No subject

- DAHDI_TRANSCODE_OP is no longer supported, instead use DAHDI_TC_ALLOCATE,
  DAHDI_TC_GETINFO, and write/read.
- Memory and stack usage is reduced (stack usage could still be reduced some
  more by continuing the process of getting rid of the users of
  wctc4xxp_send_cmd and wctc4xxp_create_cmd).
- If more than one card is in the system channels will be allocated in a
  round-robin fashion from all available cards, reducing contention for the
  supervisor channel.
- There is no longer a tc400b workqueue created that will not show up in the
  process list.
- Commands and their responses are now explicitly matched up which elimated
  certain errors caused by unsolicited messages from the transcoder confusing
  the driver.
- There is now an option to export a network interface for capturing traffic
  to/from the hardware transcoder.
- codec_test has been removed from the dadhi/linux package.


........
r4720 | kpfleming | 2008-08-06 15:53:05 -0500 (Wed, 06 Aug 2008) | 2 lines

fix a compiler warning by using the proper format string for a size_t variable

........
r4721 | kpfleming | 2008-08-06 16:07:48 -0500 (Wed, 06 Aug 2008) | 2 lines

last round of license header updates here

........
r4724 | kpfleming | 2008-08-06 17:09:22 -0500 (Wed, 06 Aug 2008) | 2 lines

major cleanup, remove userspace tools references

........
r4725 | kpfleming | 2008-08-06 17:09:53 -0500 (Wed, 06 Aug 2008) | 2 lines

minor cleanups

........
r4728 | kpfleming | 2008-08-06 17:15:18 -0500 (Wed, 06 Aug 2008) | 1 line

remove incorrect reference
........
r4742 | sruffell | 2008-08-07 15:21:07 -0500 (Thu, 07 Aug 2008) | 2 lines

Make sure types are same size on 64-bit machines.

........
r4758 | tzafrir | 2008-08-11 10:31:49 -0500 (Mon, 11 Aug 2008) | 2 lines

xpp_blink is a bit mask of ports, and not boolean anymore.

........
r4759 | tzafrir | 2008-08-11 10:33:06 -0500 (Mon, 11 Aug 2008) | 2 lines

A more robust test for bri_dchan support.

........
r4760 | tzafrir | 2008-08-11 10:34:20 -0500 (Mon, 11 Aug 2008) | 2 lines

xpp: fxo: Fix support for CID_STYLE_PASS_ALWAYS

........
r4761 | tzafrir | 2008-08-11 11:22:24 -0500 (Mon, 11 Aug 2008) | 8 lines

xpp: CAS/E1 support in the PRI module, and minor fixes.

* Add support for CAS in the PRI module. Use firmware rev. 5975.
* Debugging parameter pcmtx_chan now accepts a dahdi channel number.
* Do initialize a reserved protocol field (card_global).
* The name DAHDI as used in proc/xpp/sync has 5 (not 6) letters.
* Fix DTMF "channel leak" regression in the FXS module.

........
r4770 | dbailey | 2008-08-13 16:04:07 -0500 (Wed, 13 Aug 2008) | 2 lines

import the neon mwi detection

........
r4773 | kpfleming | 2008-08-13 20:09:35 -0500 (Wed, 13 Aug 2008) | 2 lines

remove devfs support, and use consistent include file path

........
r4776 | tzafrir | 2008-08-14 04:50:14 -0500 (Thu, 14 Aug 2008) | 5 lines

xpp: FXO: display signed voltage values (from xpp r6055)

This is only an issue with the displayed value. In case you wondered why
you have battery voltage of more than 220V.

........
r4781 | tzafrir | 2008-08-14 14:15:41 -0500 (Thu, 14 Aug 2008) | 2 lines

Also install Astribank firmwares and init scripts to /usr/share/dahdi .

........
r4784 | tzafrir | 2008-08-14 16:37:41 -0500 (Thu, 14 Aug 2008) | 2 lines

Set xpp version to 6056 and reset XPP changelog.

........
r4788 | kpfleming | 2008-08-18 18:24:19 -0500 (Mon, 18 Aug 2008) | 2 lines

minor cleanups, and allow DAHDI_BUILD_ALL to be overriden on the command line

........
r4791 | sruffell | 2008-08-19 15:14:06 -0500 (Tue, 19 Aug 2008) | 9 lines

Couple of fixes for the transcoder:

- In dahdi_transcode.c,  Embed the identifiation number, assigned sequentially
  when the transcoders are registered, in the transcoder structure. This allows
  DAHDI_TC_GETINFO to work as expected even though the transcoders are rotated
  on the list in order to spread the load. 

- In wctc4xxp, fix bug where all transcoders are named tc400b0.

........
r4794 | sruffell | 2008-08-19 15:25:29 -0500 (Tue, 19 Aug 2008) | 2 lines

Fixed calculation of the timestamp.

........
r4795 | kpfleming | 2008-08-19 15:49:43 -0500 (Tue, 19 Aug 2008) | 1 line

port over improvements to firmware Makefile from Zaptel
........
r4798 | sruffell | 2008-08-20 14:31:45 -0500 (Wed, 20 Aug 2008) | 3 lines

Change to support both 5.3kbps and 6.3kbps bit rates when using the G723.1
codec.

........
r4801 | kpfleming | 2008-08-20 16:22:12 -0500 (Wed, 20 Aug 2008) | 8 lines

improve compatibility with 2.6.26 and 2.6.27 kernels

(closes issue #13253)
Reported by: raiden
Patches:
      zap-dev.patch uploaded by smurfix on issue #13277 (license 547)
      zap-sema.patch uploaded by smurfix on issue #13277 (license 547)

........
r4804 | kpfleming | 2008-08-20 16:35:13 -0500 (Wed, 20 Aug 2008) | 1 line

update code to match version in Zaptel
........
r4805 | kpfleming | 2008-08-20 17:20:46 -0500 (Wed, 20 Aug 2008) | 7 lines

use the new separate allocation method for channel structures here too

replace "ZTD" references in channel/span names with "DYN"

(closes issue #13302)
Reported by: KNK

........
r4817 | mattf | 2008-08-24 00:53:32 -0500 (Sun, 24 Aug 2008) | 1 line

Add support for 56 KB HDLC as well as selectable rate via ioctl
........
r4823 | sruffell | 2008-08-25 09:54:04 -0500 (Mon, 25 Aug 2008) | 3 lines

Remove this warning, which could occur if the driver is loaded on a system
without a wctc4xxp, and then subsequently unloaded.

........
r4828 | qwell | 2008-08-25 12:50:48 -0500 (Mon, 25 Aug 2008) | 1 line

Make sure we remove headers that we installed in install-include
........
r4829 | tzafrir | 2008-08-25 13:03:04 -0500 (Mon, 25 Aug 2008) | 3 lines

Support MODULES_EXTRA and SUBDIRS_EXTRA to add extra modules from the
make command line.

........
r4830 | tzafrir | 2008-08-25 13:05:55 -0500 (Mon, 25 Aug 2008) | 2 lines

Make the README file more relevant to modules.

........
r4834 | tzafrir | 2008-08-25 15:31:47 -0500 (Mon, 25 Aug 2008) | 5 lines

Reverting unwanted changes in the README (from r4830)

Also fixed ZT references in the ABI compatibility section. The numerical
ioctl values need to be recalculated, though.

........
r4837 | tzafrir | 2008-08-26 07:05:29 -0500 (Tue, 26 Aug 2008) | 3 lines

Don't try to clean modules if there's no kernel source available (like
in Zaptel).

........
r4840 | tzafrir | 2008-08-26 08:03:07 -0500 (Tue, 26 Aug 2008) | 2 lines

Fix asciidoc.

........
r4841 | tzafrir | 2008-08-26 08:04:32 -0500 (Tue, 26 Aug 2008) | 2 lines

"docs" target to generate documentation. Generate docs with asciidoc.

........
r4847 | qwell | 2008-08-27 11:48:13 -0500 (Wed, 27 Aug 2008) | 1 line

Set a list of headers to install/uninstall, so the lists don't get out of sync (note the previously missing fasthdlc.h in the uninstall-include target)
........
r4848 | kpfleming | 2008-08-27 12:08:42 -0500 (Wed, 27 Aug 2008) | 2 lines

a bit of attribution cleanup

........
r4849 | kpfleming | 2008-08-27 12:12:02 -0500 (Wed, 27 Aug 2008) | 1 line

remove some more ztdummy references
........
r4856 | mattf | 2008-08-29 16:46:26 -0500 (Fri, 29 Aug 2008) | 1 line

Remove useless kzalloc
........
r4861 | sruffell | 2008-09-04 16:21:44 -0500 (Thu, 04 Sep 2008) | 5 lines

Fixes an issue where the dual-span cards are not properly configured which can
cause data loss.

Fix provided by opticron and possibly related to issue #0013393.

........
r4864 | sruffell | 2008-09-04 16:29:17 -0500 (Thu, 04 Sep 2008) | 3 lines

Remove a couple of 'magic numbers' to make it clear all the ports of the
framer should be configured.

........
r4865 | mattf | 2008-09-04 16:42:29 -0500 (Thu, 04 Sep 2008) | 1 line

Some picky switches require the LSB to be 1 for 56k links
........
r4868 | mattf | 2008-09-06 15:24:59 -0500 (Sat, 06 Sep 2008) | 1 line

Fix class_simple on old 2.6.9 kernels
........
r4869 | mattf | 2008-09-06 15:25:39 -0500 (Sat, 06 Sep 2008) | 1 line

Revert unnecessary default hardhdlc mode from 56K to 64K
........
r4870 | mattf | 2008-09-06 15:27:59 -0500 (Sat, 06 Sep 2008) | 1 line

Fix buglet in #define for 2.6.9
........
r4894 | tzafrir | 2008-09-11 15:52:21 -0500 (Thu, 11 Sep 2008) | 2 lines

Move udev rules xpp.rules from dahdi-tools to dahdi-linux .

........
r4900 | tzafrir | 2008-09-11 18:00:57 -0500 (Thu, 11 Sep 2008) | 2 lines

Fix a type used in nethdlc mode, as pointed out in #13427 .

........
r4905 | sruffell | 2008-09-15 13:31:25 -0500 (Mon, 15 Sep 2008) | 2 lines

Fixed two typos.

........
r4909 | sruffell | 2008-09-15 15:49:31 -0500 (Mon, 15 Sep 2008) | 4 lines

Edit some comments and error strings.

Issue: DAHDI-13

........
r4917 | sruffell | 2008-09-17 10:26:05 -0500 (Wed, 17 Sep 2008) | 9 lines

Turn off reference counting on the echo canceller modules in order to prevent
misconfigurations from preventing the drivers from unloading.  

NOTE: This is only a temporary workaround, since it also means that the
echocanceller can be unloaded by an administrator while in use, which would
most likely result in a kernel oops.

Related to issue #13504.

........
r4921 | sruffell | 2008-09-17 13:00:17 -0500 (Wed, 17 Sep 2008) | 3 lines

Added a note about installing support for the B410P in the installation
section of the README file.

........
r4930 | sruffell | 2008-09-17 14:07:53 -0500 (Wed, 17 Sep 2008) | 3 lines

Just moved a note into it's own paragraph so that asciidoc can make the
appropriate callout.

........
r4956 | sruffell | 2008-09-18 16:21:11 -0500 (Thu, 18 Sep 2008) | 8 lines

close_channel is called with a spin_lock held, which means that GFP_KERNEL can
not be used for the memory allocations down that call path.

Have allocations in this call path check if they are in atomic context and use
the appropriate flags.

Issue: DAHDI-195

........
r4957 | sruffell | 2008-09-18 16:23:38 -0500 (Thu, 18 Sep 2008) | 2 lines

Adding a needed header for the in_atomic call.

........
r4971 | sruffell | 2008-09-24 01:15:48 -0500 (Wed, 24 Sep 2008) | 4 lines

Adding back in the mode module parameter for the wctc4xxp driver.  This is
primarily used to increase the number of channels available when only
transcoding to/from g729 by setting it to 'g729'.

........
r4979 | tzafrir | 2008-09-25 06:45:29 -0500 (Thu, 25 Sep 2008) | 3 lines

XPP FXO: Add caller-id workaround for ESTI-DTMF (for #9096)
* Also rename the CID_STYLE_* constants to formal names.

........
r4980 | tzafrir | 2008-09-25 06:52:12 -0500 (Thu, 25 Sep 2008) | 9 lines

Fix display of indirect registers and streamline their setting.

* This commit fixes display of indirect registers through the chipregs
  (formly "slics") procfs file. Only the low byte was displayed.
* It also deprecates previous {RW}S in favour of {RW}I. The prevois
  style is still allowed but deprecated, and thus previous scripts will
  still work.


........
r4986 | tzafrir | 2008-09-25 11:50:30 -0500 (Thu, 25 Sep 2008) | 5 lines

XPP firmware: Only send out CAS D-channel messages when in CAS mode.




........
r4990 | sruffell | 2008-09-25 22:20:49 -0500 (Thu, 25 Sep 2008) | 5 lines

The channel master should not be set to 0, but rather should be 'cleared' by
setting the channel to be it's own master.  

(related to issue 11611)

........
r5002 | sruffell | 2008-09-28 01:43:51 -0500 (Sun, 28 Sep 2008) | 4 lines

Fixes failure of modular echo cancelers in DAHDI.

Reported by lots of people, fix suggested by mattf.

........
r5003 | sruffell | 2008-09-28 02:04:16 -0500 (Sun, 28 Sep 2008) | 5 lines

Fixed type of flags parameter to spin_lock_irqsave functions. Fixes
compilation issues on platforms where int and long do not have the same size.

Closes Issues #0013575. Reported by Ulmo.

........
r5007 | sruffell | 2008-09-28 12:29:12 -0500 (Sun, 28 Sep 2008) | 5 lines

Improve reliablity of UK caller ID for the TDM400P by not allowing the
ringdebounce to be decremented when it is at 0 already.

Related to issue #12531.  Reported mattbrown, fix suggested by benbrown.

........
r5017 | sruffell | 2008-09-30 15:29:59 -0500 (Tue, 30 Sep 2008) | 2 lines

Removing references to setup_timer in pre 2.6.18 kernels.

........
r5021 | sruffell | 2008-10-01 16:45:24 -0500 (Wed, 01 Oct 2008) | 4 lines

DAHDI should always make data received from the PSTN available to user mode
immediately.  Only allow the transmit buffering policy to be changed in order
to reduce the chance of underruns to the PSTN.

........
r5025 | sruffell | 2008-10-02 15:32:26 -0500 (Thu, 02 Oct 2008) | 4 lines

Mark that there is not any data waiting whenever we cleanup the private
channel structures.
Issue: DAHDI-42

........
r5029 | sruffell | 2008-10-02 16:01:07 -0500 (Thu, 02 Oct 2008) | 5 lines

wctc4xxp_cleanup_channel_private needs a pointer to the zt_transcoder_channel
and not just the private portion now in order to manage the data ready flag
state.


........
r5030 | sruffell | 2008-10-02 16:14:39 -0500 (Thu, 02 Oct 2008) | 6 lines

Keep hold of the channel lock when setting the data_ready flag for the channel
after writing to the receive queue.  Prevents a warning that data was on the
recieve queue but the data ready flag was not set.  

Issue: DAHDI-42

........
r5034 | sruffell | 2008-10-03 10:39:21 -0500 (Fri, 03 Oct 2008) | 2 lines

There are two possible valid statues when booting the TC400M.

........
r5045 | seanbright | 2008-10-03 15:09:40 -0500 (Fri, 03 Oct 2008) | 1 line

Fix some compilation problems that show up when CONFIG_DAHDI_DEBUG is defined.
........
r5046 | tzafrir | 2008-10-03 15:16:24 -0500 (Fri, 03 Oct 2008) | 5 lines

Fix building with CONFIG_DAHDI_NET . It builds, but will it run?

Patch dahdi-base.c.hdlc.patch by biohumanoid that fixes some aparant
copy&paste errors.

........
r5051 | tzafrir | 2008-10-03 15:32:29 -0500 (Fri, 03 Oct 2008) | 5 lines

Fix building with CONFIG_DAHDI_NET in kernel 2.6.22

The leftovers of issues #13542 (which was mostly resolved in previous
committ. This closes it.

........
r5056 | seanbright | 2008-10-06 11:55:37 -0500 (Mon, 06 Oct 2008) | 8 lines

Fix a few compile errors that only show up when CONFIG_DAHDI_PPP is
defined.

(closes issue #13608)
Reported by: Nik Soggia
Fix suggested by: Nik Soggia
Tested by: seanbright

........
r5060 | sruffell | 2008-10-06 12:48:07 -0500 (Mon, 06 Oct 2008) | 5 lines

Embed room for the complete packet to the DTE in the transcoder buffer
structure.  Simplifies alignment management at the cost of a little wasted
memory, but the end results is that operation is more reliable on more
systems.

........
r5064 | kpfleming | 2008-10-06 15:52:43 -0500 (Mon, 06 Oct 2008) | 2 lines

use the same logic here as elsewhere for releasing echocan module references

........
r5068 | kpfleming | 2008-10-08 21:54:15 -0500 (Wed, 08 Oct 2008) | 2 lines

a micro-optimization found while creslin and i spent four or five hours tracking down a very complex problem

........
r5084 | sruffell | 2008-10-10 17:38:44 -0500 (Fri, 10 Oct 2008) | 5 lines

- Ensure that the source format is considered when selecting a transcoder.
- When a command is to be retried, turn off the TX_COMPLETE flag before
  resubmitting it to the hardware. This should elimate some of the warnings
  printed to the kernel log in the wctc4xxp_transmit_cmd function.

........
r5090 | sruffell | 2008-10-14 17:11:45 -0500 (Tue, 14 Oct 2008) | 3 lines

If the vpmad032 firmware needs to be reloaded, make sure we use the same slot
in the ifaces array.

........
r5097 | tzafrir | 2008-10-16 12:40:22 -0500 (Thu, 16 Oct 2008) | 21 lines

xpp: start migration from procfs to sysfs.

* Sysfs representation for XPDs:
  /sys/bus/xpds/devices/<bus>:<unit>:<subunit>
* Astribanks sysfs directories now include the XPDs as subdirectories:
  e.g. /sys/bus/astribanks/devices/xbus-00/00:3:0
* procfs control interface deprecated: conditioned by OLD_PROC
  (defaults to off). Control functionality moved to sysfs:
* xbus attributes: cls connector label status timing waitfor_xpds xbus_state
* XPDs can have driver-specific attributes. Common attriubtes:
  blink chipregs span
* PRI-specific attributes:
  pri_clocking pri_dchan pri_cas pri_alarms pri_layer1 pri_localloop pri_protocol
* The Astribank attribute "xbus_state" is read/write. Reading it shows
  the current state of the Astribank. Writing "start" or "stop" allows a
  software equivalent of connect or disconnect respectively.
* When an Astribank is ready it sends an "online" event. Whenever its
  not ready (e.g. at the time of disconnect) it sends an "offline" event.
  Use astribank_hook.sample to handle those.


........
r5105 | tzafrir | 2008-10-21 12:00:03 -0500 (Tue, 21 Oct 2008) | 2 lines

Fix headers numbering.

........
r5106 | tzafrir | 2008-10-21 12:08:36 -0500 (Tue, 21 Oct 2008) | 4 lines

dahdi-linux README: better place for build requirements.

(We'll have to put a content there one of these days)

........
r5107 | tzafrir | 2008-10-21 12:17:19 -0500 (Tue, 21 Oct 2008) | 2 lines

dahdi-trunk: ignore the generated README.html .

........
r5111 | tzafrir | 2008-10-21 12:34:12 -0500 (Tue, 21 Oct 2008) | 2 lines

An experimental OSLEC echocan module.

........
r5115 | tzafrir | 2008-10-22 03:42:22 -0500 (Wed, 22 Oct 2008) | 2 lines

xpp FXS init script: Do use high-pass filter.

........
r5118 | tzafrir | 2008-10-22 06:44:54 -0500 (Wed, 22 Oct 2008) | 9 lines

wct1xxp: fix error handling at device startup.

And also give more useful error messages if things go bad.

(closes issue #13607)

Patches: wct1xxp_pci.diff uploaded by tzafrir (license 46)
Tested by: klaus3000

........
r5121 | qwell | 2008-10-22 12:19:40 -0500 (Wed, 22 Oct 2008) | 8 lines

Fix building on big endian machines.

(closes issue #13754)
Reported by: shrift
Patches:
      wctdm-powerpc.patch uploaded by irroot (license 52)
Tested by: shrift

........
r5124 | qwell | 2008-10-22 13:49:41 -0500 (Wed, 22 Oct 2008) | 5 lines

Need to make sure we check for udevadm in addition to udevinfo.
Some silly person (or people) decided that it wasn't useful to have in their distro...

*cough*

........
r5127 | tzafrir | 2008-10-27 10:14:36 -0500 (Mon, 27 Oct 2008) | 4 lines

kernel.h: cleanup DAHDI_FLAG_* defines, no functional change.

Define the DAHDI_FLAG_* using the DAHDI_FLAGBIT_* enum values.

........
r5130 | tzafrir | 2008-10-27 10:49:09 -0500 (Mon, 27 Oct 2008) | 7 lines

Fix xpp compile problems on kernel < 2.6.16

Support for kernels that use the older hotplug support rather than the
newer uevent.

Fixes the xpp issue from #13427.

........
r5131 | tzafrir | 2008-10-27 10:53:42 -0500 (Mon, 27 Oct 2008) | 4 lines

xpp: Make some definitions static, as per sparse.

Fixes the xpp warnings of #13763 (except some false alarms).

........
r5132 | tzafrir | 2008-10-27 10:57:44 -0500 (Mon, 27 Oct 2008) | 2 lines

xpp: remove an unused manual reference count field.

........
r5133 | tzafrir | 2008-10-27 11:00:49 -0500 (Mon, 27 Oct 2008) | 2 lines

xpp fxo: Add sysfs battery attribute.

........
r5138 | tzafrir | 2008-10-27 14:27:47 -0500 (Mon, 27 Oct 2008) | 2 lines

dahdi linux README: Clarify OSLEC EC build procedure.

........
r5148 | sruffell | 2008-10-28 13:12:48 -0500 (Tue, 28 Oct 2008) | 2 lines

Adding the wcb4xxp driver, a native dahdi driver for the B410P module.

........
r5149 | sruffell | 2008-10-28 13:19:53 -0500 (Tue, 28 Oct 2008) | 3 lines

Undoing commit r5148.  Copying directly from a private branch causes problems
when trying to view the history of a file. 

........
r5150 | sruffell | 2008-10-28 13:26:33 -0500 (Tue, 28 Oct 2008) | 2 lines

Adding the wcb4xxp driver, a native dahdi driver for the B410P module.

........
r5156 | kpfleming | 2008-10-28 13:55:03 -0500 (Tue, 28 Oct 2008) | 1 line

various cleanups, primarily proper capitalization
........
r5162 | kpfleming | 2008-10-28 16:49:02 -0500 (Tue, 28 Oct 2008) | 8 lines

fix a large number of warnings found by sparse, the kernel code sanity checking tool. some of these fixes are non-optimal (casting 'unsigned long' to '__user void *'), but are unavoidable in many cases. started from tzafrir's patch, did most of the work myself.

(closes issue #13763)
Reported by: tzafrir
Patches:
      sparse_fixes_1.diff uploaded by tzafrir (license 46)


........
r5163 | kpfleming | 2008-10-28 16:59:05 -0500 (Tue, 28 Oct 2008) | 1 line

update to latest octasic_api tag with NULL definition fix
........
r5167 | tzafrir | 2008-10-29 10:09:11 -0500 (Wed, 29 Oct 2008) | 6 lines

xpp: Increase the maximal size of the command queue to 500 .

Temporarily increase the maximal size of the command queue from 300 to
500 as a workaround of an issue at initialization time (mainly of
BRI+FXS).

........
r5171 | sruffell | 2008-10-29 10:31:10 -0500 (Wed, 29 Oct 2008) | 2 lines

By default, do not create the procfs entry for the wcb4xxp driver.

........
r5175 | sruffell | 2008-10-29 11:48:33 -0500 (Wed, 29 Oct 2008) | 3 lines

The /proc/wcb4xxp should not be created once for each card but rather once for
all cards.

........
r5179 | tzafrir | 2008-10-30 14:05:05 -0500 (Thu, 30 Oct 2008) | 9 lines

Use ! to hint udev about directory separator.

Udev knows how to convert a '!' in the device name to a directory
separator. Thus the name 'dahdi!ctl' will create the device
/dev/dahdi/ctl .

We still keep older udev rules in this release for compatibility, but
eventually we'll only need them to set permissions.

........
r5180 | tzafrir | 2008-10-30 14:16:08 -0500 (Thu, 30 Oct 2008) | 2 lines

Use correct length for the, well, dahdi transcoder device name.

........
r5186 | sruffell | 2008-10-31 12:11:02 -0500 (Fri, 31 Oct 2008) | 5 lines

Fix for a case where a span might not always come back up after a disconnect.

Patch provided by akohlsmith.
Issue: DAHDI-174

........
r5190 | tzafrir | 2008-10-31 15:55:08 -0500 (Fri, 31 Oct 2008) | 9 lines

wcb4xxp: Claim to provide DACS for dahdi_scan

dahdi_scan can tell that a span is digital if its first channel has DACS 
signalling capability. While this is probably not supported by the
current driver, it is also harmless.

Without this, dahdi_scan shows the spans of this card as analog.
(Done after consulting with sruffel)

........
r5191 | tzafrir | 2008-10-31 16:05:04 -0500 (Fri, 31 Oct 2008) | 2 lines

wcb4xxp: Set the spantype to "BRI" (to show e.g. in dahdi_scan)

........
r5195 | mattf | 2008-10-31 16:33:55 -0500 (Fri, 31 Oct 2008) | 1 line

Fix the FIFO configuration to use the data from the EC correctly
........
r5196 | sruffell | 2008-10-31 17:23:52 -0500 (Fri, 31 Oct 2008) | 2 lines

Changing the spantype to indicate whether it is a TE or NT BRI port.

........
r5203 | tzafrir | 2008-11-02 04:13:04 -0600 (Sun, 02 Nov 2008) | 5 lines

README: no need to generate Kbuild for OSLEC.

No need to create a Kbuild file. staging/echo has a Makefile that works
fine for us.

........
r5207 | tzafrir | 2008-11-02 11:33:10 -0600 (Sun, 02 Nov 2008) | 6 lines

dahdi-base: hw_echocancel_off: return 0 if no hardware EC

If the span has no hardware EC, return 0, rather than a random
uninitialized value (which was no harm, as that return value is always
ignored anyway).

........
r5211 | tzafrir | 2008-11-03 06:01:01 -0600 (Mon, 03 Nov 2008) | 2 lines

Adjust DAHDI to the new timers interface of kernel 2.6.28

........
r5230 | sruffell | 2008-11-05 13:17:16 -0600 (Wed, 05 Nov 2008) | 2 lines

Allow the wcb4xxp to take all the signalling types.

........
r5232 | sruffell | 2008-11-05 13:36:52 -0600 (Wed, 05 Nov 2008) | 2 lines

Remove the loopback module parameter since it's not implemented yet.

........
r5236 | sruffell | 2008-11-05 14:11:03 -0600 (Wed, 05 Nov 2008) | 11 lines

Check the return value of the down_interruptible call in order to quiet a
warning.  This semaphore does not protect any host data structures, but only
accesses to the VPMADT032 module.  The worse thing that could happen is that
the internal state of the VPM module is corrupted, and then would only happen
on module loading because otherwise access because that is the only time this
function is called in the context of a user process.  In this case, the module
would need to be reloaded again anyway.

(Closes issue #13742) 
Reported by smurfix

........
r5237 | sruffell | 2008-11-05 14:17:50 -0600 (Wed, 05 Nov 2008) | 5 lines

Fix warning messages in order to build on 2.6.27.

(Closes issue #13757)
Patch provided by tzafrir.

........
r5249 | sruffell | 2008-11-05 15:06:31 -0600 (Wed, 05 Nov 2008) | 2 lines

rate limiting a diagnostic printk.

........
r5257 | sruffell | 2008-11-05 17:45:09 -0600 (Wed, 05 Nov 2008) | 2 lines

Whitespace changes for the coding standard.

........
r5269 | seanbright | 2008-11-08 18:31:07 -0600 (Sat, 08 Nov 2008) | 4 lines

Add missing semi-colon.

Reported & Tested by: KP7 via #asterisk-dev

........
r5270 | seanbright | 2008-11-08 18:33:16 -0600 (Sat, 08 Nov 2008) | 4 lines

We only use print_debug_writebuf when CONFIG_DAHDI_NET or CONFIG_DAHDI_PPP are defined, so only define it in those cases as well.

Reported & Tested by: KP7 via #asterisk-dev

........
r5275 | sruffell | 2008-11-10 13:48:22 -0600 (Mon, 10 Nov 2008) | 11 lines

- Do not hold any locks while calling close_channel, which can result in calls
  to the echocan modules which do not necesarrily assume they are being called
  in atomic context.
- Remove the bigzaplock around calls to psuedo_alloc and pseudo_free.  The
  structures protected by this lock are already protected by the chan_lock in
  these two cases.
- Remove calls to in_atomic() that were previously added to work around this,
  but did not cover all the cases.

Issue: DAHDI-195, DAHDI-170

........
r5288 | tzafrir | 2008-11-10 14:37:52 -0600 (Mon, 10 Nov 2008) | 4 lines

xpp init_card_2_30: no need to check environment in verify mode.

Fixes #13832 in Zaptel.

........
r5303 | sruffell | 2008-11-13 16:10:51 -0600 (Thu, 13 Nov 2008) | 4 lines

Make a message about the host cacheline size being unsupported a debug only
message.  The voicebus interface still works whether the cacheline size is
supported or not, but the message is confusing to users.

........
r5310 | tzafrir | 2008-11-16 13:30:23 -0600 (Sun, 16 Nov 2008) | 9 lines

Fixed USB firmware that caused some bad bioses to hang on boot.

The BIOSes included in a number of motherboards could hang from an
Astribank firmware newer than 1.2.20.1 / 1.4.5.1 . This was due to an
some incorrect USB information in the firmware. This firmware fixes it.

Merged Zaptel revisions 4580 via svnmerge from 
http://svn.digium.com/svn/zaptel/branches/1.2

........
r5315 | sruffell | 2008-11-17 11:44:10 -0600 (Mon, 17 Nov 2008) | 5 lines

A fix for an issue with corruption on the D-Channels.

Patch provided by akohlsmith.
Issue: DAHDI-173.

........
r5320 | tzafrir | 2008-11-17 12:01:47 -0600 (Mon, 17 Nov 2008) | 3 lines

xpp_pri: Fix T1 CRC initialization
         Clock synchronization when sync is not from first port.

........
r5321 | sruffell | 2008-11-17 12:17:35 -0600 (Mon, 17 Nov 2008) | 7 lines

- Fix for race condition of encoder and decoder are allocated at the same time
  in the driver.  This would result in -EBUSY returns from the
  DAHDI_TC_ALLOCATE ioctl.
- Increase the length of the receive descriptor ring from 8 to 32 to reduce
  the probability of running out of receive descriptors.


........
r5335 | kpfleming | 2008-11-19 15:25:01 -0600 (Wed, 19 Nov 2008) | 3 lines

support wcb4xxp build on kernels that don't directly use Kbuild


........
r5340 | tzafrir | 2008-11-20 03:23:29 -0600 (Thu, 20 Nov 2008) | 4 lines

* Some extrra Zap->DAHDI (and Dahdi->DAHDI) fixes in the Astribank
  README.
* Some extra asciidoc formatting fixes.

........
r5345 | tzafrir | 2008-11-20 06:31:59 -0600 (Thu, 20 Nov 2008) | 4 lines

Make tor2 load properly. Seems to fix #13487.

Thanks to heyuqi for the testing.

........
r5350 | sruffell | 2008-11-20 10:03:50 -0600 (Thu, 20 Nov 2008) | 5 lines

Force the FIFO to reset when the file handle is closed.

Patch provided by akohlsmith.
Issue: DAHDI-178

........
r5355 | sruffell | 2008-11-20 22:42:21 -0600 (Thu, 20 Nov 2008) | 5 lines

Convert some uses of sprintf to snprintf in dahdi_proc_read in order to
eliminate a buffer overrun.

Issue: DAHDI-209

........
r5360 | qwell | 2008-11-21 14:15:47 -0600 (Fri, 21 Nov 2008) | 5 lines

Fix a think-o in numeric comparison.  Swap order to make it more clear.

(closes issue #13813)
Reported by: ys

........
r5367 | sruffell | 2008-11-23 22:09:56 -0600 (Sun, 23 Nov 2008) | 6 lines

Do not make assumptions about the number of ready HDLC frames on HDLC RX
interrupt.  This prevents libpri from becoming confused when many HDLC frames
arrive before the driver can service them or a false RX interrupt is received.

Patch provided by akolsmith. Issue DAHDI-173.

........
r5368 | sruffell | 2008-11-23 22:14:37 -0600 (Sun, 23 Nov 2008) | 5 lines

Use counter of pending HDLC frames in order to eliminate the need to send a
zero-byte frame to kick start the transmission process.

Patch provided by akohlsmith. Issue DAHDI-173.

........
r5369 | sruffell | 2008-11-23 22:16:56 -0600 (Sun, 23 Nov 2008) | 5 lines

Add a 'spanfilter' module parameter in order to isolate debugging information
to just the spans of interest.

Patch provided by akohlsmith. Issue DAHDI-173. 

........

[... 34876 lines stripped ...]