]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
26 years agoRemoved definition of _NEW_VFSCONF. The new vfsconf interface is now
bde [Tue, 20 Jan 1998 10:40:18 +0000 (10:40 +0000)]
Removed definition of _NEW_VFSCONF.  The new vfsconf interface is now
the default.

26 years agoMoved most of the (source-level) compatibility hacks for the vfsconf
bde [Tue, 20 Jan 1998 10:36:42 +0000 (10:36 +0000)]
Moved most of the (source-level) compatibility hacks for the vfsconf
interface from sys/mount.h to libc/getvfsent.c  The new interface is
now the default.

Sorted the prototypes for the library functions.

26 years agoMoved most of the (source-level) compatibility hacks for the vfsconf
bde [Tue, 20 Jan 1998 10:36:24 +0000 (10:36 +0000)]
Moved most of the (source-level) compatibility hacks for the vfsconf
interface from sys/mount.h to libc/getvfsent.c  The new interface is
now the default.

26 years ago- Move SETKLOC and CLEARKLOCK macros into uion.h.
kato [Tue, 20 Jan 1998 10:02:54 +0000 (10:02 +0000)]
- Move SETKLOC and CLEARKLOCK macros into uion.h.
- Set UN_ULOCK in union_lock() when UN_KLOCK is set.  Caller expects
  that vnode is locked correctly, and may call another function which
  expects locked vnode and may unlock the vnode.
- Do not assume the behavior of inside functions in FreeBSD's
  vfs_suber.c is same as 4.4BSD-Lite2.  Vnode may be locked in
  vget() even though flag is zero.  (Locked vnode is, of course,
  unlocked before returning from vget.)

26 years agoSuggested by: bde
jb [Tue, 20 Jan 1998 09:09:24 +0000 (09:09 +0000)]
Suggested by: bde
Move sigjmp_buf and jmp_buf structure definitions to machine/setjmp.h
so that i386 can continue to use int as the basic register type and
alpha can use long. Bruce was concerned about possible differing
alignment. I've left the definition of _JBLEN in machine/setjmp.h
even though Bruce's example used the number directly. I don't know if
any other code relies on _JBLEN, so I left it to avoid potential
breakage.

26 years agoConvert to mdoc format. Remove unused vars. Use err(3) and change exit(-1).
charnier [Tue, 20 Jan 1998 07:30:27 +0000 (07:30 +0000)]
Convert to mdoc format. Remove unused vars. Use err(3) and change exit(-1).

26 years agoAdded cross references to mincore(2) and minherit(2).
alex [Tue, 20 Jan 1998 03:59:07 +0000 (03:59 +0000)]
Added cross references to mincore(2) and minherit(2).

Obtained from: OpenBSD

26 years agoAdded #include <sys/types.h> to synopsis.
alex [Tue, 20 Jan 1998 03:52:49 +0000 (03:52 +0000)]
Added #include <sys/types.h> to synopsis.

Obtained from: OpenBSD

26 years agoAdd the PnP string for a Logitech serial mouse model.
yokota [Tue, 20 Jan 1998 03:38:36 +0000 (03:38 +0000)]
Add the PnP string for a Logitech serial mouse model.

26 years ago- Add binary compatibility support for obsolete console/mouse ioctls
yokota [Tue, 20 Jan 1998 03:37:27 +0000 (03:37 +0000)]
- Add binary compatibility support for obsolete console/mouse ioctls
  so that existing programs which were compiled before the introduction
  of the new mouse code and use these ioctls will run unmodified.

Suggested by msmith.

26 years agoupdate hub's master configuraton file
jmb [Tue, 20 Jan 1998 01:48:12 +0000 (01:48 +0000)]
update hub's master configuraton file
limit sendmail to 500 messages per run
keeps the process from growing too large
sort the queue by host, not numeric priority
connecting to a host is too expensive,
once we got the connection, ram all the mail down his throat ;>

26 years agoAlso fix ypbind to compile with ANSI-fied /usr/include/rpcsvc. One of
wpaul [Mon, 19 Jan 1998 23:31:38 +0000 (23:31 +0000)]
Also fix ypbind to compile with ANSI-fied /usr/include/rpcsvc. One of
these days I really should rerwite this creature from the ground up.

26 years agoAdd wfd style devices to detection list (we'll have to put them on the
jkh [Mon, 19 Jan 1998 23:28:41 +0000 (23:28 +0000)]
Add wfd style devices to detection list (we'll have to put them on the
boot floppy too, of course - hope I still have enough inodes! :)

26 years agoMake these programs compile and work correctly now that /usr/include/rpcsvc
wpaul [Mon, 19 Jan 1998 23:13:38 +0000 (23:13 +0000)]
Make these programs compile and work correctly now that /usr/include/rpcsvc
has been ANSI-fied.

Pointed out by: bde

26 years agodlopen(..., RTLD_NOW) rather than RTLD_LAZY.
brian [Mon, 19 Jan 1998 22:59:57 +0000 (22:59 +0000)]
dlopen(..., RTLD_NOW) rather than RTLD_LAZY.
Pointed out by: John Polstra <jdp@polstra.com>

26 years agosys/types.h -> sys/param.h
brian [Mon, 19 Jan 1998 22:34:20 +0000 (22:34 +0000)]
sys/types.h -> sys/param.h
Pointed out by: Niklas Hallqvist <niklas@petra.appli.se>

26 years agomount(8) only uses realpath() for the mountpoint, so don't look up the
bde [Mon, 19 Jan 1998 18:48:36 +0000 (18:48 +0000)]
mount(8) only uses realpath() for the mountpoint, so don't look up the
real path here for the mount device (or path).  This fixes difficulties
unmounting devices that are actually symlinks to real devices.

Also, print the original path instead of the real path in early error
messages.  nfs path handling and later error messages may still be wrong,
probably only in silly cases where the original path is both a symlink
and a remote path.

PR: 5208

26 years agoNondestructive attempts to get simple locks when SL_DEBUG is defined.
tegge [Mon, 19 Jan 1998 17:33:08 +0000 (17:33 +0000)]
Nondestructive attempts to get simple locks when SL_DEBUG is defined.

26 years agoInvoke rpcgen with -C so that headers are created with ANSI prototypes.
wpaul [Mon, 19 Jan 1998 17:19:09 +0000 (17:19 +0000)]
Invoke rpcgen with -C so that headers are created with ANSI prototypes.

Note: this may cause some problems in a few cases. With very old versions
of rpcgen, if you defined a procedure called foo, then rpcen would create
client stubs for function foo_1() and server stubs _also_ with function
foo_1(). This only worked because of the lack of ANSI prototypes: the
client side stub takes different arguments than the server side stub.
(The client side takes a CLIENT * handle, while the server side wants
an svc_req *.)

To fix this conflict, rpcgen in ANSI mode generates foo_1() for the client
and foo_1_svc() for the server. RPC server code that depends on the old
behavior might break because of this. (Fixing it is just a matter of
adding the _svc suffix onto the server procedure names.)

26 years agoDon't create superblocks with size larger than SBSIZE (8192). The
bde [Mon, 19 Jan 1998 16:55:26 +0000 (16:55 +0000)]
Don't create superblocks with size larger than SBSIZE (8192).  The
size was rounded up to a multiple of the fragment size, but this
gave invalid file systems when the fragment size was > SBSIZE (fsck
aborts early on them).  Now a fragment size of 32768 seems to work
(too-simple tests with fsck and iozone worked).

26 years agoGuard against a block size of 0 in the label. When the first
bde [Mon, 19 Jan 1998 16:28:29 +0000 (16:28 +0000)]
Guard against a block size of 0 in the label.  When the first
superblock is invalid, fsck looks at the label to help guess where
the next superblock should be.  If the partition type is 4.2BSD,
fsck assumed that the block size was valid and divided by it, so
it dumped core if the size was 0.

Initialization of the label was broken almost 3 years ago in rev.1.9
of newfs/newfs.c.  Newfs does not change the label at all, so there
is no problem (except the breakage of the automatic search for
backup superblocks) unless something else sets the partition type
to 4.2BSD.  However, it is too easy to set partition types to
4.2.BSD by copying an old label or by using a disktab entry to
create the label.

PR: 2537

26 years agoAdd a missing \ the previous committer omitted from the crunch target.
jkh [Mon, 19 Jan 1998 16:10:51 +0000 (16:10 +0000)]
Add a missing \ the previous committer omitted from the crunch target.

26 years agoFixed accesses to addresses between VM_MAXUSER_ADDRESS (normally
bde [Mon, 19 Jan 1998 15:27:56 +0000 (15:27 +0000)]
Fixed accesses to addresses between VM_MAXUSER_ADDRESS (normally
0xefbfe000) and kernel_start (normally 0xf0100000).

Things are unnecessarily (?) difficult because procfs is used to
access user addresses in the live-kernel case although we must have
access to /dev/mem to work at all, and whatever works for the
dead-kernel case should work in all cases (modulo volatility of
live kernel variables).  We used the wrong range [0, kernel_start)
for user addresses.  Procfs should only work up to VM_MAXUSER_ADDRESS,
but it bogusly works for reads up to the address 2 pages higher
(the user area, including the kernel stack, is mapped to where the
user area used to be (WTUAUTB)).  Procfs can not work at all for
addresses between WTUAUTB and kernel_start.

Now we use procfs only to access addresses up to VM_MAXUSER_ADDRESS.
Higher addresses are translated normally using kvtophys(), so the
user ptd is used for addresses below the real kernel start (0xf0000000;
see INKERNEL()) and nothing is found WTUAUTB.

Strange accesses that cross the user-kernel boundary are now handled,
but such ranges are currently always errors because they necessarily
overlap the hole WTUAUTB.

Short reads are still not handled.

26 years agoRemoved `kstack' and associated mistranslations in kvtophys().
bde [Mon, 19 Jan 1998 14:27:41 +0000 (14:27 +0000)]
Removed `kstack' and associated mistranslations in kvtophys().
Correct translations would have been null.  However, kstack was
the top of the kernel stack instead of the base of the kernel stack
like it was when the kernel exported it, so the area above the
kernel stack was mistranslated and the kernel stack was not
translated.  This bug was depended on to compensate for the wrong
value of kstack - to read the pcb, instead of just using the address
of the pcb, we used the mistranslated address of kstack, which
happened to be the same (curpcb = kstack - 0x2000).

This area is simpler than it used to be now that the kernel stack
address is per-process.  The code still seems to be more complicated
than necessary - the `found_pcb == 0' case seems to be unused.

26 years agoSet p_retval for the correct process in getpriority(). This fixes
bde [Mon, 19 Jan 1998 12:39:00 +0000 (12:39 +0000)]
Set p_retval for the correct process in getpriority().  This fixes
a null pointer panic when the pointer for the incorrect process is
NULL.  getpriority() was broken in rev.1.27.  Rev.1.28 broke the
warning instead of fixing the problem.

PR: 5495

26 years agoThe removal of a page from the free queue in vm_page_zero_idle was
tegge [Mon, 19 Jan 1998 04:16:16 +0000 (04:16 +0000)]
The removal of a page from the free queue in vm_page_zero_idle was
imcomplete. Also set m->queue, in order to prevent vm_page_select_free
from selecting the page being zeroed.

26 years agoReimplement proxy-arp address stuff using sysctl().
brian [Mon, 19 Jan 1998 02:59:34 +0000 (02:59 +0000)]
Reimplement proxy-arp address stuff using sysctl().

26 years agoAdd an example for busy mail servers, commented out.
wosch [Sun, 18 Jan 1998 22:28:06 +0000 (22:28 +0000)]
Add an example for busy mail servers, commented out.

26 years agoAllow "set vj" for changing the default number of slots
brian [Sun, 18 Jan 1998 20:49:22 +0000 (20:49 +0000)]
Allow "set vj" for changing the default number of slots
and whether slot compression is requested.
Don't show current values with "show ipcp" if IPCP isn't
"opened".

26 years agoDon't throw away bootp reply packets that are shorter than our extended-length
tegge [Sun, 18 Jan 1998 18:53:46 +0000 (18:53 +0000)]
Don't throw away bootp reply packets that are shorter than our extended-length
bootp query packet.
PR: 5512

26 years agoIncrease the minimum bootp reply packet size from 16 (bogus) to 300 (correct).
tegge [Sun, 18 Jan 1998 18:46:20 +0000 (18:46 +0000)]
Increase the minimum bootp reply packet size from 16 (bogus) to 300 (correct).

26 years agoSync with sys/i386/isa/wd.c revision 1.147.
kato [Sun, 18 Jan 1998 15:10:03 +0000 (15:10 +0000)]
Sync with sys/i386/isa/wd.c revision 1.147.

26 years agoSync with sys/i386/conf/files.i386 revision 1.186.
kato [Sun, 18 Jan 1998 15:09:34 +0000 (15:09 +0000)]
Sync with sys/i386/conf/files.i386 revision 1.186.

26 years agoSync with sys/i386/conf/files.i386 revision 1.12.
kato [Sun, 18 Jan 1998 15:09:05 +0000 (15:09 +0000)]
Sync with sys/i386/conf/files.i386 revision 1.12.

26 years agoSync with sys/i386/conf/majors.i386 revision 1.24.
kato [Sun, 18 Jan 1998 15:08:34 +0000 (15:08 +0000)]
Sync with sys/i386/conf/majors.i386 revision 1.24.

26 years agoFixed endless loop for `p/x *(int *)0xf0000000'. kvm_uread() in
bde [Sun, 18 Jan 1998 13:18:55 +0000 (13:18 +0000)]
Fixed endless loop for `p/x *(int *)0xf0000000'.  kvm_uread() in
gdb was cloned from the buggy version of kvm_uread() in libkvm and
had the same bugs.  It looped endlessly on EOF and checked errno
without setting it in the lseek() error check.  The first bug caused
gdb to loop endlessly for reads from addresses between the end of
the user area and the start of the kernel text.  kvm_uread() should
not be used for addresses beyond the end of the user area, but is
due to bugs elsewhere.

26 years agoDon't override FRAME_CHAIN(). If the current frame is valid, then
bde [Sun, 18 Jan 1998 12:35:48 +0000 (12:35 +0000)]
Don't override FRAME_CHAIN().  If the current frame is valid, then
the previous frame is in the usual place even for traps, interrupts
and syscalls in the kernel, because the assembly language stubs
don't change the frame pointer.  The previous frame is just not for
the calling function.  We may as well depend on this as on magic to
determine the trap frame address.  The magic is in FRAME_SAVED_PC()
which elides the correct number of stubs (1) to go back to a pc that
matches the previous frame.

Removing fbsd_kern_frame_chain() fixes bugs in it.  Xsyscall was
misspelled as _Xsyscall (gdb removes one leading underscore), so
the tf_syscall frame type was never found.  This was harmless
because tf_normal works in all cases in fbsd_kern_frame_chain()
and Xsyscall is spelled correctly in fbsd_kern_frame_saved_pc()
where it matters.  There were style bugs on almost every line,
starting with a primary indent of 7.

26 years agoDo not run by default. Can grow your root mailbox up to 2Mb per day
ache [Sun, 18 Jan 1998 12:06:40 +0000 (12:06 +0000)]
Do not run by default. Can grow your root mailbox up to 2Mb per day
since SPAMmers like to retry often

26 years agoFixed sloppy definitions of SIGTRAMP_START and SIGTRAMP_END. The old
bde [Sun, 18 Jan 1998 11:51:48 +0000 (11:51 +0000)]
Fixed sloppy definitions of SIGTRAMP_START and SIGTRAMP_END.  The old
range was a little too large.

26 years agoWorkarround for locking violation while recycling vnode which union fs
kato [Sun, 18 Jan 1998 08:17:48 +0000 (08:17 +0000)]
Workarround for locking violation while recycling vnode which union fs
used in freelist.

26 years agoImprove and revise fixes for locking violation.
kato [Sun, 18 Jan 1998 07:56:41 +0000 (07:56 +0000)]
Improve and revise fixes for locking violation.

Obtained from: NetBSD/pc98

26 years agoPass the system name to dmesg. Rev.1.7 only works when the symbols in
bde [Sat, 17 Jan 1998 17:07:53 +0000 (17:07 +0000)]
Pass the system name to dmesg.  Rev.1.7 only works when the symbols in
/kernel aren't too different form those in the kernel being debugged.

26 years agoVVMIO -> VOBJBUF as in pstat.c.
bde [Sat, 17 Jan 1998 16:58:55 +0000 (16:58 +0000)]
VVMIO -> VOBJBUF as in pstat.c.

26 years agoStarted getting rid of the compatibility cruft for the Lite1 mount()
bde [Sat, 17 Jan 1998 16:45:03 +0000 (16:45 +0000)]
Started getting rid of the compatibility cruft for the Lite1 mount()
and the pre-Lite2 vfsconf interfaces.

For quot, just back out revs. 1.1 and 1.2 and change MNT_FFS to
"ufs", so that vfsconf isn't used at all.  Revs. 1.1 and 1.2 were
hacks to get around f_fstypename not being in `struct statfs' in
Lite1.

26 years agoStarted getting rid of the compatibility cruft for the Lite1 mount()
bde [Sat, 17 Jan 1998 16:32:14 +0000 (16:32 +0000)]
Started getting rid of the compatibility cruft for the Lite1 mount()
and the pre-Lite2 vfsconf interfaces.

For getvfsent.c, just define _OLD_VFSCONF.  This will give the
current default macro hacks in <sys/mount.h> when the default is
reversed.  This is an intermediate step.

26 years agoStarted getting rid of the compatibility cruft for the Lite1 mount()
bde [Sat, 17 Jan 1998 16:24:27 +0000 (16:24 +0000)]
Started getting rid of the compatibility cruft for the Lite1 mount()
and the pre-Lite2 vfsconf interfaces.

For lsvfs, use the new interface for getvfsbyname(), and use the
old interface for getvfsent() explicitly instead of depending on
macro hacks in <sys/mount.h>.  This is an intermediate step.

26 years agoDon't install the tcl implementation headers. The tcl distribution
bde [Sat, 17 Jan 1998 15:52:32 +0000 (15:52 +0000)]
Don't install the tcl implementation headers.  The tcl distribution
makefile doesn't install them, and they couldn't be used without
lots of undocumented -I's in CFLAGS.  tcl.h is still installed in
/usr/include/tcl/.  Note that rev.1.24 of tcl_bmake/mkMakefile.sh
broke all the section 3 tcl man pages by putting it there instead
of in /usr/include.

26 years agoUpdated shlib version to 80.4 in the correct place.
bde [Sat, 17 Jan 1998 15:49:55 +0000 (15:49 +0000)]
Updated shlib version to 80.4 in the correct place.

Regenerated libtcl/Makefile.  This adds -DHAVE_ST_BLKSIZE=1 to CFLAGS.
It was broken by direct editing in rev.1.25.

26 years agoRemove the necessity of -ldes and -lalias etc from release/Makefile.
brian [Sat, 17 Jan 1998 14:21:21 +0000 (14:21 +0000)]
Remove the necessity of -ldes and -lalias etc from release/Makefile.
When building a release, RELEASE_CRUNCH is defined for a `make' of
the objects required by the crunch of each program.  The object list
is still obtained in the same way, so you must make sure that all
objects are built (empty if necessary) by this make.  ppp/Makefile
provides an example.

Reviewed by: jkh

26 years agoFixed syntax error in normally-unconfigured code in previous commit.
bde [Sat, 17 Jan 1998 13:39:29 +0000 (13:39 +0000)]
Fixed syntax error in normally-unconfigured code in previous commit.

26 years agoMF22: fix to install script handling.
jkh [Sat, 17 Jan 1998 12:26:19 +0000 (12:26 +0000)]
MF22: fix to install script handling.

26 years agoTie up some loose ends in vnode/object management. Remove an unneeded
dyson [Sat, 17 Jan 1998 09:17:02 +0000 (09:17 +0000)]
Tie up some loose ends in vnode/object management.  Remove an unneeded
config option in pmap.  Fix a problem with faulting in pages.  Clean-up
some loose ends in swap pager memory management.

The system should be much more stable, but all subtile bugs aren't fixed yet.

26 years agoI wrote a poor, but better-than-nothing manual page for the wfd subdriver.
pst [Sat, 17 Jan 1998 05:15:01 +0000 (05:15 +0000)]
I wrote a poor, but better-than-nothing manual page for the wfd subdriver.
Obtained from:  od.4

26 years agoFix misleading comment about major #
pst [Fri, 16 Jan 1998 22:43:33 +0000 (22:43 +0000)]
Fix misleading comment about major #

26 years agoBring in IDE ATAPI floppy support.
pst [Fri, 16 Jan 1998 22:13:07 +0000 (22:13 +0000)]
Bring in IDE ATAPI floppy support.
This is Junichi's v1.0 driver.

NOTE: Major device numbers have been changed to avoid conflict with other
      FreeBSD 3.0 devices.  The new numbers should be considered "official."
      This driver is still considered "beta" quality, although we have been
      playing with it.  Please submit bugs to junichi and myself.

Submitted by: junichi@astec.co.jp

26 years agoRemoved obsolete man pages.
bde [Fri, 16 Jan 1998 18:56:34 +0000 (18:56 +0000)]
Removed obsolete man pages.

26 years agoRemoved references to the man pages for the obsolete interfaces
bde [Fri, 16 Jan 1998 18:49:42 +0000 (18:49 +0000)]
Removed references to the man pages for the obsolete interfaces
VOP_BLKATOFF, VOP_SEEK, VOP_SELECT, VOP_TRUNCATE, VOP_UPDATE,
VOP_VALLOC and VOP_VFREE.

26 years agoFixed synopsis again. Too much was blown away by splatting a NetBSD-
bde [Fri, 16 Jan 1998 18:45:52 +0000 (18:45 +0000)]
Fixed synopsis again.  Too much was blown away by splatting a NetBSD-
based version over the top of the FreeBSD version in rev.1.6.

26 years agoFixed bitrot. The type of the "level" is intrmask_t, not int.
bde [Fri, 16 Jan 1998 18:14:34 +0000 (18:14 +0000)]
Fixed bitrot.  The type of the "level" is intrmask_t, not int.

26 years agoFixed prototype for tsleep() (const poisoning).
bde [Fri, 16 Jan 1998 18:12:57 +0000 (18:12 +0000)]
Fixed prototype for tsleep() (const poisoning).

26 years agoFixed a missing #include in the synopsis.
bde [Fri, 16 Jan 1998 18:06:23 +0000 (18:06 +0000)]
Fixed a missing #include in the synopsis.

26 years agoInclude <machine/pc/display.h>, not <i386/include/pc/display.h>. This is
bde [Fri, 16 Jan 1998 17:58:55 +0000 (17:58 +0000)]
Include <machine/pc/display.h>, not <i386/include/pc/display.h>.  This is
not quite correct, because the Makefiles in lkm/syscons don't set up the
`machine' symlink, but other, more volatile headers in <machine> are
already used.

26 years agoRemoved most unused includes of <net/if_var.h> outside the kernel.
bde [Fri, 16 Jan 1998 17:38:56 +0000 (17:38 +0000)]
Removed most unused includes of <net/if_var.h> outside the kernel.

26 years agoFixed bitrot in the prototype for logwtmp().
bde [Fri, 16 Jan 1998 16:53:30 +0000 (16:53 +0000)]
Fixed bitrot in the prototype for logwtmp().

26 years agoFixed bugs in the conversion of kvm to to use procfs in rev.1.3. All
bde [Fri, 16 Jan 1998 16:45:05 +0000 (16:45 +0000)]
Fixed bugs in the conversion of kvm to to use procfs in rev.1.3.  All
are in kvm_uread():
- the setting of errno before checking it in the lseek() was lost.
- EOF handling was lost.  kvm_uread() retried forever on EOF.  EOF is
  not really an error, but report it one as in rev.1.2.
- reporting of errno after a read error was lost.

Fixed style bugs in rev.1.3 and rev.1.12.

Not fixed: errno is not reported after lseek() failures.

26 years agoAdapt to new mount(2) argument syntax.
jkh [Fri, 16 Jan 1998 15:07:55 +0000 (15:07 +0000)]
Adapt to new mount(2) argument syntax.
Submitted by: bde

26 years agoFixed wrong prototype for history().
bde [Fri, 16 Jan 1998 14:33:47 +0000 (14:33 +0000)]
Fixed wrong prototype for history().
Don't (mis)use .Fd outside of the synopsis.

26 years agoFixed missing #include in synopsis.
bde [Fri, 16 Jan 1998 13:41:28 +0000 (13:41 +0000)]
Fixed missing #include in synopsis.

26 years agoFixed wrong prototype for clock_getres().
bde [Fri, 16 Jan 1998 13:39:49 +0000 (13:39 +0000)]
Fixed wrong prototype for clock_getres().

26 years agoFixed wrong prototypes. Most of the prototypes had missing return types,
bde [Fri, 16 Jan 1998 13:33:09 +0000 (13:33 +0000)]
Fixed wrong prototypes.  Most of the prototypes had missing return types,
or missing const's or `short *' instead of `[ug]id_t *' in argument types.

26 years agoFixed missing return type in a prototype.
bde [Fri, 16 Jan 1998 13:10:18 +0000 (13:10 +0000)]
Fixed missing return type in a prototype.

26 years agoFixed a missing #include in the synopsis.
bde [Fri, 16 Jan 1998 13:02:58 +0000 (13:02 +0000)]
Fixed a missing #include in the synopsis.
Fixed some wrong prototypes.
Fixed a misspelled function name.

The owner of this file should add a copyright and an Id.

26 years agoAdded prototypes for functions that were documented in libalias.3
bde [Fri, 16 Jan 1998 12:56:07 +0000 (12:56 +0000)]
Added prototypes for functions that were documented in libalias.3
but not prototyped here.

26 years agoBenign changes to support Justin's CAM code.
jkh [Fri, 16 Jan 1998 12:53:04 +0000 (12:53 +0000)]
Benign changes to support Justin's CAM code.

26 years agoNon-intrusive changes to support Justin's CAM stuff.
jkh [Fri, 16 Jan 1998 12:50:36 +0000 (12:50 +0000)]
Non-intrusive changes to support Justin's CAM stuff.

26 years agoAdded missing declaration of signanosleep().
bde [Fri, 16 Jan 1998 12:44:20 +0000 (12:44 +0000)]
Added missing declaration of signanosleep().

26 years agoFixed pedantic syntax error (trailing comma in enum).
bde [Fri, 16 Jan 1998 12:39:20 +0000 (12:39 +0000)]
Fixed pedantic syntax error (trailing comma in enum).

26 years agoRemoved `#pragma ident' (for a FreeBSD Id) so that the K&R support
bde [Fri, 16 Jan 1998 12:28:09 +0000 (12:28 +0000)]
Removed `#pragma ident' (for a FreeBSD Id) so that the K&R support
isn't gratuitously broken.  This also prevents ANSI compilers from
recognizing the pragma as a request to run /usr/games/hack...
FreeBSD Ids should be in comments or rcsids, not in `#pragma ident's
(which are equivalent to comments when compiled by gcc), and the
only FreeBSD change in this file seems to be adding the Id.

26 years agoCommented out `#pragma ident' (for a Sun id) so that the K&R support
bde [Fri, 16 Jan 1998 12:20:41 +0000 (12:20 +0000)]
Commented out `#pragma ident' (for a Sun id) so that the K&R support
isn't gratuitously broken.  This also prevents ANSI compilers from
recognizing the pragma as a request to run /usr/games/hack...

26 years agoRemoved redundant declarations. rpcgen generates `typedef enum foo;'
bde [Fri, 16 Jan 1998 12:17:55 +0000 (12:17 +0000)]
Removed redundant declarations.  rpcgen generates `typedef enum foo;'
from `enum foo;'.

26 years agoMade this header self-sufficent. It had unnecessary dependencies on
bde [Fri, 16 Jan 1998 12:10:13 +0000 (12:10 +0000)]
Made this header self-sufficent.  It had unnecessary dependencies on
<sys/cdefs.h>, <sys/types.h>, <rpcsvc/yp_prot.h>, and on the latter's
prerequisites (5 more).

26 years agoFixed a missing dependency.
bde [Fri, 16 Jan 1998 11:51:19 +0000 (11:51 +0000)]
Fixed a missing dependency.

26 years agoNull change to attach the correct log message for the previous commit:
bde [Fri, 16 Jan 1998 11:47:51 +0000 (11:47 +0000)]
Null change to attach the correct log message for the previous commit:

Added some forward struct declarations so that this file is less
self-insufficent.
Fixed an incomplete prototype.
Fixed some misformattings.

26 years agoAdded a forward struct declaration so that this file is less
bde [Fri, 16 Jan 1998 11:46:05 +0000 (11:46 +0000)]
Added a forward struct declaration so that this file is less
self-insufficent.

26 years agoCommented out `#pragma ident' so that the K&R support isn't gratuitously
bde [Fri, 16 Jan 1998 11:37:54 +0000 (11:37 +0000)]
Commented out `#pragma ident' so that the K&R support isn't gratuitously
broken.  This also prevents ANSI compilers from recognizing the pragma
as a request to run /usr/games/hack...

26 years agoSynchronize with NetBSD/pc98 based on NetBSD/i386 version 1.3G patch
kato [Fri, 16 Jan 1998 11:32:13 +0000 (11:32 +0000)]
Synchronize with NetBSD/pc98 based on NetBSD/i386 version 1.3G patch
level 117.

Reviewed by: Naofumi Honda <honda@Kururu.math.sci.hokudai.ac.jp>

26 years agoFixed bugs introduced when files were synchronized with
kato [Fri, 16 Jan 1998 11:20:22 +0000 (11:20 +0000)]
Fixed bugs introduced when files were synchronized with
sys/i386/isa/mse.c and sio.c.

Submitted by: Takahashi Yoshihiro <nyan@wyvern.cc.kogakuin.ac.jp>

26 years agoTranslated to `C' - commented out #ident.
bde [Fri, 16 Jan 1998 10:31:46 +0000 (10:31 +0000)]
Translated to `C' - commented out #ident.

26 years agoFixed long lines in install rule. Fixed other bogons while I was here:
bde [Fri, 16 Jan 1998 09:43:47 +0000 (09:43 +0000)]
Fixed long lines in install rule.  Fixed other bogons while I was here:
- removed vestigal rule for a Lite1 `.0' manpage.
- don't duplicate defaults in the `all' dependencies.
- removed bogus dependency of `setup' on monsters.asc.
- compile `setup' with the usual CFLAGS.
- create monster files at build time, not at install time.
- don't be so repetitive in the install rule.
- fixed too-global replacement of `install' by ${INSTALL}.

Long lines made disgustingly longer by: previous commit.

26 years agoRemoved redundant declarations. Keymap functions are declared in
bde [Fri, 16 Jan 1998 09:12:05 +0000 (09:12 +0000)]
Removed redundant declarations.  Keymap functions are declared in
<readline/keymaps.h> which is included here.

26 years agoUse the system __CONCAT() instead of /**/ for concatenation in the
bde [Fri, 16 Jan 1998 08:59:21 +0000 (08:59 +0000)]
Use the system __CONCAT() instead of /**/ for concatenation in the
non-ANSI case.  The system version is normally also /**/, but may
be different to support K&R compilers other than gcc -traditional.

26 years agoFixed copying of N_TXTADDR() from the system version in the previous
bde [Fri, 16 Jan 1998 08:46:50 +0000 (08:46 +0000)]
Fixed copying of N_TXTADDR() from the system version in the previous
revision.  N_TXTADDR() was 0x20 too high in the usual case.

Reported by: luoqi@watermarkgroup.com (Luoqi Chen)

26 years agoFixed handling of short writes. Previously, we stopped copying and
bde [Fri, 16 Jan 1998 07:37:05 +0000 (07:37 +0000)]
Fixed handling of short writes.  Previously, we stopped copying and
printed a bogus warning with a stale errno if write() returns a short
count.  Now we continue copying.  We still print a bogus warning if
write() returns an "impossible" short count of 0.

26 years agoAdd rcsid. Remove unused vars. Use err(3) and change exit(-1).
charnier [Fri, 16 Jan 1998 07:20:34 +0000 (07:20 +0000)]
Add rcsid. Remove unused vars. Use err(3) and change exit(-1).

26 years agoAdd rcsid. Change exit(-1). Use err() instead of warn()/exit().
charnier [Fri, 16 Jan 1998 07:17:45 +0000 (07:17 +0000)]
Add rcsid. Change exit(-1). Use err() instead of warn()/exit().

26 years agoFixed some spelling errors.
bde [Fri, 16 Jan 1998 06:31:23 +0000 (06:31 +0000)]
Fixed some spelling errors.

26 years agoFix wrong ---/*** names produced using -L diff option
ache [Fri, 16 Jan 1998 06:00:36 +0000 (06:00 +0000)]
Fix wrong ---/*** names produced using -L diff option
CVS gurus from info-cvs agrees that previous Index: usage is bogus

26 years agoPR: 5489
danny [Thu, 15 Jan 1998 23:28:18 +0000 (23:28 +0000)]
PR: 5489
Submitted by: Steve G. Kargl <kargl@troutmask.apl.washington.edu>
Repair corrupted text.

26 years agoupdate hub.mc to match what we are currently using:
jmb [Thu, 15 Jan 1998 17:39:50 +0000 (17:39 +0000)]
update hub.mc to match what we are currently using:

both /etc/aliases and /etc/majordomo.aliases
sendmail.cw is *not* optional
command and connection timeouts
common hoststat directory for all queues
add fakenames database
clean up spam filter rules
limit the number of message processed each queue run
dont need to check Vixie's RBL in both check_relay and check_mail

26 years agoSync with sys/i386/conf/files.i386 revision 1.185.
kato [Thu, 15 Jan 1998 15:09:00 +0000 (15:09 +0000)]
Sync with sys/i386/conf/files.i386 revision 1.185.