Sascha Wildner [Sat, 30 Aug 2014 09:50:38 +0000 (11:50 +0200)]
Sync ACPICA with Intel's version
20140828.
* Add support for runtime validation of the _DSD package.
* Add the GICC affinity subtable to the SRAT table.
* Add error detection for illegally named references within control
methods that would cause runtime failures.
* Add error checking for dependencies related to the _PSx power
methods.
* Various improvements, bug fixes and cleanups.
For a more detailed list, please see sys/contrib/dev/acpica/changes.txt.
Sascha Wildner [Fri, 29 Aug 2014 08:07:36 +0000 (10:07 +0200)]
kernel/devfs: Allow devices to devctl_notify() also during cold boot.
This fixes the issue that no moused(8) was started if a USB mouse was
plugged in already during booting.
While here, also rename the notify function.
In-collaboration-with: mneumann
Sascha Wildner [Fri, 29 Aug 2014 08:03:01 +0000 (10:03 +0200)]
rc.conf: Provide a default for rc_startmsgs and document it.
Sascha Wildner [Fri, 29 Aug 2014 07:16:50 +0000 (09:16 +0200)]
nrelease: Do not provide default make.conf files.
The git package delivered on the installation media is special in the
sense that it is built with certain options not set to prevent it from
pulling in python and perl as dependencies (to keep the ISOs small,
mainly).
Once the system is installed, storage size is no longer really an issue,
so there is no need to have those options in the make.conf we install.
After this commit, a manually built git package should match the binary
package which we provide.
Michael Neumann [Tue, 26 Aug 2014 23:24:47 +0000 (01:24 +0200)]
udev - use objcache for struct udev_event_kernel
François Tigeot [Tue, 26 Aug 2014 16:21:17 +0000 (18:21 +0200)]
drm/i915: Implement GEM GET and SET_CACHING ioctls
They can be used by userspace drivers to accelerate uploads and downloads
from the GPU and also mix CPU and GPU rendering/activity efficiently.
Obtained-from: Linux 3.8.13
Sascha Wildner [Tue, 26 Aug 2014 13:43:05 +0000 (15:43 +0200)]
rc.conf.5: Document moused_nondefault_enable (taken from FreeBSD).
Michael Neumann [Mon, 25 Aug 2014 16:59:02 +0000 (18:59 +0200)]
moused - Ease usb mouse attach/detach
WARNING: Upgrading will overwrite your /etc/devd.conf.
Extend /etc/rc.d/moused. To attach e.g. ums0, you can now call:
/etc/rc.d/moused start ums0
To detach:
/etc/rc.d/moused stop ums0
If moused is called with an explicit device name, it is enabled by default
(moused_nondefault_enable="YES").
You can explicitly override any settings in rc.conf with directives like:
moused_ums0_enable="YES"
moused_ums0_type="..."
If you have one or more ums[0-9]+ mouses, you can now make use of them by
simply setting devd_enable="YES" in rc.conf.
Matthew Dillon [Mon, 25 Aug 2014 16:51:29 +0000 (09:51 -0700)]
build - separate out the rescue initrd
* Multiple issues with users crashing at the end of an installworld because
initrd kldloads an out-of-date vn device.
* Not a good idea to update the rescue image during an installworld anyway.
* Make it a manual target called 'rescue' and remind the user at the end
of the installworld that the target is available, but suggest
rebooting first.
Matthew Dillon [Mon, 25 Aug 2014 16:40:39 +0000 (09:40 -0700)]
kernel config - Build 'vn' into the kernel
* Build vn into the kernel, we are using it in the build system and
its not a good idea to kldload it during a buildworld or installworld.
Sepherosa Ziehau [Mon, 25 Aug 2014 12:55:53 +0000 (20:55 +0800)]
test: Add bunch of tests for udp protocol
They test various combination of bind(2), connect(2) and sendto(2)
for both unicast and multicast. Unicast tests work w/ udpecho in
tools/netrate. Multicast tests work w/ mcastsend.
Sascha Wildner [Sun, 24 Aug 2014 21:54:16 +0000 (23:54 +0200)]
kernel/spinlock: Add a description to struct spinlock.
And add it to spin_init() and SPINLOCK_INITIALIZER().
Submitted-by: dclink (see <http://bugs.dragonflybsd.org/issues/2714>)
OK'd-by: dillon
Sascha Wildner [Sun, 24 Aug 2014 21:51:43 +0000 (23:51 +0200)]
unifdef.1: Use Mt.
Michael Neumann [Sat, 23 Aug 2014 23:23:40 +0000 (01:23 +0200)]
Matthew Dillon [Sat, 23 Aug 2014 17:57:13 +0000 (10:57 -0700)]
ucom - Hack unref to fix infinite 1-second wait loop.
* ucom depended on the .tsw_* switch to decomission ports, which we do
not use currently. This was preventing the ucom from being properly
dereferenced and resulted in an infinite 1-second wait loop on close.
* Move the unref elsewhere to hopefully solve the problem.
Matthew Dillon [Sat, 23 Aug 2014 17:53:25 +0000 (10:53 -0700)]
kernel - Fix sendfile() bugs / nginx
* Fix a sendfile() bug tickled by nginx. When sending partial file bodies the
nbytes test was improperly included the header bytes in its file body limit
test. This caused sendfile() to return 0 (complete success) but not
actually output the entirety of the requested file segment.
* sendfile() now holds VM pages instead of wiring them. Wiring the pages
prevents truncation operations from succeeded and can further cause a
kernel panic if the underlying file is deleted and the descriptor is
closed while mbufs related to the sendfile() are still pending on the
socket.
Massive-debugging-by: joris
Michael Neumann [Sat, 23 Aug 2014 11:03:58 +0000 (13:03 +0200)]
[iwn] Fix required calibration flags for the Centrino 1000 NIC.
Tested with Intel Centrino Wireless-N 1000.
Obtained-From: FreeBSD (commit
dd4e54a1ddf657f080a88c674e12459f404dea61)
Fixes: http://bugs.dragonflybsd.org/issues/2720
Sascha Wildner [Sat, 23 Aug 2014 10:37:59 +0000 (12:37 +0200)]
kernel/pci: Use spin_*() functions directly.
Matthew Dillon [Fri, 22 Aug 2014 21:59:46 +0000 (14:59 -0700)]
u4b - Fix panic on certain cell phone connections
* Remove incorrect unlock in ucom_close().
* Add missing usb_callout_stop() to umass_cam_detach_sim(), fixes issue with umass
if the mass-storage device quickly connects and disconnects from the bus.
Reported-by: Rolinh
Matthew Dillon [Fri, 22 Aug 2014 18:41:54 +0000 (11:41 -0700)]
sendmail - pre-fork in daemon mode to avoid stall (2)
* disconnect() and set CurrentPid on pre-fork. This fixes an issue where
the boot console is revoked and enters getty before the backgrounded
sendmail disconnects from the foreground process and tty, causing sendmail
to complain in /var/log/maillog.
This issue can occur due to myhostname() stalls when the internet is not
available (e.g. /etc/resolv.conf is present but wifi-only system has not
yet associated), or has other connectivity issues during boot.
Reported-by: Rolinh
Sascha Wildner [Thu, 21 Aug 2014 20:16:40 +0000 (22:16 +0200)]
kernel/kern_dup(): Fix comment.
Sascha Wildner [Thu, 21 Aug 2014 20:12:01 +0000 (22:12 +0200)]
kernel: Fix error for dup2 if the new file descriptor is out of bounds.
According to POSIX, when using fcntl()'s F_DUPFD or F_DUPFD_CLOEXEC
commands and the new file descriptor is out of bounds, EINVAL must be
returned. But if it is out of bounds when using dup2(), EBADF must be
returned.
So add detection of whether kern_dup() was called from kern_fcntl() or
from sys_dup2() and act accordingly.
Justin C. Sherrill [Thu, 21 Aug 2014 06:02:05 +0000 (02:02 -0400)]
Move from dns/bind98 on install img/iso to dns/bind-tools.
BIND 9.8 is reaching EOL in a month. We don't need a full version of
BIND on the default install, since bind-tools provides dig/nslookup/host
and so on.
Markus Pfeiffer [Tue, 19 Aug 2014 21:28:53 +0000 (21:28 +0000)]
sys/subr_rman: fix some issues
Some bounds checks were off which lead to bad interactions with the
ram reservation code in nexus.
Markus Pfeiffer [Sun, 17 Aug 2014 15:48:16 +0000 (16:48 +0100)]
sys/platform/pc64: port the ram device from FreeBSD's nexus.c
Our kernel didn't reserve RAM in the IO manager, which lead to
mmaped IO from devices being mapped in the same address space,
most notably this lead to problems with the i915 driver, bug
Sascha Wildner [Wed, 20 Aug 2014 18:49:11 +0000 (20:49 +0200)]
kernel/ath: Add missing #ifdef ATH_DEBUG (fixes kernel build).
Reported-by: Robin Hahling <robin.hahling@gw-computing.net>
Sascha Wildner [Wed, 20 Aug 2014 18:45:30 +0000 (20:45 +0200)]
csh(1): Use iconv().
Dragonfly-bug: <http://bugs.dragonflybsd.org/issues/2704>
Reported-by: s4
Matthew Dillon [Wed, 20 Aug 2014 17:42:24 +0000 (10:42 -0700)]
ath - move kprintf() under debug variable
* Move an unconditional debugging kprintf() under a debug variable.
Matthew Dillon [Wed, 20 Aug 2014 17:34:14 +0000 (10:34 -0700)]
dhclient - Do not force-up wlan interfaces
* Do not force a wlan interface up, this may interfere with wpa_supplicant
operation.
* Increase the default timeout for waiting for the interface to become
active from 10 seconds to 30 seconds. wlan interfaces can take
considerably longer than 10 seconds to associate and dhclient operation
now detaches without waiting so the longer timeout will not interfere
with the boot sequence.
Matthew Dillon [Wed, 20 Aug 2014 17:04:09 +0000 (10:04 -0700)]
boot - don't wait for wpa_supplicant to associate
* Do not wait for wpa_supplicant to associate as doing so will prevent
the machine from finishing the boot sequence if there are no base stations
in range.
We can do this now because the new dhclient will not blow up.
Matthew Dillon [Wed, 20 Aug 2014 16:56:03 +0000 (09:56 -0700)]
dhclient - Adjust documentation
* Document how the pidfile works.
* Document that executing a dhclient on an interface which already has a
running dhclient will kill the old dhclient and replace it with a new
instance.
Matthew Dillon [Wed, 20 Aug 2014 16:54:46 +0000 (09:54 -0700)]
sendmail - pre-fork in daemon mode to avoid stall
* Fork sendmail prior to doing the hostname lookup when run in -bd or -q<time>
mode.
* Prevents stalls during system boot when the internet is not available and
a resolv.conf is present.
Sascha Wildner [Wed, 20 Aug 2014 16:10:14 +0000 (18:10 +0200)]
bsd-family-tree: Sync with FreeBSD (for various minor DragonFly releases).
Patch-by: Zach Crownover
Matthew Dillon [Wed, 20 Aug 2014 01:57:05 +0000 (18:57 -0700)]
dhclient - Add -x and -w and adjust rc scripts
* Add the -w option which prevents dhclient from exiting on interface
down. dhclient will sleep 10 seconds and retry, forever.
* Add the -x option which kills any running dhclient on the specified
interface.
* Adjust rc.d/dhclient to use the -x option to kill the dhclient instead
of taking down the interface (which no longer kills it), and also adjust
the default flags to pass -w.
Sascha Wildner [Tue, 19 Aug 2014 20:11:05 +0000 (22:11 +0200)]
pam/pam_unix: Sync function declaration and definition.
No functional change.
Taken-from: FreeBSD
Sascha Wildner [Tue, 19 Aug 2014 20:10:10 +0000 (22:10 +0200)]
libstand: Add a missing free() in an error path.
Sascha Wildner [Tue, 19 Aug 2014 20:09:23 +0000 (22:09 +0200)]
libc/locale: Remove a wrong comma in the monetary testcase.
Sascha Wildner [Tue, 19 Aug 2014 20:08:47 +0000 (22:08 +0200)]
libc/valloc(): Use posix_memalign().
Taken-from: FreeBSD
Matthew Dillon [Mon, 18 Aug 2014 21:04:44 +0000 (14:04 -0700)]
kernel - Add EVFILT_USER
* Bring in EVFILT_USER support from FreeBSD. Document the hacked use of
EV_CLEAR.
* Implemented a bit differently. Currently not implemented as another
filter callback. Instead we explicitly test for user_filtops, which
isn't as messy (at least until some other kqueue feature needs the 'touch'
interface).
Taken-From: FreeBSD
Matthew Dillon [Mon, 18 Aug 2014 07:55:09 +0000 (00:55 -0700)]
cyapa - attempt #2 to fix boot issue
* Previous fix didn't work. It appears that the trackpad can sometimes
loose its mind even after initializing ok, so instead of doing a load-time
fix we now check the status byte in the poll.
* If the poll status indicates that cyapa is not running the device will be
bootstrapped again.
* Also, when not running do not try to interpret remaining status fields
for input.
François Tigeot [Mon, 18 Aug 2014 06:34:57 +0000 (08:34 +0200)]
drm/i915: Sync ioctl definitions with Linux 3.8.13
Many ioctls are now unlocked.
Matthew Dillon [Mon, 18 Aug 2014 06:37:20 +0000 (23:37 -0700)]
cyapa - Implement power save mode, try to fix boot issue (Acer c720 chromebook)
* Implement power save mode when the pad hasn't been touched for 10 seconds.
Return to full power on any activity.
* Try to fix an issue when cyapa is loaded from /boot/loader.conf or built
into the kernel. For some reason cyapa polls that it isn't running but
we'd already gone through the bootstrap sequence and determined that it
was running. Why it stops is not known but it appears doing a status query
after init completion puts it in the right state.
Matthew Dillon [Mon, 18 Aug 2014 06:36:46 +0000 (23:36 -0700)]
dmesg - flush stdout before sleeping
* Flush output before sleeping when dmesg -f is specified.
Sascha Wildner [Sun, 17 Aug 2014 18:29:21 +0000 (20:29 +0200)]
drm/ttm: Add missing kfree().
Sascha Wildner [Sun, 17 Aug 2014 18:17:23 +0000 (20:17 +0200)]
drm/i915: Fix a missing initialization.
Sascha Wildner [Sun, 17 Aug 2014 10:33:44 +0000 (12:33 +0200)]
<sys/taskqueue>: Remove FreeBSD specific, *_fast() related defines.
They were brought in with
d96c147595b4f9046cb3d1682071adf4f98a2204.
François Tigeot [Sun, 17 Aug 2014 09:08:58 +0000 (11:08 +0200)]
drm/i915: Enable fbc on Mobile Ironlake
Sascha Wildner [Sat, 16 Aug 2014 21:09:39 +0000 (23:09 +0200)]
unifdef(1): Update to upstream version 2.10.15.a52596a
Taken-from: http://dotat.at/prog/unifdef/
Matthew Dillon [Sat, 16 Aug 2014 20:54:17 +0000 (13:54 -0700)]
ath - Change the way the edma rxfifo is reset (2)
* Correct a debugging change that wasn't supposed to go in.
Matthew Dillon [Sat, 16 Aug 2014 20:36:21 +0000 (13:36 -0700)]
ath - Change the way the edma rxfifo is reset
* Do not unconditionally try to reload the whole FIFO. Instead set a flag
when the HAL is reset.
* Force the hal to cold-reset every time for now to match the flag.
* Currently a hack until a Adrian comes up with the correct fix.
* TODO - Currently does not re-enable the RXEOL interrupt correctly.
Matthew Dillon [Sat, 16 Aug 2014 20:32:22 +0000 (13:32 -0700)]
kernel - Adjust BUS_DMASYNC_* API
* Change the BUS_DMASYNC_* enumeration to a bitmask to allow multiple
flags to be specified in one call.
* Now more compatible with FreeBSD and one less thing to worry about when
porting a driver.
* While we are at it, allow add_bounce_page() to modify the size. This has
no effect on the current codebase since bounce pages are page-aligned (so the
page-residual will never be less than the passed address), but is more
correct.
* Fixes bounce-buffer issue with if_ath.
Sascha Wildner [Sat, 16 Aug 2014 12:56:00 +0000 (14:56 +0200)]
libc: Remove unneeded semconfig function & file.
François Tigeot [Sat, 16 Aug 2014 09:52:04 +0000 (11:52 +0200)]
drm/i915: Sync i915_drv.h and i915_debugfs.c with Linux 3.8.13
Matthew Dillon [Sat, 16 Aug 2014 06:03:53 +0000 (23:03 -0700)]
ath - incremental update ath from FreeBSD
* Update to fbsd
61ad4c7509a280c218a389f91026136cec60708e
(Aug 11 2014).
Matthew Dillon [Sat, 16 Aug 2014 05:55:43 +0000 (22:55 -0700)]
ath - incremental update ar9300 from FreeBSD
* Update to fbsd
61ad4c7509a280c218a389f91026136cec60708e
(Aug 11 2014).
* Minor adjustments to the ar9300 hal, plus sysctl support for
forced cold reset.
Matthew Dillon [Fri, 15 Aug 2014 17:05:21 +0000 (10:05 -0700)]
kernel - Reduce ig4 receive errors
* The timeout code now checks the FIFO and register status at least once
before testing whether the timeout has been exceeded, and at least once
more after waiting the timeout period.
Previously the code would check, see that no data was present, then delay
a little and check whether a timeout occurred after the delay without
again testing whether data is present.
* Appears to fix receive timeouts reported by cyapa and atmel_mxt.
Sascha Wildner [Fri, 15 Aug 2014 13:56:35 +0000 (15:56 +0200)]
Update the pciconf(8) database.
August 10, 2014 snapshot from http://pciids.sourceforge.net/
Matthew Dillon [Fri, 15 Aug 2014 07:14:38 +0000 (00:14 -0700)]
atmel - Add automatic recalibration
* Add automatic recalibration every 30 seconds when idle. I'm really not
sure if this is necessary or not but for now we do it.
* Interval can be adjusted with sysctl debug.atmel_mxt_autocalibrate,
set to 0 to disable.
Matthew Dillon [Fri, 15 Aug 2014 07:10:53 +0000 (00:10 -0700)]
ig4 - Fix rxdata shift error and add debugging
* Fix a bug where the rx fifo can wind up with spurious data that never
clears because later commands are always one-ahead of it.
* Report rx data failures and spurious rx fifo data, except when the
NOREPORT flag has been specified.
Matthew Dillon [Fri, 15 Aug 2014 07:09:32 +0000 (00:09 -0700)]
smbus - Add SMB_TRANS_NOREPORT
* Add a flag to tell subsystems not to report errors (during probing).
François Tigeot [Fri, 15 Aug 2014 06:44:44 +0000 (08:44 +0200)]
drm/i915: Only build on amd64/x86_64 machines
This driver now requires a direct memory map.
Matthew Dillon [Fri, 15 Aug 2014 01:52:07 +0000 (18:52 -0700)]
atmel_mxt - Add atmel MXT touchscreen driver (Acer c720p)
* Implements the communications infrastructure on the hardware backend and
emulates an elographics tablet on the frontend.
* X Configuration is:
Section "InputDevice"
Identifier "TouchScreen0"
Driver "elographics"
Option "Device" "/dev/atmel1-4a"
# Option "ZAxisMapping" "4 5 6 7"
EndSection
* System configuration is (in /boot/loader.conf):
ig4_load="YES"
atmel_mxt_load="YES"
* Currently only tested and only likely to work with the Acer c720p.
Can do basic touchscreen operations... hold-move-and-release for
button-1. Can move or resize windows (you may want to make your resize
widgets bigger). Can hit buttons and hotlinks in browsers. Cannot
scroll a browser window on its own.
* Noise issues. I've noticed that both my chromebooks have a lot of mouse
pointer jitter with both the trackpad and the touchscreen, more than
I can actually fix. I've noticed some complaints on the linux forums too
so it aint just us.
The jitter is reduced if you put the laptop on your lap, and appears to
go away entirely if the laptop is on your lap AND the power is disconnected
(i.e. running on battery).
Markus Pfeiffer [Thu, 14 Aug 2014 22:03:07 +0000 (22:03 +0000)]
drm/i915: grab the config log on hardware init
Markus Pfeiffer [Thu, 14 Aug 2014 21:46:24 +0000 (21:46 +0000)]
drm/i915: Use intel_gmbus_get_adapter instead of dev_priv->gmbus array
Markus Pfeiffer [Sat, 9 Aug 2014 18:20:18 +0000 (19:20 +0100)]
drm/i915: Fix multiple off-by-one errors in intel_i2c
* the i2c pin numbers are 1 based as per intel spec but
we used the unit number as pin number, which is 0 based
* the gmbus array is 0 based, but the port number is 1 based
accesses into the port array are therefore off-by-one and this
only worked by chance before.
François Tigeot [Thu, 14 Aug 2014 18:13:51 +0000 (20:13 +0200)]
drm/i915: Sync i915_gem_execbuffer.c with Linux 3.8.13
* Simpler code and more consistent performance
* Implement a workaround for a Sandy-Bridge hardware issue
Matthew Dillon [Thu, 14 Aug 2014 17:36:03 +0000 (10:36 -0700)]
smb - Increase max block size from 32 to 1024
* The SMB API can be used for I2C transactions. Do not impose the
SMB2.0 block size limit on transfers.
Sascha Wildner [Thu, 14 Aug 2014 07:46:37 +0000 (09:46 +0200)]
bsd-family-tree: Sync with FreeBSD.
François Tigeot [Thu, 14 Aug 2014 07:09:59 +0000 (09:09 +0200)]
drm/i915: Fix display corruption issues on Haswell
* The mid-level cache is setup differently on Haswell
* Haswell's PTE encoding has also changed; we have to know what
GPU generation we're running on to write the correct attributes
to the Graphics Translation Tables
Sepherosa Ziehau [Wed, 13 Aug 2014 12:58:32 +0000 (20:58 +0800)]
tools/kq_connect_client: Add option to create UDP socket before connect
And do connect and send 18B datagram before closing the UDP socket. This
simulates DNS resolving, though this UDP socket does not receive a reply.
Sepherosa Ziehau [Wed, 13 Aug 2014 12:57:59 +0000 (20:57 +0800)]
tools/udpecho: Add option to disable echo; just sink
Matthew Dillon [Tue, 12 Aug 2014 23:57:16 +0000 (16:57 -0700)]
kernel - PF SMP work - fix RDR rules
* Use the global state table for all global-flagged state. RDR
rules can apparently cause PF_SK_WIRE/PF_SK_STACK to be
reversed so our global state table tests which only used
PF_SK_WIRE for GLOBAL-flagged translations were not sufficient.
* Fixes issues with the reverse packet direction for a rdr
rule not finding the reverse state.
* Note that RDR rules will not be SMP-optimal.
* Adds a bit more overhead for NAT translations which are not
SMP optimal (however, most NAT translations select a port
that will be SMP optimal).
Reported-by: alexh
François Tigeot [Tue, 12 Aug 2014 16:44:53 +0000 (18:44 +0200)]
drm: Add Linux kmap functions
These implementations are only suitable for use with systems with
direct memory maps, which means amd64/x86-64 for now.
François Tigeot [Tue, 12 Aug 2014 09:20:08 +0000 (11:20 +0200)]
drm/i915: i915_emit_box() from Linux 3.8.13
François Tigeot [Mon, 11 Aug 2014 19:40:21 +0000 (21:40 +0200)]
agp: Add more Haswell PCI IDs
François Tigeot [Mon, 11 Aug 2014 09:58:41 +0000 (11:58 +0200)]
drm/i915: Enable Haswell support
* The drm/i915 driver includes its own set of PCI ids, making those in
drm_pciids.h redundant. Remove i915/i930 ids from this file.
* Unfortunately our old drm code base expects to directly use device
lists with a different format.
* Rewrite the i915_probe() routine to use the driver-included device
list and patch a special-purpose, one-element list in the legacy
format at runtime to make the old drm_attach() routine and associated
code happy.
* At this point, Haswell GPUs still exhibit display corruption issues
with 2D acceleration. Disabling it in xorg.conf is recommended.
François Tigeot [Mon, 11 Aug 2014 08:51:35 +0000 (10:51 +0200)]
drm/i915: Sync i915_gem_evict.c with Linux 3.8.13
Matthew Dillon [Sun, 10 Aug 2014 20:03:51 +0000 (13:03 -0700)]
kernel - Fix PF SMP issues
* Fix an address table scan race that can cause packets to not match.
* The same fix also fixes a NULL pointer dereference caused by the
same scan race.
Reported-by: alexh (and testing too)
François Tigeot [Sun, 10 Aug 2014 08:37:01 +0000 (10:37 +0200)]
agp/drm: Remove intel_gtt_get_bridge_device()
François Tigeot [Sun, 10 Aug 2014 08:04:58 +0000 (10:04 +0200)]
agp/drm: Rename agp_i810.h to intel-gtt.h
So that we use the same file name as Linux.
Sepherosa Ziehau [Sun, 10 Aug 2014 12:26:02 +0000 (20:26 +0800)]
tools/netrate: Use FIONBIO to make the socket nonblocking
Sepherosa Ziehau [Sun, 10 Aug 2014 12:22:12 +0000 (20:22 +0800)]
tcp: Fix upper half (socket code) and so_port switching race.
Use message put done receipt to change this socket's so_port, i.e. _after_
this message was put onto the target netisr's msgport but _before_ the
message could be pulled from the target netisr's msgport, so that:
- The upper half (socket code) will not see the new msgport before this
message reaches the new msgport and messages for this socket will be
ordered.
- This message will see the new msgport, when its handler is called in the
target netisr.
Add comment about it and explain why the so_port switching should be done
in this commit's fashion.
Reported-by: dillon@
Sepherosa Ziehau [Sun, 10 Aug 2014 11:42:58 +0000 (19:42 +0800)]
msgport: Add message put done receipt support
The receipt callback is called immediately after the message is put on to
the target msgport. Complex operation is NOT allowed in this callback,
since it could be called w/ spinlock for spin msgports.
It will be used to fix the race between the upper half (socket code)
and the so_port switching.
Antonio Huete Jimenez [Sat, 9 Aug 2014 16:24:54 +0000 (18:24 +0200)]
dhclient - Change config 'ignore' directive behaviour.
- Change dhclient.conf directive 'ignore' to take a list of option names
rather than list of option declarations. e.g. 'ignore routers;'
instead of 'ignore routers 1.2.3.4;' The value in the declaration
was being ignored anyway.
- While there clean up the related code a bit.
Taken-from: OpenBSD
Antonio Huete Jimenez [Thu, 7 Aug 2014 19:11:13 +0000 (21:11 +0200)]
dhclient - Do not update linkstat several times.
- In the case where two RTM_IFINFO link changes occur immediately
after each other (as in the case with trunk and probably vlan)
dhclient missed the state change of the second interface.
Taken-from: OpenBSD
Antonio Huete Jimenez [Thu, 7 Aug 2014 19:06:42 +0000 (21:06 +0200)]
dhclient - Revert
a4a242d4bd3b87eb38ccf205bb2784ada509ca3a
- Routing breaks if dhclient-script flushes the
interface's routes when the interface is down.
Taken-from: OpenBSD
Antonio Huete Jimenez [Thu, 7 Aug 2014 08:05:20 +0000 (10:05 +0200)]
dhclient - Update OBSD Ids
- RFC changes to the manual pages already there.
Taken-from: OpenBSD
Antonio Huete Jimenez [Thu, 7 Aug 2014 07:53:39 +0000 (09:53 +0200)]
dhclient - Remove saved resolv.conf to avoid problems.
- Once resolv.conf.saved has been copied back as
resolv.conf, delete it so it isn't copied again the
next time a dhclient goes away.
Taken-from: OpenBSD
Antonio Huete Jimenez [Thu, 7 Aug 2014 07:47:01 +0000 (09:47 +0200)]
dhclient - Refuse leases for already configured subnets.
Taken-from: OpenBSD
Antonio Huete Jimenez [Sat, 2 Aug 2014 12:04:09 +0000 (14:04 +0200)]
dhclient - When a link is lost, call dhclient-script with reason "FAIL".
- This does the resolv.conf dance and removes 'dead' routes, rather than
leaving these droppings behind.
- When dhclient is exiting after calling dhclient-script with "FAIL",
give the script a couple of seconds to finish. Also pass the lease
parameters (old_*) to dhclient-script so routes can be cleaned up
in this path too.
Taken-from: OpenBSD
Antonio Huete Jimenez [Sat, 2 Aug 2014 12:02:22 +0000 (14:02 +0200)]
dhclient - Update reality correspondance of some comments.
Taken-from: OpenBSD
Antonio Huete Jimenez [Sat, 2 Aug 2014 12:01:21 +0000 (14:01 +0200)]
dhclient - Missed a reinitialize_interface() declaration.
Taken from: OpenBSD
Antonio Huete Jimenez [Sat, 2 Aug 2014 11:59:14 +0000 (13:59 +0200)]
dhclient - Nuke 'interfaces_invalidated' global variable.
- Since reinitialize_interface() was simply setting interfaces_invalidated
to 0, nuke it too!
- Always update ifi->linkstat in dispatch() loop.
Taken-from: OpenBSD
Antonio Huete Jimenez [Sat, 2 Aug 2014 11:55:32 +0000 (13:55 +0200)]
dhclient - Nuke some comments.
- Nuke a bunch of annoying "/* Send a packet. */" comments above
invocations to send_packet(), and similiar combinations.
Taken-from: OpenBSD
Antonio Huete Jimenez [Fri, 1 Aug 2014 18:24:13 +0000 (20:24 +0200)]
dhclient - No point having two copies of this file
- This copy never got installed anyway.
Taken-from: OpenBSD
Antonio Huete Jimenez [Fri, 1 Aug 2014 18:19:04 +0000 (20:19 +0200)]
dhclient - Remove cur_time global.
- Junk global cur_time, and use time(NULL) or local variables where
time is checked multiple times. Add a set_timeout_interval() function
to allow setting a timeout based on an interval from current time.
- Fixes issues with initial startup where the global cur_time was
always old and caused initial DHCPDISCOVER or DHCPREQUEST packets
to be sent multiple times. And probably other timeout related
oddities.
Taken-from: OpenBSD
Antonio Huete Jimenez [Fri, 1 Aug 2014 18:09:13 +0000 (20:09 +0200)]
dhclient - Add ignore keyword to conf file.
- Allows to ignore unwanted info from the server without
necessarily speciyfing a supersede value.
Taken-from: OpenBSD
Antonio Huete Jimenez [Fri, 1 Aug 2014 17:43:37 +0000 (19:43 +0200)]
dhclient - Don't abandon time_t precision for intervals.
- Use (long long) and %lld when printing, rather than (int)and %d.
Follows idiom used for 64-bit size_t and off_t.
Taken-from: OpenBSD
Antonio Huete Jimenez [Fri, 1 Aug 2014 17:41:32 +0000 (19:41 +0200)]
dhclient - Don't store a time_t into an int.
- Make the variable time_t.
Taken-from: OpenBSD
Antonio Huete Jimenez [Fri, 1 Aug 2014 17:36:21 +0000 (19:36 +0200)]
dhclient - Don't print time_t as %d
- Cast to (int) just in case time_t ever changes size.
Values in these cases are intervals, i.e. (time_t - time_t)
so int/%d will be fine.
Taken-from: OpenBSD