Commit Graph

183059 Commits

Author SHA1 Message Date
haad
5f6671a94a Allow undescribed, direct ioctls as used by Unix. This capability was removed in BSD, presumably because nothing used it any more.
Third party system software written for Unix (like ZFS) requires this to work without significant modifications.

Ok supremeleader@
2009-08-13 08:57:43 +00:00
dholland
69c3e9d213 Pass WARNS=4, not without some gross preprocessor hackery.
XXX: does this program actually do anything useful these days?
2009-08-13 06:59:37 +00:00
dholland
e63a3e7105 Assorted minor cleanup:
- use stdbool.h (partly)
  - move extern declarations of data to header files
  - use right types for calloc() wrapper
  - remove bogus casts on return values
  - remove excessive Pascal-style parentheses in conditionals
  - a couple const fixes
  - fix some typos in comments
2009-08-13 05:53:58 +00:00
matt
ab902ee68d Move MID_MACHINE to <mips/mips_param.h> and use local values so we don't
need to include exec_aout.h
2009-08-13 05:15:08 +00:00
dholland
7d59a3fee1 pass -Wshadow 2009-08-13 04:09:53 +00:00
matt
69d946c278 Don't include EXEC_AOUT stuff by default. 2009-08-13 03:56:32 +00:00
matt
baf57f2df6 #include "opt_execfmt.h" and only compile innards if EXEC_AOUT is defined. 2009-08-13 03:53:13 +00:00
dholland
72efe4fb6f Sprinkle const. 2009-08-13 03:50:02 +00:00
matt
a15cb55d9e Fix KASSERT() failure reported by Geoff Wing. 2009-08-13 03:21:03 +00:00
dholland
6c23c8ddec woops (doh!) 2009-08-13 03:10:03 +00:00
dholland
f42113e362 Whitespace. 2009-08-13 03:07:49 +00:00
dholland
b74af21b24 sprinkle static 2009-08-13 02:10:50 +00:00
dyoung
94981d88f4 Postpone to a workqueue adding link-local and loopback IPv6 addresses
to an interface.  This keeps the kernel from entering ifp->if_ioctl
recursively, which can deadlock if if_ioctl takes locks.  This will
fix deadlocks & LOCKDEBUG errors in agr(4) (kern/39940) and in
gre(4).
2009-08-13 00:34:04 +00:00
dyoung
7a9941f8e8 Use sysctl(9) to expose to userland each interface transmission
queue's maximum length, current length, and number of drops.  E.g.,

% sysctl net.interfaces.bnx0
net.interfaces.bnx0.sndq.len = 0
net.interfaces.bnx0.sndq.maxlen = 509
net.interfaces.bnx0.sndq.drops = 0

Let userland adjust the maximum queue length.

While I'm here, add a 64-bit generation number, if_index_gen, to
ifnet; the pair [ifp->if_index, ifp->if_index_gen] can serve to
identify an ifnet for the lifetime of the system.  I will use this
in an upcoming change.

Ok matt@.
2009-08-13 00:23:31 +00:00
enami
cc63c84008 Make nanosleep cancelable again. 2009-08-12 23:51:23 +00:00
matt
28623809e3 Note remove of a.out support for MIPS 2009-08-12 23:41:10 +00:00
matt
5f89c92891 Don't build for MIPS anymore 2009-08-12 23:39:13 +00:00
matt
1191a19f81 We don't install <machine/bsd-aout.h> and <machine/aout_machdep.h> on MIPS
anymore and we don't build elf2aout for MIPS either.
2009-08-12 23:38:20 +00:00
matt
9b2fb10da6 Nuke a.out support for MIPS. 2009-08-12 23:29:18 +00:00
skrll
89ec5e3d9c Put back random.S. Hi dsl! 2009-08-12 22:49:37 +00:00
dyoung
6c7a849f95 Don't require the gateway address to have room for both an interface
name and address.  Room for an address will do.  This should fix
a regression in 'arp -s ...' on interfaces such as xennet0 with
unusually long names.

I will request a pull-up to netbsd-5.
2009-08-12 22:16:15 +00:00
haad
732a349764 Register callbacks for reclaiming kvm kernel space to arc. Unregister both
kvm address space and memory callbacks from kernel during arc_fini call.
2009-08-12 21:52:41 +00:00
dsl
e46010532d Remove some .c files accidentally left on the .S lines 2009-08-12 21:24:57 +00:00
dsl
cfac8834d7 Split the lines that add .c to SRCS from those that add .S
No changes (apart from the order) intended.
2009-08-12 21:20:40 +00:00
dsl
6f5236b423 Use stuff from libc/Makefile to auto-remove .c files if a .S has been added.
Use it to dispose of tne .c files that were already only conditionally
added.
2009-08-12 21:18:42 +00:00
seanb
612dfc4de9 Watch getservbyport_r() semantics. sv needs scope
while results are being looked at which in turn
are no longer valid after a endservent_r().
2009-08-12 20:24:30 +00:00
skrll
6213e5fd5f Define HPPA_FLEX_SIZE 2009-08-12 20:10:41 +00:00
macallan
5326bdc37e mention voyagerfb 2009-08-12 19:43:00 +00:00
macallan
c36cf5306b use voyagerfb 2009-08-12 19:39:51 +00:00
macallan
5f495e4be4 add voyagerfb, a driver for the Gdium Liberty 1000's onboard graphics chip 2009-08-12 19:28:00 +00:00
tsutsui
c144cf4a36 Fix a botch in the previous rev 1.21:
We have to mask higher bits on NCR register access via bus_space_read_4(9),
as the previous NCR_READ_REG() does.
Problem reported by Erik Bertelsen on port-pmax@.

I tested a different version on 3MIN and I also tested committed version
on gxemul, but 3MAX doesn't have asc_ioasic but asc_tc..
2009-08-12 14:06:13 +00:00
simonb
598bb4aec3 Match the Broadcom BCM5464 PHY. 2009-08-12 13:34:34 +00:00
simonb
7971cd7c5f Regen. 2009-08-12 13:33:42 +00:00
simonb
1bb23ef36e Add Broadcom BCM5464 PHY. 2009-08-12 13:33:02 +00:00
simonb
00bf4c87d7 Remove many magic numbers for addresses and interrupt numbers, and use
constants defined in SiByte/Broadcom standard header files.  Switch from
using offsets for locators to actual addresses.

Tested on a swarm (my rhone is dead, but should be OK there too).
2009-08-12 12:56:29 +00:00
sborrill
713116a751 Update to reflect new open-source ath hal 2009-08-12 12:15:49 +00:00
msaitoh
7dc1329052 Add 1000BASE-BX10. 2009-08-12 11:17:17 +00:00
skrll
32444302bd Fix non-DEBUG builds. 2009-08-12 10:03:30 +00:00
dholland
da121b3331 sprinkle static 2009-08-12 08:54:54 +00:00
dholland
1e99780e5d sprinkle static 2009-08-12 08:44:45 +00:00
dholland
5305281b0c sprinkle static 2009-08-12 08:21:41 +00:00
dholland
f2a20f5fe4 sprinkle static 2009-08-12 08:04:05 +00:00
skrll
c352a1ce0c Remove NOP flushes. pvh_list == NULL. 2009-08-12 07:46:36 +00:00
skrll
3e1a0e3c2d Pull across a fix from OpenBSD, but use PDE_SIZE instead of ~PDE_MASK + 1.
Make sure that if a PDE isn't there, we skip to the start of the address
range covered by the next PDE and not somewhere in the middle.  The old could
have skipped over some valid PTE's causing them to stay behind in the pmap.
Since we would not flush the cache for those pages either this could also
cause memory corruption when dirty cache lines would be written back to
memory at a later stage.
2009-08-12 07:42:36 +00:00
dholland
75b3905d60 sprinkle static and prune some dead code 2009-08-12 07:42:11 +00:00
dholland
9b92b18917 sprinkle static 2009-08-12 07:28:40 +00:00
dholland
bc0970cff2 sprinkle static 2009-08-12 06:19:17 +00:00
dholland
6613b31163 sprinkle static 2009-08-12 06:06:28 +00:00
dholland
e0ba63fe83 sprinkle static 2009-08-12 05:48:04 +00:00
dholland
f59d107e64 sprinkle static 2009-08-12 05:35:44 +00:00