Joerg Sonnenberger [Thu, 5 Feb 2004 17:51:44 +0000 (17:51 +0000)]
Second part of the firewire sync. Add defined(__DragonFly__) or
defined(__FreeBSD__) to all checks of __FreeBSD_version to prepare
the kernel for GCC 3.3.
Joerg Sonnenberger [Thu, 5 Feb 2004 14:32:53 +0000 (14:32 +0000)]
Add missing sbp.h to complete the firewire sync
Joerg Sonnenberger [Thu, 5 Feb 2004 13:32:08 +0000 (13:32 +0000)]
Sync DragonFly and FreeBSD-current's FireWire driver.
Patch submitted by Hidetoshi Shimokawa <simokawa@sat.t.u-tokyo.ac.jp>
Modified to leave the sbp hack inplace.
Joerg Sonnenberger [Thu, 5 Feb 2004 11:51:54 +0000 (11:51 +0000)]
Update style(9) to reflect current code practise.
Patch submitted by Chris Pressey <cpressey@catseye.mine.nu>
Robert Garrett [Thu, 5 Feb 2004 04:22:22 +0000 (04:22 +0000)]
Allow world to build with gcc3, basically force openssl to use gettimeofday
instead of ftime()
Robert Garrett [Wed, 4 Feb 2004 21:09:17 +0000 (21:09 +0000)]
Disable -fformat-extensions temporarily
Joerg Sonnenberger [Wed, 4 Feb 2004 17:40:01 +0000 (17:40 +0000)]
Fix various warnings.
Patch submitted by Simon "corecode" Schubert.
Robert Garrett [Wed, 4 Feb 2004 15:39:54 +0000 (15:39 +0000)]
more __FreeBSD__ and __FreeBSD__ -> __DragonFly__ work.
with this commit, world now builds with gcc3,
however this is not completed, until the kernel
files are cleaned, and the __FreeBSD__ definitions removed,
as well as contrib being fixed.
Robert Garrett [Wed, 4 Feb 2004 14:36:11 +0000 (14:36 +0000)]
define __FreeBSD__ for things that use kernel source that needs it for now
Robert Garrett [Wed, 4 Feb 2004 14:34:36 +0000 (14:34 +0000)]
__FreeBSD__ defines
Robert Garrett [Wed, 4 Feb 2004 01:02:14 +0000 (01:02 +0000)]
Define __FreeBSD__ in the make file for ipmon
Robert Garrett [Wed, 4 Feb 2004 00:52:55 +0000 (00:52 +0000)]
__FreeBSD__ -> __DragonFly__
Robert Garrett [Wed, 4 Feb 2004 00:48:39 +0000 (00:48 +0000)]
dhclient needs __FreeBSD__ defined
Robert Garrett [Tue, 3 Feb 2004 23:24:50 +0000 (23:24 +0000)]
__FreeBSD__ -> __DragonFly__
Matthew Dillon [Tue, 3 Feb 2004 19:22:59 +0000 (19:22 +0000)]
More __FreeBSD__ -> __DragonFly__ translation
Matthew Dillon [Tue, 3 Feb 2004 19:06:06 +0000 (19:06 +0000)]
This should fix the dependancy loop for sure.
Matthew Dillon [Tue, 3 Feb 2004 18:54:01 +0000 (18:54 +0000)]
Fake __FreeBSD__ for sendmail contrib use.
Matthew Dillon [Tue, 3 Feb 2004 18:06:06 +0000 (18:06 +0000)]
Try to clean up more of GCC's xrealloc and xmalloc wrapper mess so
buildworld builds.
Robert Garrett [Tue, 3 Feb 2004 10:14:15 +0000 (10:14 +0000)]
grep needs __FreeBSD__ defined
Robert Garrett [Tue, 3 Feb 2004 10:04:04 +0000 (10:04 +0000)]
define __FreeBSD__ for diff
Robert Garrett [Tue, 3 Feb 2004 09:21:23 +0000 (09:21 +0000)]
Add warning about memory usage to resident.conf and examples to both files
Robert Garrett [Tue, 3 Feb 2004 08:43:41 +0000 (08:43 +0000)]
change FreeBSD in examples to DragonFly
Robert Garrett [Tue, 3 Feb 2004 08:37:33 +0000 (08:37 +0000)]
Update a few macros to print DragonFly instead of FreeBSD
Matthew Dillon [Tue, 3 Feb 2004 07:34:10 +0000 (07:34 +0000)]
Remove obsolete __FreeBSD_version checks, add __DragonFly__ tests to
existing __FreeBSD__ checks.
Robert Garrett [Tue, 3 Feb 2004 07:13:10 +0000 (07:13 +0000)]
Add DragonFly partition type and change FreeBSD in warnings to DragonFly
Matthew Dillon [Tue, 3 Feb 2004 07:11:53 +0000 (07:11 +0000)]
Change __FreeBSD__ -> __DragonFly__
Matthew Dillon [Tue, 3 Feb 2004 06:26:32 +0000 (06:26 +0000)]
Fake __FreeBSD__ for various contrib/ code that needs it.
Matthew Dillon [Tue, 3 Feb 2004 05:51:51 +0000 (05:51 +0000)]
Adjust osreldate.h to only set __FreeBSD_version if __FreeBSD__ is set,
so we don't pollute the namespace when only __DragonFly__ is set.
__DragonFly_version is always set.
Robert Garrett [Tue, 3 Feb 2004 05:26:48 +0000 (05:26 +0000)]
Forgot the contribited by line
Contributed by:
Dylan Reinhold
Matthew Dillon [Tue, 3 Feb 2004 05:12:51 +0000 (05:12 +0000)]
Add -DSETPWENT_VOID to CFLAGS to work around a __FreeBSD__ conditional in
contrib/bind.
Robert Garrett [Tue, 3 Feb 2004 04:50:47 +0000 (04:50 +0000)]
update this file correctly
I munged the patch the first time round
Matthew Dillon [Tue, 3 Feb 2004 04:26:01 +0000 (04:26 +0000)]
Make the 'realclean' target respect KERNCONF.
Submitted-by: Peter Kadau <peter.kadau@tuebingen.mpg.de>
Matthew Dillon [Tue, 3 Feb 2004 04:23:25 +0000 (04:23 +0000)]
Adjust __FreeBSD__ -> __DragonFly_
Matthew Dillon [Tue, 3 Feb 2004 03:47:12 +0000 (03:47 +0000)]
Undo the xmalloc->malloc optimization FreeBSD made in certain cases
where xmalloc gets defined to malloc which gets defined to xmalloc which
gets defined to malloc again which gets defined to ... and so on.
Culprits: parser files, gcc/system.h, dragonfly-native.h, and various
Makefile's.
Add [DragonFly] to gcc3's version ident.
Matthew Dillon [Mon, 2 Feb 2004 23:35:12 +0000 (23:35 +0000)]
Update the GCC3 infrastructure Stage 1/2. This commit generates the basic
infrastructure needed to support /usr/src/contrib/gcc-3.3 (a pristine gcc
dist) intead of /usr/src/contrib/gcc3 (munged version taken from FreeBSD-5).
This stage does not have all the fbsd/dfly specific extensions, such as
-fformat-extensions.
Matthew Dillon [Mon, 2 Feb 2004 08:05:56 +0000 (08:05 +0000)]
AMD64 infrastructure stage 1/99. This is just a preliminary commit. Many
of these files are being brought in from FreeBSD-5. The noteable exception
is sys/amd64/pmap.h, which contains some #if 0'd FreeBSD-5 stuff but has
the beginnings of a new MMU infrastructure. NOTHING WILL BUILD YET!
The DFly 64 bit AMD work will have a lot of similarities to the FreeBSD-5
work, but the PMAP/MMU/VM subsystem support is going to be written from
scratch and diverge wildly from F-5.
In particular, it is my intent to entirely do-away with %fs for per-cpu
storage references and instead use the top level PML4 page table (512GB
per entry, 256TB total VM) as a fixed per-cpu entity. Each cpu will get
its own PML4 table. Initially only one entry (512GB of VM) in PML4 will
be used to represent UVM. This is the entry that the thread switch code
changes out. Other entries will be used:
PML4 TABLE (each cpu gets its own table)
* 1 switched entry 512GB of UVM (expandable to more)
* 1 static entry 512GB of KVM
* 1 static entry to map physical memory
* 1 per-cpu entry to map the per-cpu globaldata space.
* Additional entries for recursive page tables, third party
process address space caches (userspace<->userspace copying),
and other features.
This should greatly simplify the PMAP code... no copies of the 'high' kernel
PDE entries, for example, and we still have the flexibility of switching out
the whole PML4 table at a later date without interfering with the basic
feature set.
Matthew Dillon [Mon, 2 Feb 2004 07:32:30 +0000 (07:32 +0000)]
binutils214 stage 2/4 (continued). remove /usr/bin/gcc{2,3} from
BSD.usr.dist.
Matthew Dillon [Mon, 2 Feb 2004 07:29:28 +0000 (07:29 +0000)]
Merge from vendor branch BINUTILS:
Bring GNU binutils-2.14. This commit is an exact copy of the contents of
binutils-2.14.tar.gz. No files have been modified. Files that we do not
currently used have been deleted and two files, README.DRAGONFLY and
README.DELETED have been added.
Matthew Dillon [Mon, 2 Feb 2004 07:29:28 +0000 (07:29 +0000)]
Bring GNU binutils-2.14. This commit is an exact copy of the contents of
binutils-2.14.tar.gz. No files have been modified. Files that we do not
currently used have been deleted and two files, README.DRAGONFLY and
README.DELETED have been added.
Matthew Dillon [Mon, 2 Feb 2004 05:43:16 +0000 (05:43 +0000)]
binutils214 stage 2/4.
* Normalize binutils212 and binutils214. Remove /usr/bin/gcc{2,3}/*
and move those binaries to /usr/libexec/gcc{2,3}. Adjust objformat
to suit.
* Hack up contrib/gcc3 to accomodoate library path requirements and fix
STANDARD_BINDIR_PREFIX. Add a missing trailing '/' in STANDARD_EXEC_PREFIX.
* Adjust /usr/src/etc/Makefile (upgrade_etc) to remove old GCC and BINUTILS
junk.
* Cleanup etc/mtree/BSD.usr.dist
* Only build gcc2 for non-cross-compiled i386 native builds. Do not
build libc_r for non-i386 architectures for now. Allow on sysctl based
setproctitle() updates for non-i386 architectures for now.
* Move ldscripts from /usr/libdata/gcc{2,3}/ldscripts into
/usr/libexec/binutilsXXX/ldscripts and remove /usr/libdata/gcc{2,3}.
* GCC vendor obrien->dillon, change TARGET_OS in various places from
freebsd -> dragonfly.
* Consolidate config.h differences for cross builds in the Makefiles so
we need only have one config.h for binutils214/as.
* Cleanup and fix default target BFD for the assembler and linker when
cross-building.
* Build both the 32 and 64 bit support BFD's for binutils214 even if
just doing a native build.
* Remove a bunch of alpha junk (incomplete).
* Note gnu/usr.bin/binutils214/libbfd/dfly_targmatch.h contains aliases
to match dragonfly BFD targets. Some are probably unnecessary but I
had to move on.
* Bring in a bunch of AMD64 related files (incomplete).
* Update rpcgen to locate cpp with the current path instead of /usr/bin/cpp.
Add rpcgen to the buildworld bootstrap tools.
Obtained-From: A good chunk of the amd64-specific files were adapted from
FreeBSD-5 (authored by Peter Wemm and others).
Matthew Dillon [Sun, 1 Feb 2004 19:51:45 +0000 (19:51 +0000)]
Don't use wildcards in csh expansions to avoid early termination of
the upgrade.
Robert Garrett [Sun, 1 Feb 2004 10:25:48 +0000 (10:25 +0000)]
Add much needed scripts to /etc/rc.d
resident
takes a list of pathnames or filenames from /etc/resident.conf
and makes them resident
varsym
takes a list of variable definitions from /etc/varsym.conf
and makes them system wide varsyms
these are currently off by default
thanks to Eirik Nygaard for his assistance
Matthew Dillon [Sun, 1 Feb 2004 08:53:05 +0000 (08:53 +0000)]
binutils214 stage 1/4. Bring in the build infrastructure (left untied from
system). This infrastructure can handle i386 but the amd64 code is not
yet complete.
Robert Garrett [Sun, 1 Feb 2004 03:32:08 +0000 (03:32 +0000)]
only update motd if the first word of the first line is Dragonfly
Submitted by:
Dylan Reinhold
Eirik Nygaard [Sat, 31 Jan 2004 18:08:03 +0000 (18:08 +0000)]
Readd ending '\' so it compiles the cross build tools
Joerg Sonnenberger [Sat, 31 Jan 2004 17:14:40 +0000 (17:14 +0000)]
Build lwkt_process_ipiq_frame only for the kernel, since
struct intrframe doesn't exist in userspace and breaks libcaps.
Joerg Sonnenberger [Sat, 31 Jan 2004 13:38:48 +0000 (13:38 +0000)]
Add getopt_long from NetBSD
Eirik Nygaard [Sat, 31 Jan 2004 11:56:26 +0000 (11:56 +0000)]
Remove genassym and gensetdefs from the cross build as well.
Matthew Dillon [Sat, 31 Jan 2004 09:14:24 +0000 (09:14 +0000)]
Make gcc3 the default for non-i386 architectures, leave gcc2 the default
for i386 architectures.
Matthew Dillon [Sat, 31 Jan 2004 08:59:01 +0000 (08:59 +0000)]
remove genassym and gensetdefs in the upgrade_etc target.
Matthew Dillon [Sat, 31 Jan 2004 08:56:18 +0000 (08:56 +0000)]
Remove genassym and gensetdefs. These programs are obsolete and no longer
used.
Jeroen Ruigrok/asmodai [Sat, 31 Jan 2004 08:52:16 +0000 (08:52 +0000)]
Cut our umbilical cord from mother FreeBSD.
Reword some sentences.
Add some new directories.
Matthew Dillon [Sat, 31 Jan 2004 06:56:46 +0000 (06:56 +0000)]
MAINTAINER lines in Makefile's are no longer applicable, remove them.
Matthew Dillon [Sat, 31 Jan 2004 03:26:56 +0000 (03:26 +0000)]
Remove limitations on the 'machine' specification.
Loosely-based-on: FreeBSD-5.x
David Rhodus [Sat, 31 Jan 2004 01:09:19 +0000 (01:09 +0000)]
* Sync pages with FreeBSD 4.x
Sent in by: Christian Brueffer
Matthew Dillon [Fri, 30 Jan 2004 22:38:47 +0000 (22:38 +0000)]
Update miscellanious firewire manual pages from FreeBSD-4.x.
Submitted-by: Christian Brueffer <chris@unixpages.org>
Taken-from: FreeBSD-4.x
Matthew Dillon [Fri, 30 Jan 2004 18:21:18 +0000 (18:21 +0000)]
Clean up the sys_nsig verses NSIG code to handle additional fault
cases and reformat 'trap -l' output to handle the additional signals.
Report-by: Nate Lawson <nate@root.org>, Bruce Evans <bde@zeta.org.au>
David Rhodus [Fri, 30 Jan 2004 13:23:55 +0000 (13:23 +0000)]
* Add kern_systimer.c to the kernel build process.
Jeroen Ruigrok/asmodai [Fri, 30 Jan 2004 11:31:17 +0000 (11:31 +0000)]
Update to latest version.
Submitted/requested by: Xin LI <delphij@frontfree.net>
Matthew Dillon [Fri, 30 Jan 2004 06:20:08 +0000 (06:20 +0000)]
This commit represents a major revamping of the clock interrupt and timebase
infrastructure in DragonFly.
(fix syntax error in conditionalized section of kern_ntptime.c)
Matthew Dillon [Fri, 30 Jan 2004 06:18:28 +0000 (06:18 +0000)]
This commit represents a major revamping of the clock interrupt and timebase
infrastructure in DragonFly.
(missing nfs files)
Matthew Dillon [Fri, 30 Jan 2004 05:42:18 +0000 (05:42 +0000)]
This commit represents a major revamping of the clock interrupt and timebase
infrastructure in DragonFly.
* Rip out the existing 8254 timer 0 code, and also disable the use of
Timer 2 (which means that the PC speaker will no longer go beep). Timer 0
used to represent a periodic interrupt and a great deal of code was in
place to attempt to obtain a timebase off of that periodic interrupt.
Timer 0 is now used in software retriggerable one-shot mode to produce
variable-delay interrupts. A new hardware interrupt clock abstraction
called SYSTIMERS has been introduced which allows threads to register
periodic or one-shot interrupt/IPI callbacks at approximately 1uS
granularity.
Timer 2 is now set in continuous periodic mode with a period of 65536
and provides the timebase for the system, abstracted to 32 bits.
All the old platform-integrated hardclock() and statclock() code has
been rewritten. The old IPI forwarding code has been #if 0'd out and
will soon be entirely removed (the systimer abstraction takes care of
multi-cpu registrations now). The architecture-specific clkintr() now
simply calls an entry point into the systimer and provides a Timer 0
reload and Timer 2 timebase function API.
* On both UP and SMP systems, cpus register systimer interrupts for the Hz
interrupt, the stat interrupt, and the scheduler round-robin interrupt.
The abstraction is carefully designed to allow multiple interrupts occuring
at the same time to be processed in a single hardware interrupt. While
we currently use IPI's to distribute requested interrupts from other cpu's,
the intent is to use the abstraction to take advantage of per-cpu timers
when available (e.g. on the LAPIC) in the future.
systimer interrupts run OUTSIDE THE MP LOCK. Entry points may be called
from the hard interrupt or via an IPI message (IPI messages have always
run outside the MP lock).
* Rip out timecounters and disable alternative timecounter code for other
time sources. This is temporary. Eventually other time sources, such as
the TSC, will be reintegrated as independant, parallel-running entities.
There will be no 'time switching' per-say, subsystems will be able to
select which timebase they wish to use. It is desireable to reintegrate
at least the TSC to improve [get]{micro,nano}[up]time() performance.
WARNING: PPS events may not work properly. They were not removed, but
they have not been retested with the new code either.
* Remove spl protection around [get]{micro,nano}[up]time() calls, they are
now internally protected.
* Use uptime instead of realtime in certain CAM timeout tests
* Remove struct clockframe. Use struct intrframe everywhere where clockframe
used to be used.
* Replace most splstatclock() protections with crit_*() protections, because
such protections must now also protect against IPI messaging interrupts.
* Add fields to the per-cpu globaldata structure to access timebase related
information using only a critical section rather then a mutex. However,
the 8254 Timer 2 access code still uses spin locks. More work needs to
be done here, the 'realtime' correction is still done in a single global
'struct timespec basetime' structure.
* Remove the CLKINTR_PENDING icu and apic interrupt hacks.
* Augment the IPI Messaging code to make an intrframe available to callbacks.
* Document 8254 timing modes in i386/sai/timerreg.h. Note that at the
moment we assume an 8254 instead of an 8253 as we are using TIMER_SWSTROBE
mode. This may or may not have to be changed to an 8253 mode.
* Integrate the NTP correction code into the new timebase subsystem.
* Separate boottime from basettime. Once boottime is believed to be stable
it is no longer effected by NTP or other time corrections.
CAVETS:
* PC speaker no longer works
* Profiling interrupt rate not increased (it needs work to be
made operational on a per-cpu basis rather then system-wide).
* The native timebase API is function-based, but currently hardwired.
* There might or might not be issues with 486 systems due to the
timer mode I am using.
Matthew Dillon [Fri, 30 Jan 2004 02:35:02 +0000 (02:35 +0000)]
This should hopefully fix current issues with bootstrap buildworlds from
4.x and older versions of DFly, and with user-specified CCVER and CPUTYPE
environment variables.
Explicitly override CCVER in buildworld in a backwards compatible fashion
when building gcc2's and gcc3's startup libraries. Also override CCVER
in gnu/lib/gcc{2,3}/Makefile when bootstrapping from older DFly or FBsd
systems which do not implement the '.makeenv' make directive.
In bsd.init.mk detect when CCVER is changed by a Makefile or Makefile.inc and
rerun bsd.cpu.mk to update the _CPUCFLAGS set, which is assigned in bsd.cpu.mk.
We relay on the fact that ${} variables are not resolved until rules accessing
them are actually run.
Clean up objformat (no operational changes).
Matthew Dillon [Thu, 29 Jan 2004 22:40:36 +0000 (22:40 +0000)]
Set a variable indicating that we support .makeenv, so we can conditionalize
its use.
Matthew Dillon [Thu, 29 Jan 2004 02:26:46 +0000 (02:26 +0000)]
oops, undo part of the last commit. cpp, f77 need their gccX frontends in
/usr/bin/gccX
Matthew Dillon [Thu, 29 Jan 2004 01:56:05 +0000 (01:56 +0000)]
Add back a directory path that buildworld needs, remove /usr/libexec/gcc2
from mtree.
Matthew Dillon [Thu, 29 Jan 2004 01:43:47 +0000 (01:43 +0000)]
have upgrade / upgrade_etc remove /usr/libexec/gcc2 now as well, it has
been moved to /usr/libexec/binutils212.
Matthew Dillon [Thu, 29 Jan 2004 01:39:48 +0000 (01:39 +0000)]
Remove the remainder of Makefile.sub handling. Make objformat aware of the
new binutils location for cpp and f77. Make contrib/gcc aware of the new
binutils locations as specified in /usr/src/gnu/usr.bin/cc/Makefile.inc.
This is hardwired at the moment.
This fixes dependancies on the old /usr/libexec/gcc2 directory.
Matthew Dillon [Wed, 28 Jan 2004 21:26:55 +0000 (21:26 +0000)]
Rename .sub files back to .inc.
Matthew Dillon [Wed, 28 Jan 2004 19:27:01 +0000 (19:27 +0000)]
Rename Makefile.sub back to Makefile.inc to fix cross builds.
Matthew Dillon [Wed, 28 Jan 2004 19:24:14 +0000 (19:24 +0000)]
Undo some of the previously made changes to deal with cross build issues
from 4.x and older versions of DFly.
sys.mk should *NOT* include bsd.init.mk. That is way, way too early and
blows up .inc files, e.g. such as those used by sys/boot.
Rename .sub back to .inc in sys/boot (more on the way) so it will cross build
from 4.x.
Joerg Sonnenberger [Wed, 28 Jan 2004 16:41:18 +0000 (16:41 +0000)]
Move binutils from /usr/libexec/gcc2/{aout,elf} to
/usr/libexec/binutils212/{aout,elf}.
Move GCC 2.95.4 C++ headers from /usr/include/g++ to
/usr/include/c++/2.95.
Joerg Sonnenberger [Wed, 28 Jan 2004 16:25:29 +0000 (16:25 +0000)]
Fix various warnings and other glitches.
Patch submitted by Simon "corecode" Schubert.
Joerg Sonnenberger [Wed, 28 Jan 2004 14:36:00 +0000 (14:36 +0000)]
Change functin definitions to conform to style(9)
Joerg Sonnenberger [Wed, 28 Jan 2004 12:48:49 +0000 (12:48 +0000)]
Add support for the AMD 8111 chipset
Matthew Dillon [Wed, 28 Jan 2004 03:52:28 +0000 (03:52 +0000)]
Fix the DDB 'trace' command. When segment address adjustments were added
in 1.7 so VM86 code could be properly traced, the routine that pulls out
%ebp was accidently changed to return %esp instead, causing DDB
backtraces to produce incorrect results.
Matthew Dillon [Tue, 27 Jan 2004 23:56:48 +0000 (23:56 +0000)]
Try to work-around a DFly-specific crash that can occur in ufs_ihashget()
if the underlying vnode is being reclaimed at the same time. Bump the
vnodes ref count to interlock against vget's VXLOCK test.
Matthew Dillon [Tue, 27 Jan 2004 21:03:08 +0000 (21:03 +0000)]
func() -> func(void) style.
Robert Garrett [Tue, 27 Jan 2004 19:01:21 +0000 (19:01 +0000)]
remove ifwatchd we don't have the binary we don't need the script
Matthew Dillon [Tue, 27 Jan 2004 18:36:52 +0000 (18:36 +0000)]
Describe variant symlinks in the 'ln' manual page.
With-some-changes-by: dillon
Submitted-by: Chris Pressey <cpressey@catseye.mine.nu>
Matthew Dillon [Tue, 27 Jan 2004 18:28:22 +0000 (18:28 +0000)]
Fix type-o's and add a few files to the cleanup list.
Submitted-by: Chris Pressey <cpressey@catseye.mine.nu>
Matthew Dillon [Tue, 27 Jan 2004 18:26:43 +0000 (18:26 +0000)]
USB storage devices are standard fare these days, add device atapicam to
GENERIC.
Joerg Sonnenberger [Tue, 27 Jan 2004 16:26:34 +0000 (16:26 +0000)]
Rename Makefile.inc to Makefile.sub
Robert Garrett [Tue, 27 Jan 2004 01:26:25 +0000 (01:26 +0000)]
fix bug where i accidently left ;; at the end of two case switches
Robert Garrett [Tue, 27 Jan 2004 00:59:21 +0000 (00:59 +0000)]
finish updating rcng to remove FreeBSD and NetBSD code
Robert Garrett [Tue, 27 Jan 2004 00:55:43 +0000 (00:55 +0000)]
update rcng to remove FreeBSD and NetBSD
Robert Garrett [Tue, 27 Jan 2004 00:42:46 +0000 (00:42 +0000)]
update rcng to remove FreeBSD and NetBSD
Matthew Dillon [Mon, 26 Jan 2004 20:58:11 +0000 (20:58 +0000)]
Fix a serious bug in the NTPD loopfilter. Basically what happens is that
when the frequency error is low (e.g. like 15ppm), but an offset correction
must be made (e.g. like 3 seconds), the rstclock() call loopfilter makes
in the S_NSET state, after adjusting the system time, will reload a
clock_offset value PRIOR to the system time adjustment actually having taken
place, because rstclock() uses offset information prior to the time step, not
after, and because the kernel may slew the time step over several seconds.
The result is that clock_offset will wind up being very high (e.g. like 3
seconds) as the loopfilter enters into the S_FREQ state. This will cause
the loopfilter to make a wildly incorrect clock_frequency calculation and
in my tests it can take ntpd and the kernel upwards of one to two DAYS to
correct for the bad calculation.
The fix is to introduce an offset settling state, S_PFREQ, which resets
the offset calculations CLOCK_MINSEC seconds after the correction was made
in order to give the next state, S_FREQ, a good basis for calculating the
frequency drift. clock_offset will be far more reasonable when the S_FREQ
state is entered, a good clock_frequency calculation will be made, and
ntpd will lock onto the frequency is less then half an hour (instead of
24+ hours).
Joerg Sonnenberger [Mon, 26 Jan 2004 19:57:36 +0000 (19:57 +0000)]
Remove Makefile.inc
Joerg Sonnenberger [Mon, 26 Jan 2004 19:49:32 +0000 (19:49 +0000)]
Rename Makefile.in to Makefile.sub
Joerg Sonnenberger [Mon, 26 Jan 2004 19:22:47 +0000 (19:22 +0000)]
Rename Makefile.inc to Makefile.sub
Joerg Sonnenberger [Mon, 26 Jan 2004 19:09:49 +0000 (19:09 +0000)]
To let the crazy Perl Makefiles work, add the support for ../Makefile.sub
which is included at the time ../Makefile.inc originally was.
David Rhodus [Mon, 26 Jan 2004 18:16:18 +0000 (18:16 +0000)]
* DragonFly<-FreeBSD Name clean ups
David Rhodus [Mon, 26 Jan 2004 18:14:43 +0000 (18:14 +0000)]
* Change FreeBSD ifdef over to DragonFly
* Clean up compiler warning in phantasia
David Rhodus [Mon, 26 Jan 2004 18:12:54 +0000 (18:12 +0000)]
* Change FreeBSD ifdefs to DragonFly
Joerg Sonnenberger [Mon, 26 Jan 2004 18:04:03 +0000 (18:04 +0000)]
Really remove C++ from libc_r
Joerg Sonnenberger [Mon, 26 Jan 2004 17:50:34 +0000 (17:50 +0000)]
Remove C++ from libc_r. Use __attribute__((constructor)) instead.
Robert Garrett [Mon, 26 Jan 2004 17:21:15 +0000 (17:21 +0000)]
remove NetBSD and FreeBSD from rcng scripts
Joerg Sonnenberger [Mon, 26 Jan 2004 16:13:49 +0000 (16:13 +0000)]
Always include bsd.init.mk to fix CPUTYPE evaluation
Joerg Sonnenberger [Mon, 26 Jan 2004 15:42:36 +0000 (15:42 +0000)]
Fix make complaining about CCVER being not defined