Sam Leffler [Thu, 28 Dec 2006 01:31:26 +0000 (01:31 +0000)]
o unbreak rate set defaulting
o mark 11g mode support on finding 11g or pure 11g (OFDM-only)
channels; was requiring pure 11g which caused some contortions
in drivers that manually setup their channel lists
Bruce Evans [Wed, 27 Dec 2006 20:26:00 +0000 (20:26 +0000)]
Avoid an instruction in atomic_cmpset_{int_long)() in most cases.
These functions are used a lot for mutexes, so this reduces the text
size of an average kernel by about 0.75%. This wasn't intended to
be a significant optimization, but it somehow increased the maximum
number of packets per second that can be transmitted by my bge hardware
from 320000 to 460000 (this benchmark is CPU-bound and remarkably
sensitive to changes in the text section).
Details: we would prefer to leave the result of the cmpxchg in %al,
but cannot tell gcc that it is there, so we have to convert it to an
integer register. We converted to %al, then to %[re]ax, but the
latter step is usually wasted since gcc usually only wants the condition
code and can recover it from %al just as easily as from %[re]ax. Let
gcc promote %al in the few cases where this is needed.
Nearby style fixes;
- let gcc manage the load of `res', and don't abuse `res' for a copy of `exp'
- don't echo `res's name in comments
- consistently spell the condition code as 'e' after comparison for equality
- don't hard-code %al anywhere except in constraints
- for the version that doesn't use cmpxchg, there is no requirement to use
%al anywhere, so don't hard-code it in the constraints either.
Style non-fix:
- for the versions that use cmpxchg, keep using "a" (was %[re]ax, now %al)
for the main output operand, although this is not required. The input
and output operands that use the "a" constraint are now decoupled, and
this makes things clearer except for the reason that the output register
is hard-coded. It is now just a hack to tell gcc that the input "a" has
been clobbered without increasing the number of operands.
Sam Leffler [Wed, 27 Dec 2006 19:07:09 +0000 (19:07 +0000)]
Add half/quarter rate 11a channel support:
o change handling of regdomain-related mib knobs so they can be set
post-attach: regdomain, countrycode, outdoor, and xchanmode; the
hal will not permit changing the regdomain but we expose it for now
o on regdomain/countrycode change recalculate the channel list and
push it to the net80211 layer (NB: looks to need more tweaking)
o setup rate tables for half/quarter rate channels
o honor half/quarter rate channel configs when changing channels
o honor half/quarter rate channel configs when setting the slot time
o use hack/nonstandard channel numbering scheme for the public safety
band to avoid overlapping 2.4G channels on dual-band cards
o remove setup of ic_sup_rates; the net80211 layer can do this for us
and it simplifies handling of half/quarter rate channels
Tested only in Public Safety Band with cards that have RF5112.
Sam Leffler [Wed, 27 Dec 2006 18:48:50 +0000 (18:48 +0000)]
Add half/quarter rate 11a channel support:
o add hack/nonstandard channel mapping for public safety band channels to
mirror kernel (temporary until we have proper 802.11 state)
o change ieee80211_mhz2ieee to take channel flags (unused right now)
While here do some minor fixups like using IEEE80211_IS_CHAN_ANYG.
Sam Leffler [Wed, 27 Dec 2006 18:46:18 +0000 (18:46 +0000)]
First cut at half/quarter-rate 11a channel support (e.g. for use
in the Public Safety Band):
o add channel flags to identify half/quarter-rate operation
o add rate sets (need to check spec on 4Mb/s in 1/4 rate)
o add if_media definitions for new rates
o split net80211 channel setup out into ieee80211_chan_init
o fixup ieee80211_mhz2ieee and ieee80211_ieee2mhz to understand half/quarter
rate channels: note we temporarily use a nonstandard/hack numbering that
avoids overlap with 2.4G channels because we don't (yet) have enough
state to identify and/or map overlapping channel sets
o fixup ieee80211_ifmedia_init so it can be called post attach and will
recalculate the channel list and associated state; this enables changing
channel-related state like the regulatory domain after attach (will be
needed for 802.11d support too)
o add ieee80211_get_suprates to return a reference to the supported rate
set for a given channel
o add 3, 4.5, and 27 MB/s tx rates to rate <-> media conversion routines
o const-poison channel arg to ieee80211_chan2mode
Yaroslav Tykhiy [Wed, 27 Dec 2006 13:15:33 +0000 (13:15 +0000)]
MFp4:
Implement the checks for required_* objects as two functions, one
to be run before precmd and the other after it. They get the current
rc command as an argument so they can choose what requirement tests
to perform. As of now, only "start" needs such tests.
Implement a new requirement variable, required_modules. It can
list kernel modules that need to be loaded after start_precmd
indicated success. Each name in the list can be just "file", or
"file:module", or "file~regex". This will allow us to remove a lot
of duplicated code from rc.d scripts.
Perform the checks not only for the default start method, but for
any method. This allows for more flexibility and fixes a few rc.d
scripts (namely newsyslog, pf, sendmail) that rely on a required_*
variable while providing a non-default start method.
To be able to call the new check_required* functions naturally,
remove lots of crufty duplicated code pieces from run_rc_command
and replace each of them by a call to the helper function providing
a single corrected instance of the respective code snippet. Now
run_rc_command isn't as scary as it used to be, and it even appears
to have quite a nice logic that was obscured by the old crufty code.
In the default handler for restart, run start from a subshell to
protect global varibles, e.g., _postcmd, from modification by the
start handler. This enables using restart_postcmd. [x]
PR: conf/98734 [x]
Submitted by: Rick van der Zwet <rick@wzoeterwoude.net> [x]
Reviewed by: freebsd-rc (silence for an older version)
MFC after: 1 month
Bruce Evans [Tue, 26 Dec 2006 18:33:55 +0000 (18:33 +0000)]
After rev.1.169, the "interrupt" coalescing parameters are not used in
bge_intr(). Some of them are used in bge_poll(). Simplify by only
initializing these for polling mode and not toggling them when switching
modes. This also fixes missing synchronization with the coalescing
engine in the toggling.
Kip Macy [Tue, 26 Dec 2006 02:48:42 +0000 (02:48 +0000)]
turn non-INVARIANT KASSERT into an empty but real C
statement so that syntax errors will still be caught
even if INVARIANTS aren't enabled in the config file
Ruslan Ermilov [Mon, 25 Dec 2006 08:59:09 +0000 (08:59 +0000)]
Pull up current -mdoc version from vendor.
- New Darwin, FreeBSD, and NetBSD versions.
- DragonFly support including the new .Dx macro.
- New .St strings: -isoC-amd1, -isoC-tcor1, -isoC-tcor2, and -ieee1275-94.
Joseph Koshy [Mon, 25 Dec 2006 02:22:22 +0000 (02:22 +0000)]
Keep shadow copies of the `e_shnum', `e_phnum' and `e_shstrndx'
members of the ELF Executable Header inside the library-private
`struct _Elf' descriptor and only update the underlying Elf{32,64}_Ehdr
structure on an elf_update(3) call. These fields of the Ehdr
structure are technically `out of bounds' for an application program
per the ELF(3) API, but we've seen applications that initialize
a new Ehdr structure using memcpy(), messing up the library's
invariants. [1]
Implement elf_getphnum() and handle ELF objects with more than
64K program header table entries.
Joseph Koshy [Mon, 25 Dec 2006 02:06:32 +0000 (02:06 +0000)]
Use strncpy() instead of strlcpy() when copying members of
a `struct ar_hdr'. These members do not use NUL-termination
while strlcpy() expects its source buffer to be NUL-terminated.
Marius Strobl [Sun, 24 Dec 2006 22:24:39 +0000 (22:24 +0000)]
Fix a typo in __fpu_ftox() that caused long double to long (and long long)
conversion of negative numbers to always result in -1.
While at it, rearrange the nearby comment so it fits in 80 chars per line,
like the rest of this file does.
Yaroslav Tykhiy [Sun, 24 Dec 2006 08:52:13 +0000 (08:52 +0000)]
Note that rev. 1.221 introduced a local workaround for a general problem.
Add a pointer to the relevant PR for future reference. The whole comment
will be OK to remove as soon as the general solution is applied.
Kip Macy [Sun, 24 Dec 2006 08:03:27 +0000 (08:03 +0000)]
- remove all calls to sched_pin and sched_unpin as they are only useful to
pmap on i386
- check for change in executable status in pmap_enter
- pmap_qenter and pmap_qremove only need to invalidate the range if one
of the pages has been referenced
- remove pmap_kenter/pmap_kremove as they were only used by pmap_qenter
and pmap_qremove
- in pmap_copy don't copy wired bit to destination pmap
- mpte was unused in pmap_enter_object - remove
- pmap_enter_quick_locked is not called on the kernel_pmap, remove check
- move pmap_remove_write specific logic out of tte_clear_phys_bit
- in pmap_protect check for removal of execute bit
- panic in the presence of a wired page in pmap_remove_all
- pmap_zero_range can call hwblkclr if offset is zero and size is PAGE_SIZE
- tte_clear_virt_bit is only used by pmap_change_wiring - thus it can be
greatly simplified
- pmap_invalidate_page need only be called in tte_clear_phys_bit if there
is a match with flags
- lock the pmap in tte_clear_phys_bit so that clearing the page bits is
atomic with invalidating the page
- these changes result in 100s reduction in buildworld from a malloc backed
disk to a malloc backed disk - ~2.5%
Yaroslav Tykhiy [Sat, 23 Dec 2006 18:54:49 +0000 (18:54 +0000)]
Dynamically resize the Disk column. It was too narrow for modern
disk device names such as da0s1b. So we also get rid of the nasty
constant 5 scattered over the code.
Implementing this change is a good chance to improve other bits
around it: init saved lengths early, always check return value from
kvm_getswapinfo().
Alfred Perlstein [Sat, 23 Dec 2006 17:18:18 +0000 (17:18 +0000)]
Fix a deadlock in detach/shutdown.
The problem was that I was acquiring the driver sx lock and then waiting
for a taskqueue to drain, however the taskqueue itself would try to
acquire the lock as well leading to a deadlock.
To fix the problem roll my own exclusive lock that allows for lock
cancellation. This is a normal exclusive lock, however if someone
marks it as "dead" then all waiters who request an error return will
get back an error instead of continuing to wait for the lock.
In this particular case, the shutdown and detach functions kill the
lock while the async task thread tries to acquire the lock but will
abort if the lock returns an error.
The other option was to drop the driver lock mid-detach and mid-shutdown,
mid-detach was a ok, however mid-shutdown was not.
While I'm here, fix a bug in what appears to be the mii link status
word in the softc going out to lunch. Explicitly set the status
word to 1 after initializing the mii. This would result in an interface
that would never respond to "if_start" requests as the mii interface
would always look down.
Yaroslav Tykhiy [Sat, 23 Dec 2006 15:40:41 +0000 (15:40 +0000)]
Fix the swap display further:
1) Resize the Used column to avoid screen overflow if BLOCKSIZE is long.
2) Track the current swap configuration so that its changes don't break
the display.
Yaroslav Tykhiy [Sat, 23 Dec 2006 14:25:17 +0000 (14:25 +0000)]
Start fixing the "swap" display by saving one horizontal position.
Now the display won't overflow the 80-char row if BLOCKSIZE=1024.
The new spacing is also consistent with the "pigs" display.
David Xu [Sat, 23 Dec 2006 03:30:50 +0000 (03:30 +0000)]
Fix a panic when rebooting a SMP machine, when option STOP_NMI is used,
nmi handler is used to stop other processors, nmi hander calls trap(),
however, trap() now accepts a pointer rather than a reference, this was
changed by kmacy@.
Jason Evans [Sat, 23 Dec 2006 00:18:51 +0000 (00:18 +0000)]
Implement chunk allocation/deallocation hysteresis by caching one spare
chunk per arena, rather than immediately deallocating all unused chunks.
This fixes a potential performance issue when allocating/deallocating
an object of size (4kB..1MB] in a loop.
Robert Watson [Fri, 22 Dec 2006 23:36:13 +0000 (23:36 +0000)]
Forced commit to recognize the repo-copy of vfs_syscalls.c to
vfs_extattr.c. This follows a naming convention similar to the break-out
of the extended attribute system calls (etc) in NetBSD.
Robert Watson [Fri, 22 Dec 2006 23:34:47 +0000 (23:34 +0000)]
Move src/sys/sys/mac_policy.h, the kernel interface between the MAC
Framework and security modules, to src/sys/security/mac/mac_policy.h,
completing the removal of kernel-only MAC Framework include files from
src/sys/sys. Update the MAC Framework and MAC policy modules. Delete
the old mac_policy.h.
Third party policy modules will need similar updating.
Warner Losh [Fri, 22 Dec 2006 05:54:19 +0000 (05:54 +0000)]
Avoid 'bogus' uninitialized warning by initializing rather than
playing cute games. It is much simpler, clearer and easier to follow
and understand. Besides, gcc4 likes it better.
Warner Losh [Fri, 22 Dec 2006 00:57:26 +0000 (00:57 +0000)]
ACPIIO_BATT_GET_UNITS would always return ENXIO. However, it should never
return an error since it returns a count of battery devices in the system.
Set it to 0 explicitly, since it is the only switch branch that doesn't set
it.
Yaroslav Tykhiy [Thu, 21 Dec 2006 21:35:49 +0000 (21:35 +0000)]
Allow this module to get its options from the kernel build directory
instead of always hard coding them in CFLAGS. POLA is kept here:
The module file built with GENERIC stays the same.