ƒ·>213482
150
178
163
346
1289
145
208
146
768
286
233
387
277
327
1065
665
158
154
153
273
143
241
563
563
1443
1443
371
485
173
184
218
257
172
301
220
140
145
219
144
219
145
558
169
198
148
663
248
327
433
244
169
164
234
247
1497
305
377
160
154
367
134
122
198
143
98
182
176
171
175
161
181
179
163
159
159
166
166
181
181
149
149
108
105
117
108
105
154
198
588
139
137
145
193
539
167
430
273
331
356
265
156
196
164
109
228
157
313
213
213
903
903
199
1979
1979
573
573
365
365
177
177
261
261
385
385
247
247
300
473
149
300
274
2467
2467
134
134
1167
138
1167
503
251
503
236
236
640
640
640
640
167
167
294
294

K 10
svn:author
V 7
pluknet
K 8
svn:date
V 27
2010-10-06T14:43:37.380361Z
K 7
svn:log
V 82
Add myself to calendar.freebsd and committers-src.dot.

Approved by:	avg (mentor)

END
K 10
svn:author
V 7
pluknet
K 8
svn:date
V 27
2010-10-06T15:01:37.558193Z
K 7
svn:log
V 67
Fix wrong slashes in a previous commit.

Approved by:	avg (mentor)

END
K 10
svn:author
V 8
jchandra
K 8
svn:date
V 27
2010-10-06T15:37:55.574203Z
K 7
svn:log
V 248
PCI fix for XLR C revision chips, limit DMA address to the first 2GB
physical address.  Adds a dma tag to the XLR/XLS pci bus with the
lowaddr if the CPU happens to be a XLR C rev.

Submitted by:	Sreekanth M. S. (kanthms at netlogicmicro dot com))

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-06T17:35:27.805527Z
K 7
svn:log
V 1191
Overhaul MII register access routine and remove unnecessary
BGE_MI_MODE register accesses. Previously bge(4) used to read
BGE_MI_MODE register to detect whether it needs to disable
autopolling feature or not. Because we don't touch autopolling in
other part of driver there is no reason to read BGE_MI_MODE
register given that we know default value in advance. In order to
achieve the goal, check whether the controller has CPMU(Central
Power Mangement Unit) capability. If controller has CPMU feature,
use 500KHz MII management interface(mdio/mdc) frequency regardless
core clock frequency. Otherwise use default MII clock. While I'm
here, add CPMU register definition.

In bge_miibus_readreg(), rearrange code a bit and remove goto
statement. In bge_miibus_writereg(), make sure to restore
autopolling even if MII write failed. The delay time inserted after
accessing BGE_MI_MODE register increased from 40us to 80us.

The default PHY address is now stored in softc. All PHYs supported
by bge(4) currently uses PHY address 1 but it will be changed when
we add newer controllers. This change will make it easier to change
default PHY address depending on PHY models.

Submitted by:	davidch

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-06T17:47:13.466220Z
K 7
svn:log
V 49
Remove bge_tx_buf_ratio, unused member of softc.

END
K 10
svn:author
V 7
delphij
K 8
svn:date
V 27
2010-10-06T18:20:07.773298Z
K 7
svn:log
V 111
Add definition for FreeBSD 7.4, which will be used in a manual page change.

Reviewed by:	ru
MFC after:	3 days

END
K 10
svn:author
V 7
delphij
K 8
svn:date
V 27
2010-10-06T18:20:39.409346Z
K 7
svn:log
V 50
Use a defined FreeBSD version.

MFC after:	3 days

END
K 10
svn:author
V 8
ambrisko
K 8
svn:date
V 27
2010-10-06T18:36:50.455370Z
K 7
svn:log
V 670
Add the capability to read the complete contents of the NVRAM via sysctl
	dev.bce.<unit>.nvram_dump
Add the capability to write the complete contents of the NVRAM via sysctl
	dev.bce.<unit>.nvram_write
These are only available if the kernel option BCE_DEBUG is enabled.
The nvram_write sysctl also requires the kernel option
BCE_NVRAM_WRITE_SUPPORT to be enabled.  These are to be used at your
own caution.  Since the MAC addresses are stored in the NVRAM, if you
dump one NIC and restore it on another NIC the destination NIC's
MAC addresses will not be preserved.  A tool can be made using these
sysctl's to manage the on-chip firmware.

Reviewed by:	davidch, yongari

END
K 10
svn:author
V 4
jkim
K 8
svn:date
V 27
2010-10-06T18:51:22.616858Z
K 7
svn:log
V 192
Simplify timeout check in futex_wait() using itimerfix() and return error
if the given timeout is invalid.  Consistently use int type for timeout and
correct a format string in futex_sleep().

END
K 10
svn:author
V 7
weongyo
K 8
svn:date
V 27
2010-10-06T19:32:12.074124Z
K 7
svn:log
V 136
A code cleanup of usbd_transfer_setup() to remove `while (1)'.  It's
seperated into two step and added comments what each step's doing.

END
K 10
svn:author
V 3
dim
K 8
svn:date
V 27
2010-10-06T20:35:07.305319Z
K 7
svn:log
V 294
Apply r207674 from the clangbsd project branch:

Make "clang -print-multi-os-directory" return "." on amd64, matching
gcc's behaviour.  This is needed because some ports use the option to
determine the installation directory for their libraries.

Requested by:	kwm
Approved by:	rpaulo (mentor)

END
K 10
svn:author
V 3
imp
K 8
svn:date
V 27
2010-10-06T20:49:45.959739Z
K 7
svn:log
V 184
Use more portable errx instead of errc.  The latter doesn't buy us
anyting anyway, since the error EAGAIN's error message doesn't add
anything to the error strings that are there now.

END
K 10
svn:author
V 3
imp
K 8
svn:date
V 27
2010-10-06T20:54:17.924935Z
K 7
svn:log
V 234
Remove hack needed by 6.x machines and older machines to run newer
makes on FreeBSD/pc98.  The need for this hack has passed.  If you are
one of the rare people that may need this, then you should setenv
MACHINE=pc98 as a workaround.

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-06T21:23:57.360568Z
K 7
svn:log
V 968
Add more comments to rings supported by the controller. Different
versions of controller support different number of ring control
blocks such that adjust code a bit to access known number of
send/receive ring control blocks. Previously bge(4) blindly
accessed 16 send/receive RCBs. Also move initializing standard
receive producer ring producer index, jumbo receive producer ring
producer index and mini receive producer ring producer index to
the end of each receive producer ring initialization.

Do not assume mini receive producer ring is available only when
controller has jumbo frame capability, instead explicitly check
ASIC version BCM5700 to disable mini receive producer ring.

Additionally always enable send ring 0 regardless of controller
versions. Previously bge(4) didn't enable send ring 0 if controller
is BGE_IS_5705_PLUS. Becase bge(4) need 1 send ring to send frames
at least, I have no idea how it would have worked so far.

Submitted by:	davidch

END
K 10
svn:author
V 6
cognet
K 8
svn:date
V 27
2010-10-06T22:25:21.670497Z
K 7
svn:log
V 569
if_ate.c:

    * Support for sam9 "EMAC" controller.
    * Support for rmii interface to phy.

at91.c & at91sam9.c:

    * Eliminate separate at91sam9.c file.
    * Add new devices to at91sam9_devs table.

at91_machdep.c & at at91sam9_machdep.c:

    * Automatic chip type determination.
    * Remove compile time chip dependencies.
    * Eliminate separate at91sam9_machdep.c file.

at91_pmc.c:

    * Corrected support for all of the sam926? and sam9g20 chips.
    * Remove compile time chip dependencies.

My apologies to Greg for taking so long to take care of it.

END
K 10
svn:author
V 6
cognet
K 8
svn:date
V 27
2010-10-06T22:32:31.059757Z
K 7
svn:log
V 63
Add the AT91SAM9G20EK config files.

Submitted by:	Greg Ansley

END
K 10
svn:author
V 6
cognet
K 8
svn:date
V 27
2010-10-06T22:40:27.900151Z
K 7
svn:log
V 59
Add support for the AT91SAM9260

Submitted by:	Greg Ansley

END
K 10
svn:author
V 6
cognet
K 8
svn:date
V 27
2010-10-06T22:41:32.547869Z
K 7
svn:log
V 58
Add the QILA9G20 config files.

Submitted by:	Greg Ansley

END
K 10
svn:author
V 7
weongyo
K 8
svn:date
V 27
2010-10-06T23:04:07.339272Z
K 7
svn:log
V 176
Rewrites USB_DMATAG_TO_XROOT macro to remove linux-like container_of()
macro.  Sometimes it's useful because it don't need to handle the parent
pointer but it's not intuitive.

END
K 10
svn:author
V 7
weongyo
K 8
svn:date
V 27
2010-10-06T23:13:30.282928Z
K 7
svn:log
V 47
Uses USB_P2U macro that it's doing same thing.

END
K 10
svn:author
V 7
weongyo
K 8
svn:date
V 27
2010-10-06T23:47:52.165324Z
K 7
svn:log
V 144
Cleans up ..._BUS2SC() macros that don't need to use container_of()-like
macros because the usb_bus already knows how to get the softc pointer.

END
K 10
svn:author
V 7
delphij
K 8
svn:date
V 27
2010-10-07T00:29:07.970605Z
K 7
svn:log
V 466
MFC r213044:

In the past gunzip(1) write()'s after each inflate return.  This is
not optimal from a performance standpoint since the write buffer is
not necessarily be filled up when the inflate rountine reached the
end of input buffer and it's not the end of file.

This problem gets uncovered by trying to pipe gunzip -c output to
a GEOM device directly, which enforces the writes be multiple of
sector size.

Sponsored by: iXsystems, Inc.
Reported by:  jpaetzel

END
K 10
svn:author
V 7
delphij
K 8
svn:date
V 27
2010-10-07T00:30:29.937285Z
K 7
svn:log
V 466
MFC r210344:

In the past gunzip(1) write()'s after each inflate return.  This is
not optimal from a performance standpoint since the write buffer is
not necessarily be filled up when the inflate rountine reached the
end of input buffer and it's not the end of file.

This problem gets uncovered by trying to pipe gunzip -c output to
a GEOM device directly, which enforces the writes be multiple of
sector size.

Sponsored by: iXsystems, Inc.
Reported by:  jpaetzel

END
K 10
svn:author
V 7
delphij
K 8
svn:date
V 27
2010-10-07T00:36:58.653411Z
K 7
svn:log
V 1345
MFC r211059:

Address an edge condition that we found at work, where the carp(4)
interface goes to issue LINK_UP, then LINK_DOWN, then LINK_UP at
cold boot.  This behavior is not observed when carp(4) interface
is created slightly later, when the underlying interface is fully
up.

Before this change what happen at boot is roughly:

 - ifconfig creates em0 interface;
 - ifconfig clones a carp device using em0;
   (em0's link state is DOWN at this point)
 - carp state: INIT -> BACKUP [*]
 - carp state: BACKUP -> MASTER
 - [Some negotiate between em0 and switch]
 - em0 kicks up link state change event
   (em0's link state is now up DOWN at this point)
 - do_link_state_change() -> carp_carpdev_state()
 - carp state: MASTER -> INIT (via carp_set_state(sc, INIT)) [+]
 - carp state: INIT -> BACKUP
 - carp state: BACKUP -> MASTER

At the [*] stage, em0 did not received any broadcast message from other
node, and assume our node is the master, thus carp(4) sets the link
state to "UP" after becoming a master.  At [+], the master status
is forcely set to "INIT", then an election is casted, after which our
node would actually become a master.

We believe that at the [*] stage, the master status should remain as
"INIT" since the underlying parent interface's link state is not up.

Obtained from:  iXsystems, Inc.
Reported by:    jpaetzel

END
K 10
svn:author
V 7
delphij
K 8
svn:date
V 27
2010-10-07T00:37:40.199520Z
K 7
svn:log
V 1345
MFC r211059:

Address an edge condition that we found at work, where the carp(4)
interface goes to issue LINK_UP, then LINK_DOWN, then LINK_UP at
cold boot.  This behavior is not observed when carp(4) interface
is created slightly later, when the underlying interface is fully
up.

Before this change what happen at boot is roughly:

 - ifconfig creates em0 interface;
 - ifconfig clones a carp device using em0;
   (em0's link state is DOWN at this point)
 - carp state: INIT -> BACKUP [*]
 - carp state: BACKUP -> MASTER
 - [Some negotiate between em0 and switch]
 - em0 kicks up link state change event
   (em0's link state is now up DOWN at this point)
 - do_link_state_change() -> carp_carpdev_state()
 - carp state: MASTER -> INIT (via carp_set_state(sc, INIT)) [+]
 - carp state: INIT -> BACKUP
 - carp state: BACKUP -> MASTER

At the [*] stage, em0 did not received any broadcast message from other
node, and assume our node is the master, thus carp(4) sets the link
state to "UP" after becoming a master.  At [+], the master status
is forcely set to "INIT", then an election is casted, after which our
node would actually become a master.

We believe that at the [*] stage, the master status should remain as
"INIT" since the underlying parent interface's link state is not up.

Obtained from:  iXsystems, Inc.
Reported by:    jpaetzel

END
K 10
svn:author
V 6
gordon
K 8
svn:date
V 27
2010-10-07T06:34:47.066994Z
K 7
svn:log
V 275
Add the ability to display specific manual pages if passed on the
commandline. This mirrors the old (undocumented) GNU man functionality.
Also document this feature in the implementation notes section of
the manpage.

Submitted by:	arundel
Approved by:	wes (mentor implicit)

END
K 10
svn:author
V 3
kib
K 8
svn:date
V 27
2010-10-07T08:36:02.132673Z
K 7
svn:log
V 392
The msdosfs lookup is case insensitive. Several aliases may be inserted for
a single directory entry. As a consequnce, name cache purge done by lookup
for fvp when DELETE op for namei is specified, might be not enough to
expunge all namecache entries that were installed for this direntry.

Explicitely call cache_purge(fvp) when msdosfs_rename() succeeded.

PR:	kern/93634
MFC after:	1 week

END
K 10
svn:author
V 3
mav
K 8
svn:date
V 27
2010-10-07T09:07:48.413428Z
K 7
svn:log
V 81
Add ID for Vodafone (ZTE) Mobile Broadband K3565-Z modem.

Reviewed by:	hselasky

END
K 10
svn:author
V 6
cognet
K 8
svn:date
V 27
2010-10-07T09:30:35.467001Z
K 7
svn:log
V 89
Add options I missed in the additionnal AT91 support commits.

Submitted by:	Greg Ansley

END
K 10
svn:author
V 7
davidxu
K 8
svn:date
V 27
2010-10-07T12:20:50.227423Z
K 7
svn:log
V 121
MFC r213258:
- kern_sched_rr_get_interval should return interval for thread 1 in
    target process.
- eliminate a goto.

END
K 10
svn:author
V 7
rwatson
K 8
svn:date
V 27
2010-10-07T12:34:54.624719Z
K 7
svn:log
V 160
Add Jonathan Anderson <jonathan> as a src committer.  Jon has been working
on a number of things, not least Capsicum.  I will be his mentor.

Approved by:	core

END
K 10
svn:author
V 7
davidxu
K 8
svn:date
V 27
2010-10-07T12:57:16.041609Z
K 7
svn:log
V 76
MFC r211833:
Decrease rdlock count only when thread unlocked a reader lock.

END
K 10
svn:author
V 7
pluknet
K 8
svn:date
V 27
2010-10-07T12:58:09.919282Z
K 7
svn:log
V 204
Cosmetic change:
Wrap lines for SiS and Promise entries to 80 columns.
This doesn't change appearance in page rendered with troff or man.cgi.

PR:		docs/146195
Approved by:	kib (mentor)
MFC after:	1 week

END
K 10
svn:author
V 7
davidxu
K 8
svn:date
V 27
2010-10-07T13:13:02.089643Z
K 7
svn:log
V 123
MFC r213258:
- kern_sched_rr_get_interval should return interval for thread 1 in
      target process.
- eliminate a goto.

END
K 10
svn:author
V 7
attilio
K 8
svn:date
V 27
2010-10-07T15:04:16.042059Z
K 7
svn:log
V 44
Fix a reversed use of arguments in bcopy().

END
K 10
svn:author
V 7
attilio
K 8
svn:date
V 27
2010-10-07T15:17:16.265277Z
K 7
svn:log
V 49
Remove bcopy usage and use memcpy when possible.

END
K 10
svn:author
V 3
dim
K 8
svn:date
V 27
2010-10-07T16:32:35.489295Z
K 7
svn:log
V 126
Vendor import of llvm 2.8 release:
http://llvm.org/svn/llvm-project/llvm/tags/RELEASE_28@115866

Approved by:	rpaulo (mentor)

END
K 10
svn:author
V 3
dim
K 8
svn:date
V 27
2010-10-07T16:34:12.861419Z
K 7
svn:log
V 52
Tag llvm 2.8 release.

Approved by:	rpaulo (mentor)

END
K 10
svn:author
V 3
dim
K 8
svn:date
V 27
2010-10-07T16:35:05.127376Z
K 7
svn:log
V 126
Vendor import of clang 2.8 release:
http://llvm.org/svn/llvm-project/cfe/tags/RELEASE_28@115870

Approved by:	rpaulo (mentor)

END
K 10
svn:author
V 3
dim
K 8
svn:date
V 27
2010-10-07T16:35:50.455231Z
K 7
svn:log
V 53
Tag clang 2.8 release.

Approved by:	rpaulo (mentor)

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-07T17:14:50.747929Z
K 7
svn:log
V 461
Fix a long standing bug which regarded some revisions of controller
as 5788. This caused BGE_MISC_LOCAL_CTL register is used to
generate link state change interrupt for non-5788 controllers. The
interrupt handler may or may not detect link state attention as
status block wouldn't be updated when an interrupt was generated
with BGE_MISC_LOCAL_CTL register. All controllers except 5700 and
5788 should use host coalescing mode register to trigger an
interrupt.

END
K 10
svn:author
V 6
emaste
K 8
svn:date
V 27
2010-10-07T17:26:22.193869Z
K 7
svn:log
V 74
Add a note on the removal of copyright strings from login(1) and sshd(8).

END
K 10
svn:author
V 7
weongyo
K 8
svn:date
V 27
2010-10-07T17:35:10.146807Z
K 7
svn:log
V 101
Removes recurse_1 and recurse_2 variables of `struct usb_xfer_queue'
that it's replaced as `status'.

END
K 10
svn:author
V 7
weongyo
K 8
svn:date
V 27
2010-10-07T17:49:19.975603Z
K 7
svn:log
V 52
Enhances a comment what xfer_setup is really doing.

END
K 10
svn:author
V 2
jh
K 8
svn:date
V 27
2010-10-07T18:00:55.304545Z
K 7
svn:log
V 571
Check the device name validity on device registration.

A new function prep_devname() sanitizes a device name by removing
leading and redundant sequential slashes. The function returns an error
for names which already exist or are considered invalid.

A new flag MAKEDEV_CHECKNAME for make_dev_p(9) and make_dev_credf(9)
indicates that the caller is prepared to handle an error related to the
device name. An invalid name triggers a panic if the flag is not
specified.

Document the MAKEDEV_CHECKNAME flag in the make_dev(9) manual page.

Idea from:	kib
Reviewed by:	kib

END
K 10
svn:author
V 3
avg
K 8
svn:date
V 27
2010-10-07T18:11:33.348729Z
K 7
svn:log
V 155
vm.kmem_map_size: a sysctl to query current kmem_map->size

Based on a patch from Sandvine Incorporated via emaste.

Reviewed by:	emaste
MFC after:	1 week

END
K 10
svn:author
V 3
avg
K 8
svn:date
V 27
2010-10-07T18:16:14.014059Z
K 7
svn:log
V 234
opensolaris_kmem kmem_size(): report lesser of vm_kmem_size and available
physical memory

This is needed to correctly autotune ZFS ARC size when vm_kmem_size is
set to value larger than available physical memory.

MFC after:	2 weeks

END
K 10
svn:author
V 3
pjd
K 8
svn:date
V 27
2010-10-07T18:16:22.958461Z
K 7
svn:log
V 340
Don't close local component on exit as we can hang waiting on g_waitidle.
I'm unable to reproduce the race described in comment anymore and also the
comment is incorrect - localfd represents local component from configuration
file, eg. /dev/da0 and not HAST provider.

Reported by:	Mikolaj Golub <to.my.trociny@gmail.com>
MFC after:	1 week

END
K 10
svn:author
V 3
pjd
K 8
svn:date
V 27
2010-10-07T18:19:02.576827Z
K 7
svn:log
V 151
Start the guard thread first, so we can handle signals from the very begining.

Reported by:	Mikolaj Golub <to.my.trociny@gmail.com>
MFC after:	1 week

END
K 10
svn:author
V 3
pjd
K 8
svn:date
V 27
2010-10-07T18:20:16.108779Z
K 7
svn:log
V 77
Log error message when we fail to destroy ggate provider.

MFC after:	3 days

END
K 10
svn:author
V 8
jonathan
K 8
svn:date
V 27
2010-10-07T18:20:51.754829Z
K 7
svn:log
V 67
Add myself (jonathan@) to committers-src.dot

Reviewed by:	rwatson

END
K 10
svn:author
V 3
pjd
K 8
svn:date
V 27
2010-10-07T18:23:28.556507Z
K 7
svn:log
V 141
Clear ggate structures before using them. We don't initialize all the field
and there can be some garbage from the stack.

MFC after:	1 week

END
K 10
svn:author
V 3
dim
K 8
svn:date
V 27
2010-10-07T20:31:07.391191Z
K 7
svn:log
V 154
Upgrade Clang and LLVM to the 2.8 release.  See here for release notes:
http://llvm.org/releases/2.8/docs/ReleaseNotes.html

Approved by:	rpaulo (mentor)

END
K 10
svn:author
V 3
ken
K 8
svn:date
V 27
2010-10-07T21:56:10.256537Z
K 7
svn:log
V 1403
Turn on serialization of task management commands going down to the
controller, but make it optional.

After a problem report from Andrew Boyer, it looks like the LSI
chip may have issues (the watchdog timer fired) if too many aborts
are sent down to the chip at the same time.  We know that task
management commands are serialized, and although the manual doesn't
say it, it may be a good idea to just send one at a time.

But, since I'm not certain that this is necessary, add a tunable
and sysctl variable (hw.mps.%d.allow_multiple_tm_cmds) to control
the driver's behavior.

mps.c:		Add support for the sysctl and tunable, and add a
		comment about the possible return values to
		mps_map_command().

mps_sas.c:	Run all task management commands through two new
		routines, mpssas_issue_tm_request() and
		mpssas_complete_tm_request().

		This allows us to optionally serialize task
		management commands.  Also, change things so that
		the response to a task management command always
		comes back through the callback.  (Before it could
		come via the callback or the return value.)

mpsvar.h:	Add softc variables for the list of active task
		management commands, the number of active commands,
		and whether we should allow multiple active task
		management commands.  Add an active command flag.

mps.4:		Describe the new sysctl/loader tunable variable.

Sponsored by:	Spectra Logic Corporation

END
K 10
svn:author
V 6
emaste
K 8
svn:date
V 27
2010-10-08T00:44:53.255395Z
K 7
svn:log
V 209
Make a thread's address available via the kern proc sysctl, just like the
process address.

Add "tdaddr" keyword to ps(1) to display this thread address.

Distilled from Sandvine's patch set by Mark Johnston.

END
K 10
svn:author
V 6
emaste
K 8
svn:date
V 27
2010-10-08T01:03:30.937343Z
K 7
svn:log
V 281
In r207768 I silenced a console warning from rum(4).  There was legitimate
opposition to the change, since really we need to implement missing
functionality in drivers or the 802.3 layer.

For now, restore a reminder message for a missing rum_update_mcast, but
print it only once.

END
K 10
svn:author
V 7
weongyo
K 8
svn:date
V 27
2010-10-08T01:17:22.175527Z
K 7
svn:log
V 64
Defines the helper macro SLEEPOUT_RUN_TASK to enqueue the task.

END
K 10
svn:author
V 7
davidxu
K 8
svn:date
V 27
2010-10-08T01:47:14.068472Z
K 7
svn:log
V 58
unwind.h was imported, gcc directory is no longer needed.

END
K 10
svn:author
V 7
weongyo
K 8
svn:date
V 27
2010-10-08T01:52:01.359797Z
K 7
svn:log
V 270
o fixes a regression that setting the promiscuous mode should be
  happened at the taskqueue.  It's to avoid a `sleepable after
  non-sleepable' because ioctl handler could be called with holding bpf
  mtx which is a default mutex.
o defines SLEEPOUT_DRAIN_TASK helper.

END
K 10
svn:author
V 7
davidxu
K 8
svn:date
V 27
2010-10-08T01:53:33.467411Z
K 7
svn:log
V 38
Oops, don't remove -fexceptions flag.

END
K 10
svn:author
V 7
weongyo
K 8
svn:date
V 27
2010-10-08T01:54:27.479343Z
K 7
svn:log
V 26
A trivial macro renaming.

END
K 10
svn:author
V 3
kib
K 8
svn:date
V 27
2010-10-08T07:17:22.163376Z
K 7
svn:log
V 105
Add a comment describing the reason for calling cache_purge(fvp).

Requested by:	danfe
MFC after:	6 days

END
K 10
svn:author
V 3
kib
K 8
svn:date
V 27
2010-10-08T07:18:44.238975Z
K 7
svn:log
V 51
Fix typo.

Submitted by:	arundel
MFC after:	3 days

END
K 10
svn:author
V 3
kib
K 8
svn:date
V 27
2010-10-08T07:19:05.751097Z
K 7
svn:log
V 7
Regen.

END
K 10
svn:author
V 3
avg
K 8
svn:date
V 27
2010-10-08T07:26:56.091030Z
K 7
svn:log
V 90
MFC r212950: zfs_vnops: use zfs_map_page/zfs_unmap_page helper functions
in another place

END
K 10
svn:author
V 3
avg
K 8
svn:date
V 27
2010-10-08T07:29:26.659199Z
K 7
svn:log
V 84
MFC r212951: zfs_map_page/zfs_unmap_page: do not use sched_pin() and
SFB_CPUPRIVATE

END
K 10
svn:author
V 3
avg
K 8
svn:date
V 27
2010-10-08T07:31:32.842012Z
K 7
svn:log
V 79
MFC r212657: zfs vn_has_cached_data: take into account v_object->cache != NULL

END
K 10
svn:author
V 3
avg
K 8
svn:date
V 27
2010-10-08T07:33:22.500514Z
K 7
svn:log
V 83
MFC r212780: zfs arc_reclaim_needed: more reasonable threshold for
available pages

END
K 10
svn:author
V 3
avg
K 8
svn:date
V 27
2010-10-08T07:34:32.699093Z
K 7
svn:log
V 69
MFC r212783: zfs arc_reclaim_needed: fix typo in mismerge in r212780

END
K 10
svn:author
V 3
avg
K 8
svn:date
V 27
2010-10-08T07:36:30.966007Z
K 7
svn:log
V 89
MFC r212781: zfs arc_reclaim_needed: remove redundant checks for
arc_c_max and arc_c_max

END
K 10
svn:author
V 3
avg
K 8
svn:date
V 27
2010-10-08T07:38:44.927392Z
K 7
svn:log
V 87
MFC r212964: kdb_backtrace: stack(9)-based code to print backtrace
without any backend

END
K 10
svn:author
V 3
avg
K 8
svn:date
V 27
2010-10-08T07:49:05.996334Z
K 7
svn:log
V 71
MFC r212994: kdb_backtrace: use stack_print_ddb instead of stack_print

END
K 10
svn:author
V 3
avg
K 8
svn:date
V 27
2010-10-08T07:52:00.668643Z
K 7
svn:log
V 67
MFC r213303: kmem_size* sysctls: hint that these are also tunables

END
K 10
svn:author
V 3
avg
K 8
svn:date
V 27
2010-10-08T07:54:55.175376Z
K 7
svn:log
V 67
MFC r213303: kmem_size* sysctls: hint that these are also tunables

END
K 10
svn:author
V 3
avg
K 8
svn:date
V 27
2010-10-08T07:57:19.691360Z
K 7
svn:log
V 74
MFC r213304: debug.kdb.stop_cpus sysctl: hint that this is also a
tunable

END
K 10
svn:author
V 3
avg
K 8
svn:date
V 27
2010-10-08T07:58:57.734842Z
K 7
svn:log
V 74
MFC r213304: debug.kdb.stop_cpus sysctl: hint that this is also a
tunable

END
K 10
svn:author
V 3
avg
K 8
svn:date
V 27
2010-10-08T08:01:36.873800Z
K 7
svn:log
V 89
MFC r213305: there must be only one SYSINIT with
SI_SUB_RUN_SCHEDULER+SI_ORDER_ANY order

END
K 10
svn:author
V 3
avg
K 8
svn:date
V 27
2010-10-08T08:09:45.722091Z
K 7
svn:log
V 89
MFC r213305: there must be only one SYSINIT with
SI_SUB_RUN_SCHEDULER+SI_ORDER_ANY order

END
K 10
svn:author
V 3
avg
K 8
svn:date
V 27
2010-10-08T08:13:22.332816Z
K 7
svn:log
V 57
MFC r213322: sysctls in kern_shutdown: add twin tunables

END
K 10
svn:author
V 3
avg
K 8
svn:date
V 27
2010-10-08T08:14:43.154909Z
K 7
svn:log
V 57
MFC r213322: sysctls in kern_shutdown: add twin tunables

END
K 10
svn:author
V 2
ed
K 8
svn:date
V 27
2010-10-08T09:32:37.958615Z
K 7
svn:log
V 17
Import ee 1.5.1.

END
K 10
svn:author
V 2
ed
K 8
svn:date
V 27
2010-10-08T09:33:21.854827Z
K 7
svn:log
V 14
Tag ee 1.5.1.

END
K 10
svn:author
V 2
ed
K 8
svn:date
V 27
2010-10-08T09:33:52.389759Z
K 7
svn:log
V 26
Make tag name consistent.

END
K 10
svn:author
V 2
ed
K 8
svn:date
V 27
2010-10-08T09:34:59.273838Z
K 7
svn:log
V 17
Import ee 1.5.2.

END
K 10
svn:author
V 2
ed
K 8
svn:date
V 27
2010-10-08T09:35:23.027202Z
K 7
svn:log
V 14
Tag ee 1.5.2.

END
K 10
svn:author
V 2
ed
K 8
svn:date
V 27
2010-10-08T09:43:06.575890Z
K 7
svn:log
V 63
Import ee 1.5.2 into HEAD.

PR:		bin/145921
MFC after:	2 weeks

END
K 10
svn:author
V 3
pho
K 8
svn:date
V 27
2010-10-08T10:27:52.918529Z
K 7
svn:log
V 105
Ensure that a make from sys/boot does not pick up include files from
/usr/include.

Discussed with:	 kib

END
K 10
svn:author
V 7
attilio
K 8
svn:date
V 27
2010-10-08T10:42:16.721776Z
K 7
svn:log
V 491
Add the support for polling on ixgbe. It is mostly modelled on what
ixgb already does.

Small questions:
- does the IFF_DRV_RUNNING check might be kept unlocked? I can't find
  an uniform locking pattern for it looking around and it seems
  mostly unlocked.
- reading IXGBE_EICR and _LSC should be unlocked? It seems functions do
  this way.
- In case ixgbe_rxeof() returns 'more' should we schedule the taskqueue?
  ixgb polling doesn't do that, so I'd like to hear more opinion in return.

END
K 10
svn:author
V 7
attilio
K 8
svn:date
V 27
2010-10-08T11:16:48.078016Z
K 7
svn:log
V 43
Add netdump support for ixgb(4) interface.

END
K 10
svn:author
V 7
attilio
K 8
svn:date
V 27
2010-10-08T12:39:05.977952Z
K 7
svn:log
V 41
Add netdump support for ixgbe(4) driver.

END
K 10
svn:author
V 3
uqs
K 8
svn:date
V 27
2010-10-08T12:39:49.760329Z
K 7
svn:log
V 53
mdoc: fix manlint warnings by unbreaking mdoc syntax

END
K 10
svn:author
V 3
uqs
K 8
svn:date
V 27
2010-10-08T12:40:16.194835Z
K 7
svn:log
V 100
mdoc: drop redundant .Pp and .LP calls

They have no effect when coming in pairs, or before .Bl/.Bd

END
K 10
svn:author
V 7
pluknet
K 8
svn:date
V 27
2010-10-08T14:31:49.874656Z
K 7
svn:log
V 442
Clean up tools in tools/tools/netrate.

- tcpconnect incorrectly uses err() in usage() with errx() semantics [1]
  That produces dirty error message:
  tcpconnect: usage: tcpconnect [ip]: Unknown error: 0
- 64-bit aware fixes in printf() usage [2]
o   netrate/juggle: time_t has arch-dependent size
o   netrate/tcpconnect: don't assume %ll has always 64bit.

PR:		146088 [1], 146086 [2] (modified)
Approved by:	kib (mentor)
MFC after:	1 week

END
K 10
svn:author
V 6
marius
K 8
svn:date
V 27
2010-10-08T14:50:58.751788Z
K 7
svn:log
V 72
Just use the sparc64 version of this header rather than duplicating it.

END
K 10
svn:author
V 6
emaste
K 8
svn:date
V 27
2010-10-08T14:56:39.030937Z
K 7
svn:log
V 334
MFC r212570:

  Allow a kernel config to specify a set but empty value via
  'makeoptions OPTION=' for consistency with the make commandline.

  Previously 'makeoptions WERROR=' would result in a syntax error; now
  it produces the same effect as 'makeoptions WERROR'.  Both forms now
  result in 'WERROR=' in the generated Makefile.

END
K 10
svn:author
V 6
emaste
K 8
svn:date
V 27
2010-10-08T14:59:14.811474Z
K 7
svn:log
V 177
MFC r213013:

  Move test for zero bufp or size before rseq and wseq calculation.  This
  avoids spinning in an infinite loop for some (possibly corrupt?) core
  files at work.

END
K 10
svn:author
V 6
marius
K 8
svn:date
V 27
2010-10-08T14:59:45.454870Z
K 7
svn:log
V 235
In the replacement text of the __bswapN_const() macros cast the argument
to the expected type so they work like the corresponding __bswapN_var()
functions and the compiler doesn't complain when arguments of different
width are passed.

END
K 10
svn:author
V 3
pjd
K 8
svn:date
V 27
2010-10-08T15:02:15.577028Z
K 7
svn:log
V 263
We close the event socketpair early in the mainloop to prevent spaming with
error messages, so when we clean up after child process, we have to check if
the event socketpair is still there.

Submitted by:	Mikolaj Golub <to.my.trociny@gmail.com>
MFC after:	3 days

END
K 10
svn:author
V 3
pjd
K 8
svn:date
V 27
2010-10-08T15:05:39.062410Z
K 7
svn:log
V 172
We can't zero out ggio request, as we have some fields in there we initialize
once during start-up.

Reported by:	Mikolaj Golub <to.my.trociny@gmail.com>
MFC after:	3 days

END
K 10
svn:author
V 7
attilio
K 8
svn:date
V 27
2010-10-08T15:10:30.569499Z
K 7
svn:log
V 60
Add tunables that can set up netdump earlier than userland.

END
K 10
svn:author
V 3
uqs
K 8
svn:date
V 27
2010-10-08T15:20:20.622750Z
K 7
svn:log
V 103
rm(1): clarify that -P works only when blocks are updated in-place

Suggested by:	pjd, ivoras, arundel

END
K 10
svn:author
V 7
attilio
K 8
svn:date
V 27
2010-10-08T15:59:02.091487Z
K 7
svn:log
V 68
- Lock properly ifnet list for scanning
- virtualize the ifnet list

END
K 10
svn:author
V 7
attilio
K 8
svn:date
V 27
2010-10-08T16:33:25.388272Z
K 7
svn:log
V 13
MFC @ 213582

END
K 10
svn:author
V 5
markm
K 8
svn:date
V 27
2010-10-08T17:42:09.850350Z
K 7
svn:log
V 133
Don't blow away /bin/rmail symlink if we are keeping mailwrapper.
Mailwrapper can provide a perfectly good rmail with other
mailers.

END
K 10
svn:author
V 7
attilio
K 8
svn:date
V 27
2010-10-08T17:46:51.661208Z
K 7
svn:log
V 61
Remove changes that crept-in during development and testing.

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T17:58:07.239438Z
K 7
svn:log
V 216
Do not blindly UP the interface when interface's MTU is changed. If
driver is not running there is no need to up the interface. While
I'm here hold driver lock before modifying MTU as it is referenced
in RX handler.

END
K 10
svn:author
V 3
mdf
K 8
svn:date
V 27
2010-10-08T18:06:15.065482Z
K 7
svn:log
V 120
MFC r212926:

Use destroy_dev_sched(9) instead of destroy_dev(9) in passcleanup() as
it is indirectly a d_close method.

END
K 10
svn:author
V 3
mdf
K 8
svn:date
V 27
2010-10-08T18:06:30.190027Z
K 7
svn:log
V 120
MFC r212926:

Use destroy_dev_sched(9) instead of destroy_dev(9) in passcleanup() as
it is indirectly a d_close method.

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T18:19:05.988454Z
K 7
svn:log
V 806
MFC r213306:
  Rename rl_setmulti() to rl_rxfilter() as rl_rxfilter() will handle
  IFF_ALLMULTI/IFF_PROMISC as well as multicast filter configuration.
  Rewrite RX filter logic to reduce number of register accesses and
  make it handle promiscuous/allmulti toggling without controller
  reinitialization.

  Previously rl(4) counted on controller reinitialization to reprogram
  promiscuous configuration but r211767 resulted in avoiding
  controller reinitialization whenever promiscuous mode is toggled.
  To address this, keep track of driver's view of interface state and
  handle IFF_ALLMULTI/IFF_PROMISC changes without reinitializing
  controller. This should fix a regression introduced in r211267.

  While I'm here remove unnecessary variable reassignment in ioctl
  handler.

  PR:	kern/151079

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T18:26:14.578446Z
K 7
svn:log
V 806
MFC r213306:
  Rename rl_setmulti() to rl_rxfilter() as rl_rxfilter() will handle
  IFF_ALLMULTI/IFF_PROMISC as well as multicast filter configuration.
  Rewrite RX filter logic to reduce number of register accesses and
  make it handle promiscuous/allmulti toggling without controller
  reinitialization.

  Previously rl(4) counted on controller reinitialization to reprogram
  promiscuous configuration but r211767 resulted in avoiding
  controller reinitialization whenever promiscuous mode is toggled.
  To address this, keep track of driver's view of interface state and
  handle IFF_ALLMULTI/IFF_PROMISC changes without reinitializing
  controller. This should fix a regression introduced in r211267.

  While I'm here remove unnecessary variable reassignment in ioctl
  handler.

  PR:	kern/151079

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T18:31:56.026117Z
K 7
svn:log
V 102
MFC r188392:
  Destroy TX tag outside of loop scope.

  Found with:	Coverity Prevent(tm)
  CID:		3886

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T18:43:06.089629Z
K 7
svn:log
V 1881
MFC r212061,212065,212302:
r212061:
  Split common parent DMA tag into ring DMA tag and TX/RX mbuf DMA
  tag. All controllers that are not BCM5755 or higher have 4GB
  boundary DMA bug. Previously bge(4) used 32bit DMA address to
  workaround the bug(r199670). However this caused the use of bounce
  buffers such that it resulted in poor performance for systems which
  have more than 4GB memory. Because bus_dma(9) honors boundary
  restriction requirement of DMA tag for dynamic buffers, having a
  separate TX/RX mbuf DMA tag will greatly reduce the possibility of
  using bounce buffers. For DMA buffers allocated with
  bus_dmamem_alloc(9), now bge(4) explicitly checks whether the
  requested memory region crossed the boundary or not.
  With this change, only the DMA buffer that crossed the boundary
  will use 32bit DMA address. Other DMA buffers are not affected as
  separate DMA tag is created for each DMA buffer.
  Even if 32bit DMA address space is used for a buffer, the chance to
  use bounce buffer is still very low as the size of buffer is small.
  This change should eliminate most usage of bounce buffers on
  systems that have more than 4GB memory.

  More correct fix would be teaching bus_dma(9) to honor boundary
  restriction for buffers created with bus_dmamem_alloc(9) but it
  seems that is not easy.

  While I'm here cleanup bge_dma_map_addr() and remove unnecessary
  member variables in bge_dmamap_arg structure.

  Tested by:	marcel

r212065:
  Handle PAE case correctly. You cannot effectively specify a 4GB
  boundary in PAE case so use a 2GB boundary for PAE as suggested by
  jhb.

  Pointed out by:	jhb
  Reviewed by:	jhb

r212302:
  Make sure to create DMA'able memory for statistics block. This was
  missed in r212061 and it caused crashes for 570x controllers as
  controller DMAed statistics to physical address 0.

  Reported by:	kan

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T18:46:02.426300Z
K 7
svn:log
V 1881
MFC r212061,212065,212302:
r212061:
  Split common parent DMA tag into ring DMA tag and TX/RX mbuf DMA
  tag. All controllers that are not BCM5755 or higher have 4GB
  boundary DMA bug. Previously bge(4) used 32bit DMA address to
  workaround the bug(r199670). However this caused the use of bounce
  buffers such that it resulted in poor performance for systems which
  have more than 4GB memory. Because bus_dma(9) honors boundary
  restriction requirement of DMA tag for dynamic buffers, having a
  separate TX/RX mbuf DMA tag will greatly reduce the possibility of
  using bounce buffers. For DMA buffers allocated with
  bus_dmamem_alloc(9), now bge(4) explicitly checks whether the
  requested memory region crossed the boundary or not.
  With this change, only the DMA buffer that crossed the boundary
  will use 32bit DMA address. Other DMA buffers are not affected as
  separate DMA tag is created for each DMA buffer.
  Even if 32bit DMA address space is used for a buffer, the chance to
  use bounce buffer is still very low as the size of buffer is small.
  This change should eliminate most usage of bounce buffers on
  systems that have more than 4GB memory.

  More correct fix would be teaching bus_dma(9) to honor boundary
  restriction for buffers created with bus_dmamem_alloc(9) but it
  seems that is not easy.

  While I'm here cleanup bge_dma_map_addr() and remove unnecessary
  member variables in bge_dmamap_arg structure.

  Tested by:	marcel

r212065:
  Handle PAE case correctly. You cannot effectively specify a 4GB
  boundary in PAE case so use a 2GB boundary for PAE as suggested by
  jhb.

  Pointed out by:	jhb
  Reviewed by:	jhb

r212302:
  Make sure to create DMA'able memory for statistics block. This was
  missed in r212061 and it caused crashes for 570x controllers as
  controller DMAed statistics to physical address 0.

  Reported by:	kan

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T18:49:59.447317Z
K 7
svn:log
V 476
MFC r212069,212071:
r212069:
  bge_txeof() already checks whether it has to free transmitted mbufs
  or not by comparing reported TX consumer index with saved index. So
  remove unnecessary check done after freeing transmitted mbufs.
  While I'm here nuke unnecessary variable initializations.

r212071:
  Remove unnecessary atomic operation in bge_poll. bge(4) always
  holds a driver lock in the function entry and
  memory synchronization is handled by bus_dmamap_sync(9).

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T18:51:28.439637Z
K 7
svn:log
V 476
MFC r212069,212071:
r212069:
  bge_txeof() already checks whether it has to free transmitted mbufs
  or not by comparing reported TX consumer index with saved index. So
  remove unnecessary check done after freeing transmitted mbufs.
  While I'm here nuke unnecessary variable initializations.

r212071:
  Remove unnecessary atomic operation in bge_poll. bge(4) always
  holds a driver lock in the function entry and
  memory synchronization is handled by bus_dmamap_sync(9).

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T18:58:01.471435Z
K 7
svn:log
V 268
MFC r212306-212307,212342:
r212306:
  Remove trailing CR at EOL.

r212307:
  Consistently use tab characters instead of tab + space characters.
  No functional changes.

r212342:
  Correct definition of T2 mode bit of MRBE Message Page 5 Next Page
  Control Register.

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T19:00:36.299927Z
K 7
svn:log
V 268
MFC r212306-212307,212342:
r212306:
  Remove trailing CR at EOL.

r212307:
  Consistently use tab characters instead of tab + space characters.
  No functional changes.

r212342:
  Correct definition of T2 mode bit of MRBE Message Page 5 Next Page
  Control Register.

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T19:19:51.283153Z
K 7
svn:log
V 81
MFC r212968:
  Remove unnecessary controller reinitialization.

  PR:	kern/87506

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T19:21:16.091799Z
K 7
svn:log
V 81
MFC r212968:
  Remove unnecessary controller reinitialization.

  PR:	kern/87506

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T19:24:38.037085Z
K 7
svn:log
V 164
MFC r212969:
  Make sure to clear IFF_DRV_RUNNING to reinitialize controller.
  While I'm here update if_oerrors counter when driver encounters
  watchdog timeout.

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T19:25:46.978025Z
K 7
svn:log
V 164
MFC r212969:
  Make sure to clear IFF_DRV_RUNNING to reinitialize controller.
  While I'm here update if_oerrors counter when driver encounters
  watchdog timeout.

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T19:27:34.775745Z
K 7
svn:log
V 288
MFC r212971:
  Remove unnecessary controller reinitialization.
  StarFire controller does not require controller reinitialization to
  program perfect filters. While here, make driver immediately exit
  from interrupt/polling handler if driver reinitialized controller.

  PR:	kern/87506

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T19:29:17.127963Z
K 7
svn:log
V 288
MFC r212971:
  Remove unnecessary controller reinitialization.
  StarFire controller does not require controller reinitialization to
  program perfect filters. While here, make driver immediately exit
  from interrupt/polling handler if driver reinitialized controller.

  PR:	kern/87506

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T19:32:39.536117Z
K 7
svn:log
V 150
MFC r212972,21300:
r212972:
  Remove unnecessary controller reinitialization.

  PR:	kern/87506

r213000:
  Fix build breakage introduced in r212972.

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T19:34:23.162297Z
K 7
svn:log
V 150
MFC r212972,21300:
r212972:
  Remove unnecessary controller reinitialization.

  PR:	kern/87506

r213000:
  Fix build breakage introduced in r212972.

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T19:39:35.046726Z
K 7
svn:log
V 203
MFC r212103-212105:
r212103:
  Remove leading and trailing white spaces.
  No functional changes.

r212104:
  style(9) - space after keywords.

r212105:
  Nuke unnecessary return at the end of function.

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T20:12:00.725484Z
K 7
svn:log
V 376
MFC r182065:
  There actually were bugs in the original handling that I missed last
  night.

  Free the children after each pci bus that is searched.  Otherwise we
  leak them.  With free in the new place, we also have to free children
  before going to done when we find the device we're looking for.

  Also, if we can't get the children of a device, just ignore that bus.

END
K 10
svn:author
V 5
markm
K 8
svn:date
V 27
2010-10-08T20:13:12.079994Z
K 7
svn:log
V 55
Extend the "example" section a bit, for other mailers.

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T20:13:42.578062Z
K 7
svn:log
V 203
MFC r212103-212105:
r212103:
  Remove leading and trailing white spaces.
  No functional changes.

r212104:
  style(9) - space after keywords.

r212105:
  Nuke unnecessary return at the end of function.

END
K 10
svn:author
V 5
markm
K 8
svn:date
V 27
2010-10-08T20:17:17.943500Z
K 7
svn:log
V 179
Don't let the "available"/"not available" logs flap quite so badly. In the
case of a multi-interfaced host, sub-second flapping can fill up the logs
too quickly. This fixes that.

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T20:18:44.983811Z
K 7
svn:log
V 2369
MFC r212109,212124,212185:
r212109:
  bus_dma(9) cleanup.
   o Enforce TX/RX descriptor ring alignment. NS data sheet says the
     controller needs 4 bytes alignment but use 16 to cover both SiS
     and NS controllers. I don't have SiS data sheet so I'm not sure
     what is alignment restriction of SiS controller but 16 would be
     enough because it's larger than the size of a TX/RX descriptor.
     Previously sis(4) ignored the alignment restriction.
   o Enforce RX buffer alignment, 4.
     Previously sis(4) ignored RX buffer alignment restriction.
   o Limit number of TX DMA segment to be used to 16. It seems
     controller has no restriction on number of DMA segments but
     using more than 16 looks resource waste.
   o Collapse long mbuf chains with m_collapse(9) instead of calling
     expensive m_defrag(9).
   o TX/RX side bus_dmamap_load_mbuf_sg(9) support and remove
     unnecessary callbacks.
   o Initial endianness support.
   o Prefer local alignment fixup code to m_devget(9).
   o Pre-allocate TX/RX mbuf DMA maps instead of creating/destroying
     these maps in fast TX/RX path. On non-x86 architectures, this is
     very expensive operation and there is no need to do that.
   o Add missing bus_dmamap_sync(9) in TX/RX path.
   o watchdog is now unarmed only when there are no pending frames
     on controller. Previously sis(4) blindly unarmed watchdog
     without checking the number of queued frames.
   o For efficiency, loaded DMA map is reused for error frames.
   o DMA map loading failure is now gracefully handled. Previously
     sis(4) ignored any DMA map loading errors.
   o Nuke unused macros which are not appropriate for endianness
     operation.
   o Stop embedding driver maintained structures into descriptor
     rings. Because TX/RX descriptor structures are shared between
     host and controller, frequent bus_dmamap_sync(9) operations are
     required in fast path. Embedding driver structures will increase
     the size of DMA map which in turn will slow down performance.

r212124:
  Fix stupid error in r212109 which didn't swap DMA maps. This caused
  IOMMU panic on sparc64 under high TX load.

r212185:
  Fix another bug introduced in r212109. We should unload DMA maps
  only after sending the last fragment of a frame so the mbuf pointer
  also should be stored in the last descriptor index.

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T20:27:51.416115Z
K 7
svn:log
V 2369
MFC r212109,212124,212185:
r212109:
  bus_dma(9) cleanup.
   o Enforce TX/RX descriptor ring alignment. NS data sheet says the
     controller needs 4 bytes alignment but use 16 to cover both SiS
     and NS controllers. I don't have SiS data sheet so I'm not sure
     what is alignment restriction of SiS controller but 16 would be
     enough because it's larger than the size of a TX/RX descriptor.
     Previously sis(4) ignored the alignment restriction.
   o Enforce RX buffer alignment, 4.
     Previously sis(4) ignored RX buffer alignment restriction.
   o Limit number of TX DMA segment to be used to 16. It seems
     controller has no restriction on number of DMA segments but
     using more than 16 looks resource waste.
   o Collapse long mbuf chains with m_collapse(9) instead of calling
     expensive m_defrag(9).
   o TX/RX side bus_dmamap_load_mbuf_sg(9) support and remove
     unnecessary callbacks.
   o Initial endianness support.
   o Prefer local alignment fixup code to m_devget(9).
   o Pre-allocate TX/RX mbuf DMA maps instead of creating/destroying
     these maps in fast TX/RX path. On non-x86 architectures, this is
     very expensive operation and there is no need to do that.
   o Add missing bus_dmamap_sync(9) in TX/RX path.
   o watchdog is now unarmed only when there are no pending frames
     on controller. Previously sis(4) blindly unarmed watchdog
     without checking the number of queued frames.
   o For efficiency, loaded DMA map is reused for error frames.
   o DMA map loading failure is now gracefully handled. Previously
     sis(4) ignored any DMA map loading errors.
   o Nuke unused macros which are not appropriate for endianness
     operation.
   o Stop embedding driver maintained structures into descriptor
     rings. Because TX/RX descriptor structures are shared between
     host and controller, frequent bus_dmamap_sync(9) operations are
     required in fast path. Embedding driver structures will increase
     the size of DMA map which in turn will slow down performance.

r212124:
  Fix stupid error in r212109 which didn't swap DMA maps. This caused
  IOMMU panic on sparc64 under high TX load.

r212185:
  Fix another bug introduced in r212109. We should unload DMA maps
  only after sending the last fragment of a frame so the mbuf pointer
  also should be stored in the last descriptor index.

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T20:30:08.173076Z
K 7
svn:log
V 38
MFC r212113:
  s/u_intXX_t/uintXX_t/g

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T20:31:35.689724Z
K 7
svn:log
V 38
MFC r212113:
  s/u_intXX_t/uintXX_t/g

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T20:33:43.001013Z
K 7
svn:log
V 1069
MFC r212116:
  Overhaul link state change handling. Previously sis(4) blindly
  configured TX/RX MACs before getting a valid link. After that, when
  link state change callback is called, it called device
  initialization again to reconfigure TX/RX MACs depending on
  resolved link state. This hack created several bad side effects and
  it required more hacks to not collide with sis_tick callback as
  well as disabling switching to currently selected media in device
  initialization. Also it seems sis(4) was used to be a template
  driver for long time so other drivers which was modeled after
  sis(4) also should be changed.

  TX/RX MACs are now reconfigured after getting a valid link. Fix for
  short cable error is also applied after getting a link because it's
  only valid when the resolved speed is 100Mbps.

  While I'm here slightly reorganize interrupt handler such that
  sis(4) always read SIS_ISR register to see whether the interrupt is
  ours or not. This change removes another hack and make it possible
  to nuke sis_stopped variable in softc.

END
K 10
svn:author
V 7
weongyo
K 8
svn:date
V 27
2010-10-08T20:34:23.836112Z
K 7
svn:log
V 42
Adds a comment why the taskqueue is used.

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T20:37:13.441207Z
K 7
svn:log
V 1069
MFC r212116:
  Overhaul link state change handling. Previously sis(4) blindly
  configured TX/RX MACs before getting a valid link. After that, when
  link state change callback is called, it called device
  initialization again to reconfigure TX/RX MACs depending on
  resolved link state. This hack created several bad side effects and
  it required more hacks to not collide with sis_tick callback as
  well as disabling switching to currently selected media in device
  initialization. Also it seems sis(4) was used to be a template
  driver for long time so other drivers which was modeled after
  sis(4) also should be changed.

  TX/RX MACs are now reconfigured after getting a valid link. Fix for
  short cable error is also applied after getting a link because it's
  only valid when the resolved speed is 100Mbps.

  While I'm here slightly reorganize interrupt handler such that
  sis(4) always read SIS_ISR register to see whether the interrupt is
  ours or not. This change removes another hack and make it possible
  to nuke sis_stopped variable in softc.

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T20:39:45.029173Z
K 7
svn:log
V 406
MFC r212117,212119:
rr212117:
  Report result of link state change to caller. Previously it always
  returned success.

r212119:
  Do not reinitialize controller whenever promiscuous mode or
  allmulti is toggled. Controller does not require reinitialization.
  This removes unnecessary controller reinitialization whenever
  tcpdump is used.
  While I'm here remove unnecessary variable reinitialization.

END
K 10
svn:author
V 3
dim
K 8
svn:date
V 27
2010-10-08T20:40:05.308448Z
K 7
svn:log
V 158
Remove superfluous cast in usr.bin/netstat/sctp.c.

Found by:	clang
Submitted by:	Norberto Lopes, nlopes dot ml at gmail dot com
Approved by:	rpaulo (mentor)

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T20:41:15.108538Z
K 7
svn:log
V 406
MFC r212117,212119:
rr212117:
  Report result of link state change to caller. Previously it always
  returned success.

r212119:
  Do not reinitialize controller whenever promiscuous mode or
  allmulti is toggled. Controller does not require reinitialization.
  This removes unnecessary controller reinitialization whenever
  tcpdump is used.
  While I'm here remove unnecessary variable reinitialization.

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T20:42:59.669314Z
K 7
svn:log
V 139
MFC r212120:
  Avoid controller reinitialization which could be triggered by
  dhclient(8) or alias addresses are added.

  PR:	kern/87506

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T20:44:35.002508Z
K 7
svn:log
V 139
MFC r212120:
  Avoid controller reinitialization which could be triggered by
  dhclient(8) or alias addresses are added.

  PR:	kern/87506

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T20:48:09.312057Z
K 7
svn:log
V 543
MFC r212121,212156:
r212121:
  Move sis_reset() to sis_initl(). This ensures driver starts with
  known good state of controller.

r212156:
  Fix the last endianness issue on handling station address which
  prevented driver from working on big-endian machines. Also rewrite
  station address programming to make it work on strict-alignment
  architectures. With this change, sis(4) now works on sparc64 and
  performance number looks good even though sis(4) have to apply
  fixup code to align received frames on 2 bytes boundary on sparc64.

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T20:49:44.061067Z
K 7
svn:log
V 543
MFC r212121,212156:
r212121:
  Move sis_reset() to sis_initl(). This ensures driver starts with
  known good state of controller.

r212156:
  Fix the last endianness issue on handling station address which
  prevented driver from working on big-endian machines. Also rewrite
  station address programming to make it work on strict-alignment
  architectures. With this change, sis(4) now works on sparc64 and
  performance number looks good even though sis(4) have to apply
  fixup code to align received frames on 2 bytes boundary on sparc64.

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T20:51:33.880155Z
K 7
svn:log
V 543
MFC r212157:
  Unlike most other controllers, NS DP83815/DP83816 controllers seem
  to pad with 0xFF when it encounter short frames.  According to RFC
  1042 the pad bytes should be 0x00.
  Because manual padding consumes extra CPU cycles, introduce a new
  tunable which controls the padding behavior. Turning this tunable
  on will have driver pad manually but it's disabled by default. Users
  can enable software padding by setting the following tunable to
  non-zero value.

  dev.sis.%d.manual_pad="1"

  PR:	kern/35422 (patch not used)

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T20:52:47.821272Z
K 7
svn:log
V 543
MFC r212157:
  Unlike most other controllers, NS DP83815/DP83816 controllers seem
  to pad with 0xFF when it encounter short frames.  According to RFC
  1042 the pad bytes should be 0x00.
  Because manual padding consumes extra CPU cycles, introduce a new
  tunable which controls the padding behavior. Turning this tunable
  on will have driver pad manually but it's disabled by default. Users
  can enable software padding by setting the following tunable to
  non-zero value.

  dev.sis.%d.manual_pad="1"

  PR:	kern/35422 (patch not used)

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T20:54:09.448203Z
K 7
svn:log
V 71
MFC r212158:
  Enable sis(4). sis(4) should work on all architectures.

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T20:55:27.556345Z
K 7
svn:log
V 71
MFC r212158:
  Enable sis(4). sis(4) should work on all architectures.

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T20:58:05.639015Z
K 7
svn:log
V 197
MFC r212159,212161:
r212159:
  Document tunable dev.sis.%unit.manual_pad
  While I'm here Xref vlan(4) as sis(4) supports VLAN oversized
  frames.

r212161:
  Better wording.

  Submitted by:	jkim

END
K 10
svn:author
V 7
yongari
K 8
svn:date
V 27
2010-10-08T20:58:48.981967Z
K 7
svn:log
V 197
MFC r212159,212161:
r212159:
  Document tunable dev.sis.%unit.manual_pad
  While I'm here Xref vlan(4) as sis(4) supports VLAN oversized
  frames.

r212161:
  Better wording.

  Submitted by:	jkim

END