65107 Commits

Author SHA1 Message Date
cognet
6b6ebc8c90 Add a driver for the 7seg found on the CRB board, largely based on the
IQ31244 version.

Approved by:	re (blanket)
2007-09-22 16:25:43 +00:00
cognet
910d500785 Twist the RAS logic a bit to avoid branching.
MFC After:	1 week
Approved by:	re (blanket)
2007-09-22 14:23:52 +00:00
jeff
280a86b2ee - Improve grammar. s/it's/its/.
- Improve load long-term load balancer by always IPIing exactly once.
   Previously the delay after rebalancing could cause problems with
   uneven workloads.
 - Allow nice to have a linear effect on the interactivity score.  This
   allows negatively niced programs to stay interactive longer.  It may be
   useful with very expensive Xorg servers under high loads.  In general
   it should not be necessary to alter the nice level to improve interactive
   response.  We may also want to consider never allowing positively niced
   processes to become interactive at all.
 - Initialize ccpu to 0 rather than 0.0.  The decimal point was leftover
   from when the code was copied from 4bsd.  ccpu is 0 in ULE because ULE
   only exports weighted cpu values.

Reported by:	Steve Kargl (Load balancing problem)
Approved by:	re
2007-09-22 02:20:14 +00:00
rodrigc
b2b7d089f7 Disable multiple ntfs mounts to the same mountpoint.
Eliminates panics due to locking issues.
Idea taken from src/sys/gnu/fs/xfs/FreeBSD/xfs_super.c.

PR:	89966, 92000, 104393
Reported by:	H. Matsuo <hiroshi50000 yahoo co jp>,
		Chris <m2chrischou gmail.com>,
		Andrey V. Elsukov <bu7cher yandex ru>,
		Jan Henrik Sylvester <me janh de>
Approved by:	re (kensmith)
2007-09-21 23:50:15 +00:00
pjd
763f7449dc Fix some locking cases where we ask for exclusively locked vnode, but we get
shared locked vnode in instead when vfs.lookup_shared is set to 1.

Discussed with:	kib, kris
Tested by:	kris
Approved by:	re (kensmith)
2007-09-21 10:16:56 +00:00
mav
1f0cbf5a23 Dead code removal.
Approved by:	re (kensmith), glebius (mentor)
2007-09-21 08:25:55 +00:00
mav
01f9f69635 This is optimization of ether and debug hooks determination. It
simplifies code and should speedup pppoe_findsession() function which is
called for every incoming packet.

Approved by:	re (kensmith), glebius (mentor)
2007-09-21 08:24:08 +00:00
mav
14a66bdd4b This patch fixes thread unsafe usage of global pkt_hdr
variable. Second part is not so important, but IMO is also good.

Approved by:	re (kensmith), glebius (mentor)
2007-09-21 08:16:33 +00:00
jeff
e132f56a45 - Redefine p_swtime and td_slptime as p_swtick and td_slptick. This
changes the units from seconds to the value of 'ticks' when swapped
   in/out.  ULE does not have a periodic timer that scans all threads in
   the system and as such maintaining a per-second counter is difficult.
 - Change computations requiring the unit in seconds to subtract ticks
   and divide by hz.  This does make the wraparound condition hz times
   more frequent but this is still in the range of several months to
   years and the adverse effects are minimal.

Approved by:    re
2007-09-21 05:07:07 +00:00
rrs
66d80bdf93 - fix (global) address handling in the presence of duplicates, the
last interface should own the address, but the current code
  fumbles the handoff. This fixes that.
- move address related debugs to PCB4 and add additional ones to
  help in debugging address problems.

Approved by:	re@freebsd.org (K Smith)
2007-09-21 04:19:33 +00:00
jeff
bc0eadb21d - Redefine p_swtime and td_slptime as p_swtick and td_slptick. This
changes the units from seconds to the value of 'ticks' when swapped
   in/out.  ULE does not have a periodic timer that scans all threads in
   the system and as such maintaining a per-second counter is difficult.
 - Change computations requiring the unit in seconds to subtract ticks
   and divide by hz.  This does make the wraparound condition hz times
   more frequent but this is still in the range of several months to
   years and the adverse effects are minimal.

Approved by:	re
2007-09-21 04:10:23 +00:00
jeff
13d3160ef5 - Call sched_sleep() before we suspend threads. sched_wakeup() is already
called via setrunnable().  This allows time slept while suspended to
   be accounted for swap.

Approved by:	re
2007-09-21 04:04:22 +00:00
kevlo
be5effe157 - Add the device ID for the VIA VT3324 (CX700) chipset.
- Set and Get aperture size correctly for VIA's AGP3 chipsets.

Approved by: re (kensmith)
2007-09-21 02:10:13 +00:00
attilio
e25b203061 Fix some entries in the locks static table of witness.
In particular:
- smp_tlb_mtx is no longer used, so it is axed.
- smp rendezvous lock isn't really a leaf spin-mutex. Its bad placement in
  the table, however, has been the source of a false positive LOR reporting
  with the dt_lock.  However, smp rendezvous lock would have had sched_lock
  there for older lock, so it wasn't still a leaf lock.
- allpmaps is only used in ia32 architecture, so it is inserted in the
  appropriate stub.

Addictionally:
- kse_zombie_lock is no longer present, so its definition is axed out.
- zombie_lock doesn't need to have an exported symbol, so just let's it be
  declared as static.

Tested by: kris
Approved by: jeff (mentor)
Approved by: re
2007-09-20 20:38:43 +00:00
kib
038cf0387b Fill in cr2 in the signal context from ksi->ksi_addr.
Together with the sys/i386/i386/trap.c rev. 1.306 it fixes the PR.

Submitted by:	rdivacky
Suggested by:	jhb
Sponsored by:	Google Summer of Code 2007
PR:		kern/77710
Approved by:	re (kensmith)
2007-09-20 13:46:26 +00:00
simokawa
db835aec4a Serialize output routine of terminal emulator (te_puts()) by a lock.
- The output routine of low level console is not protected by any lock
by default.
- Increment and decrement of sc->write_in_progress are not atomic and
this may cause console hang.
- We also have many other states used by emulator that should be protected
by the lock.
- This change does not fix interspersed messages which PRINTF_BUFR_SIZE
kernel option should fix.

Approved by: re (bmah)
MFC after: 1 week
2007-09-20 04:05:59 +00:00
imp
099a7307bb Add mmc and mmcsd, and correct a couple of comments. They are
commented out until I can re-test them on all our architectures.  I
had re@ approval to commit this a long time ago, but that's before we
were this close to the branch.

Approved by: re@
2007-09-19 18:12:44 +00:00
jhb
9eab711109 - Only assume that a PCI device on the same slot is a "neighbor" USB
controller if it's sole child device has the "usb" device class.
  Previously ehci(4) would think that PCI-ISA bridges on the same slot
  (such as in some Intel ICHs) were "neighbors" resulting in spurious
  warnings about neighbor count mismatches.
- Fix a memory leak when looking for neighbors.

MFC after:	1 week
Approved by:	re (kensmith)
Tested by:	phk
2007-09-19 16:21:34 +00:00
cognet
1c8ede1513 Remove dead code.
Approved by:	re (blanket)
Beer from:	jadawin
2007-09-19 15:30:25 +00:00
sepotvin
f21ad79fb8 - Fix a small bit slip in PCIM_PCAP_D[0-2]PME defines.
- Add the definitions for D3PME_COLD and D3PME_HOT capabilities.

Reviewed by:	njl (mentor), imp
Approved by:	re (kensmith)
MFC after:	1 week
2007-09-19 13:05:58 +00:00
marcus
85efdc8988 Fix a bug that will cause a process that calls the VT_WAITACTIVE ioctl
to become unkillable when that process is sent a termination signal.  The
process will sit in waitvt looping in the kernel, and chewing up all
available CPU until the system is rebooted.

Submitted by:	Jilles Tjoelker <jilles@stack.nl>
Reviewed by:	bde
Approved by:	re (kensmith)
MFC after:	1 week
2007-09-19 03:59:33 +00:00
sam
8faa6081c7 invalidate ic_prevchan when constructing a new channel list
Approved by:	re (blanket wireless)
2007-09-18 21:54:27 +00:00
sam
6c988f9d11 honor IEEE80211_C_TXFRAG; drivers should never get fragmented packets
unless they indicate they're able to handle them

Approved by:	re (blanket wireless)
2007-09-18 21:09:26 +00:00
sam
70cb2f8614 tag mgmt and null data frames w/ a WME priority so drivers can assume
only data frames require classification

Approved by:	re (blanket wireless)
2007-09-18 21:07:41 +00:00
sam
a965065ed5 mgmt subtype 13 is action
Approved by:	re (blanket wireless)
MFC after:	2 weeks
2007-09-18 21:02:23 +00:00
sam
27063b31a0 Close longstanding race dropping inactive stations; grab a reference
to the node before starting the work, otherwise the node may go
away before a reference is made in ieee80211_send_mgmt.

Approved by:	re (blanket wireless)
Obtained from:	Atheros
2007-09-18 21:01:04 +00:00
sam
ce0001ed04 Safeguard count of stations in power save; don't touch it unless we're
operating in a mode where it should be maintained.

Approved by:	re (blanket wireless)
MFC after:	2 weeks
2007-09-18 20:55:51 +00:00
sam
c5c86cd09d skip IEEE80211_MODE_AUTO in announcing supported rates in ieee80211_announce;
there won't be any

Approved by:	re (blanket wireless)
2007-09-18 20:51:55 +00:00
sam
98df412892 o add IEEE80211_RATE_MCS to use instead of naked constant (for marking MCS)
o correct ieee80211_rate2media handling of MCS
o correct rate HT announcements for 11n devices

Approved by:	re (blanket wireless)
2007-09-18 20:46:36 +00:00
sam
20c40e534c o move IEEE80211_NONQOS_TID into a public location
o add IEEE80211_TID_SIZE and replace all the nake constants

Approved by:	re (blanket wireless)
2007-09-18 20:40:21 +00:00
sam
1f33ce6db9 remove IFM_IEEE80211_HT40PLUS and IFM_IEEE80211_HT40MINUS; they
never got used so nuke 'em before we branch

Approved by:	re (blanket wireless)
2007-09-18 20:30:40 +00:00
dwmalone
11cf0c8f4a regen.
Approved by: re (kensmith)
2007-09-18 19:51:49 +00:00
dwmalone
37c880369b The kernel version of Linux statfs64 is actually supposed to take
3 arguments, but we had forgotten the second argument. Also make the
Linux statfs64 struct depend on the architecture because it has an
extra 4 bytes padding on amd64 compared to i386.

The three argument fix is from David Taylor, the struct statfs64
stuff is my fault. With this patch I can install i386 Linux matlab
on an amd64 machine.

Submitted by: David Taylor <davidt_at_yadt.co.uk>
Approved by: re (kensmith)
2007-09-18 19:50:33 +00:00
ambrisko
ec62043bda Fix an incorrect PCI device id. The current value conflicts with
the mfi(4) LSI MegaSAS RAID card.  Looking at the Linux driver for the
mpt(4) it should be 0x0062 and not 0x0060.  Tested with an mfi card
of this device id.

Approved by:    re (bmah)
Reviewed by:    scottl
MFC after:      3 days
2007-09-18 16:39:24 +00:00
rrs
af4581daa3 - The address lock is changed to a rwlock. This
also involves macro changes to have a RLOCK and a WLOCK
  and placing the correct version within the code.
- The INP-INFO lock is changed to a rwlock.
- When sctp_shutdown() is called on Mac OS X, the socket lock is held.
  So call sctp_chunk_output with SCTP_SO_LOCKED and
  not SCTP_SO_NOT_LOCKED.
- Add SCTP_IPI_ADDR_[RW]LOCK and SCTP_IPI_ADDR_[RW]UNLOCK for Mac OS X.
- u_int64_t -> uint64_t
- add missing addr unlock for error return path
Approved by:	re@freebsd.org (K Smith)
2007-09-18 15:16:39 +00:00
alfred
7100da0db2 Reserve AF_ constants for vendors by giving them the odd numbered
AF_ constants ranging from 39 to 133.

Approved by: re (kensmith)
2007-09-18 09:22:16 +00:00
phk
8ffb12a0b6 Recognize the Soekris NET5501 and configure the error led.
Add watchdog(4) support by using the MFGPT0 in the Geode LX CX5536.
(Supported range: 2^30 .. 2^44 ns = 1s ... 5h)

Approved by:	re (bmah)
2007-09-18 09:19:44 +00:00
peter
4e0e5f8fee Fix an undefined symbol that as/ld neglected to flag as a problem. It
was used in assembler code in such a way that no unresolved relocation
records were generated, so ld didn't flag the problem.   You can see
this with an 'nm' of the kernel.  There will be 'U MAXCPU' on SMP systems.

The impact of this is that the intrcount/intrnames arrays do not have
the intended amount of space reserved.  This could lead to interesting
problems due to the arrays being present in the middle of kernel code.
An overflow would be rather interesting as executable code would be used
as per-cpu incrementing interrupt counters.

This fixes it for now by exporting MAXCPU to the assembler.  A better fix
might be to define these data structures in C - they're only referenced
in the kernel from C code these days anyway.

Approved by:  re (kensmith)
2007-09-17 21:55:28 +00:00
sam
d2e6b6eac7 Update beacon handling to sync w/ vap code base:
o add driver callback to handle notification of beacon changes;
  this is required for devices that manage beacon frames themselves
  (devices must override the default handler which does nothing)
o move beacon update-related flags from ieee80211com to the beacon
  offsets storage (or handle however a driver wants)
o expand beacon offsets structure with members needed for 11h/dfs
  and appie's
o change calling convention for ieee80211_beacon_alloc and
  ieee80211_beacon_update
o add overlapping bss support for 11g; requires driver to pass
  beacon frames from overlapping bss up to net80211 which is not
  presently done by any driver
o move HT beacon contents update to a routine in the HT code area

Reviewed by:	avatar, thompsa, sephe
Approved by:	re (blanket wireless)
2007-09-17 19:07:24 +00:00
sam
b6c4c36f72 convert hardware rate codes to IEEE rate codes with a lookup table
instead of a linear search

Reviewed by:	sephe, avatar
Approved by:	re (blanket wireless)
MFC after:	2 weeks
2007-09-17 16:25:02 +00:00
jeff
3fc0f8b973 - Move all of the PS_ flags into either p_flag or td_flags.
- p_sflag was mostly protected by PROC_LOCK rather than the PROC_SLOCK or
   previously the sched_lock.  These bugs have existed for some time.
 - Allow swapout to try each thread in a process individually and then
   swapin the whole process if any of these fail.  This allows us to move
   most scheduler related swap flags into td_flags.
 - Keep ki_sflag for backwards compat but change all in source tools to
   use the new and more correct location of P_INMEM.

Reported by:	pho
Reviewed by:	attilio, kib
Approved by:	re (kensmith)
2007-09-17 05:31:39 +00:00
sam
74666fdfce bandaid Dynamic Turbo A operation with old hal's: HAL_MODE_108A
does not have a rate table in older hal's so if we scan such a
channel the driver will hit an assertion or crash; for old hal's
fallback to using the static turbo rate table for this mode
(not correct but good enough for now given none of the rate
control algorithms understand how to switch between base+boost)

Approved by:	re (blanket wireless)
2007-09-17 05:16:38 +00:00
sam
0874cc7a2f fix led blinking in RUN state: the addition of the CAC state moved
IEEE80211_S_RUN and broke the array lookup used to find the LED flags

Approved by:	re (blanket wireless)
2007-09-17 05:12:19 +00:00
sam
0addef77a4 correct netband skip check in constructing a country ie
Approved by:	re (blanket net80211)
2007-09-17 03:48:32 +00:00
rrs
44d85d753b - For the 1-to-1 model, fix an off by one error that
allowed an extra connection over the backlog (by one)
Approved by:	re@freebsd.org (B. Mah)
2007-09-16 23:03:38 +00:00
thompsa
1498d5a44d Allow additional packet filtering on the physical interface for locally
destined packets, disabled by default.

PR:		kern/116051
Submitted by:	Eygene Ryabinkin
Approved by:	re (bmah)
MFC after:	2 weeks
2007-09-16 21:09:15 +00:00
thompsa
ad84066093 Make wi_set_channel() actually change the channel in hardware so that hostap
mode works properly, previously the hostap channel could not be changed off #3.

Fix an ifp/sc misuse while I am here.

Reported by:	many
Approved by:	re (bmah)
2007-09-16 20:02:29 +00:00
imp
cf6bc93f9b Kill bogus printf debugs.
Approved by: re@ (blanket)
2007-09-16 07:51:02 +00:00
imp
3bac12ad89 Kill overly verbose messages about setting bus width.
Approved by: re@ (blanket)
2007-09-16 07:48:58 +00:00
imp
23050ff30a Zyd needs be under Giant because USB is under Giant.
Submitted by: Weongyo Jeong
Reviewed by: sam@
Approved by: re@ (blanket)
2007-09-15 23:30:10 +00:00