Commit Graph

108599 Commits

Author SHA1 Message Date
Poul-Henning Kamp
a76d8f4ec9 Move the VI_BWAIT flag into no bo_flag element of bufobj and call it BO_WWAIT
Add bufobj_wref(), bufobj_wdrop() and bufobj_wwait() to handle the write
count on a bufobj.  Bufobj_wdrop() replaces vwakeup().

Use these functions all relevant places except in ffs_softdep.c where
the use if interlocked_sleep() makes this impossible.

Rename b_vnbufs to b_bobufs now that we touch all the relevant files anyway.
2004-10-21 15:53:54 +00:00
Poul-Henning Kamp
1bca607b9f Add BO_* macros parallel to VI_* macros for manipulating the bo_mtx.
Initialize the bo_mtx when we allocate a vnode i getnewvnode() For
now we point to the vnodes interlock mutex, that retains the exact
same locking sematics.

Move v_numoutput from vnode to bufobj.  Add renaming macro to
postpone code sweep.
2004-10-21 14:42:31 +00:00
Poul-Henning Kamp
67647b2312 Polish vtruncbuf() to improve readability and style a bit. 2004-10-21 14:13:54 +00:00
Marc Fonvieille
7fb04a6a8d Remove NOLIBPTHREAD=yes since BIND related binaries (dig etc.) require
LIBPTHREAD.

Approved by:	phk
2004-10-21 13:56:24 +00:00
Poul-Henning Kamp
e163395619 Simplify buf_vlist_remove().
Now that we have encapsulated the splaytree related information
into a structure we can eliminate the half of this function.
2004-10-21 13:48:50 +00:00
Poul-Henning Kamp
4c95d87652 Forced commit to get the right commit message:
Add new include file <sys/bufobj.h> which will contain the gory
details on the new buffer-cache object. (see comments in file
about the direction this is moving).

Include it from <sys/vnode.h> for now to avoid munging a lot of files
which can later be munged back.

Embed a bufobj in vnode.

Move the buf splay trees from the vnode to the bufobj.

Alias the fields to avoid sweeping code yet.

Hide vnode and bufobj behind

	#if defined(_KERNEL) || defined(_KVM_VNODE)

to discourage userland voyeurism.
2004-10-21 12:51:36 +00:00
Poul-Henning Kamp
0bf874246b Add new function ttyinitmode() which sets our systemwide default
modes on a tty structure.  Both the ".init" and the current settings
are initialized allowing the function to be used both at attach and
open time.

The function takes an argument to decide if echoing should be enabled
by default.  Echoing should not be enabled for regular physical
serial ports unless they are consoles, in which case they should
be configured by ttyconsolemode() instead.

Use the new function throughout.
2004-10-21 12:24:38 +00:00
Poul-Henning Kamp
4ef77ed68a Add -D_KVM_VNODE to indicate that despite being a userland program,
we want to know the vnode structures internals.
2004-10-21 12:09:45 +00:00
Olivier Houchard
99cf590376 We want to ignore BUS_DMASYNC_POSTWRITE, not BUS_DMASYNC_POSTREAD.
Spotted out by:	mux
Pointy hat to:	cognet
2004-10-21 11:59:33 +00:00
Robert Watson
bda3709718 Bump copyright dates for NETA on these files. 2004-10-21 11:29:56 +00:00
Robert Watson
9dc981da8d Remove unnecessary include of vnode.h.
Requested by:	phk
2004-10-21 11:22:07 +00:00
Robert Watson
f4194603ce Modify libugidfw(3) to use MBI_* permission flags from mac_bsdextended.h
instead of using the V* permission flags from vnode.h.  Remove include
of vnode.h.

Requested by:	phk
2004-10-21 11:21:13 +00:00
Robert Watson
2e74bca132 Modify mac_bsdextended policy so that it defines its own vnode access
right bits rather than piggy-backing on the V* rights defined in
vnode.h.  The mac_bsdextended bits are given the same values as the V*
bits to make the new kernel module binary compatible with the old
version of libugidfw that uses V* bits.  This avoids leaking kernel
API/ABI to user management tools, and in particular should remove the
need for libugidfw to include vnode.h.

Requested by:	phk
2004-10-21 11:19:02 +00:00
Ruslan Ermilov
d885836c32 Pass WORLD_FLAGS down to the distributeworld target, so it's
possible to build releases with customizations such as with
NOPROFILE.

Problem found and fix tested by Roman Neuhauser.
2004-10-21 08:54:10 +00:00
David Xu
39454d368f Decrease reference count if we won't use the thread, this avoids memory
leak under some cases.
2004-10-21 03:42:24 +00:00
Stephan Uphoff
f742a1edcd Zero terminate empty sting in kdb_sysctl_available.
Approved by:    sam (mentor)
MFC after: 1 week
2004-10-21 01:11:25 +00:00
Giorgos Keramidas
8e8b31001c Bump the document date, s/filesystem/file system/ and add missing word.
Noticed by:	ru
2004-10-20 19:05:50 +00:00
Alan Cox
0f777d7d9b Modify the vm object locking in do_sendfile() so that the containing object
is locked when vm_page_io_finish() is called on a page.  This is to satisfy
a new, post-RELENG_5 assertion in vm_page_io_finish().  (I am in the
process of transitioning the responsibility for synchronizing access to
various fields/flags on the page from the global page queues lock to the
per-object lock.)

Tripped over by: obrien@
2004-10-20 17:44:40 +00:00
Giorgos Keramidas
6e1b2264bd Introduce root_rw_mount as a new variable in defaults/rc.conf to
unbreak /etc/rc.d/root for diskless systems that get their root
filesystem from a read-only NFS mount.

PR:		conf/72927
Submitted by:	Ralf Wenk <RZ-FreeBSD1004@fh-karlsruhe.de>
Reviewed by:	brooks
2004-10-20 16:58:28 +00:00
John Baldwin
c05c4140e1 Fix a typo so that this compiles. 2004-10-20 16:22:53 +00:00
John Baldwin
9b94662d59 Fix a typo so that this compiles again. 2004-10-20 16:22:00 +00:00
Maxim Konovalov
49c2dc64da -c flag (cache stats) was killed in rev. 1.68. Fix getopt(3) and
usage() accordingly.

Submitted by:	Pawel Worach
2004-10-20 11:18:31 +00:00
Simon L. B. Nielsen
3127ce5d60 Fix typo.
Submitted by:	Patrick Oonk <patrick@cookie.monster.org>
PR:		docs/72912
MFC after:	1 week
2004-10-20 10:11:12 +00:00
Søren Schmidt
208a8dc227 Do not retry on requests that has lost thier device during reinit.
Should fix hangs on IBM's etc with the fake slave problem.

MFC:
	asap
2004-10-20 10:11:05 +00:00
Robert Watson
60c9762920 Explicitly break out NETA license from Berkeley license to clearly
indicate license grant, as well as to indicate that NETA is asserting
only two clauses, not four clauses.

Requested by:	imp
2004-10-20 08:05:02 +00:00
Ruslan Ermilov
26cd344960 Don't prepend "/dev" if an argument given is an absolute pathname.
Reviewed by:	jhb
2004-10-20 05:43:45 +00:00
Warner Losh
54df531327 Rumor has it that ACCTON EN2216 is also an ne2000 compatible. Make it so. 2004-10-20 04:54:50 +00:00
Maxime Henrion
2534c4f99c Add missing bus_dmamap_sync() calls. If you are using an architecture
with a weak memory model or x86 + PAE (or more specifically, your
driver is using bounce pages) and you have had problems with em(4),
this may fix it.  At least this is needed to have em(4) work properly
on FreeBSD/arm.

Original version by:	cognet
Reviewed by:		tackerman
Tested by:		cognet
2004-10-19 23:31:44 +00:00
Andre Oppermann
57bbe2e1ab Destroy the UMA zone on unload. 2004-10-19 22:51:20 +00:00
Andre Oppermann
2de1a9eb6e Slightly extend the locking during unload to fully cover the protocol
deregistration.  This does not entirely close the race but narrows the
even previously extremely small chance of a race some more.
2004-10-19 22:08:13 +00:00
Robert Watson
279128e295 Annotate a newly introduced race present due to the unloading of
protocols: it is possible for sockets to be created and attached
to the divert protocol between the test for sockets present and
successful unload of the registration handler.  We will need to
explore more mature APIs for unregistering the protocol and then
draining consumers, or an atomic test-and-unregister mechanism.
2004-10-19 21:35:42 +00:00
Andre Oppermann
72584fd2c0 Convert IPDIVERT into a loadable module. This makes use of the dynamic loadability
of protocols.  The call to divert_packet() is done through a function pointer.  All
semantics of IPDIVERT remain intact.  If IPDIVERT is not loaded ipfw will refuse to
install divert rules and  natd will complain about 'protocol not supported'.  Once
it is loaded both will work and accept rules and open the divert socket.  The module
can only be unloaded if no divert sockets are open.  It does not close any divert
sockets when an unload is requested but will return EBUSY instead.
2004-10-19 21:14:57 +00:00
Andre Oppermann
969bb53e80 Properly declare the "net.inet" sysctl subtree. 2004-10-19 21:06:14 +00:00
Andre Oppermann
539be79a9d Pre-emptively define IPPROTO_SPACER to 32767, the same value as PROTO_SPACER
to document that this value is globally assigned for a special purpose and
may not be reused within the IPPROTO number space.
2004-10-19 20:59:01 +00:00
Justin T. Gibbs
a9c6886af6 aic7xxx.h:
Add constants for SPI protocol delays that are needed for
	target mode.

aic7xxx.c:
	Correct a target mode issue that caused an occassional
	spurious REQ to be seen on the bus when performing manual
	message processing (e.g. transfer rate negotiation).

	Enforce phase change bus settle rules with explicit
	delays when performing manual message processing in
	target mode.  The sequencer already did this for
	"fast-path", target mode message processing.
2004-10-19 20:48:06 +00:00
Ruslan Ermilov
09a25e1fb1 - Removed the .CURDIR/.OBJDIR magic, it is not necessary here.
- Let the built-in sys.mk rule produce the "yearistype" script.
- Install zone files with mode 444 (now that the -m option of
  zic(8) has been fixed).
2004-10-19 20:38:49 +00:00
Ruslan Ermilov
a4270a4241 Make the -m option actually work. 2004-10-19 20:30:09 +00:00
Søren Schmidt
61efd2a8b3 Idle the channel earlier in reinit().
Cosmetic change to suspend, dont call tsleep an extra time at exit.
2004-10-19 20:13:38 +00:00
Søren Schmidt
53f19be8fd Cosmetics 2004-10-19 20:11:23 +00:00
Bruce M Simpson
9a3b0b8cdf Really really fix typo this time. 2004-10-19 20:02:07 +00:00
Robert Watson
31302ebf9d Define IFF_LOCKGIANT() and IFF_UNLOCKGIANT() macros, which conditionally
acquire Giant if the passed interface has IFF_NEEDSGIANT set on it.
Modify calls into (ifp)->if_ioctl() in if.c to use these macros in order
to ensure that Giant is held.

MFC after:	3 days
Bumped into by:	jmg
2004-10-19 18:11:55 +00:00
David E. O'Brien
5282e73fd7 This is now WARNS 6 clean on 64-bit platforms. 2004-10-19 18:09:29 +00:00
David E. O'Brien
a152ca4f14 libsmb is now WARNS 2 clean on 64-bit platforms. 2004-10-19 18:07:26 +00:00
David E. O'Brien
ec27f11411 style.Makefile(5) 2004-10-19 18:06:29 +00:00
David E. O'Brien
1ac62e0b31 + Get prototypes for libc functions.
+ Use the correct printf format for size_t.
2004-10-19 17:44:31 +00:00
David E. O'Brien
caef65b837 Size matters. Correctly use a size_t so 64-bit hosts can mount SMB FS's
when using character set conversions.
Also include POSIX <string.h> vs. BSD <strings.h> now that we've broken
traditional BSD behavior [and compatibility with our BSD brethren].

PR:		72445
Submitted by:	Vladimir Nechitailo <nechit@lpi.ru>
Patch by:	Stasys Smailys <ssmailys@komvista.lt>
2004-10-19 17:39:15 +00:00
Ruslan Ermilov
4c2cbf47fb Apply README guidelines (no tabs). 2004-10-19 17:32:56 +00:00
David E. O'Brien
d41f6f569c Define "I386_CPU" if CPUTYPE is 'i386'. Userland bits can check for "I386_CPU"
to determine if they should select code paths suitable for the 80386 CPU.

Suggested by:	ru
2004-10-19 17:25:33 +00:00
Bruce M Simpson
80f7b487a1 Fix typo sc -> dev. 2004-10-19 16:47:53 +00:00
Andre Oppermann
dff3237ee5 Make use of the PROTO_SPACER functionality for dynamically loadable
protocols in inetsw[] and define initially eight spacer slots.

Remove conflicting declaration 'struct pr_usrreqs nousrreqs'.  It is
now declared and initialized in kern/uipc_domain.c.
2004-10-19 15:58:22 +00:00