Commit Graph

24001 Commits

Author SHA1 Message Date
John Baldwin
1aab03a584 Proc locking to protect p_ucred while we obtain an additional reference. 2001-01-23 22:38:15 +00:00
John Baldwin
f5343b3219 - FreeBSD doesn't have an abortop vnop as far as I can tell, so #ifdef
references to the hpf op out.
- Remove a lockdestroy() on a non-existent variable.
2001-01-23 22:37:30 +00:00
John Baldwin
1c827e2d15 Tag unused functions with __unused. 2001-01-23 22:24:26 +00:00
John Baldwin
0e9f2d667c Proc locking to protect p_ucred while we obtain another reference to it. 2001-01-23 22:23:25 +00:00
John Baldwin
4957d1b4cb Use suser(9) instead of checking p->p_ucred->cr_uid directly. 2001-01-23 22:22:25 +00:00
John Baldwin
6db2e28142 Remove a prototype for an unused and undefined debugging function. 2001-01-23 22:21:14 +00:00
John Baldwin
399fde35e8 Tag unused functions with __unused. 2001-01-23 22:18:41 +00:00
John Baldwin
b787813727 - Move stg_card_intr()'s prototype and definition inside appropriate
#ifdef's as it is is only used on older versions of FreeBSD.
- #ifdef a local variable only used on older versions of FreeBSD.
2001-01-23 22:18:07 +00:00
Justin T. Gibbs
98ba52eadc When flipping the first entry in the qinfifo with the "next queued SCB",
we must also inform the card of this change.  Otherwise the sequencer
will traverse a corrupt list of SCBS.  The side effects of this problem
were unknown SCBs completing in the qoutfifo or worse yet, panics due
to sequencer interrupts that referenced what, to the kernel, were invalid
SCB ids.
2001-01-23 22:17:03 +00:00
John Baldwin
4a08af58fe Relocate the 'hwfuncs' functions below the inlined functions they attempt
to call so that the other functions can actually be inlined.
2001-01-23 22:16:18 +00:00
John Baldwin
ec447af191 - Proc locking.
- Use queue macros.
- Use NULL instead of 0 for pointers.

Reviewed by:	des
2001-01-23 22:13:07 +00:00
John Baldwin
f5f7ddc668 Move nsp_card_intr()'s prototype and definition inside the appropriate
#ifdef's as it is only used in older versions of FreeBSD.
2001-01-23 22:09:04 +00:00
John Baldwin
acf5cad5aa Move the definition of nsp_start_timer() earlier in the file before it is
called so that it can be inlined.
2001-01-23 22:08:21 +00:00
John Baldwin
32c98125c9 The ncv_card_intr function is only used in older versions of FreeBSD, so
move its prototype and definition inside the appropriate #ifdef's.
2001-01-23 22:06:10 +00:00
John Baldwin
5d3e8edac4 Use the 'slp' local variable that was otherwise unused instead of casting
the driver softc under the assumption that the first part of it was a
struct scsi_low_softc.
2001-01-23 22:05:07 +00:00
John Baldwin
5928fa5cc7 #if 0 out local variables only used in #if 0'd code and remove unused
local variables.
2001-01-23 22:02:08 +00:00
John Baldwin
e54e93e21a Add curly braces ({}) around each struct speedtab to quiet warnings. 2001-01-23 22:00:05 +00:00
John Baldwin
14c3599a2d Proc locking in the form of using the proc lock to protect p_ucred while
we obtain another reference to it for vnode operations.
2001-01-23 21:58:44 +00:00
John Baldwin
d1a147af8b Axe unused local variable. 2001-01-23 21:53:44 +00:00
John Baldwin
4cd9657860 - Mark an unused function with __unused.
- Temporarily #if 0 some unused local variables.
2001-01-23 21:52:34 +00:00
John Baldwin
705a513a29 Axe an unused static softc. 2001-01-23 21:51:26 +00:00
John Baldwin
9697124fc8 Move a temporary #ifdef of code (just the #ifdef part) up so that it is
above the local variable declaration to quiet warnings about unused
variables.
2001-01-23 21:45:05 +00:00
John Baldwin
65baa27ce6 Kill a warning by marking an unused function with __unused. 2001-01-23 21:43:47 +00:00
John Baldwin
97eba215ac Argh, atomic_store_rel -> atomic_store_rel_int. 2001-01-23 21:40:07 +00:00
John Baldwin
505eb4d591 Woops, add in missing headers. 2001-01-23 21:39:15 +00:00
John Baldwin
1fed5f0326 Proc locking. 2001-01-23 21:33:55 +00:00
John Baldwin
c41e968cbd Use queue macros. 2001-01-23 21:32:02 +00:00
John Baldwin
bc4afa1f45 - Add proc locking.
- Fix several bugs in the wait syscall, including freeing the actual
  proc start, freeing the args, freeing the prison, and other minor
  nits.
- Use appropriate queue(3) macros.
- Use zpfind() instead of walking zombproc ourselves.
2001-01-23 21:30:25 +00:00
Kris Kennaway
1b2556e4a4 Fix the vulnerability with TCP ECE packets recently fixed in ipfw.
This is untested, but believed to work.
2001-01-23 21:11:28 +00:00
John Baldwin
48f8ca448a - Use proper atomic operations to make the run time initialization
controlled by svr_str_initialized be MP safe.
2001-01-23 21:07:16 +00:00
John Baldwin
f843c385b1 - Add necessary proc locking.
- Use proper atomic operations to make the run time initialization
  controlled by svr_str_initialized be MP safe.
- Use appropriate queue(3) macros where needed.
2001-01-23 21:06:30 +00:00
John Baldwin
37bf0fc1f8 FreeBSD doesn't have p_emuldata, and our stackgap_init() doesn't take an
argument.
2001-01-23 21:02:44 +00:00
John Baldwin
07a1f31cd0 Use proc lock to safely obtain references to p_ucred before vnode
operations.
2001-01-23 20:52:50 +00:00
John Baldwin
fb29c3e083 Protect calcru() with sched_lock. 2001-01-23 20:50:40 +00:00
Matt Jacob
15516f16d2 Do not do the commenting out the way that saves bytes and looks cleaner
to you. Do it the way Vox Populi wants it.
2001-01-23 16:35:33 +00:00
Takanori Watanabe
fd8433ab80 Map BSS section in PECOFF executable.
Submitted by:	 KUROSAWA Takahiro <fwkg7679@mb.infoweb.ne.jp>
2001-01-23 16:19:37 +00:00
Hajimu UMEMOTO
5d22597f3a Add mibs to hold the number of forks since boot. New mibs are:
vm.stats.vm.v_forks
	vm.stats.vm.v_vforks
	vm.stats.vm.v_rforks
	vm.stats.vm.v_kthreads
	vm.stats.vm.v_forkpages
	vm.stats.vm.v_vforkpages
	vm.stats.vm.v_rforkpages
	vm.stats.vm.v_kthreadpages

Submitted by:	Paul Herman <pherman@frenchfries.net>
Reviewed by:	alfred
2001-01-23 14:32:01 +00:00
Yoshihiro Takahashi
f2334c4082 Added pc98 apm driver.
Submitted by:	MURAMATSU Atsushi <amura@ma3.seikyou.ne.jp>
2001-01-23 12:37:15 +00:00
Joerg Wunsch
3b9d81dc1a Mark the (ancient) MaxOptix Tahiti 1 drive as a known rogue that would
respond to all LUNs.

Reviewed by:	ken
2001-01-23 10:49:30 +00:00
Hellmuth Michaelis
74acef9ae7 Remove useless variable vjlen.
Submitted by:	Sergio de Souza Prallon <prallon@tmp.com.br>
2001-01-23 10:01:55 +00:00
Peter Wemm
9f3e2250ca In answer to the comment: /* XXX is it OK to block here? */, the answer
is definately NO! as we are in interrupt context and malloc() does a
KASSERT() to be sure.
2001-01-23 09:43:23 +00:00
Wes Peters
550b151850 When attempting to bind to an ephemeral port, if no such port is
available, the error return should be EADDRNOTAVAIL rather than
EAGAIN.

PR:		14181
Submitted by:	Dima Dorfman <dima@unixfreak.org>
Reviewed by:	Garrett Wollman <wollman@khavrinen.lcs.mit.edu>
2001-01-23 07:27:56 +00:00
Garrett Wollman
e8f5d21723 This file declares `struct timespec' so that we can have a single,
centralized declaration (this will become useful later when we declare
struct timespec in all the places it is supposed to be).
2001-01-23 04:44:04 +00:00
Garrett Wollman
a25e45d40c Synchronize with some of my local changes. To quote:
* This file must be kept synchronized with <sys/timespec.h>.
 * It defines a structure which must be a type pun for
 * `struct timespec'; this structure is used in header files where
 * the ABI uses a `struct timespec' but standards prohibit its
 * definition.  (Currently only <sys/stat.h>.)
2001-01-23 04:40:54 +00:00
Robert Watson
02b65ffb64 o The move to using VADMIN under vaccess() resulted in some system
calls returning EACCES instead of EPERM.  This patch modifies vaccess()
  to return EPERM instead of EACCES if VADMIN is among the requested
  rights.  This affects functions normally limited to the owners of
  a file, such as chmod(), as EPERM is the error indicating that
  privilege would allow the operation, rather than a chance in mandatory
  or discretionary rights.

Reported by:	bde
2001-01-23 04:15:19 +00:00
Kirk McKusick
64d9f3b982 Add function prototype for libc function sysctlnametomib() 2001-01-23 03:42:20 +00:00
Jake Burkholder
43be6e2fa2 Sigh. atomic_add_int takes a pointer, not an integer.
Pointy-hat-to:	des
2001-01-23 03:40:27 +00:00
Dag-Erling Smørgrav
ac2e223868 Use atomic operations to update the stat counters. 2001-01-23 01:11:11 +00:00
Luigi Rizzo
8b2cd62d7d Change critical section protection for dummynet from splnet() to
splimp() -- we need it because dummynet can be invoked by the
bridging code at splimp().

This should cure the pipe "stalls" that several people have been
reporting on -stable while using bridging+dummynet (the problem
would not affect routers using dummynet).
2001-01-22 23:04:13 +00:00
David Malone
d2db050252 Free the kbd pointer when it isn't NULL, as opposed to when it is.
This was a typo in the M_ZERO patches.

Submitted by:	Mike Silbersack <silby@silby.com>
2001-01-22 22:54:02 +00:00