]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
20 years agoResolve conflicts.
mbr [Sat, 26 Jun 2004 10:37:42 +0000 (10:37 +0000)]
Resolve conflicts.

20 years agoMFNetBSD.
le [Sat, 26 Jun 2004 10:35:10 +0000 (10:35 +0000)]
MFNetBSD.

uhid.c (1.61), author: jdolecek
   add support for USB_GET_DEVICEINFO and USB_GET_STRING_DESC ioctls,
   with same meaning as for ugen(4)

usbdi_util.h (1.29), usb_quirks.c (1.50), uhid.c (1.62),
ugen.c (1.68), usb_subr.c (1.114) author: mycroft
   Yes, some devices return incorrect lengths in their string
   descriptors.  Rather than losing, do what Windows does: just
   request the maximum size, and allow a shorter response.  Obsoletes
   the need for UQ_NO_STRINGS, and therefore these "quirks" are removed.

usb_subr.c (1.116), author: mycroft
    In the "seemed like a good idea until I found the fatal flaw"
    department...  Attempting to read a maximum-size string descriptor
    causes my kue device to go completely apeshit.  So, go back to the
    original method, but allow the device to return a shorter string than
    it claimed.

Obtained from:   NetBSD

20 years agoThis commit was generated by cvs2svn to compensate for changes in r131136,
mbr [Sat, 26 Jun 2004 10:34:24 +0000 (10:34 +0000)]
This commit was generated by cvs2svn to compensate for changes in r131136,
which included commits to RCS files with non-trunk default branches.

20 years agoImport ISC DHCP 3.0.1 RC14 client.
mbr [Sat, 26 Jun 2004 10:34:24 +0000 (10:34 +0000)]
Import ISC DHCP 3.0.1 RC14 client.

20 years agoMake the stop command respect the 'fast' prefix.
mtm [Sat, 26 Jun 2004 09:27:30 +0000 (09:27 +0000)]
Make the stop command respect the 'fast' prefix.

Most notably, this cleans up messages when shutting down from
single user. In such a case there are usually no daemons running,
but their pid files are still in /var/run. This causes rc.d to
output diagnostics about daemons with pidfiles, but that are
not running.

20 years agoPick the hotchar out of the tty structure instead of caching private
phk [Sat, 26 Jun 2004 09:20:07 +0000 (09:20 +0000)]
Pick the hotchar out of the tty structure instead of caching private
copies.

No current line disciplines have a dynamically changing hotchar, and
expecting to receive anything sensible during a change in ldisc is
insane so no locking of the hotchar field is necessary.

20 years agoMFNetBSD.
le [Sat, 26 Jun 2004 09:19:31 +0000 (09:19 +0000)]
MFNetBSD.

ohci.c (1.147), author: mycroft
   Failure to properly mask off UE_DIR_IN from the endpoint address
   was causing OHCI_ED_FORMAT_ISO and EHCI_QH_HRECL to get set
   spuriously, causing rather interesting lossage.

   Suddenly I get MUCH better performance with ehci...

ohci.c (1.148), author: mycroft
   Adjust a couple of comments to make it clear WTF is going on.

Obtained from:   NetBSD

20 years agoFixed a style bug in the previous commit.
bde [Sat, 26 Jun 2004 09:02:44 +0000 (09:02 +0000)]
Fixed a style bug in the previous commit.

20 years agoFixed world breakage in the NOSHARED=yes case (missing library).
bde [Sat, 26 Jun 2004 08:48:50 +0000 (08:48 +0000)]
Fixed world breakage in the NOSHARED=yes case (missing library).

Fixed some style bugs.

20 years agoFix line discipline switching issues: If opening a new ldisc fails,
phk [Sat, 26 Jun 2004 08:44:04 +0000 (08:44 +0000)]
Fix line discipline switching issues: If opening a new ldisc fails,
we have to revert to TTYDISC which we know will successfully open
rather than try the previous ldisc which might also fail to open.

Do not let ldisc implementations muck about with ->t_line, and remove
code which checks for reopens, it should never happen.

Move ldisc->l_hotchar to tty->t_hotchar and have ldisc implementation
initialize it in their open routines.  Reset to zero when we enter
TTYDISC.  ("no" should really be -1 since zero could be a valid
hotchar for certain old european mainframe protocols.)

20 years agoMFp4:
imp [Sat, 26 Jun 2004 06:02:06 +0000 (06:02 +0000)]
MFp4:

Now that the devs files are marked before-depend, we can remvoe them
from a few places they were explicitly mentioned (along with
BEFORE_DEPEND).

Noticed by: bde

20 years agoMFp4:
imp [Sat, 26 Jun 2004 05:58:43 +0000 (05:58 +0000)]
MFp4:

Reduce the need for hard coded *devs in various makefiles by declaring
them before-depend.

Other bugs in the handling of *devs remain, but this is the start of
the cleanup.  These will be address in future commits.

Cleanup Motivator: bde

20 years agoEmbellish the getopt(3) example with mixed case.
obrien [Sat, 26 Jun 2004 05:15:00 +0000 (05:15 +0000)]
Embellish the getopt(3) example with mixed case.

Reviewed by: bde

20 years agoMerged from sys/dev/sio/sio.c revision 1.442.
nyan [Sat, 26 Jun 2004 04:00:46 +0000 (04:00 +0000)]
Merged from sys/dev/sio/sio.c revision 1.442.
(Use generic support for modemcontrol and BREAK ioctls.)

20 years agoNot quite sure how that one got past me..
julian [Sat, 26 Jun 2004 01:22:29 +0000 (01:22 +0000)]
Not quite sure how that one got past me..

20 years agoIf HISMACADDR is set in the environment (by pppoed), pass the value to
brian [Sat, 26 Jun 2004 01:02:31 +0000 (01:02 +0000)]
If HISMACADDR is set in the environment (by pppoed), pass the value to
the RADIUS server as RAD_CALLING_STATION_ID.

PR: 44310
Submitted by: Gleb Smirnoff <glebius@cell.sick.ru>
MFC after: 2 weeks

20 years agoMFNetBSD ehci.c and ehcireg.h
le [Sat, 26 Jun 2004 00:52:37 +0000 (00:52 +0000)]
MFNetBSD ehci.c and ehcireg.h

ehci.c (1.55), ehcireg.h (1.16); author: mycroft
     Set the data toggle correctly, and use EHCI_QTD_DTC.  This fixes
     problems with my ALi-based drive enclosure (it works now, rather
     than failing to attach).  Also seems to work with a GL811-based
     enclosure and an ASUS enclosure with a CD-RW, on both Intel and
     NEC controllers.

     Note: The ALi enclosure is currently very SLOW, due to some issue
     with taking too long to notice that the QTD is complete.  This
     requires more investigation.

ehci.c (1.56); author: mycroft
     Failure to properly mask off UE_DIR_IN from the endpoint address
     was causing OHCI_ED_FORMAT_ISO and EHCI_QH_HRECL to get set
     spuriously, causing rather interesting lossage.

     Suddenly I get MUCH better performance with ehci...

ehci.c (1.58); author: mycroft
     Fix a stupid bug in ehci_check_intr() that caused use to try to
     complete a transaction that was still running.  Now ehci can
     handle multiple devices being active at once.

ehci.c (1.59); author: enami
     As the ehci_idone() now uses the variable `epipe'
     unconditionally, always declare it (in other words, make this
     file compile w/o EHCI_DEBUG).

ehci.c (1.60); author: mycroft
     Remove comment about the data toggle being borked.

ehci.c (1.61); author: mycroft
     Update comment.

ehci.c (1.62); author: mycroft
     Adjust a couple of comments to make it clear WTF is going on.

ehci.c (1.63); author: mycroft
     Fix an error in a debug printf().

ehci.c (1.64), ehcireg.h (1.17); author: mycroft
     Further cleanup of toggle handling.  Now that we use EHCI_QH_DTC,
     we don't need to fiddle with the TOGGLE bit in the overlay
     descriptor, so minimize how much we fuss with it.

Obtained from:   NetBSD

20 years agoRemove dependency on dev/usb/usbdevs.h, in preparation for its
imp [Fri, 25 Jun 2004 23:14:06 +0000 (23:14 +0000)]
Remove dependency on dev/usb/usbdevs.h, in preparation for its
disappearing from the tree.  We already were splitting the baby (using
the symbol for the vendor BROADCOM, but not for the device).  Use
#defines for both.

20 years agoAdd type definitions for prgregset_t and psaddr_t. Both are used by
marcel [Fri, 25 Jun 2004 23:06:20 +0000 (23:06 +0000)]
Add type definitions for prgregset_t and psaddr_t. Both are used by
the proc services API. The prfpregset_t type already existed.

20 years agoAdd '#include <sys/mbuf.h>' to fix the kernel build.
emax [Fri, 25 Jun 2004 23:03:33 +0000 (23:03 +0000)]
Add '#include <sys/mbuf.h>' to fix the kernel build.

20 years agooops from Gleb..
julian [Fri, 25 Jun 2004 22:59:59 +0000 (22:59 +0000)]
oops from Gleb..
This shouldn't be visible from userland.

Submitted by: Gleb Smirnoff <glebius@cell.sick.ru>

20 years agoRemove fbsd-thread.o, it's not for consumption yet.
marcel [Fri, 25 Jun 2004 22:57:56 +0000 (22:57 +0000)]
Remove fbsd-thread.o, it's not for consumption yet.

20 years agoGah! commit from wrong tree.
phk [Fri, 25 Jun 2004 22:10:20 +0000 (22:10 +0000)]
Gah!  commit from wrong tree.

Remove now unused variables from last commit.

20 years agoRetire the TIOC_REMOTE ioctl.
phk [Fri, 25 Jun 2004 21:54:49 +0000 (21:54 +0000)]
Retire the TIOC_REMOTE ioctl.

It was added 22 years ago for emacs to use, but emacs gave up on it
it 17 years ago.

20 years agoAdd support for LSI type software RAID's.
sos [Fri, 25 Jun 2004 21:21:59 +0000 (21:21 +0000)]
Add support for LSI type software RAID's.

Made possible by: John Cagle @ HP

20 years agoMake the frameworkl responsible for not passing the nodes a NULL mbuf pointer.
julian [Fri, 25 Jun 2004 21:11:14 +0000 (21:11 +0000)]
Make the frameworkl responsible for not passing the nodes a NULL mbuf pointer.
this allows the nodes to not test for this..

Submitted by: Gleb Smirnoff <glebius@cell.sick.ru>

20 years agoUse generic routines for BREAK and modem control ioctls.
phk [Fri, 25 Jun 2004 20:33:52 +0000 (20:33 +0000)]
Use generic routines for BREAK and modem control ioctls.
This eliminates the need for a local ioctl method.

20 years agoUse generic modem control and BREAK ioctl handling. This eliminated
phk [Fri, 25 Jun 2004 20:32:51 +0000 (20:32 +0000)]
Use generic modem control and BREAK ioctl handling.  This eliminated
the need for a local ioctl handler.

20 years agoRelease UNIX domain socket subsystem lock earlier -- don't need to
rwatson [Fri, 25 Jun 2004 20:12:06 +0000 (20:12 +0000)]
Release UNIX domain socket subsystem lock earlier -- don't need to
hold it over free of unp_addr if we've already removed all references
to unp.

20 years agoConvert Netgraph to use mbuf tags to pass its meta information around.
julian [Fri, 25 Jun 2004 19:22:05 +0000 (19:22 +0000)]
Convert Netgraph to use mbuf tags to pass its meta information around.
Thanks to Sam for importing tags in a way that allowed this to be done.

Submitted by: Gleb Smirnoff <glebius@cell.sick.ru>
Also allow the sr and ar drivers to create netgraph versions of their modules.
Document the change to the ksocket node.

20 years agoMark a plex as 'newborn' when it is created. This is used to indicate
le [Fri, 25 Jun 2004 18:04:33 +0000 (18:04 +0000)]
Mark a plex as 'newborn' when it is created.  This is used to indicate
that new RAID5 plexes need to be initialized first.

20 years agoFix uninitialized variable, that breaks the build.
josef [Fri, 25 Jun 2004 16:34:33 +0000 (16:34 +0000)]
Fix uninitialized variable, that breaks the build.

Approved by: hmp
Reviewed by: chris
Pointy hat to: josef (for ignoring warning)

20 years agoCatch up with sparc64 OFWCONS_POLL_HZ change
grehan [Fri, 25 Jun 2004 13:44:34 +0000 (13:44 +0000)]
Catch up with sparc64 OFWCONS_POLL_HZ change

20 years ago- set resid correctly so that a failed seek (e.g. end of file) returns
grehan [Fri, 25 Jun 2004 13:43:56 +0000 (13:43 +0000)]
- set resid correctly so that a failed seek (e.g. end of file) returns
  correctly
- included required <sys/module.h>

20 years agoCatchup to now-required <sys/module.h> for PowerPC
grehan [Fri, 25 Jun 2004 13:42:48 +0000 (13:42 +0000)]
Catchup to now-required <sys/module.h> for PowerPC

20 years agoo Fix semantics of comparison function for qsort(3). According to qsort(3)
sobomax [Fri, 25 Jun 2004 13:04:49 +0000 (13:04 +0000)]
o Fix semantics of comparison function for qsort(3). According to qsort(3)
  manpage:

     The comparison function must return an integer less than, equal to, or
     greater than zero if the first argument is considered to be respectively
     less than, equal to, or greater than the second.

  Therefore, simply returning "arg1 > arg2" is incorrect. Actually it works
  but for the number of items to be sorted less than 7 due to special case
  handling in qsort(3);

o add missing '\n' to one of usage() calls.

Approved by: phk

20 years agoRevert the last change. There are more 64bit platforms than amd64, and
kan [Fri, 25 Jun 2004 12:32:45 +0000 (12:32 +0000)]
Revert the last change. There are more 64bit platforms than amd64, and
they break due to diferent alignment restrictions.

20 years agoAdd qsort, mergesort and heapsort regression tests.
sobomax [Fri, 25 Jun 2004 12:31:12 +0000 (12:31 +0000)]
Add qsort, mergesort and heapsort regression tests.

20 years agoMake the relationship between modem control bits and their delta cousins
phk [Fri, 25 Jun 2004 10:56:43 +0000 (10:56 +0000)]
Make the relationship between modem control bits and their delta cousins
public.

20 years agoUse generic support for BREAK and modem control ioctls
phk [Fri, 25 Jun 2004 10:54:46 +0000 (10:54 +0000)]
Use generic support for BREAK and modem control ioctls

20 years agoUse generic support for BREAK and modem control ioctls.
phk [Fri, 25 Jun 2004 10:54:05 +0000 (10:54 +0000)]
Use generic support for BREAK and modem control ioctls.

20 years agoUse generic support for modemcontrol and BREAK ioctls.
phk [Fri, 25 Jun 2004 10:51:33 +0000 (10:51 +0000)]
Use generic support for modemcontrol and BREAK ioctls.

20 years agoDo not count loobacks as other fuilures.
rik [Fri, 25 Jun 2004 10:25:33 +0000 (10:25 +0000)]
Do not count loobacks as other fuilures.
As a result magic will not be rejected any more in case of loopback.

Discussed with: joerg@

20 years agoAdd two new methods to struct tty: One for manipulating BREAK condition
phk [Fri, 25 Jun 2004 10:24:10 +0000 (10:24 +0000)]
Add two new methods to struct tty:  One for manipulating BREAK condition
and one for fiddling modem-control signals.

Add generic code to deal with the relevant ioctls if these methods are
present.

20 years agoNuke COMPAT_43
phk [Fri, 25 Jun 2004 10:11:28 +0000 (10:11 +0000)]
Nuke COMPAT_43

20 years agoGDB has moved up one level.
obrien [Fri, 25 Jun 2004 07:34:56 +0000 (07:34 +0000)]
GDB has moved up one level.

20 years agoHook gdb into the build. Exclude gdb on arm and powerpc as support
marcel [Fri, 25 Jun 2004 07:15:13 +0000 (07:15 +0000)]
Hook gdb into the build. Exclude gdb on arm and powerpc as support
does not exist yet. The NO_GDB knob is respected.

20 years agoRemove building gdb info files. The gdb info files are built as part
marcel [Fri, 25 Jun 2004 06:55:49 +0000 (06:55 +0000)]
Remove building gdb info files. The gdb info files are built as part
of building gdb itself, which has been decoupled from binutils for
as much as is reasonable.

20 years agoAdd bmake glue for gdb 6.1.1. Supports alpha, amd64, i386, ia64
marcel [Fri, 25 Jun 2004 06:48:54 +0000 (06:48 +0000)]
Add bmake glue for gdb 6.1.1. Supports alpha, amd64, i386, ia64
and sparc64.

Note that the debugger doesn't support threading, nor kernel
debugging yet.

20 years agoUpdate HEAD with support for ia64.
marcel [Fri, 25 Jun 2004 05:37:08 +0000 (05:37 +0000)]
Update HEAD with support for ia64.

20 years agoDon't continue if the '-C' option (a-la GNUtar; commited in options.c
obrien [Fri, 25 Jun 2004 05:27:59 +0000 (05:27 +0000)]
Don't continue if the '-C' option (a-la GNUtar; commited in options.c
rev 1.18, but not documented in the man page) caused a failed chdir.
Otherwise, one can easily overwrite files.

Submitted by: Robert Nagy <robert@openbsd.org>
Obtained from: OpenBSD

20 years agoThis commit was generated by cvs2svn to compensate for changes in r131082,
marcel [Fri, 25 Jun 2004 05:19:25 +0000 (05:19 +0000)]
This commit was generated by cvs2svn to compensate for changes in r131082,
which included commits to RCS files with non-trunk default branches.

20 years agoAdd support for ia64. This code will be contributed to GDB as soon
marcel [Fri, 25 Jun 2004 05:19:25 +0000 (05:19 +0000)]
Add support for ia64. This code will be contributed to GDB as soon
as the legal prerequisites are met. To avoid future merge conflicts,
this is added to the vendor branch. The changes are seperately kept
in FREEBSD-diffs for easy reference.

20 years agoUpdate the upgrade instructions.
marcel [Fri, 25 Jun 2004 05:04:09 +0000 (05:04 +0000)]
Update the upgrade instructions.

20 years agoAdd readme and installation to the build
den [Fri, 25 Jun 2004 05:03:56 +0000 (05:03 +0000)]
Add readme and installation to the build

Obtained from: The FreeBSD Russian Documentation Project

20 years agoWhite space & spelling fixes
ps [Fri, 25 Jun 2004 04:11:26 +0000 (04:11 +0000)]
White space & spelling fixes

Submitted by: Xin LI <delphij@frontfree.net>

20 years agoWhitespace.
bms [Fri, 25 Jun 2004 02:29:58 +0000 (02:29 +0000)]
Whitespace.

20 years agoRemove the use of cast as lvalue.
kan [Fri, 25 Jun 2004 01:54:26 +0000 (01:54 +0000)]
Remove the use of cast as lvalue.

20 years agoAdd support for multibyte characters in input files and delimiter
tjr [Fri, 25 Jun 2004 01:48:43 +0000 (01:48 +0000)]
Add support for multibyte characters in input files and delimiter
strings (arguments to the -d option.) This involves backing out paste.c
rev. 1.13 until we have a version of fgetln() that operates on wide
character streams.

20 years agoRemove the use of cast as lvalue. GCC 3.4 isses a deprecation warning
kan [Fri, 25 Jun 2004 01:16:02 +0000 (01:16 +0000)]
Remove the use of cast as lvalue. GCC 3.4 isses a deprecation warning
for this now.

20 years agoCorrect the tracking of various bits of the process's vmspace and vm_map
green [Thu, 24 Jun 2004 22:43:46 +0000 (22:43 +0000)]
Correct the tracking of various bits of the process's vmspace and vm_map
when not propogated on fork (due to minherit(2)).  Consistency checks
otherwise fail when the vm_map is freed and it appears to have not been
emptied completely, causing an INVARIANTS panic in vm_map_zdtor().

PR: kern/68017
Submitted by: Mark W. Krentel <krentel@dreamscape.com>
Reviewed by: alc

20 years agoAnnotate that we don't check the returned data length from ufs_readdir()
rwatson [Thu, 24 Jun 2004 18:31:23 +0000 (18:31 +0000)]
Annotate that we don't check the returned data length from ufs_readdir()
because UFS uses fixed-size directory blocks.  When using this code with
other file systems, such as HFS+, the value of auio.uio_resid will need
to be taken into account.

20 years agoFix a typo that made the busdma tag have a 0 highaddr. This driver heavily
scottl [Thu, 24 Jun 2004 18:30:08 +0000 (18:30 +0000)]
Fix a typo that made the busdma tag have a 0 highaddr.  This driver heavily
abuses busdma in other ways, and those will likely be fixed another day.

20 years agoAdd PCI Device ID for 6300ESB ICH. If allocating 16 bytes fails then try 32.
ambrisko [Thu, 24 Jun 2004 18:21:28 +0000 (18:21 +0000)]
Add PCI Device ID for 6300ESB ICH.  If allocating 16 bytes fails then try 32.
This is also require for 6300ESB.

20 years agoRemove unnecessary setting of VV_SYSTEM on extended attribute backing
rwatson [Thu, 24 Jun 2004 18:17:41 +0000 (18:17 +0000)]
Remove unnecessary setting of VV_SYSTEM on extended attribute backing
files.  When this flag is used in our port of this code to Darwin, it
caused remarkable pain, and doesn't offer a benefit in FreeBSD.

20 years agoRemove the entry for configuring pcm0 mixer channels.
josef [Thu, 24 Jun 2004 17:49:18 +0000 (17:49 +0000)]
Remove the entry for configuring pcm0 mixer channels.
This is accomplished via device.hints now.

20 years agoProtect a non-text comment with a '-'.
rwatson [Thu, 24 Jun 2004 17:45:45 +0000 (17:45 +0000)]
Protect a non-text comment with a '-'.

20 years agoWhite space cleanup: use spaces instead of tabs in variable declarations
rwatson [Thu, 24 Jun 2004 17:44:14 +0000 (17:44 +0000)]
White space cleanup: use spaces instead of tabs in variable declarations
local to a function.  Remove a couple of blank lines in variable
declarations.

In one case, explicitly test against NULL rather than using a pointer
as a boolean directly.

20 years agoDocument the possibility to set default values for pcm mixer channels
josef [Thu, 24 Jun 2004 17:43:12 +0000 (17:43 +0000)]
Document the possibility to set default values for pcm mixer channels
by setting appropriate hints.

E.g.: hint.pcm.0.line="0"

Reviewed by: ru

20 years agoChange the possibility to configure pcm(4) via boot/loader.conf
josef [Thu, 24 Jun 2004 17:41:20 +0000 (17:41 +0000)]
Change the possibility to configure pcm(4) via boot/loader.conf
to use boot/device.hints now

As discussed on cvs-src@

Reviewed by: Hiten Pandya <hmp@backplane.com>

20 years agoDon't cuddle else's so much as we removed additional parts of each
rwatson [Thu, 24 Jun 2004 17:22:29 +0000 (17:22 +0000)]
Don't cuddle else's so much as we removed additional parts of each
block.

20 years agoRemove temporary API bandage that allowed applications speaking the
rwatson [Thu, 24 Jun 2004 17:14:28 +0000 (17:14 +0000)]
Remove temporary API bandage that allowed applications speaking the
older API to list attributes on a file (zero-length attribute name)
to function.  extattr_list_*() are now the only available APIs to
use when listing attributes.

20 years agoFix typo.
mtm [Thu, 24 Jun 2004 16:57:49 +0000 (16:57 +0000)]
Fix typo.

Submitted by: kuriyama

20 years agoDocument the fact that uniq(1) does not recognize multibyte characters.
tjr [Thu, 24 Jun 2004 16:29:28 +0000 (16:29 +0000)]
Document the fact that uniq(1) does not recognize multibyte characters.

20 years agoDocument the fact that join(1) does not recognize multibyte characters.
tjr [Thu, 24 Jun 2004 16:24:11 +0000 (16:24 +0000)]
Document the fact that join(1) does not recognize multibyte characters.

20 years agoCopy the warning about incorrect multibyte character handling from vis(3).
tjr [Thu, 24 Jun 2004 16:08:02 +0000 (16:08 +0000)]
Copy the warning about incorrect multibyte character handling from vis(3).

20 years agoDocument the fact that comm(1) does not recognize multibyte characters
tjr [Thu, 24 Jun 2004 15:57:56 +0000 (15:57 +0000)]
Document the fact that comm(1) does not recognize multibyte characters
in its input. Although doing so would require only trivial changes,
it would be incompatible with the ordering used by sort(1), which is the
primary source of comm's input.

20 years agoAdd support for multibyte characters and characters that take up more
tjr [Thu, 24 Jun 2004 15:12:29 +0000 (15:12 +0000)]
Add support for multibyte characters and characters that take up more
than one column position.

20 years agoAdd support for multibyte characters and for characters that take up
tjr [Thu, 24 Jun 2004 13:48:28 +0000 (13:48 +0000)]
Add support for multibyte characters and for characters that take up
more than one column position.

20 years agoAdd russian translation for readme
den [Thu, 24 Jun 2004 13:45:38 +0000 (13:45 +0000)]
Add russian translation for readme

Obtained from: The FreeBSD Russian Documentation Project

20 years agoThe description of the -S option in the man page says we won't fail if the
des [Thu, 24 Jun 2004 13:43:57 +0000 (13:43 +0000)]
The description of the -S option in the man page says we won't fail if the
remote size is unknown, but we do.  Resolve this in the man page's favor.

Requested by:   Andre Albsmeier <andre.albsmeier@siemens.com>
MFC after: 1 week

20 years agoAdd support for multibyte characters and for characters that take up
tjr [Thu, 24 Jun 2004 13:42:26 +0000 (13:42 +0000)]
Add support for multibyte characters and for characters that take up
more than one column position.

20 years agoAdd a couple of #ifdef DEBUG printf()s in vlan_input() I found to be
joerg [Thu, 24 Jun 2004 12:32:41 +0000 (12:32 +0000)]
Add a couple of #ifdef DEBUG printf()s in vlan_input() I found to be
useful when debugging the ether_demux() problem (when bridging over
VLANs).

20 years agoWhen considering an ethernet frame that is not destined for us, do not
joerg [Thu, 24 Jun 2004 12:31:44 +0000 (12:31 +0000)]
When considering an ethernet frame that is not destined for us, do not
only allow this to be further processed when bridging is active on
that interface, but also if the current packet has a VLAN tag and
VLANs are active on our interface.  This gives the VLAN layers a
chance to also consider the packet (and perhaps drop it instead of the
main dispatcher).

This fixes a situation where bridging was only active on VLAN
interfaces but ether_demux() called on behalf of the main interface
had already thrown the packet away.

MFC after: 4 weeks

20 years agoMake dependencies on the TCP/IP stack conditional on INET / INET6. This
des [Thu, 24 Jun 2004 10:58:08 +0000 (10:58 +0000)]
Make dependencies on the TCP/IP stack conditional on INET / INET6.  This
makes it possible to build a kernel with NIC drivers but no TCP/IP stack.

Sponsored by: Teleplan AS

20 years agos/SS_CANTSENDMORE/SBS_CANTSENDMORE/
rik [Thu, 24 Jun 2004 10:53:44 +0000 (10:53 +0000)]
s/SS_CANTSENDMORE/SBS_CANTSENDMORE/

20 years agoDon't force class to give a valid softc to g_slice_new(), it is not always
pjd [Thu, 24 Jun 2004 10:50:20 +0000 (10:50 +0000)]
Don't force class to give a valid softc to g_slice_new(), it is not always
needed.

Approved by: phk

20 years ago#include <sys/serial.h>
phk [Thu, 24 Jun 2004 10:32:30 +0000 (10:32 +0000)]
#include <sys/serial.h>

20 years agoUse the correct names for modem control signals as the primary and make
phk [Thu, 24 Jun 2004 10:20:53 +0000 (10:20 +0000)]
Use the correct names for modem control signals as the primary and make
the aliases aliases.

20 years agoUse the new serial port definitions for modemsignals.
phk [Thu, 24 Jun 2004 10:07:28 +0000 (10:07 +0000)]
Use the new serial port definitions for modemsignals.

20 years agoUse CTASSERT to enforce the relationship between the new serial port
phk [Thu, 24 Jun 2004 10:06:55 +0000 (10:06 +0000)]
Use CTASSERT to enforce the relationship between the new serial port
modem definitions and the old definitions from ioctls.

20 years agoAdd a tty-independent include file for definitions pertaining to serial
phk [Thu, 24 Jun 2004 10:05:58 +0000 (10:05 +0000)]
Add a tty-independent include file for definitions pertaining to serial
ports.

Define the most commonly used modem signals, and delta-bit definitions
for them.

20 years agoAdd russian translation for install notes
den [Thu, 24 Jun 2004 08:39:20 +0000 (08:39 +0000)]
Add russian translation for install notes

Build sparc64/install.sgml from english sources

Obtained from: The FreeBSD Russian Documentation Project

20 years agoUpdate some comments and remove non-standard notation.
njl [Thu, 24 Jun 2004 06:37:23 +0000 (06:37 +0000)]
Update some comments and remove non-standard notation.

20 years agoHook acpi_quirks up to the build for kernel and modules.
njl [Thu, 24 Jun 2004 06:29:22 +0000 (06:29 +0000)]
Hook acpi_quirks up to the build for kernel and modules.

20 years agoAdd a script to parse acpi_quirks into a header file that we can compile
njl [Thu, 24 Jun 2004 06:28:05 +0000 (06:28 +0000)]
Add a script to parse acpi_quirks into a header file that we can compile
into ACPI.  Script written by Mark Santcroos with some edits from myself.

Submitted by: marks

20 years agoAdd a quirks array based on matching ACPI tables. The entries in this file
njl [Thu, 24 Jun 2004 06:27:05 +0000 (06:27 +0000)]
Add a quirks array based on matching ACPI tables.  The entries in this file
are currently all bad BIOS revisions that will never be able to support
ACPI.  They were derived by examining which BIOS's are blacklisted by other
operating systems.  Other types of quirks will be possible here as well.

20 years agoSync to usbdevs 1.183
jb [Thu, 24 Jun 2004 05:25:04 +0000 (05:25 +0000)]
Sync to usbdevs 1.183

20 years agoInsert device foo for a few drivers that don't have it. Replace the
imp [Thu, 24 Jun 2004 05:09:15 +0000 (05:09 +0000)]
Insert device foo for a few drivers that don't have it.  Replace the
NetBSD config syntax with the FreeBSD syntax.

20 years agoAdd Delorme Publishing Earthmate GPS
jb [Thu, 24 Jun 2004 05:05:56 +0000 (05:05 +0000)]
Add Delorme Publishing Earthmate GPS