]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
29 years ago1. Mark file as seriously out of date with reality.
rgrimes [Sun, 9 Oct 1994 21:22:32 +0000 (21:22 +0000)]
1.  Mark file as seriously out of date with reality.

2.  Update the list of library names and variables.

3.  Update to reflect forth coming bsd.inc.mk file.

4.  Update which .mk files include other .mk files.

Submitted by: Bruce Evans (partial, enhanced by me)

29 years ago1. Remove comments about broken procfs, it has been fixed.
rgrimes [Sun, 9 Oct 1994 20:31:01 +0000 (20:31 +0000)]
1.  Remove comments about broken procfs, it has been fixed.

2.  Remove comments about exit status of 2 from mtree, that won't
    happen any more.

3.  Update mtree commands to use new -U option.

29 years agoAdd -U option which does the same things as -u except exits with 0
rgrimes [Sun, 9 Oct 1994 20:28:31 +0000 (20:28 +0000)]
Add -U option which does the same things as -u except exits with 0
instead of 2 on MISMATCH.

29 years agoBacked out part of the last change that prevents the rpos table from
dg [Sun, 9 Oct 1994 20:10:56 +0000 (20:10 +0000)]
Backed out part of the last change that prevents the rpos table from
being output if <= 1 rpos; there is a bug in the kernel which doesn't
quite get along with this. Changed default #rpos to 1, and fixed up
manual page. Converted nrpos to 1 if user specifies 0.

29 years agoGenerate term.h only once, not each beforeinstall.
ache [Sun, 9 Oct 1994 18:39:01 +0000 (18:39 +0000)]
Generate term.h only once, not each beforeinstall.
Add 'update_term_h' target for manual update

29 years agoSplit all binaries from makefile now
ache [Sun, 9 Oct 1994 17:50:50 +0000 (17:50 +0000)]
Split all binaries from makefile now

29 years agoThis commit was generated by cvs2svn to compensate for changes in r3462,
ache [Sun, 9 Oct 1994 17:40:30 +0000 (17:40 +0000)]
This commit was generated by cvs2svn to compensate for changes in r3462,
which included commits to RCS files with non-trunk default branches.

29 years agoSplitted from libmytinfo
ache [Sun, 9 Oct 1994 17:40:30 +0000 (17:40 +0000)]
Splitted from libmytinfo

29 years agoPreparation to split binary away, move caps to TESTS
ache [Sun, 9 Oct 1994 16:37:35 +0000 (16:37 +0000)]
Preparation to split binary away, move caps to TESTS
Add some testing bits to TESTS directory

29 years agoBack out part of ctype fixes, unneded with new ctypoe
ache [Sun, 9 Oct 1994 15:40:07 +0000 (15:40 +0000)]
Back out part of ctype fixes, unneded with new ctypoe

29 years agoBack out ctype fix, unneded with new ctype
ache [Sun, 9 Oct 1994 15:25:23 +0000 (15:25 +0000)]
Back out ctype fix, unneded with new ctype

29 years agoBack out part of ctype fixes, unneded with new ctype
ache [Sun, 9 Oct 1994 15:22:35 +0000 (15:22 +0000)]
Back out part of ctype fixes, unneded with new ctype

29 years agoSync with ctype.h (remove EOF handling)
ache [Sun, 9 Oct 1994 11:20:55 +0000 (11:20 +0000)]
Sync with ctype.h (remove EOF handling)

29 years agoRemove EOF handling after Bruce explanation. This step returns
ache [Sun, 9 Oct 1994 11:18:44 +0000 (11:18 +0000)]
Remove EOF handling after Bruce explanation. This step returns
to 4.4 way to not allow EOF in ctype and now all signed chars
(including '\377' which becomes EOF) converted to (unsigned char) properly.

29 years agoAdd copyright.c to be always linked in
ache [Sun, 9 Oct 1994 09:27:55 +0000 (09:27 +0000)]
Add copyright.c to be always linked in

29 years agoModified to use swaplist instead of swapmap.
dg [Sun, 9 Oct 1994 07:45:58 +0000 (07:45 +0000)]
Modified to use swaplist instead of swapmap.

29 years ago#if 0'd out the meat of the swap code until I get a chance to rewrite it.
dg [Sun, 9 Oct 1994 07:42:32 +0000 (07:42 +0000)]
#if 0'd out the meat of the swap code until I get a chance to rewrite it.

29 years ago#if 0'd out the meat of the swap code until I get a chance to rewrite it.
dg [Sun, 9 Oct 1994 07:37:18 +0000 (07:37 +0000)]
#if 0'd out the meat of the swap code until I get a chance to rewrite it.

29 years agoGot rid of map.h. It's a leftover from the rmap code, and we use rlists.
dg [Sun, 9 Oct 1994 07:35:18 +0000 (07:35 +0000)]
Got rid of map.h. It's a leftover from the rmap code, and we use rlists.
Changed swapmap into swaplist.

29 years agoRestore leading tabs
ache [Sun, 9 Oct 1994 02:09:14 +0000 (02:09 +0000)]
Restore leading tabs

29 years agoCosmetics: unused vars, ()'s, #include's &c &c to silence gcc.
phk [Sun, 9 Oct 1994 01:52:19 +0000 (01:52 +0000)]
Cosmetics:  unused vars, ()'s, #include's &c &c to silence gcc.
Reviewed by: davidg

29 years agoUpdate multicasting FAQ entry
pst [Sun, 9 Oct 1994 01:14:51 +0000 (01:14 +0000)]
Update multicasting FAQ entry

29 years agoCosmetic cleanup
pst [Sun, 9 Oct 1994 00:46:16 +0000 (00:46 +0000)]
Cosmetic cleanup

29 years agoCall resetpriority, not setpriority() ...oops.
dg [Sun, 9 Oct 1994 00:18:22 +0000 (00:18 +0000)]
Call resetpriority, not setpriority() ...oops.

Submitted by: John Dyson

29 years agoBe optionally verbose about free-space checking. I need this to ascertian
jkh [Sat, 8 Oct 1994 23:55:36 +0000 (23:55 +0000)]
Be optionally verbose about free-space checking.  I need this to ascertian
whether or not this always works.

29 years agoCosmetics: silences gcc -Wall.
phk [Sat, 8 Oct 1994 22:39:58 +0000 (22:39 +0000)]
Cosmetics:  silences gcc -Wall.

29 years agoCosmetics: to silence gcc -wall.
phk [Sat, 8 Oct 1994 22:38:27 +0000 (22:38 +0000)]
Cosmetics:  to silence gcc -wall.

29 years agoCosmetics: added a #include and a static prototype to silence gcc.
phk [Sat, 8 Oct 1994 22:37:00 +0000 (22:37 +0000)]
Cosmetics:  added a #include and a static prototype to silence gcc.

29 years agoCosmetics: added ()'s and fixed prinf-formats to make gcc silent.
phk [Sat, 8 Oct 1994 22:33:43 +0000 (22:33 +0000)]
Cosmetics: added ()'s and fixed prinf-formats to make gcc silent.

29 years agoA couple of prototypes moved out from here.
phk [Sat, 8 Oct 1994 22:27:58 +0000 (22:27 +0000)]
A couple of prototypes moved out from here.

29 years agoRemoved unused vars. Silenced gcc -Wall. Some of this stuff is virtually
phk [Sat, 8 Oct 1994 22:26:40 +0000 (22:26 +0000)]
Removed unused vars.  Silenced gcc -Wall.  Some of this stuff is virtually
type-less.  Everything is "u_int32" :-(

29 years agoAdded prototypes here and there. Moved pfctlinput into socket.h.
phk [Sat, 8 Oct 1994 22:23:00 +0000 (22:23 +0000)]
Added prototypes here and there.  Moved pfctlinput into socket.h.

29 years agoAdded prototypes.
phk [Sat, 8 Oct 1994 22:21:34 +0000 (22:21 +0000)]
Added prototypes.

29 years agodb_disasm.c: Unused var zapped.
phk [Sat, 8 Oct 1994 22:19:51 +0000 (22:19 +0000)]
db_disasm.c: Unused var zapped.
pmap.c: tons of unused vars zapped, various other warnings silenced.
trap.c: unused vars zapped.
vm_machdep.c:  A wrong argument, which by chance did the right thing, was
corrected.

29 years agoThis file describes new features, conventions or newly introduced "gotchas"
jkh [Sat, 8 Oct 1994 17:57:45 +0000 (17:57 +0000)]
This file describes new features, conventions or newly introduced "gotchas"
that a user coming from a 1.x environment should know about.

See the notes in the file for more info.

29 years agoSync with ctype.h (EOF, sign extention fixes)
ache [Sat, 8 Oct 1994 17:42:45 +0000 (17:42 +0000)]
Sync with ctype.h (EOF, sign extention fixes)

29 years agoHandle EOF case in all macros by ANSI standard.
ache [Sat, 8 Oct 1994 17:36:44 +0000 (17:36 +0000)]
Handle EOF case in all macros by ANSI standard.
Cast all ints < 0 to (unsigned char) to fix common problem
with sign extention on signed char.

29 years agoDialog added
ache [Sat, 8 Oct 1994 15:19:12 +0000 (15:19 +0000)]
Dialog added

29 years agoThis commit was generated by cvs2svn to compensate for changes in r3430,
ache [Sat, 8 Oct 1994 15:16:41 +0000 (15:16 +0000)]
This commit was generated by cvs2svn to compensate for changes in r3430,
which included commits to RCS files with non-trunk default branches.

29 years agoMoved from ports, arrow keys allowed to edit input,
ache [Sat, 8 Oct 1994 15:16:41 +0000 (15:16 +0000)]
Moved from ports, arrow keys allowed to edit input,
back tab key understanded now

29 years agolibmytinfo, libncurses added to 'make beforeinstall' sequence
ache [Sat, 8 Oct 1994 15:08:14 +0000 (15:08 +0000)]
libmytinfo, libncurses added to 'make beforeinstall' sequence

29 years agoCompletely removed trailer support. The only reason I wrote that code in
dg [Sat, 8 Oct 1994 09:24:20 +0000 (09:24 +0000)]
Completely removed trailer support. The only reason I wrote that code in
the first place was so that BPF could grok trailer packets. I've since
decided that this is a job for tcpdump to decipher (if at all). Also
fixed up checks for received packet length to better cope with ancient
starlan boards.

29 years agoPOSSIBLE BOGUS CODE found, (related to dos-partitions) in ufs_disksubr.c,
phk [Sat, 8 Oct 1994 06:57:29 +0000 (06:57 +0000)]
POSSIBLE BOGUS CODE found, (related to dos-partitions) in ufs_disksubr.c,
look for CC_WALL.
Cosmetics, a couple of unused vars.

29 years agoCorrect #ifdef for nfs_disless support is #ifdef NFS, there will be no
rgrimes [Sat, 8 Oct 1994 06:20:52 +0000 (06:20 +0000)]
Correct #ifdef for nfs_disless support is #ifdef NFS, there will be no
option DISKLESS for the 2.0 nfs diskless support.   A 2.0 diskless kernel
simple needs NFS linked in statically.

29 years agoCosmetics for gcc -Wall. A couple of unused "int i"'s removed and a couple of
phk [Sat, 8 Oct 1994 06:20:06 +0000 (06:20 +0000)]
Cosmetics for gcc -Wall.  A couple of unused "int i"'s removed and a couple of
prototypes added.  And the usual () work.

29 years agoI got my patch-file applied twice it seems. Fixed.
phk [Sat, 8 Oct 1994 03:31:53 +0000 (03:31 +0000)]
I got my patch-file applied twice it seems.  Fixed.

29 years agoAdd HW.TROUBLE to ROADMAP.
jkh [Sat, 8 Oct 1994 03:15:53 +0000 (03:15 +0000)]
Add HW.TROUBLE to ROADMAP.

29 years agoPut this here, moving it from /usr/src.
jkh [Sat, 8 Oct 1994 03:10:34 +0000 (03:10 +0000)]
Put this here, moving it from /usr/src.

29 years agoPrototypes of today. Brought to you by a 28 minute transit time on BART :-)
phk [Sat, 8 Oct 1994 01:45:30 +0000 (01:45 +0000)]
Prototypes of today.  Brought to you by a 28 minute transit time on BART :-)

(For the SF-unaware:  I ride the BART (The Bay-area subway) for half an hour
each way to work.  I use the time to shut up gcc -Wall on my handbook).

29 years agoCosmetics.
phk [Sat, 8 Oct 1994 01:41:15 +0000 (01:41 +0000)]
Cosmetics.

29 years agoMostly Cosmetics. Some of the procedures in if_sl.c was void, but should
phk [Sat, 8 Oct 1994 01:40:23 +0000 (01:40 +0000)]
Mostly Cosmetics.  Some of the procedures in if_sl.c was void, but should
be int.  I made them int, and let them return 0.  Will have to find out
what the return-val is used for.

29 years agoMake inflate() reentrant - no more global variables.
csgr [Fri, 7 Oct 1994 23:18:18 +0000 (23:18 +0000)]
Make inflate() reentrant - no more global variables.
(The gzip image activator now needs a good code tidy up.)

29 years agoFirst stage of getting imgact_gzip reentrant:
csgr [Fri, 7 Oct 1994 22:27:00 +0000 (22:27 +0000)]
First stage of getting imgact_gzip reentrant:
1) cut this up into /sys/sys/inflate.h, sys/kern/inflate.c
sys/kern/ingact_gzip.c
2) make a lot more things static
3) make a lot of globals const
4) make some args const
5) first stage of making globals into a struct (not used yet)

The vm_allocate() call which was introduced between revisions 1.4 and
1.5 of imagact_gzip.c broke things.  I have backed that out for the time
being.  (Davidg: help please)

WARNING: if you have gzip enabled in your kernel, you must now run
config again, as another source file has been added.  Otherwise your
kernel compile will fall over.

This is all still WIP.  More commits to come.

Suggestions from: phk.

29 years agoI forgot this bit of APM stuff last week.
phk [Fri, 7 Oct 1994 21:23:33 +0000 (21:23 +0000)]
I forgot this bit of APM stuff last week.

29 years agoReviewed by: phk
phk [Fri, 7 Oct 1994 21:17:41 +0000 (21:17 +0000)]
Reviewed by:  phk
Submitted by:  Thomas David Rivers <rivers%ponds@ncren.net>

WARNING: might hide some bug below!  I commit this to improve the stability
of 2.0.

Thomas wrote:
-------------
 I have been running a kernel with this change since October 4th; barring
unrelated network router troubles, the pitiful little machine has
completed several builds without any interaction from me, and continues
to chug along.

 I re-read wd.c, and added appropriate printfs() to look for references
to dk_badsect[].  My changes should have printed something when dk_badsect[]
was referenced.

 I got no output :-(

 Thus, I'm forced to concluded that something else is examining some
spurious memory... which happened to be in dk_badsect[] of the disk structure
in wd.c.  I can find no other explanation of why this unnecessary
initialization causes things to operate correctly.

 On the premise that such an initialization isn't going to hurt anything,
I'm going to suggest it go into 2.0.

 I'd like to thank everyone for there assistance, particularly David,
John and Bruce.

29 years agoFix many problems with 8bit chars (sign extend in ctype macros)
ache [Fri, 7 Oct 1994 13:13:32 +0000 (13:13 +0000)]
Fix many problems with 8bit chars (sign extend in ctype macros)
Fix main problem with 8-bit chars in directories names: because
signed_sum left uninitialized, wrong checksum occurse

29 years agolibncurses added
ache [Fri, 7 Oct 1994 09:28:33 +0000 (09:28 +0000)]
libncurses added

29 years agoThis commit was generated by cvs2svn to compensate for changes in r3410,
ache [Fri, 7 Oct 1994 08:58:58 +0000 (08:58 +0000)]
This commit was generated by cvs2svn to compensate for changes in r3410,
which included commits to RCS files with non-trunk default branches.

29 years agoMoved from ports with several enhancements
ache [Fri, 7 Oct 1994 08:58:58 +0000 (08:58 +0000)]
Moved from ports with several enhancements

29 years agoMoved from ports with several enhancements
ache [Fri, 7 Oct 1994 08:58:58 +0000 (08:58 +0000)]
Moved from ports with several enhancements

29 years agoBack out the fluff, leave the meat
pst [Fri, 7 Oct 1994 07:48:18 +0000 (07:48 +0000)]
Back out the fluff, leave the meat

29 years agoCosmetics. Added four prototypes.
phk [Fri, 7 Oct 1994 07:33:51 +0000 (07:33 +0000)]
Cosmetics.  Added four prototypes.

29 years agoCosmetics. Unused vars and other warnings.
phk [Fri, 7 Oct 1994 07:23:04 +0000 (07:23 +0000)]
Cosmetics.  Unused vars and other warnings.

29 years ago#ifdef DISKLESS the copying of the nfs_diskless structure. Not the best
dg [Fri, 7 Oct 1994 05:45:27 +0000 (05:45 +0000)]
#ifdef DISKLESS the copying of the nfs_diskless structure. Not the best
solution, but the only one I have time for at the moment.

29 years agoChange first CFLAGS+= to CFLAGS= or we got really bad results
ache [Fri, 7 Oct 1994 05:36:01 +0000 (05:36 +0000)]
Change first CFLAGS+= to CFLAGS= or we got really bad results
especially with -m486 from make.conf

29 years agoLIBMYTINFO and LIBNCURSES defines added
ache [Fri, 7 Oct 1994 03:10:05 +0000 (03:10 +0000)]
LIBMYTINFO and LIBNCURSES defines added

29 years ago(a) there's no reason for PASSIVEMODE to be conditionally compiled since it's
pst [Thu, 6 Oct 1994 23:57:21 +0000 (23:57 +0000)]
(a) there's no reason for PASSIVEMODE to be conditionally compiled since it's
    controlled by a runtime switch.
(b) add '-P' to toggle passive mode from the command line
(c) turn on passive mode by default
- passive mode ftp works on all but a few servers out there
- it's easy to disable
- no, standard ftp should not run with passive enabled by default,
  but that doesn't matter because I've already fixed standard mode
  ftp to take a switch

This really should have been implemented as a ncftp "set" command instead
of carrying over the same hack that was done to ftp.  I will go back and fix
it some day.

29 years agolibmytinfo added.
ache [Thu, 6 Oct 1994 23:41:24 +0000 (23:41 +0000)]
libmytinfo added.
PS: don't forget 'make beforeinstall' if you want to play with it

29 years agoMoved from ports + several enhacmenets
ache [Thu, 6 Oct 1994 23:38:41 +0000 (23:38 +0000)]
Moved from ports + several enhacmenets

29 years agoThis commit was generated by cvs2svn to compensate for changes in r3400,
ache [Thu, 6 Oct 1994 23:38:41 +0000 (23:38 +0000)]
This commit was generated by cvs2svn to compensate for changes in r3400,
which included commits to RCS files with non-trunk default branches.

29 years agoSync with termcap.src
ache [Thu, 6 Oct 1994 21:40:42 +0000 (21:40 +0000)]
Sync with termcap.src

29 years agoChange do= from ^J to \E[B, this escape must works in all terminal modes
ache [Thu, 6 Oct 1994 21:37:48 +0000 (21:37 +0000)]
Change do= from ^J to \E[B, this escape must works in all terminal modes
and old variant don't work properly, if \n --> \r\n (cons25,pc3)

29 years agoBack out ospeed change
ache [Thu, 6 Oct 1994 21:07:55 +0000 (21:07 +0000)]
Back out ospeed change

29 years agoUse tsleep() rather than sleep so that 'ps' is more informative about
dg [Thu, 6 Oct 1994 21:07:04 +0000 (21:07 +0000)]
Use tsleep() rather than sleep so that 'ps' is more informative about
the wait.

29 years agoBack out ospeed change (why all bugs discovered immediately _after_ commits
ache [Thu, 6 Oct 1994 21:05:23 +0000 (21:05 +0000)]
Back out ospeed change (why all bugs discovered immediately _after_ commits
and no one before?)

29 years agoOops, back out ospeed fix, I forget that flushing can occurse _after_
ache [Thu, 6 Oct 1994 21:02:21 +0000 (21:02 +0000)]
Oops, back out ospeed fix, I forget that flushing can occurse _after_
usleep, too quick commit :-(

29 years agoMake multicasts go out default interface
pst [Thu, 6 Oct 1994 20:52:00 +0000 (20:52 +0000)]
Make multicasts go out default interface

29 years agoEnable PC back, because it is ouside visible
ache [Thu, 6 Oct 1994 20:44:51 +0000 (20:44 +0000)]
Enable PC back, because it is ouside visible

29 years agoSince Rod insists on having this idiotic error message, at least call
wollman [Thu, 6 Oct 1994 20:43:35 +0000 (20:43 +0000)]
Since Rod insists on having this idiotic error message, at least call
warnx() correctly so we don't get the double newline.

29 years agoPC/ospeed code ifdefed out, our libtermcap use usleep now
ache [Thu, 6 Oct 1994 20:41:56 +0000 (20:41 +0000)]
PC/ospeed code ifdefed out, our libtermcap use usleep now

29 years agoospeed/PC code ifdefed out, our libtermcap use usleep now
ache [Thu, 6 Oct 1994 20:36:34 +0000 (20:36 +0000)]
ospeed/PC code ifdefed out, our libtermcap use usleep now

29 years agoChange padding mechanism to use usleep, old variant not works
ache [Thu, 6 Oct 1994 20:32:08 +0000 (20:32 +0000)]
Change padding mechanism to use usleep, old variant not works
on terminals with no pad char (cons25) and quote from tputs.c says so too:
!   * Too bad there are no user program accessible programmed delays.
!   * Transmitting pad characters slows many
!   * terminals down and also loads the system.

29 years agoSteven Wallace provided a program which broke this stuff. I guess there are
phk [Thu, 6 Oct 1994 18:22:24 +0000 (18:22 +0000)]
Steven Wallace provided a program which broke this stuff.  I guess there are
more weird kinds of a.out than anyone can argue for.  This code failed to
load the first 28K of the text-segment, in the case where the first page
of the a.out contains only the a.out-header, and the text is still at 0x0.
Thanks Steven !

29 years ago1. BOOTSEG and BOOTSTACK are now set from the Makefile, the boot code has
rgrimes [Thu, 6 Oct 1994 09:41:05 +0000 (09:41 +0000)]
1.  BOOTSEG and BOOTSTACK are now set from the Makefile, the boot code has
    been relocated to run in the 64k segment at 0x10000 with the stack at
    the top of this segment.  This corrects the problems machines with 512K
    base memory had booting.

2.  startprog routing rewritten to convert the BOOTSEG ss to a KERNELSEG
    ss, this eliminated the last of the >512K memory references.  Additional
    cleanup in here included a better way to copy the arguments to the
    kernel stack.

3.  Elimination of argv and esym cruft saved a few bytes.

4.  Only need to truncate the head.a_entry to a meg boundary once intead
    of every time we used it!  [Saving more bytes].

5.  Addition of version 1 bootinfo structure support.  These boot blocks
    pass the kernel name in to the kernel now.

6.  Removed historical comments about MACH argv stuff, as it is useless now.

29 years ago1. bootinfo.h defines the structure passed in to the kernel by the
rgrimes [Thu, 6 Oct 1994 09:25:25 +0000 (09:25 +0000)]
1.  bootinfo.h defines the structure passed in to the kernel by the
    new boot code.

29 years ago1. Eliminate unused esym global from locore, our boot code never supported
rgrimes [Thu, 6 Oct 1994 09:22:47 +0000 (09:22 +0000)]
1.  Eliminate unused esym global from locore, our boot code never supported
    that and when it does it will be done differently.

2.  The kernel now does a frame setup on entry so it ``looks'' like a
    real function call.  This will be needed by future boot code and
    debuggers.

3.  Clean up stack offsets to all be in decimal and use %ebp when copying
    parameters in from the boot code.

4.  Implement version 1 of the uniform boot code passing mechanism with
    support for kernelname passing and nfs_diskless structure passing.

5.  Document the 3 different ways the kernel is called depending on what code
    is calling it.

29 years agoZap 4.0 libkdb as well.
wollman [Thu, 6 Oct 1994 00:48:55 +0000 (00:48 +0000)]
Zap 4.0 libkdb as well.

29 years agoA few fixes:
wollman [Thu, 6 Oct 1994 00:32:42 +0000 (00:32 +0000)]
A few fixes:

- register, registerd, and make_keypair don't compile (and are bogus anyway)
- don't forget to put back the obj directory when doing `kprog'
- while we're at it make the `kprog' commands overrideable from the command
  line
- add a bootstrap target which does the following:
install includes
cleandir and obj
zap old version 4.0 shared libraries (these will screw the build)
depend all install
rebuild stuff in the main source tree which depends on kerberos

29 years agoFollowing changes from Robert Withrow (+ a few mods):
jkh [Wed, 5 Oct 1994 22:28:45 +0000 (22:28 +0000)]
Following changes from Robert Withrow (+ a few mods):

1. Make DEPENDS fully qualified, and not implictly assume relative
   to ${PORTSDIR}.  This allows more arbitrary dependencies to be
   specified.  This also means that DEPENDS= x11/foo needs to be changed
   to DEPENDS= ${PORTSDIR}/x11/foo in any Makefiles.  I'll try to do
   these changes myself.

2. Add an option NO_DEPENDS to disable the automagic building of depended
   ports.
Submitted by: rww

29 years agoDefine a new macro. PSEUDO_SET, to hide TEXT_SET(pseudo_set, foo)
wollman [Wed, 5 Oct 1994 21:23:38 +0000 (21:23 +0000)]
Define a new macro. PSEUDO_SET, to hide TEXT_SET(pseudo_set, foo)
from users.  Eventually this will be used for LKM support.

29 years agoInstall line discipline the new way.
wollman [Wed, 5 Oct 1994 21:22:45 +0000 (21:22 +0000)]
Install line discipline the new way.

29 years agoAllow for loadable line disciplines (eventually).
wollman [Wed, 5 Oct 1994 21:22:24 +0000 (21:22 +0000)]
Allow for loadable line disciplines (eventually).

29 years agoA number of bug-fixes inspired by Mark Treacy:
wollman [Wed, 5 Oct 1994 20:11:28 +0000 (20:11 +0000)]
A number of bug-fixes inspired by Mark Treacy:
- Allow PPP to run multicasts natively.
- Deal properly with lots of similarly-named interfaces.
- Don't sign-extend if_flags.

NB: the last fix (to rtsock.c) must be reversed when we expand if_flags to a
reasonable size.

Submitted by: Mark Treacy

29 years agoAfter a comment from Rod Grimes about buf.h, I went back and looked at this
dg [Wed, 5 Oct 1994 10:20:17 +0000 (10:20 +0000)]
After a comment from Rod Grimes about buf.h, I went back and looked at this
and found that swapinfo doesn't need that include file and five others (!).
Sheesh.

29 years agoChanges to buf.h now require that param.h be included.
dg [Wed, 5 Oct 1994 09:57:49 +0000 (09:57 +0000)]
Changes to buf.h now require that param.h be included.

29 years agoStuff object into v_vmdata rather than pager. Not important which at
dg [Wed, 5 Oct 1994 09:48:45 +0000 (09:48 +0000)]
Stuff object into v_vmdata rather than pager. Not important which at
the moment, but will be in the future. Other changes mostly cosmetic,
but are made for future VMIO considerations.

Submitted by: John Dyson

29 years agoFixed minor bug caused by some missing parens that can result in slightly
dg [Wed, 5 Oct 1994 09:02:53 +0000 (09:02 +0000)]
Fixed minor bug caused by some missing parens that can result in slightly
reduced paging performance by missing a clustering opportunity. Found
by Poul-Henning Kamp with gcc -Wall.

29 years agoRealigned the output of "vmstat -m", "MSDOSFS mount" was too wide for the
phk [Wed, 5 Oct 1994 04:52:47 +0000 (04:52 +0000)]
Realigned the output of "vmstat -m", "MSDOSFS mount" was too wide for the
field.  Saved some space and gained a little clarity by printing "128K"
instead of 131072 (and so on).

29 years agoDavid Greenman told me to do this: (Thanks!) use vm_allocate to allocate
phk [Wed, 5 Oct 1994 00:58:33 +0000 (00:58 +0000)]
David Greenman told me to do this: (Thanks!) use vm_allocate to allocate
the uncompression buffer.  Now malloc(M_GZIP) is used for all the Huffman-
tree stuff only.  Numbers so far indicate < 15Kb Malloc use + 32 Kb for
the abovementioned buffer while uncompressing.

29 years agoCorrectly interpret the code field of IGMP membership queries.
wollman [Tue, 4 Oct 1994 22:47:11 +0000 (22:47 +0000)]
Correctly interpret the code field of IGMP membership queries.