]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
20 years agoAvoid accidental use of ANSI C trigraphs.
Maxim Konovalov [Sat, 21 Aug 2004 07:23:41 +0000 (07:23 +0000)]
Avoid accidental use of ANSI C trigraphs.

Spotted by: Pawel Worach
Obtained from: NetBSD (rev. 1.11, 1.13)

20 years agoLet GCC know that ___runetype(), ___tolower() and ___toupper() are pure
Tim J. Robbins [Sat, 21 Aug 2004 07:00:40 +0000 (07:00 +0000)]
Let GCC know that ___runetype(), ___tolower() and ___toupper() are pure
functions, allowing it to generate better code for the <ctype.h> and
<wctype.h> functions. For example, it can now keep _CurrentRuneLocale
in a register across calls to these functions, and can delete calls to
___runetype() if the result is already known or not used.

20 years agoFix aperture size detection on some ALi chipsets by only using the lowest 4 bits
Eric Anholt [Sat, 21 Aug 2004 06:24:21 +0000 (06:24 +0000)]
Fix aperture size detection on some ALi chipsets by only using the lowest 4 bits
to check aperture size, avoiding hangs.  Maintain the rest of the bits when
setting/unsetting ATTBASE.  This essentially matches Linux's AGP driver as well.

PR: kern/70037
Submitted by: Mark Tinguely <tinguely at casselton dot net>
Obtained from: NetBSD

20 years agoApply some stylistic changes based off of kern/70037 (content changes to
Eric Anholt [Sat, 21 Aug 2004 06:18:11 +0000 (06:18 +0000)]
Apply some stylistic changes based off of kern/70037 (content changes to
follow).

20 years agoEnter the long awaited start of FreeBSD 6.0!
David E. O'Brien [Sat, 21 Aug 2004 03:58:39 +0000 (03:58 +0000)]
Enter the long awaited start of FreeBSD 6.0!

20 years agoAdd some details about our wonderful system.
David E. O'Brien [Sat, 21 Aug 2004 03:36:47 +0000 (03:36 +0000)]
Add some details about our wonderful system.

20 years agoDocument when this function came into FreeBSD.
David E. O'Brien [Sat, 21 Aug 2004 03:31:17 +0000 (03:31 +0000)]
Document when this function came into FreeBSD.

20 years agoDon't bother calling the module event handlers from module_shutdown()
Don Lewis [Fri, 20 Aug 2004 21:47:48 +0000 (21:47 +0000)]
Don't bother calling the module event handlers from module_shutdown()
in the shutdown_final state if the RB_NOSYNC flag is set.

The specific motivation in this case is that a system panic in an
interrupt context results in a call to module_shutdown(), which
calls g_modevent(), which calls g_malloc(..., M_WAITOK), which
results in a second panic.   While g_modevent() could be fixed to
not call malloc() for MOD_SHUTDOWN events (which it doesn't handle
in any case), it is probably also a good idea to entirely skip the
execution of the module shutdown handlers after a panic.

This may be a MFC candidate for RELENG_5.

20 years agoDon't attempt to trigger the syncer thread final sync code in the
Don Lewis [Fri, 20 Aug 2004 19:21:47 +0000 (19:21 +0000)]
Don't attempt to trigger the syncer thread final sync code in the
shutdown_pre_sync state if the RB_NOSYNC flag is set.  This is the
likely cause of hangs after a system panic that are keeping crash
dumps from being done.

This is a MFC candidate for RELENG_5.

MFC after: 3 days

20 years agoCorrect the args to busdma, mostly cosmetic.
Søren Schmidt [Fri, 20 Aug 2004 19:05:22 +0000 (19:05 +0000)]
Correct the args to busdma, mostly cosmetic.

20 years agoRemove some dead code under a straggling APIC_IO #ifdef that I missed
John Baldwin [Fri, 20 Aug 2004 17:24:52 +0000 (17:24 +0000)]
Remove some dead code under a straggling APIC_IO #ifdef that I missed
back before 5.2.

20 years agoUse the new start for the offset, not the old end.
Nate Lawson [Fri, 20 Aug 2004 17:04:49 +0000 (17:04 +0000)]
Use the new start for the offset, not the old end.

20 years agoAdd description of raidtest tool.
Pawel Jakub Dawidek [Fri, 20 Aug 2004 16:56:43 +0000 (16:56 +0000)]
Add description of raidtest tool.

20 years agoCorrectly handle BIOS resources that are duplicated (!). There are many
Nate Lawson [Fri, 20 Aug 2004 16:52:44 +0000 (16:52 +0000)]
Correctly handle BIOS resources that are duplicated (!).  There are many
systems that have overlapping regions specified in their sysresource
objects.  This patch fixes ATA DMA and acpi_timer allocation for such
sysctems.  It should eventually be moved to resource_list_add() if it is
a valid generalized approach.  The minimal approach for 5.3 is:

"Loop through all current resources to see if the new one overlaps
any existing ones.  If so, the old one always takes precedence and
the new one is adjusted (or rejected).  We check for three cases:

1. Tail of new resource overlaps head of old resource:  truncate the
   new resource so it is contiguous with the start of the old.
2. New resource wholly contained within the old resource:  error.
3. Head of new resource overlaps tail of old resource:  truncate the
   new resource so it is contiguous, following the old."

Tested by: Radek Kozlowski <radek_at_raadradd.com>
Discussed with: imp
MFC after: 4 days

20 years agoRemove a check that is too strict. With BIOSen that specify an IO/ctl port
Nate Lawson [Fri, 20 Aug 2004 16:34:30 +0000 (16:34 +0000)]
Remove a check that is too strict.  With BIOSen that specify an IO/ctl port
of 0x3f2-0x3f5,0x3f7 the ports are not 7 bytes apart.  This should fix
floppy probing on such systems.  (We handle the case of adjusting for
a start of 0x3f2 -> 0x3f0 separately, although that code should still be
checked if there are still floppy problems for others.)

Tested by: Sarunas Vancevicius <vsarunas_at_eircom.net>
MFC after: 3 days

20 years agoBack out uipc_socket.c:1.208, as it incorrectly assumes that all
Robert Watson [Fri, 20 Aug 2004 16:24:23 +0000 (16:24 +0000)]
Back out uipc_socket.c:1.208, as it incorrectly assumes that all
sockets are connection-oriented for the purposes of kqueue
registration.  Since UDP sockets aren't connection-oriented, this
appeared to break a great many things, such as RPC-based
applications and services (i.e., NFS).  Since jmg isn't around I'm
backing this out before too many more feet are shot, but intend to
investigate the right solution with him once he's available.

Apologies to: jmg
Discussed with: imp, scottl

20 years agoRemove unused file.
Poul-Henning Kamp [Fri, 20 Aug 2004 15:15:48 +0000 (15:15 +0000)]
Remove unused file.

20 years agoRewrite of the floppy driver to make it MPsafe & GEOM friendly:
Poul-Henning Kamp [Fri, 20 Aug 2004 15:14:25 +0000 (15:14 +0000)]
Rewrite of the floppy driver to make it MPsafe & GEOM friendly:

Centralize the fdctl_wr() function by adding the offset in
the resource to the softc structure.

Bugfix: Read the drive-change signal from the correct place:
same place as the ctl register.

Remove the cdevsw{} related code and implement a GEOM class.

Ditch the state-engine and park a thread on each controller
to service the queue.

Make the interrupt FAST & MPSAFE since it is just a simple
wakeup(9) call.

Rely on a per controller mutex to protect the bioqueues.
Grab GEOMs topology lock when we have to and Giant when
ISADMA needs it.  Since all access to the hardware is
isolated in the per controller thread, the rest of the
driver is lock & Giant free.

Create a per-drive queue where requests are parked while
the motor spins up.  When the motor is running the requests
are purged to the per controller queue.  This allows
requests to other drives to be serviced during spin-up.

Only setup the motor-off timeout when we finish the last
request on the queue and cancel it when a new request
arrives.  This fixes the bug in the old code where the motor
turned off while we were still retrying a request.

Make the "drive-change" work reliably.  Probe the drive on
first opens.  Probe with a recal and a seek to cyl=1 to
reset the drive change line and check again to see if we
have a media.

When we see the media disappear we destroy the geom provider,
create a new one, and flag that autodetection should happen
next time we see a media (unless a specific format is configured).

Add sysctl tunables for a lot of drive related parameters.
If you spend a lot of time waiting for floppies you can
grab the i82078 pdf from Intels web-page and try tuning
these.

Add sysctl debug.fdc.debugflags which will enable various
kinds of debugging printfs.

Add central definitions of our well known floppy formats.

Simplify datastructures for autoselection of format and
call the code at the right times.

Bugfix: Remove at least one piece of code which would have
made 2.88M floppies not work.

Use implied seeks on enhanced controllers.

Use multisector transfers on all controllers.  Increase
ISADMA bounce buffers accordingly.

Fall back to single sector when retrying.  Reset retry count
on every successful transaction.

Sort functions in a more sensible order and generally tidy
up a fair bit here and there.

Assorted related fixes and adjustments in userland utilities.

WORKAROUNDS:
Do allow r/w opens of r/o media but refuse actual write
operations.  This is necessary until the p4::phk_bufwork
branch gets integrated (This problem relates to remounting
not reopening devices, see sys/*/*/${fs}_vfsops.c for details).

Keep PC98's private copy of the old floppy driver compiling
and presumably working (see below).

TODO (planned)

Move probing of drives until after interrupts/timeouts work
(like for ATA/SCSI drives).

TODO (unplanned)

This driver should be made to work on PC98 as well.

Test on YE-DATA PCMCIA floppy drive.

Fix 2.88M media.

This is a MT5 candidate (depends on the bioq_takefirst() addition).

20 years agoImmortalize the words of a -current reader.
Dag-Erling Smørgrav [Fri, 20 Aug 2004 13:39:28 +0000 (13:39 +0000)]
Immortalize the words of a -current reader.

20 years ago- fflush() standard output before fork()ing.
Pawel Jakub Dawidek [Fri, 20 Aug 2004 12:49:09 +0000 (12:49 +0000)]
- fflush() standard output before fork()ing.
- Remove redundant 'break'.

20 years agoAdd the raidtest tool, which can be used for performance tests of storage devices.
Pawel Jakub Dawidek [Fri, 20 Aug 2004 12:02:34 +0000 (12:02 +0000)]
Add the raidtest tool, which can be used for performance tests of storage devices.
It uses random offsets, random requests size and random operation type (READ or
WRITE). It also allows to run many processes to send I/O requests in parallel.

20 years agoAdd my birthday.
Dejan Lesjak [Fri, 20 Aug 2004 12:01:19 +0000 (12:01 +0000)]
Add my birthday.

Approved by: anholt (mentor)

20 years agoRemove a couple of warnings and tweak an error message.
Dag-Erling Smørgrav [Fri, 20 Aug 2004 10:23:22 +0000 (10:23 +0000)]
Remove a couple of warnings and tweak an error message.

20 years agoCopy open_locked() from tinderbox.pl and use it to optionally acquire a
Dag-Erling Smørgrav [Fri, 20 Aug 2004 08:10:30 +0000 (08:10 +0000)]
Copy open_locked() from tinderbox.pl and use it to optionally acquire a
lock file upon startup.  If this fails, tbmaster will simply terminate.

20 years agodocument that the sk driver now supports the Belkin F5D5005 gige card.
John-Mark Gurney [Fri, 20 Aug 2004 06:27:58 +0000 (06:27 +0000)]
document that the sk driver now supports the Belkin F5D5005 gige card.

20 years agofix LOR's in sk. Original patch from dwhite. This moves the memory
John-Mark Gurney [Fri, 20 Aug 2004 06:22:04 +0000 (06:22 +0000)]
fix LOR's in sk.  Original patch from dwhite.  This moves the memory
allocation earlier on in sk_attach so we don't have to lock until a bit
later.

PR: 69752

20 years agoDONT PANIC
Søren Schmidt [Fri, 20 Aug 2004 06:19:25 +0000 (06:19 +0000)]
DONT PANIC

Only call dmainit() if there is a valid busmaster resource.

20 years agoRevert the previous change. It works great for 4BSD but causes major
Scott Long [Fri, 20 Aug 2004 05:58:38 +0000 (05:58 +0000)]
Revert the previous change.  It works great for 4BSD but causes major
problems for ULE.  The reason is quite unknown and worrisome.

20 years agoReplace the ERRORS section with a DIAGNOSTICS section.
Tom Rhodes [Fri, 20 Aug 2004 05:25:11 +0000 (05:25 +0000)]
Replace the ERRORS section with a DIAGNOSTICS section.

Requested by: ru

20 years agoIn maybe_preempt(), ignore threads that are in an inconsistent state. This
Scott Long [Fri, 20 Aug 2004 05:18:50 +0000 (05:18 +0000)]
In maybe_preempt(), ignore threads that are in an inconsistent state.  This
is an effective band-aid for at least some of the scheduler corruption seen
recently.  The real fix will involve protecting threads while they are
inconsistent, and will come later.

Submitted by: julian

20 years agoput function's name at begining of column...
John-Mark Gurney [Fri, 20 Aug 2004 05:09:31 +0000 (05:09 +0000)]
put function's name at begining of column...

20 years agoReplace the ERRORS section with a DIAGNOSTICS section.
Tom Rhodes [Fri, 20 Aug 2004 05:07:02 +0000 (05:07 +0000)]
Replace the ERRORS section with a DIAGNOSTICS section.

Requested by: ru

20 years agoadd pci id for Belkin F5D5005 Gigabit ethernet card.
John-Mark Gurney [Fri, 20 Aug 2004 05:06:10 +0000 (05:06 +0000)]
add pci id for Belkin F5D5005 Gigabit ethernet card.

20 years agoFix a few style nits.
Tom Rhodes [Fri, 20 Aug 2004 04:52:00 +0000 (04:52 +0000)]
Fix a few style nits.

Noticed by: Mr. mdoc(7) himself, ru

20 years agomake sure that the socket is either accepting connections or is connected
John-Mark Gurney [Fri, 20 Aug 2004 04:15:30 +0000 (04:15 +0000)]
make sure that the socket is either accepting connections or is connected
when attaching a knote to it...  otherwise return EINVAL...

Pointed out by: benno

20 years agoAlign netgraph message fields ready for 64-bit (and 128 bit :-) machines.
Julian Elischer [Fri, 20 Aug 2004 01:24:23 +0000 (01:24 +0000)]
Align netgraph message fields ready for 64-bit (and 128 bit :-) machines.
requires a recompile of netgraph users.
Also change the size of a field in the bluetooth code
that was waiting for the next change that needed recompiles so
it could piggyback its way in.

Submitted by: jdp, maksim
MFC after: 2 days

20 years agoFix a stupid typo which prevented an ipfw KLD unload from successfully cleaning
Andre Oppermann [Fri, 20 Aug 2004 00:36:55 +0000 (00:36 +0000)]
Fix a stupid typo which prevented an ipfw KLD unload from successfully cleaning
up its remains.  Do not terminate 'if' lines with ';'.

Spotted by: claudio@OpenBSD.ORG (sitting 3m from my desk)
Pointy hat to: andre

20 years agobuild libthr on amd64.
David Xu [Thu, 19 Aug 2004 23:51:42 +0000 (23:51 +0000)]
build libthr on amd64.

20 years agoAdjust code to support AMD64, on AMD64, thread needs to set fsbase by
David Xu [Thu, 19 Aug 2004 23:49:04 +0000 (23:49 +0000)]
Adjust code to support AMD64, on AMD64, thread needs to set fsbase by
itself before it can execute any other code, so new thread should be
created with all signals are masked until after fsbase is set.

20 years agoAdd AMD64 support code.
David Xu [Thu, 19 Aug 2004 23:41:12 +0000 (23:41 +0000)]
Add AMD64 support code.

20 years agoWhen unloading ipfw module use callout_drain() to make absolutely sure that
Andre Oppermann [Thu, 19 Aug 2004 23:31:40 +0000 (23:31 +0000)]
When unloading ipfw module use callout_drain() to make absolutely sure that
all callouts are stopped and finished.  Move it before IPFW_LOCK() to avoid
deadlocking when draining callouts.

20 years ago- Remove note about device listings going away.
Simon L. B. Nielsen [Thu, 19 Aug 2004 22:03:20 +0000 (22:03 +0000)]
- Remove note about device listings going away.
- Add a note about possibility for duplicate listing of devices.
- Auto generate device listings for the following drivers: ncr, sym,
  umodem, and uscanner

MFC after: 3 days

20 years agoAdd a HARDWARE section which lists supported devices based on the
Simon L. B. Nielsen [Thu, 19 Aug 2004 21:37:32 +0000 (21:37 +0000)]
Add a HARDWARE section which lists supported devices based on the
Hardware Notes.

MFC after: 3 days

20 years agoAdd a HARDWARE section which lists supported devices.
Simon L. B. Nielsen [Thu, 19 Aug 2004 21:34:31 +0000 (21:34 +0000)]
Add a HARDWARE section which lists supported devices.

MFC after: 3 days

20 years agoAdd a HARDWARE section which lists supported devices.
Simon L. B. Nielsen [Thu, 19 Aug 2004 21:24:45 +0000 (21:24 +0000)]
Add a HARDWARE section which lists supported devices.

MFC after: 3 days

20 years agoUnconditionally support the AMD64 GART HW.
David E. O'Brien [Thu, 19 Aug 2004 20:58:24 +0000 (20:58 +0000)]
Unconditionally support the AMD64 GART HW.

20 years agoMove listing of DTC 3290 to aha(4).
Simon L. B. Nielsen [Thu, 19 Aug 2004 20:49:39 +0000 (20:49 +0000)]
Move listing of DTC 3290 to aha(4).

MFC after: 3 days

20 years agoFor IPv6 access pointer to tcpcb only after we have checked it is valid.
Andre Oppermann [Thu, 19 Aug 2004 20:16:17 +0000 (20:16 +0000)]
For IPv6 access pointer to tcpcb only after we have checked it is valid.

Found by: Coverity's automated analysis (via Ted Unangst)

20 years agoAdd a newline.
Nate Lawson [Thu, 19 Aug 2004 20:16:09 +0000 (20:16 +0000)]
Add a newline.

20 years agoTemporary bandaid to help sparc64 systems with ATA disks boot. Recent
Ken Smith [Thu, 19 Aug 2004 20:13:31 +0000 (20:13 +0000)]
Temporary bandaid to help sparc64 systems with ATA disks boot.  Recent
changes to the ATA driver cause a kernel crash, no fault of the ATA
code.  Work is in progress to add the necessary feature to the sparc64
kernel and this commit will be backed out when it is complete.  This
bandaid is being put in mostly in the interests of getting the first
release snapshot done and out the door.

Tested on: Ultra-10 exhibiting the insta-panic.
MFC: Real Soon

20 years agoAdd bioq_takefirst().
Poul-Henning Kamp [Thu, 19 Aug 2004 19:51:51 +0000 (19:51 +0000)]
Add bioq_takefirst().

If the bioq is empty, NULL is returned.  Otherwise the front element
is removed and returned.

This can simplify locking in many drivers from:

lock()
bp = bioq_first(bq);
if (bp == NULL) {
unlock()
return
}
bioq_remove(bp, bq)
unlock
to:
lock()
bp = bioq_takefirst(bq);
unlock()
if (bp == NULL)
return;

20 years agoFix spelling error in my last blurb.
Andre Oppermann [Thu, 19 Aug 2004 19:45:28 +0000 (19:45 +0000)]
Fix spelling error in my last blurb.

Pointed out by: ru

20 years ago- Note that Adaptec 3940UW won't be recognised by the 164sx SRM
Wilko Bulte [Thu, 19 Aug 2004 19:27:33 +0000 (19:27 +0000)]
- Note that Adaptec 3940UW won't be recognised by the 164sx SRM
- Warn that 164SX is a picky customer with regards to DIMMs.

20 years ago86 new vendor ID's from USB.org, and 3 corrections of existing vendor ID's.
David E. O'Brien [Thu, 19 Aug 2004 19:21:14 +0000 (19:21 +0000)]
86 new vendor ID's from USB.org, and 3 corrections of existing vendor ID's.

20 years agoPut in a note about the ipfw to pfil_hooks conversion and the requirement
Andre Oppermann [Thu, 19 Aug 2004 18:54:54 +0000 (18:54 +0000)]
Put in a note about the ipfw to pfil_hooks conversion and the requirement
of having said in the kernel to be able to use ipfw.

20 years agoDisable interrupts after using pmap_enter() to add the identity mapping.
Nate Lawson [Thu, 19 Aug 2004 18:48:17 +0000 (18:48 +0000)]
Disable interrupts after using pmap_enter() to add the identity mapping.
Since pmap_enter() calls pmap_invalidate_page(), which needs interrupts
enabled in the SMP case, we defer the disable to right before saving the
register context.  This has been incorrect for about a year but caused no
real problems because the identity page never actually replaces a previously
mapped page and suspend/resume on SMP systems has been uncommon.

Tested by: sos
MFC after: 3 days

20 years agoGive a useful error message if someone tries to compile IPFIREWALL into the
Andre Oppermann [Thu, 19 Aug 2004 18:38:23 +0000 (18:38 +0000)]
Give a useful error message if someone tries to compile IPFIREWALL into the
kernel without specifying PFIL_HOOKS as well.

20 years agoAdd comment that IPFIREWALL now requires option PFIL_HOOKS.
Andre Oppermann [Thu, 19 Aug 2004 18:29:55 +0000 (18:29 +0000)]
Add comment that IPFIREWALL now requires option PFIL_HOOKS.

20 years agoNote that IPFIREWALL depends on PFIL_HOOKS compiled into the kernel as well.
Andre Oppermann [Thu, 19 Aug 2004 18:04:10 +0000 (18:04 +0000)]
Note that IPFIREWALL depends on PFIL_HOOKS compiled into the kernel as well.

Submitted by: ceri, simon (mdoc fixes)

20 years agoDo not unconditionally ignore IPDIVERT and IPFIREWALL_FORWARD when building
Andre Oppermann [Thu, 19 Aug 2004 17:59:26 +0000 (17:59 +0000)]
Do not unconditionally ignore IPDIVERT and IPFIREWALL_FORWARD when building
the ipfw KLD.

 For IPFIREWALL_FORWARD this does not have any side effects.  If the module
 has it but not the kernel it just doesn't do anything.

 For IPDIVERT the KLD will be unloadable if the kernel doesn't have IPDIVERT
 compiled in too.  However this is the least disturbing behaviour.  The user
 can just recompile either module or the kernel to match the other one.  The
 access to the machine is not denied if ipfw refuses to load.

20 years agoBring back the sysctl 'net.inet.ip.fw.enable' to unbreak the startup scripts
Andre Oppermann [Thu, 19 Aug 2004 17:38:47 +0000 (17:38 +0000)]
Bring back the sysctl 'net.inet.ip.fw.enable' to unbreak the startup scripts
and to be able to disable ipfw if it was compiled directly into the kernel.

20 years agoAdd debugging to rman_manage_region() as well. This is useful since we
Nate Lawson [Thu, 19 Aug 2004 16:41:12 +0000 (16:41 +0000)]
Add debugging to rman_manage_region() as well.  This is useful since we
manage subregions in ACPI.

MFC after: 3 days

20 years agoReword recent addition about memory moving.
Alfred Perlstein [Thu, 19 Aug 2004 16:34:31 +0000 (16:34 +0000)]
Reword recent addition about memory moving.

Requested by: keramida

Bump .Dd

Requested by: ru

20 years agoAdd my birthday.
Cheng-Lung Sung [Thu, 19 Aug 2004 16:18:39 +0000 (16:18 +0000)]
Add my birthday.

Approved by: leeym (mentor)

20 years agoRemove GIANT_REQUIRED from setugidsafety() as knote_fdclose() no longer
Robert Watson [Thu, 19 Aug 2004 14:59:51 +0000 (14:59 +0000)]
Remove GIANT_REQUIRED from setugidsafety() as knote_fdclose() no longer
requires Giant.

20 years agoAdd a HARDWARE section and bump the document date
Christian Brueffer [Thu, 19 Aug 2004 14:33:53 +0000 (14:33 +0000)]
Add a HARDWARE section and bump the document date

MFC after: 3 days

20 years agoRemove an unnecessary .It.
Maxim Konovalov [Thu, 19 Aug 2004 13:41:06 +0000 (13:41 +0000)]
Remove an unnecessary .It.

Reviewed by: ru

20 years agoCatch up to recent API changes including the removal of the signal_caught
John Baldwin [Thu, 19 Aug 2004 12:46:02 +0000 (12:46 +0000)]
Catch up to recent API changes including the removal of the signal_caught
argument to sleepq_timedwait() and the enhancements to the flags argument
passed to sleepq_add().

20 years agoA volume can be up if it has a degraded RAID5 plex.
Lukas Ertl [Thu, 19 Aug 2004 12:03:27 +0000 (12:03 +0000)]
A volume can be up if it has a degraded RAID5 plex.

20 years agoNow that the return value semantics of cv's for multithreaded processes
John Baldwin [Thu, 19 Aug 2004 11:31:42 +0000 (11:31 +0000)]
Now that the return value semantics of cv's for multithreaded processes
have been unified with that of msleep(9), further refine the sleepq
interface and consolidate some duplicated code:
- Move the pre-sleep checks for theaded processes into a
  thread_sleep_check() function in kern_thread.c.
- Move all handling of TDF_SINTR to be internal to subr_sleepqueue.c.
  Specifically, if a thread is awakened by something other than a signal
  while checking for signals before going to sleep, clear TDF_SINTR in
  sleepq_catch_signals().  This removes a sched_lock lock/unlock combo in
  that edge case during an interruptible sleep.  Also, fix
  sleepq_check_signals() to properly handle the condition if TDF_SINTR is
  clear rather than requiring the callers of the sleepq API to notice
  this edge case and call a non-_sig variant of sleepq_wait().
- Clarify the flags arguments to sleepq_add(), sleepq_signal() and
  sleepq_broadcast() by creating an explicit submask for sleepq types.
  Also, add an explicit SLEEPQ_MSLEEP type rather than a magic number of
  0.  Also, add a SLEEPQ_INTERRUPTIBLE flag for use with sleepq_add() and
  move the setting of TDF_SINTR to sleepq_add() if this flag is set rather
  than sleepq_catch_signals().  Note that it is the caller's responsibility
  to ensure that sleepq_catch_signals() is called if and only if this flag
  is passed to the preceeding sleepq_add().  Note that this also removes a
  sched_lock lock/unlock pair from sleepq_catch_signals().  It also ensures
  that for an interruptible sleep, TDF_SINTR is always set when
  TD_ON_SLEEPQ() is true.

20 years agoBump document date.
Dag-Erling Smørgrav [Thu, 19 Aug 2004 11:25:03 +0000 (11:25 +0000)]
Bump document date.

Reminded by: ru

20 years agoGeneralize the UFS bad magic value used to determine when a filesystem
John Baldwin [Thu, 19 Aug 2004 11:09:13 +0000 (11:09 +0000)]
Generalize the UFS bad magic value used to determine when a filesystem
has only been partly initialized via newfs(8) so that it applies to both
UFS1 and UFS2.

Submitted by: "Xin LI" delphij at frontfree dot net
MFC: maybe?

20 years agoDocument NO_PF and NO_AUTHPF
Christian Brueffer [Thu, 19 Aug 2004 11:00:26 +0000 (11:00 +0000)]
Document NO_PF and NO_AUTHPF

MFC after: 3 days

20 years agoMeet the 6.0-CURRENT era. While here, teach mdoc(7) about the upcoming
Ruslan Ermilov [Thu, 19 Aug 2004 10:41:54 +0000 (10:41 +0000)]
Meet the 6.0-CURRENT era.  While here, teach mdoc(7) about the upcoming
FreeBSD 4.11-RELEASE.

20 years agoMeet the new era of 6.0-CURRENT.
Ruslan Ermilov [Thu, 19 Aug 2004 10:34:59 +0000 (10:34 +0000)]
Meet the new era of 6.0-CURRENT.

20 years agoAdd a NO_BOOT knob to prevent building the boot blocks and loader.
Dag-Erling Smørgrav [Thu, 19 Aug 2004 09:54:28 +0000 (09:54 +0000)]
Add a NO_BOOT knob to prevent building the boot blocks and loader.

Reviewed by: ru
MFC after: 3 days

20 years agoAlways quote variables in tests, to ensure correct evaluation even when
Dag-Erling Smørgrav [Thu, 19 Aug 2004 08:55:24 +0000 (08:55 +0000)]
Always quote variables in tests, to ensure correct evaluation even when
they are empty or undefined.

MFC after: 3 days

20 years agoPrint the patch file name in the warning message for missing patch files.
Dag-Erling Smørgrav [Thu, 19 Aug 2004 07:31:10 +0000 (07:31 +0000)]
Print the patch file name in the warning message for missing patch files.

20 years agoClean up, add update_releng_[45].rc.
Dag-Erling Smørgrav [Thu, 19 Aug 2004 07:28:30 +0000 (07:28 +0000)]
Clean up, add update_releng_[45].rc.

20 years agoRELENG_5 is now available.
Dag-Erling Smørgrav [Thu, 19 Aug 2004 07:27:42 +0000 (07:27 +0000)]
RELENG_5 is now available.

20 years agoBranches of interest are RELENG_4_{8,9,10}.
Dag-Erling Smørgrav [Thu, 19 Aug 2004 07:27:21 +0000 (07:27 +0000)]
Branches of interest are RELENG_4_{8,9,10}.

20 years agoComment out the timeout setting; it doesn't work properly.
Dag-Erling Smørgrav [Thu, 19 Aug 2004 07:26:31 +0000 (07:26 +0000)]
Comment out the timeout setting; it doesn't work properly.

20 years agoadd options MPROF_BUFFERS and MPROF_HASH_SIZE that adjust the sizes of
John-Mark Gurney [Thu, 19 Aug 2004 06:38:26 +0000 (06:38 +0000)]
add options MPROF_BUFFERS and MPROF_HASH_SIZE that adjust the sizes of
the mutex profiling buffers.  Document them in the man page and in NOTES.
Ensure _HASH_SIZE is larger than _BUFFERS with a cpp error.

20 years agoAcquire and release Giant around a call to VOP_BMAP(). (This is a
Alan Cox [Thu, 19 Aug 2004 02:37:12 +0000 (02:37 +0000)]
Acquire and release Giant around a call to VOP_BMAP().  (This is a
prerequisite to any further reduction in Giant's use by vm_fault().)

20 years agoAdd UNP_UNLOCK_ASSERT() to asser that the UNIX domain socket subsystem
Robert Watson [Thu, 19 Aug 2004 01:45:16 +0000 (01:45 +0000)]
Add UNP_UNLOCK_ASSERT() to asser that the UNIX domain socket subsystem
lock is not held.

Rather than annotating that the lock is released after calls to
unp_detach() with a comment, annotate with an assertion.

Assert that the UNIX domain socket subsystem lock is not held when
unp_externalize() and unp_internalize() are called.

20 years agoPush down pcbinfo and inpcb locking from udp_send() into udp_output().
Robert Watson [Thu, 19 Aug 2004 01:13:10 +0000 (01:13 +0000)]
Push down pcbinfo and inpcb locking from udp_send() into udp_output().
This provides greater context for the locking and allows us to avoid
locking the pcbinfo structure if not binding operations will take
place (i.e., already bound, connected, and no expliti sendto()
address).

20 years agoIn in_pcbrehash(), do assert the inpcb lock as well as the pcbinfo lock.
Robert Watson [Thu, 19 Aug 2004 01:11:17 +0000 (01:11 +0000)]
In in_pcbrehash(), do assert the inpcb lock as well as the pcbinfo lock.

20 years agoWe really don't want to receive spoil event for synchroniztion consumers.
Pawel Jakub Dawidek [Wed, 18 Aug 2004 23:33:37 +0000 (23:33 +0000)]
We really don't want to receive spoil event for synchroniztion consumers.

20 years agoBring ia64 back from the dead. After a call one needs to restore the
Marcel Moolenaar [Wed, 18 Aug 2004 23:06:47 +0000 (23:06 +0000)]
Bring ia64 back from the dead. After a call one needs to restore the
GP register, because it's clobbered for calls across load modules. The
previous commit inserted the call to _init_tls() between the call to
atexit() and the restoration of the GP register clobbered by it. Fix:
restore GP before we call _init_tls().

Pointy hat: dfr@

20 years agoSkip entries for GBDE swap devices if they are commented out in /etc/fstab.
Thomas Quinot [Wed, 18 Aug 2004 21:54:40 +0000 (21:54 +0000)]
Skip entries for GBDE swap devices if they are commented out in /etc/fstab.

Reviewed by: des

20 years agoDo not override the class provided dumpconf function.
Poul-Henning Kamp [Wed, 18 Aug 2004 21:42:08 +0000 (21:42 +0000)]
Do not override the class provided dumpconf function.

20 years agoClarify that realloc and reallocf may move the memory allocation.
Alfred Perlstein [Wed, 18 Aug 2004 21:13:15 +0000 (21:13 +0000)]
Clarify that realloc and reallocf may move the memory allocation.

20 years agoPretty print some informational messages.
Lukas Ertl [Wed, 18 Aug 2004 20:43:56 +0000 (20:43 +0000)]
Pretty print some informational messages.

20 years agoFix a stupid bug in the drive taste function: when checking if a
Lukas Ertl [Wed, 18 Aug 2004 20:34:45 +0000 (20:34 +0000)]
Fix a stupid bug in the drive taste function: when checking if a
drive is known to the configuration check also if it already has a geom.
Without this check several needless geoms are created and valid
configuration data was overwritten.

This change obsoletes the need for a separate geom to taste an
offered provider and the consumer doesn't need to be opened with the
exclusive bit set.

20 years agoInitialize iobase from the resource allocated by bus_alloc_resource_any()
Justin T. Gibbs [Wed, 18 Aug 2004 17:17:43 +0000 (17:17 +0000)]
Initialize iobase from the resource allocated by bus_alloc_resource_any()
rather than with isa_get_port().  This value is only used in diagnostics,
but the value we want to print is the value in our resource, not in any
hint.

20 years agoNOP class doesn't operate on metadata, so the spoil event can be safely
Pawel Jakub Dawidek [Wed, 18 Aug 2004 16:58:42 +0000 (16:58 +0000)]
NOP class doesn't operate on metadata, so the spoil event can be safely
ignored.

20 years agoss if_vx through indent, and use ANSI function definitions, prior to adding
Justin T. Gibbs [Wed, 18 Aug 2004 16:56:54 +0000 (16:56 +0000)]
ss if_vx through indent, and use ANSI function definitions, prior to adding
if_media and DMA support to the driver.  The previous style was inconsistent
making it difficult to emulate existing style.

20 years agoDump device status on 'list' command.
Pawel Jakub Dawidek [Wed, 18 Aug 2004 16:46:51 +0000 (16:46 +0000)]
Dump device status on 'list' command.

20 years ago- Add a manual page for graid3(8) utility.
Pawel Jakub Dawidek [Wed, 18 Aug 2004 16:41:30 +0000 (16:41 +0000)]
- Add a manual page for graid3(8) utility.
- Connect it to the build.
- Inform geom(8) about it.

20 years agoIf _CRS fails, assume that it succeeded. The ASUS K8V (and others) defines
Nate Lawson [Wed, 18 Aug 2004 16:39:59 +0000 (16:39 +0000)]
If _CRS fails, assume that it succeeded.  The ASUS K8V (and others) defines
single-entry irq links even though it uses an APIC.  It appears that it
ignores _SRS when in APIC mode but returns a valid irq at other times.