1994 07:28:28 +0000 (07:28 +0000)]
Add --unlink flag to all tar commands, but each for different reasons.

1. Add to secr and bindists to possibly save the occasional fool who
   doesn't RTFM and uses the wrong command to extract this (or even someone
   who's legitimately using this to extract on top of a bindist somewhere
   *else*).

2. Do the right thing with any symlinks in the src tree.  Right now, we're
   free of the buggers, but just in case.

30 years agoRemove Rod's hacks - they're no longer necessary with static tar/gzip and
jkh [Tue, 28 Jun 1994 07:20:30 +0000 (07:20 +0000)]
Remove Rod's hacks - they're no longer necessary with static tar/gzip and
tar --unlink.

30 years agoAdd Chris Demetriou's --unlink changes.
jkh [Tue, 28 Jun 1994 07:12:43 +0000 (07:12 +0000)]
Add Chris Demetriou's --unlink changes.

I know that I said earlier that this should be unconditional behaviour,
but I thought about it a little more and concluded that the principle of least
surprise dictates that I make it an option.

30 years agoCommit Scott's changes to the bad144 stuff - bad144 takes a list of bad
jkh [Tue, 28 Jun 1994 05:56:13 +0000 (05:56 +0000)]
Commit Scott's changes to the bad144 stuff - bad144 takes a list of bad
blocks as *multiple* arguments, not one argument (as was incorrectly being
passed).

30 years agoFix a simple makefile brain-o; didn't effect 1.1.5A since we don't compile
jkh [Tue, 28 Jun 1994 05:37:04 +0000 (05:37 +0000)]
Fix a simple makefile brain-o; didn't effect 1.1.5A since we don't compile
with CRYPT initially set.

30 years agoFixed spelling errors.
wollman [Fri, 24 Jun 1994 23:50:16 +0000 (23:50 +0000)]
Fixed spelling errors.

30 years ago1. Remove a rather strangely gratuitous bit of profanity
jkh [Fri, 24 Jun 1994 22:18:46 +0000 (22:18 +0000)]
1. Remove a rather strangely gratuitous bit of profanity
2. Use elvis instead of vi for the editor on the cpio floppy; dmesg is back.

30 years agoAdd proper links so that - and -sh are catered for.
jkh [Fri, 24 Jun 1994 16:39:30 +0000 (16:39 +0000)]
Add proper links so that - and -sh are catered for.

30 years agoJust talked with James - there's another, better way to go about this.
jkh [Fri, 24 Jun 1994 15:14:16 +0000 (15:14 +0000)]
Just talked with James - there's another, better way to go about this.
Back out my earlier change.  Note that this is just for the 1.1.5R floppies;
the 1.1.5A ones still have the work-around method (which works fine and
doesn't hurt anything, it's just kludge!).

30 years agoThis is a kludge. Unless I make a link from sh to - (which would leave a file
jkh [Fri, 24 Jun 1994 14:32:36 +0000 (14:32 +0000)]
This is a kludge.  Unless I make a link from sh to - (which would leave a file
called `-' lying around on the users system forever) login shells will fail.
Just special-case the handling of `-' for now until/unless I find a more
palatable solution.

30 years agoMy last few changes to the release code plus a small termcap file
jkh [Fri, 24 Jun 1994 14:27:25 +0000 (14:27 +0000)]
My last few changes to the release code plus a small termcap file
for the cpio floppy.

30 years agoWhoops - missed a couple of intermediate files. Nuke 'em.
jkh [Wed, 22 Jun 1994 08:37:28 +0000 (08:37 +0000)]
Whoops - missed a couple of intermediate files.  Nuke 'em.

30 years agoOk, this should about wrap it up.
jkh [Wed, 22 Jun 1994 08:33:28 +0000 (08:33 +0000)]
Ok, this should about wrap it up.

1. All the crunched files work now; don't install copies, used the crunched
   versions.
2. Make the clean rules DTRT now that we have stuff to clean.

30 years agoCommit a whole cluster of last minute critical (and one cosmetic) fixes
jkh [Wed, 22 Jun 1994 05:52:25 +0000 (05:52 +0000)]
Commit a whole cluster of last minute critical (and one cosmetic) fixes
from David Greenman, Bruce Evans and Julian Elischer.

They are:

[vnode pager - David/Bruce]:
  The following patch fixes a problem where some data could be lost in a
delayed-write buffer if the cached buffer was larger than a page. This fix was
provided by Bruce Evans and modified slightly by me.

[st.c - Julian]:
  My fix for "bad request, must be between 0 and 0"

RTFS if you're interested).

[gnu/fpemul - David/Bruce]:
  These changes fix single stepping of emulated FPU instructions.
Previously, the instruction after an emulated instruction was
executed without causing a SIGTRAP ...

The also fix the initial control word being different for the
GPL emulator (it is still wrong for the old emulator) and remove
an unnecessary panic when emulation is not configured (I hope at
least init, sh and reboot will run without floating point.  I
remember only df and mkfs being broken by the lack of FP in 0.0).

[Various fixes described below - Bruce/David]:
sys/i386/boot/boot2.S:
        Yet another attempt to propagate the correct fix for 16 vs
        32-bit mode bugs.  [verified]

sys/i386/i386/db_interface.c:
        Protect against reentering Debugger().

sys/kern/kern_time.c:
        Don't allow 'time.tv_usec == 0' except at clock interrupts.

sys/pcfs/pcfs_fat.c:
        Make it compile without -O.

sys/scsi/sd.c:
        Fix as posted to some freebsd mailing list.
        (changes the order of the assignment of "sectors" because it earlier
          value is needed first -DG)

sys/vm/vm_glue.c:
        Fix stale comments and verbose code.

sys/vm/vm_mmap.c
        Fix off by 1 errors and verbose code.

[From Nate - cosmetic but non-intrusive and useful enough to go in]
sys/i386/isa/isa.c:
Appended you'll find a patch to the NMI error log routine in isa/isa.c.
The below patch just adds some additional information when an NMI occurs
which can help debug the hardware problem.

30 years agoUpdate this to the latest version from Steve Gerakines. This is an easy
jkh [Wed, 22 Jun 1994 04:49:04 +0000 (04:49 +0000)]
Update this to the latest version from Steve Gerakines.  This is an easy
drop-in for me and looks substantailly neater than the previous version,
so I'll give the floppy tape users a break (but just this once :).

30 years agoFix serious off by one error for FreeBSD keyword, this has been driving
rgrimes [Wed, 22 Jun 1994 00:51:42 +0000 (00:51 +0000)]
Fix serious off by one error for FreeBSD keyword, this has been driving
me nuts as it was on by default and that is NOT what I wanted.

30 years agoWhoops - I missed a commit; EXTRACT scripts were misnamed. Correct.
jkh [Tue, 21 Jun 1994 12:42:56 +0000 (12:42 +0000)]
Whoops -  I missed a commit; EXTRACT scripts were misnamed.  Correct.

30 years ago1. Increase number of inodes for kcopy floppy; we've currently run out.
jkh [Tue, 21 Jun 1994 11:02:38 +0000 (11:02 +0000)]
1. Increase number of inodes for kcopy floppy; we've currently run out.
2. Get kcopy and filesystem images from current directory since we
   now build them here; a clean rule is now all that's needed to make
   the crunch stuff complete.

30 years agoCommit a small change to allow FreeBSD to install on a different drive.
jkh [Mon, 20 Jun 1994 06:05:16 +0000 (06:05 +0000)]
Commit a small change to allow FreeBSD to install on a different drive.
The kernel configs already support this, so with a boot floppy or a utility
like booteasy, the user should be able to install and boot off the second drive.
Hurrah.

30 years agoMake BOOTWAIT a truly tunable parameter. It was hard-set before.
jkh [Mon, 20 Jun 1994 04:32:40 +0000 (04:32 +0000)]
Make BOOTWAIT a truly tunable parameter.  It was hard-set before.

30 years agoCommit some additional changes for making this work with obj dirs.
jkh [Mon, 20 Jun 1994 03:39:59 +0000 (03:39 +0000)]
Commit some additional changes for making this work with obj dirs.

30 years agoRemove bogus -lcrypt line.
jkh [Sun, 19 Jun 1994 20:39:30 +0000 (20:39 +0000)]
Remove bogus -lcrypt line.

30 years agoFixes from Bruce Evans:
jkh [Fri, 17 Jun 1994 21:50:07 +0000 (21:50 +0000)]
Fixes from Bruce Evans:

o insecure and inconsistent permissions for tape devices.
o didn't make the DOSpartition wd devices.

30 years agoAdd entry for the new PCMCIA ethernet adapter code (ze0).
jkh [Fri, 17 Jun 1994 19:29:05 +0000 (19:29 +0000)]
Add entry for the new PCMCIA ethernet adapter code (ze0).