]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
13 years agoWhile inspecting the disklabel check that start offset of partition is
ae [Thu, 27 Jan 2011 08:02:26 +0000 (08:02 +0000)]
While inspecting the disklabel check that start offset of partition is
within provider's bounds. If not then reject this disklabel.
Mark bbarea as NULL to do not free it again in destroy method.

MFC after: 1 week

13 years agoMake a note to re-check whether that particular check is needed.
adrian [Thu, 27 Jan 2011 07:33:17 +0000 (07:33 +0000)]
Make a note to re-check whether that particular check is needed.

13 years agoQuick fix to a comment.
gnn [Thu, 27 Jan 2011 03:32:16 +0000 (03:32 +0000)]
Quick fix to a comment.

13 years agoWriting to the analog registers on the AR9220 (Merlin PCI) seems to require a delay.
adrian [Thu, 27 Jan 2011 02:56:03 +0000 (02:56 +0000)]
Writing to the analog registers on the AR9220 (Merlin PCI) seems to require a delay.

This, along with an initval change which will appear in a subsequent commit,
fixes bus panics that I have been seing with the AR9220 on a Routerstation Pro
(AR7161 MIPS board.)

Obtained from: Linux ath9k
PR: kern/154220

13 years agoExplicitly wire the user buffer rather than doing it implicitly in
mdf [Thu, 27 Jan 2011 00:34:12 +0000 (00:34 +0000)]
Explicitly wire the user buffer rather than doing it implicitly in
sbuf_new_for_sysctl(9).  This allows using an sbuf with a SYSCTL_OUT
drain for extremely large amounts of data where the caller knows that
appropriate references are held, and sleeping is not an issue.

Inspired by: rwatson

13 years agoRemove the CTLFLAG_NOLOCK as it seems to be both unused and
mdf [Wed, 26 Jan 2011 22:48:09 +0000 (22:48 +0000)]
Remove the CTLFLAG_NOLOCK as it seems to be both unused and
unfunctional.  Wiring the user buffer has only been done explicitly
since r101422.

Mark the kern.disks sysctl as MPSAFE since it is and it seems to have
been mis-using the NOLOCK flag.

Partially break the KPI (but not the KBI) for the sysctl_req 'lock'
field since this member should be private and the "REQ_LOCKED" state
seems meaningless now.

13 years agoDocument newly added tunables.
yongari [Wed, 26 Jan 2011 21:59:59 +0000 (21:59 +0000)]
Document newly added tunables.
 hw.re.intr_filter
 hw.re.msix_disable
 dev.re.%d.int_rx_mod

13 years ago* Use 300 ms as the default for RTO_MIN.
tuexen [Wed, 26 Jan 2011 21:38:17 +0000 (21:38 +0000)]
* Use 300 ms as the default for RTO_MIN.
* Disable burst mitigation by default.
* Remove unused constant.
Discussed with rrs.
MFC after: 3 months.

13 years agoAdd support for RTL8105E PCIe Fast Ethernet controller. It seems
yongari [Wed, 26 Jan 2011 21:14:20 +0000 (21:14 +0000)]
Add support for RTL8105E PCIe Fast Ethernet controller. It seems
the controller has a kind of embedded controller/memory and vendor
applies a large set of magic code via undocumented PHY registers in
device initialization stage. I guess it's a firmware image for the
embedded controller in RTL8105E since the code is too big compared
to other DSP fixups. However I have no idea what that magic code
does and what's purpose of the embedded controller. Fortunately
driver seems to still work without loading the firmware.

While I'm here change device description of RTL810xE controller.

H/W donated by: Realtek Semiconductor Corp.

13 years agoAdd Realtek RTL8201E 10/100 PHY found in RTL8105E controller. The
yongari [Wed, 26 Jan 2011 21:07:44 +0000 (21:07 +0000)]
Add Realtek RTL8201E 10/100 PHY found in RTL8105E controller. The
exact model name is not clear yet. All previous RTL8201 10/100 PHYs
used 0x8201 in MII_PHYIDR2 which in turn makes model number 0x20
but this PHY used new model number 0x08.

13 years agoDo not use interrupt taskqueue on controllers with MSI/MSI-X
yongari [Wed, 26 Jan 2011 20:25:40 +0000 (20:25 +0000)]
Do not use interrupt taskqueue on controllers with MSI/MSI-X
capability. One of reason using interrupt taskqueue in re(4) was
to reduce number of TX/RX interrupts under load because re(4)
controllers have no good TX/RX interrupt moderation mechanism.
Basic TX interrupt moderation is done by hardware for most
controllers but RX interrupt moderation through undocumented
register showed poor RX performance so it was disabled in r215025.
Using taskqueue to handle RX interrupt greatly reduced number of
interrupts but re(4) consumed all available CPU cycles to run the
taskqueue under high TX/RX network load.  This can happen even with
RTL810x fast ethernet controller and I believe this is not
acceptable for most systems.

To mitigate the issue, use one-shot timer register to moderate RX
interrupts. The timer register provides programmable one-shot timer
and can be used to suppress interrupt generation. The timer runs at
125MHZ on PCIe controllers so the minimum time allowed for the
timer is 8ns. Data sheet says the register is 32 bits but
experimentation shows only lower 13 bits are valid so maximum time
that can be programmed is 65.528us. This yields theoretical maximum
number of RX interrupts that could be generated per second is about
15260. Combined with TX completion interrupts re(4) shall generate
less than 20k interrupts. This number is still slightly high
compared to other intelligent ethernet controllers but system is
very responsive even under high network load.

Introduce sysctl variable dev.re.%d.int_rx_mod that controls amount
of time to delay RX interrupt processing in units of us. Value 0
completely disables RX interrupt moderation. To provide old
behavior for controllers that have MSI/MSI-X capability, introduce
a new tunable hw.re.intr_filter. If the tunable is set to non-zero
value, driver will use interrupt taskqueue. The default value of
the tunable is 0. This tunable has no effect on controllers that
has no MSI/MSI-X capability or if MSI/MSI-X is explicitly disabled
by administrator.

While I'm here cleanup interrupt setup/teardown since re(4) uses
single MSI/MSI-X message at this moment.

13 years agoAdd macro to test the sv_flags of any process. Change some places to test
dchagin [Wed, 26 Jan 2011 20:03:58 +0000 (20:03 +0000)]
Add macro to test the sv_flags of any process. Change some places to test
the flags instead of explicit comparing with address of known sysentvec
structures.

MFC after: 1 month

13 years agoMake SCTP_MAX_BURST compliant with the latest version of
tuexen [Wed, 26 Jan 2011 19:55:54 +0000 (19:55 +0000)]
Make SCTP_MAX_BURST compliant with the latest version of
the socket API ID. This is not compatible with the API
in stable/8.

13 years agoChange infrastructure for SCTP_MAX_BURST to allow compliance
tuexen [Wed, 26 Jan 2011 19:49:03 +0000 (19:49 +0000)]
Change infrastructure for SCTP_MAX_BURST to allow compliance
with the latest socket API ID. Especially it can be disabled.

Full compliance needs changing the structure used in the
socket option. Since this breaks the API, it will be a
seperate commit which will not be MFCed to stable/8.

MFC after: 3 months.

13 years agoUnroll vgrind filter to no longer depend on vgrind(1) during buildworld.
uqs [Wed, 26 Jan 2011 19:39:31 +0000 (19:39 +0000)]
Unroll vgrind filter to no longer depend on vgrind(1) during buildworld.

The source files haven't been touched in ages and this is unlikely
to change in the future.

13 years agoFix typo in example getopt(1) script: $i vs $1 [1]
uqs [Wed, 26 Jan 2011 18:43:15 +0000 (18:43 +0000)]
Fix typo in example getopt(1) script: $i vs $1 [1]
While here apply style hammer.

PR: docs/154289 [1]
Submitted by: Jamie Landeg Jones <jamie@bishopston.net>
MFC after: 1 week

13 years agoPrison check addresses set with multicast interface options.
deischen [Wed, 26 Jan 2011 17:31:03 +0000 (17:31 +0000)]
Prison check addresses set with multicast interface options.

Reviewed by: bz
MFC after: 1 week

13 years agoSet td_kstack_pages for thread0. This was already being done for most
mdf [Wed, 26 Jan 2011 17:06:13 +0000 (17:06 +0000)]
Set td_kstack_pages for thread0.  This was already being done for most
architectures, but i386 and amd64 were missing it.

Submitted by: Mohd Fahadullah <mfahadullah AT isilon DOT com>

13 years agoAdd missing part of r217877.
mav [Wed, 26 Jan 2011 13:10:07 +0000 (13:10 +0000)]
Add missing part of r217877.

13 years agoAdd ar5416RestoreChainMask() which will undo any AR5416 specific chainmask
adrian [Wed, 26 Jan 2011 10:48:29 +0000 (10:48 +0000)]
Add ar5416RestoreChainMask() which will undo any AR5416 specific chainmask
overriding after calibration.

This will get set for other two chain radios, such as AR9280 and later on,
AR9287. It should however be a nul operation.

13 years agoAdd an AR5416 workaround - force a different bias based on 2.4ghz channel frequency.
adrian [Wed, 26 Jan 2011 10:36:43 +0000 (10:36 +0000)]
Add an AR5416 workaround - force a different bias based on 2.4ghz channel frequency.

Obtained from: Linux ath9k

13 years agoTreat async buffer writes from the gjournal switcher thread the same as
kib [Wed, 26 Jan 2011 10:34:21 +0000 (10:34 +0000)]
Treat async buffer writes from the gjournal switcher thread the same as
from syncer. We shall not sleep on running buffer space when suspending.

Reproduced and tested by: pho
PR: kern/154228
MFC after: 1 week

13 years agoBreak out the chainmask init code into a new function - ar5416InitChainMasks() .
adrian [Wed, 26 Jan 2011 10:08:37 +0000 (10:08 +0000)]
Break out the chainmask init code into a new function - ar5416InitChainMasks() .

ath9k does a few different things here during config - if it's an early
AR5416 with two chains, it enables all three chains for calibration and
then restores the chainmask to the original values after initial
calibration has completed.

The reason behind this commit is to begin breaking out the chainmask
configuration for this specific reason; follow-up commits will add
the chainmask restore in the ar5416Reset() routine.

13 years ago* fix HAL_DEBUG_INTERRUPT to be a separate bit, it was overlapping with
adrian [Wed, 26 Jan 2011 09:37:43 +0000 (09:37 +0000)]
* fix HAL_DEBUG_INTERRUPT to be a separate bit, it was overlapping with
  something else
* add HAL_DEBUG_GPIO, for some GPIO related debugging I'm tinkering with
  at the moment.

13 years agoHardware supported by siis(4) allows software control over activity LEDs.
mav [Wed, 26 Jan 2011 08:54:10 +0000 (08:54 +0000)]
Hardware supported by siis(4) allows software control over activity LEDs.
Expose that functionality to led(4) OR-ing it with regular LED activity.

13 years agoIn addition to r217444 ignore also ATA status errors on DMA Auto-Activation
mav [Wed, 26 Jan 2011 06:57:48 +0000 (06:57 +0000)]
In addition to r217444 ignore also ATA status errors on DMA Auto-Activation
enabling request. Some HP disks reported to return ABORT error there while
declaring support for this feature.

13 years agoMake device initialization sequence shorter when possible. Do not enable/
mav [Wed, 26 Jan 2011 06:37:51 +0000 (06:37 +0000)]
Make device initialization sequence shorter when possible. Do not enable/
disable already enabled/disabled SATA features.

13 years agoAdd -H flag to print thread id.
dchagin [Wed, 26 Jan 2011 06:36:14 +0000 (06:36 +0000)]
Add -H flag to print thread id.

13 years agoFix another broken date
dougb [Wed, 26 Jan 2011 06:07:24 +0000 (06:07 +0000)]
Fix another broken date

13 years agoClarify the availability of the noatime option on network file systems
dougb [Wed, 26 Jan 2011 05:06:11 +0000 (05:06 +0000)]
Clarify the availability of the noatime option on network file systems

13 years agoFix my fix to nfe.4, and also fix re.4
dougb [Wed, 26 Jan 2011 01:07:56 +0000 (01:07 +0000)]
Fix my fix to nfe.4, and also fix re.4

So just when I thought my pointy hat collection was going down ...

13 years agoFix date in .Dd
dougb [Wed, 26 Jan 2011 01:02:39 +0000 (01:02 +0000)]
Fix date in .Dd

13 years agoRemove TX taskqueue and directly invoke re_start in interrupt task.
yongari [Tue, 25 Jan 2011 23:27:28 +0000 (23:27 +0000)]
Remove TX taskqueue and directly invoke re_start in interrupt task.

13 years agoFix compilation with debug on.
cracauer [Tue, 25 Jan 2011 22:25:16 +0000 (22:25 +0000)]
Fix compilation with debug on.
Fix segfault when TargetAddress is missing or mis-spelled in config file.

13 years agoRemove dead code.
jhb [Tue, 25 Jan 2011 22:21:05 +0000 (22:21 +0000)]
Remove dead code.

MFC after: 2 weeks

13 years agoPrefer MSI-X to MSI on controllers that support MSI-X. All
yongari [Tue, 25 Jan 2011 22:18:00 +0000 (22:18 +0000)]
Prefer MSI-X to MSI on controllers that support MSI-X. All
recent PCIe controllers(RTL8102E or later and RTL8168/8111C or
later) supports either 2 or 4 MSI-X messages. Unfortunately vendor
did not publicly release RSS related information yet. However
switching to MSI-X is one-step forward to support RSS.

13 years agoWhen loading dso without PT_GNU_STACK phdr, only call
kib [Tue, 25 Jan 2011 21:12:31 +0000 (21:12 +0000)]
When loading dso without PT_GNU_STACK phdr, only call
__pthread_map_stacks_exec() on architectures that allow executable
stacks.

Reported and tested by: marcel (ia64)

13 years agoEmit .note.GNU-stack for the syscall stubs generated by libc only on
kib [Tue, 25 Jan 2011 21:06:49 +0000 (21:06 +0000)]
Emit .note.GNU-stack for the syscall stubs generated by libc only on
architectures that support this .note. In particular, do not unneccessary
emit the notes on ia64 and sparc64, which ABI require non-executable stacks.

Tested by: marcel

13 years agoFix get_program_var_addr() when type of the resolved symbol is functional.
kib [Tue, 25 Jan 2011 21:04:55 +0000 (21:04 +0000)]
Fix get_program_var_addr() when type of the resolved symbol is functional.
Use make_function_pointer then, otherwise ia64 is broken.

Reported and tested by: marcel

13 years agosh: Clean up some old comments:
jilles [Tue, 25 Jan 2011 20:56:18 +0000 (20:56 +0000)]
sh: Clean up some old comments:
* There is no plan for an alternative to the command "set".
* Attempting to unset a readonly variable has not raised an error for quite
  a while, so the order of unsetting a variable and a function with the same
  name does not matter.

MFC after: 1 week

13 years agoDo not allocate buffer to hold data for zero-sized sections.
kan [Tue, 25 Jan 2011 19:17:50 +0000 (19:17 +0000)]
Do not allocate buffer to hold data for zero-sized sections.

13 years agoDisable TSO for all Realtek controllers. Experimentation showed
yongari [Tue, 25 Jan 2011 19:05:46 +0000 (19:05 +0000)]
Disable TSO for all Realtek controllers. Experimentation showed
RTL8111C generated corrupted frames where TCP option header was
broken. All other sample controllers I have did not show such
problem so it could be RTL8111C specific issue. Because there are
too many variants it's hard to tell how many controllers have such
issue. Just disable TSO by default but have user override it.

13 years agoUse test(1) operators and test for the catpage not being older than the
uqs [Tue, 25 Jan 2011 18:32:21 +0000 (18:32 +0000)]
Use test(1) operators and test for the catpage not being older than the
manpage.

Identical mtimes (as generated by buildworld for these files) precluded
catpages from working.

Approved by: gordon

13 years agoDocument sbuf_new_for_sysctl(9).
mdf [Tue, 25 Jan 2011 17:39:52 +0000 (17:39 +0000)]
Document sbuf_new_for_sysctl(9).

Pointed out by: lstewart

13 years agoWhen matching an incoming ARP against a bridge, ensure both interfaces belong
thompsa [Tue, 25 Jan 2011 17:15:23 +0000 (17:15 +0000)]
When matching an incoming ARP against a bridge, ensure both interfaces belong
to the same bridge.

Submitted by: Alexander Zagrebin

13 years agoRemove DOS-style EOLs
sem [Tue, 25 Jan 2011 15:06:50 +0000 (15:06 +0000)]
Remove DOS-style EOLs

Approved by: kib

13 years agoWhen vtruncbuf() iterates over the vnode buffer list, lock buffer object
kib [Tue, 25 Jan 2011 14:04:02 +0000 (14:04 +0000)]
When vtruncbuf() iterates over the vnode buffer list, lock buffer object
before checking the validity of the next buffer pointer. Otherwise, the
buffer might be reclaimed after the check, causing iteration to run into
wrong buffer.

Reported and tested by: pho
MFC after: 1 week

13 years agoBio shall not be accessed after g_io_deliver(9).
kib [Tue, 25 Jan 2011 14:00:30 +0000 (14:00 +0000)]
Bio shall not be accessed after g_io_deliver(9).

Reported and tested by: pho
Reviewed by: ae, phk
MFC after: 1 week

13 years agoFix logic error introduced in previous commit.
jpaetzel [Tue, 25 Jan 2011 13:41:48 +0000 (13:41 +0000)]
Fix logic error introduced in previous commit.
Along the way make some efficiency improvements.

Submitted by: jilles
Approved by: kib (mentor)
MFC after: 3 days

13 years agoDocument P_FOLLOWFORK.
kib [Tue, 25 Jan 2011 11:04:16 +0000 (11:04 +0000)]
Document P_FOLLOWFORK.

MFC after: 2 weeks

13 years agoDocument PT_FLAG_FORKED, PT_FOLLOW_FORK, pl_tdname and pl_child_pid.
kib [Tue, 25 Jan 2011 11:02:12 +0000 (11:02 +0000)]
Document PT_FLAG_FORKED, PT_FOLLOW_FORK, pl_tdname and pl_child_pid.

MFC after: 2 weeks

13 years agoAllow debugger to specify that children of the traced process should be
kib [Tue, 25 Jan 2011 10:59:21 +0000 (10:59 +0000)]
Allow debugger to specify that children of the traced process should be
automatically traced. Extend the ptrace(PL_LWPINFO) to report that child
just forked.

Reviewed by: davidxu, jhb
MFC after: 2 weeks

13 years agoBack out r217734. [1]
bz [Tue, 25 Jan 2011 08:13:31 +0000 (08:13 +0000)]
Back out r217734. [1]

Properly document what `make targets` is supposed to list to avoid
further confusion given the place the target sits. Should have happened
with r217125.

Requested by: imp [1]
Reviewed by: rwatson

13 years ago* Re-format the v4k header to be consistent
adrian [Tue, 25 Jan 2011 07:37:12 +0000 (07:37 +0000)]
* Re-format the v4k header to be consistent
* Re-do the structure size/component math to make sure the struct matches
  the expected size
* Just to be clear that we care about bitmask ordering, revert my previous
  change and instead define that macro if we're on big-endian.

13 years agoBring over a fix from ath9k - zero some of the TX descriptors for Kite/AR9285.
adrian [Tue, 25 Jan 2011 05:47:50 +0000 (05:47 +0000)]
Bring over a fix from ath9k - zero some of the TX descriptors for Kite/AR9285.

Kite doesn't have per-chain control (it has one chain) or antenna control; so
don't try to set those descriptor entries.

13 years agoRename this linux-ism __BIG_ENDIAN_BITFIELD macro to something suitable for FreeBSD.
adrian [Tue, 25 Jan 2011 05:41:36 +0000 (05:41 +0000)]
Rename this linux-ism __BIG_ENDIAN_BITFIELD macro to something suitable for FreeBSD.

Warner has pointed out that FreeBSD's bit orders follow byte orders.

13 years agoCommit updated AR9285 (Kite) v2 initvals from ath9k.
adrian [Tue, 25 Jan 2011 05:36:29 +0000 (05:36 +0000)]
Commit updated AR9285 (Kite) v2 initvals from ath9k.

13 years agoFix this tool to use the updated (corrected) v4k eeprom definition introduced in...
adrian [Tue, 25 Jan 2011 05:35:43 +0000 (05:35 +0000)]
Fix this tool to use the updated (corrected) v4k eeprom definition introduced in a previous commit.

13 years agoFix the Atheros V4K EEPROM definitions to match those in ath9k.
adrian [Tue, 25 Jan 2011 05:35:09 +0000 (05:35 +0000)]
Fix the Atheros V4K EEPROM definitions to match those in ath9k.

It turns out that the V4K eeprom definitions (used by the AR9285 and
its derivatives) is wrong. These values are at least causing issues
on my AR2427.

With this fix (and initvals in a subsequent commit), the AR2427 behaves
a lot better.

Note - there's still significant drift between the ath9k v4k eeprom
init code (again, used by AR9285 and derivatives) and what's in this
tree. That needs to be investigated and resolved.

13 years agoSupply maximum value as an argument to the decimal() function
sobomax [Tue, 25 Jan 2011 04:35:07 +0000 (04:35 +0000)]
Supply maximum value as an argument to the decimal() function
instead of supplying number of bits.

Submitted by: bde

13 years agoImport the ERTT (Enhanced Round Trip Time) Khelp module. ERTT uses the
lstewart [Mon, 24 Jan 2011 23:08:38 +0000 (23:08 +0000)]
Import the ERTT (Enhanced Round Trip Time) Khelp module. ERTT uses the
Khelp/Hhook KPIs to hook into the TCP stack and maintain a per-connection, low
noise estimate of the instantaneous RTT. ERTT's implementation is robust even in
the face of delayed acknowledgements and/or TSO being in use for a connection.

A high quality, low noise RTT estimate is a requirement for applications such as
delay-based congestion control, for which we will be importing some algorithm
implementations shortly.

In collaboration with: David Hayes <dahayes at swin edu au> and
Grenville Armitage <garmitage at swin edu au>
Sponsored by: FreeBSD Foundation
Reviewed by: bz and others along the way
MFC after: 3 months

13 years agoFix a LOR by dropping the global ifnet locks while allocating a new ifnet
jhb [Mon, 24 Jan 2011 22:21:58 +0000 (22:21 +0000)]
Fix a LOR by dropping the global ifnet locks while allocating a new ifnet
table in if_grow().  The order of the SYSINIT's for ifnet state were swapped
so that the various locks were initialized before being used.

Reviewed by: pluknet, bz
MFC after: 2 weeks

13 years agoIn uart_tty_outwakeup(), check CTS/RTS flow control settings and
marcel [Mon, 24 Jan 2011 18:34:16 +0000 (18:34 +0000)]
In uart_tty_outwakeup(), check CTS/RTS flow control settings and
prevent sending data when CTS is de-asserted.
In uart_tty_intr(), call uart_tty_outwakeup() when the CTS signal
changed, knowing that uart_tty_outwakeup() will do the right
thing for flow control. This avoids redundant conditionals.

PR: kern/148644
Submitted by: John Wehle <john@feith.com>
MFC after: 3 days

13 years agoReinitialize driver when MTU is changed and driver is running.
yongari [Mon, 24 Jan 2011 17:53:58 +0000 (17:53 +0000)]
Reinitialize driver when MTU is changed and driver is running.

Reported by: Huang, Yusheng ( yusheng.huang <> bluecoat com )

13 years agoAdd more sanity checks for USB_HOST_ALIGN input values. Re-factor existing
hselasky [Mon, 24 Jan 2011 17:38:37 +0000 (17:38 +0000)]
Add more sanity checks for USB_HOST_ALIGN input values. Re-factor existing
checks for readability.

Approved by: thompsa (mentor)

13 years agoReplace spaces with tabs.
jh [Mon, 24 Jan 2011 17:08:26 +0000 (17:08 +0000)]
Replace spaces with tabs.

13 years agoRemove an invalid register setup; this is likely a holdover from
adrian [Mon, 24 Jan 2011 17:03:22 +0000 (17:03 +0000)]
Remove an invalid register setup; this is likely a holdover from
the AR5212 code. It doesn't exist in ath9k and I've been told it
doesn't exist in the Atheros internal driver.

13 years agoAdd missing section number for .Xr rc.
gjb [Mon, 24 Jan 2011 15:11:08 +0000 (15:11 +0000)]
Add missing section number for .Xr rc.

Pointed out by: keramida
Approved by: keramida (mentor)

13 years agoDon't open configuration file from worker process. Handle SIGHUP in the
pjd [Mon, 24 Jan 2011 15:04:15 +0000 (15:04 +0000)]
Don't open configuration file from worker process. Handle SIGHUP in the
master process only and pass changes to the worker processes over control
socket. This removes access to global namespace in preparation for capsicum
sandboxing.

MFC after: 2 weeks

13 years agoAdded a blurb about thin provisioning, fixed some formatting.
ivoras [Mon, 24 Jan 2011 14:24:10 +0000 (14:24 +0000)]
Added a blurb about thin provisioning, fixed some formatting.

13 years agolibkvm: Unbreak build for powerpc64/powerpc
uqs [Mon, 24 Jan 2011 11:06:40 +0000 (11:06 +0000)]
libkvm: Unbreak build for powerpc64/powerpc

This slipped through my testing due to the recent roto-tiling of the
ARCH/TARGET parts that I need to catch up with. Mea culpa.

13 years agoWhile printing out the WSID and CPU ID only the first time it appears for
rwatson [Mon, 24 Jan 2011 10:58:36 +0000 (10:58 +0000)]
While printing out the WSID and CPU ID only the first time it appears for
each workstream, rather than on every protocol, is prettier, it makes
machine-parsing of netstat -Q output a lot harder.  Repeat the information
and hope that the user forgives us slightly dense formatting.

MFC after: 3 days
Reported by: bz
Sponsored by: Juniper Networks

13 years agoFix off-by-one whitespace error in netstat -Q workstream listing.
rwatson [Mon, 24 Jan 2011 10:54:09 +0000 (10:54 +0000)]
Fix off-by-one whitespace error in netstat -Q workstream listing.

Reported by: bz
MFC after: 3 days
Sponsored by: Juniper Networks

13 years agoICH7 SATA controller in legacy mode can provide access to SATA registers
mav [Mon, 24 Jan 2011 09:24:20 +0000 (09:24 +0000)]
ICH7 SATA controller in legacy mode can provide access to SATA registers
via AHCI-like memory resource at BAR(5). Use it if BIOS was so kind to
allocate memory for that BAR. This allows hot-plug support and connection
speed reporting.

MFC after: 2 weeks

13 years agoAdd build infrastructure for Khelp modules.
lstewart [Mon, 24 Jan 2011 07:50:29 +0000 (07:50 +0000)]
Add build infrastructure for Khelp modules.

Sponsored by: FreeBSD Foundation
Reviewed by: bz
MFC after: 3 months

13 years agoUndo a local option which mistakenly crept into a commit.
adrian [Mon, 24 Jan 2011 07:45:14 +0000 (07:45 +0000)]
Undo a local option which mistakenly crept into a commit.

13 years agoo Cylinder numbers are 10 bits in the MBR;
sobomax [Mon, 24 Jan 2011 07:16:20 +0000 (07:16 +0000)]
o Cylinder numbers are 10 bits in the MBR;

o Sector numbers are only 6 bits in the MBR;

o bde'cize name of the local variable.

Submitted by: bde

13 years agoAdd a simple tool to print the contents of a v4k EEPROM dump.
adrian [Mon, 24 Jan 2011 06:46:03 +0000 (06:46 +0000)]
Add a simple tool to print the contents of a v4k EEPROM dump.

13 years agoThe dump, fsck_ffs, fsdb, fsirand, newfs, makefs, and quot utilities
mckusick [Mon, 24 Jan 2011 06:17:05 +0000 (06:17 +0000)]
The dump, fsck_ffs, fsdb, fsirand, newfs, makefs, and quot utilities
include sys/time.h instead of time.h. This include is incorrect as
per the manpages for the APIs and the POSIX definitions. This commit
replaces sys/time.h where necessary with time.h.

The commit also includes some minor style(9) header fixup in newfs.

This commit is part of a larger effort by Garrett Cooper started in
//depot/user/gcooper/posix-conformance-work/ -- to make FreeBSD more
POSIX compliant.

Submitted by:  Garrett Cooper   yanegomi at gmail dot com

13 years agoHonor $PAGER instead of always calling more.
gordon [Mon, 24 Jan 2011 04:32:59 +0000 (04:32 +0000)]
Honor $PAGER instead of always calling more.

Approved by: colin@
MFC after: 1 week

13 years agoApply TX interrupt moderation to all RTL810xE PCIe Fast Ethernet
yongari [Mon, 24 Jan 2011 00:01:06 +0000 (00:01 +0000)]
Apply TX interrupt moderation to all RTL810xE PCIe Fast Ethernet
controllers. Experimentation with RTL8102E, RTL8103E and RTL8105E
showed dramatic decrement of TX completion interrupts under high TX
load(e.g.  from 147k interrupts/second to 10k interrupts/second)
With this change, TX interrupt moderation is applied to all
controllers except RTL8139C+.

13 years agoAdd stream scheduling support.
tuexen [Sun, 23 Jan 2011 19:36:28 +0000 (19:36 +0000)]
Add stream scheduling support.
This work is based on a patch received from Robin Seggelmann.

MFC after: 3 months.

13 years agoDisable ATAPI DMA unconditionally on Apple Kauai ATA controllers, like it
nwhitehorn [Sun, 23 Jan 2011 17:20:11 +0000 (17:20 +0000)]
Disable ATAPI DMA unconditionally on Apple Kauai ATA controllers, like it
is on the MacIO ones. It appears to be unreliable on all DBDMA-based
controllers for unknown reasons, which should be figured out eventually.

Tested by: Torfinn Ingolfsen
MFC after: 1 week

13 years agoThis script parses output of userland tools. In the case of a faulted
jpaetzel [Sun, 23 Jan 2011 17:13:29 +0000 (17:13 +0000)]
This script parses output of userland tools.  In the case of a faulted
zpool the output causes the script to bail out with syntax errors.
Since a scrub of a faulted zpool is pointless, just skip over any pools
marked as such.

PR: conf/150228
Submitted by: jpaetzel
Approved by: kib (mentor)
MFC after: 3 days
MFC note: only for RELENG_8

13 years agoFor `make tinderbox` there is no need to print the extra commands.
bz [Sun, 23 Jan 2011 16:28:44 +0000 (16:28 +0000)]
For `make tinderbox` there is no need to print the extra commands.

MFC after: 5 days

13 years agoinit(8): Document that login(1) is now responsible for recording logouts.
jilles [Sun, 23 Jan 2011 14:54:45 +0000 (14:54 +0000)]
init(8): Document that login(1) is now responsible for recording logouts.

init(8) only uses utmpx for recording reboots and shutdowns.

13 years agoEnable the 11n PHY by default whether or not 11n is configured.
adrian [Sun, 23 Jan 2011 14:49:50 +0000 (14:49 +0000)]
Enable the 11n PHY by default whether or not 11n is configured.

The linux ath9k driver and (from what I've been told) the atheros reference
driver does this; it then leaves discarding 11n frames to the 802.11 layer.

Whilst I'm here, merge in a fix from ath9k which maintains a turbo register
setting when enabling the 11n register; and remove an un-needed (duplicate)
flag setting.

13 years agoUpdate the AR9280v2 inivals to match what is in Linux ath9k.
adrian [Sun, 23 Jan 2011 14:30:35 +0000 (14:30 +0000)]
Update the AR9280v2 inivals to match what is in Linux ath9k.

This repairs the behaviour of my AR9280 - both radio chains now seem
to correctly be receiving.

13 years agoinit: Only run /etc/rc.shutdown if /etc/rc was run.
jilles [Sun, 23 Jan 2011 14:22:26 +0000 (14:22 +0000)]
init: Only run /etc/rc.shutdown if /etc/rc was run.

It does not make sense to shut down daemons that were not started. In
particular, this fixes loss of mixer settings when shutting down using
shutdown(8), init(8) or ctrl+alt+del from single-user mode.

If /etc/rc reboots, /etc/rc.shutdown is not run.

Also fix segfaults and other erratic behaviour if init receives SIGHUP or
SIGTSTP while in single-user mode.

This commit does not attempt to fix any badness with signal handlers
(assumption that pointers can be read and written atomically, EINTR race
condition). I believe it does not make this badness any worse.

Silence on: -arch@

13 years agoUse string literal format strings to quiet clang warnings.
emaste [Sun, 23 Jan 2011 14:08:59 +0000 (14:08 +0000)]
Use string literal format strings to quiet clang warnings.

13 years agoAn sbuf configured with SBUF_AUTOEXTEND will call malloc with M_WAITOK when a
lstewart [Sun, 23 Jan 2011 13:00:25 +0000 (13:00 +0000)]
An sbuf configured with SBUF_AUTOEXTEND will call malloc with M_WAITOK when a
write to the buffer causes it to overflow. We therefore can't hold the CC list
rwlock over a call to sbuf_printf() for an sbuf configured with SBUF_AUTOEXTEND.

Switch to a fixed length sbuf which should be of sufficient size except in the
very unlikely event that the sysctl is being processed as one or more new
algorithms are loaded. If that happens, we accept the race and may fail the
sysctl gracefully if there is insufficient room to print the names of all the
algorithms.

This should address a WITNESS warning and the potential panic that would occur
if the sbuf call to malloc did sleep whilst holding the CC list rwlock.

Sponsored by: FreeBSD Foundation
Reported by: Nick Hibma
Reviewed by: bz
MFC after: 3 weeks
X-MFC with: r215166

13 years agoFix typo.
kib [Sun, 23 Jan 2011 12:44:17 +0000 (12:44 +0000)]
Fix typo.

MFC after: 3 days

13 years agoTouch up the sample memory usage numbers a bit, to avoid wrapping
keramida [Sun, 23 Jan 2011 11:08:39 +0000 (11:08 +0000)]
Touch up the sample memory usage numbers a bit, to avoid wrapping
on terminal boundary.  While here add definition for 'G' and fix
the indentation of 'K' units.

Submitted by: plunket
PR: docs/153614
MFC after: 3 days

13 years agolibkvm: fix process runtime calculation on crashdumps
uqs [Sun, 23 Jan 2011 11:08:36 +0000 (11:08 +0000)]
libkvm: fix process runtime calculation on crashdumps

Fix a long standing bug, where the procs ticks where assumed to be in
us. Instead, read cpu_tick_frequency from the kernel and use the same
logic to convert runtime. This is still too optimistic in that it
assumes cpu_tick_frequency is available and fixed. Since this function
is only called on crashdumps, I think we can live with that. Testing has
shown the values to be correct for different kern.hz inside Virtualbox.

Bump WARNS. Alignment issues on some archs mean this is still at 3.

Reviewed by: bde

13 years agolibkvm code janitoring
uqs [Sun, 23 Jan 2011 11:08:28 +0000 (11:08 +0000)]
libkvm code janitoring

- make WARNS=6 clean for archs w/o strict alignment requirments
- add const, ANSIfy, remove unused vars, cast types for comparison
- thanks to differing definitions of VM_MIN_ADDRESS across our archs, we
  need to trick the compiler to not complain about signedness. We could
  either fix VM_MIN_ADDRESS to always be a simple integer or make the
  check conditional on $ARCH.

Closes PRs: kern/42386, kern/83364
Reviewed by: bde

13 years agoStyle(9) fix.
dchagin [Sun, 23 Jan 2011 09:50:39 +0000 (09:50 +0000)]
Style(9) fix.

Approved by: kib(mentor)
MFC after: 1 month

13 years agoRemove unnecessary checking of variable.
tuexen [Sun, 23 Jan 2011 07:27:35 +0000 (07:27 +0000)]
Remove unnecessary checking of variable.

MFC after: 3 months.

13 years agoAdd a new tool which takes a text hexdump of the current EEPROM contents.
adrian [Sat, 22 Jan 2011 23:44:56 +0000 (23:44 +0000)]
Add a new tool which takes a text hexdump of the current EEPROM contents.

13 years agoPatch the athstats code/Makefile to cross-compile correctly.
adrian [Sat, 22 Jan 2011 23:37:42 +0000 (23:37 +0000)]
Patch the athstats code/Makefile to cross-compile correctly.

13 years agoAdd missing logs.
pjd [Sat, 22 Jan 2011 23:30:01 +0000 (23:30 +0000)]
Add missing logs.

MFC after: 1 week