Peter Avalos [Wed, 8 Nov 2006 21:21:34 +0000 (21:21 +0000)]
Merge from vendor branch TNFTP:
Import of NetBSD's ftp client under its new name (tnftp)
Peter Avalos [Wed, 8 Nov 2006 21:21:34 +0000 (21:21 +0000)]
Import of NetBSD's ftp client under its new name (tnftp)
Matthew Dillon [Wed, 8 Nov 2006 20:59:52 +0000 (20:59 +0000)]
Add a missing #undef to properly compile atomic.c functions into the
kernel so modules can access them.
Matthew Dillon [Wed, 8 Nov 2006 20:43:31 +0000 (20:43 +0000)]
bmake uses /usr/share/mk/sys.mk, so we cannot require that MACHINE_CPU be
defined by the make program.
Matthew Dillon [Wed, 8 Nov 2006 16:40:00 +0000 (16:40 +0000)]
Get the MI portions of VKERNEL to build, start linking it against libc.
Matthew Dillon [Wed, 8 Nov 2006 16:38:47 +0000 (16:38 +0000)]
Move the Maxmem extern to systm.h
Peter Avalos [Wed, 8 Nov 2006 16:27:42 +0000 (16:27 +0000)]
Add patch from FreeBSD to fix an infinite loop described at:
http://security.FreeBSD.org/advisories/FreeBSD-SA-06:24.libarchive.asc
Sascha Wildner [Wed, 8 Nov 2006 13:46:33 +0000 (13:46 +0000)]
Sync zoneinfo database with tzdata2006o from elsie.
antarctica: 8.2 -> 8.3
northamerica: 8.9 -> 8.10
southamerica: 8.5 -> 8.6
From Paul Eggert's comments:
* Changes affecting current and future time stamps.
Cuba has ended its three years of permanent DST.
(Thanks to Jesper Norgaard Welen for this.)
* Changes affecting past time stamps.
Many transitions in Chile from 1918 through 1998 have been updated
to reflect <http://www.horaoficial.cl/horaof.htm>. (Thanks to
Jesper Norgaard Welen for this.) This also affects Palmer base
in Antarctica. I also updated some incorrect rules for use of
Argentina daylight saving rules at Palmer before 1982.
* Changes affecting commentary only.
The URLs for Paraguay's daylight saving decrees have been updated.
(Thanks to Jesper Norgaard Welen and Carlos Raul Perasso for this.)
Matthew Dillon [Tue, 7 Nov 2006 20:48:15 +0000 (20:48 +0000)]
More Machine-dependant/Machine-independant code and header file separation.
Numerous machine interfaces have MI APIs and should be declared in MI headers
even though the routines are defined in MD sources.
* Improve rdtsc()'s API so it can be used in MI code.
* Add an explicit enable in machine/${MACHINE}/Makefile.inc for syscons/apm.
* Abstract <machine/reg.h> and provide a MI API for it via <sys/reg.h>.
* Move additional MI API calls from <machine/md_var.h> to <sys/systm.h>.
Matthew Dillon [Tue, 7 Nov 2006 19:56:27 +0000 (19:56 +0000)]
Move <machine/dvcfg.h> to the one device that actually uses it, remove
other inclusions of dvcfg.h that are not used.
Give each platform architecture a "Makefile.inc" file which the build
system can use to determine device support when building devices.
Matthew Dillon [Tue, 7 Nov 2006 19:23:54 +0000 (19:23 +0000)]
Move <machine/ccbque.h> to <sys/ccbque.h>. ccbque.h is not a
machine-specific header file.
Matthew Dillon [Tue, 7 Nov 2006 18:50:07 +0000 (18:50 +0000)]
Remove system dependancies on <machine/ipl.h>. Only architecture files
need it now. SWI_* defines moved from the MD <machine/ipl.h> to the
MI <sys/interrupt.h> directory.
Matthew Dillon [Tue, 7 Nov 2006 17:51:24 +0000 (17:51 +0000)]
Misc cleanups and CVS surgery. Move a number of header and source files
from machine/pc32 to cpu/i386 as part of the ongoing architectural separation
work and do a bit of cleanup.
Matthew Dillon [Tue, 7 Nov 2006 06:57:02 +0000 (06:57 +0000)]
Major kernel build infrastructure changes, part 2/2 (user).
These changes are primarily designed to create a 2-layer machine and cpu
build hierarchy in order to support virtual kernel builds in the near term
and future porting efforts in the long term.
* The make program now pulls MACHINE, MACHINE_ARCH, and MACHINE_CPU from
the kernel hw.machine, hw.machine_arch, and hw.machine_cpu sysctl
variables, unless overridden by environment variables.
* The config program now generates additional softlinks in the kernel build
directory:
machine points to the platform architecture header files
machine_base points to the baes of the platform architecture directory.
cpu points to the cpu architecture header files
cpu_base points to the base of the cpu architecture directory
Additionally, the config program generates forwarding header files as
includes/machine/*.h. Any cpu architecture header files that are not
overridden by the platform architecture are generated in this directory.
The kernel build's Makefile adjusts the -I paths to effectively merge
machine/ and includes/machine/, presenting a uniform <machine/*.h>
include file architecture to the kernel build.
The kernel build's Makefile now automatically exports MACHINE and
MACHINE_ARCH as environment variables.
The kernel build's Makefile now makes opt_*, use_*, and IF header files
directly available to the module build subsystem so the module build
subsystem does not have to generate them in each module subdirectory.
* Kernel configuration files now require both a 'machine' and a 'machine_arch'
directive. Typically:
machine pc32 (subject to change)
machine_arch i386
* /usr/include/cpu has been added to augment /usr/include/machine. The
buildworld infrastructure also automatically merges any <cpu/*.h> header
files that were not overridden by the <machine/*.h> header files into
/usr/include/machine.
Note that direct access to /usr/include/cpu is not recommended.
* A number of weird cases that generate extranious 'machine' softlinks
(for <machine/*.h> header access) have been removed.
* The MACHINE_ARCH default to i386 has been removed. If the make program
does not define the MACHINE_ environment and sys.mk is included, sys.mk
will now fail with an error.
* cpdup has been added to the bootstrap tools.
Matthew Dillon [Tue, 7 Nov 2006 06:43:25 +0000 (06:43 +0000)]
Major kernel build infrastructure changes, part 1/2 (sys).
These changes are primarily designed to create a 2-layer machine and cpu
build hierarchy in order to support virtual kernel builds in the near term
and future porting efforts in the long term.
* Split arch/ into a set of platform architectures under machine/ and a
set of cpu architectures under cpu/. All platform and cpu header files
will be accessible via <machine/*.h>. Platform header files may override
cpu header files (the platform header file then typically #include's the
cpu header file).
* Any cpu header files that are not overridden will be copied directly
into /usr/include/machine/, allowing the platform to omit those header
files (not have to create degenerate forwarding header files).
* All source files access platform and cpu architecture files via the
<machine/*.h> path. The <cpu/*.h> path should only be used by
platform header files when including the lower level cpu header files.
* Require both the 'machine' and the 'machine_arch' directives in the
kernel config file.
* When building modules in the presence of a kernel config, use the IF
files, use*.h files, and opt*.h files provided by the kernel config
and do not generate them in each module's object directory. This
streamlines the module build considerably.
Matthew Dillon [Sun, 5 Nov 2006 21:49:23 +0000 (21:49 +0000)]
test 5
Matthew Dillon [Sun, 5 Nov 2006 21:47:49 +0000 (21:47 +0000)]
test 4
Dragonfly-bug:
Submitted-by:
Reviewed-by:
Approved-by:
Obtained-from:
Simon Schubert [Sun, 5 Nov 2006 11:40:45 +0000 (11:40 +0000)]
Abort if the device doesn't exist.
Matthew Dillon [Sun, 5 Nov 2006 02:26:39 +0000 (02:26 +0000)]
Sync our rm -P option with OpenBSD - if the file has a hardlink count
greater then one do not overwrite it or remove it, and issue a warning.
Simon Schubert [Fri, 3 Nov 2006 16:33:38 +0000 (16:33 +0000)]
re-add vinumsize, so that vinum volumes can be used to swap on
Jeffrey Hsu [Fri, 3 Nov 2006 00:54:55 +0000 (00:54 +0000)]
Fix userland compilation error.
Pointed out by Simon Schubert.
Sascha Wildner [Thu, 2 Nov 2006 19:53:56 +0000 (19:53 +0000)]
Add wcstod(3) and wcstol(3) manual pages.
Adapted-from: NetBSD
Sascha Wildner [Thu, 2 Nov 2006 19:48:55 +0000 (19:48 +0000)]
Sync strtol(3) and strtoul(3) with FreeBSD.
This fixes some mdoc nits, adds documentation about when EINVAL is returned
and makes the existance of strtoimax(3) and strtoumax(3) known to the world.
Jeffrey Hsu [Thu, 2 Nov 2006 05:16:35 +0000 (05:16 +0000)]
Clean up code.
Sascha Wildner [Wed, 1 Nov 2006 22:41:21 +0000 (22:41 +0000)]
There is no securenets(5) manual page. Use .Pa and add a reference
to ypserv(8).
Peter Avalos [Wed, 1 Nov 2006 22:33:47 +0000 (22:33 +0000)]
Upgrade awk to the latest version (
20050424).
Peter Avalos [Wed, 1 Nov 2006 21:34:07 +0000 (21:34 +0000)]
Add our READMEs.
Peter Avalos [Wed, 1 Nov 2006 21:20:57 +0000 (21:20 +0000)]
Merge from vendor branch AWK:
Import of awk
20050424
Peter Avalos [Wed, 1 Nov 2006 21:20:57 +0000 (21:20 +0000)]
Peter Avalos [Tue, 31 Oct 2006 20:22:22 +0000 (20:22 +0000)]
WARNS6 cleanup. Add a length modifier to a printf format argument since
inode numbers are 64 bits.
Sascha Wildner [Mon, 30 Oct 2006 13:47:59 +0000 (13:47 +0000)]
Fix parameter name in prototype.
Noticed-by: Steve Mynott <steve.mynott@gmail.com>
Sascha Wildner [Mon, 30 Oct 2006 10:02:10 +0000 (10:02 +0000)]
Remove recently deleted header files via 'make upgrade'.
Peter Avalos [Sat, 28 Oct 2006 20:42:54 +0000 (20:42 +0000)]
Add a __DECONST to bring this up to WARNS6.
Sepherosa Ziehau [Sat, 28 Oct 2006 12:28:25 +0000 (12:28 +0000)]
Bug fixing:
device_delete_child() is called unconditionally in bge_detach(), due to the
';' after 'if()'. It is dangerous here, since bge_detach() serves as failure
cleanup routine too.
Sascha Wildner [Fri, 27 Oct 2006 23:02:59 +0000 (23:02 +0000)]
Fix incorrect function name (kobj_free -> kobj_delete) and remove references
to kobj_class_compile_static() which was removed 2.5 years ago.
Sascha Wildner [Fri, 27 Oct 2006 22:39:05 +0000 (22:39 +0000)]
Comment out the remaining references to rman_reserve_resource_bound().
Sascha Wildner [Fri, 27 Oct 2006 18:15:28 +0000 (18:15 +0000)]
Fix function name and add MLINKS for all functions.
Sascha Wildner [Fri, 27 Oct 2006 10:53:59 +0000 (10:53 +0000)]
NgSendMsgReply() doesn't exist. Should be NgSendReplyMsg().
Matthew Dillon [Fri, 27 Oct 2006 05:04:35 +0000 (05:04 +0000)]
Adjust fstat to properly traverse mount points when constructing a
file or directory path, based on the new namecache topology.
Matthew Dillon [Fri, 27 Oct 2006 04:56:34 +0000 (04:56 +0000)]
Major namecache work primarily to support NULLFS.
* Move the nc_mount field out of the namecache{} record and use a new
namecache handle structure called nchandle { mount, ncp } for all
API accesses to the namecache.
* Remove all mount point linkages from the namecache topology. Each mount
now has its own namecache topology rooted at the root of the mount point.
Mount points are flagged in their underlying filesystem's namecache
topology but instead of linking the mount into the topology, the flag
simply triggers a mountlist scan to locate the mount. ".." is handled
the same way... when the root of a topology is encountered the scan
can traverse to the underlying filesystem via a field stored in the
mount structure.
* Ref the mount structure based on the number of nchandle structures
referencing it, and do not kfree() the mount structure during a forced
unmount if refs remain.
These changes have the following effects:
* Traversal across mount points no longer require locking of any sort,
preventing process blockages occuring in one mount from leaking across
a mount point to another mount.
* Aliased namespaces such as occurs with NULLFS no longer duplicate the
namecache topology of the underlying filesystem. Instead, a NULLFS
mount simply shares the underlying topology (differentiating between
it and the underlying topology by the fact that the name cache
handles { mount, ncp } contain NULLFS's mount pointer.
This saves an immense amount of memory and allows NULLFS to be used
heavily within a system without creating any adverse impact on kernel
memory or performance.
* Since the namecache topology for a NULLFS mount is shared with the
underyling mount, the namecache records are in fact the same records
and thus full coherency between the NULLFS mount and the underlying
filesystem is maintained by design.
* Future efforts, such as a unionfs or shadow fs implementation, now
have a mount structure to work with. The new API is a lot more
flexible then the old one.
Sascha Wildner [Thu, 26 Oct 2006 16:50:42 +0000 (16:50 +0000)]
Fix typo.
Sascha Wildner [Thu, 26 Oct 2006 12:58:30 +0000 (12:58 +0000)]
Remove obsolete prototype and documentation of the suser_proc()
function (which was never actually implemented).
OK'd-by: dillon
Manpage-fix-by: Matthias Schmidt <schmidtm@mathematik.uni-marburg.de>
Matthew Dillon [Thu, 26 Oct 2006 02:27:19 +0000 (02:27 +0000)]
Fix a stack overflow due to recursion. When the namecache must invalidate
an entire subtree the depth of the tree can cause the kernel stack to
overflow. Implement a simple algorithm which limits the recursion depth.
When the recursion depth is reached the current namecache node is recorded
and further recursion does not occur. The top level then recalls the
deletion at the recorded node, then restarts.
When removing a file or directory, invalidate all namecache records
associated with the vnode rather then just the one led to by the current
path. Note that RENAME is not fixed and requires a more sophisticated
solution. These fixes effect NULLFS.
Matthew Dillon [Wed, 25 Oct 2006 20:56:03 +0000 (20:56 +0000)]
Do a major clean-up of the BUSDMA architecture. A large number of
essentially machine-independant drivers use the structures and definitions
in machine-dependant directories that are really machine-independant in
nature.
Split <machine/bus_dma.h> into machine-depdendant and machine-independant parts
and make the primary access run through <sys/bus_dma.h>.
Remove <machine/bus.h>, <machine/bus_memio.h> and <machine/bus_pio.h>. The
optimizations related to bus_memio.h and bus_pio.h made a huge mess,
introduced machine-specific knowledge into essentially machine-independant
drivers, and required specific #include file orderings to do their job.
They may be reintroduced in some other form later on.
Move <machine/resource.h> to <sys/bus_resource.h>. The contents of the file
is machine-independant or can be made a superset across many platforms.
Make <sys/bus.h> include <sys/bus_dma.h> and <sys/bus_resource.h> and
include <sys/bus.h> where necessary. Remove all #include's of
<machine/resource.h> and <machine/bus.h>. That is, make the BUSDMA
infrastructure integral to I/O-mapped and memory-mapped accesses to devices
and remove a large chunk of machine-specific dependancies from drivers.
bus_if.h and device_if.h are now required to be present when using <sys/bus.h>.
Sascha Wildner [Wed, 25 Oct 2006 08:27:27 +0000 (08:27 +0000)]
Fix typos in some messages: writting -> writing
Jeffrey Hsu [Wed, 25 Oct 2006 07:41:45 +0000 (07:41 +0000)]
Localize some variables.
Simon Schubert [Wed, 25 Oct 2006 07:02:27 +0000 (07:02 +0000)]
Unbreak compilation without IPSEC: move variables into conditional
Matthew Dillon [Tue, 24 Oct 2006 17:16:13 +0000 (17:16 +0000)]
Purge the IFQ when associating a new altq. Packets that have already been
queued are not visible to the new altq and consequently not removed when
the altq is disassociated, resulting in a panic.
Reported-by: Kyle Butt <kylebutt@gmail.com>
Sascha Wildner [Tue, 24 Oct 2006 17:09:46 +0000 (17:09 +0000)]
Fix file-/pathnames which have changed now.
Jeffrey Hsu [Tue, 24 Oct 2006 16:33:11 +0000 (16:33 +0000)]
Fix compilation error with IPSEC.
Use boolean type and values for boolean variables.
Sepherosa Ziehau [Tue, 24 Oct 2006 14:39:45 +0000 (14:39 +0000)]
Add support in ONOE/AMRR for drivers that can't provide per TX statistics:
- Add callback function pointer ieee80211_ratectl_state.rc_st_state, which
will be called periodically by ONOE/AMRR during their internal state
updating. This callback function pointer should return a
ieee80211_ratectl_stats struct, which provides statistics concerning TX.
- Add ieee80211_ratectl_state.rc_st_valid_stats bitmask. It is used to
tell ONOE/AMRR which fields of ieee80211_ratectl_stats are valid.
- Make ONOE/AMRR spit out a warning message if driver can't provide per TX
statistics and the statistics in ieee80211_ratectl_stats are not enough to
do proper TX rate control.
Drivers that can't provide per TX statistics should setup
ieee80211_ratectl_state.rc_st_state and
ieee80211_ratectl_state.rc_st_valid_stats before ieee80211_ifattach().
Add ieee80211_ratectl_state.rc_st_param, it could be used by drivers to
override TX rate control algorithm's default tunable values. This field should
be allocated, setup and freed in ieee80211_ratectl_state.rc_st_change. Drivers
that is going to override TX rate control algorithm's default tunable values
should include specific TX rate control algorithm's parameter header file
e.g. netproto/802_11/wlan_ratectl/ieee80211_amrr_param.h
Victor Balada Diaz [Tue, 24 Oct 2006 09:24:15 +0000 (09:24 +0000)]
Now that we support foo_enable="YES|NO" as well as foo="YES|NO" some
variables collide because they use foo="" for something else.
Rename those variables to non-conflicting names.
Noticed-by: Sascha Wildner
The securelevel stuff is based on Kevin L. Kane work.
Jeffrey Hsu [Tue, 24 Oct 2006 07:27:26 +0000 (07:27 +0000)]
Remove nested block.
Jeffrey Hsu [Tue, 24 Oct 2006 06:49:52 +0000 (06:49 +0000)]
Cosmetic changes.
Jeffrey Hsu [Tue, 24 Oct 2006 06:18:42 +0000 (06:18 +0000)]
Cosmetic changes.
Jeffrey Hsu [Tue, 24 Oct 2006 02:52:10 +0000 (02:52 +0000)]
Eliminate a macro layer of indirection to clear up the control flow.
Matthew Dillon [Mon, 23 Oct 2006 21:59:18 +0000 (21:59 +0000)]
Further normalize the _XXX_H_ symbols used to conditionalize header file
inclusion.
Use _MACHINE_BLAH_H_ for headers found in "/usr/src/sys/arch/<arch>/include".
Most headers already did this, but some did not.
Use _ARCH_SUBDIR_BLAH_H_ for headers found in "/usr/src/sys/arch/<arch>/subdir
"
instead of _I386_SUBDIR_BLAH_H_.
Change #include's made in architecture-specific directories to use
<machine/blah.h> instead of "blah.h", allowing the included header files
to be overrdden by another architecture. For example, a virtual kernel
architecture might include a header from arch/i386/include which then
includes some other header in arch/i386/include. But really we want that
other header to also go via the arch/vkernel/include, so the header files
in arch/i386/include must use <machine/blah.h> instead of "blah.h" for most
of their sub-includes.
Change most architecture-specific includes such as <i386/icu/icu.h> to
use a generic path through the "arch" softlink, such as <arch/icu/icu.h>.
Remove the temporary -I@/arch shim made in a recent commit, the <arch/...>
mechanism replaces it.
These changes allow us to implement heirarchical architectural overrides,
primarily intended for virtual kernel support. A virtual kernel uses an
architecture of 'vkernel' but must be able to access actual cpu-specific
header files such as those found in arch/i386. It does this using a
"cpu" softlink. For example, someone including <machine/atomic.h> in a
vkernel build would hit the "arch/vkernel/include/atomic.h" header, and this
header could then #include <cpu/atomic.h> to access the actual cpu's
atomic.h file: "arch/i386/include/atomic.h".
The ultimate effect is that an architecture can build on another
architecture's header and source files.
Matthew Dillon [Mon, 23 Oct 2006 21:50:33 +0000 (21:50 +0000)]
Further normalize the _XXX_H_ symbols used to conditionalize header file
inclusion.
Use _MACHINE_BLAH_H_ for headers found in "/usr/src/sys/arch/<arch>/include".
Most headers already did this, but some did not.
Use _ARCH_SUBDIR_BLAH_H_ for headers found in "/usr/src/sys/arch/<arch>/subdir"
instead of _I386_SUBDIR_BLAH_H_.
Change #include's made in architecture-specific directories to use
<machine/blah.h> instead of "blah.h", allowing the included header files
to be overrdden by another architecture. For example, a virtual kernel
architecture might include a header from arch/i386/include which then
includes some other header in arch/i386/include. But really we want that
other header to also go via the arch/vkernel/include, so the header files
in arch/i386/include must use <machine/blah.h> instead of "blah.h" for most
of their sub-includes.
Change most architecture-specific includes such as <i386/icu/icu.h> to
use a generic path through the "arch" softlink, such as <arch/icu/icu.h>.
Remove the temporary -I@/arch shim made in a recent commit, the <arch/...>
mechanism replaces it.
These changes allow us to implement heirarchical architectural overrides,
primarily intended for virtual kernel support. A virtual kernel uses an
architecture of 'vkernel' but must be able to access actual cpu-specific
header files such as those found in arch/i386. It does this using a
"cpu" softlink. For example, someone including <machine/atomic.h> in a
vkernel build would hit the "arch/vkernel/include/atomic.h" header, and this
header could then #include <cpu/atomic.h> to access the actual cpu's
atomic.h file: "arch/i386/include/atomic.h".
The ultimate effect is that an architecture can build on another
architecture's header and source files.
Matthew Dillon [Mon, 23 Oct 2006 18:01:15 +0000 (18:01 +0000)]
Bump the config version. Add a 'cpu_arch' directive that allows the
cpu architecture for the virtual machine to be specified. This also
creates additional softlinks for <cpu/*.h> and <cpu_arch/...> to allow
a virtual machine to access cpu-architecture-specific files.
Matthew Dillon [Mon, 23 Oct 2006 15:42:50 +0000 (15:42 +0000)]
Get rid of the indirect function pointer for bzero(). We haven't used it
to install a more 'optimized' version of bzero in a long time, and it
isn't compatible with libc.
Sascha Wildner [Mon, 23 Oct 2006 09:27:37 +0000 (09:27 +0000)]
Rather than making static declarations global, get rid of the extern.
Pointed-out-by: dillon
Additional-clue: corecode
Simon Schubert [Mon, 23 Oct 2006 09:14:55 +0000 (09:14 +0000)]
The uthreads code previously didn't work for debugging while
uthreads was not initialized yet. Fix this by using _thread_kern_thread
if the _thread_list isn't populated yet.
Simon Schubert [Mon, 23 Oct 2006 09:14:47 +0000 (09:14 +0000)]
reorder libs to allow static linking
Sascha Wildner [Mon, 23 Oct 2006 07:45:47 +0000 (07:45 +0000)]
Update configuration instructions.
Matthew Dillon [Sun, 22 Oct 2006 18:43:52 +0000 (18:43 +0000)]
Fix paths to arch/i386, related to the recent architecture topology changes.
Matthew Dillon [Sun, 22 Oct 2006 18:42:12 +0000 (18:42 +0000)]
memset must be a real procedure rather then an indirect pointer because
GCC-4.x calls it directly.
Matthew Dillon [Sun, 22 Oct 2006 16:26:44 +0000 (16:26 +0000)]
Add advice if a kernel config file cannot be found to remind people that
the config directory has changed.
Matthew Dillon [Sun, 22 Oct 2006 16:12:50 +0000 (16:12 +0000)]
Reorganize the way machine architectures are handled. Consolidate the
kernel configurations into a single generic directory. Move machine-specific
Makefile's and loader scripts into the appropriate architecture directory.
Kernel and module builds also generally add sys/arch to the include path so
source files that include architecture-specific headers do not have to
be adjusted.
sys/<ARCH> -> sys/arch/<ARCH>
sys/conf/*.<ARCH> -> sys/arch/<ARCH>/conf/*.<ARCH>
sys/<ARCH>/conf/<KERNEL> -> sys/config/<KERNEL>
Matthew Dillon [Sun, 22 Oct 2006 16:09:22 +0000 (16:09 +0000)]
Reorganize the way machine architectures are handled. Consolidate the
kernel configurations into a single generic directory. Move machine-specific
Makefile's and loader scripts into the appropriate architecture directory.
Kernel and module builds also generally add sys/arch to the include path so
source files that include architecture-specific headers do not have to
be adjusted.
sys/<ARCH> -> sys/arch/<ARCH>
sys/conf/*.<ARCH> -> sys/arch/<ARCH>/conf/*.<ARCH>
sys/<ARCH>/conf/<KERNEL> -> sys/config/<KERNEL>
Sascha Wildner [Sat, 21 Oct 2006 22:44:34 +0000 (22:44 +0000)]
Fix typos.
Sascha Wildner [Sat, 21 Oct 2006 22:30:25 +0000 (22:30 +0000)]
Use .Ft for function types.
Sascha Wildner [Sat, 21 Oct 2006 22:17:09 +0000 (22:17 +0000)]
Fix typo.
Sepherosa Ziehau [Sat, 21 Oct 2006 09:25:58 +0000 (09:25 +0000)]
WOL_MAGIC 0x7770 -> 0x1111
Obtained-from: FreeBSD
# This "magic" change of the "magic" value is submitted by
# Peer Chen at nVIDIA
# to FreeBSD
Sascha Wildner [Sat, 21 Oct 2006 08:57:31 +0000 (08:57 +0000)]
Adjust to post Citrus import reality.
Sepherosa Ziehau [Sat, 21 Oct 2006 08:37:04 +0000 (08:37 +0000)]
Change the AMRR TX rate control algorithm:
- Bookkeeping number of TX (counter1), and number of "fail to TX at
desired rate" (counter2), instead of bookkeeping number of TX at
each rate (old_counter[1-4]).
- Calculate "failure" and "success" conditions using 'counter1' and
'counter2', instead of only using 'old_counter1' and 'old_counter2'.
The mistake of the old way is that 'old_counter3' and 'old_counter4'
should also be taken as "fail to TX at desired rate".
- Nuke no longer needed macro after above changes.
This change is tested with acx(4). The testing result shows much
better TX performance under bad enviroment (e.g. lots of concrete
obstacles between STA and AP, long distance etc.).
Sepherosa Ziehau [Sat, 21 Oct 2006 05:21:30 +0000 (05:21 +0000)]
Nuke M_HASFCS processing in ieee80211_input(), since
- It is easy for drivers to process trailing FCS after calling bpf_ptap().
- As mentioned by y0netan1@ in
http://leaf.dragonflybsd.org/mailarchive/bugs/2006-06/msg00034.html ,
leaving it there probably will bite us in the future.
Matthew Dillon [Sat, 21 Oct 2006 04:28:22 +0000 (04:28 +0000)]
Reformulate the way the kernel updates the PMAPs in the system when adding
a new page table page to expand kernel memory. Keep track of the PMAPs in
their own list rather then scanning the process list to locate them. This
allows PMAPs managed on behalf of virtual kernels to be properly updated.
VM spaces can now be allocated from scratch and may not have a parent
template to inherit certain fields from. Make sure these fields are
properly cleared.
Peter Avalos [Sat, 21 Oct 2006 04:10:02 +0000 (04:10 +0000)]
Add some casts to bring us up to WARNS2. This is almost WARNS6 except for
1 warning coming from the dblksize macro in <vfs/ufs/fs.h>.
Sascha Wildner [Fri, 20 Oct 2006 19:11:25 +0000 (19:11 +0000)]
Fix typo.
Simon Schubert [Fri, 20 Oct 2006 18:30:12 +0000 (18:30 +0000)]
Add the -D flag to specify an alternative dumpdev.
mdoc-help-by: swildner
Matthew Dillon [Fri, 20 Oct 2006 17:02:19 +0000 (17:02 +0000)]
Add a ton of infrastructure for VKERNEL support. Add code for intercepting
traps and system calls, for switching to and executing a foreign VM space,
and for accessing trap frames.
Peter Avalos [Fri, 20 Oct 2006 14:50:42 +0000 (14:50 +0000)]
Remove an unused argument and mark it WARNS6.
Peter Avalos [Fri, 20 Oct 2006 14:18:26 +0000 (14:18 +0000)]
Move WANT_LUKEMFTPD up a few lines so it's not grouped with the cvsup
options.
Reported by: Trevor Kendall <trevorjk@gmail.com>
Jeffrey Hsu [Fri, 20 Oct 2006 11:12:17 +0000 (11:12 +0000)]
Refactor internal ip6_splithdr() API to make it more of a pure function
by removing side-effects.
Peter Avalos [Thu, 19 Oct 2006 21:22:13 +0000 (21:22 +0000)]
WARNS6 cleanups for quotacheck.
Peter Avalos [Thu, 19 Oct 2006 21:11:13 +0000 (21:11 +0000)]
Remove a check that is always false.
Peter Avalos [Thu, 19 Oct 2006 20:51:40 +0000 (20:51 +0000)]
Remove variable names from function prototypes, and add a complete
prototype for inodesc.id_func.
There will be further WARNS cleanups to follow...
Sascha Wildner [Thu, 19 Oct 2006 20:11:52 +0000 (20:11 +0000)]
s/free/kfree/
Sascha Wildner [Thu, 19 Oct 2006 19:00:47 +0000 (19:00 +0000)]
Adjust stuff to the renaming of free to kfree and realloc to krealloc
also.
Sascha Wildner [Thu, 19 Oct 2006 18:44:01 +0000 (18:44 +0000)]
Adjust man pages, comments, messages and some defunct driver generation
scripts to the recent malloc(9) -> kmalloc(9) renaming.
Sascha Wildner [Thu, 19 Oct 2006 13:14:32 +0000 (13:14 +0000)]
Add reallocf.9 for removal with 'make upgrade'.
Jeffrey Hsu [Thu, 19 Oct 2006 07:12:14 +0000 (07:12 +0000)]
Cosmetic cleanups.
Victor Balada Diaz [Wed, 18 Oct 2006 21:44:41 +0000 (21:44 +0000)]
Support enabling daemons in base and pkgsrc like NetBSD and FreeBSD, ie:
foo="YES|NO" and foo_enable="YES|NO"
Victor Balada Diaz [Wed, 18 Oct 2006 21:38:23 +0000 (21:38 +0000)]
Update to new API and make it compile and work again
Victor Balada Diaz [Wed, 18 Oct 2006 21:30:06 +0000 (21:30 +0000)]
update man page:
malloc - kmalloc
free - kfree
realloc - krealloc
reallocf - no longer in tree, deleted
Jeffrey Hsu [Wed, 18 Oct 2006 06:27:43 +0000 (06:27 +0000)]
OpenBSD rev 1.66:
date: 2003-07-08 04:01:20 -0700; author: markus; state: Exp; lines: +31 -3;
Make sure the packets contains a complete inner header for ip{4,6}-in-ip{4,6}
encapsulation. Fixes panic for truncated ip-in-ip over ipsec.
Same as FreeBSD rev 1.5.
Peter Avalos [Tue, 17 Oct 2006 19:20:49 +0000 (19:20 +0000)]
Fix spelling error.
Submitted-by: Francis GUDIN <fgudin@nerim.net>
Sepherosa Ziehau [Tue, 17 Oct 2006 12:38:25 +0000 (12:38 +0000)]
regen
Sepherosa Ziehau [Tue, 17 Oct 2006 12:37:46 +0000 (12:37 +0000)]
Add device id for USR Robotics 997902 Gigabit Ethernet
Sepherosa Ziehau [Tue, 17 Oct 2006 12:26:52 +0000 (12:26 +0000)]
regen