Brian Feldman [Wed, 18 Sep 2002 16:25:59 +0000 (16:25 +0000)]
Update the usage string in the limits(1) manpage to include -b for
sbsize. Also, correct the format string in getopt(3) usage to reflect
that -b takes an argument, and correct another case of RLIMIT_SBSIZE
having been forgotten.
Juli Mallett [Wed, 18 Sep 2002 16:13:03 +0000 (16:13 +0000)]
Make the DEBUGF() macro portable by (ugh) adding a Debug() function, which
is merely printf() but to stderr. This takes care of the caveat which lead
to the use of a vararg macro -- getting everything to stderr.
Jacques Vidrine [Wed, 18 Sep 2002 14:17:14 +0000 (14:17 +0000)]
These RFCs and internet-drafts are not really needed in the base
system, and I've not been importing them lately. cvs rm them now
so they can be cleaned out of the attic later.
Tim J. Robbins [Wed, 18 Sep 2002 12:17:28 +0000 (12:17 +0000)]
Optimise the common case where no special encoding is in use (LC_CTYPE is "C"
or "POSIX", other European locales). Use __sgetc() and __sputc() where
possible to avoid a wasteful lock and unlock for each byte and to avoid
function call overhead.
Boris Popov [Wed, 18 Sep 2002 09:54:16 +0000 (09:54 +0000)]
Always open file in the DENYNONE mode and let the server to decide what is
good for this file.
This should allow read only access to file which is already opened on server.
Don Lewis [Wed, 18 Sep 2002 09:34:01 +0000 (09:34 +0000)]
Style and markup changes only.
Capitalize the first letter of the descriptions for the entries in the ERRORS
section if they are complete sentences and end the sentences with periods.
Fix implementation of rc variables $amd_flags and $amd_map_program in
rcNG. The $amd_flags variable was already taken over into $rc_flags
by run_rc_command() when amd_precmd() is executed, so changing
$amd_flags there no longer effects the actual execution of amd.
Hence in amd_precmd() the $rc_flags have to be adjusted instead.
Jeff Roberson [Wed, 18 Sep 2002 08:26:30 +0000 (08:26 +0000)]
- Split UMA_ZFLAG_OFFPAGE into UMA_ZFLAG_OFFPAGE and UMA_ZFLAG_HASH.
- Remove all instances of the mallochash.
- Stash the slab pointer in the vm page's object pointer when allocating from
the kmem_obj.
- Use the overloaded object pointer to find slabs for malloced memory.
Mike Barcroft [Wed, 18 Sep 2002 07:51:46 +0000 (07:51 +0000)]
ARIN has annoyingly started using caps when referencing
whois.apnic.net. To properly receive hints for recursive IP searches,
we convert a buffer to lowercase before searching for magic words.
PR: 42834
Submitted by: Andre Albsmeier <andre.albsmeier@mchp.siemens.de> (mostly)
X-MFC-After: re approval
Tim J. Robbins [Wed, 18 Sep 2002 06:11:21 +0000 (06:11 +0000)]
Mark mbmb(), mbrune(), and mbrrune() as deprecated functions. We want to
get applications to move to the ISO C interfaces as well as have the
freedom to replace the rune interfaces with ones that support stateful
conversions some time in the future.
Tim J. Robbins [Wed, 18 Sep 2002 05:58:11 +0000 (05:58 +0000)]
Reimplement the functionality of fgetrune(), fputrune(), and fungetrune()
here in terms of mbrtowc(), wcrtomb(), and the single-byte I/O functions.
The rune I/O functions are about to become deprecated in favour of the
ones provided by ISO C90 Amd. 1 and C99.
Make obrien happy. Add a bad awk script which emulates as much of
gperf's behavior as we ever actually needed here. This generates
a much-less-efficient keyword recognizer, but it's not like that matters
in this application. Makefile changes coming once this passes the world
test.
Sam Leffler [Tue, 17 Sep 2002 22:39:57 +0000 (22:39 +0000)]
MFC: Elan support;
o Recognize AMD Elan SC520 hostbridge.
o Add initialization code for the AMD Elan sc520 which maps the MMCR
into KVM and sets the i8254 frequency to the correct value.
o MAKEDEV entry for elan-mmcr device
Juli Mallett [Tue, 17 Sep 2002 22:31:26 +0000 (22:31 +0000)]
Move common use of if (DEBUG(FOO)) printf... to DEBUGF(FOO, ...), using
variable length arguments to a macro. Bump version as this makes DEBUG
statements *always* go to stderr rather than sometimes stdout. There are
a few stragglers, which I will take care of as soon as I can. Mostly these
relate to the need-for-death-of some of the remote job code.
Nearby stylistic nits and XXX added/fixed where appropriate.
Mike Barcroft [Tue, 17 Sep 2002 22:22:50 +0000 (22:22 +0000)]
o Add typedefs for mode_t, off_t, pid_t rather than including
<sys/types.h>.
o Use the relatively new visibility primitives for conditionals.
o Make O_SYNC an alias for O_FSYNC.
o Mark the F* names as deprecated.
o Add some comments to note missing POSIX requirements or options.
Matthew Dillon [Tue, 17 Sep 2002 22:21:37 +0000 (22:21 +0000)]
Guido reported an interesting bug where an FTP connection between a
Windows 2000 box and a FreeBSD box could stall. The problem turned out
to be a timestamp reply bug in the W2K TCP stack. FreeBSD sends a
timestamp with the SYN, W2K returns a timestamp of 0 in the SYN+ACK
causing FreeBSD to calculate an insane SRTT and RTT, resulting in
a maximal retransmit timeout (60 seconds). If there is any packet
loss on the connection for the first six or so packets the retransmit
case may be hit (the window will still be too small for fast-retransmit),
causing a 60+ second pause. The W2K box gives up and closes the
connection.
In rare cases when there is no room for ip options ip_insertoptions()
can fail and corrupt a header length. Initialize len and check what
ip_insertoptions() returns.
Reviewed by: archie, silence on -net
MFC after: 5 days
Bruce Evans [Tue, 17 Sep 2002 10:01:02 +0000 (10:01 +0000)]
Make `as' compile before it is axed. It still uses the archaic BSD
interface setbuffer(), and emulates setbuffer() on USG systems using a
#define of setbuffer() in terms of setvbuf(). The #define is correctly
ifdefed in some places but was not correctly ifdefed here -- i.e., BSD
was essentially configured as USG here. This became fatal when <stdio.h>
was de-__P(())ified without testing. This file gets included before
<stdio.h>, so the #define now affects (and breaks)
`setbuffer<left parentheses>' in <stdio.h> where it didn't affect
`setbuffer<whitespace>'.
Peter Wemm [Tue, 17 Sep 2002 09:07:06 +0000 (09:07 +0000)]
The previous commit also removed the addition of userland include files to
the kernel #include path. While this was not ready at the time (sorry
folks!), it is a good thing. I think all the loose ends have been tied up
on at least for i386 (LINT compiles) and alpha.
Peter Wemm [Tue, 17 Sep 2002 07:40:15 +0000 (07:40 +0000)]
Stub out the calls to get_mcontext and set_mcontext which only exist on
i386. This stuff should not be prototyped in MD inludes if the interface
is expected to be MI.
Peter Wemm [Tue, 17 Sep 2002 07:32:36 +0000 (07:32 +0000)]
Add i386 to the list of architectures that libc_r is broken on. This
effectively removes pppctl from the build for now. It only compiles on
alpha now (now ironic).
Peter Wemm [Tue, 17 Sep 2002 07:22:23 +0000 (07:22 +0000)]
It seems I accidently committed a change that removed the userland includes
from the kernel build. This broke linux_genassym on the alpha. For the
kernel, the correct place to get offsetof() is not in /usr/include/stddef.h
but rather <sys/types.h>
Mike Barcroft [Tue, 17 Sep 2002 05:14:09 +0000 (05:14 +0000)]
Include <sys/types.h> directly rather than depending on <sys/fcntl.h>
to include it. This could be avoided by adding the necessary typedefs
here, or by making users of <sys/file.h> include <sys/types.h> first.
Juli Mallett [Tue, 17 Sep 2002 02:52:44 +0000 (02:52 +0000)]
Conglomerate printing of ps_pgtok'd data into a PLONG type. I couldn't think
of a better name, except PINT, but I decided to go with assuming LONG to
be safe, rather than assuming INT.
David E. O'Brien [Tue, 17 Sep 2002 01:54:56 +0000 (01:54 +0000)]
BANG! BANG! BANG! Put these bits out of their misery.
Murdered by members of: a.out.die.die.die
ELF is the 1 true path now. So make good on the src/Makefile threat that
building a.out will not be supported post 4.x. These bits should either
resurface as a port, or a new port using the latest Binutils bits. The
later will not support our SunOS-style shared a.out libs; but we shouldn't
need such support by this point in time.
Peter Wemm [Tue, 17 Sep 2002 01:49:00 +0000 (01:49 +0000)]
Initiate deorbit burn for the i386-only a.out related support. Moves are
under way to move the remnants of the a.out toolchain to ports. As the
comment in src/Makefile said, this stuff is deprecated and one should not
expect this to remain beyond 4.0-REL. It has already lasted WAY beyond
that.
Notable exceptions:
gcc - I have not touched the a.out generation stuff there.
ldd/ldconfig - still have some code to interface with a.out rtld.
old as/ld/etc - I have not removed these yet, pending their move to ports.
some includes - necessary for ldd/ldconfig for now.
Bill Fenner [Mon, 16 Sep 2002 21:45:37 +0000 (21:45 +0000)]
Fix documentation of clnt_control()'s CL{GET|SET}_{VERS|XID} to
reflect that they actually require a u_int32_t *, which is not
necessarily the same as an unsigned long *.