d/freebsd-skq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
83,059 Commits 4 Branches 49 Tags 2 GiB
1ff96af1ea
Commit Graph

36365 Commits

Author SHA1 Message Date
sam
1ff96af1ea a better solution to building FAST_IPSEC w/o INET6 
Submitted by:	Jeffrey Hsu <hsu@FreeBSD.org>
 2002-11-10 17:17:32 +00:00
mjacob
d00e94bfc6 Use atomic_set_8 on the us_freelist maps as they are not otherwise 
protected. Furthermore, in some RISC architectures with no normal
byte operations, the surrounding 3 bytes are also affected by the
read-modify-write that has to occur.
 2002-11-10 16:16:44 +00:00
alc
fc8a5bc419 When prot is VM_PROT_NONE, call pmap_page_protect() directly rather than 
indirectly through vm_page_protect().  The one remaining page flag that
is updated by vm_page_protect() is already being updated by our various
pmap implementations.

Note: A later commit will similarly change the VM_PROT_READ case and
eliminate vm_page_protect().
 2002-11-10 07:12:04 +00:00
mdodd
55c878faa1 Replace {disable,enable}_intr() with critical_{enter,exit}(). 2002-11-10 03:45:49 +00:00
alc
d577ab06c6 Fix an error case in vm_map_wire(): unwiring of an entry during cleanup 
after a user wire error fails when the entry is already system wired.

Reported by:	tegge
 2002-11-09 21:26:49 +00:00
iwasaki
eec4b9c40a Add a new loader tunable, hw.hasbrokenint12, to indicate that BIOS 
has broken int 12H.
If hw.hasbrokenint12="1" in loader environment, kernel never use BIOS
INT 12 call to determine base memory size.
Otherwise, kernel use INT 12 in old behaviour.
This should fix kernel panic problem caused by 1.544 changes.

MFC after:	1 day
 2002-11-09 21:17:41 +00:00
alfred
6adba48a65 Fix an unparenthasized macro argument. md5s differ but this is likely 
to order of operations that are actually fixed by the proper parenthasizing.
 2002-11-09 20:13:16 +00:00
des
4e2288959e Print real / avail memory in megabytes rather than kilobytes. 2002-11-09 16:19:14 +00:00
alfred
533541c47b Fix instances of macros with improperly parenthasized arguments. 
Verified by: md5
 2002-11-09 12:55:07 +00:00
alc
5f7be03df2 In vm_page_remove(), avoid calling vm_page_splay() if the object's memq 
is empty.
 2002-11-09 08:27:42 +00:00
sam
e5864cae6b temporarily disallow FAST_IPSEC and INET6 to avoid potential panics; 
will correct this before 5.0 release
 2002-11-08 23:50:32 +00:00
sam
ef62292833 FAST_IPSEC fixups: 
o fix #ifdef typo
o must use "bounce functions" when dispatched from the protosw table

don't know how this stuff was missed in my testing; must've committed
the wrong bits

Pointy hat:	sam
Submitted by:	"Doug Ambrisko" <ambrisko@verniernetworks.com>
 2002-11-08 23:37:50 +00:00
sam
6019e3c767 fixup FAST_IPSEC build w/o INET6 2002-11-08 23:33:59 +00:00
sam
f9821d8021 correct fast ipsec logic: compare destination ip address against the 
contents of the SA, not the SP

Submitted by:	"Doug Ambrisko" <ambrisko@verniernetworks.com>
 2002-11-08 23:11:02 +00:00
sam
e23e06eb6f correct minor # in make_dev call 
Submitted by:	Doug Ambrisko" <ambrisko@verniernetworks.com>
 2002-11-08 23:07:41 +00:00
sam
01475299d2 change load order so module is present before crypto drivers 
Submitted by:	Doug Ambrisko" <ambrisko@verniernetworks.com>
 2002-11-08 23:07:03 +00:00
jhb
66d9abef31 - Move netsmb entries over to MI files. netsmb appears to be MI code. 
- Add 'nowerror' to pci/simos.c to help LINT builds.
 2002-11-08 22:33:14 +00:00
jhb
b8daeabf59 Print daddr_t's with %j and intmax_t. 2002-11-08 22:28:35 +00:00
jhb
686416bfa5 Fix some sizeof(int) != sizeof(void *) warnings. 2002-11-08 21:30:26 +00:00
jhb
6c4366c376 - Change mb_copy_t to take a size_t as the length argument instead of an 
int.
- Change the local variable in smb_copy_iconv() from an int to a size_t.

These make smb_copy_iconv() happy in a 64-bit world.
 2002-11-08 21:26:32 +00:00
jhb
4481b3e997 Fix a sizeof(int) != sizeof(void *) warning. 2002-11-08 21:16:27 +00:00
jhb
5c9fa9d32a Use intptr_t to fix various sizeof(int) != sizeof(void *) warnings. 2002-11-08 21:13:18 +00:00
jhb
eb9c5b7d92 - Move opt_fb.h earlier so it covers included headers. 
- Update cdevsw.  Block majors are out, kqfilters are in.
- No need to bzero softc's that we get from new-bus.  They come that way to
  begin with.
 2002-11-08 21:10:53 +00:00
jhb
f47ce099d4 Move include of opt_fb.h earlier so that FB_INSTALL_CDEV is defined in 
included headers.
 2002-11-08 21:04:44 +00:00
jhb
7dbfeb6da1 Fix printf warnings with %j and uintmax_t. 2002-11-08 20:59:23 +00:00
jhb
40fae53b3b Fix warnings when compiled with SIMOS defined. 2002-11-08 20:52:21 +00:00
rwatson
a563c04c4a Assign value of NULL to imgp->execlabel when imgp is initialized 
in the ELF code.  Missed in earlier merge from the MAC tree.

Approved by:	re
Obtained from:	TrustedBSD Project
Sponsored by:	DARPA, Network Associates Laboratories
 2002-11-08 20:49:50 +00:00
jhb
ba93b1c02a Cast pointers in mem*_io() compat macros to uintptr_t so that they can be 
used as bus handles by the bus_space functions implementing these macros.
 2002-11-08 20:49:43 +00:00
jhb
ad9e680296 Make 3dfx i386-only. The memrange API it uses may be defined in an MI 
header, but it is only implemented on i386.
 2002-11-08 20:04:38 +00:00
jhb
257ad7107b Move rc(4) over to MI notes and enable it as a MI module. 2002-11-08 19:34:26 +00:00
rwatson
551263d9d4 To reduce per-return overhead of userret(), call into 
mac_thread_userret() only if PS_MACPEND is set in the process AST mask.
This avoids the cost of the entry point in the common case, but
requires policies interested in the userret event to set the flag
(protected by the scheduler lock) if they do want the event.  Since
all the policies that we're working with which use mac_thread_userret()
use the entry point only selectively to perform operations deferred
for locking reasons, this maintains the desired semantics.

Approved by:	re
Requested by:	bde
Obtained from:	TrustedBSD Project
Sponsored by:	DARPA, Network Associates Laboratories
 2002-11-08 19:00:17 +00:00
jhb
32c37d4a13 Use critical_enter/exit instead of disable/enable_intr to make this 
driver more portable.
 2002-11-08 18:32:57 +00:00
jhb
b72917b520 - Change the ATM stack functions to use intptr_t instead of int for opaque 
arguments.
- Fix a few other places that assumed that sizeof(int) == sizeof(void *).

Reviewed by:	mdodd
 2002-11-08 18:27:30 +00:00
pdeuskar
9918575c8a - Set RS (Report Status) bit on all descriptors of a packet instead of just the last one. 
- Set RDTR to zero by default instead of 28.
- Fixed a problem with TX hangs with jumbo frames when number of fragments in the mbuf chain
is large.
- Added support for 82540EP based cards.

MFC after:	3 days
 2002-11-08 18:14:17 +00:00
rwatson
f3748b0c0b Update MAC modules for changes in arguments for exec MAC policy 
entry points to include an explicit execlabel.

Approved by:	re
Obtained from:	TrustedBSD Project
Sponsored by:	DARPA, Network Associates Laboratories
 2002-11-08 18:04:36 +00:00
rwatson
95c4afbed0 Add an explicit execlabel argument to exec-related MAC policy entry 
points, rather than relying on policies to grub around in the
image activator instance structure.

Approved by:	re
Obtained from:	TrustedBSD Project
Sponsored by:	DARPA, Network Associates Laboratories
 2002-11-08 18:04:00 +00:00
phk
38900b3ee3 Remove harmless but irritating printf. 2002-11-08 15:35:16 +00:00
phk
875ed4eea7 Always recalculate the SRM checksum if the label is at 64 bytes offset. 
Tested by:	jhb
 2002-11-08 15:31:34 +00:00
jhb
1027780c50 - Update the SIMOS fp-stealing code to KSE-II and beyond. 
- Wrap a variable only used in !SIMOS case in #ifndef SIMOS.
 2002-11-08 15:24:32 +00:00
jhb
8ff6cd442c Remove #ifdef DEBUG signal debugging code brought over from NetBSD with the 
original signal handling code.  It doesn't compile and it seems the rest of
the infrastructure was never brought over from NetBSD.
 2002-11-08 15:16:38 +00:00
jhb
2c9bc07b31 Make xrpu(4) i386-only. Consumers of i386_btop() are not MI. 2002-11-08 15:05:16 +00:00
jhb
bf90740990 Wrap a device_printf() that violates bus space abstractions to figure out 
if it's IO port resource is IO or memory mapped for the sake of a printf
using i386-specific values in #ifdef __i386__.
 2002-11-08 15:01:02 +00:00
jhb
d102d4721a Use %z to print a size_t value. 2002-11-08 14:58:35 +00:00
jhb
d39f72b79f Use %z to print size_t values. 2002-11-08 14:57:18 +00:00
jhb
980dea1922 Cast a ptrdiff_t to an int to printf. 2002-11-08 14:52:26 +00:00
jhb
4d0d9edb13 Use %z to print a size_t value. 2002-11-08 14:50:44 +00:00
tmm
6b29f7e140 Remove physmem from here, too, as it is defined in vm_init.c since 
r1.35 (forgotten in my last commit due to a botched patch).

Pointy hat to:	tmm
 2002-11-08 11:10:50 +00:00
mr
10ff71fdf9 Reviewed by: n_hibma 
MFC after:	2 weeks
fix XPT_CALC_GEOMETRY to fill in some reasonable values.
Inspired by aic/sbp.
 2002-11-08 07:57:42 +00:00
davidxu
8274c76b58 adjust critical section to only wrap around vm86_bioscall(). 2002-11-08 03:07:52 +00:00
davidxu
9ca980d0e7 use critical_enter/exit to add a critical section around BIOS call, 
this unbreaks WITNESS.

Pointed out by: jhb
 2002-11-08 01:09:16 +00:00