]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
22 years agoFully exterminate nfsd_srvargs and nfsd_cargs. They were either unused
Peter Wemm [Wed, 24 Jul 2002 22:27:35 +0000 (22:27 +0000)]
Fully exterminate nfsd_srvargs and nfsd_cargs.  They were either unused
or giant NOP's.  There was a credential in srvargs that was giving
rwatson some heartburn. :-)

22 years agoRemove WARNS and WFORMAT overrides.
Mark Peek [Wed, 24 Jul 2002 22:26:44 +0000 (22:26 +0000)]
Remove WARNS and WFORMAT overrides.

22 years agoIf NO_MAILWRAPPER is set, and sendmail is still enabled, then install
Peter Wemm [Wed, 24 Jul 2002 22:17:22 +0000 (22:17 +0000)]
If NO_MAILWRAPPER is set, and sendmail is still enabled, then install
a direct symlink to sendmail.

22 years agoenable fdisk for ia64 as well as i386
Peter Wemm [Wed, 24 Jul 2002 22:16:11 +0000 (22:16 +0000)]
enable fdisk for ia64 as well as i386

22 years agoOnly install the mbr code on i386.. not ia64. This changes it from
Peter Wemm [Wed, 24 Jul 2002 22:15:51 +0000 (22:15 +0000)]
Only install the mbr code on i386.. not ia64.  This changes it from
#ifndef __alpha__ to #ifdef __i386__

22 years agoTurn nfsiod back on. It is basically now just a sysctl wrapper that also
Peter Wemm [Wed, 24 Jul 2002 22:14:27 +0000 (22:14 +0000)]
Turn nfsiod back on.  It is basically now just a sysctl wrapper that also
arranges a kldload of the nfs client file system code if necessary.

22 years agoWARNS clean
Peter Wemm [Wed, 24 Jul 2002 22:13:08 +0000 (22:13 +0000)]
WARNS clean

22 years agoFake a resume method for this chip via calling detach/attach. This lets
Doug Ambrisko [Wed, 24 Jul 2002 21:27:22 +0000 (21:27 +0000)]
Fake a resume method for this chip via calling detach/attach.  This lets
ThinkPads using this chip to have sound after a suspend/resume sequence
without resorting to kldunload/kldload.

MFC after: 1 week

22 years agoCopy the right dhclient-script(8).
Ruslan Ermilov [Wed, 24 Jul 2002 20:37:21 +0000 (20:37 +0000)]
Copy the right dhclient-script(8).

22 years agoWhen single threading a multithreaded program, awaken the
Julian Elischer [Wed, 24 Jul 2002 19:50:08 +0000 (19:50 +0000)]
When single threading a multithreaded program, awaken the
'single threading thread' when the last other thread suspends.
I had this code in there before but it seems to have been
accidentally deleted somewhere along the way.  This would only affect
multithreaded processes.

Reviewed by: David Xu <bsddiy@yahoo.com>

22 years agoFix a stupid bug where I wasn't initializing the names
Maxime Henrion [Wed, 24 Jul 2002 19:50:00 +0000 (19:50 +0000)]
Fix a stupid bug where I wasn't initializing the names
of 0-length mount options.

22 years ago o Merge vm_fault_wire() and vm_fault_user_wire() by adding a new parameter,
Alan Cox [Wed, 24 Jul 2002 19:47:56 +0000 (19:47 +0000)]
 o Merge vm_fault_wire() and vm_fault_user_wire() by adding a new parameter,
   user_wire.

22 years agomake sure to set/unset INP_IPV4 according to a value
Hajimu UMEMOTO [Wed, 24 Jul 2002 19:19:53 +0000 (19:19 +0000)]
make sure to set/unset INP_IPV4 according to a value
of IN6P_IPV6_V6ONLY

Reviewed by: Keiichi SHIMA <keiichi@iij.ad.jp>

22 years ago- ntohs() returns unsigned value.
Hajimu UMEMOTO [Wed, 24 Jul 2002 18:58:32 +0000 (18:58 +0000)]
- ntohs() returns unsigned value.
- use strlcpy.
- snprintf can return negative value, so cope with it.
- tweak interface index on interface locals (ff01::/16).
- removed unused macros.
- removed a macro that uses only once (in a trivial context).
- explicitly say goodbye to ENI_xxx.
- constify struct afd.

Obtained from: KAME
MFC after: 1 week

22 years agostyle(9): - Put a space after the 'return' keyword.
Robert Drehmel [Wed, 24 Jul 2002 18:54:59 +0000 (18:54 +0000)]
style(9): - Put a space after the 'return' keyword.
          - Wrap lines longer than 80 characters.

22 years agoUse an ANSI-C function definition rather than a
Robert Drehmel [Wed, 24 Jul 2002 18:39:51 +0000 (18:39 +0000)]
Use an ANSI-C function definition rather than a
K&R one to avoid a warning.

22 years ago - Add a COMPATIBILITY section explaining that the default
Robert Drehmel [Wed, 24 Jul 2002 18:37:22 +0000 (18:37 +0000)]
 - Add a COMPATIBILITY section explaining that the default
   behaviour of obtaining the affected terminal has changed.
 - Refer to the user's terminal instead of the current
   terminal session.
 - Tell the reader what the utility does when it is invoked
   without arguments.

Submitted by: johan

22 years agoDocument upgrade procedure.
Mark Peek [Wed, 24 Jul 2002 16:42:58 +0000 (16:42 +0000)]
Document upgrade procedure.

22 years agoThis commit was generated by cvs2svn to compensate for changes in r100616,
Mark Peek [Wed, 24 Jul 2002 16:23:10 +0000 (16:23 +0000)]
This commit was generated by cvs2svn to compensate for changes in r100616,
which included commits to RCS files with non-trunk default branches.

22 years agoImport of tcsh-6.12.00
Mark Peek [Wed, 24 Jul 2002 16:23:10 +0000 (16:23 +0000)]
Import of tcsh-6.12.00

22 years agoRe-use passive data ports with the SO_REUSEADDR
Yaroslav Tykhiy [Wed, 24 Jul 2002 16:11:34 +0000 (16:11 +0000)]
Re-use passive data ports with the SO_REUSEADDR
socket option to avoid exausting the passive port
space by TIME_WAIT'ing connections.

PR: bin/36955
Submitted by: Maxim Konovalov <maxim@FreeBSD.org>
MFC after: 2 weeks

22 years agoAllow selection of the affected terminal using redirection
Robert Drehmel [Wed, 24 Jul 2002 15:48:17 +0000 (15:48 +0000)]
Allow selection of the affected terminal using redirection
of standard input or standard output, like mesg(1) does.

Suggested by: sheldonh

22 years agoUnder #ifdef DIAGNOSTIC, NULL out componentname pointers if we free the
Robert Watson [Wed, 24 Jul 2002 15:42:22 +0000 (15:42 +0000)]
Under #ifdef DIAGNOSTIC, NULL out componentname pointers if we free the
pnbuf to increase the chances of detecting use of a free'd name buffer
if SAVENAME or SAVESTART wasn't passed in.  Curiously, running with these
changes doesn't panic the kernel, and should.

22 years agoRemove the outdated casts to "char *" from the setsockopt(2),
Yaroslav Tykhiy [Wed, 24 Jul 2002 15:30:53 +0000 (15:30 +0000)]
Remove the outdated casts to "char *" from the setsockopt(2),
write(2), and getipnodebyaddr(3) calls.  Now all the above functions
accept "void *" in that arguments and have prototypes.  Thus, the
casts are useless under the normal circumstances (and would be harmful
if the functions had no prototypes.)

MFC after: 2 weeks

22 years agoMove m_freem() from uipc_mbuf.c to subr_mbuf.c so it can take advantage
Bosko Milekic [Wed, 24 Jul 2002 15:11:23 +0000 (15:11 +0000)]
Move m_freem() from uipc_mbuf.c to subr_mbuf.c so it can take advantage
of the inlines, like its cousin, m_free().  Also, make a small (first
step?) optimisation of m_free() to use the MBP_PERSIST{,ENT} interface
to hold the lock across frees when possible.  The thing is that right
now, we can only do this easily for at most across one mbuf + one
cluster free, as the comment mentions (it also explains why).  Anyway,
some basic tests revealed a 5-10% overall improvement.  Some of the
results can be found here:
http://people.freebsd.org/~bmilekic/code/measure.txt

22 years agoMove whereis(1) relnote to 'userland changes' section.
Johan Karlsson [Wed, 24 Jul 2002 14:51:24 +0000 (14:51 +0000)]
Move whereis(1) relnote to 'userland changes' section.

Approved by: bmah, sheldonh (mentor)

22 years agoClean up the syslog(3) messages on the setsockopt(2) errors:
Yaroslav Tykhiy [Wed, 24 Jul 2002 14:50:17 +0000 (14:50 +0000)]
Clean up the syslog(3) messages on the setsockopt(2) errors:

o Always check a setsockopt(2) return value
o Use a consistent message format
o Don't abort if the failed setsockopt(2) was actually not vital
o Use LOG_WARNING, not LOG_ERR, in non-fatal cases

MFC after: 1 week

22 years agoTeach whereis(1) about games.
Johan Karlsson [Wed, 24 Jul 2002 14:35:29 +0000 (14:35 +0000)]
Teach whereis(1) about games.

Approved by: joerg, sheldonh (mentor)

22 years agoStick a dark comment in about the fact that the NFS server code allocates
Robert Watson [Wed, 24 Jul 2002 14:24:16 +0000 (14:24 +0000)]
Stick a dark comment in about the fact that the NFS server code allocates
a ucred by itself as part of an nfs descriptor, then bzero's the ucred,
fails to initialize the mutex, etc.  This is very bad, but I don't have
time to fix it right now.  nfsd should instead hold a cred pointer,
and the credential should be properly initialized, probably from a
descendent of a kernel process credential.

22 years agoCatch up to rev 1.87 of sys/sys/socketvar.h (sb_cc changed from u_long
Mike Barcroft [Wed, 24 Jul 2002 14:21:41 +0000 (14:21 +0000)]
Catch up to rev 1.87 of sys/sys/socketvar.h (sb_cc changed from u_long
to u_int).

Noticed by: sparc64 tinderbox

22 years agoMove sio's ebus attachment to the MI files section so it is compiled in
John Baldwin [Wed, 24 Jul 2002 12:43:27 +0000 (12:43 +0000)]
Move sio's ebus attachment to the MI files section so it is compiled in
for any machines that use ebus.

22 years agoMove sio_isa.c back to MD files files due to PC98 brain damage.
John Baldwin [Wed, 24 Jul 2002 12:35:52 +0000 (12:35 +0000)]
Move sio_isa.c back to MD files files due to PC98 brain damage.

22 years agoMFi386: revision 1.407 (move the em driver entry to MI file)
Yoshihiro Takahashi [Wed, 24 Jul 2002 12:33:39 +0000 (12:33 +0000)]
MFi386: revision 1.407 (move the em driver entry to MI file)

22 years agoRemove the word IPSec to respect RFCs (for example
Marc Fonvieille [Wed, 24 Jul 2002 10:24:52 +0000 (10:24 +0000)]
Remove the word IPSec to respect RFCs (for example
http://www.ietf.org/rfc/rfc2401.txt), IPsec is the right word and
already in the dictionary.

PR: in part docs/38668
Reviewed by: murray
MFC after: 10 days

22 years agoWhen suspending a thread, update the appropriate (sic) statistic.
Julian Elischer [Wed, 24 Jul 2002 07:29:16 +0000 (07:29 +0000)]
When suspending a thread, update the appropriate (sic) statistic.

22 years agorevert some of the handling of STOP signals in
Julian Elischer [Wed, 24 Jul 2002 07:23:41 +0000 (07:23 +0000)]
revert some of the handling of STOP signals in
issignal(). Let thread_suspend_check() actually do the suspension
at the user boundary.

Submitted by: David Xu <bsddiy@yahoo.com>

22 years agoWiden struct sockbuf's sb_timeo member to int from short. With
John Polstra [Wed, 24 Jul 2002 03:02:43 +0000 (03:02 +0000)]
Widen struct sockbuf's sb_timeo member to int from short.  With
non-default but reasonable values of hz this member overflowed,
breaking NFS over UDP.

Also, as long as I'm plowing up struct sockbuf ... Change certain
members from u_long/long to u_int/int in order to reduce wasted
space on 64-bit machines.  This change was requested by Andrew
Gallatin.

Netstat and systat need to be rebuilt.  I am incrementing
__FreeBSD_version in case any ports need to change.

22 years agoOnly log things net.inet.ip.fw.verbose is set
Luigi Rizzo [Wed, 24 Jul 2002 02:41:19 +0000 (02:41 +0000)]
Only log things net.inet.ip.fw.verbose is set

22 years agoAvoid calling el_resize() from a signal handler, even though libedit
Tim J. Robbins [Wed, 24 Jul 2002 02:06:07 +0000 (02:06 +0000)]
Avoid calling el_resize() from a signal handler, even though libedit
itself does that if you set EL_SIGNAL. Instead, set a flag and check it
before calling el_gets(). This is safer, but slower to respond to changes.

Pointed out by: mp

22 years agoFix warning.
Jake Burkholder [Wed, 24 Jul 2002 01:09:38 +0000 (01:09 +0000)]
Fix warning.

22 years agoAttempt to clarify comment in selrecord.
Alfred Perlstein [Wed, 24 Jul 2002 00:29:22 +0000 (00:29 +0000)]
Attempt to clarify comment in selrecord.

22 years agoMore userland section translation
Sebastien Gioria [Tue, 23 Jul 2002 20:16:09 +0000 (20:16 +0000)]
More userland section translation

22 years agoInstall more man pages - I thought I'd committed this ages ago...
Dag-Erling Smørgrav [Tue, 23 Jul 2002 17:59:46 +0000 (17:59 +0000)]
Install more man pages - I thought I'd committed this ages ago...

22 years agoWhitespace nit.
Dag-Erling Smørgrav [Tue, 23 Jul 2002 17:57:17 +0000 (17:57 +0000)]
Whitespace nit.

22 years agoUpdated release notes: lint 20020719, smbfs-1.4.5, texinfo 4.2.
Bruce A. Mah [Tue, 23 Jul 2002 16:00:19 +0000 (16:00 +0000)]
Updated release notes:  lint 20020719, smbfs-1.4.5, texinfo 4.2.

22 years agoRevert previous delta for cut(1) and calendar(1) as they were not applied
Tom Rhodes [Tue, 23 Jul 2002 15:09:56 +0000 (15:09 +0000)]
Revert previous delta for cut(1) and calendar(1) as they were not applied
correctly.  My fault, forgot that -current was checked out.

22 years agoCatch SIGWINCH in interactive shells and call el_resize() to update
Tim J. Robbins [Tue, 23 Jul 2002 15:05:00 +0000 (15:05 +0000)]
Catch SIGWINCH in interactive shells and call el_resize() to update
libedit's idea of the window size.

22 years agoIntroduce mb_free() to the MBP_PERSIST{,ENT} interface. What this means
Bosko Milekic [Tue, 23 Jul 2002 14:55:33 +0000 (14:55 +0000)]
Introduce mb_free() to the MBP_PERSIST{,ENT} interface.  What this means
is that grouped frees will be done as most often as possible without
dropping the cache lock in between.  So, for the most part, they'll be
done without the lock being dropped.  This is particularly true if you
have something that does a grouped m_getm() or m_getcl() (a cluster and
mbuf at the same time) - most likely getting the buffers from the
same per-CPU cache - and then frees them with m_free{,m}().  Unless
the buffers' underlying buckets were moved, the free will be done without
the lock getting dropped in between.  So far, only m_free() has been
shown how to do this, and m_freem() will shortly follow.

Since I'm here, I also fixed a small (but mostly harmless) type-mismatch
introduced in the last commit.

22 years agoFix the sound driver vchan support to work when hw.snd.maxautovchans
Alexander Kabaev [Tue, 23 Jul 2002 14:50:51 +0000 (14:50 +0000)]
Fix the sound driver vchan support to work when hw.snd.maxautovchans
has been specified through /boot/loader.conf as opposed to setting it
in /etc/sysctl.conf. Only PCMDIR_PLAY channel can be used as a parent
of virtual channel. Do not initialize a new vchan for a given physical
channel if other physical channel already has one created.

PR: 31597
Approved by: obrien (mentor)

22 years agoRemove duplicate __FBSDID() - spotted by trying to build with a non-GNU
Juli Mallett [Tue, 23 Jul 2002 14:39:20 +0000 (14:39 +0000)]
Remove duplicate __FBSDID() - spotted by trying to build with a non-GNU
compiler.

22 years agoMention the new bind(1) sh builtin.
Tim J. Robbins [Tue, 23 Jul 2002 14:38:07 +0000 (14:38 +0000)]
Mention the new bind(1) sh builtin.

22 years agoFix DIOCGMEDIASIZE and DIOCGSECTORSIZE ioctls to work for all
Alexander Kabaev [Tue, 23 Jul 2002 14:30:27 +0000 (14:30 +0000)]
Fix DIOCGMEDIASIZE and DIOCGSECTORSIZE ioctls to work for all
disk devices. This fixes the problem with these ioctls returning
EINVAL for plain slice devices with no disklabel on them.

The patch incorporates improvements and style fixes from BDE.

Reviewed by: bde
Approved by: obrien (mentor)

22 years agoMention SAVENAME and SAVESTART in the namei(9) operation flag description
Robert Watson [Tue, 23 Jul 2002 13:16:36 +0000 (13:16 +0000)]
Mention SAVENAME and SAVESTART in the namei(9) operation flag description
since apparently people were missing that you aren't supposed to access
the name buffer following namei() unless you specify one of these flags.

Pointed out by: green
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs

22 years agoUpdate ucred.9 to talk a bit more about correct use of credentials,
Robert Watson [Tue, 23 Jul 2002 13:09:24 +0000 (13:09 +0000)]
Update ucred.9 to talk a bit more about correct use of credentials,
including documenting that ucreds must not be pulled out of thin air,
when to use td_cred vs. p_ucred, how to avoid race conditions in
credential updates, and why to use p_ucred when targetting a thread
or process in an access control operation involving two processes.

Reviewed by: julian, jhb (earlier revision)

22 years agoTurn on emacs line editing mode by default, as this is what most shells do
Juli Mallett [Tue, 23 Jul 2002 12:28:16 +0000 (12:28 +0000)]
Turn on emacs line editing mode by default, as this is what most shells do
these days, and the average user expects ^A and arrow keys to work, however
if they know nothing of editing modes, they will think sh(1) just sucks.  It
is likely that because of defaults on most systems and with most shells that
anyone who actually wants vi(1) editing mode will have 'set -o vi'.  This
won't affect existing accounts, this way, of course.  Only accounts with
.shrc from new etc/skel will be affected.  This is much better than making
the change in sh(1).

22 years agoCall el_source() to read the contents of .editrc when line editing is enabled.
Tim J. Robbins [Tue, 23 Jul 2002 12:26:34 +0000 (12:26 +0000)]
Call el_source() to read the contents of .editrc when line editing is enabled.

Obtained from: NetBSD
MFC after: 2 weeks

22 years agos/ifdef/ifndef/.
Daniel C. Sobral [Tue, 23 Jul 2002 12:25:34 +0000 (12:25 +0000)]
s/ifdef/ifndef/.

Fenner got this inverted through a misunderstanding between us.

22 years agoAdd a `bind' builtin command, which is simply a wrapper around libedit's
Tim J. Robbins [Tue, 23 Jul 2002 11:50:53 +0000 (11:50 +0000)]
Add a `bind' builtin command, which is simply a wrapper around libedit's
builtin command of the same name. This allows the key bindings for the
shell's line editor to be changed.

MFC after: 2 weeks

22 years agoDependencies are delimited by space not tab.
Ruslan Ermilov [Tue, 23 Jul 2002 09:15:45 +0000 (09:15 +0000)]
Dependencies are delimited by space not tab.

22 years agos/IPSEC/IPsec according to RFCs
Marc Fonvieille [Tue, 23 Jul 2002 08:38:03 +0000 (08:38 +0000)]
s/IPSEC/IPsec according to RFCs

PR: in part docs/38668
Reviewed by: charnier
MFC after: 10 days

22 years agode-count pci
Peter Wemm [Tue, 23 Jul 2002 06:38:47 +0000 (06:38 +0000)]
de-count pci

22 years agode-count pci now that ahc/ahd are taken care of.
Peter Wemm [Tue, 23 Jul 2002 06:36:23 +0000 (06:36 +0000)]
de-count pci now that ahc/ahd are taken care of.

22 years agopci is (finally) no longer a 'count' device. ahc/ahd were the last
Peter Wemm [Tue, 23 Jul 2002 06:33:08 +0000 (06:33 +0000)]
pci is (finally) no longer a 'count' device.  ahc/ahd were the last
holdouts.

22 years agoMinimal tweak to not depend on NPCI, but still maintain portability
Peter Wemm [Tue, 23 Jul 2002 06:31:45 +0000 (06:31 +0000)]
Minimal tweak to not depend on NPCI, but still maintain portability
to 4.x which still does need it (since pci_if.h isn't generated without
pci being present)

22 years agoHandle directories correctly.
Eric Melville [Tue, 23 Jul 2002 04:54:57 +0000 (04:54 +0000)]
Handle directories correctly.

PR: 40801
MFC after: 5 days

22 years ago o Lock page queue accesses by vm_page_dontneed().
Alan Cox [Tue, 23 Jul 2002 04:39:48 +0000 (04:39 +0000)]
 o Lock page queue accesses by vm_page_dontneed().
 o Assert that the page queue lock is held in vm_page_dontneed().

22 years ago- Pass the VM_ALLOC_WIRED flag to vm_page_alloc() in pmap_growkernel() so
Andrew R. Reiter [Tue, 23 Jul 2002 04:18:10 +0000 (04:18 +0000)]
- Pass the VM_ALLOC_WIRED flag to vm_page_alloc() in pmap_growkernel() so
  that we can avoid a call to vm_page_lock_queues().

Approved by: peter

22 years ago o Extend the scope of the page queues lock in vm_pageout_scan()
Alan Cox [Tue, 23 Jul 2002 02:42:25 +0000 (02:42 +0000)]
 o Extend the scope of the page queues lock in vm_pageout_scan()
   to cover the traversal of the cache queue.

22 years ago- Make use of the VM_ALLOC_WIRED flag in the call to vm_page_alloc() in
Andrew R. Reiter [Tue, 23 Jul 2002 01:09:34 +0000 (01:09 +0000)]
- Make use of the VM_ALLOC_WIRED flag in the call to vm_page_alloc() in
  do_sendfile().  This allows us to rearrange an if statement in order to
  avoid doing an unnecesary call to vm_page_lock_queues(), and an attempt
  at re-wiring the pages (which were wired in the vm_page_alloc() call).

Reviewed by: alc, jhb

22 years agoAdd '+FreeBSD' to the version string, to reflect that this program
Bill Fenner [Tue, 23 Jul 2002 00:56:52 +0000 (00:56 +0000)]
Add '+FreeBSD' to the version string, to reflect that this program
 bears little relationship to the one from LBL with the same
 name and version number.

22 years agoAdd NULL check for the output buffer from AcpiEvaluateObject().
Mitsuru IWASAKI [Tue, 23 Jul 2002 00:47:13 +0000 (00:47 +0000)]
Add NULL check for the output buffer from AcpiEvaluateObject().

Submitted by: jon

22 years agoAdd the -n option, which automatically answers "no" to the overwrite question.
Johan Karlsson [Tue, 23 Jul 2002 00:42:56 +0000 (00:42 +0000)]
Add the -n option, which automatically answers "no" to the overwrite question.

PR: 7828
Suggested by: Daniel O'Connor <doconnor@gsoft.com.au>
Approved by: sheldonh (mentor)
MFC after: 2 weeks

22 years agoDon't forget to recalculate the IP checksum of the original
Ruslan Ermilov [Tue, 23 Jul 2002 00:16:19 +0000 (00:16 +0000)]
Don't forget to recalculate the IP checksum of the original
IP datagram embedded into ICMP error message.

Spotted by: tcpdump 3.7.1 (-vvv)
MFC after: 3 days

22 years agoAdd an ICMP protocol handler, partly based on LBL's traceroute 1.4 .
Bill Fenner [Mon, 22 Jul 2002 23:01:11 +0000 (23:01 +0000)]
Add an ICMP protocol handler, partly based on LBL's traceroute 1.4 .

Submitted by: dcs

22 years agoDon't shrink socket buffers in tcp_mss(), application might have already
Ruslan Ermilov [Mon, 22 Jul 2002 22:31:09 +0000 (22:31 +0000)]
Don't shrink socket buffers in tcp_mss(), application might have already
configured them with setsockopt(SO_*BUF), for RFC1323's scaled windows.

PR: kern/11966
MFC after: 1 week

22 years agoActually use gen_prep() and gen_check(), instead of udp_prep() and
Bill Fenner [Mon, 22 Jul 2002 22:07:23 +0000 (22:07 +0000)]
Actually use gen_prep() and gen_check(), instead of udp_prep() and
 udp_check(), for unknown protocols.

22 years agoRevert last commit. Sorry for breaking things
Tom Rhodes [Mon, 22 Jul 2002 21:45:16 +0000 (21:45 +0000)]
Revert last commit.  Sorry for breaking things

22 years agointegrate from p4 repo:
Warner Losh [Mon, 22 Jul 2002 21:28:46 +0000 (21:28 +0000)]
integrate from p4 repo:
o Honor NE2000DVF_{AX88190,DL10019} flags by setting the ED_FLAGS_xxxx
  flag.
o Mark linksys combo_ecard as ax88190
o Set the type_str to AX88190 for the ax88190 cards.

This fixes ax88190 based cards, for the most part, but doesn't seem to fix
the mii based dl10019 cards (aka linksys cards).

22 years agoThree minor changes:
Warner Losh [Mon, 22 Jul 2002 21:24:12 +0000 (21:24 +0000)]
Three minor changes:
1) Add support for compaq cpq 10/100 card (a rebadged ce3)
2) Minor protection for failure on vendor lookup
3) Minor style(9) nits.

22 years agoMFC: Fix some grammar, greater reduce diffs to HEAD
Tom Rhodes [Mon, 22 Jul 2002 21:14:16 +0000 (21:14 +0000)]
MFC: Fix some grammar, greater reduce diffs to HEAD

22 years agoRemove unneeded caddr_t casts.
Alfred Perlstein [Mon, 22 Jul 2002 19:05:44 +0000 (19:05 +0000)]
Remove unneeded caddr_t casts.

22 years agoFix the transfer size calculation.
Søren Schmidt [Mon, 22 Jul 2002 18:37:33 +0000 (18:37 +0000)]
Fix the transfer size calculation.

22 years agoSync with the spl changes in -stable.
Søren Schmidt [Mon, 22 Jul 2002 18:36:14 +0000 (18:36 +0000)]
Sync with the spl changes in -stable.
Add start of support for DVD+RW.

22 years agoUpdate the tags handling a bit, which makes support for the
Søren Schmidt [Mon, 22 Jul 2002 18:35:01 +0000 (18:35 +0000)]
Update the tags handling a bit, which makes support for the
older IBM DTTA series of drives possible.
Update error handling a bit now we are here.

22 years agoCleanup:
Alfred Perlstein [Mon, 22 Jul 2002 18:27:54 +0000 (18:27 +0000)]
Cleanup:
Define a debug printf macro rather than wrapping all calls to printf
with #ifdefs.

22 years agouse IPV6_V6ONLY instead of non standard IPV6_BINDV6ONLY.
Hajimu UMEMOTO [Mon, 22 Jul 2002 18:16:05 +0000 (18:16 +0000)]
use IPV6_V6ONLY instead of non standard IPV6_BINDV6ONLY.

Approved by: gad

22 years agoadd -4 and -6 option to be able to specify an address family.
Hajimu UMEMOTO [Mon, 22 Jul 2002 17:19:54 +0000 (17:19 +0000)]
add -4 and -6 option to be able to specify an address family.

Requested by: matusita
MFC after: 1 week

22 years agoThis commit was generated by cvs2svn to compensate for changes in r100519,
Ruslan Ermilov [Mon, 22 Jul 2002 17:06:28 +0000 (17:06 +0000)]
This commit was generated by cvs2svn to compensate for changes in r100519,
which included commits to RCS files with non-trunk default branches.

22 years agoUpdate for GNU texinfo 4.2.
Ruslan Ermilov [Mon, 22 Jul 2002 16:58:54 +0000 (16:58 +0000)]
Update for GNU texinfo 4.2.

22 years agoMerge texinfo 4.2 changes onto the trunk.
Ruslan Ermilov [Mon, 22 Jul 2002 16:55:14 +0000 (16:55 +0000)]
Merge texinfo 4.2 changes onto the trunk.

22 years agoUse stock (FSF) version of this file.
Ruslan Ermilov [Mon, 22 Jul 2002 16:54:53 +0000 (16:54 +0000)]
Use stock (FSF) version of this file.

22 years agoThis commit was generated by cvs2svn to compensate for changes in r100513,
Ruslan Ermilov [Mon, 22 Jul 2002 16:52:57 +0000 (16:52 +0000)]
This commit was generated by cvs2svn to compensate for changes in r100513,
which included commits to RCS files with non-trunk default branches.

22 years agoImport of stripped down GNU texinfo 4.2
Ruslan Ermilov [Mon, 22 Jul 2002 16:52:57 +0000 (16:52 +0000)]
Import of stripped down GNU texinfo 4.2

22 years agoChange struct vmspace->vm_shm from void * to struct shmmap_state *, this
Alfred Perlstein [Mon, 22 Jul 2002 16:22:27 +0000 (16:22 +0000)]
Change struct vmspace->vm_shm from void * to struct shmmap_state *, this
removes the need for casts in several cases.

22 years agoRemove caddr_t.
Alfred Perlstein [Mon, 22 Jul 2002 16:12:55 +0000 (16:12 +0000)]
Remove caddr_t.

22 years agoSurround prototypes with __BEGIN_DECLS / __END_DECLS.
Dag-Erling Smørgrav [Mon, 22 Jul 2002 16:11:39 +0000 (16:11 +0000)]
Surround prototypes with __BEGIN_DECLS / __END_DECLS.

PR: misc/40399
MFC after: 3 days

22 years agosync comment with reality. IN6P_BINDV6ONLY -> IN6P_IPV6_V6ONLY.
Hajimu UMEMOTO [Mon, 22 Jul 2002 15:55:50 +0000 (15:55 +0000)]
sync comment with reality.  IN6P_BINDV6ONLY -> IN6P_IPV6_V6ONLY.

22 years agodo not refer to IN6P_BINDV6ONLY anymore.
Hajimu UMEMOTO [Mon, 22 Jul 2002 15:51:02 +0000 (15:51 +0000)]
do not refer to IN6P_BINDV6ONLY anymore.

Obtained from: KAME
MFC after: 1 week

22 years agoremove caddr_t from fo_ioctl calls
Alfred Perlstein [Mon, 22 Jul 2002 15:46:51 +0000 (15:46 +0000)]
remove caddr_t from fo_ioctl calls

22 years agoremove caddr_t
Alfred Perlstein [Mon, 22 Jul 2002 15:44:27 +0000 (15:44 +0000)]
remove caddr_t