Commit Graph

36422 Commits

Author SHA1 Message Date
Greg Lehey
1ae808e860 Correct ioctl calls VINUM_STARTCONFIG. The last commit broke the
'vinum start' command.
1999-03-23 23:24:00 +00:00
Bill Fumerola
26bb956563 Remove duplicate line.
Reviewed by:	eivind
1999-03-23 23:01:15 +00:00
Nick Hibma
d0ef08ba15 Check whether the PIRQD enable bit is set in LegSup and if not, set it.
On my PIIX4 chip rev 1 this is required. Have not had any complaints from
other people so it might be a problem with this stepping.
1999-03-23 21:37:45 +00:00
Nick Hibma
7dfd509563 1) clean up debugging output
2) remove erroneous querying of PCI LegSup register while it should be
   fetched from PCI config space
3) change name of define for this register
1999-03-23 21:35:57 +00:00
Doug Rabson
6ca34d85be Call the module's unload handler before removing the device from the
cdevsw list.  This allows a handler to veto the load without losing its
place in the list.

PR:	kern/10653
1999-03-23 21:11:47 +00:00
Peter Wemm
ce04fea445 Remove last remaining references to malloc/realloc and functions that
call them.  All the execX() libc functions should be vfork() safe now.
Specifically:
- execlp() does the argument count-and-build into a vector from alloca
    (like the others) - buildargv() is no longer used (and gone).
- execvp() uses alloca/strcpy rather than strdup().
- the ENOEXEC handler uses alloca rather than malloc.
- a couple of free() calls removed - alloca works on the local stack and
    the allocations are freed on function exit (which is why buildargv
    wasn't useful - it's alloca() context would disappear on return).
Along the way:
- If alloca() fails (can it?), set errno = ENOMEM explicitly.
- The ENOEXEC recovery routine that trys again with /bin/sh appeared to
    not be terminating the new argv[] array for /bin/sh, allowing it to
    walk off the end of the list.

I dithered a bit about using alloca() even more as it's most commonly
associated with gcc.  However, standalone portable (using malloc) and
machine-specific assembler alloca implementations appear to be available
on just about all the architectures we're likely to want to port to.
alloca will be the least of our problems if ever going to another compiler.
1999-03-23 16:40:34 +00:00
Poul-Henning Kamp
cc7532aaf0 Add a sysctl variable which can help stop chroot(2) escapes.
kern.chroot_allow_open_directories = 0
	chroot(2) fails if there are open directories.

kern.chroot_allow_open_directories = 1 (default)
	chroot(2) fails if there are open directories and the process
	is subject of a previous chroot(2).

kern.chroot_allow_open_directories = anything else
	filedescriptors are not checked.  (old behaviour).

I'm very interested in reports about software which breaks when
running with the default setting.
1999-03-23 14:26:40 +00:00
Poul-Henning Kamp
7f4173cc09 Fix some nasty hangs if garbage were passed.
Noticed by:	Emmanuel DELOGET <pixel@DotCom.FR>
Remembered by:	msmith
1999-03-23 14:23:15 +00:00
Bruce Evans
3aacd339f0 Fixed a couple of missing drive selects in wdreset(). This mainly
affects cases where there is a slave but no master.  These bugs
were usually recovered from provided ATAPI was configured, but only
after lengthy delays.  Configuring ATAPI still fixes some bugs for
non-atapi drives.

Don't wait for seek complete in wdreset().  If necessary for pre-ata
drives, it can be waited for later (we got it wrong by only looking
at it for drive 0 anyway).  It is set as part of a historical
signature for ata drives but doesn't say anything useful about the
reset state.  It is cleared as part of a non-historical signature
for atapi drives so that drivers which don't understand atapi drives
seem to see no drive.  Waiting for it caused lengthy delays and
broke the status returned by wdreset() in cases where the master
was not an ata drive.  Then the whole wdprobe() failed in some
cases where the recovery code didn't work.

Don't wait for drive ready in wdreset().  The considerations are
the same as for seek complete, except drive ready does say something
useful about the reset state of ata drives, and waiting for it
later is required anyway for such drives.

Lengthy delays can now be avoided by not configuring nonexistent
(ata) drives.  Unfortunately, this breaks detection of atapi drives
in some configurations.
1999-03-23 10:22:19 +00:00
Jordan K. Hubbard
bdd9342bad Set X11BASE to something deliberately bogus in elf_buildworld so that
it doesn't link with a.out X11 libs.

Noticed by:	Arno J. Klaassen <arno@heho.snv.jussieu.fr>
1999-03-23 08:09:25 +00:00
Justin T. Gibbs
1652f91eb5 Fix a typo.
PR: 10722
Submitted by:	Chris Costello <chris@holly.dyndns.org>
1999-03-23 07:27:38 +00:00
Justin T. Gibbs
73c9c6dd62 Add support for the 3950U2B controller.
Recognize aic7895 controllers that have been "acquired" by a RAIDPort
card as normal aic7895s.

Recognize the aic7815 Raid Parity/Memory controller chip and notify
the user that it's RAID functionality will be ignored.
1999-03-23 07:26:41 +00:00
Justin T. Gibbs
e287abc83c Use TCL_CHANNEL/SCSI_ID/etc. macros in more locations.
Don't mess with the IRQMS bit in the host control register unless
we are an aic7770 chip.

Use calling context to determine if the card is already paused when
we update the target message request bit field in controller scratch
ram.  Looking at the paused bit in the HCNTRL register opened up a
race condition.

Insert delays in the target message request update routine as a temporary
work around for what looks like a chip bug.  I'm still investigating this
one.

Fix the Abort/Abort Tag/BDR handler to pull its message from the message
buffer in our softc instead of attempting to get it from a register on
the controller.  The message is never recorded by the controller in the
new message scheme.

Don't rely on having an SCB when a BDR occurs.  We can issue these during
invalid reconnects to.

Fix a few cases where we were restarting the sequencer but then still
falling out of a switch statement to unpause the sequencer again.
This could cause us to mess up sequencer state if it generated another
pausing interrupt between the time of the restart and unpause.

Kill the 'transceiver settle' loop during card initialization.  I
failed to realize that a controller that is not connected to any
cables will never settle or enable the SCSI transceivers at all.
The correct solution is to monitor the IOERR interrupt which indicates
that the transceiver state has changed (UW<->LVD).

Modify the aic7xxx assembler to properly echo input when stdin is not
a tty.
1999-03-23 07:24:29 +00:00
John Birrell
b847980f15 o Runnable threads are now maintained in priority queues. The
implementation requires two things:

      1.) The priority queues must be protected during insertion
          and removal of threads.  Since the kernel scheduler
          must modify the priority queues, a spinlock for
          protection cannot be used.   The functions
          _thread_kern_sched_defer() and _thread_kern_sched_undefer()
          were added to {un}defer kernel scheduler activation.

      2.) A thread (active) priority change can be performed only
          when the thread is removed from the priority queue.  The
          implementation uses a threads active priority when
          inserting it into the queue.

    A by-product is that thread switches are much faster.  A
    separate queue is used for waiting and/or blocked threads,
    and it is searched at most 2 times in the kernel scheduler
    when there are active threads.  It should be possible to
    reduce this to once by combining polling of threads waiting
    on I/O with the loop that looks for timed out threads and
    the minimum timeout value.

  o Functions to defer kernel scheduler activation were added.  These
    are _thread_kern_sched_defer() and _thread_kern_sched_undefer()
    and may be called recursively.  These routines do not block the
    scheduling signal, but latch its occurrence.  The signal handler
    will not call the kernel scheduler when the running thread has
    deferred scheduling, but it will be called when running thread
    undefers scheduling.

  o Added support for _POSIX_THREAD_PRIORITY_SCHEDULING.  All the
    POSIX routines required by this should now be implemented.
    One note, SCHED_OTHER, SCHED_FIFO, and SCHED_RR are required
    to be defined by including pthread.h.  These defines are currently
    in sched.h.  I modified pthread.h to include sched.h but don't
    know if this is the proper thing to do.

  o Added support for priority protection and inheritence mutexes.
    This allows definition of _POSIX_THREAD_PRIO_PROTECT and
    _POSIX_THREAD_PRIO_INHERIT.

  o Added additional error checks required by POSIX for mutexes and
    condition variables.

  o Provided a wrapper for sigpending which is marked as a hidden
    syscall.

  o Added a non-portable function as a debugging aid to allow an
    application to monitor thread context switches.  An application
    can install a routine that gets called everytime a thread
    (explicitly created by the application) gets context switched.
    The routine gets passed the pthread IDs of the threads that are
    being switched in and out.  I found this useful, but we can
    get rid of it if you want.

Submitted by: Dan Eischen <eischen@vigrid.com>
1999-03-23 05:11:30 +00:00
John Birrell
58a7cc5d1b [ The author's description... ]
o Runnable threads are now maintained in priority queues.  The
    implementation requires two things:

      1.) The priority queues must be protected during insertion
          and removal of threads.  Since the kernel scheduler
          must modify the priority queues, a spinlock for
          protection cannot be used.   The functions
          _thread_kern_sched_defer() and _thread_kern_sched_undefer()
          were added to {un}defer kernel scheduler activation.

      2.) A thread (active) priority change can be performed only
          when the thread is removed from the priority queue.  The
          implementation uses a threads active priority when
          inserting it into the queue.

    A by-product is that thread switches are much faster.  A
    separate queue is used for waiting and/or blocked threads,
    and it is searched at most 2 times in the kernel scheduler
    when there are active threads.  It should be possible to
    reduce this to once by combining polling of threads waiting
    on I/O with the loop that looks for timed out threads and
    the minimum timeout value.

  o Functions to defer kernel scheduler activation were added.  These
    are _thread_kern_sched_defer() and _thread_kern_sched_undefer()
    and may be called recursively.  These routines do not block the
    scheduling signal, but latch its occurrence.  The signal handler
    will not call the kernel scheduler when the running thread has
    deferred scheduling, but it will be called when running thread
    undefers scheduling.

  o Added support for _POSIX_THREAD_PRIORITY_SCHEDULING.  All the
    POSIX routines required by this should now be implemented.
    One note, SCHED_OTHER, SCHED_FIFO, and SCHED_RR are required
    to be defined by including pthread.h.  These defines are currently
    in sched.h.  I modified pthread.h to include sched.h but don't
    know if this is the proper thing to do.

  o Added support for priority protection and inheritence mutexes.
    This allows definition of _POSIX_THREAD_PRIO_PROTECT and
    _POSIX_THREAD_PRIO_INHERIT.

  o Added additional error checks required by POSIX for mutexes and
    condition variables.

  o Provided a wrapper for sigpending which is marked as a hidden
    syscall.

  o Added a non-portable function as a debugging aid to allow an
    application to monitor thread context switches.  An application
    can install a routine that gets called everytime a thread
    (explicitly created by the application) gets context switched.
    The routine gets passed the pthread IDs of the threads that are
    being switched in and out.

Submitted by: Dan Eischen <eischen@vigrid.com>

Changes by me:

  o Added a PS_SPINBLOCK state to deal with the priority inversion
    problem most often (I think) seen by threads calling malloc/free/realloc.

  o Dispatch signals to the running thread directly rather than at a
    context switch to avoid the situation where the switch never occurs.
1999-03-23 05:07:56 +00:00
Joseph Koshy
f5a7833449 Typo fix.
PR:		docs/10733
Submitted by:	Steve Coltrin <spcoltri@io.com>
1999-03-23 05:01:11 +00:00
Greg Lehey
240b9c4449 Describe the new semantics of 'create':
When creating a new drive, if the drive already has a vinum label,
  and name doesn't match the specified drive, do it anyway if the
  'force' flag is specified.

Continually-tripped-over-by:  Karl Pielorz <kpielorz@tdx.co.uk>

Update information about partition type for Vinum drives.  They
*should* be of type 'vinum', but currently we still accept (and
complain about) partitions of type 'unused'.  At a later date, only
'vinum' will be accepted.

Threatened-since: over a year
1999-03-23 04:55:50 +00:00
Greg Lehey
170cfb0311 Print correct message if the wrong module is loaded. 1999-03-23 04:55:23 +00:00
Greg Lehey
7a4bec173b vinum_create: pass 'force' flag to ioctl VINUM_STARTCONFIG.
Semantics:

  When creating a new drive, if the drive already has a vinum label,
  and name doesn't match the specified drive, do it anyway if the
  'force' flag is specified.

Continually-tripped-over-by:  Karl Pielorz <kpielorz@tdx.co.uk>
1999-03-23 04:54:59 +00:00
Greg Lehey
6100b4d0c5 Change the definition of VINUM_STARTCONFIG to take an int write parameter. 1999-03-23 04:50:40 +00:00
Greg Lehey
6d723f4eb0 Add flag VF_FORCECONFIG for use by the config_drive and friends. 1999-03-23 04:49:32 +00:00
Greg Lehey
ec79ff6a85 vinum_start:
remove the splbio() around the call to launch read requests.

launch_requests:
  Move the splbio() protection outside the entire launch_loop.  The
  previous location was causing problems with IDE drives, where the
  call to the strategy routine often did not complete until after
  complete_rqe deallocated the request structure.

Solution-independently-found-by: Russell Neeper <r-neeper@tamu.edu>
Problem-reported-by: Vallo Kallaste <vallo@matti.ee>
		     John Saunders <john@nlc.net.au>
		     Bernd Walter <ticso@cicely.de> (maybe)
1999-03-23 04:49:08 +00:00
Greg Lehey
8d2eda6481 VINUM_STARTCONFIG:
Pass the 'data' parameter to start_config, which uses it as the
  parameter 'force.
1999-03-23 04:48:31 +00:00
Greg Lehey
dae797bc58 init_drive:
Check for partition types FS_VINUM and FS_UNUSED.  Accept both, but
  complain about FS_UNUSED.  At a later date, only FS_VINUM will be
  accepted.

Threatened-since: over a year
1999-03-23 04:48:05 +00:00
Greg Lehey
b555657ffa Change the declaration of start_config. 1999-03-23 04:47:33 +00:00
Greg Lehey
7c0052fe03 start_config:
Add a flag `force' (VF_FORCECONFIG) to force name changes of
  existing drives.
config_drive:
  If the drive already has a vinum label, and name doesn't match the
  specified drive, do it anyway if the 'force' flag is specified.
finish_config:
  Reset the `force' flag.

Continually-tripped-over-by:  Karl Pielorz <kpielorz@tdx.co.uk>

give_sd_to_drive:
  If the drive is down, take the subdisk down and don't try to fix
  things.

update_plex_config:
  Don't try to update the config parameters of a plex which isn't
  fully configured (state plex_init or plex_unallocated).

  Correctly calculate the amount to trim off a striped or RAID-5 plex
  whose size is not a multiple of the stripe size.
1999-03-23 04:47:09 +00:00
Greg Lehey
84ac4f6bad Add partition type for Vinum.
Offered-for-inspection-to: jkh
1999-03-23 04:08:42 +00:00
Bruce Evans
3c697e0a62 Simplified using new SYMLINKS macro, mainly to test this macro. The
ifdefs are too ugly for this to be much of a simplification.  The
existence tests are even uglier now.  Note that the previous commit
was not submitted by me.  It missed the point and just added a second
layer of unused removals.

Fixed hard-coded "libcrypt"s.  The LCRYPTBASE macro mainly makes
things hard to read, but use it while we have it.
1999-03-23 03:41:09 +00:00
Bruce Evans
2336a7ea75 Removed all traces of LN_FLAGS. It was only used to produce a link
/usr/sbin/sysctl -> ${DESTDIR}/sbin/sysctl in some versions of 2.2,
and this link was broken if DESTDIR was set.

Added a SYMLINKS macro.  This works the same as LINKS, except it
creates symlinks and the linked-to pathname may be relative.  This
is more flexible than LN_FLAGS, since it supports installing
symlinks independently of hard links.

Use `ln -f[s] ...' instead of `rm -f ...; ln [-s] ...' for LINKS and
SYMLINKS.  This is equivalent if the target is neither a directory nor
a symlink to a directory.

PR:		8279
1999-03-23 03:06:25 +00:00
Alexander Langer
da69a97ff0 Typo fix (set --> get).
Obtained from:	OpenBSD (David Leonard)
1999-03-22 23:13:37 +00:00
Nick Hibma
83fac6ab3f Cosmetical changes. 1999-03-22 19:58:59 +00:00
Nick Hibma
86df1e04cc Implementation of the keyboard driver.
Done by: Kazutaka YOKOTA <yokota@zodiac.mech.utsunomiya-u.ac.jp>
1999-03-22 19:55:30 +00:00
Nick Hibma
5e294a20cf Moved timeout initialisation a bit earlier in the process. The
untimeout function triggered panics once in a while.
1999-03-22 19:52:28 +00:00
Poul-Henning Kamp
3a25914cfd Make the same size rounding error both ways. 1999-03-22 14:01:58 +00:00
Poul-Henning Kamp
7f29144bc9 Nuke old copies of /usr/include/timepps.h 1999-03-22 12:58:15 +00:00
Poul-Henning Kamp
057cfd176a Address our users in a civilized manner. (They may know better, but probably
just forgot it, it is certainly no reason to don't insult them.)
1999-03-22 10:38:07 +00:00
Bruce Evans
0af0a4b0f4 Display floats with format %*.0f instead of as "*****" if there is
enough space for this but not enough space for the normal %*.*f
format.  Similarly for long doubles.
1999-03-22 03:44:01 +00:00
David E. O'Brien
fc8fc1d234 fix conflicts. 1999-03-22 01:29:39 +00:00
David E. O'Brien
2fee01e716 Update to latest version. 1999-03-22 01:28:54 +00:00
David E. O'Brien
21062dd2aa This commit was generated by cvs2svn to compensate for changes in r44930,
which included commits to RCS files with non-trunk default branches.
1999-03-22 01:26:43 +00:00
David E. O'Brien
cb1dd5d1b3 Virgin import of ISC-DHCP v2.0b1pl18 1999-03-22 01:26:43 +00:00
Jordan K. Hubbard
f1a3bb13d8 The attached patch to /usr/src/sbin/dump/optr.c changes the comparison
so that dumps are treated by dump -w as having been done on midnight
of the day they were actually run. This makes dump -w behave as
expected for regularly scheduled daily dumps - if they all run the
same day. It makes dump -w behave strangely if you dump late in the
day and check again after midnight, but that is the lesser of two
evils.

Submitted by:	Mike Meyer <mwm@phone.net>
PR:		9429
1999-03-21 23:47:38 +00:00
Alan Cox
99c81ca94d Two changes:
Remove more (redundant) map timestamp increments from properly
synchronized routines.  (Changed: vm_map_entry_link, vm_map_entry_unlink,
and vm_map_pageable.)

Micro-optimize vm_map_entry_link and vm_map_entry_unlink, eliminating
unnecessary dereferences.  At the same time, converted them from macros
to inline functions.
1999-03-21 23:37:00 +00:00
Jordan K. Hubbard
8abc4f6dad Add cdev 120 as the audit device.
Requested by:	Robert Watson <robert+freebsd@cyrus.watson.org>
1999-03-21 21:53:09 +00:00
Joerg Wunsch
0e20a72001 Finally remove the ancient `-exec rm -f {} ;' cruft that used to be
used to cleanup old files, and replace it by -delete which has been
present in our find(1) for ages now.
1999-03-21 12:34:06 +00:00
Poul-Henning Kamp
2fcf3276d9 Now bb structures are linked together. 1999-03-21 12:32:17 +00:00
Poul-Henning Kamp
48c619d0ee Link the bb structures together as we find them. 1999-03-21 12:30:50 +00:00
Poul-Henning Kamp
aca6858651 Remove the linker-set hack we used for BB profiling. 1999-03-21 12:28:14 +00:00
Bruce Evans
760a0960be Restored objlink. The previous commit was confused about the difference
between OBJLINK and objlink.
1999-03-21 06:43:40 +00:00
Warner Losh
43b4fd733d Don't be so selective about which errors cause us to continue and
which ones cause us to fail.  Now all open errors on the databse file
will cause the next file in the list to be tried.

Submitted by: Arne Henrik Juul <arnej@math.ntnu.no>
PR: 4585
1999-03-21 03:45:58 +00:00