dragonfly.git
9 years agokernel/ath: Add missing braces (FreeBSD's r277290).
Sascha Wildner [Sat, 17 Jan 2015 08:05:56 +0000 (09:05 +0100)]
kernel/ath: Add missing braces (FreeBSD's r277290).

9 years agokernel - Augment resource_string_value() to check kenv.
Matthew Dillon [Sat, 17 Jan 2015 01:58:21 +0000 (17:58 -0800)]
kernel - Augment resource_string_value() to check kenv.

* resource_string_value() now checks for a kenv hint called
  <devname><unit>.<resourcename>, similar to how resource_int_value()
  and resource_long_value() work.

9 years agopowerd - Add an EXAMPLES section to the manual
Matthew Dillon [Sat, 17 Jan 2015 00:08:44 +0000 (16:08 -0800)]
powerd - Add an EXAMPLES section to the manual

* Add an EXAMPLES section with a suggestion for workstation settings.

9 years agopowerd - Add options to set poll_rate and sample_time, adjust idle algorithm
Matthew Dillon [Fri, 16 Jan 2015 23:44:41 +0000 (15:44 -0800)]
powerd - Add options to set poll_rate and sample_time, adjust idle algorithm

* Add -r poll_rate and -T sample_time.  The user can adjust the values
  to e.g. make a workstation react more quickly to UI operations coming
  out of idle and to take longer to return to idle.

* Use a shorter averaging time (3 poll intervals) for ratcheting up the
  speed and a longer averaging time (adjustable by -T) for ratcheting down
  the speed.

* Enhance the manual page, describe default values and make suggestions.

9 years agoypclnt.3: Add missing parameter type.
Sascha Wildner [Fri, 16 Jan 2015 19:21:12 +0000 (20:21 +0100)]
ypclnt.3: Add missing parameter type.

9 years agolibusb20.3: Add missing function type.
Sascha Wildner [Fri, 16 Jan 2015 17:05:58 +0000 (18:05 +0100)]
libusb20.3: Add missing function type.

9 years agoLINT64: Document the kernel options of the sound stack.
Sascha Wildner [Fri, 16 Jan 2015 08:55:48 +0000 (09:55 +0100)]
LINT64: Document the kernel options of the sound stack.

Also, comment out SND_DIAGNOSTIC which is currently broken.

9 years agokernel: Define HAVE_KERNEL_OPTION_HEADERS when building kernel/modules.
Sascha Wildner [Fri, 16 Jan 2015 08:54:35 +0000 (09:54 +0100)]
kernel: Define HAVE_KERNEL_OPTION_HEADERS when building kernel/modules.

It gets defined when building the kernel and when building modules
along with the kernel.

FreeBSD wraps some of their #include "opt_blah.h" in it and it should
ease porting $stuff from them.

It also makes the kernel options of the sound stack actually work.

9 years agouio.9: uiomove() takes size_t, not int.
Sascha Wildner [Fri, 16 Jan 2015 08:43:14 +0000 (09:43 +0100)]
uio.9: uiomove() takes size_t, not int.

9 years agopfil.9: pfil_{add,remove}_hook() return int.
Sascha Wildner [Fri, 16 Jan 2015 08:20:45 +0000 (09:20 +0100)]
pfil.9: pfil_{add,remove}_hook() return int.

9 years agoeditline.3: Ansify two prototypes.
Sascha Wildner [Fri, 16 Jan 2015 08:01:44 +0000 (09:01 +0100)]
editline.3: Ansify two prototypes.

9 years agoopenssl: Adjust manual pages for 1.0.1l.
Sascha Wildner [Thu, 15 Jan 2015 20:25:46 +0000 (21:25 +0100)]
openssl: Adjust manual pages for 1.0.1l.

9 years agoMerge branch 'vendor/OPENSSL'
Sascha Wildner [Thu, 15 Jan 2015 21:18:26 +0000 (22:18 +0100)]
Merge branch 'vendor/OPENSSL'

9 years agodrm - Fix a second X lockup w/radeon ttm
Matthew Dillon [Thu, 15 Jan 2015 20:59:51 +0000 (12:59 -0800)]
drm - Fix a second X lockup w/radeon ttm

* Code which releases bo->reserved and wakes up waiters was
  not interlocked against wait_event_common() used to wait
  for the release.  This can result in a race where the
  release occurs inbetween the wait's test and its sleep,
  preventing the wait from ever waking up.

* This is different from the recent dev_pager_mtx deadlock which
  was recently fixed, but in the same code path.

* Roll the release of bo->reserved into a static function
  and throw the bo->event_queue.lock around the actual
  release to fix the race.

9 years agoImport OpenSSL-1.0.1l.
Sascha Wildner [Thu, 15 Jan 2015 20:22:00 +0000 (21:22 +0100)]
Import OpenSSL-1.0.1l.

* Fixes for CVE-2014-3571, CVE-2015-0206, CVE-2014-3569, CVE-2014-3572,
  CVE-2015-0204, CVE-2015-0205, CVE-2014-8275 and CVE-2014-3570.

* Ensure that the session ID context of an SSL is updated when its
  SSL_CTX is updated via SSL_set_SSL_CTX.

* Do not resume sessions on the server if the negotiated protocol
  version does not match the session's version.

* Tighten handling of the ChangeCipherSpec (CCS) message.

* Tighten client-side session ticket handling during renegotiation.

Also, while here, remove the doc/ subdirectory on the vendor branch.
We don't need to distribute it in contrib/.

9 years agokernel/sound: Fix typos in user visible messages etc.
Sascha Wildner [Thu, 15 Jan 2015 15:57:51 +0000 (16:57 +0100)]
kernel/sound: Fix typos in user visible messages etc.

9 years agoumtx.2/quick_exit.3: Fix some prototypes.
Sascha Wildner [Thu, 15 Jan 2015 13:19:38 +0000 (14:19 +0100)]
umtx.2/quick_exit.3: Fix some prototypes.

9 years agotruss(1): Retry opening /proc/%d/mem when errno == EAGAIN.
Imre Vadasz [Thu, 15 Jan 2015 11:04:10 +0000 (12:04 +0100)]
truss(1): Retry opening /proc/%d/mem when errno == EAGAIN.

9 years agolibedit: Install and use the correct <histedit.h>.
Sascha Wildner [Thu, 15 Jan 2015 10:15:19 +0000 (11:15 +0100)]
libedit: Install and use the correct <histedit.h>.

Ever since libedit was moved to contrib/ back in May 2012 (c22639f2f9ab7)
we were still installing the previous <histedit.h> from /usr/src/include.

9 years agoUpdate the pciconf(8) database.
Sascha Wildner [Thu, 15 Jan 2015 08:56:34 +0000 (09:56 +0100)]
Update the pciconf(8) database.

January 5, 2015 snapshot from http://pciids.sourceforge.net/

9 years agocam: Disable verbose "Unretryable error" log for probe pseudo device.
Sepherosa Ziehau [Wed, 14 Jan 2015 12:50:30 +0000 (20:50 +0800)]
cam: Disable verbose "Unretryable error" log for probe pseudo device.

Discussed-with: swildner@

9 years agoacpi/pstate: Allow missing _PCT and _PSS, as long as _PSD exists
Sepherosa Ziehau [Wed, 14 Jan 2015 12:14:07 +0000 (20:14 +0800)]
acpi/pstate: Allow missing _PCT and _PSS, as long as _PSD exists

Some BIOSes does not export _PCT and _PSS for the second thread of CPU
cores.  As long as _PSD exists and control/status registers are "fixed
hardware", e.g. for most of the modern Intel CPUs, we could reconstruct
P-state for the CPU even if there are no _PCT or _PSS.

In case that _PSD is wrong, e.g. 2 threads of a same core are in different
P-state domain, we try to put them to the same P-state domain.  This could
be disabled by setting tunable hw.acpi.cpu.pst.ht_reuse_domain to 0; this
tunable is 1 by default.

9 years agokernel: rename MAP_ANON to MAP_ANONYMOUS, add alias
Markus Pfeiffer [Wed, 14 Jan 2015 15:00:02 +0000 (15:00 +0000)]
kernel: rename MAP_ANON to MAP_ANONYMOUS, add alias

* rename MAP_ANON to MAP_ANONYMOUS
* add MAP_ANON as an alias of MAP_ANONYMOUS for backward
  compatibility. This should ease porting of some software which
  just fails, because MAP_ANONYMOUS is not defined

9 years agoaltq: Dispatch tbr_timeout to netisr0 to run
Sepherosa Ziehau [Tue, 13 Jan 2015 13:33:55 +0000 (21:33 +0800)]
altq: Dispatch tbr_timeout to netisr0 to run

This function now access ifnet global variable in netisr.

9 years agoinet6: Defer in6_tmpaddrtimer
Sepherosa Ziehau [Tue, 13 Jan 2015 12:56:33 +0000 (20:56 +0800)]
inet6: Defer in6_tmpaddrtimer

Same as before fd89b323a2536d2a3a0c7cfff93719a2aec48323

9 years agoinet6: Dispatch nd6_slowtimo to netisr0 to run
Sepherosa Ziehau [Tue, 13 Jan 2015 12:54:42 +0000 (20:54 +0800)]
inet6: Dispatch nd6_slowtimo to netisr0 to run

This function now access ifnet global variable in netisr.

9 years agokernel: Hide the intrmask_t typedef from userspace with #ifdef _KERNEL.
Imre Vadasz [Tue, 13 Jan 2015 21:26:32 +0000 (22:26 +0100)]
kernel: Hide the intrmask_t typedef from userspace with #ifdef _KERNEL.

9 years agosshlockout - Handle IPv6 and don't block localhost
Michael Neumann [Tue, 13 Jan 2015 21:15:28 +0000 (22:15 +0100)]
sshlockout - Handle IPv6 and don't block localhost

9 years agosshlockout - fix typo
Michael Neumann [Tue, 13 Jan 2015 18:13:49 +0000 (19:13 +0100)]
sshlockout - fix typo

9 years agosshlockout - Handle invalid user login attempts
Michael Neumann [Tue, 13 Jan 2015 18:10:46 +0000 (19:10 +0100)]
sshlockout - Handle invalid user login attempts

9 years agosshlockout - Prevent DoS attacks
Michael Neumann [Tue, 13 Jan 2015 14:00:58 +0000 (15:00 +0100)]
sshlockout - Prevent DoS attacks

9 years agosshlockout - refactor
Michael Neumann [Tue, 13 Jan 2015 12:44:20 +0000 (13:44 +0100)]
sshlockout - refactor

9 years agosshlockout - use a PF table instead of IPFW
Michael Neumann [Tue, 13 Jan 2015 12:04:29 +0000 (13:04 +0100)]
sshlockout - use a PF table instead of IPFW

9 years agodebug - Add bufqueues program
Antonio Huete Jimenez [Tue, 13 Jan 2015 10:49:20 +0000 (11:49 +0100)]
debug - Add bufqueues program

- A program to list buf(9) stats on a live system, vkernel or a
  system dump.

root@andromeda:/root# bufqueues
NONE  LOCKED  CLEAN  DIRTY  DIRTY_HW  EMPTYKVA  EMPTY  OFF-QUEUE KVMFAIL
   0       2   3903      2        61        34     12          0      14
   0       2   3903      2        61        34     12          0      14
   0       2   3903      2        61        34     12          0      14
   0       2   3979      0         0        17     12          0      10

...

9 years agodebug - Build evfilt_user as well
Antonio Huete Jimenez [Mon, 12 Jan 2015 12:17:22 +0000 (13:17 +0100)]
debug - Build evfilt_user as well

9 years agoinet6: Dispatch in6_tmpaddrtimer to netisr0 to run
Sepherosa Ziehau [Mon, 12 Jan 2015 13:37:07 +0000 (21:37 +0800)]
inet6: Dispatch in6_tmpaddrtimer to netisr0 to run

This function now access ifnet global variable in netisr.

9 years agokernel - Flag cdevpriv use (2)
Matthew Dillon [Mon, 12 Jan 2015 19:57:25 +0000 (11:57 -0800)]
kernel - Flag cdevpriv use (2)

* Don't throw away fp->f_flag bits set prior to the call to
  vop_stdopen().

9 years agokernel - Flag cdevpriv use
Matthew Dillon [Mon, 12 Jan 2015 18:56:54 +0000 (10:56 -0800)]
kernel - Flag cdevpriv use

* Flag cdevpriv use in devfs, allowing device drivers to
  use the fp->f_data1 field without having to use the
  cdevpriv mechanism if they desire.

9 years agokernel - remove O_FBUFFERED and O_FUNBUFFERED
Matthew Dillon [Mon, 12 Jan 2015 18:49:03 +0000 (10:49 -0800)]
kernel - remove O_FBUFFERED and O_FUNBUFFERED

* Remove these fcntl / special flags.  They are no longer being
  used by anything.

9 years agodebug - Update zallocinfo
Antonio Huete Jimenez [Mon, 12 Jan 2015 11:46:33 +0000 (12:46 +0100)]
debug - Update zallocinfo

9 years agoRevert "drm: Handle drm masters and minors like Linux"
François Tigeot [Mon, 12 Jan 2015 07:50:18 +0000 (08:50 +0100)]
Revert "drm: Handle drm masters and minors like Linux"

This reverts commit 2aa08105f11f5815d087c0f6baea9af394bd323f.

Many people reported this commit breaked their Xorg setup.

9 years agodebug - Fix wildcardinfo
Antonio Huete Jimenez [Sun, 11 Jan 2015 23:18:46 +0000 (00:18 +0100)]
debug - Fix wildcardinfo

9 years agoAdd commented-out rule allowing ICMP.
Justin C. Sherrill [Sun, 11 Jan 2015 23:01:22 +0000 (18:01 -0500)]
Add commented-out rule allowing ICMP.

Leaving it disabled can cause silent weirdness.

9 years agodrm: Handle drm masters and minors like Linux
François Tigeot [Sat, 10 Jan 2015 20:06:46 +0000 (21:06 +0100)]
drm: Handle drm masters and minors like Linux

* Make drm data structures more like their Linux versions

* Reduce differences with Linux 3.9.11 and use the same
  idr-based mechanisms when possible

* Create devices with the right minor numbers

* Use the same struct file.private_data mechanism Linux uses to pass
  drm_file information, removing the need for the FreeBSD-originated
  devfs cdevpriv code.

9 years agokernel/dev_ops: Make default dev_ops (noopen(), etc.) static.
Sascha Wildner [Sun, 11 Jan 2015 11:05:17 +0000 (12:05 +0100)]
kernel/dev_ops: Make default dev_ops (noopen(), etc.) static.

They should only ever be needed by compile_dev_ops() (via default_dev_ops).

9 years agokernel/cam: Remove unneeded dev_ops assignments.
Sascha Wildner [Sun, 11 Jan 2015 10:43:24 +0000 (11:43 +0100)]
kernel/cam: Remove unneeded dev_ops assignments.

compile_dev_ops() takes care of assigning no*() to uninitialized ops.

9 years agoinet6: Dispatch nd6_setmtu0 to netisr0 to run
Sepherosa Ziehau [Fri, 9 Jan 2015 14:50:12 +0000 (22:50 +0800)]
inet6: Dispatch nd6_setmtu0 to netisr0 to run

This function now access ifnet global variable in netisr.

9 years agoinet6: Dispatch nigroup attach/detach to netisr0 to run
Sepherosa Ziehau [Fri, 9 Jan 2015 14:11:57 +0000 (22:11 +0800)]
inet6: Dispatch nigroup attach/detach to netisr0 to run

These two functions now access ifnet global variable in netisr.

9 years agonetmap: Remove unnecessary ifunit call
Sepherosa Ziehau [Fri, 9 Jan 2015 13:27:57 +0000 (21:27 +0800)]
netmap: Remove unnecessary ifunit call

9 years agoSync mixer rc script with FreeBSD and enable mixer state save/restore.
Sascha Wildner [Sun, 11 Jan 2015 09:23:06 +0000 (10:23 +0100)]
Sync mixer rc script with FreeBSD and enable mixer state save/restore.

9 years agomixer(8): Sync up with FreeBSD.
Sascha Wildner [Sun, 11 Jan 2015 09:23:03 +0000 (10:23 +0100)]
mixer(8): Sync up with FreeBSD.

9 years agokernel/usb4bsd: Sync usb_quirk(4) with FreeBSD.
Sascha Wildner [Sat, 10 Jan 2015 19:25:58 +0000 (20:25 +0100)]
kernel/usb4bsd: Sync usb_quirk(4) with FreeBSD.

9 years agodrm - Fix deadlock in ttm pager
Matthew Dillon [Sat, 10 Jan 2015 17:35:24 +0000 (09:35 -0800)]
drm - Fix deadlock in ttm pager

* Fix a deadlock which most often occurs via the ttm (radeon) VM pager.
  A similar path is also used by i915 (all intel).

* Basically removes an unnecessary lock in the paging path which was creating
  the deadlock.

Reported-by: ivadasz, ftigeot
9 years agodrm: drm_detach() is drm_release()
François Tigeot [Sat, 10 Jan 2015 08:31:00 +0000 (09:31 +0100)]
drm: drm_detach() is drm_release()

Rename it in order to reduce differences with Linux.

9 years agousb4bsd: Sync with FreeBSD r276791 - revert broken 64bit DMA
Markus Pfeiffer [Fri, 9 Jan 2015 20:40:30 +0000 (20:40 +0000)]
usb4bsd: Sync with FreeBSD r276791 - revert broken 64bit DMA

* Revert a change that lead to an uninitialised value being used
  for the number of bits in a DMA address.

9 years agoStop removing the snd_emu10kx(4) manual page via 'make upgrade'.
Sascha Wildner [Fri, 9 Jan 2015 17:14:53 +0000 (18:14 +0100)]
Stop removing the snd_emu10kx(4) manual page via 'make upgrade'.

9 years agokernel/hda: Add missing file to 'files' to be able to compile it in.
Sascha Wildner [Fri, 9 Jan 2015 13:49:10 +0000 (14:49 +0100)]
kernel/hda: Add missing file to 'files' to be able to compile it in.

9 years agousb4bsd: Sync with FreeBSD r276791
Markus Pfeiffer [Thu, 8 Jan 2015 23:51:31 +0000 (23:51 +0000)]
usb4bsd: Sync with FreeBSD r276791

* Make sure the 64bit DMA bits are all there

9 years agousb4bsd: Sync with FreeBSD r276791
Markus Pfeiffer [Thu, 8 Jan 2015 23:47:46 +0000 (23:47 +0000)]
usb4bsd: Sync with FreeBSD r276791

9 years agousb4bsd: Sync with FreeBSD r276791
Markus Pfeiffer [Thu, 8 Jan 2015 23:40:22 +0000 (23:40 +0000)]
usb4bsd: Sync with FreeBSD r276791

* Sync usb_request

9 years agousb4bsd: Sync with FreeBSD r276791
Markus Pfeiffer [Thu, 8 Jan 2015 23:31:00 +0000 (23:31 +0000)]
usb4bsd: Sync with FreeBSD r276791

* sync usb_msctest

9 years agousb4bsd: Sync with FreeBSD r276791
Markus Pfeiffer [Thu, 8 Jan 2015 23:23:58 +0000 (23:23 +0000)]
usb4bsd: Sync with FreeBSD r276791

* sync the usb_hub code

9 years agousb4bsd: Sync with FreeBSD r276791
Markus Pfeiffer [Thu, 8 Jan 2015 23:10:52 +0000 (23:10 +0000)]
usb4bsd: Sync with FreeBSD r276791

9 years agousb4bsd: Sync with FreeBSD r276791
Markus Pfeiffer [Thu, 8 Jan 2015 23:03:36 +0000 (23:03 +0000)]
usb4bsd: Sync with FreeBSD r276791

9 years agousb4bsd: Sync with FreeBSD r276791
Markus Pfeiffer [Thu, 8 Jan 2015 22:09:17 +0000 (22:09 +0000)]
usb4bsd: Sync with FreeBSD r276791

* Update some files' SVN ids

9 years agousb4bsd: Sync with FreeBSD r276791
Markus Pfeiffer [Thu, 8 Jan 2015 21:58:50 +0000 (21:58 +0000)]
usb4bsd: Sync with FreeBSD r276791

* Sync up usb_busdma

9 years agousb4bsd: Remove some porting gunk
Markus Pfeiffer [Thu, 8 Jan 2015 21:51:43 +0000 (21:51 +0000)]
usb4bsd: Remove some porting gunk

* I wrapped the callout structure in a layer of indirection to
  simulate FreeBSD's callout that acquires a lock before calling
  into the callout function. This functionality is now available
  in DragonFly directly, so use it.

9 years agousb4bsd: Sync with FreeBSD r276791
Markus Pfeiffer [Thu, 8 Jan 2015 21:38:39 +0000 (21:38 +0000)]
usb4bsd: Sync with FreeBSD r276791

* update usb_transfer.{c,h} so uaudio can be ported and tested

9 years agodrm: Add wake_up_interruptible_all()
François Tigeot [Fri, 9 Jan 2015 10:40:42 +0000 (11:40 +0100)]
drm: Add wake_up_interruptible_all()

9 years agokernel: Remove stray SYSCTL_XUNLOCK from sysctl_sysctl_debug_dump_node().
Imre Vadasz [Thu, 8 Jan 2015 22:53:22 +0000 (23:53 +0100)]
kernel: Remove stray SYSCTL_XUNLOCK from sysctl_sysctl_debug_dump_node().

9 years agokernel/acpi: Add missing ACPI_LOCK_INIT in acpi_thermal.
Imre Vadasz [Thu, 8 Jan 2015 21:04:02 +0000 (22:04 +0100)]
kernel/acpi: Add missing ACPI_LOCK_INIT in acpi_thermal.

9 years agoMake pthread_*_destroy() more standards compliant
Michael Neumann [Thu, 8 Jan 2015 23:40:06 +0000 (00:40 +0100)]
Make pthread_*_destroy() more standards compliant

Function pthread_{mutex,cond,rwlock}_destroy() returned EINVAL when the
mutex/cond/rwlock was initialized statically via one of the
PTHREAD_*_INITIALIZER macros and not being used before destruction. We now
return success (0) instead, as it would have been the case when the *_init()
function were used for initialization. This is also the behaviour Linux
exhibits.

Note that we now can no longer detect multiple calls to *_destroy(). Multiple
calls will do no harm, but return success.

While there, fix some potential null pointer derefs in cond and rwlock.

Fixes: #2763

9 years agosound: Re-add Acer C720 support with some adjustments
Matthew Dillon [Thu, 8 Jan 2015 23:25:02 +0000 (15:25 -0800)]
sound: Re-add Acer C720 support with some adjustments

* Add the Acer C720 infrastructure back in.

* Move the core register programming patch into
  gnu/dev/sound/pci/hda/hda_acer_c720_patch.c along with the original
  GPL copyright from the linux file.

  This is the only part of the patch which could reasonably be said to
  have been taken from the linux driver.

* The rest of the patch is basically two things.

  First, mandatory adjustments to the NID routing to detect
  headphone insertion and removal, which was done through trial
  an error with my Acer and then verified against the linux driver.

  Second, we did include the disabling of the remixer based on a
  comment in the linux driver.  Basically NID 11 is disabled, two
  lines of code, and there's only one way to do that.

9 years agohammer - Fix timeout for 'dedup' directive
Antonio Huete Jimenez [Thu, 8 Jan 2015 11:30:05 +0000 (12:30 +0100)]
hammer - Fix timeout for 'dedup' directive

- 'hammer dedup' did not stop on timeout set by -t option.
- As the deduplication itself works by comparing one reference
  base element against a candidate base element, hammer(8) collects
  all the elements to be deduplicated and actually perform the needed
  ioctl() calls to make it happen. This is why the SIGALRM code has to
  be handled from within hammer(8).

Reported by: ftigeot
DragonFly-bug: <http://bugs.dragonflybsd.org/issues/2759>

9 years agodloader: Adjust some bits for the sound upgrade.
Sascha Wildner [Thu, 8 Jan 2015 18:33:47 +0000 (19:33 +0100)]
dloader: Adjust some bits for the sound upgrade.

9 years agoLINT: Move snd_uaudio out of the list of PCI devices.
Sascha Wildner [Thu, 8 Jan 2015 18:19:17 +0000 (19:19 +0100)]
LINT: Move snd_uaudio out of the list of PCI devices.

9 years agoQuick sound manpage upgrade (syncs with FreeBSD).
Sascha Wildner [Thu, 8 Jan 2015 18:16:51 +0000 (19:16 +0100)]
Quick sound manpage upgrade (syncs with FreeBSD).

9 years agomake.conf: WITH_NEW_XORG and WITH_KMS are no longer required
François Tigeot [Thu, 8 Jan 2015 17:34:13 +0000 (18:34 +0100)]
make.conf: WITH_NEW_XORG and WITH_KMS are no longer required

9 years agokernel/sound: Shorten the .PATH's in module Makefiles a bit.
Sascha Wildner [Thu, 8 Jan 2015 17:29:56 +0000 (18:29 +0100)]
kernel/sound: Shorten the .PATH's in module Makefiles a bit.

9 years agokernel/sound: Remove unused Makefile.
Sascha Wildner [Thu, 8 Jan 2015 17:26:38 +0000 (18:26 +0100)]
kernel/sound: Remove unused Makefile.

9 years agoFix up Makefile_upgrade.inc a bit.
Sascha Wildner [Thu, 8 Jan 2015 15:31:41 +0000 (16:31 +0100)]
Fix up Makefile_upgrade.inc a bit.

* Remove removed modules too.

* Remove cat4 manpages too.

9 years agosound: Remove Acer C720 support
François Tigeot [Thu, 8 Jan 2015 13:05:03 +0000 (14:05 +0100)]
sound: Remove Acer C720 support

There are concerns parts of this code have been derived from
GPL-licensed files.

Remove it.

9 years agosound: Bump __DragonFly_version, add notes to UPDATING
François Tigeot [Thu, 8 Jan 2015 10:03:34 +0000 (11:03 +0100)]
sound: Bump __DragonFly_version, add notes to UPDATING

9 years agosound: Remove debugging messages
François Tigeot [Wed, 7 Jan 2015 22:24:30 +0000 (23:24 +0100)]
sound: Remove debugging messages

9 years agosound: Make device cloning work
Matthew Dillon [Mon, 5 Jan 2015 10:40:32 +0000 (11:40 +0100)]
sound: Make device cloning work

* Some lockstatus() calls were wrong and needed to be compared
  with LK_EXCLUSIVE

* The sound device caches dynamically created devices and tries
  to reuse them, but devfs got angry if the clone code passed
  it a device which was already in the filesystem namespace. devfs
  had to be modified in 2c94b9eeb85c6cfa5ff918aa58ce484b27fc7c91
  and eb1c7cbc9e1052820efc6b0b7a94c40a7d45b329

* The clone code was designed to clone to new unit numbers, not
  clone an existing unit into multiple channels

9 years agosound: Make uaudio compile
François Tigeot [Sun, 4 Jan 2015 21:41:46 +0000 (22:41 +0100)]
sound: Make uaudio compile

9 years agosound: Device cloning tentative
François Tigeot [Sun, 7 Dec 2014 15:06:33 +0000 (16:06 +0100)]
sound: Device cloning tentative

* New /dev/dspx.y devices are properly created

* Sound can even be played, sometimes

* The most common errors are:
  - Can't open audio device /dev/dsp: Operation not supported by device
  - Can't open audio device /dev/dsp: Bad file descriptor

9 years agosound: Create default devfs entries for the best device
François Tigeot [Sat, 6 Dec 2014 20:07:37 +0000 (21:07 +0100)]
sound: Create default devfs entries for the best device

9 years agosound: Implement hw.snd.default_unit sysctl
François Tigeot [Thu, 4 Dec 2014 08:29:52 +0000 (09:29 +0100)]
sound: Implement hw.snd.default_unit sysctl

9 years agosound - haswell / ALC283 - Acer C720 support
Matthew Dillon [Sun, 10 Aug 2014 20:14:20 +0000 (13:14 -0700)]
sound - haswell / ALC283 - Acer C720 support

* Get headphone auto-sensing and speaker/headphone selection changes working.
  I couldn't figure out how to do it with the hpredir mechanism so I added
  a new field, w->senseredir, which tells the driver to also mess around
  with mixer selections.

  In the case of the Acer C720 nid 33 (the headphone) has two mixer
  selections, allowing either DMA channel to be selected while nid 20
  (the speaker) only has one.  The default selection winds up being
  wrong.  The senseredir field allows the sense mechanism to determine
  the correct selection and do change the active selection.

* WARNING!  We currently do not parse the acer out-of-band config data
  to figure out the headphone sense.

9 years agosound - Add haswell / ALC283 support (prelim)
Matthew Dillon [Sun, 10 Aug 2014 08:26:32 +0000 (01:26 -0700)]
sound - Add haswell / ALC283 support (prelim)

* Add haswell / ALC283 support (preliminary), taken from the linux driver.

9 years agosound: Fix pcm unit translation
Matthew Dillon [Sun, 10 Aug 2014 08:26:32 +0000 (01:26 -0700)]
sound: Fix pcm unit translation

Fix pcm unit translation so /dev/dspN actually maps to pcmN. Before
e.g. /dev/dsp* all mapped to pcm0.

9 years agosound: Fix compile issues when INVARIANTs is not specified
Matthew Dillon [Sun, 10 Aug 2014 08:26:32 +0000 (01:26 -0700)]
sound: Fix compile issues when INVARIANTs is not specified

9 years agosound: M_NOWAIT -> M_WAITOK, add missing M_ZERO
Matthew Dillon [Sun, 10 Aug 2014 08:26:32 +0000 (01:26 -0700)]
sound: M_NOWAIT -> M_WAITOK, add missing M_ZERO

The missing M_ZERO was causing kernel data leakage issues.

9 years agosound: Run KNOTE() from a taskqueue callback
Joris Giovannangeli [Sat, 9 Aug 2014 21:13:49 +0000 (23:13 +0200)]
sound: Run KNOTE() from a taskqueue callback

This is necessary to make pulseaudio work correctly.

9 years agosound: Create the first /dev/dsp* links
François Tigeot [Sat, 7 Jul 2012 21:00:02 +0000 (23:00 +0200)]
sound: Create the first /dev/dsp* links

At this point the sound device can be used, albeit by only one
application at the same time.

9 years agosound: Change dsp_clone() to the dragonfly way
François Tigeot [Sat, 7 Jul 2012 16:39:41 +0000 (18:39 +0200)]
sound: Change dsp_clone() to the dragonfly way

9 years agosound/hda: fix PCIe cache snooping
Sepherosa Ziehau [Wed, 11 Jul 2012 11:16:08 +0000 (13:16 +0200)]
sound/hda: fix PCIe cache snooping

* The PCIe snoop register is no longer vendor-specific, do not use
  hard-coded register values

* CPU architecture checks were also irrelevant, remove them

9 years agosound: Port to DragonFly - mixer_ioctl_cmd() parameters
François Tigeot [Thu, 5 Jul 2012 21:32:37 +0000 (23:32 +0200)]
sound: Port to DragonFly - mixer_ioctl_cmd() parameters

9 years agosound: Port to DragonFly - CTLFLAG_RWTUN
François Tigeot [Sat, 29 Nov 2014 08:46:41 +0000 (09:46 +0100)]
sound: Port to DragonFly - CTLFLAG_RWTUN