openSUSE Build Service

Overview Repositories Revisions Requests Users Attributes Meta

File ChangeLog of Package ffado (Revision af6b1d434d875699d420d4a174942169)

Currently displaying revision af6b1d434d875699d420d4a174942169 , Show latest

------------------------------------------------------------------------
r2040 | jwoithe | 2012-02-13 13:58:09 +0200 (Mon, 13 Feb 2012) | 1 line

rme: added TCO option to the preferred sync ref control set (it's disabled for the ff400 since the TCO can't be fitted to this interface - it's ff800 only).  The preferred sync ref control in ffado-mixer is now active.  As for the other sync controls it is yet to be tested with real sync sources connected to the ff400.  This completes the first pass of the device controls for the ff400 within ffado-mixer.
------------------------------------------------------------------------
r2039 | jwoithe | 2012-02-07 14:23:56 +0200 (Tue, 07 Feb 2012) | 1 line

rme: for completeness, include the sysclock mode control type in the write case; it was overlooked earlier.
------------------------------------------------------------------------
r2038 | jwoithe | 2012-02-07 14:11:19 +0200 (Tue, 07 Feb 2012) | 1 line

rme: various information panes are now active in ffado-mixer: sync check, autosync reference, spdif freqency and system clock.  The function of these in the presence of actual external sync sources is yet to be confirmed.  The system clock information has been confirmed to update when the clock mode and/or sample rate controls are altered.
------------------------------------------------------------------------
r2037 | jwoithe | 2012-02-05 13:47:14 +0200 (Sun, 05 Feb 2012) | 1 line

rme: make clock mode control active.  Note that this isn't overly useful without the other clock-related controls and indicators being active.  They will follow soon.
------------------------------------------------------------------------
r2036 | jwoithe | 2012-02-04 08:09:33 +0200 (Sat, 04 Feb 2012) | 1 line

Adjust my email address
------------------------------------------------------------------------
r2035 | jwoithe | 2012-02-03 14:03:09 +0200 (Fri, 03 Feb 2012) | 1 line

rme: clock source selection and management on the RME devices is a bit trickier than a simple list implies.  Down the track it might be possible to use the standard clock source list functionality in some way, but for now fake it out with a dummy clock source to prevent warnings about no clock sources being active (and so on).
------------------------------------------------------------------------
r2034 | jwoithe | 2012-02-03 13:20:28 +0200 (Fri, 03 Feb 2012) | 1 line

rme: a return value of 0 from set*() indicates no error, so fix those locations where it was assumed a zero denoted an error.  This is really only to prevent confusion later on: at present none of the affected controls rely on m_value being set.
------------------------------------------------------------------------
r2033 | jwoithe | 2012-02-03 13:15:10 +0200 (Fri, 03 Feb 2012) | 1 line

rme: implement SPDIF output option control via ffado-mixer.  Again, I haven't the means to confirm whether these settings have the desired effect on the SPDIF bitstream yet.
------------------------------------------------------------------------
r2032 | jwoithe | 2012-02-03 12:35:02 +0200 (Fri, 03 Feb 2012) | 1 line

rme: implement SPDIF input mode selection via ffado-mixer.  Whether this has the desired effect on the device remains to be seen - I do not have anything to test the mode of the optical port right now.
------------------------------------------------------------------------
r2031 | jwoithe | 2012-02-02 14:36:30 +0200 (Thu, 02 Feb 2012) | 1 line

matrixmixer: don't allow column headers to expand beyond the size required to fit the text.  Otherwise the vertical size of the column headers can be very large in the case where a matrix has only a few rows.
------------------------------------------------------------------------
r2030 | jwoithe | 2012-02-02 14:19:11 +0200 (Thu, 02 Feb 2012) | 1 line

matrixmixer: comment out debug output
------------------------------------------------------------------------
r2029 | jwoithe | 2012-02-02 14:18:37 +0200 (Thu, 02 Feb 2012) | 1 line

matrixmixer: include graphical indicators for mute and invert status.  Muted faders are displayed with a dark gray background, while inverted channels have a phi (the universally recognised phase symbol) in the top left corner.
------------------------------------------------------------------------
r2028 | jwoithe | 2012-02-02 13:33:52 +0200 (Thu, 02 Feb 2012) | 1 line

rme: remove mixer debug output which  is no longer required
------------------------------------------------------------------------
r2027 | jwoithe | 2012-02-02 12:38:03 +0200 (Thu, 02 Feb 2012) | 1 line

rme: add mute/invert controls to playback mixer faders, and mute controls to the output faders
------------------------------------------------------------------------
r2026 | jwoithe | 2012-02-02 12:23:55 +0200 (Thu, 02 Feb 2012) | 1 line

Minor updates to MOTU references in README.  The official spelling of Traveler uses the US convention (ie: one l).  Remove references to the qt3 mixer since that's no longer in trunk.
------------------------------------------------------------------------
r2025 | adi | 2012-02-01 23:18:59 +0200 (Wed, 01 Feb 2012) | 13 lines

Update requirements in README

Patch provided by Stefan Richter. He writes:

While libraw1394 v1.3.0 is indeed the minimum build requirement, and
v2.0.5 the minimum runtime requirement in absence of the raw1394 kernel
driver of kernel 2.6.36 and older, the only realistic requirement today
is libraw1394 v2.0.7.  The older versions lack bug fixes that are
important for use with FFADO.

Not sure whether the current libiec61883 v1.2.0 contains anything
relevant
to FFADO, compared to v1.1.0.
------------------------------------------------------------------------
r2024 | adi | 2012-02-01 19:24:55 +0200 (Wed, 01 Feb 2012) | 1 line

ffado-diag: check the juju device IDs and group membership
------------------------------------------------------------------------
r2023 | adi | 2012-02-01 19:24:44 +0200 (Wed, 01 Feb 2012) | 1 line

ffado-diag: helper functions for juju device permissions and group membership
------------------------------------------------------------------------
r2022 | adi | 2012-02-01 19:24:32 +0200 (Wed, 01 Feb 2012) | 1 line

ffado-diag: [cosmetics] remove code duplication
------------------------------------------------------------------------
r2021 | jwoithe | 2012-01-30 15:02:09 +0200 (Mon, 30 Jan 2012) | 4 lines

matrixmixer: complete initial implementation of optional mute functionality.  A graphical indication of mute state (beyond the toggled menu item) is still needed.
matrixmixer: implement optional phase inversion interface.  Again, a graphical indication of this state is still required.
rme: make use of the mute/invert functionality of the matrix mixer.  Currently this is only connected for the input faders.

------------------------------------------------------------------------
r2020 | adi | 2012-01-19 20:09:35 +0200 (Thu, 19 Jan 2012) | 4 lines

Remove .gitignore.

This file is only necessary in the git repository.

------------------------------------------------------------------------
r2019 | jwoithe | 2012-01-11 10:54:59 +0200 (Wed, 11 Jan 2012) | 8 lines

From Orcan via trac ticket #344.  Writes Orcan:

It is that time of the year that the gcc folks decide to impose the
 standards at a higher level. Attached is a trivial patch that will make
 libffado compile against new gcc.

Thanks for the patch Orcan.

------------------------------------------------------------------------
r2018 | jwoithe | 2011-12-22 00:00:37 +0200 (Thu, 22 Dec 2011) | 2 lines

Apply fix to mono Saffire mixer as suggested by John Vaites.  Logically this seems like the correct thing to do but as I don't have a Saffire I can't test this myself.  If this breaks other interfaces please revert and apply something more appropriate.  The error this fixes in John's case is "AttributeError: 'SaffireMixerMono' object has no attribute 'sldIN1Out910'".

------------------------------------------------------------------------
r2017 | jwoithe | 2011-12-20 13:41:58 +0200 (Tue, 20 Dec 2011) | 1 line

matrixmixer: more infrastructure to permit control of muting if supported by the hardware
------------------------------------------------------------------------
r2016 | jwoithe | 2011-12-19 14:17:30 +0200 (Mon, 19 Dec 2011) | 3 lines

rme: refine mute hardware mute controls.  Add mute support for mixer outputs.  Create dbus mixer objects to control channel muting and inversion.
matrixmixer: prepare infrastructure to allow channel muting to be controlled if enabled by the caller.

------------------------------------------------------------------------
r2015 | jwoithe | 2011-12-09 13:33:05 +0200 (Fri, 09 Dec 2011) | 1 line

motu: more 828mk1 debugging
------------------------------------------------------------------------
r2014 | jwoithe | 2011-12-09 01:03:59 +0200 (Fri, 09 Dec 2011) | 1 line

motu: additional debug output to aid ongoing exploration of the 828mk1
------------------------------------------------------------------------
r2013 | jwoithe | 2011-12-06 13:34:53 +0200 (Tue, 06 Dec 2011) | 1 line

motu: fix a typo introduced in r2012
------------------------------------------------------------------------
r2012 | jwoithe | 2011-12-06 12:34:16 +0200 (Tue, 06 Dec 2011) | 1 line

motu: allocate the send iso channel before the receive channel to see if the 828mk1 - as currently set up - is sensitive to the precise channels allocated.  If an 828mk1 is in use, show the iso control register value as read from the device for debug purposes.
------------------------------------------------------------------------
r2011 | jwoithe | 2011-12-05 14:19:21 +0200 (Mon, 05 Dec 2011) | 1 line

rme: implement the output fader controls in ffado-mixer.  There's still some work to do on sizing and labelling but the basic fader functionality is there.
------------------------------------------------------------------------
r2010 | jwoithe | 2011-12-05 14:03:40 +0200 (Mon, 05 Dec 2011) | 1 line

matrixmixer GUI: was there any reason why the column count was used as the number of rows, and vice versa?  This has been fixed, subject to feedback from others.
------------------------------------------------------------------------
r2009 | jwoithe | 2011-12-02 14:34:58 +0200 (Fri, 02 Dec 2011) | 2 lines

MOTU: fix G1 monitor-in control bit defines (what *was* I thinking when I coded those the first time around?).  Clarify the registers used with various G1 defines in the header file.  At present the original 828 streams successfully but the physical inputs and outputs are muted; enable an additional 2 control register bits when streaming is started to see if the device is unmuted by these.

------------------------------------------------------------------------
r2008 | jwoithe | 2011-11-22 14:21:59 +0200 (Tue, 22 Nov 2011) | 2 lines

rme: refine input matrix mixer channel names.  Add low level support for mixer channel phase inversion and muting (untested).  Add playback mixer to RME ffado-mixer.

------------------------------------------------------------------------
r2007 | jwoithe | 2011-11-22 11:37:38 +0200 (Tue, 22 Nov 2011) | 1 line

matrixmixer widget: label channels starting at channel 1 rather than 0, since this almost certainly corresponds better with channel labels used on almost all interfaces.  If there's a problem let me know.
------------------------------------------------------------------------
r2006 | jwoithe | 2011-11-21 14:26:26 +0200 (Mon, 21 Nov 2011) | 4 lines

* rme: implement a first cut at the input matrix mixer.  This permits inputs to be mixed to outputs directly in the device for latency-free monitoring applications.  The playback matrix mixer and output fader controls will follow soon now this initial proof-of-concept is done.
* matrixmixer: added new optional parameter to the this widget's constructor allowing the control maximums to be set.  If not given the default is 65536/+12dB as before.

------------------------------------------------------------------------
r2005 | jwoithe | 2011-11-20 13:25:56 +0200 (Sun, 20 Nov 2011) | 1 line

rme: initial experiments to hook the MatrixMixer widget into the RME ffado-mixer module.  More to be done.
------------------------------------------------------------------------
r2004 | jwoithe | 2011-11-18 14:28:04 +0200 (Fri, 18 Nov 2011) | 1 line

rme: complete mixer glue code.  Create dbus controls for the onboard matrix mixer (these will ultimately be used by ffado-mixer).
------------------------------------------------------------------------
r2003 | jwoithe | 2011-11-17 13:54:44 +0200 (Thu, 17 Nov 2011) | 1 line

rme: implement some glue needed to make mixer controls available via ffado-mixer
------------------------------------------------------------------------
r2002 | jwoithe | 2011-11-07 15:08:55 +0200 (Mon, 07 Nov 2011) | 1 line

Add configuration entry for Presonus Firestudio Mobile.  Thanks to "Pule" for the forum post with this information.
------------------------------------------------------------------------
r2001 | jwoithe | 2011-11-04 13:21:08 +0200 (Fri, 04 Nov 2011) | 1 line

motu: add debugging output to assist in tracking down issues with the 828mk1
------------------------------------------------------------------------
r2000 | adi | 2011-10-29 23:39:24 +0200 (Sat, 29 Oct 2011) | 2 lines

Fix copy&paste error in debug message.

------------------------------------------------------------------------
r1999 | adi | 2011-10-10 09:10:52 +0200 (Mon, 10 Oct 2011) | 6 lines

Fix segfault upon termination.

Kudos to David Henningsson for the patch.

Closes: #329

------------------------------------------------------------------------
r1998 | adi | 2011-10-05 21:07:21 +0200 (Wed, 05 Oct 2011) | 2 lines

Fix spelling mistake "usefull"

------------------------------------------------------------------------
r1997 | adi | 2011-10-05 21:04:39 +0200 (Wed, 05 Oct 2011) | 2 lines

Fix spelling mistake "connnection"

------------------------------------------------------------------------
r1996 | adi | 2011-10-05 21:02:51 +0200 (Wed, 05 Oct 2011) | 2 lines

Set compiler define when using EXPAT.

------------------------------------------------------------------------
r1995 | jwoithe | 2011-09-13 02:22:20 +0200 (Tue, 13 Sep 2011) | 1 line

Ensure the IsoHandler disable lock is initialised regardless of which constructor is called.  Its omission from these was a big oversight of mine in r1985, and probably explains the ongoing problems noted by various people in issue 306 after r1985 was committed.  A big thanks to Holger Dehnhardt for lending his eyes to the problem and spotting my silly obvious error.
------------------------------------------------------------------------
r1994 | jwoithe | 2011-09-13 01:05:55 +0200 (Tue, 13 Sep 2011) | 1 line

Add configuration entry for Phonic HelixBoard 24 Universal, provided by Steffen Klein.  With this entry this device is reported by Steffen to "work flawlessly".
------------------------------------------------------------------------
r1993 | jwoithe | 2011-09-04 14:03:42 +0200 (Sun, 04 Sep 2011) | 2 lines

rme: clean up debug output by replacing "temporary" printf() calls with debugOutput().  Remove unnecessary register reads used during early stages of development.

------------------------------------------------------------------------
r1992 | jwoithe | 2011-08-18 14:29:54 +0200 (Thu, 18 Aug 2011) | 1 line

rme: removal of commented out tests which are no longer required.
------------------------------------------------------------------------
r1991 | jwoithe | 2011-08-10 07:44:17 +0200 (Wed, 10 Aug 2011) | 1 line

Add configuration info for Presonus StudioLive 16.4.2, thanks to Johan Landman
------------------------------------------------------------------------
r1990 | adi | 2011-07-28 17:24:17 +0200 (Thu, 28 Jul 2011) | 1 line

git: Ignore autogenerated files.
------------------------------------------------------------------------
r1989 | adi | 2011-07-28 17:24:08 +0200 (Thu, 28 Jul 2011) | 1 line

Enable unmute-ozonic in tests/SConscript
------------------------------------------------------------------------
r1988 | adi | 2011-07-28 17:23:59 +0200 (Thu, 28 Jul 2011) | 3 lines

Add utility to unmute M-Audio Ozonic

Patch provided by Mark Brand (orania). Re: #341
------------------------------------------------------------------------
r1987 | adi | 2011-07-28 17:23:50 +0200 (Thu, 28 Jul 2011) | 3 lines

Add mixer for M-Audio Ozonic

Patch provided by Mark Brand (orania). Re: #341
------------------------------------------------------------------------
r1986 | adi | 2011-07-28 17:23:38 +0200 (Thu, 28 Jul 2011) | 3 lines

Add configuration entry for M-Audio Ozonic

Patch provided by Mark Brand (orania). Re: #341
------------------------------------------------------------------------
r1985 | jwoithe | 2011-05-22 14:52:01 +0200 (Sun, 22 May 2011) | 8 lines

Fix double-free on exit under the new firewire stack.  It seems that with the new kernel firewire stack, raw1394_destroy_handle() can take upwards of 20 milliseconds(!) to return.  Therefore the IsoHandler's disable() call invoked by the IsoTask (FW_ISORCV or FW_ISOXMT) may not have completed before the "jackd" thread calls ~IsoHandler().  ~IsoHandler() thus infers that the handler is still running and calls disable() itself.  The practical upshot is that raw1394_destroy_handle() gets called on the same object twice, and a double-free results.

The fix I've implemented is a touch crude, but it appears to work.  A mutex is introduced to track the progress of disable(), and this is checked by ~IsoHandler() before the state of the handler is tested.  Any in-progress disable() call is allowed to complete before ~IsoHandler() tests the state.  This prevents the second call of raw1394_destroy_handle() and therefore the double-free cannot occur.

Perhaps as a result of the delays caused by raw1394_destroy_handle(), it seems the handler list can be altered by other threads while updateShadowMapHelper() (called by the IsoTask threads) is running.  A crude test has been added to this function to prevent out-of-range exceptions in most cases.

None of this is particularly elegant but it should work around the double-free issue for the moment.  The correct approach is to work out precisely why these concurrency issues are occuring and fix them.  However, given that all this will be obsoleted by the in-kernel streaming work at some point in the future, it's arguable that the solution in this patch is sufficient in practice.

------------------------------------------------------------------------
r1984 | jwoithe | 2011-05-16 12:59:13 +0200 (Mon, 16 May 2011) | 2 lines

MOTU: tweaks to streaming startup on G1 devices to see if there's a trick to getting audio out of the physical outputs.

------------------------------------------------------------------------
r1983 | jwoithe | 2011-04-17 15:30:58 +0200 (Sun, 17 Apr 2011) | 1 line

RME: big cleanup of streaming code, removing old debug tests and so on.  At least in theory the driver is now ready for wider testing.
------------------------------------------------------------------------
r1982 | jwoithe | 2011-04-16 17:04:18 +0200 (Sat, 16 Apr 2011) | 2 lines

RME: tweaks to timing synthesis to improve stability of operation at 44.1k.

------------------------------------------------------------------------
r1981 | jwoithe | 2011-04-16 16:41:26 +0200 (Sat, 16 Apr 2011) | 1 line

When calculating the difference signal in the receive DLL, allow the actual and predicted timestamps to wrap at slightly different times.  For most interfaces, the chances of hitting a problem without this allowance is extremely small since the DLL tracks the incoming timestamp tightly.  However for RME devices the difference is grossly miscalculated at every wraparound without this fix due to the way the timestamp is synthesised for these devices.
------------------------------------------------------------------------
r1980 | jwoithe | 2011-04-14 14:54:02 +0200 (Thu, 14 Apr 2011) | 2 lines

RME: make the bandwidth limit have the desired effect on the channels created.  Note that there is presently a separate issue at 2x rates which is yet to be worked through (it's associated with the full channel list we now have).

------------------------------------------------------------------------
r1979 | adi | 2011-04-08 19:07:21 +0200 (Fri, 08 Apr 2011) | 5 lines

[cosmetics] Fix compiler warning

By moving mb_head's and mb_tail's declaration further down, a compiler
warning about initialization order can be avoided.

------------------------------------------------------------------------
r1978 | adi | 2011-04-08 18:59:54 +0200 (Fri, 08 Apr 2011) | 5 lines

[cosmetics] Replace AND mask by modulo operation.

The assembler output is the same, so let's not complicate things by
using bitmasks together with negations.

------------------------------------------------------------------------
r1977 | adi | 2011-04-08 18:50:33 +0200 (Fri, 08 Apr 2011) | 6 lines

More unlikely() in midi decoding.

We don't expect any 2X/3X data to be present, hence it's utterly
unlikely to hit this branch. Let's give a hint to the compiler so he can
properly optimize.

------------------------------------------------------------------------
r1976 | adi | 2011-04-08 17:39:07 +0200 (Fri, 08 Apr 2011) | 2 lines

Remove just another leftover printf statement

------------------------------------------------------------------------
r1975 | adi | 2011-04-08 17:32:56 +0200 (Fri, 08 Apr 2011) | 2 lines

Fix leftover printf statement.

------------------------------------------------------------------------
r1974 | adi | 2011-04-08 17:31:11 +0200 (Fri, 08 Apr 2011) | 2 lines

[cosmetics] Code beautification, no semantic change.

------------------------------------------------------------------------
r1973 | adi | 2011-04-08 17:27:05 +0200 (Fri, 08 Apr 2011) | 11 lines

Fix MIDI receive on DICE (Re: #330)

Code more or less copied from MOTU. It seems MIDI bytes in the input
buffer might occur at all locations, not just 8-sample-aligned, so we'll
have to step through each incoming quadlet and see if it's MIDI.

On the jackd side, we'll have to obey the 8-sample-alignment rule.

An upcoming patch will indent the code, just for reviewing, I decided to
have no whitespace fixes.

------------------------------------------------------------------------
r1972 | jwoithe | 2011-04-05 14:48:58 +0200 (Tue, 05 Apr 2011) | 3 lines

- Parameterise selected low-level error detection thresholds.  The default values are those which have been used up to now, so no functional changes should result from this.
- RME: attempt to use the main DLL without a smoothing DLL in front of it since this may track the current timestamp better.  Broaden various lower level error thresholds of RME objects accordingly.  Testing continues.

------------------------------------------------------------------------
r1971 | adi | 2011-04-03 15:50:41 +0200 (Sun, 03 Apr 2011) | 10 lines

Set fThread to 0 instead of NULL

fThread is of type pthread_t which is unsigned long int, at least
according to /usr/include/bits/pthreadtypes.h.

This commit fixes a compiler warning:

src/libutil/PosixThread.cpp:199:19: warning: converting to non-pointer
type 'pthread_t' from NULL

------------------------------------------------------------------------
r1970 | adi | 2011-04-03 15:44:45 +0200 (Sun, 03 Apr 2011) | 10 lines

Fix compiler warning

When the default branch in the switch-case statement is taken,
receive_mode would be uninitialised.

src/libieee1394/IsoHandlerManager.cpp:754:40: warning: 'receive_mode'
may be used uninitialized in this functio

Let's at least assign something to it.

------------------------------------------------------------------------
r1969 | adi | 2011-04-03 15:41:02 +0200 (Sun, 03 Apr 2011) | 11 lines

Remove duplicate code.

When compiling, this line caused a warning:

src/bebob/focusrite/focusrite_saffirepro.cpp:1163:16: warning:
converting 'false' to pointer type

Obviously, the default branch in the switch-case statement does
everything, so the code in question looks like a leftover.

------------------------------------------------------------------------
r1968 | jwoithe | 2011-03-22 14:14:39 +0200 (Tue, 22 Mar 2011) | 1 line

RME: first pass at creating ports for all audio channels.  Copyright year addition.
------------------------------------------------------------------------
r1967 | jwoithe | 2011-03-21 14:43:06 +0200 (Mon, 21 Mar 2011) | 1 line

RME: more streaming stability fixes.  Further extended testing is needed across all sample rates, but initial indications are that the streaming system is finally approaching a state of stability.
------------------------------------------------------------------------
r1966 | jwoithe | 2011-03-19 15:53:42 +0200 (Sat, 19 Mar 2011) | 1 line

RME: more work on improving robustness of streaming.  48k and 192k seem to work reasonably reliably now.  96k can take some time to stabilise but seems good once this has occurred.  44.1k still has issues, most of which are still to be investigated.
------------------------------------------------------------------------
r1965 | jwoithe | 2011-03-11 17:32:51 +0200 (Fri, 11 Mar 2011) | 3 lines

- RME: further tweaks to the streaming system to make it more reliable (finally some progress is being made).
- Generic: reset timestamps in StreamProcessor::doDryRunning() in case the stream processor has previously been running (eg: before an xrun).  Timestamps acquired during a previous run won't apply once the processor starts again.  This change makes sense to me and helps xrun recovery on RME devices but there may be subtleties I don't see - so please double check the reasoning and behaviour with other interfaces.

------------------------------------------------------------------------
r1964 | jwoithe | 2011-03-04 14:57:40 +0200 (Fri, 04 Mar 2011) | 1 line

digidesign: fix bug identified by compile-testing the new skeleton driver
------------------------------------------------------------------------
r1963 | jwoithe | 2011-03-04 14:40:38 +0200 (Fri, 04 Mar 2011) | 1 line

digidesign: integrate into build system
------------------------------------------------------------------------
r1962 | jwoithe | 2011-03-04 14:31:36 +0200 (Fri, 04 Mar 2011) | 1 line

digidesign: add configuration options for the skeleton digidesign driver
------------------------------------------------------------------------
r1961 | jwoithe | 2011-03-04 14:29:41 +0200 (Fri, 04 Mar 2011) | 1 line

digidesign: further foundational work for the digidesign driver skeleton.  This revision adds in the streaming-related objects.  Coming next will be integration of this new driver into the build system and compile-testing.
------------------------------------------------------------------------
r1960 | arnonym | 2011-03-03 21:06:05 +0200 (Thu, 03 Mar 2011) | 2 lines

Rename the tabs of the generic dice mixer to make it a bit more clear which one provides the mixer and which is the router.

------------------------------------------------------------------------
r1959 | jwoithe | 2011-03-03 14:21:40 +0200 (Thu, 03 Mar 2011) | 1 line

digidesign: the incomplete beginnings of a skeleton driver for Digidesign interfaces.  This revision includes the base-level device object.  Skeleton port and streaming objects will follow when they're ready, as will the addition of this new driver into the build system.
------------------------------------------------------------------------
r1958 | adi | 2011-02-25 11:48:51 +0200 (Fri, 25 Feb 2011) | 4 lines

Fix garbage in transmitted CIP headers

Patch by Clemens Ladisch. TNX.

------------------------------------------------------------------------
r1957 | adi | 2011-02-22 11:32:41 +0200 (Tue, 22 Feb 2011) | 4 lines

Remove 'U' from Onyx configuration entry

According to Holger, 'U' is causing a parse error.

------------------------------------------------------------------------
r1956 | adi | 2011-02-16 23:23:09 +0200 (Wed, 16 Feb 2011) | 5 lines

Fix Onyx-i series.

It was broken after r1945 (and maybe also r1950), but according to
Holger Dehnhardt, overriding the transfer delay make it work again.

------------------------------------------------------------------------
r1955 | adi | 2011-02-16 11:11:42 +0200 (Wed, 16 Feb 2011) | 4 lines

Make missing EAP a warning instead of an error.

Closes: #322

------------------------------------------------------------------------
r1954 | jwoithe | 2011-02-11 13:05:32 +0200 (Fri, 11 Feb 2011) | 2 lines

Add configuration entry for the TC Electronic Studio Konnekt 48 device.  Thanks to "Juanramon" for the information.

------------------------------------------------------------------------
r1953 | jwoithe | 2011-02-10 16:33:51 +0200 (Thu, 10 Feb 2011) | 2 lines

RME: more testing to get reliable timestamp synthesis.

------------------------------------------------------------------------
r1952 | adi | 2011-02-10 14:06:06 +0200 (Thu, 10 Feb 2011) | 8 lines

Allow Echo Audiofire Pre8 to work above 48kHz

Patch provided by Euan in ticket #310.

Confirmed to work on my Alesis io14 (DICE). If it breaks other devices,
we might need to revert the change.

------------------------------------------------------------------------
r1951 | arnonym | 2011-02-03 00:51:32 +0200 (Thu, 03 Feb 2011) | 2 lines

Print a message when the dbus service file can not be installed...

------------------------------------------------------------------------
r1950 | adi | 2011-01-27 11:22:25 +0200 (Thu, 27 Jan 2011) | 4 lines

Fix AMDTP_TRANSMIT_TRANSFER_DELAY

As pointed out bx Clemens Ladisch in #321, we had 125us too much.

------------------------------------------------------------------------
r1949 | adi | 2011-01-23 11:53:26 +0200 (Sun, 23 Jan 2011) | 4 lines

Fix MIDI channel detection on Echo AF series

Patch provided in #233

------------------------------------------------------------------------
r1948 | adi | 2011-01-16 19:01:31 +0200 (Sun, 16 Jan 2011) | 9 lines

Fix ListDevices on Juju stack

Patch by Stefan Richter. He writes:

The fix is straightforward:  Let ListDevices continue at the next node ID       
when the configuration ROM of a lower node could not be fetched or parsed       
successfully.

------------------------------------------------------------------------
r1947 | adi | 2011-01-14 17:53:13 +0200 (Fri, 14 Jan 2011) | 2 lines

Fix typo.

------------------------------------------------------------------------
r1946 | adi | 2011-01-14 16:28:09 +0200 (Fri, 14 Jan 2011) | 7 lines

Finally enable Mackie Onyx i series.

The previous commit did all the infrastructure work, this is a small
quirk required only for the Mackie. Since I don't have devices at hand
to test, this commit might well break other things. If so, revert it and
special case for onyx.

------------------------------------------------------------------------
r1945 | adi | 2011-01-14 16:17:19 +0200 (Fri, 14 Jan 2011) | 8 lines

Infrastructure for Mackie Onyx i.

Original patch provided by Holger Dehnhardt.

Note that this commit alone isn't sufficient to make the Onyx work. The
next commit to avc_plug.cpp will be required, but I wanted to separate
commits just in case the next one breaks something.

------------------------------------------------------------------------
r1944 | adi | 2010-12-31 19:02:29 +0200 (Fri, 31 Dec 2010) | 7 lines

Add Russian translation to desktop file

Thanks to prokoudine for providing the updated file

Re: #316
Closes: #316

------------------------------------------------------------------------
r1943 | jwoithe | 2010-12-20 15:39:11 +0200 (Mon, 20 Dec 2010) | 2 lines

RME: fix for lengthy startup delay which prevented reliable startup in approximately 20% attempts.  This may not be the approach adopted long-term since there seems to be something amiss with the way iso handlers are started in general (in short, the start cycle - m_switch_on_cycle in IsoHandler - is never set to anything other than zero).

------------------------------------------------------------------------
r1942 | arnonym | 2010-12-14 23:32:48 +0200 (Tue, 14 Dec 2010) | 2 lines

Add preliminary support for the saffire pro14. The extended mixer is not enabled yet, that needs work inside libffado. But the generic dice-stuff should just work. Of course it will show a to high number of channels in the router...

------------------------------------------------------------------------
r1941 | jwoithe | 2010-12-13 16:31:30 +0200 (Mon, 13 Dec 2010) | 2 lines

RME: more debugging of streaming system.

------------------------------------------------------------------------
r1940 | jwoithe | 2010-12-10 14:40:57 +0200 (Fri, 10 Dec 2010) | 2 lines

RME: incidental changes for further debugging

------------------------------------------------------------------------
r1939 | adi | 2010-12-07 18:32:38 +0200 (Tue, 07 Dec 2010) | 3 lines

Refactor startStreamByIndex and stopStreamByIndex.

------------------------------------------------------------------------
r1938 | adi | 2010-12-07 18:17:21 +0200 (Tue, 07 Dec 2010) | 2 lines

Refactor stopStreamByIndex.

------------------------------------------------------------------------
r1937 | adi | 2010-12-07 17:45:07 +0200 (Tue, 07 Dec 2010) | 13 lines

Heavy refactoring.

I moved writeFunc and readFunc to .h and merged the two for-loops in
prepare(). This code is actually now in prepareSP(), prepare() only
contains the calling loops.

Note that in contrast to the previous code, I conditionally set dll_bw
depending whether it's RX or TX. I guess it was a copy-and-paste error
before. (I'm talking about line 1001 in r1936)

Works for my Alesis device.

------------------------------------------------------------------------
r1936 | adi | 2010-12-07 15:29:51 +0200 (Tue, 07 Dec 2010) | 10 lines

Refactor startStreamByIndex.

A lot of code in dice_avdevice.cpp violates the DRY principle (don't
repeat yourself), thus making development unnecessary hard and
error-prone.

This is a first attempt to improve the situation. Clearly, there are
more parts to be refactored.

------------------------------------------------------------------------
r1935 | adi | 2010-12-07 13:50:47 +0200 (Tue, 07 Dec 2010) | 4 lines

Remove obsolete vendor-specific hacks.

The code wasn't active for a while, it's not necessary anymore.

------------------------------------------------------------------------
r1934 | adi | 2010-12-07 13:45:52 +0200 (Tue, 07 Dec 2010) | 4 lines

Shebang-line cleanup. Closes: #292

Patch provided by Orcan Ogetbil (Fedora).

------------------------------------------------------------------------
r1933 | adi | 2010-12-07 13:37:36 +0200 (Tue, 07 Dec 2010) | 2 lines

Use gcc builtin atomic functions. Closes #280.

------------------------------------------------------------------------
r1930 | jwoithe | 2010-12-05 08:46:34 +0200 (Sun, 05 Dec 2010) | 2 lines

Quieten a false warning from Valgrind.

------------------------------------------------------------------------
r1928 | jwoithe | 2010-12-04 14:24:26 +0200 (Sat, 04 Dec 2010) | 2 lines

ffado.h: remove references to old ffado_streaming_{read,write}() functions as reported by Adrian Knoth.  Partial fix to ticket:313.  I'm not certain I've got all the details correct - others may wish to check.

------------------------------------------------------------------------
r1927 | jwoithe | 2010-12-04 14:13:26 +0200 (Sat, 04 Dec 2010) | 2 lines

Remove old, obsolete code from teststreaming3.cpp which may confuse new users of libffado through references to removed functions such as ffado_streaming_{read,write}().  Partial fix to ticket:313.

------------------------------------------------------------------------
r1926 | jwoithe | 2010-12-04 13:54:55 +0200 (Sat, 04 Dec 2010) | 2 lines

Fix error description from pthread_setschedparam().  Reported and patch supplied by Adrian Knoth in ticket:312.

------------------------------------------------------------------------
r1925 | jwoithe | 2010-12-03 14:37:41 +0200 (Fri, 03 Dec 2010) | 3 lines

RME: more experiments to avoid warnings from the main ffado timestamp DLL by smoothing out the synthetic timestamps generated from the packet arrival cycle timer value.
RME: fix an uninitialised variable.

------------------------------------------------------------------------
r1924 | jwoithe | 2010-11-20 13:24:11 +0200 (Sat, 20 Nov 2010) | 2 lines

MOTU: 828mk1: implement a first cut at streaming control for the 828mk1 based on details supplied by Martin Peach.  This has not been tested on an 828mk1 yet.

------------------------------------------------------------------------
r1923 | jwoithe | 2010-11-20 12:52:54 +0200 (Sat, 20 Nov 2010) | 2 lines

MOTU: 828mk1: update optical port mode control based on more recent dumps provided by Martin Peach.  These changes are yet to be tested with an 828mk1.

------------------------------------------------------------------------
r1922 | arnonym | 2010-10-28 20:18:21 +0200 (Thu, 28 Oct 2010) | 2 lines

Use lscpu if present, de-activate the rt-check which wasn't implemented. Change the wording if the new stack is loaded.

------------------------------------------------------------------------
r1921 | jwoithe | 2010-10-24 07:35:47 +0200 (Sun, 24 Oct 2010) | 2 lines

MOTU: 896-mk3: For cases where optical output is off, fix playback channel count at 2x and correctly label a few others.  Thanks to Max Tucci for running the tests and experiments to make this possible.

------------------------------------------------------------------------
r1920 | jwoithe | 2010-10-23 11:40:36 +0200 (Sat, 23 Oct 2010) | 2 lines

MOTU: 896-mk3: we're still sending the wrong number of playback channels at 2x rates.  Add another 2 to see what happens.

------------------------------------------------------------------------
r1919 | jwoithe | 2010-10-22 15:42:30 +0200 (Fri, 22 Oct 2010) | 2 lines

MOTU: 896mk3: without optical ports enabled, 1x and 4x playback is now ok.  2x still gives trouble though, so add another two channels to that case to see if that fixes it.

------------------------------------------------------------------------
r1918 | jwoithe | 2010-10-22 15:32:45 +0200 (Fri, 22 Oct 2010) | 2 lines

RME: refinement of debugging output

------------------------------------------------------------------------
r1917 | jwoithe | 2010-10-22 15:10:46 +0200 (Fri, 22 Oct 2010) | 2 lines

RME: fix a typo, although it's doubtful it had any practical impact on anything.

------------------------------------------------------------------------
r1916 | jwoithe | 2010-10-22 01:38:40 +0200 (Fri, 22 Oct 2010) | 2 lines

MOTU: 896mk3: add two playback channels for 2x and 4x rates since the device expects a different number than what we currently send.

------------------------------------------------------------------------
r1915 | jwoithe | 2010-10-20 13:43:37 +0200 (Wed, 20 Oct 2010) | 2 lines

RME: streamline debug messages.  Try to make startup more reliable (testing is ongoing in this regard).

------------------------------------------------------------------------
r1914 | jwoithe | 2010-10-19 15:03:03 +0200 (Tue, 19 Oct 2010) | 2 lines

MOTU: tweak 896mk3 channel layout for 2x and 4x rates to try to get things running with optical ports disabled.  Also note that r1913 applies to 896mk3 devices too.

------------------------------------------------------------------------
r1913 | jwoithe | 2010-10-14 02:41:19 +0200 (Thu, 14 Oct 2010) | 2 lines

MOTU: tweak to 1x output packets to explore the format of these packets.

------------------------------------------------------------------------
r1912 | jwoithe | 2010-10-11 01:34:47 +0200 (Mon, 11 Oct 2010) | 2 lines

MOTU: fix the setting of 48k sampling rate for the 828 Mk 1.  Thanks to Martin Peach for pointing out this failure.

------------------------------------------------------------------------
r1911 | jwoithe | 2010-10-11 00:58:32 +0200 (Mon, 11 Oct 2010) | 2 lines

MOTU: separately track packet offsets of input and output ports in the new portgroup code.  This should restore the creation of the correct input port names for all MOTU devices.

------------------------------------------------------------------------
r1910 | arnonym | 2010-10-10 19:39:07 +0200 (Sun, 10 Oct 2010) | 2 lines

Fix #308: Do the sample-rate changing the way the echo devices need them. I can't test this due to a lack of echo devices, if this breaks anything (specially the non pre8-devices), please re-open the ticket #308.

------------------------------------------------------------------------
r1909 | jwoithe | 2010-10-10 13:55:14 +0200 (Sun, 10 Oct 2010) | 2 lines

MOTU: a tweak to the 896mk3 channel list for 1x rates; it seems there's 4 "unknown" channels sent by the device at present, not just two.

------------------------------------------------------------------------
r1908 | jwoithe | 2010-10-09 12:25:53 +0200 (Sat, 09 Oct 2010) | 2 lines

Always read the ConfigROM using quadlet reads, since some devices advertise larger supported read requests but reject anything other than quadlet reads.  This is known to affect Echo Audiofire pre8 devices and there may be others too.

------------------------------------------------------------------------
r1907 | jwoithe | 2010-10-08 14:54:53 +0200 (Fri, 08 Oct 2010) | 2 lines

MOTU: a first cut at defining the audio port layout for the 896mk3.  This is presently nothing more than an educated guess and is extremely unlikely to work.  Testing is welcomed though.

------------------------------------------------------------------------
r1906 | jwoithe | 2010-10-07 14:59:51 +0200 (Thu, 07 Oct 2010) | 2 lines

MOTU: allow FFADO to detect (but not use) the 896mk3.  Port details will follow in time.

------------------------------------------------------------------------
r1905 | jwoithe | 2010-09-30 14:24:17 +0200 (Thu, 30 Sep 2010) | 2 lines

MOTU: switch to using the new portgroups functionality by default.  Please test.  If there's a problem one can revert to the old code for the moment by commenting out the "USE_PORTGROUPS" #define in line 63 of motu_avdevice.cpp.

------------------------------------------------------------------------
r1904 | jwoithe | 2010-09-23 14:27:10 +0200 (Thu, 23 Sep 2010) | 2 lines

MOTU: step 2 of the process started in r1903.  If the new port groups functionality is activated the groups are only scanned once during the prepare() call via initDirPortGroups().  Calls to getEventSize() and addDirPortGroups() then use this information without having to scan the port list themselves.  This means the scanning and flag testing code is no longer duplicated.

------------------------------------------------------------------------
r1903 | jwoithe | 2010-09-20 14:52:20 +0200 (Mon, 20 Sep 2010) | 2 lines

MOTU: step 1 in a process to eliminate code duplication presently seen in addDirPorts()/addDirPortGroups() and getEventSize().  The new code is not active yet and is still a work in progress.

------------------------------------------------------------------------
r1902 | jwoithe | 2010-09-19 14:00:27 +0200 (Sun, 19 Sep 2010) | 3 lines

MOTU: Create a define (USE_PORTGROUPS) which if defined will switch to the new PortGroups methods of identifying audio streams in data packets.  This is currently not defined, meaning the old PortEntry method continues to be used by default.
MOTU: Make MotuDevice::getEventSize aware of the PortGroups mechanism when necessary.

------------------------------------------------------------------------
r1901 | arnonym | 2010-09-16 23:29:41 +0200 (Thu, 16 Sep 2010) | 4 lines

Lets add an experimental mixer for presonus firepod/firebox for the time being. On my firepod the volume-faders ''work'' but no audible change happens as it seems. But maybe this helps firebox users to get the device unmuted (see #191).

Note that this mixer might also work on other bebob-based devices. But for that I need to spend some more time with it...

------------------------------------------------------------------------
r1900 | arnonym | 2010-09-16 22:15:35 +0200 (Thu, 16 Sep 2010) | 2 lines

Apply adi's patch to re-use streams (on dice-based devices) if jack/ffado crashed and is re-started. see #287

------------------------------------------------------------------------
r1899 | arnonym | 2010-09-16 21:38:55 +0200 (Thu, 16 Sep 2010) | 2 lines

Take a shot at fixing #233: I think it is okay to rely on streamformat==-1 (==0xff) to be safe for replacement. According to avc_extended_stream_format.h 0xff is AVC1394_STREAM_FORMAT_AM824_DONT_CARE.

------------------------------------------------------------------------
r1898 | jwoithe | 2010-09-14 15:50:25 +0200 (Tue, 14 Sep 2010) | 2 lines

MOTU: protocol documentation update - added summary of audio channel location within frames of iso data packets.  This has been documented mainly because the explicit coding of locations is going away with the new portgroup mechanism being introduced.  Preserving the knowledge of the location of the channels may be helpful when debugging issues in the future.

------------------------------------------------------------------------
r1897 | arnonym | 2010-09-12 22:27:15 +0200 (Sun, 12 Sep 2010) | 2 lines

Fix #214: Replace ''Options'' by ''Variables'' and thus remove the deprecation warnings. I think it is save to do this now. Packagers for distros where scons is to old "simply" need to apply the reverse of this patch. but I don't think there are that many distros/distro-versions using that old scons and still maintaining current copies of ffado...

------------------------------------------------------------------------
r1896 | arnonym | 2010-09-12 22:15:27 +0200 (Sun, 12 Sep 2010) | 2 lines

Replace os.popen2 by subprocess. Removes two ugly warnings.

------------------------------------------------------------------------
r1895 | arnonym | 2010-09-12 22:08:32 +0200 (Sun, 12 Sep 2010) | 2 lines

Simplify the check for the external dbus-stuff. And add the DBus-specific flags to one of the dbus-variables. For building libffado.so, this flag is not needed...

------------------------------------------------------------------------
r1894 | arnonym | 2010-09-12 21:48:12 +0200 (Sun, 12 Sep 2010) | 2 lines

Fix #295: Aid debugging by naming the threads.

------------------------------------------------------------------------
r1893 | arnonym | 2010-09-12 21:34:52 +0200 (Sun, 12 Sep 2010) | 2 lines

Take a shot at "fixing" #285. Does this make the profire2625 work? Is the mixer working too? (Hard to tell without the hardware...)

------------------------------------------------------------------------
r1892 | jwoithe | 2010-09-12 13:15:49 +0200 (Sun, 12 Sep 2010) | 3 lines

MOTU: make specification of port groups in the DevicesProperty[] array somewhat neater with an intermediate macro.
MOTU: add port groups to the DevicesProperty[] for all devices not previously done.  Note that the old PortEntry specifications are still the ones used by default.

------------------------------------------------------------------------
r1891 | jwoithe | 2010-09-11 15:38:09 +0200 (Sat, 11 Sep 2010) | 3 lines

MOTU: add port_num_offset to the port group entry structure.  This allows a group of channels to start their numbering at a number other than 1 (eg: Analog 3-8 instead of 1-6).  This is needed for some MOTU devices which have mic 1-2 and analog 3-8.
MOTU: construct new port group tables for all interfaces not done previously.  Note that these new tables are not yet used by ffado (testing/verification continues).

------------------------------------------------------------------------
r1890 | jwoithe | 2010-09-11 14:49:25 +0200 (Sat, 11 Sep 2010) | 2 lines

MOTU: fix a typo which made the new port configuration system think there were more groups than there really were.

------------------------------------------------------------------------
r1889 | arnonym | 2010-09-11 00:02:30 +0200 (Sat, 11 Sep 2010) | 4 lines

Make the desktop-file nicer (and probably translatable).

Thanks for the patch! Fixes #291.

------------------------------------------------------------------------
r1888 | arnonym | 2010-09-10 23:50:58 +0200 (Fri, 10 Sep 2010) | 4 lines

Drop the libs in external/ and use the system-libs for that. Easier on administration, easier on compilation time, easier for packagers.

This is basically the patches provided by packagers in #290.

------------------------------------------------------------------------
r1887 | arnonym | 2010-09-10 18:51:49 +0200 (Fri, 10 Sep 2010) | 2 lines

Fix #303 and #298: Using tool-buttons saves space. The normal QPushButton has more space left and right of the text...

------------------------------------------------------------------------
r1886 | jwoithe | 2010-09-10 15:59:10 +0200 (Fri, 10 Sep 2010) | 3 lines

MOTU: implement a new way to specify the audio ports present in packets and how this varies with device state.  Instead of specifying packet offsets statically these are now computed dynamically based on device state and the order of the new PortGroupEntry tables.  Example tables for the Traveler and 828mk3 have been written.  This new code is not yet used by ffado because it is still yet to undergo thorough testing.  Once that's done all devices will be converted to use the new method.
The advantages of this new approach are clear by looking at the 828mk3 table: its size reduces from 150 entries down to a much more manageable 17.

------------------------------------------------------------------------
r1885 | arnonym | 2010-09-05 21:53:22 +0200 (Sun, 05 Sep 2010) | 2 lines

Try to save some space in the mixer of the profire610.

------------------------------------------------------------------------
r1884 | arnonym | 2010-09-05 20:13:04 +0200 (Sun, 05 Sep 2010) | 2 lines

Add the generic mixer entry for the ProFire 610.

------------------------------------------------------------------------
r1883 | jwoithe | 2010-09-03 14:33:30 +0200 (Fri, 03 Sep 2010) | 3 lines

RME: when initialising the device, create a 1:1 mapping from playback channels to physical outputs.  This means audio sent to the device will emerge from the respective outputs.
RME: as for MOTU, only create the 1 kHz test tone (in Phones-L here) if a non-zero debug level is specified.

------------------------------------------------------------------------
r1882 | jwoithe | 2010-08-26 14:27:41 +0200 (Thu, 26 Aug 2010) | 2 lines

MOTU: correctly read the sample rate from G3 devices.

------------------------------------------------------------------------
r1881 | jwoithe | 2010-08-12 13:36:34 +0200 (Thu, 12 Aug 2010) | 2 lines

MOTU: include IDs for the Traveler Mk3 and 896HD Mk3 in the mixer and count these as generation-3 devices.

------------------------------------------------------------------------
r1880 | jwoithe | 2010-08-06 13:43:39 +0200 (Fri, 06 Aug 2010) | 2 lines

MOTU: The Traveler Mk3 has two optical ports, so the port active flags must reflect this when masking allowable optical port modes.  Up to now Traveler Mk3 devices have failed to create any audio ports at all; this change may fix that.

------------------------------------------------------------------------
r1879 | jwoithe | 2010-08-06 13:35:50 +0200 (Fri, 06 Aug 2010) | 2 lines

MOTU: fix a typo affecting the correct detection of G3 optical port modes.

------------------------------------------------------------------------
r1878 | jwoithe | 2010-08-05 16:35:41 +0200 (Thu, 05 Aug 2010) | 2 lines

MOTU: add some debugging output to assist with Traveler Mk3 streaming format testing.

------------------------------------------------------------------------
r1877 | jwoithe | 2010-08-05 03:24:30 +0200 (Thu, 05 Aug 2010) | 2 lines

MOTU: revert r1876.  The ultralite hybrid is meant to have 14 outputs at 1x rates.  Call r1876 a brain fade.

------------------------------------------------------------------------
r1876 | jwoithe | 2010-08-05 03:11:03 +0200 (Thu, 05 Aug 2010) | 2 lines

MOTU: tweaks to the ultralite-hybrid audio port map

------------------------------------------------------------------------
r1875 | jwoithe | 2010-08-05 02:20:23 +0200 (Thu, 05 Aug 2010) | 2 lines

MOTU: some refining of Traveler Mk3 audio port details.  Optical port info is to follow later.

------------------------------------------------------------------------
r1874 | jwoithe | 2010-08-02 13:14:44 +0200 (Mon, 02 Aug 2010) | 2 lines

MOTU: define a separate port map for the Ultralite hybrid based on that of the ultralite mk3.  Make some changes to the input direction based on observations reported by another hybrid user.

------------------------------------------------------------------------
r1873 | jwoithe | 2010-08-01 13:51:40 +0200 (Sun, 01 Aug 2010) | 2 lines

MOTU: fix a bug in the MOTU mixer which caused ffadomixer to crash out with a python error.  The bug was added in r1840 - sorry about that.  Thanks to Luke Yelavich for posting a debug trace which unearthed this bug.

------------------------------------------------------------------------
r1872 | jwoithe | 2010-07-29 15:30:31 +0200 (Thu, 29 Jul 2010) | 2 lines

RME: trivial addition to existing test code.  Nothing to see here, move along please.

------------------------------------------------------------------------
r1871 | jwoithe | 2010-07-19 15:50:37 +0200 (Mon, 19 Jul 2010) | 2 lines

RME: put test tone onto all channels in case ffado has the wrong idea about channel ordering.  It made no difference, so the test code has been commented out for now, leaving the test tone theoretically on Phones-L only.

------------------------------------------------------------------------
r1870 | jwoithe | 2010-07-19 15:42:58 +0200 (Mon, 19 Jul 2010) | 2 lines

RME: further playback streaming tests.  Device outputs are still silent.

------------------------------------------------------------------------
r1869 | jwoithe | 2010-07-18 16:42:58 +0200 (Sun, 18 Jul 2010) | 3 lines

RME: more receive streaming changes.  The resulting code has enabled the first successful audio capture from an RME firewire interface to Linux (test was done using FF400).
RME: work towards getting playback working (no success yet - just silence).

------------------------------------------------------------------------
r1868 | jwoithe | 2010-07-15 16:20:13 +0200 (Thu, 15 Jul 2010) | 3 lines

RME: more work to stabilise streaming startup.
RME: delete the tx/rx stream processors when the RME device object is destroyed (fixes several iso handler related issues at shutdown).

------------------------------------------------------------------------
r1867 | jwoithe | 2010-07-12 01:36:26 +0200 (Mon, 12 Jul 2010) | 2 lines

Address a potential false error report from IsoHandlerManager::stopHandlers() when called from the IsoHandlerManager destructor after a user shutdown.

------------------------------------------------------------------------
r1866 | jwoithe | 2010-07-10 15:42:33 +0200 (Sat, 10 Jul 2010) | 2 lines

RME: yet more streaming experiments.

------------------------------------------------------------------------
r1865 | jwoithe | 2010-07-09 15:02:44 +0200 (Fri, 09 Jul 2010) | 2 lines

RME: more experimental work with the streaming system.

------------------------------------------------------------------------
r1864 | arnonym | 2010-07-06 21:43:42 +0200 (Tue, 06 Jul 2010) | 2 lines

Add the IDs of the Weiss devices.

------------------------------------------------------------------------
r1863 | jwoithe | 2010-07-01 16:21:28 +0200 (Thu, 01 Jul 2010) | 2 lines

MOTU: don't print receive stream debugging dump unless a non-zero debug level is in use.  This keeps output clean for production use while still making it relatively easy to get at the data dump when necessary.

------------------------------------------------------------------------
r1862 | jwoithe | 2010-07-01 14:55:52 +0200 (Thu, 01 Jul 2010) | 2 lines

test-volume and test-enhanced-mixer depend on the generic AVC support, so only attempt to compile them if generic AVC support is configured.

------------------------------------------------------------------------
r1861 | jwoithe | 2010-06-29 02:17:11 +0200 (Tue, 29 Jun 2010) | 2 lines

MOTU: include the Traveler-mk3 and 896HD-mk3 as generation-3 devices.

------------------------------------------------------------------------
r1860 | jwoithe | 2010-06-29 02:14:53 +0200 (Tue, 29 Jun 2010) | 2 lines

MOTU: add ID information for the MOTU Traveler mk3.  Thanks to Bo-Erik Sandholm for providing the information.  Note that the Traveler Mk3 will almost certainly not work yet: this is the first of many steps along that road.

------------------------------------------------------------------------
r1859 | arnonym | 2010-06-19 20:34:16 +0200 (Sat, 19 Jun 2010) | 2 lines

Result of several days of searching a bug (that turned out to be a pebkac): some modifications of the debug-strings.

------------------------------------------------------------------------
r1858 | arnonym | 2010-06-19 20:31:27 +0200 (Sat, 19 Jun 2010) | 2 lines

Less warnings.

------------------------------------------------------------------------
r1857 | jwoithe | 2010-06-18 15:50:55 +0200 (Fri, 18 Jun 2010) | 2 lines

RME: more streaming-related tests

------------------------------------------------------------------------
r1856 | jwoithe | 2010-06-14 14:50:38 +0200 (Mon, 14 Jun 2010) | 2 lines

RME: more assistance for initial debugging

------------------------------------------------------------------------
r1853 | arnonym | 2010-06-13 16:34:09 +0200 (Sun, 13 Jun 2010) | 4 lines

Make it easier for distributions to define their own additional built-flags. (Importing all the various shell-variables is a pain in the a** with scons as the gcc-tools of scons overwrite half of them.)

See #278

------------------------------------------------------------------------
r1851 | jwoithe | 2010-06-13 15:15:50 +0200 (Sun, 13 Jun 2010) | 2 lines

RME: accept attempts to change the hardware streaming state to the state already active.  This makes some exit paths happier.

------------------------------------------------------------------------
r1850 | jwoithe | 2010-06-12 15:53:01 +0200 (Sat, 12 Jun 2010) | 2 lines

RME: inform the IsoHandlerManager associated with the device that missed packets are acceptable for this interface.  This prevents the IsoHandlerManager from attempting to correct for something which isn't an issue with these interfaces.

------------------------------------------------------------------------
r1849 | jwoithe | 2010-06-12 15:30:33 +0200 (Sat, 12 Jun 2010) | 2 lines

IsoHandlerManager: some interfaces (eg: RME) do not send placeholder iso packets when there is no data to send.  For such interfaces IsoHandler wrongly interprets this as a dropped packet.  Therefore provide a mechanism to disable the dropped packet check.  The default state is unchanged from the original code: a missing packet is regarded as a dropped packet.

------------------------------------------------------------------------
r1848 | jwoithe | 2010-06-11 15:22:02 +0200 (Fri, 11 Jun 2010) | 3 lines

RME: nothing exciting - just some additional code to facilitate development of the streaming system.
RME: minor documentation updates.

------------------------------------------------------------------------
r1847 | jwoithe | 2010-06-07 16:37:38 +0200 (Mon, 07 Jun 2010) | 2 lines

RME: fix some streaming setup errors.  Add debug output to assist with further development of the streaming interface.

------------------------------------------------------------------------
r1845 | jwoithe | 2010-06-02 03:11:42 +0200 (Wed, 02 Jun 2010) | 2 lines

MOTU: The bits in the rate control register used to control the sample rate have changed in G3 devices, so allow for this.

------------------------------------------------------------------------
r1844 | jwoithe | 2010-05-31 14:30:56 +0200 (Mon, 31 May 2010) | 2 lines

MOTU: add some debugging output to gain some insights into the streaming data format used by the Ultralite mk3 hybrid.

------------------------------------------------------------------------
r1843 | jwoithe | 2010-05-31 14:11:42 +0200 (Mon, 31 May 2010) | 2 lines

MOTU: some G3 devices don't have any optical ports, so allow for this.

------------------------------------------------------------------------
r1842 | jwoithe | 2010-05-31 13:32:34 +0200 (Mon, 31 May 2010) | 2 lines

MOTU: tidy up clock source availability, especially for G3 devices.

------------------------------------------------------------------------
r1841 | jwoithe | 2010-05-28 04:15:26 +0200 (Fri, 28 May 2010) | 2 lines

MOTU: ultralite mk3: logically the pad channels should only be present when SPDIF is available (ie: at 1x and 2x rates).  Whether this is true in practice remains to be seen.

------------------------------------------------------------------------
r1840 | jwoithe | 2010-05-28 04:08:11 +0200 (Fri, 28 May 2010) | 3 lines

MOTU: add detection of the Ultralite Mk3 hybrid device (thanks to Tomas R Cantu for providing the ID information for this device).
MOTU: ffado-mixer: don't attempt to support generation-3 devices yet.  Log a debug message to this effect.

------------------------------------------------------------------------
r1839 | jwoithe | 2010-05-26 01:34:16 +0200 (Wed, 26 May 2010) | 2 lines

MOTU: 8pre: the 8pre data packets don't change format with rate changes (all channels are always available), so the padding needed on the output packets at 1x rate will almost certainly be required at 2x rates as well.

------------------------------------------------------------------------
r1838 | jwoithe | 2010-05-25 16:30:36 +0200 (Tue, 25 May 2010) | 2 lines

MOTU: 8pre: oops, the extra pad fields in the audio data packets added in r1837 should have been in the output (playback) direction, not the input.

------------------------------------------------------------------------
r1837 | jwoithe | 2010-05-24 14:53:00 +0200 (Mon, 24 May 2010) | 4 lines

MOTU: 8pre: disable the phones / main fader mixer widgets because these controls are not software controllable on the 8pre and thus aren't accessible via the dbus interface.
MOTU: 8pre: add pad controls to the channel matrix mixers so the ADAT controls show up at the same place in the matrix mixers as they do on other devices.  See also r1384 where a similar issue was addressed for the Ultralite.
MOTU: 8pre: add some pad channels to enlarge the 1x rate data packets to the sizes seen under other systems.

------------------------------------------------------------------------
r1836 | jwoithe | 2010-05-21 17:51:45 +0200 (Fri, 21 May 2010) | 2 lines

MOTU: as far as I can tell the 8pre doesn't have software-controlled boost/level controls over the analog inputs, so ffado doesn't export them via dbus.  Make sure the mixer doesn't request them.

------------------------------------------------------------------------
r1835 | jwoithe | 2010-05-21 17:07:01 +0200 (Fri, 21 May 2010) | 7 lines

MOTU:
 - Implement optical mode control for G3/Mk3 devices.  This is currently untested.
 - Fix bug from r1834 which broke streaming for all non-G3 MOTU devices (the wrong MOTU_PA_MK3_OPT_* value was used to force the port B test in addDirPorts() and getEventSize() to True for these devices).
 - Allow MOTU registers to be optionally absolute rather than relative to a notional base address.  This is needed to accommodate some G3 registers which do not share the same base address as traditional MOTU registers.
 - Update G3 protocol document to include input and output control of optical port modes.
 - Note that streaming for non-G3 devices was broken by r1831 (optical_mode was used uninitialised to set flags).  This was implicitly fixed by changes in r1834, but as noted above r1834 introduced a new bug which kept G1/G2 streaming broken until now.

------------------------------------------------------------------------
r1834 | jwoithe | 2010-05-17 16:39:19 +0200 (Mon, 17 May 2010) | 6 lines

MOTU: rename the ADAT channels on the 828mk3 to be more consistent with the hardware naming convention.
MOTU: if present on the interface, utilise the modes of the two independent ADAT interfaces when configuring such things as audio channels (some "mark 3" devices have 2 ADAT interfaces).  Note that getting/setting the optical modes on Mark3/G3 devices is yet to be coded.
MOTU: attempt to be more complete with the 828mk3 channel specification with respect to optical port modes.
MOTU: streaming to/from the 828Mk3 device seems to be working at 1x rates if the device is configured with both ADAT interfaces set for ADAT input/output.
MOTU: document toslink/adat channel assignment within the "mark 3" data stream.

------------------------------------------------------------------------
r1833 | jwoithe | 2010-05-17 13:42:30 +0200 (Mon, 17 May 2010) | 2 lines

MOTU: initial implementation of clock source control for "Mark 3" devices.

------------------------------------------------------------------------
r1832 | jwoithe | 2010-05-16 14:07:57 +0200 (Sun, 16 May 2010) | 4 lines

MOTU: decouple clock source identifiers from the values sent to devices.  This will make it much easier to add support for new sources added in the "mark 3" devices.
MOTU: the active clock source should be correctly retrieved from original 828 interfaces now.
MOTU: update "mark 3" protocol details, correcting errors and adding SMTPE clock source information.

------------------------------------------------------------------------
r1831 | jwoithe | 2010-05-11 15:38:27 +0200 (Tue, 11 May 2010) | 2 lines

MOTU: make optical mode control API easier to use when support for devices with multiple optical ports is added.

------------------------------------------------------------------------
r1830 | arnonym | 2010-05-10 21:34:07 +0200 (Mon, 10 May 2010) | 2 lines

Fix the check for the c-compiler. The old one produced a warning actually...

------------------------------------------------------------------------
r1828 | jwoithe | 2010-05-10 16:44:28 +0200 (Mon, 10 May 2010) | 10 lines

MOTU: begin expansion of optical mode functions needed to support the 828Mk3:
 - expand optical port mode control function interface to allow for the two ports on some of the G3 devices (eg: the 828Mk3).  Actual G3 support code is yet to be written.
 - consolidate optical port mode code to device methods and have control objects call these.
 - optical port mode identifiers passed to optical mode functions are now decoupled from the values sent to (G2) device registers making them truly device independent.
 - improved readability of the G2 optical mode code.
MOTU: fix retrieval of optical port mode for "generation 1" devices (ie: the original 828).
MOTU: added clock source defines for the G3 devices.
MOTU: added function to return device generation.
MOTU: minor Mark3 documentation fixes.

------------------------------------------------------------------------
r1827 | arnonym | 2010-05-03 23:07:53 +0200 (Mon, 03 May 2010) | 4 lines

Replace usage of QStyledItemDelegate by QItemDelegate, at least for now. The later is available on Qt4.3, the styled version only since Qt4.4. And ubuntu hardy only ships 4.3.

Should fix #276.

------------------------------------------------------------------------
r1824 | arnonym | 2010-05-03 22:01:01 +0200 (Mon, 03 May 2010) | 4 lines

Apply the patch from #255 (also from #278) to build on gcc4.5.

Tested here with gcc4.4 where it doesn't break:-)

------------------------------------------------------------------------
r1822 | arnonym | 2010-05-03 21:49:58 +0200 (Mon, 03 May 2010) | 2 lines

Fix #279 (part one) by activating the hand-coded atomic operations for ppc.

------------------------------------------------------------------------
r1821 | arnonym | 2010-05-03 21:37:15 +0200 (Mon, 03 May 2010) | 2 lines

Some layouting (spaces between the words and the dots). And fix #277 by checking for pyuic4 to see if a Qt4 version of pyqt is installed.

------------------------------------------------------------------------
r1820 | jwoithe | 2010-04-30 14:35:24 +0200 (Fri, 30 Apr 2010) | 2 lines

Some systems can have the XDG_TOOLS installed but not have a "system" menu directory.  In this case, installation as root will give an unnecessary error.  Tell scons that errors from the xdg tools are acceptable so the build process isn't aborted because of them.

------------------------------------------------------------------------
r1819 | jwoithe | 2010-04-30 13:58:48 +0200 (Fri, 30 Apr 2010) | 2 lines

- MOTU: make audio streaming work for Mark 3 devices.  There are still details to be sorted out regarding the optical port mode, but it's a good start.

------------------------------------------------------------------------
r1818 | jwoithe | 2010-04-30 12:45:02 +0200 (Fri, 30 Apr 2010) | 4 lines

- MOTU: initial set of changes to streaming code for "mark 3" devices.  More to come.
- MOTU: updates to "mark 3" documentation.

------------------------------------------------------------------------
r1817 | arnonym | 2010-04-27 23:59:10 +0200 (Tue, 27 Apr 2010) | 2 lines

fix #268 by adding the modelid of the newer version of the audiofire8

------------------------------------------------------------------------
r1816 | arnonym | 2010-04-27 23:34:59 +0200 (Tue, 27 Apr 2010) | 4 lines

Support more minimalistic installations:

When the dbus-devel-stuff is not found, the dbus-server is not built. Of course this stops the mixer (if it can be built) from being usable. But when you don't need that because the device doesn't need/have mixer control, this simplifies things for low-disk-space systems...

------------------------------------------------------------------------
r1815 | arnonym | 2010-04-27 21:35:18 +0200 (Tue, 27 Apr 2010) | 2 lines

Looks as if the DSP-version has the 0x08 as id.

------------------------------------------------------------------------
r1814 | jwoithe | 2010-04-20 15:46:11 +0200 (Tue, 20 Apr 2010) | 2 lines

MOTU: further updates to "mark 3" device protocol documentation.

------------------------------------------------------------------------
r1813 | jwoithe | 2010-04-19 16:04:13 +0200 (Mon, 19 Apr 2010) | 2 lines

RME: fix some format specifiers in debug messages.

------------------------------------------------------------------------
r1812 | jwoithe | 2010-04-19 16:01:18 +0200 (Mon, 19 Apr 2010) | 4 lines

MOTU: Start expanding the optical port control infrastructure to allow for the dual optical ports on some Mark3 devices (eg: the 828mk3).
MOTU: First cut at refining the 828mk3 port list.
MOTU: Update Mark-3 documentation.

------------------------------------------------------------------------
r1810 | arnonym | 2010-04-16 22:53:18 +0200 (Fri, 16 Apr 2010) | 2 lines

andrezino reports the tc elektronic konnekt6 to be working with this entry.

------------------------------------------------------------------------
r1809 | arnonym | 2010-04-14 22:34:23 +0200 (Wed, 14 Apr 2010) | 4 lines

Catch up with things on my todo list:

I am told this fixes the two-channel devices of weiss.ch. And it doesn't seem to break my saffire, so lets wait for the crowd to find errors from this change...

------------------------------------------------------------------------
r1808 | jwoithe | 2010-04-08 14:54:16 +0200 (Thu, 08 Apr 2010) | 2 lines

MOTU: minor corrections to "mark 3" protocol documentation.

------------------------------------------------------------------------
r1807 | jwoithe | 2010-03-29 15:59:45 +0200 (Mon, 29 Mar 2010) | 2 lines

MOTU: add initial documentation of the protocol used in MOTU "mark 3" devices.  Derived using an 828 Mk 3 device.  Thanks to Rob Joyner for the loan of an 828 Mk 3 to allow this work to be done.  Refinements (and code based on this information) to follow.

------------------------------------------------------------------------
r1806 | arnonym | 2010-03-17 21:34:03 +0200 (Wed, 17 Mar 2010) | 4 lines

Fix #266

And some stuff I had on disk: When showDevice is called, the device info should be shown. Not depending on the debug-verbosity but just shown... Makes test-dice-eap _much_ nicer to use.

------------------------------------------------------------------------
r1805 | arnonym | 2010-03-08 23:59:51 +0200 (Mon, 08 Mar 2010) | 2 lines

Fix some warnings...

------------------------------------------------------------------------
r1804 | arnonym | 2010-02-18 00:25:46 +0200 (Thu, 18 Feb 2010) | 2 lines

Less code.

------------------------------------------------------------------------
r1803 | arnonym | 2010-02-18 00:11:45 +0200 (Thu, 18 Feb 2010) | 2 lines

First try to start the ffado-dbus-server, then show the message...

------------------------------------------------------------------------
r1802 | arnonym | 2010-02-18 00:08:15 +0200 (Thu, 18 Feb 2010) | 3 lines

- Start ffado-dbus-daemon if its not running. (old)
 - Retry if ffado-dbus-daemon stops while the mixer runs. (new)

------------------------------------------------------------------------
r1801 | arnonym | 2010-02-17 01:19:19 +0200 (Wed, 17 Feb 2010) | 2 lines

Finally make the nickname work. The saffires use a special place in application-space to store the nickname. but it doesn't work if you don't name the functions correctly.

------------------------------------------------------------------------
r1800 | arnonym | 2010-02-16 21:53:32 +0200 (Tue, 16 Feb 2010) | 2 lines

Don't reset the volume of the neighbors output when changing one output. This only affects the monitoring section of the saffire pro24 and pro40...

------------------------------------------------------------------------
r1799 | ppalmers | 2010-02-12 17:43:45 +0200 (Fri, 12 Feb 2010) | 1 line

fix wrong read function bug
------------------------------------------------------------------------
r1798 | jwoithe | 2010-02-11 11:51:44 +0200 (Thu, 11 Feb 2010) | 2 lines

RME: remove commented out debug lines which are no longer required.

------------------------------------------------------------------------
r1797 | jwoithe | 2010-02-11 11:46:06 +0200 (Thu, 11 Feb 2010) | 2 lines

MOTU: add some commented out code which could be used for future debugging

------------------------------------------------------------------------
r1796 | arnonym | 2010-02-03 00:31:11 +0200 (Wed, 03 Feb 2010) | 4 lines

Re-factor the generic-dice parts of the mixer into its own gui.

For devices that have a dice EAP but are not focusrite, using 'mixer = "Generic_Dice_EAP"' in the configuration should give a nice mixer and router.

------------------------------------------------------------------------
r1795 | arnonym | 2010-02-03 00:29:11 +0200 (Wed, 03 Feb 2010) | 2 lines

Change the wording a bit.

------------------------------------------------------------------------
r1794 | arnonym | 2010-01-18 23:38:33 +0200 (Mon, 18 Jan 2010) | 9 lines

This might (or might not) fix #256 by actually implementing EAP-support for the Pro40.

Please someone test this and report back.

Everything should work except:
 - The volume knob doesn't seem to have an effect on the Pro24, don't know about the Pro40. It writes to the right registers but somehow the hardware doesn't recognise it.
 - The "Use ADAT-Out as S/PDIF" and "Pad Output" are not implemented yet.
 - I don't know whether changing to nickname works with the Pro40. It doesn't with the Pro24 altough the correct registers are written to. Probably sending an OP-code is missing...

------------------------------------------------------------------------
r1793 | ppalmers | 2010-01-18 20:36:44 +0200 (Mon, 18 Jan 2010) | 1 line

add config entry for the flex radio SDR
------------------------------------------------------------------------
r1792 | arnonym | 2010-01-18 02:06:07 +0200 (Mon, 18 Jan 2010) | 2 lines

Remove the additional "_p" and "_c" in the channel names. And remove the "in" and "out" in the names of the midi ports.

------------------------------------------------------------------------
r1791 | arnonym | 2010-01-18 02:05:06 +0200 (Mon, 18 Jan 2010) | 2 lines

Remove that MIDI-name mangling again. (Anyone noticed that I got it wrong? :-)

------------------------------------------------------------------------
r1790 | arnonym | 2010-01-16 22:10:22 +0200 (Sat, 16 Jan 2010) | 2 lines

If a channel is only named 'MIDI', add a 'IN' or 'OUT' to indicate the direction. Otherwise only one channel gets registered at jack when jack doesn't add P_ and C_ before the names (and there is no reason it should).

------------------------------------------------------------------------
r1789 | arnonym | 2010-01-10 22:59:02 +0200 (Sun, 10 Jan 2010) | 2 lines

Use the GUID as the id and get rid of setId(). This is part of making the port names unique and persistent per device.

------------------------------------------------------------------------
r1788 | jwoithe | 2010-01-07 00:26:22 +0200 (Thu, 07 Jan 2010) | 2 lines

MOTU: fix a typo in r1785 patchset (fixes #252).  Thanks to pdhales for reporting.

------------------------------------------------------------------------
r1787 | ppalmers | 2010-01-06 19:16:22 +0200 (Wed, 06 Jan 2010) | 1 line

ensure that non-eap dice devices are discovered properly (fixes #248)
------------------------------------------------------------------------
r1785 | jwoithe | 2010-01-06 08:05:21 +0200 (Wed, 06 Jan 2010) | 2 lines

MOTU: the MOTU driver doesn't currently implement boost/level for the 896HD, probably because that interface doesn't have them.  Therefore disable them in ffado-mixer.

------------------------------------------------------------------------
r1784 | jwoithe | 2010-01-06 00:19:48 +0200 (Wed, 06 Jan 2010) | 2 lines

MOTU: add temporary debugging output to help track down 896HD issues.  It should be removed once the problems have been identified and fixed.

------------------------------------------------------------------------
r1782 | jwoithe | 2010-01-05 06:46:19 +0200 (Tue, 05 Jan 2010) | 2 lines

MOTU: possibly fix random segfaults/crashes when using the 896HD at 1x or 2x sampling rates due to references beyond the end of a tx/rx packet's data allocation

------------------------------------------------------------------------
r1781 | arnonym | 2010-01-05 00:20:56 +0200 (Tue, 05 Jan 2010) | 2 lines

Adopt the test:-)

------------------------------------------------------------------------
r1780 | arnonym | 2010-01-05 00:10:14 +0200 (Tue, 05 Jan 2010) | 9 lines

Bring back the peaks.

And *flush* fix compilation from the commit before.
I will not commit without a succesfull build.
I will not commit without a succesfull build.
I will not commit without a succesfull build.
I will not commit without a succesfull build.
I will not commit without a succesfull build.

------------------------------------------------------------------------
r1779 | arnonym | 2010-01-04 23:10:45 +0200 (Mon, 04 Jan 2010) | 6 lines

Rewrite the RouterConfig and save lots of code.

Plus, it seems I fixed a "bug" where the device would output a high pitched sound when an output was disconnected. Now this seems gone...

Next stop: PeakSpace

------------------------------------------------------------------------
r1777 | arnonym | 2010-01-04 18:09:43 +0200 (Mon, 04 Jan 2010) | 2 lines

fix #249: Add all needed library-flags to satisfy the pedantic debian tools...

------------------------------------------------------------------------
r1776 | arnonym | 2010-01-03 01:07:26 +0200 (Sun, 03 Jan 2010) | 14 lines

Reduce the LOC:

- Use only strings as identifiers for the Dice::EAP::Router. Updating the router doesn't happen that often, using strings only is acceptable. And it eases handling so much.
 - Adapt the control interface.
 - Adapt the dbus interface.
 - Adapt the routers gui.

- The peak-space is not yet working (I didn't actually test it), the peak-functions of EAP::Router return nothing.
 - And the test for the dice has to be adopted when the peaks are working again.

TODO:
 - Re-activate the peaks.
 - Adopt the mixer interface.

------------------------------------------------------------------------
r1775 | arnonym | 2010-01-02 01:16:03 +0200 (Sat, 02 Jan 2010) | 4 lines

''Saffire Pro24'': Display only the connected ports in the mixer.

This one shows that somewhere in the router or mixer a blockname is used from blockToName, while a real name was given in code. Will look at this tomorrow.

------------------------------------------------------------------------
r1774 | arnonym | 2010-01-02 00:50:51 +0200 (Sat, 02 Jan 2010) | 2 lines

Be more consistent.

------------------------------------------------------------------------
r1773 | arnonym | 2010-01-02 00:10:10 +0200 (Sat, 02 Jan 2010) | 2 lines

Adopt to the api change in src/dice.

------------------------------------------------------------------------
r1772 | arnonym | 2010-01-02 00:09:19 +0200 (Sat, 02 Jan 2010) | 5 lines

- Use stringlist.
 - Shuffle lots of code around.
 - Change public/private access.
 - Make setupSources() and setupDestinations() virtual functions in Dice::EAP for easier subclassing.

------------------------------------------------------------------------
r1771 | arnonym | 2010-01-02 00:07:30 +0200 (Sat, 02 Jan 2010) | 2 lines

Move the include to the right place.

------------------------------------------------------------------------
r1770 | arnonym | 2010-01-02 00:04:37 +0200 (Sat, 02 Jan 2010) | 2 lines

Change the fbtypes into ffadotypes.h and add a generic stringlist.

------------------------------------------------------------------------
r1769 | arnonym | 2010-01-02 00:02:54 +0200 (Sat, 02 Jan 2010) | 2 lines

Less warnings with doxygen.

------------------------------------------------------------------------
r1767 | jwoithe | 2009-12-29 01:14:36 +0200 (Tue, 29 Dec 2009) | 2 lines

MOTU: fix ffado-mixer bug triggered at 4x sampling rates (fixes part of #247)

------------------------------------------------------------------------
r1766 | arnonym | 2009-12-29 01:07:32 +0200 (Tue, 29 Dec 2009) | 5 lines

Start some cleanup in the dice-code.

No need for the EAP to be defined inside Dice::Device.

------------------------------------------------------------------------
r1765 | ppalmers | 2009-12-23 21:10:49 +0200 (Wed, 23 Dec 2009) | 1 line

put everything related to iso streaming into one CPP file. this should give the compiler some more freedom to optimize.
------------------------------------------------------------------------
r1764 | ppalmers | 2009-12-19 18:26:20 +0200 (Sat, 19 Dec 2009) | 1 line

fix 32bit format string warnings
------------------------------------------------------------------------
r1763 | ppalmers | 2009-12-19 18:12:53 +0200 (Sat, 19 Dec 2009) | 6 lines

Merged revisions 1536,1541,1544-1546,1549,1554-1562,1571,1579-1581,1618,1632,1634-1635,1661,1677-1679,1703-1704,1715,1720-1723,1743-1745,1755 via svnmerge from 
svn+ssh://ffadosvn@ffado.org/ffado/branches/libffado-2.0

Also fix remaining format string warnings.

------------------------------------------------------------------------
r1762 | ppalmers | 2009-12-19 15:08:45 +0200 (Sat, 19 Dec 2009) | 1 line

move the svnmerge properties to where they belong
------------------------------------------------------------------------
r1753 | arnonym | 2009-12-18 20:52:13 +0200 (Fri, 18 Dec 2009) | 2 lines

some doc formatting.

------------------------------------------------------------------------
r1752 | arnonym | 2009-12-09 23:47:35 +0200 (Wed, 09 Dec 2009) | 2 lines

Fix #245: remove the duplicate multimix16 entry.

------------------------------------------------------------------------
r1749 | jwoithe | 2009-12-07 14:03:36 +0200 (Mon, 07 Dec 2009) | 4 lines

RME:
- minor code clean-up to avoid unnecessary duplication
- crudely make the RME device start streaming data (now it just needs to be controlled)

------------------------------------------------------------------------
r1748 | arnonym | 2009-12-05 00:44:08 +0200 (Sat, 05 Dec 2009) | 4 lines

Split the main-part of ffado-mixer.in into ffado/ffadowindow.py.

Use that new independency to create ffado-mixer-profiler (which isn't installed system-wide). Maybe that helps to determine where time is ticking away.

------------------------------------------------------------------------
r1747 | arnonym | 2009-12-04 23:48:48 +0200 (Fri, 04 Dec 2009) | 6 lines

Some buildsystem stuff:

- Replace more AppendUnique() by MergeFlags()
 - Manage the GENERIC_AVC dependancy in src/SConscript where it is used
 - Forward-port the creation of libffado.so.* from the 2.0-branch.

------------------------------------------------------------------------
r1746 | jwoithe | 2009-12-04 13:50:31 +0200 (Fri, 04 Dec 2009) | 2 lines

RME: add a transmit stream processor to the device.  A functional tx stream is required before rx data will be seen; this is the first step towards that.

------------------------------------------------------------------------
r1742 | arnonym | 2009-12-02 00:54:27 +0200 (Wed, 02 Dec 2009) | 5 lines

Extend the interface of the MatrixMixer:

- Have booleans indicating if names and/or connections are supported. These functions are pure virtual.
 - Allow names and connections to be read and written, the default implementations of these (not anymore pure) virtual functions do nothing and return false or nothing.

------------------------------------------------------------------------
r1741 | arnonym | 2009-12-02 00:50:48 +0200 (Wed, 02 Dec 2009) | 2 lines

ignore me

------------------------------------------------------------------------
r1740 | arnonym | 2009-12-02 00:49:23 +0200 (Wed, 02 Dec 2009) | 2 lines

Minor gui stuff: Smaller names for the mixer.

------------------------------------------------------------------------
r1739 | arnonym | 2009-12-01 23:16:07 +0200 (Tue, 01 Dec 2009) | 2 lines

Reducing the number of lines while maintaining the functionality is good. Improving the gui is even better :-)

------------------------------------------------------------------------
r1738 | arnonym | 2009-11-30 23:44:23 +0200 (Mon, 30 Nov 2009) | 2 lines

No need to name this a "test" server.

------------------------------------------------------------------------
r1737 | arnonym | 2009-11-30 23:43:03 +0200 (Mon, 30 Nov 2009) | 2 lines

Be less noisy. And output a wanring if writing to the application space fails.

------------------------------------------------------------------------
r1736 | arnonym | 2009-11-30 23:31:56 +0200 (Mon, 30 Nov 2009) | 2 lines

Add a widget to control the monitoring section of the Saffire Pro24 and Pro40. This also controls the line/inst and level switches of the pro24.

------------------------------------------------------------------------
r1735 | arnonym | 2009-11-25 23:09:04 +0200 (Wed, 25 Nov 2009) | 2 lines

Lots more controls. Should be ready for the gui now...

------------------------------------------------------------------------
r1734 | arnonym | 2009-11-25 00:30:28 +0200 (Wed, 25 Nov 2009) | 2 lines

Controlling the dim-level works now. Controlling the DAC-level works too, but switching the dac-level to control channels is still missing.

------------------------------------------------------------------------
r1733 | arnonym | 2009-11-25 00:27:33 +0200 (Wed, 25 Nov 2009) | 2 lines

Make the debug protected.

------------------------------------------------------------------------
r1732 | arnonym | 2009-11-24 00:08:40 +0200 (Tue, 24 Nov 2009) | 2 lines

Forgot to commit this file. Otherwise r1730 doesn't compile.

------------------------------------------------------------------------
r1731 | arnonym | 2009-11-24 00:05:19 +0200 (Tue, 24 Nov 2009) | 2 lines

Looks as if this doesn't break using the old stack. see #240

------------------------------------------------------------------------
r1730 | arnonym | 2009-11-23 23:55:36 +0200 (Mon, 23 Nov 2009) | 2 lines

Save code by creating another class. Next step is unified monitoring and switches for both the pro40 and the pro24.

------------------------------------------------------------------------
r1729 | arnonym | 2009-11-23 23:50:44 +0200 (Mon, 23 Nov 2009) | 3 lines

This should enable nickname changes for the saffire pro40. ppalmers can you test this?
On the pro24 the changes work and are written to the application space. Only they don't seem to last over a power-cycle. Don't yet know why...

------------------------------------------------------------------------
r1728 | arnonym | 2009-11-23 23:40:01 +0200 (Mon, 23 Nov 2009) | 2 lines

ignore me

------------------------------------------------------------------------
r1727 | arnonym | 2009-11-23 23:38:45 +0200 (Mon, 23 Nov 2009) | 2 lines

Introduce a virtual factory function. That way subclasses can have subclassed EAPs created. Like EAPs dealing with the special Focusrite extensions...

------------------------------------------------------------------------
r1726 | arnonym | 2009-11-22 23:42:57 +0200 (Sun, 22 Nov 2009) | 2 lines

Access more switches of the pro24. Dim and Mute work, Mono-monitoring somehow doesn't (yet).

------------------------------------------------------------------------
r1725 | arnonym | 2009-11-22 23:40:24 +0200 (Sun, 22 Nov 2009) | 2 lines

As far as I can say this function works. Don't display the warning. But I think the control-interface needs to be extended to allow to call save and restore from gui.

------------------------------------------------------------------------
r1724 | arnonym | 2009-11-22 22:55:51 +0200 (Sun, 22 Nov 2009) | 2 lines

Introduce a Control::Boolean for simple switches. To be used in the SaffirePro24 first, but maybe others will follow? Using this instead of Control::Enum saves code.

------------------------------------------------------------------------
r1719 | ppalmers | 2009-11-22 15:36:34 +0200 (Sun, 22 Nov 2009) | 1 line

fix some compilation issues with the PRO24 code
------------------------------------------------------------------------
r1718 | arnonym | 2009-11-18 22:53:43 +0200 (Wed, 18 Nov 2009) | 4 lines

Switching line/inst on channels 1+2 and low/high in channels 3+4 works now with the dbus-interface. Mixer is coming soon.

Gotta ask back whether I am allowed to add the full description of the application-space here for all to see.

------------------------------------------------------------------------
r1717 | arnonym | 2009-11-17 00:37:28 +0200 (Tue, 17 Nov 2009) | 2 lines

Make reading the whole application space its own special thing.

------------------------------------------------------------------------
r1716 | arnonym | 2009-11-16 22:44:40 +0200 (Mon, 16 Nov 2009) | 6 lines

- Update the doxygen file to a new doxygen-version.

- De-activate the LaTeX-output.

- Enable output for QAssistant files.

------------------------------------------------------------------------
r1714 | arnonym | 2009-11-16 00:56:42 +0200 (Mon, 16 Nov 2009) | 2 lines

ignore me

------------------------------------------------------------------------
r1713 | arnonym | 2009-11-16 00:54:26 +0200 (Mon, 16 Nov 2009) | 5 lines

Print the hexdump of the quadlets to stderr.

Print the stuff from show() as normal messages, not as debug.
And print the first 512 bytes of the application space of the dice-chip. Trying to figure out which bits to set to control line/instrument and high/low switches.

------------------------------------------------------------------------
r1712 | arnonym | 2009-11-16 00:51:57 +0200 (Mon, 16 Nov 2009) | 2 lines

Start a special class for the saffire24...

------------------------------------------------------------------------
r1711 | arnonym | 2009-11-16 00:20:11 +0200 (Mon, 16 Nov 2009) | 2 lines

Fix deletion.

------------------------------------------------------------------------
r1710 | arnonym | 2009-11-15 21:05:30 +0200 (Sun, 15 Nov 2009) | 2 lines

Not needed anymore. Advanced panning will be resurrected at a different place.

------------------------------------------------------------------------
r1709 | arnonym | 2009-11-15 21:02:29 +0200 (Sun, 15 Nov 2009) | 2 lines

Some more gui-niceness.

------------------------------------------------------------------------
r1708 | arnonym | 2009-11-15 17:45:56 +0200 (Sun, 15 Nov 2009) | 4 lines

Add a context menu with some predefined values and a spinbox for direct entry of custom values.

And display the values in dB. The calculation is currently fixed on the values of the DICE, for more flexibility the dbus-interface needs an extension. Then the range and the definition of 0 dB could be device-dependant.

------------------------------------------------------------------------
r1707 | arnonym | 2009-11-14 23:17:22 +0200 (Sat, 14 Nov 2009) | 4 lines

Use QSettings to store the state of the VU-switch.

And do some playing with deleting panels when the devices are removed from the bus. Otherwise the timer checking the VU-values will run forever and complain about missing dbus-objects.

------------------------------------------------------------------------
r1706 | arnonym | 2009-11-14 23:15:10 +0200 (Sat, 14 Nov 2009) | 2 lines

Panning multiple channels to multiple channels is rather complicated. Probably more complicated then what is needed for ffado-mixer. So go back to individual controls for each junction. They look like colored labels but are in fact sliders. A context-menu with predefined values and a direct input for numbers will come soon...

------------------------------------------------------------------------
r1705 | arnonym | 2009-11-12 22:57:18 +0200 (Thu, 12 Nov 2009) | 2 lines

Somehow there where some "Mixer"s left...

------------------------------------------------------------------------
r1702 | arnonym | 2009-11-09 22:34:33 +0200 (Mon, 09 Nov 2009) | 4 lines

Change whitespaces.

And don't crash when a configuration-file doesn't have a "device_definitions" section.

------------------------------------------------------------------------
r1701 | arnonym | 2009-11-09 22:04:57 +0200 (Mon, 09 Nov 2009) | 2 lines

As far as I can tell these patches don't break the old stack. So apply them... see #240

------------------------------------------------------------------------
r1700 | arnonym | 2009-11-03 23:08:56 +0200 (Tue, 03 Nov 2009) | 2 lines

Use the libraw1394 in the correct way by only checking errno if an error actually occured.

------------------------------------------------------------------------
r1699 | arnonym | 2009-11-02 23:48:55 +0200 (Mon, 02 Nov 2009) | 2 lines

Fix alsa-dependancy.

------------------------------------------------------------------------
r1698 | jwoithe | 2009-10-24 13:23:19 +0200 (Sat, 24 Oct 2009) | 2 lines

RME: hacks to try to convince the device to start streaming.  They didn't work.

------------------------------------------------------------------------
r1697 | jwoithe | 2009-10-19 14:37:15 +0200 (Mon, 19 Oct 2009) | 5 lines

RME: 
* implement stub functions for more device initialisation processes
* continued work towards convincing the interface to start streaming

------------------------------------------------------------------------
r1696 | jwoithe | 2009-10-18 13:07:09 +0200 (Sun, 18 Oct 2009) | 2 lines

RME: add some device initialisation details.

------------------------------------------------------------------------
r1695 | arnonym | 2009-10-17 22:40:31 +0200 (Sat, 17 Oct 2009) | 2 lines

stupid error(s)...

------------------------------------------------------------------------
r1694 | arnonym | 2009-10-17 21:15:36 +0200 (Sat, 17 Oct 2009) | 2 lines

Process the dbus-service file even if it can't be installed. That way one can copy it by hand and have the ffado-dbus-server start automaticly when ffado-mixer starts. Or when some other app requests that service.

------------------------------------------------------------------------
r1693 | arnonym | 2009-10-17 21:13:56 +0200 (Sat, 17 Oct 2009) | 2 lines

Set a sane name.

------------------------------------------------------------------------
r1692 | arnonym | 2009-10-17 21:13:09 +0200 (Sat, 17 Oct 2009) | 2 lines

Fix the dummy after last commit.

------------------------------------------------------------------------
r1691 | arnonym | 2009-10-17 20:14:32 +0200 (Sat, 17 Oct 2009) | 8 lines

Only load the mixerwidgets that are actually needed.

To automate this, a rules are imposed:

The filename has to be the lower-case variant of the mixer name

This should speed up loading time of ffado-mixer. Please everybody test this and report any errors as I don't have all the devices to test everything fully.

------------------------------------------------------------------------
r1690 | jwoithe | 2009-10-17 15:15:46 +0200 (Sat, 17 Oct 2009) | 5 lines

RME:
* work on streaming setup continues
* some whitespace cleanups
* update documentation

------------------------------------------------------------------------
r1689 | jwoithe | 2009-10-16 14:14:30 +0200 (Fri, 16 Oct 2009) | 2 lines

RME: more streaming control tweaks.

------------------------------------------------------------------------
r1688 | jwoithe | 2009-10-15 15:55:37 +0200 (Thu, 15 Oct 2009) | 2 lines

RME: more work on the streaming system.

------------------------------------------------------------------------
r1687 | arnonym | 2009-10-12 23:03:04 +0200 (Mon, 12 Oct 2009) | 2 lines

Apply the patch to recognise the maudio pro610.

------------------------------------------------------------------------
r1685 | arnonym | 2009-10-12 22:32:00 +0200 (Mon, 12 Oct 2009) | 2 lines

Implement a rather old patch from stefan richter to better check for the firewire stacks presence.

------------------------------------------------------------------------
r1684 | jwoithe | 2009-10-12 14:11:09 +0200 (Mon, 12 Oct 2009) | 2 lines

RME: more preparatory work on the streaming system

------------------------------------------------------------------------
r1683 | arnonym | 2009-10-11 23:03:41 +0200 (Sun, 11 Oct 2009) | 4 lines

Kind of a usable panner. In theory this should allow panning several sources to several sinks. In practice only mono-to-stereo and stereo-to-stereo is currently possible. Not because of the panner but because of the way it works in the matrixmixer. Will change that soon.

One of the current draw-backs is that I am using numpy. Standard pythons matrix handling doesn't seem so good.

------------------------------------------------------------------------
r1681 | arnonym | 2009-10-09 21:39:42 +0200 (Fri, 09 Oct 2009) | 2 lines

Apply the atomic.patch of #197 to trunk.

------------------------------------------------------------------------
r1680 | arnonym | 2009-10-08 22:59:43 +0200 (Thu, 08 Oct 2009) | 4 lines

Some older changes to the crossbarrouter. Most prominent: The VU have to be hand-activated...

And start a panning widget that allows to pan an arbitary number of sources to be panned to an arbitary number of sinks. In 2D.

------------------------------------------------------------------------
r1675 | arnonym | 2009-10-02 23:47:44 +0200 (Fri, 02 Oct 2009) | 2 lines

Lets see how many setups this breaks. see #199

------------------------------------------------------------------------
r1674 | jwoithe | 2009-09-28 14:36:17 +0200 (Mon, 28 Sep 2009) | 2 lines

RME: define some packet dimensional details

------------------------------------------------------------------------
r1673 | jwoithe | 2009-09-20 13:44:43 +0200 (Sun, 20 Sep 2009) | 2 lines

RME: include new RME configuration defines in config.h.in rather than in the local generated copy of config.h as was accidently done initially.

------------------------------------------------------------------------
r1672 | jwoithe | 2009-09-20 13:40:33 +0200 (Sun, 20 Sep 2009) | 2 lines

RME: create templates for streaming objects for the RME driver.  They are based on the MOTU stream processor and still contain a lot of MOTU-specific code.  This will be elimintated over time, but for now these objects should not be used as a reference for the RME devices.

------------------------------------------------------------------------
r1671 | jwoithe | 2009-09-19 14:02:54 +0200 (Sat, 19 Sep 2009) | 4 lines

RME: add some device-level support for stream configuration and control.
RME: make some comments clearer.
RME: remove some useless unreachable code.

------------------------------------------------------------------------
r1670 | arnonym | 2009-09-18 23:08:49 +0200 (Fri, 18 Sep 2009) | 2 lines

Maybe thats an incentive to implement names for the mixer-channels.

------------------------------------------------------------------------
r1669 | arnonym | 2009-09-18 23:01:58 +0200 (Fri, 18 Sep 2009) | 2 lines

some cleanup I forgot.

------------------------------------------------------------------------
r1668 | arnonym | 2009-09-18 23:00:59 +0200 (Fri, 18 Sep 2009) | 3 lines

- Use logging for the debug-output.
 - Add the ability to completely disconnect outputs from their input. This removes (should remove?) routes from the routing table.

------------------------------------------------------------------------
r1667 | arnonym | 2009-09-17 23:51:24 +0200 (Thu, 17 Sep 2009) | 2 lines

ignore me

------------------------------------------------------------------------
r1666 | arnonym | 2009-09-17 23:47:03 +0200 (Thu, 17 Sep 2009) | 4 lines

Simplify (de-)coupling of channels a bit.

And start with getting the mixer values on startup and on coupling. Currently only the maximum is used, no balance/pan is computed...

------------------------------------------------------------------------
r1664 | arnonym | 2009-09-17 22:26:12 +0200 (Thu, 17 Sep 2009) | 2 lines

Remember to regularly check the registration dialog. For example by getting a new device with a different serial number ;-)

------------------------------------------------------------------------
r1663 | arnonym | 2009-09-16 23:06:12 +0200 (Wed, 16 Sep 2009) | 2 lines

Try to fix the problem of #232 by compiling a static file of pkg-config output during compile-time to be included in the ffado-diag output.

------------------------------------------------------------------------
r1662 | arnonym | 2009-09-16 23:02:18 +0200 (Wed, 16 Sep 2009) | 2 lines

From now on, devs have to remove the qt3-mixer by hand :-)

------------------------------------------------------------------------
r1660 | jwoithe | 2009-09-07 13:41:57 +0200 (Mon, 07 Sep 2009) | 3 lines

ieee1394Service: new method allocateFixedIsoChannelGeneric() to allocate a specific channel number instead of scanning for the next available channel number.  This functionality is needed for the RME devices.
RME: assign/detect iso channels to use and allocate required bandwidth.

------------------------------------------------------------------------
r1659 | arnonym | 2009-09-06 23:13:08 +0200 (Sun, 06 Sep 2009) | 2 lines

Make the input- and output-balance work.

------------------------------------------------------------------------
r1658 | arnonym | 2009-09-06 21:57:28 +0200 (Sun, 06 Sep 2009) | 10 lines

Taking everything apart and recombine it:
 - Hiding channels works.
 - Combining channels to stereo-pairs works.
 - Controlling the volume works.

Left for the next commits (to be seen on a per mixer-element basis):
 - Control input- and output-balance.
 - Allow for more than two inputs or outputs.
 - Change the mapping of inputs->outputs.

------------------------------------------------------------------------
r1657 | jwoithe | 2009-09-06 14:27:13 +0200 (Sun, 06 Sep 2009) | 2 lines

RME: lay some groundwork for the management of iso resources used for audio data transfer to/from the device.

------------------------------------------------------------------------
r1656 | arnonym | 2009-09-05 14:56:05 +0200 (Sat, 05 Sep 2009) | 2 lines

Allow to deactivate the peaks.

------------------------------------------------------------------------
r1655 | arnonym | 2009-09-05 14:07:54 +0200 (Sat, 05 Sep 2009) | 2 lines

Thinking about getting structured info about the router-channels.

------------------------------------------------------------------------
r1654 | arnonym | 2009-09-04 21:44:45 +0200 (Fri, 04 Sep 2009) | 2 lines

Implement the vu-meters for the router.

------------------------------------------------------------------------
r1653 | arnonym | 2009-09-04 21:41:36 +0200 (Fri, 04 Sep 2009) | 2 lines

Implement getting a simple list of all current peaks.

------------------------------------------------------------------------
r1652 | arnonym | 2009-09-04 21:40:19 +0200 (Fri, 04 Sep 2009) | 2 lines

Getting the destination index for each peak is enough. Its always only one source per destination, so the destination is a unique identifier.

------------------------------------------------------------------------
r1651 | arnonym | 2009-09-03 23:08:00 +0200 (Thu, 03 Sep 2009) | 6 lines

Lets see if this works out: Work for ppalmers.

A new way to get the peakvalues. Getting each value individual is really cpu-heavy. Getting them all toghether should be better.

(My first real code commit to the internals:)

------------------------------------------------------------------------
r1650 | arnonym | 2009-09-02 23:17:12 +0200 (Wed, 02 Sep 2009) | 2 lines

More ints that are actually bools.

------------------------------------------------------------------------
r1649 | arnonym | 2009-09-01 21:47:15 +0200 (Tue, 01 Sep 2009) | 2 lines

Just flesh out a widget for each destination. Getting the peak-meter is postponed to tomorrow.

------------------------------------------------------------------------
r1648 | arnonym | 2009-09-01 20:30:51 +0200 (Tue, 01 Sep 2009) | 2 lines

Since installing kde4.3 from ppa, some involved package was updated. This newer version seems to not like QWidget.handle() and logging.handle() having different signatures...

------------------------------------------------------------------------
r1647 | arnonym | 2009-08-31 23:35:35 +0200 (Mon, 31 Aug 2009) | 4 lines

Replace the big matrix of buttons by only one button per destination and choose the source per menu. The menu is filled when first shown to reduce startup load.

This is even real working. I have successfully switched the routing here :-)

------------------------------------------------------------------------
r1646 | arnonym | 2009-08-31 23:33:10 +0200 (Mon, 31 Aug 2009) | 2 lines

These functions return bool. Adopt the dbus-interface to the reality.

------------------------------------------------------------------------
r1645 | arnonym | 2009-08-31 21:20:41 +0200 (Mon, 31 Aug 2009) | 4 lines

Something got lost on the way of times...

Don't add -m32 in 64bit builts.

------------------------------------------------------------------------
r1644 | arnonym | 2009-08-31 00:36:59 +0200 (Mon, 31 Aug 2009) | 2 lines

Smaller buttons for the router-states. The group names have an increasing off-by-one error I can't put my finger on now...

------------------------------------------------------------------------
r1643 | arnonym | 2009-08-30 23:41:28 +0200 (Sun, 30 Aug 2009) | 4 lines

Implement a very first interface for the crossbarrouter. currently its read-only.

With 81 sources and 83 destinations initialization of the mixer takes its time. Mostly because each connection is checked individually. Would be faster to use the getConnectionMap function defined in the dbus interface. Unfortunately this doesn't seem to be implemented currently.

------------------------------------------------------------------------
r1642 | arnonym | 2009-08-30 21:21:55 +0200 (Sun, 30 Aug 2009) | 2 lines

ignore me

------------------------------------------------------------------------
r1641 | arnonym | 2009-08-30 21:08:49 +0200 (Sun, 30 Aug 2009) | 6 lines

Unify tabs->space.

Remove the python stuff from the installs of the developers.

Compute the directory where to install the python modules.

------------------------------------------------------------------------
r1640 | arnonym | 2009-08-30 21:05:51 +0200 (Sun, 30 Aug 2009) | 6 lines

Make it more pythonic.

All ffado stuff is now in packages getting installed to the official python dirs. Ideally this would allow other apps to use the stuff from us.

Maybe the generic widgets (that are used by multiple mixers) should go into ffado/widgets...

------------------------------------------------------------------------
r1639 | arnonym | 2009-08-29 16:00:15 +0200 (Sat, 29 Aug 2009) | 2 lines

More on the mixingmatrix. channels can be hidden now.

------------------------------------------------------------------------
r1638 | arnonym | 2009-08-29 13:34:30 +0200 (Sat, 29 Aug 2009) | 4 lines

Remove the old qt3 mixer. Only ports from 2.0 branch where happening here. And the new work in trunk is in the qt4 mixer. And Qt3 is unsupported since quite some time already. ppalmers okayed it...

I added a temporary code-segment to sconscript to remove the qt3 mixer stuff from the installation on the devs machines. Hopefully I remember to remove that code in two or three weeks.

------------------------------------------------------------------------
r1637 | arnonym | 2009-08-28 20:18:28 +0200 (Fri, 28 Aug 2009) | 2 lines

Let the pro40 use the same mixer. Probably more devices can use that, but that is for ppalmers to decide.

------------------------------------------------------------------------
r1636 | arnonym | 2009-08-28 20:17:04 +0200 (Fri, 28 Aug 2009) | 6 lines

Mix it baby!

A first matrix with dials (extra for ppalmers:-P). For screen-space its limited to 8x8 elements currently (or less, if the device has less).
Have to think about how to hide channels that aren't used. Or hide groups of channels with similar names. Hint: this would currently result in one group named "FIXME" :-)

------------------------------------------------------------------------
r1633 | arnonym | 2009-08-27 23:42:43 +0200 (Thu, 27 Aug 2009) | 2 lines

The mixer works. Will look at streaming tomorrow.

------------------------------------------------------------------------
r1631 | ppalmers | 2009-08-27 22:55:41 +0200 (Thu, 27 Aug 2009) | 1 line

fix byte-swapping SSE code bug
------------------------------------------------------------------------
r1630 | ppalmers | 2009-08-27 21:25:28 +0200 (Thu, 27 Aug 2009) | 1 line

add base level support for the DICE EAP based mixers (incomplete)
------------------------------------------------------------------------
r1629 | jwoithe | 2009-08-27 14:18:06 +0200 (Thu, 27 Aug 2009) | 2 lines

RME: refine shared device configuration locking.  Rename some data objects for clarity.  Move frequency settings into the shared configuration object since ultimately these need to be accessible to all FFADO/RME processes.  Other minor cleanups.

------------------------------------------------------------------------
r1628 | jwoithe | 2009-08-23 14:03:54 +0200 (Sun, 23 Aug 2009) | 2 lines

RME: If shared memory allocation fails, use a local duplicate of the shared memory structure.  This alleviates the need to test for valid shared data before any of the housekeeping fields are used.

------------------------------------------------------------------------
r1627 | jwoithe | 2009-08-22 08:55:43 +0200 (Sat, 22 Aug 2009) | 2 lines

RME: implement initial use of shared data object for configuration data.  Some rough edges remain and are being addressed.

------------------------------------------------------------------------
r1626 | jwoithe | 2009-08-22 06:32:03 +0200 (Sat, 22 Aug 2009) | 3 lines

RME: refinement of shared memory management.  Removal of race conditions (even though they were unlikely in practice).
RME: include shared memory management module in compilation in readiness for use.

------------------------------------------------------------------------
r1625 | arnonym | 2009-08-21 21:47:15 +0200 (Fri, 21 Aug 2009) | 5 lines

- add a small line to make vim use spaces for indentation
 - Don't compile ui-files to python-code at install-time. Load them from the pythondir during runtime.

Now the only reason to run scons (during ffado-mixer development) is when one of the .in-files where changed.

------------------------------------------------------------------------
r1624 | jwoithe | 2009-08-21 15:58:02 +0200 (Fri, 21 Aug 2009) | 2 lines

RME: prepare for the use of a shared memory object to share device configuration between FFADO-enabled applications.  This is required because significant parts of the device configuration cannot be retrieved from the device and must be managed in software.

------------------------------------------------------------------------
r1623 | arnonym | 2009-08-16 19:50:21 +0200 (Sun, 16 Aug 2009) | 2 lines

Build-system patrol is back on duty :-)

------------------------------------------------------------------------
r1622 | jwoithe | 2009-08-16 15:53:12 +0200 (Sun, 16 Aug 2009) | 2 lines

RME: input, output and phones mixer/control elements now control respective device parameters

------------------------------------------------------------------------
r1621 | jwoithe | 2009-08-16 14:50:26 +0200 (Sun, 16 Aug 2009) | 2 lines

Increase debug log level of selected method notifiers in the controlserver from VERBOSE to VERY_VERBOSE.  This prevents messages associated with periodic idle handler action getting in the way of other more interesting messages at the verbose level.

------------------------------------------------------------------------
r1620 | jwoithe | 2009-08-16 14:37:38 +0200 (Sun, 16 Aug 2009) | 2 lines

Trunk: Note changes to URLs for SIP and PyQt

------------------------------------------------------------------------
r1617 | jwoithe | 2009-08-13 16:27:35 +0200 (Thu, 13 Aug 2009) | 2 lines

bebob/focusrite: debugOutput() was missing an argument in the mixer show() methods.

------------------------------------------------------------------------
r1616 | jwoithe | 2009-08-13 16:21:51 +0200 (Thu, 13 Aug 2009) | 3 lines

RME: minor documentation corrections
RME: the Fireface-400 input gain mixer controls now drive the hardware.  Control of the output gains is also included in this infrastructure but is not connected to any mixer controls yet - they will be controlled by the mixer proper once that's implemented.

------------------------------------------------------------------------
r1615 | jwoithe | 2009-08-10 15:35:03 +0200 (Mon, 10 Aug 2009) | 5 lines

RME: Rearrange control pane in mixer.  Add controls for input gains on Fireface 400.
RME: Fix bug which prevented driver from correctly reporting the available samplerates.
RME: Correct a misunderstanding about the device streaming status' availability from hardware (essentially it's not available and needs to be cached in software).
RME: Minor corrections to device documentation.

------------------------------------------------------------------------
r1614 | jwoithe | 2009-08-08 13:32:20 +0200 (Sat, 08 Aug 2009) | 2 lines

RME: documentation corrections and minor updates

------------------------------------------------------------------------
r1613 | jwoithe | 2009-08-06 15:18:49 +0200 (Thu, 06 Aug 2009) | 2 lines

RME: refine mixer control tab arrangement

------------------------------------------------------------------------
r1612 | jwoithe | 2009-08-06 14:30:32 +0200 (Thu, 06 Aug 2009) | 2 lines

RME: initial documentation of Fireface 400 matrix mixer controls and input preamp gain controls.

------------------------------------------------------------------------
r1611 | jwoithe | 2009-08-04 10:11:39 +0200 (Tue, 04 Aug 2009) | 2 lines

RME: incidental tweaks to widget names in the mixer.

------------------------------------------------------------------------
r1610 | jwoithe | 2009-08-02 16:26:58 +0200 (Sun, 02 Aug 2009) | 2 lines

RME: initial implementation of ff400 channel 3/4 input options in mixer control tab

------------------------------------------------------------------------
r1609 | jwoithe | 2009-08-02 15:48:07 +0200 (Sun, 02 Aug 2009) | 3 lines

RME: provide low-level support for the channel 3/4 pad/instrument options of the Fireface-400
RME: update device documentation

------------------------------------------------------------------------
r1608 | jwoithe | 2009-08-02 14:17:20 +0200 (Sun, 02 Aug 2009) | 2 lines

RME: refine available mixer device controls depending on the model in use.

------------------------------------------------------------------------
r1607 | jwoithe | 2009-07-26 15:08:27 +0200 (Sun, 26 Jul 2009) | 5 lines

RME:
 * only show two mic phantom channels for the FF400 to avoid user confusion
 * label the phantom channels correctly when a FF800 is in use
 *  add graphical mixer elements for all required device controls (note that only the "phantom" controls are connected to the device hardware at present).  Top-level layout is not dynamic: this will be fixed once a final layout is decided on.  For the moment controls exist to facilitate testing of controls.

------------------------------------------------------------------------
r1606 | jwoithe | 2009-07-25 15:58:45 +0200 (Sat, 25 Jul 2009) | 3 lines

RME: phantom switches in mixer GUI now control the respective Fireface hardware.  Tested only on FF400 to date.
RME: add some more TCO glue.

------------------------------------------------------------------------
r1605 | jwoithe | 2009-07-23 13:57:42 +0200 (Thu, 23 Jul 2009) | 3 lines

RME: implement the beginnings of a device control interface via the ffado dbus framework.
MOTU: fix error handling code paths in buildMixer() and destroyMixer().

------------------------------------------------------------------------
r1604 | jwoithe | 2009-07-23 13:18:42 +0200 (Thu, 23 Jul 2009) | 2 lines

RME: remove some macros which are unlikely to be used anymore.

------------------------------------------------------------------------
r1603 | jwoithe | 2009-07-23 13:12:27 +0200 (Thu, 23 Jul 2009) | 2 lines

Restore scan-devreg functionality accidently removed in revision 1550.

------------------------------------------------------------------------
r1602 | jwoithe | 2009-07-20 15:27:00 +0200 (Mon, 20 Jul 2009) | 2 lines

RME: remove unused defines.  Correct model defines in the RME mixer application.

------------------------------------------------------------------------
r1601 | jwoithe | 2009-07-20 15:19:31 +0200 (Mon, 20 Jul 2009) | 2 lines

RME: implement mixer skeleton to allow device configuration functionality to be tested.  Note that none of the included controls are functional yet.

------------------------------------------------------------------------
r1600 | jwoithe | 2009-07-19 13:48:15 +0200 (Sun, 19 Jul 2009) | 3 lines

RME: put setting control type defines in a more intelligent place.  Add a separate type to denote an unset type field.
RME: Export a "model" info control to report the model via dbus (to be used by ffadomixer).  Handle unknown control types in set/get methods.

------------------------------------------------------------------------
r1599 | jwoithe | 2009-07-19 08:22:51 +0200 (Sun, 19 Jul 2009) | 3 lines

RME: implement highlevel function for setting phantom power status. 
RME: add phantom power control to the generic device control mixer element.

------------------------------------------------------------------------
r1598 | jwoithe | 2009-07-18 14:12:07 +0200 (Sat, 18 Jul 2009) | 2 lines

RME: minor boilerplate additions to the settings control object.

------------------------------------------------------------------------
r1597 | jwoithe | 2009-07-13 14:45:54 +0200 (Mon, 13 Jul 2009) | 2 lines

RME: begin high-level framework for device control via FFADO dbus interface.

------------------------------------------------------------------------
r1596 | jwoithe | 2009-07-12 15:30:55 +0200 (Sun, 12 Jul 2009) | 5 lines

RME:
 * implement lowlevel streaming control/setup/status functions
 * fill in further status register details
 * update documentation

------------------------------------------------------------------------
r1595 | ppalmers | 2009-07-11 11:22:46 +0200 (Sat, 11 Jul 2009) | 1 line

disable running status for SCS
------------------------------------------------------------------------
r1594 | jwoithe | 2009-07-10 15:43:04 +0200 (Fri, 10 Jul 2009) | 8 lines

RME driver:
 * add some missing status bit defines, fix naming of some others
 * add full decoding of device status registers
 * cache device run status
 * take external clock rates and device run status into account when evaluating the validity of sampling rates
 * minor bugfixes
 * implement function to configure a fixed DDS frequency on the device

------------------------------------------------------------------------
r1593 | jwoithe | 2009-07-09 14:47:19 +0200 (Thu, 09 Jul 2009) | 2 lines

RME: add missing state defines, fix an incorrect one and make others more consistent.

------------------------------------------------------------------------
r1592 | jwoithe | 2009-07-09 14:42:30 +0200 (Thu, 09 Jul 2009) | 3 lines

RME: begin tidy-up of sampling rate control.
RME: start implementation of device status retrieval.

------------------------------------------------------------------------
r1591 | jwoithe | 2009-07-03 13:23:20 +0200 (Fri, 03 Jul 2009) | 3 lines

RME: implement writing to device settings flash block
RME: correctly initialise the entire settings structure to zero

------------------------------------------------------------------------
r1590 | jwoithe | 2009-06-27 16:10:52 +0200 (Sat, 27 Jun 2009) | 4 lines

RME: clarify comments in source code
RME: add explicit define to use when determining DDS status from configuration flash RAM
RME: add TCO information to the RME documentation

------------------------------------------------------------------------
r1589 | jwoithe | 2009-06-27 15:25:55 +0200 (Sat, 27 Jun 2009) | 5 lines

RME: tidied up software settings structure and removed unused fields.
RME: map configuration flash contents into software settings structure after flash is read.
RME: rename some methods to better reflect their function.
RME: documentation update to reflect the information embodied in these changes.

------------------------------------------------------------------------
r1588 | jwoithe | 2009-06-27 14:15:51 +0200 (Sat, 27 Jun 2009) | 2 lines

RME: differentiate between general sampling rate and that programmed into the device's DDS.  It's still not completely clear how these two distinct rates interact though.

------------------------------------------------------------------------
r1587 | jwoithe | 2009-06-26 15:32:57 +0200 (Fri, 26 Jun 2009) | 4 lines

RME: add functions to erase and program device flash.
RME: add stub for function to send software configuration to device flash.
RME: minor source formatting fixups.

------------------------------------------------------------------------
r1586 | jwoithe | 2009-06-26 14:43:09 +0200 (Fri, 26 Jun 2009) | 2 lines

RME: obtain and interpret TCO status from device.

------------------------------------------------------------------------
r1585 | jwoithe | 2009-06-25 16:26:17 +0200 (Thu, 25 Jun 2009) | 2 lines

RME: write user TCO settings to the device.

------------------------------------------------------------------------
r1584 | jwoithe | 2009-06-22 14:31:17 +0200 (Mon, 22 Jun 2009) | 2 lines

RME: Fill out TCO definitions.  Start work on TCO interface methods.

------------------------------------------------------------------------
r1583 | ppalmers | 2009-06-21 22:28:03 +0200 (Sun, 21 Jun 2009) | 1 line

fix regression introduced by r1577
------------------------------------------------------------------------
r1578 | jwoithe | 2009-06-21 07:46:03 +0200 (Sun, 21 Jun 2009) | 2 lines

MOTU: start documentation of the Cuemix protocol used in "Mark-3" devices.  Protocol analysis by MG and GQ.

------------------------------------------------------------------------
r1576 | ppalmers | 2009-06-20 11:14:50 +0200 (Sat, 20 Jun 2009) | 1 line

fix bug related to #227 where the user config file is not picked up by ffadomixer
------------------------------------------------------------------------
r1575 | jwoithe | 2009-06-15 14:29:38 +0200 (Mon, 15 Jun 2009) | 2 lines

RME: whitespace fix, clarify some comments

------------------------------------------------------------------------
r1574 | jwoithe | 2009-06-14 13:48:47 +0200 (Sun, 14 Jun 2009) | 3 lines

RME: Fix some typos.
RME: Separate device state initialisation and hardware parameter setting since the latter is needed whenever a single parameter is changed.

------------------------------------------------------------------------
r1573 | jwoithe | 2009-06-11 15:58:35 +0200 (Thu, 11 Jun 2009) | 2 lines

RME: add I/O functions for the TCO.  Clarify some comments.  Fix compile issue which wasn't picked up earlier to do an incorrect scons setting.

------------------------------------------------------------------------
r1572 | jwoithe | 2009-06-08 15:08:29 +0200 (Mon, 08 Jun 2009) | 2 lines

RME: more register address/value defines, mostly associated with the TCO

------------------------------------------------------------------------
r1570 | jwoithe | 2009-05-29 16:18:55 +0200 (Fri, 29 May 2009) | 3 lines

RME: add more details to device initialisation code.  The basic structure is now in place.
RME: update documentation.

------------------------------------------------------------------------
r1569 | holin | 2009-05-18 14:13:17 +0200 (Mon, 18 May 2009) | 2 lines

tidy up

------------------------------------------------------------------------
r1568 | holin | 2009-05-18 14:10:40 +0200 (Mon, 18 May 2009) | 2 lines

gcc 4.4 fixes (r1566, r1567, DICE) to trunk

------------------------------------------------------------------------
r1553 | jwoithe | 2009-05-16 14:29:27 +0200 (Sat, 16 May 2009) | 2 lines

RME: expand software status structure and begin to add details to device initialisation function.

------------------------------------------------------------------------
r1552 | ppalmers | 2009-05-10 17:04:32 +0200 (Sun, 10 May 2009) | 1 line

remove the requirement for the -p parameter to test-scs
------------------------------------------------------------------------
r1551 | ppalmers | 2009-05-10 16:52:04 +0200 (Sun, 10 May 2009) | 1 line

add dice EAP test tool
------------------------------------------------------------------------
r1550 | ppalmers | 2009-05-10 16:50:57 +0200 (Sun, 10 May 2009) | 5 lines

- Implement basic HSS1394 support for the Stanton SCS devices
- Start of the implementation of a generic TCAT DICE EAP control
- Reworked part of the busreset / ARM handler code

------------------------------------------------------------------------
r1548 | jwoithe | 2009-04-28 01:19:13 +0200 (Tue, 28 Apr 2009) | 2 lines

RME: add a brief comment to the configuration file explaining how the modelid field is used for RME device discovery.

------------------------------------------------------------------------
r1547 | jwoithe | 2009-04-28 01:13:39 +0200 (Tue, 28 Apr 2009) | 6 lines

RME:
- quickly fix device detection by comparing config file modelid against the unit's unit version since RME use unit version to differentiate models
- documentation update
- refinements to device defines
- begin to fill in device initialisation details

------------------------------------------------------------------------
r1543 | ppalmers | 2009-04-25 19:14:57 +0200 (Sat, 25 Apr 2009) | 4 lines

- Clean up class names
- Change probe code for all devices (except MOTU) to use the config file based approach

------------------------------------------------------------------------
r1542 | jwoithe | 2009-04-24 01:33:11 +0200 (Fri, 24 Apr 2009) | 2 lines

RME: more register definitions

------------------------------------------------------------------------
r1540 | jwoithe | 2009-04-21 08:22:27 +0200 (Tue, 21 Apr 2009) | 2 lines

RME: some additional register bit definitions

------------------------------------------------------------------------
r1539 | ppalmers | 2009-04-19 15:39:40 +0200 (Sun, 19 Apr 2009) | 4 lines

- Fix bug in dice transmit channel detection
- Implement snoop mode for DICE

------------------------------------------------------------------------
r1538 | ppalmers | 2009-04-14 09:18:13 +0200 (Tue, 14 Apr 2009) | 1 line

stale include
------------------------------------------------------------------------
r1537 | jwoithe | 2009-04-14 02:20:37 +0200 (Tue, 14 Apr 2009) | 2 lines

RME: more foundational work.  Confirmed that reading device flash works.

------------------------------------------------------------------------
r1535 | ppalmers | 2009-04-12 15:51:07 +0200 (Sun, 12 Apr 2009) | 1 line

add support for the FCA-202, and possibly other Oxford FW-92x devices
------------------------------------------------------------------------
r1534 | jwoithe | 2009-04-06 02:36:45 +0200 (Mon, 06 Apr 2009) | 2 lines

RME: more low-level infrastructure development and experimentation

------------------------------------------------------------------------
r1533 | jwoithe | 2009-03-31 00:48:16 +0200 (Tue, 31 Mar 2009) | 2 lines

RME: more low-level infrastructure

------------------------------------------------------------------------
r1532 | jwoithe | 2009-03-30 03:00:32 +0200 (Mon, 30 Mar 2009) | 4 lines

RME: documentation updates
RME: clean up some compile warnings
RME: introduce new defines for future work

------------------------------------------------------------------------
r1531 | ppalmers | 2009-03-29 22:59:09 +0200 (Sun, 29 Mar 2009) | 1 line

svn merge -r 1506:HEAD svn+ssh://ffadosvn@ffado.org/ffado/branches/libffado-2.0
------------------------------------------------------------------------
r1523 | jwoithe | 2009-03-23 01:37:36 +0200 (Mon, 23 Mar 2009) | 2 lines

MOTU: update protocol document to include "main out" and "phones" volume control registers.

------------------------------------------------------------------------
r1519 | jwoithe | 2009-03-16 01:14:54 +0200 (Mon, 16 Mar 2009) | 2 lines

RME: minor updates to the RME register map.

------------------------------------------------------------------------
r1515 | holin | 2009-02-11 13:41:24 +0200 (Wed, 11 Feb 2009) | 2 lines

Alesis Multimix16 support from adi (ticket #196)

------------------------------------------------------------------------
r1514 | ppalmers | 2008-12-27 21:50:51 +0200 (Sat, 27 Dec 2008) | 1 line

add Weiss engineering device ID's
------------------------------------------------------------------------
r1512 | arnonym | 2008-12-25 21:04:49 +0200 (Thu, 25 Dec 2008) | 4 lines

- Somehow there where two checks for the xdg-tools. One is enough.
 - Reword the failure message for the mixers a bit.
 - Don't touch the tests-dir when BUILD_TESTS is false. See #189

------------------------------------------------------------------------
r1511 | ppalmers | 2008-12-21 18:12:38 +0200 (Sun, 21 Dec 2008) | 3 lines

fix stream alignment

------------------------------------------------------------------------
r1510 | ppalmers | 2008-12-21 18:12:05 +0200 (Sun, 21 Dec 2008) | 3 lines

add Focusrite Saffire PRO 40 device id

------------------------------------------------------------------------
r1507 | jwoithe | 2008-12-09 06:25:41 +0200 (Tue, 09 Dec 2008) | 2 lines

MOTU: device documentation addition for Ultralite.

------------------------------------------------------------------------
r1504 | arnonym | 2008-12-07 21:08:50 +0200 (Sun, 07 Dec 2008) | 1 line

it looks nicer when there is only one layout. Otherwise there are three horizontal lines deviding upper and lower part, now its one.
------------------------------------------------------------------------
r1502 | ppalmers | 2008-12-07 18:04:10 +0200 (Sun, 07 Dec 2008) | 1 line

add chunk missing in merge
------------------------------------------------------------------------
r1501 | ppalmers | 2008-12-07 17:51:28 +0200 (Sun, 07 Dec 2008) | 1 line

fix merge issues with the QT4 mixer
------------------------------------------------------------------------
r1499 | ppalmers | 2008-12-07 16:00:32 +0200 (Sun, 07 Dec 2008) | 1 line

fix the build/install of the diag tool
------------------------------------------------------------------------
r1498 | ppalmers | 2008-12-07 15:50:41 +0200 (Sun, 07 Dec 2008) | 5 lines

Merge all changes from 2.0 branch into trunk (since r1361). This _should_ contain all forward merges done in the mean time. At this moment in time both branches should be in sync.

------------------------------------------------------------------------
r1493 | arnonym | 2008-12-01 00:53:26 +0200 (Mon, 01 Dec 2008) | 2 lines

Try to install a service-file for dbus if possible. This allows for dbus to start the ffado-dbus-server automaticly when someone wants to connect to org.ffado.Control.

------------------------------------------------------------------------
r1492 | arnonym | 2008-11-30 23:44:43 +0200 (Sun, 30 Nov 2008) | 2 lines

Install the logginghandler. And don't print the logmessage to stdout (which effectively doubled the debug-output on the console).

------------------------------------------------------------------------
r1491 | arnonym | 2008-11-30 23:43:16 +0200 (Sun, 30 Nov 2008) | 8 lines

multiple things:

- use MergeFlags instead of Append(Unique)
 - Add an PEDANTIC option to add -Werror to the flags
 - Execute the mixer-checks on clean, so the mixers actually get uninstalled
 - Remove the "HAVE_ALSA" thing. "ALSA_FLAGS" is enough to check for

------------------------------------------------------------------------
r1490 | jwoithe | 2008-11-28 07:06:35 +0200 (Fri, 28 Nov 2008) | 2 lines

Further updates to device documentation files.

------------------------------------------------------------------------
r1482 | arnonym | 2008-11-24 01:40:52 +0200 (Mon, 24 Nov 2008) | 1 line

ignore ffado-mixer instead of ffadomixer
------------------------------------------------------------------------
r1481 | arnonym | 2008-11-24 01:39:52 +0200 (Mon, 24 Nov 2008) | 2 lines

Add two logging-handlers: One to log to the statusbar (INFO and higher) and one to log to a textwidget. Which is included as a dock-widget.

------------------------------------------------------------------------
r1479 | arnonym | 2008-11-23 19:43:04 +0200 (Sun, 23 Nov 2008) | 2 lines

Forward port the python-module check from r1476.

------------------------------------------------------------------------
r1444 | jwoithe | 2008-11-17 04:35:56 +0200 (Mon, 17 Nov 2008) | 2 lines

More device documentation updates.

------------------------------------------------------------------------
r1442 | arnonym | 2008-11-17 01:30:46 +0200 (Mon, 17 Nov 2008) | 1 line

forward-port r1401: call the new ffado-mixer
------------------------------------------------------------------------
r1441 | arnonym | 2008-11-17 01:21:53 +0200 (Mon, 17 Nov 2008) | 2 lines

Forward-port r1429: Make the error more verbose

------------------------------------------------------------------------
r1440 | arnonym | 2008-11-17 01:20:47 +0200 (Mon, 17 Nov 2008) | 2 lines

Forward-port r1403: Install the xdg-icon the right way.

------------------------------------------------------------------------
r1439 | arnonym | 2008-11-17 01:20:17 +0200 (Mon, 17 Nov 2008) | 2 lines

I should go to bed now...

------------------------------------------------------------------------
r1438 | arnonym | 2008-11-17 01:17:22 +0200 (Mon, 17 Nov 2008) | 2 lines

Forward-port r1402 and r1404

------------------------------------------------------------------------
r1437 | arnonym | 2008-11-17 01:11:26 +0200 (Mon, 17 Nov 2008) | 2 lines

Forward-port r1421 and r1428

------------------------------------------------------------------------
r1435 | arnonym | 2008-11-17 01:05:38 +0200 (Mon, 17 Nov 2008) | 2 lines

forward port the mixer-changes from 2.0-branch r1361:HEAD

------------------------------------------------------------------------
r1424 | jwoithe | 2008-11-10 05:26:27 +0200 (Mon, 10 Nov 2008) | 2 lines

MOTU: add identification information for the 828 Mk 3

------------------------------------------------------------------------
r1420 | jwoithe | 2008-11-07 00:32:07 +0200 (Fri, 07 Nov 2008) | 2 lines

Updates and refinements to device protocol documentation.  More to come eventually.

------------------------------------------------------------------------
r1395 | arnonym | 2008-10-30 00:54:17 +0200 (Thu, 30 Oct 2008) | 2 lines

Forward-port the changes to the configure-checks to fix #69 (r1389 - r1393).

------------------------------------------------------------------------
r1361 | ppalmers | 2008-10-19 12:19:04 +0200 (Sun, 19 Oct 2008) | 5 lines

Merge 2.0 branch changes.

svn merge -r1349:HEAD svn+ssh://ffadosvn@ffado.org/ffado/branches/libffado-2.0

------------------------------------------------------------------------
r1358 | wagi | 2008-10-13 19:53:42 +0200 (Mon, 13 Oct 2008) | 1 line

fix #165
------------------------------------------------------------------------
r1354 | jwoithe | 2008-10-01 03:56:51 +0200 (Wed, 01 Oct 2008) | 2 lines

Minor documentation update.

------------------------------------------------------------------------
r1353 | ppalmers | 2008-09-30 13:36:20 +0200 (Tue, 30 Sep 2008) | 1 line

fix config file install location
------------------------------------------------------------------------
r1349 | wagi | 2008-09-27 13:29:43 +0200 (Sat, 27 Sep 2008) | 1 line

enalbe test-volumne build
------------------------------------------------------------------------
r1348 | ppalmers | 2008-09-24 18:45:50 +0200 (Wed, 24 Sep 2008) | 1 line

merge 2.0 branch changes to trunk. svn merge -r 1337:HEAD svn+ssh://ffadosvn@ffado.org/ffado/branches/libffado-2.0
------------------------------------------------------------------------
r1347 | ppalmers | 2008-09-24 18:42:55 +0200 (Wed, 24 Sep 2008) | 1 line

re #152: apply patch
------------------------------------------------------------------------
r1340 | wagi | 2008-09-23 23:51:42 +0200 (Tue, 23 Sep 2008) | 2 lines

add test program for seting the pan.

------------------------------------------------------------------------
r1336 | ppalmers | 2008-09-23 12:42:04 +0200 (Tue, 23 Sep 2008) | 13 lines

Bring trunk up to date with branches/libffado-2.0:

"""
svn merge -r 1254:1299 svn+ssh://ffadosvn@ffado.org/ffado/branches/libffado-2.0
svn merge -r 1301:1320 svn+ssh://ffadosvn@ffado.org/ffado/branches/libffado-2.0
svn merge -r 1322:1323 svn+ssh://ffadosvn@ffado.org/ffado/branches/libffado-2.0
svn merge -r 1329:HEAD svn+ssh://ffadosvn@ffado.org/ffado/branches/libffado-2.0
"""

Add getSupportedSamplingFrequencies() to DICE, RME and Metric Halo AvDevices

------------------------------------------------------------------------
r1333 | ppalmers | 2008-09-23 11:13:16 +0200 (Tue, 23 Sep 2008) | 1 line

closes #151: fix typo.
------------------------------------------------------------------------
r1332 | ppalmers | 2008-09-23 11:10:33 +0200 (Tue, 23 Sep 2008) | 1 line

re #152: applied the patch.
------------------------------------------------------------------------
r1328 | wagi | 2008-09-21 13:07:14 +0200 (Sun, 21 Sep 2008) | 2 lines

merge r1327 back to trunk

------------------------------------------------------------------------
r1326 | wagi | 2008-09-20 12:45:18 +0200 (Sat, 20 Sep 2008) | 2 lines

Merge r1325 back to trunk.

------------------------------------------------------------------------
r1324 | wagi | 2008-09-20 11:40:19 +0200 (Sat, 20 Sep 2008) | 2 lines

Merge changeset r1321 into mainline

------------------------------------------------------------------------
r1314 | jwoithe | 2008-08-19 01:21:21 +0200 (Tue, 19 Aug 2008) | 2 lines

MOTU: additions to protocol documentation, mainly related to the 896HD.

------------------------------------------------------------------------
r1308 | jwoithe | 2008-08-07 04:48:49 +0200 (Thu, 07 Aug 2008) | 2 lines

DOC: add some documentation about the RME firewire protocol which I've been working on occasionally.  There's still work to be done but getting some information out there at this stage seems sensible.  The last update to these documents was on 24 March 2008.

------------------------------------------------------------------------
r1274 | jwoithe | 2008-07-08 07:26:05 +0200 (Tue, 08 Jul 2008) | 4 lines

* MOTU: add a model identifer to the dbus interface to assist with per-model mixer customisations.
* MOTU: Begin some model-specific mixer variations.  Users of the 828Mk2 should now be able to set the level and boost for analog channels 1-4.
* MOTU: On models without AES/EBU the relevant entry in the mix destination now reads "MainOut" (since that's what "AES/EBU" did on those interfaces).

------------------------------------------------------------------------
r1265 | jwoithe | 2008-06-23 01:43:55 +0200 (Mon, 23 Jun 2008) | 2 lines

MOTU: major reworking of mixer dbus interface structure.  Matrix mixers are used (where it makes sense) to greatly reduce the number of dbus nodes required to a number well below the default limit.

------------------------------------------------------------------------
r1258 | jwoithe | 2008-06-17 01:08:48 +0200 (Tue, 17 Jun 2008) | 2 lines

Fix "scons -c".  The ALSA test must be run for cleanup as well as build.

------------------------------------------------------------------------
r1257 | jwoithe | 2008-06-16 01:31:33 +0200 (Mon, 16 Jun 2008) | 2 lines

MOTU: add controls for Cuemix mixers 2-4.  This adds a *lot* of additional dbus controls and depending on your system setup you may have to increase the max_match_rules_per_connection resource limit in your dbus session configuration file.  5120 works for me (but is probably gross overkill).

------------------------------------------------------------------------
r1256 | holin | 2008-06-13 23:19:02 +0200 (Fri, 13 Jun 2008) | 2 lines

gcc 4.3 fix

------------------------------------------------------------------------
r1255 | ppalmers | 2008-06-08 19:17:32 +0200 (Sun, 08 Jun 2008) | 1 line

improve ALSA plugin
------------------------------------------------------------------------
r1254 | ppalmers | 2008-06-08 16:34:59 +0200 (Sun, 08 Jun 2008) | 1 line

split config.h into config/version/debug_config to allow for faster compilation (splits dependencies)
------------------------------------------------------------------------
r1253 | ppalmers | 2008-06-08 16:12:45 +0200 (Sun, 08 Jun 2008) | 1 line

merge 2.0 changes back (svn merge -r 1245:HEAD svn+ssh://ffadosvn@ffado.org/ffado/branches/libffado-2.0)
------------------------------------------------------------------------
r1246 | ppalmers | 2008-06-08 14:19:39 +0200 (Sun, 08 Jun 2008) | 1 line

add dead handler detection
------------------------------------------------------------------------
r1245 | ppalmers | 2008-06-05 20:45:16 +0200 (Thu, 05 Jun 2008) | 1 line

merge 2.0 changes to trunk ( svn merge -r 1238:HEAD svn+ssh://ffadosvn@ffado.org/ffado/branches/libffado-2.0 )
------------------------------------------------------------------------
r1240 | ppalmers | 2008-06-01 18:15:34 +0200 (Sun, 01 Jun 2008) | 1 line

fix bugs in IPC comms. Add preliminary FFADO-IPC ALSA plugin
------------------------------------------------------------------------
r1239 | ppalmers | 2008-06-01 13:35:14 +0200 (Sun, 01 Jun 2008) | 4 lines

- Merge 2.0 branch changes r1225:1238 (svn merge -r 1225:HEAD svn+ssh://ffadosvn@ffado.org/ffado/branches/libffado-2.0)
 - Fix dummy MH device to compile nicely

------------------------------------------------------------------------
r1234 | holin | 2008-05-29 10:24:21 +0200 (Thu, 29 May 2008) | 2 lines

fix gcc 4.3 compile errors and some warnings (largely from Adrian Knoth)

------------------------------------------------------------------------
r1233 | holin | 2008-05-28 15:07:06 +0200 (Wed, 28 May 2008) | 2 lines

take CC and CXX from the environment if defined there

------------------------------------------------------------------------
r1232 | jwoithe | 2008-05-28 03:30:02 +0200 (Wed, 28 May 2008) | 2 lines

Add some #include directives seemingly required for newer versions of gcc.

------------------------------------------------------------------------
r1230 | jwoithe | 2008-05-28 01:49:31 +0200 (Wed, 28 May 2008) | 2 lines

Fix an incomplete sentence which somehow made it in to my previous README patch.

------------------------------------------------------------------------
r1225 | ppalmers | 2008-05-26 10:34:40 +0200 (Mon, 26 May 2008) | 1 line

merge libffado-2.0 changes r12121:1222 to trunk (svn merge -r 1221:1222 svn+ssh://ffadosvn@ffado.org/ffado/branches/libffado-2.0)
------------------------------------------------------------------------
r1218 | ppalmers | 2008-05-25 16:55:01 +0200 (Sun, 25 May 2008) | 1 line

add Alesis io|14 quirk (thx Adrian Knoth)
------------------------------------------------------------------------
r1217 | ppalmers | 2008-05-25 16:43:32 +0200 (Sun, 25 May 2008) | 1 line

merge libffado-2.0 changes r1213:1216 to trunk (svn merge -r 1213:HEAD svn+ssh://ffadosvn@ffado.org/ffado/branches/libffado-2.0)
------------------------------------------------------------------------
r1212 | ppalmers | 2008-05-25 11:12:22 +0200 (Sun, 25 May 2008) | 1 line

add a config option to globally disable the cache
------------------------------------------------------------------------
r1211 | ppalmers | 2008-05-25 11:07:01 +0200 (Sun, 25 May 2008) | 1 line

merge libffado-2.0 r1199:1206 back to trunk (svn merge -r 1199:1206 svn+ssh://ffadosvn@ffado.org/ffado/branches/libffado-2.0)
------------------------------------------------------------------------
r1210 | ppalmers | 2008-05-25 11:00:07 +0200 (Sun, 25 May 2008) | 1 line

add Onyx Mixer FireWire expansion device
------------------------------------------------------------------------
r1208 | holin | 2008-05-25 07:20:28 +0200 (Sun, 25 May 2008) | 2 lines

fix typo

------------------------------------------------------------------------
r1207 | holin | 2008-05-24 21:15:42 +0200 (Sat, 24 May 2008) | 2 lines

compiler warning fixes (from Adrian Knoth)

------------------------------------------------------------------------
r1205 | arnonym | 2008-05-23 23:08:59 +0200 (Fri, 23 May 2008) | 5 lines

work-around for #118: don't "wait" for 20 seconds (which is really a 20-times-retry-one-second-wait, but di a 2-tries-two-second-wait...

TODO: backport to the 2.0-branch
TODO: Implement a nice progressbar to give a gui-feedback...

------------------------------------------------------------------------
r1199 | wagi | 2008-05-22 19:44:44 +0200 (Thu, 22 May 2008) | 2 lines

merge libffado-2.0 changes 1195:1198 to trunk

------------------------------------------------------------------------
r1188 | ppalmers | 2008-05-19 19:09:59 +0200 (Mon, 19 May 2008) | 1 line

increase the time span used for debug timestamps
------------------------------------------------------------------------
r1187 | ppalmers | 2008-05-18 22:29:07 +0200 (Sun, 18 May 2008) | 1 line

get nb devices inside exception block
------------------------------------------------------------------------
r1186 | ppalmers | 2008-05-18 21:47:01 +0200 (Sun, 18 May 2008) | 1 line

slow down the reading of the config rom, it seems to confuse older bebob devices
------------------------------------------------------------------------
r1185 | arnonym | 2008-05-18 21:13:16 +0200 (Sun, 18 May 2008) | 3 lines

env.Copy() -> env.Clone()
  Makes scons 0.97 happy...

------------------------------------------------------------------------
r1184 | ppalmers | 2008-05-18 21:09:44 +0200 (Sun, 18 May 2008) | 1 line

add some extra saffire pro control elements
------------------------------------------------------------------------
r1183 | wagi | 2008-05-18 20:06:33 +0200 (Sun, 18 May 2008) | 3 lines

Slow down AV/C model discovering for the FA101. Maybe this fixes
the problem reported in #15 #108 and #109

------------------------------------------------------------------------
r1182 | arnonym | 2008-05-18 19:21:31 +0200 (Sun, 18 May 2008) | 1 line

somehow the tabs got moved. re-sort them...
------------------------------------------------------------------------
r1181 | ppalmers | 2008-05-18 19:14:33 +0200 (Sun, 18 May 2008) | 1 line

fixes #113
------------------------------------------------------------------------
r1180 | arnonym | 2008-05-18 19:08:54 +0200 (Sun, 18 May 2008) | 2 lines

Looks as if the old checks for gcc/g++ don't work reliably. Try a new one...

------------------------------------------------------------------------
r1179 | arnonym | 2008-05-18 18:54:15 +0200 (Sun, 18 May 2008) | 2 lines

Should fix the CFLAGS-issue...

------------------------------------------------------------------------
r1178 | ppalmers | 2008-05-18 18:48:58 +0200 (Sun, 18 May 2008) | 1 line

remove stale debug code
------------------------------------------------------------------------
r1177 | ppalmers | 2008-05-18 18:48:27 +0200 (Sun, 18 May 2008) | 1 line

bring maudio device up to date on config id changes
------------------------------------------------------------------------
r1176 | ppalmers | 2008-05-18 18:12:04 +0200 (Sun, 18 May 2008) | 1 line

fix small bug
------------------------------------------------------------------------
r1175 | ppalmers | 2008-05-18 18:09:51 +0200 (Sun, 18 May 2008) | 1 line

make the configuration ID longer, add terratec specific support for it
------------------------------------------------------------------------
r1174 | arnonym | 2008-05-17 15:31:30 +0200 (Sat, 17 May 2008) | 1 line

fix the layout? (see #110
------------------------------------------------------------------------
r1173 | ppalmers | 2008-05-15 19:37:07 +0200 (Thu, 15 May 2008) | 1 line

Fix issue with the AV/C function blocks not being set correctly. Fix clocksource selection bug. Expose control of verbose level through DBUS
------------------------------------------------------------------------
r1172 | ppalmers | 2008-05-15 06:38:31 +0200 (Thu, 15 May 2008) | 1 line

lay down the foundations for easy ALSA/Pulse support
------------------------------------------------------------------------
r1171 | ppalmers | 2008-05-13 21:11:59 +0200 (Tue, 13 May 2008) | 1 line

implement save to device for saffire LE
------------------------------------------------------------------------
r1170 | ppalmers | 2008-05-13 19:47:08 +0200 (Tue, 13 May 2008) | 1 line

fix bug introduced in r1167
------------------------------------------------------------------------
r1169 | ppalmers | 2008-05-13 19:30:00 +0200 (Tue, 13 May 2008) | 3 lines

-fix 64bit compilation issue

------------------------------------------------------------------------
r1168 | jwoithe | 2008-05-13 01:20:21 +0200 (Tue, 13 May 2008) | 2 lines

* MOTU: greatly increase the chances of having a clean shutdown of MOTU interfaces (that is, without the high-pitched squeal).

------------------------------------------------------------------------
r1167 | ppalmers | 2008-05-12 22:05:18 +0200 (Mon, 12 May 2008) | 1 line

impement level control for saffire LE outputs (cfr ticket #56)
------------------------------------------------------------------------
r1166 | ppalmers | 2008-05-12 19:37:08 +0200 (Mon, 12 May 2008) | 2 lines

- increase startup timeout for slow devices

------------------------------------------------------------------------
r1165 | ppalmers | 2008-05-12 02:58:19 +0200 (Mon, 12 May 2008) | 1 line

improve library behavior in bad situations
------------------------------------------------------------------------
r1164 | ppalmers | 2008-05-12 02:16:54 +0200 (Mon, 12 May 2008) | 1 line

fix small bug
------------------------------------------------------------------------
r1163 | ppalmers | 2008-05-12 00:04:25 +0200 (Mon, 12 May 2008) | 1 line

make dbus server handle busresets cleanly (fixes #102)
------------------------------------------------------------------------
r1162 | ppalmers | 2008-05-11 23:49:44 +0200 (Sun, 11 May 2008) | 1 line

forgot to actually call the busreset
------------------------------------------------------------------------
r1161 | ppalmers | 2008-05-11 18:44:58 +0200 (Sun, 11 May 2008) | 1 line

add busreset functionality
------------------------------------------------------------------------
r1160 | ppalmers | 2008-05-10 21:47:45 +0200 (Sat, 10 May 2008) | 1 line

remove debug leftovers
------------------------------------------------------------------------
r1159 | ppalmers | 2008-05-10 21:44:10 +0200 (Sat, 10 May 2008) | 1 line

reread the active sync info after setting it
------------------------------------------------------------------------
r1158 | ppalmers | 2008-05-10 21:38:38 +0200 (Sat, 10 May 2008) | 1 line

make control more thread safe (unfinished)
------------------------------------------------------------------------
r1157 | ppalmers | 2008-05-10 18:49:27 +0200 (Sat, 10 May 2008) | 1 line

fix race condition in debug module
------------------------------------------------------------------------
r1156 | ppalmers | 2008-05-10 18:21:47 +0200 (Sat, 10 May 2008) | 4 lines

- implement clocksource control for the phase88
- restructure the function block mess a bit.

------------------------------------------------------------------------
r1155 | ppalmers | 2008-05-10 17:40:39 +0200 (Sat, 10 May 2008) | 1 line

add some files forgotten in the previous commit
------------------------------------------------------------------------
r1154 | ppalmers | 2008-05-10 17:39:39 +0200 (Sat, 10 May 2008) | 1 line

add expat based parsing of the device cache. add compile-time selection between libxml++ and expat. will allow to get rid of the libxml++ dependency on the long run. scons SERIALIZE_USE_EXPAT=0Only compile testedscons SERIALIZE_USE_EXPAT=0
------------------------------------------------------------------------
r1153 | holin | 2008-05-10 08:51:40 +0200 (Sat, 10 May 2008) | 2 lines

add 64-bit build target

------------------------------------------------------------------------
r1152 | holin | 2008-05-09 22:43:37 +0200 (Fri, 09 May 2008) | 2 lines

fix optimized compile

------------------------------------------------------------------------
r1151 | holin | 2008-05-09 22:02:47 +0200 (Fri, 09 May 2008) | 2 lines

correct cpu type

------------------------------------------------------------------------
r1150 | holin | 2008-05-09 20:29:07 +0200 (Fri, 09 May 2008) | 2 lines

initial powerpc support for the build system

------------------------------------------------------------------------
r1149 | ppalmers | 2008-05-09 16:58:28 +0200 (Fri, 09 May 2008) | 1 line

lower debug level on too late packets
------------------------------------------------------------------------
r1148 | ppalmers | 2008-05-09 16:54:27 +0200 (Fri, 09 May 2008) | 1 line

add system peak load pulse generator
------------------------------------------------------------------------
r1147 | ppalmers | 2008-05-09 14:24:17 +0200 (Fri, 09 May 2008) | 3 lines

fix namespace issue

------------------------------------------------------------------------
r1146 | ppalmers | 2008-05-09 14:02:40 +0200 (Fri, 09 May 2008) | 1 line

reset errno when parsing arguments. fixes #107.
------------------------------------------------------------------------
r1145 | holin | 2008-05-09 13:12:05 +0200 (Fri, 09 May 2008) | 2 lines

error types are returned in errno not the return value

------------------------------------------------------------------------
r1144 | holin | 2008-05-09 12:21:47 +0200 (Fri, 09 May 2008) | 2 lines

cut out local integer typedefs

------------------------------------------------------------------------
r1143 | holin | 2008-05-09 11:14:05 +0200 (Fri, 09 May 2008) | 2 lines

use consistent types

------------------------------------------------------------------------
r1142 | holin | 2008-05-09 09:57:07 +0200 (Fri, 09 May 2008) | 2 lines

use correct types (fixes 64-bit breakage)

------------------------------------------------------------------------
r1141 | jwoithe | 2008-05-09 01:53:50 +0200 (Fri, 09 May 2008) | 2 lines

* MOTU: Activate the mix-1 channel controls for ADAT, AES/EBU and SPDIF channels to facilitate further testing.

------------------------------------------------------------------------
r1140 | holin | 2008-05-08 17:34:38 +0200 (Thu, 08 May 2008) | 2 lines

add byte swapping macros for constants, which avoids #ifdefs in other code and still preserves performance

------------------------------------------------------------------------
r1139 | holin | 2008-05-08 17:11:35 +0200 (Thu, 08 May 2008) | 2 lines

at least on T.C. Konnekts, strings seem to be stored as little endian quads

------------------------------------------------------------------------
r1138 | holin | 2008-05-08 13:09:10 +0200 (Thu, 08 May 2008) | 2 lines

fix big endian cip header

------------------------------------------------------------------------
r1137 | holin | 2008-05-08 12:55:19 +0200 (Thu, 08 May 2008) | 2 lines

fix typo

------------------------------------------------------------------------
r1136 | ppalmers | 2008-05-08 11:38:15 +0200 (Thu, 08 May 2008) | 1 line

keep bus direction in endian swapping functions
------------------------------------------------------------------------
r1135 | ppalmers | 2008-05-08 11:00:38 +0200 (Thu, 08 May 2008) | 1 line

centralize byteswapping and make it conditional
------------------------------------------------------------------------
r1134 | ppalmers | 2008-05-08 10:38:46 +0200 (Thu, 08 May 2008) | 1 line

revert r1131 since it's does unconditional byteswapping
------------------------------------------------------------------------
r1133 | ppalmers | 2008-05-07 22:54:22 +0200 (Wed, 07 May 2008) | 1 line

convert fireworks poll lock to stack based locking
------------------------------------------------------------------------
r1132 | ppalmers | 2008-05-07 22:39:26 +0200 (Wed, 07 May 2008) | 1 line

fix locking bug, convert to monitor
------------------------------------------------------------------------
r1131 | ppalmers | 2008-05-07 22:03:16 +0200 (Wed, 07 May 2008) | 4 lines

implement endian fixes (thx Heikki Lindholm)
switch from network host order byteswap routines to size based swap routines to remove ambiguity in definition of "long"

------------------------------------------------------------------------
r1130 | ppalmers | 2008-05-07 21:44:02 +0200 (Wed, 07 May 2008) | 1 line

fix some scons things
------------------------------------------------------------------------
r1129 | jwoithe | 2008-05-07 09:42:05 +0200 (Wed, 07 May 2008) | 2 lines

Correct a typo in the typo fix. :-)

------------------------------------------------------------------------
r1128 | jwoithe | 2008-05-07 09:39:51 +0200 (Wed, 07 May 2008) | 2 lines

Typo correction: it's SMPTE, not SMTPE.

------------------------------------------------------------------------
r1127 | arnonym | 2008-05-06 22:34:54 +0200 (Tue, 06 May 2008) | 2 lines

Try to introduce layouts for some mixers. Please test wether this works for you. Otherwise feel free to roll back...

------------------------------------------------------------------------
r1126 | ppalmers | 2008-05-06 11:02:30 +0200 (Tue, 06 May 2008) | 1 line

introduce smtpe clock source
------------------------------------------------------------------------
r1125 | jwoithe | 2008-05-06 01:04:08 +0200 (Tue, 06 May 2008) | 3 lines

* MOTU: fix mixer font sizes which got inadvertently removed in rev 1088 (but do it in a better way than the original).
* MOTU: implement basic clock source selection infrastructure.  Not all options apply to all devices - I will refine this soon.

------------------------------------------------------------------------
r1124 | wagi | 2008-05-05 23:29:40 +0200 (Mon, 05 May 2008) | 1 line

make it compile again and usable
------------------------------------------------------------------------
r1123 | ppalmers | 2008-05-05 11:18:18 +0200 (Mon, 05 May 2008) | 1 line

fix up clocksource definition a bit
------------------------------------------------------------------------
r1122 | jwoithe | 2008-05-05 03:18:03 +0200 (Mon, 05 May 2008) | 2 lines

* RME: interpret return values of read()/write() from Ieee1394Service correctly.

------------------------------------------------------------------------
r1121 | jwoithe | 2008-05-05 01:09:30 +0200 (Mon, 05 May 2008) | 3 lines

* MOTU: the Ieee1394Service read/write functions indicate an error by a return value <=0, not <0.
* tests/scan-devreg created to assist in identifying register usage when protocol analysis is not available.

------------------------------------------------------------------------
r1120 | ppalmers | 2008-05-05 00:38:49 +0200 (Mon, 05 May 2008) | 1 line

fixes #56. Adds some extra controls for the saffire mixer. remove clock widget, is global now.
------------------------------------------------------------------------
r1119 | ppalmers | 2008-05-05 00:24:17 +0200 (Mon, 05 May 2008) | 1 line

development version number
------------------------------------------------------------------------
r1117 | ppalmers | 2008-05-04 23:47:52 +0200 (Sun, 04 May 2008) | 1 line

2nd try for beta4
------------------------------------------------------------------------
r1116 | ppalmers | 2008-05-04 22:15:09 +0200 (Sun, 04 May 2008) | 1 line

properly fix [1060] and [1115]
------------------------------------------------------------------------
r1115 | wagi | 2008-05-04 21:44:12 +0200 (Sun, 04 May 2008) | 1 line

revert r1060. it looks rather wrong in my eyes. at least it won't work at the moment this way.
------------------------------------------------------------------------
r1113 | ppalmers | 2008-05-04 21:03:36 +0200 (Sun, 04 May 2008) | 1 line

switch to development version number
------------------------------------------------------------------------
r1111 | ppalmers | 2008-05-04 20:59:54 +0200 (Sun, 04 May 2008) | 1 line

version increase for beta4
------------------------------------------------------------------------
r1110 | wagi | 2008-05-03 22:04:27 +0200 (Sat, 03 May 2008) | 6 lines

Lot's of small and big bug fixes allong the caching path.
There were several parts not correctly deserialized.

This should fix ticket #65

------------------------------------------------------------------------
r1109 | wagi | 2008-05-03 22:01:37 +0200 (Sat, 03 May 2008) | 1 line

enable bebob discovering code for fa-101
------------------------------------------------------------------------
r1108 | ppalmers | 2008-05-03 17:07:52 +0200 (Sat, 03 May 2008) | 1 line

detect failure in setting clock source, and display error message.
------------------------------------------------------------------------
r1107 | ppalmers | 2008-05-03 16:50:36 +0200 (Sat, 03 May 2008) | 1 line

attempt to implement direct monitoring control and digital channel enable for saffire pro
------------------------------------------------------------------------
r1106 | arnonym | 2008-05-03 16:44:02 +0200 (Sat, 03 May 2008) | 2 lines

Not my day. Remove the modal parameter from the arguments.

------------------------------------------------------------------------
r1105 | arnonym | 2008-05-03 16:32:54 +0200 (Sat, 03 May 2008) | 4 lines

Add the copyright to the global mixer.

And maybe the mixers work when they aren't QDialogs but QWidgets. (Hard to debug such a thing with a device that has no mixer:-/)

------------------------------------------------------------------------
r1104 | arnonym | 2008-05-03 15:25:50 +0200 (Sat, 03 May 2008) | 2 lines

Connect the nickname changes to the dbus-action...

------------------------------------------------------------------------
r1103 | ppalmers | 2008-05-03 14:58:54 +0200 (Sat, 03 May 2008) | 1 line

lower debug output level of too late message
------------------------------------------------------------------------
r1102 | ppalmers | 2008-05-03 14:58:16 +0200 (Sat, 03 May 2008) | 1 line

version increase to indicate SVN
------------------------------------------------------------------------
r1100 | ppalmers | 2008-05-03 01:11:13 +0200 (Sat, 03 May 2008) | 1 line

update version for beta3. going to adapt the even=release, uneven=devel versioning scheme from now.
------------------------------------------------------------------------
r1099 | ppalmers | 2008-05-03 00:58:56 +0200 (Sat, 03 May 2008) | 6 lines

- transmit AMDTP packets earlier (3 to 4 packets instead of 6)
  adheres to the spec.
- extra's in debug code
- switched off all extended debugging features by default

------------------------------------------------------------------------
r1098 | ppalmers | 2008-05-03 00:53:20 +0200 (Sat, 03 May 2008) | 1 line

fix conceptually wrong use of timestamp value. re-enable late transmission since device should be able to handle that. If not they should generate an error. small fixes.
------------------------------------------------------------------------
r1097 | ppalmers | 2008-05-03 00:51:36 +0200 (Sat, 03 May 2008) | 1 line

fix conceptually wrong use of timestamp value
------------------------------------------------------------------------
r1096 | ppalmers | 2008-05-02 23:34:04 +0200 (Fri, 02 May 2008) | 1 line

fix up commit [1091]
------------------------------------------------------------------------
r1095 | ppalmers | 2008-05-02 23:32:04 +0200 (Fri, 02 May 2008) | 1 line

fixes #99. for real this time (committed wrong file)
------------------------------------------------------------------------
r1094 | arnonym | 2008-05-02 22:01:15 +0200 (Fri, 02 May 2008) | 1 line

White space changes
------------------------------------------------------------------------
r1093 | arnonym | 2008-05-02 21:50:43 +0200 (Fri, 02 May 2008) | 5 lines

* Add a generic part for each mixer controlling the nickname and the clocksource.
 * Don't install the generic mixer as its only intended for developers.
 * ignore the generated ffadomixer_config.py
 * Rework/resort the initialization of mixer-widgets...

------------------------------------------------------------------------
r1092 | ppalmers | 2008-05-02 21:00:59 +0200 (Fri, 02 May 2008) | 1 line

fixes #100
------------------------------------------------------------------------
r1091 | ppalmers | 2008-05-02 20:57:48 +0200 (Fri, 02 May 2008) | 1 line

fixes #99
------------------------------------------------------------------------
r1090 | ppalmers | 2008-05-02 12:23:59 +0200 (Fri, 02 May 2008) | 1 line

fix CTR reconstruction bug
------------------------------------------------------------------------
r1089 | arnonym | 2008-05-02 12:16:38 +0200 (Fri, 02 May 2008) | 2 lines

Its no problem if the generic mixer (which is only intended for devs) doesn't give anything useful for motu devices. But as the generic mixer is not one widget per device but one widget for all connected devices, it should generally get disabled just because one motu is detected.

------------------------------------------------------------------------
r1088 | arnonym | 2008-05-02 09:53:50 +0200 (Fri, 02 May 2008) | 4 lines

If the ffadomixer-window isn't updating to the mixers size, the reason is because the mixer-widget isn't telling its size. So either set the minimumSize of the widget or use layouts to build a dynamic-resizing gui (welcome to the 21th century :).

But don't work around it with some strange hacks of setting the size of the ffadomixer. These actually break when more then one device is connected and the two widgets have different sizes...

------------------------------------------------------------------------
r1087 | arnonym | 2008-05-02 09:44:10 +0200 (Fri, 02 May 2008) | 6 lines

Basic rules of design: _If_ you define a background color you have to
 a define it for the whole widget
 a define the foreground color too

And as both of these don't really blend with the desktops settings, we should do none of them...

------------------------------------------------------------------------
r1086 | jwoithe | 2008-05-02 01:17:18 +0200 (Fri, 02 May 2008) | 2 lines

Remove duplicate configuration directory existence check since Pieter beat me to it.

------------------------------------------------------------------------
r1085 | jwoithe | 2008-05-02 01:09:34 +0200 (Fri, 02 May 2008) | 7 lines

MOTU: Name channel pair control widgets in mixer.  Disable channel pair widgets for the moment since they are not implemented.
MOTU: In mixer python glue code, renamed pan object handlers to generic control handlers since they can be used for more than just the pan controls.
Fix minor grammar issues in registration dialog.
In registration module, ensure the ffado configuration directory exists before trying to save information into it.  The test is still a bit rough and could be made more robust.
ffadomixer: explicitly set the size of the toplevel tab widget to that of the mixer it contains.  Without this for me on Qt 3.3.8 the ffado mixer window is created way too small.  The solution is a bit hacky and could possibly do with refinement.
ffadomixer: don't include the generic mixer controls for MOTU devices since they don't give us anything useful yet and could be a source of confusion for users.

------------------------------------------------------------------------
r1084 | ppalmers | 2008-05-01 22:43:13 +0200 (Thu, 01 May 2008) | 1 line

update
------------------------------------------------------------------------
r1083 | ppalmers | 2008-05-01 21:09:32 +0200 (Thu, 01 May 2008) | 1 line

reduce verbosity of logging. re-enable the use of clock_nanosleep.
------------------------------------------------------------------------
r1082 | ppalmers | 2008-05-01 16:39:29 +0200 (Thu, 01 May 2008) | 1 line

revert some things checked in by accident
------------------------------------------------------------------------
r1081 | ppalmers | 2008-05-01 16:36:26 +0200 (Thu, 01 May 2008) | 1 line

work around clock_nanosleep issue (related to ticket #97) 
------------------------------------------------------------------------
r1080 | ppalmers | 2008-05-01 16:33:48 +0200 (Thu, 01 May 2008) | 1 line

use Mutex class for locking
------------------------------------------------------------------------
r1079 | ppalmers | 2008-05-01 16:32:32 +0200 (Thu, 01 May 2008) | 1 line

add nanosleep test
------------------------------------------------------------------------
r1078 | ppalmers | 2008-05-01 16:31:51 +0200 (Thu, 01 May 2008) | 1 line

improve lock trace debugging
------------------------------------------------------------------------
r1077 | ppalmers | 2008-05-01 16:18:48 +0200 (Thu, 01 May 2008) | 1 line

create config save directory if not present. change registration window looks.
------------------------------------------------------------------------
r1076 | wagi | 2008-05-01 15:42:33 +0200 (Thu, 01 May 2008) | 1 line

install also the generic mixer
------------------------------------------------------------------------
r1075 | wagi | 2008-05-01 15:33:44 +0200 (Thu, 01 May 2008) | 1 line

fix class definition syntax
------------------------------------------------------------------------
r1074 | ppalmers | 2008-04-30 17:24:55 +0200 (Wed, 30 Apr 2008) | 1 line

implement debug lock tracing
------------------------------------------------------------------------
r1073 | ppalmers | 2008-04-30 17:03:33 +0200 (Wed, 30 Apr 2008) | 1 line

workaround MOTU not setting a vendor/model name
------------------------------------------------------------------------
r1072 | ppalmers | 2008-04-30 16:30:39 +0200 (Wed, 30 Apr 2008) | 1 line

fix bug in saffire LE mixer code
------------------------------------------------------------------------
r1070 | ppalmers | 2008-04-29 17:29:27 +0200 (Tue, 29 Apr 2008) | 1 line

some small cosmetic fixes
------------------------------------------------------------------------
r1069 | ppalmers | 2008-04-29 14:28:14 +0200 (Tue, 29 Apr 2008) | 1 line

revert since it doesn't seem to work
------------------------------------------------------------------------
r1068 | ppalmers | 2008-04-29 14:25:00 +0200 (Tue, 29 Apr 2008) | 1 line

try and fix SVN version tracking (2)
------------------------------------------------------------------------
r1067 | ppalmers | 2008-04-29 14:21:45 +0200 (Tue, 29 Apr 2008) | 1 line

try and fix SVN version tracking
------------------------------------------------------------------------
r1066 | ppalmers | 2008-04-29 14:11:27 +0200 (Tue, 29 Apr 2008) | 1 line

the quatafire doesn't support setting a clock source, altough discovery gives some.
------------------------------------------------------------------------
r1065 | arnonym | 2008-04-29 12:42:24 +0200 (Tue, 29 Apr 2008) | 2 lines

Support the Element.Text in the generic mixer. So ppalmers can test wether his nicknames for devices work or not ;-)

------------------------------------------------------------------------
r1064 | jwoithe | 2008-04-29 01:56:06 +0200 (Tue, 29 Apr 2008) | 4 lines

* MOTU: disable optical SPDIF mode if device doesn't support it.
* MOTU: update protocol documentation.
* MOTU: rearrange mixer to put the channel pair switches in a more intelligent position (note that pairing functionality is still to be implemented).

------------------------------------------------------------------------
r1063 | ppalmers | 2008-04-29 00:26:08 +0200 (Tue, 29 Apr 2008) | 1 line

implement Nickname control element
------------------------------------------------------------------------
r1062 | arnonym | 2008-04-28 23:37:56 +0200 (Mon, 28 Apr 2008) | 8 lines

* Introduce a top-level QTabWidget that contains all the mixers.

* If the specific mixer has now specialized gui, the dummy is shown instead.

* The generic mixer is not installed anymore, if it is found (because ffadomixer is executed in that directory or because the files of the generic mixer aren't delete by hand), it will be added as well.

PS: Seems to be a good commit +237 lines, -242 lines

------------------------------------------------------------------------
r1061 | arnonym | 2008-04-28 20:51:44 +0200 (Mon, 28 Apr 2008) | 1 line

fix install and whitespaces
------------------------------------------------------------------------
r1060 | ppalmers | 2008-04-28 20:21:21 +0200 (Mon, 28 Apr 2008) | 4 lines

- implement (basic) user feedback (fixes #77)
- fix some small issues in the mixer code

------------------------------------------------------------------------
r1056 | jwoithe | 2008-04-28 02:56:21 +0200 (Mon, 28 Apr 2008) | 4 lines

* MOTU: Make optical mode control functional.  It still needs to have the toslink option disabled for the likes of the 896HD which don't have this capability.
* MOTU: Implement channel pad and trimgain controls.
* MOTU: Minor updates to protocol documentation.

------------------------------------------------------------------------
r1055 | ppalmers | 2008-04-27 21:07:42 +0200 (Sun, 27 Apr 2008) | 3 lines

correct bogus text

------------------------------------------------------------------------
r1054 | ppalmers | 2008-04-27 21:05:30 +0200 (Sun, 27 Apr 2008) | 3 lines

Simplify instructions

------------------------------------------------------------------------
r1053 | ppalmers | 2008-04-27 19:22:50 +0200 (Sun, 27 Apr 2008) | 1 line

fix small debug message bug
------------------------------------------------------------------------
r1052 | ppalmers | 2008-04-26 21:46:58 +0200 (Sat, 26 Apr 2008) | 1 line

remove alsa dependency. update readme file. add some extra stuff to ffado-diag.py
------------------------------------------------------------------------
r1048 | ppalmers | 2008-04-26 19:37:52 +0200 (Sat, 26 Apr 2008) | 1 line

pick a nice version number for alpha1
------------------------------------------------------------------------
r1047 | ppalmers | 2008-04-26 19:30:41 +0200 (Sat, 26 Apr 2008) | 1 line

remove compiler warnings. change non-critical warnings to debug messages.
------------------------------------------------------------------------
r1046 | ppalmers | 2008-04-26 18:42:39 +0200 (Sat, 26 Apr 2008) | 1 line

debug message cleanup. move #defines into config.h.in.
------------------------------------------------------------------------
r1045 | ppalmers | 2008-04-26 17:45:43 +0200 (Sat, 26 Apr 2008) | 1 line

use MutexLockHelper in SPM
------------------------------------------------------------------------
r1044 | ppalmers | 2008-04-26 17:35:32 +0200 (Sat, 26 Apr 2008) | 1 line

convert to Util::Mutex
------------------------------------------------------------------------
r1043 | ppalmers | 2008-04-26 17:20:56 +0200 (Sat, 26 Apr 2008) | 4 lines

- fix octlet byteswap bug for non-debug builds
- make Ieee1394Service thread safe

------------------------------------------------------------------------
r1042 | ppalmers | 2008-04-26 13:30:55 +0200 (Sat, 26 Apr 2008) | 4 lines

- use cluster infoblock information for stream setup
- implement graphviz output for music plugs

------------------------------------------------------------------------
r1041 | arnonym | 2008-04-26 10:55:48 +0200 (Sat, 26 Apr 2008) | 2 lines

Be more verbose by missing dependencies...

------------------------------------------------------------------------
r1040 | ppalmers | 2008-04-25 23:13:55 +0200 (Fri, 25 Apr 2008) | 1 line

lower number of interrupts generated. remove leftover debug line
------------------------------------------------------------------------
r1039 | ppalmers | 2008-04-25 22:59:49 +0200 (Fri, 25 Apr 2008) | 3 lines

- Fix non-debug build

------------------------------------------------------------------------
r1038 | ppalmers | 2008-04-25 22:39:03 +0200 (Fri, 25 Apr 2008) | 1 line

reduce the amount of system time read kernel calls
------------------------------------------------------------------------
r1037 | ppalmers | 2008-04-25 21:07:32 +0200 (Fri, 25 Apr 2008) | 1 line

use new pkt_ctr in receive SP, calculate it fully in the IsoHandler
------------------------------------------------------------------------
r1036 | ppalmers | 2008-04-25 20:52:18 +0200 (Fri, 25 Apr 2008) | 1 line

remove cycle var where not needed. calculate full pkt_ctr in IsoHandler
------------------------------------------------------------------------
r1035 | ppalmers | 2008-04-25 20:32:04 +0200 (Fri, 25 Apr 2008) | 1 line

remove unused arguments for receive SP's
------------------------------------------------------------------------
r1034 | ppalmers | 2008-04-25 20:08:09 +0200 (Fri, 25 Apr 2008) | 7 lines

- remove unused function call parameters in streamprocessors
- change cycle parameter to pkt_ctr parameter. pkt_ctr contains the CTR value when a packet was received, or an approximation of the CTR value when a packet will be sent. the idea is 
that this can be used for (instead of) the syt reconstruction now done in the SP's themselves. passing a complete CTR moves the reconstruction up in the hierarchy, hence allows for a 
more efficient implementation. especially since the hardware can usually provide these values automatically.
NOTE: at this time, the CTR only contains a valid cycle field. the seconds field is for a next commit.

------------------------------------------------------------------------
r1033 | ppalmers | 2008-04-25 19:38:22 +0200 (Fri, 25 Apr 2008) | 1 line

eliminate unused dropped argument
------------------------------------------------------------------------
r1032 | ppalmers | 2008-04-25 19:21:07 +0200 (Fri, 25 Apr 2008) | 1 line

some more shuffling. Eliminate the CTR read from the generic transmit SP.
------------------------------------------------------------------------
r1031 | ppalmers | 2008-04-25 18:39:51 +0200 (Fri, 25 Apr 2008) | 4 lines

reorganize handler code a bit. prepare to move the cycle timer read calls out of the SP callbacks into the 1394 subsystem. remove bit rot and clean up compiler warnings.

------------------------------------------------------------------------
r1030 | ppalmers | 2008-04-25 18:16:06 +0200 (Fri, 25 Apr 2008) | 1 line

improve backtrace functionality
------------------------------------------------------------------------
r1029 | ppalmers | 2008-04-25 17:09:56 +0200 (Fri, 25 Apr 2008) | 1 line

compile fix
------------------------------------------------------------------------
r1028 | ppalmers | 2008-04-25 17:07:43 +0200 (Fri, 25 Apr 2008) | 1 line

fix no-data packets for bebob's
------------------------------------------------------------------------
r1027 | ppalmers | 2008-04-25 16:04:52 +0200 (Fri, 25 Apr 2008) | 3 lines

simplify system time source class. should give a performance increase due to less function calls.

------------------------------------------------------------------------
r1026 | ppalmers | 2008-04-25 15:51:24 +0200 (Fri, 25 Apr 2008) | 1 line

performance optimization
------------------------------------------------------------------------
r1025 | ppalmers | 2008-04-25 15:37:21 +0200 (Fri, 25 Apr 2008) | 1 line

implement backtrace debugging
------------------------------------------------------------------------
r1024 | ppalmers | 2008-04-25 13:14:52 +0200 (Fri, 25 Apr 2008) | 1 line

implement verbose listing of cluster info and music plug info
------------------------------------------------------------------------
r1023 | ppalmers | 2008-04-24 20:47:08 +0200 (Thu, 24 Apr 2008) | 3 lines

add konnek8 id

------------------------------------------------------------------------
r1022 | ppalmers | 2008-04-24 20:46:43 +0200 (Thu, 24 Apr 2008) | 1 line

add ListDevices operation
------------------------------------------------------------------------
r1021 | ppalmers | 2008-04-24 19:23:02 +0200 (Thu, 24 Apr 2008) | 1 line

Allow to disable sending payload on no-data packets though config.h.in. The DICE-II devices cannot cope with them.
------------------------------------------------------------------------
r1020 | ppalmers | 2008-04-24 19:20:56 +0200 (Thu, 24 Apr 2008) | 3 lines

cope with situation where a cc_cycle < rcv_cycle

------------------------------------------------------------------------
r1019 | ppalmers | 2008-04-24 16:31:34 +0200 (Thu, 24 Apr 2008) | 3 lines

fix sanity check

------------------------------------------------------------------------
r1018 | ppalmers | 2008-04-24 16:29:04 +0200 (Thu, 24 Apr 2008) | 1 line

update tests
------------------------------------------------------------------------
r1017 | ppalmers | 2008-04-24 14:19:58 +0200 (Thu, 24 Apr 2008) | 3 lines

- add false wraparound detection and notification (not corrected yet)

------------------------------------------------------------------------
r1016 | ppalmers | 2008-04-24 14:13:02 +0200 (Thu, 24 Apr 2008) | 5 lines

- move lock to improve timing of ctr read. grabbing locks can take some time (esp. on non-rt kernels). we can read the data before locking anyway.
- add some debugging to track the quality of the CTR DLL updates
- relax the filter coeffs

------------------------------------------------------------------------
r1015 | ppalmers | 2008-04-24 13:49:33 +0200 (Thu, 24 Apr 2008) | 3 lines

add random sleep function to timesource (used for debugging)

------------------------------------------------------------------------
r1014 | ppalmers | 2008-04-23 22:19:57 +0200 (Wed, 23 Apr 2008) | 1 line

increase verbosity when printing device info
------------------------------------------------------------------------
r1013 | ppalmers | 2008-04-23 20:55:47 +0200 (Wed, 23 Apr 2008) | 1 line

fix bug introduced by newer ECHO firmware (>= 4.3)
------------------------------------------------------------------------
r1012 | ppalmers | 2008-04-23 20:09:33 +0200 (Wed, 23 Apr 2008) | 1 line

fix segfault when init fails
------------------------------------------------------------------------
r1011 | ppalmers | 2008-04-23 19:10:02 +0200 (Wed, 23 Apr 2008) | 1 line

finish ECHO firmware tools
------------------------------------------------------------------------
r1010 | ppalmers | 2008-04-23 16:14:45 +0200 (Wed, 23 Apr 2008) | 1 line

first set of ECHO firmware manipulation code
------------------------------------------------------------------------
r1009 | ppalmers | 2008-04-23 16:13:49 +0200 (Wed, 23 Apr 2008) | 3 lines

fiddle around with the argument parsing to ease the reuse by other firmware loaders

------------------------------------------------------------------------
r1008 | ppalmers | 2008-04-23 16:12:05 +0200 (Wed, 23 Apr 2008) | 3 lines

fix verbose level

------------------------------------------------------------------------
r1007 | ppalmers | 2008-04-23 12:23:04 +0200 (Wed, 23 Apr 2008) | 5 lines

move conditional compilation switch for RT-safe debug buffer to config.h.in

make the use of a non-rt buffer an error condition for non-debug builds. people (esp packagers) should use the rt-safe versions for normal builds.

------------------------------------------------------------------------
r1006 | jwoithe | 2008-04-22 00:56:53 +0200 (Tue, 22 Apr 2008) | 3 lines

MOTU: implement "phones source" control.
MOTU: commence work on optical mode control.  For various reasons this isn't functional yet.

------------------------------------------------------------------------
r1005 | ppalmers | 2008-04-21 10:27:47 +0200 (Mon, 21 Apr 2008) | 13 lines

Improve thread synchronisation. Switch back to separate threads for transmit and
receive since it is not possible to statically schedule things properly. One 
of the threads (i.e. the client thread) is out of our control, hence it's 
execution can't be controlled. Using separate threads and correct priorities 
will shift this problem to the OS. Note that the priority of the packet 
receive thread should be lower than the client thread (such that the client 
thread is woken ASAP), and the priority of the transmit thread should be 
higher than the client thread (such that packets are queued ASAP).
Extra benefit: multi-cores are used.

Some other startup improvements.

------------------------------------------------------------------------
r1004 | ppalmers | 2008-04-21 10:22:37 +0200 (Mon, 21 Apr 2008) | 1 line

add konnekt live device id
------------------------------------------------------------------------
r1003 | jwoithe | 2008-04-21 01:51:50 +0200 (Mon, 21 Apr 2008) | 11 lines

MOTU updates:
 * cleanup of "define" namespace.  All MOTU-related defines now start with "MOTU_" rather than a mix of this and "MOTUFW_".
 * Significant cleanup of motu mixer UI definition and python code.  Far less glue code is now needed.
 * Use generic binary switch control in mixer dbus interface where possibe.
 * Implement proof-of-concept input level/boost switches.
 * Provide mechanism to feed some device status back to the mixer application.  Currently this is done only at startup but in time we'll need a way to poll for some of it as the mixer runs.
 * When streaming is active, disable controls whose operation is incompatible with an active streaming system.
 * Adapt active channels in the mixer to the current device state.  The handling of optical input mode is still to be done.
 * Minor updates to MOTU protocol documentation.
 * Whitespace cleanup in mixer_motu.py for consistency with the "tab is 4 spaces" rule used elsewhere in FFADO's source code.

------------------------------------------------------------------------
r1002 | ppalmers | 2008-04-20 13:06:26 +0200 (Sun, 20 Apr 2008) | 1 line

small optimization
------------------------------------------------------------------------
r1001 | ppalmers | 2008-04-19 14:18:01 +0200 (Sat, 19 Apr 2008) | 3 lines

Improve streaming startup for better initial timestamps and locking.

------------------------------------------------------------------------
r1000 | ppalmers | 2008-04-18 17:42:24 +0200 (Fri, 18 Apr 2008) | 4 lines

add some error condition debugging

------------------------------------------------------------------------
r999 | ppalmers | 2008-04-18 17:36:39 +0200 (Fri, 18 Apr 2008) | 1 line

start of user diag (ticket #13)
------------------------------------------------------------------------
r998 | arnonym | 2008-04-16 22:46:05 +0200 (Wed, 16 Apr 2008) | 6 lines

Try to improve the generic mixer.

Moving the sliders now sends values to the ffado-dbus-server.

And the interface of Continuous is extended by getMinimum() and getMaximum() which should return the range of values this value can have. For now its just [-100,10]. The generic mixer gui adopts to the values returned by these functions...

------------------------------------------------------------------------
r997 | ppalmers | 2008-04-14 08:59:17 +0200 (Mon, 14 Apr 2008) | 1 line

fixes #93
------------------------------------------------------------------------
r996 | jwoithe | 2008-04-14 02:39:16 +0200 (Mon, 14 Apr 2008) | 2 lines

MOTU: add protocol notes to doc directory to ensure they are available for reference.  Having them only in the old motufw sf project doesn't make a lot of sense these days.

------------------------------------------------------------------------
r995 | jwoithe | 2008-04-14 02:35:31 +0200 (Mon, 14 Apr 2008) | 3 lines

MOTU: more additions to mixer to test required functionality.  Analog channel solo/mute switchs work now.  Mix bus mute, destination and fader controls are also functional.
MOTU: Added prototype mixer widgets for device control.  These don't do anything yet.

------------------------------------------------------------------------
r994 | arnonym | 2008-04-13 22:58:08 +0200 (Sun, 13 Apr 2008) | 6 lines

No more "Sorry, no supported mixer found."!

Implement a first generic mixer ui.

The layout isn't yet that good, the faders don't work (since I don't have a device with a supported mixer, I can't test that) and there is not yet special support for ConfigRom or DiscreteValues. More to come in the next days...

------------------------------------------------------------------------
r993 | ppalmers | 2008-04-09 19:04:46 +0200 (Wed, 09 Apr 2008) | 1 line

fix deadlock (hackish but working)
------------------------------------------------------------------------
r992 | jwoithe | 2008-04-07 01:19:31 +0200 (Mon, 07 Apr 2008) | 4 lines

Added support for getting firewire unit version via Dbus (needed by ffadomixer to differentiate MOTU models down the track).
Created a motu mixer in ffadomixer and commenced implementation of controls.  The precise nature of the implementation is still being fine-tuned and is subject to change.
Minor whitespace cleanups and spelling corrections.

------------------------------------------------------------------------
r991 | jwoithe | 2008-04-01 00:20:01 +0200 (Tue, 01 Apr 2008) | 3 lines

MOTU: allow incoming MIDI events to extend into the next period buffer if needed to maintain the 8-quadlet alignment requirement.
MOTU: optimise MIDI execution path with pipeline hints on selected conditionals.

------------------------------------------------------------------------
r990 | jwoithe | 2008-03-31 00:09:15 +0200 (Mon, 31 Mar 2008) | 4 lines

MOTU: make MIDI transmission actually work on the Traveler.
MOTU: parse and store device status stream.
MOTU: minor whitespace cleanups.

------------------------------------------------------------------------
r989 | jwoithe | 2008-03-28 07:48:53 +0200 (Fri, 28 Mar 2008) | 2 lines

MOTU: yet more MIDI treaks.  Thanks to Francois Ernoult for the extensive testing of the 828MkII in this regard.

------------------------------------------------------------------------
r988 | jwoithe | 2008-03-28 04:31:09 +0200 (Fri, 28 Mar 2008) | 2 lines

MOTU: more MIDI tweaks.

------------------------------------------------------------------------
r987 | jwoithe | 2008-03-28 01:43:31 +0200 (Fri, 28 Mar 2008) | 2 lines

Include information about the dependencies of ffado-mixer.  Those more knowledgeable in the required packages may need to make tweaks to the stated version numbers - the ones included are just wild guesses on my part.

------------------------------------------------------------------------
r986 | jwoithe | 2008-03-28 00:33:44 +0200 (Fri, 28 Mar 2008) | 2 lines

MOTU: more tweaks to MIDI code.

------------------------------------------------------------------------
r985 | jwoithe | 2008-03-27 05:51:28 +0200 (Thu, 27 Mar 2008) | 3 lines

* MOTU: Fix interaction with FFADO MIDI stream buffer system so incoming MIDI data has a chance of being recognised and outgoing data is sent in its entirety.
* MOTU: Make test for incoming MIDI data compatible with the 828MkII.

------------------------------------------------------------------------
r984 | jwoithe | 2008-03-27 04:56:34 +0200 (Thu, 27 Mar 2008) | 2 lines

MOTU: it seems the MIX1 return isn't present at 4x rates on the 896HD as for the Traveler, so the analog channels appear at different packet offsets at 4x rates also.

------------------------------------------------------------------------
r983 | jwoithe | 2008-03-27 01:12:15 +0200 (Thu, 27 Mar 2008) | 2 lines

fb_nodeid_t is 16 bit and fb_nodeaddr_t is 64 bit.  Correct the format strings in debug output to account for this so the correct length is reported.

------------------------------------------------------------------------
r982 | ppalmers | 2008-03-26 23:33:31 +0200 (Wed, 26 Mar 2008) | 1 line

small extra debug info
------------------------------------------------------------------------
r981 | jwoithe | 2008-03-25 01:00:49 +0200 (Tue, 25 Mar 2008) | 15 lines

config.h.in: reworked MOTU_* constants to get "nearly correct" timestamps in outgoing packets.  This makes playback at 2x and 4x samplerates work.  More fine-tuning may be needed.

motu_avdevice.cpp: account for different port locations in packet data at 4x rates.

motu_avdevice.cpp: MotuDevice::probe(), MotuDevice::createDevice(): remove commented out references to the ConfigRom ModelID field; with MOTUs this is useless for differentiating models.

motu_avdevice.cpp: MotuDevice::setOpticalMode(): the 896HD doesn't have an SPDIF/TOSLINK optical mode, so don't try to set it.

RME: implemented first cut at sample rate control.  This has about a 50% chance of working as it currently stands.

MOTU: minor whitespace fixes for consistency.

MOTU: start implementation of device status tracking.

------------------------------------------------------------------------
r980 | ppalmers | 2008-03-24 19:33:26 +0200 (Mon, 24 Mar 2008) | 3 lines

- add runaway checks to more RT threads

------------------------------------------------------------------------
r979 | ppalmers | 2008-03-24 19:22:54 +0200 (Mon, 24 Mar 2008) | 1 line

use efficient rounding
------------------------------------------------------------------------
r978 | ppalmers | 2008-03-24 19:10:24 +0200 (Mon, 24 Mar 2008) | 1 line

try to detect and survive ISO thread hangup
------------------------------------------------------------------------
r977 | ppalmers | 2008-03-24 19:09:11 +0200 (Mon, 24 Mar 2008) | 1 line

switch back to DMA bufferfill for receive
------------------------------------------------------------------------
r976 | ppalmers | 2008-03-24 19:08:43 +0200 (Mon, 24 Mar 2008) | 1 line

fix verbose level
------------------------------------------------------------------------
r975 | ppalmers | 2008-03-24 17:56:23 +0200 (Mon, 24 Mar 2008) | 1 line

fix debug message verbose level
------------------------------------------------------------------------
r974 | ppalmers | 2008-03-23 19:42:42 +0200 (Sun, 23 Mar 2008) | 1 line

fix small issues with quatafire mixer
------------------------------------------------------------------------
r973 | ppalmers | 2008-03-23 19:35:01 +0200 (Sun, 23 Mar 2008) | 1 line

make quatafire mixer more useful
------------------------------------------------------------------------
r972 | ppalmers | 2008-03-23 17:59:49 +0200 (Sun, 23 Mar 2008) | 3 lines

- fix error checking

------------------------------------------------------------------------
r971 | ppalmers | 2008-03-23 13:53:18 +0200 (Sun, 23 Mar 2008) | 1 line

first experimental quatafire mixer
------------------------------------------------------------------------
r970 | ppalmers | 2008-03-23 12:42:13 +0200 (Sun, 23 Mar 2008) | 1 line

catch problems with non-rt capable systems and the watchdog thread
------------------------------------------------------------------------
r969 | ppalmers | 2008-03-23 12:32:19 +0200 (Sun, 23 Mar 2008) | 1 line

fixes #85
------------------------------------------------------------------------
r968 | ppalmers | 2008-03-23 12:17:34 +0200 (Sun, 23 Mar 2008) | 1 line

revert function change for compatibility with dbus 1.0
------------------------------------------------------------------------
r967 | ppalmers | 2008-03-22 23:48:09 +0200 (Sat, 22 Mar 2008) | 2 lines

- first attempt at not causing total havoc when devices are removed from the bus.

------------------------------------------------------------------------
r966 | ppalmers | 2008-03-22 20:37:22 +0200 (Sat, 22 Mar 2008) | 3 lines

fixes ticket #83

------------------------------------------------------------------------
r965 | ppalmers | 2008-03-22 20:33:29 +0200 (Sat, 22 Mar 2008) | 1 line

add generic register read for focusrite devices
------------------------------------------------------------------------
r964 | ppalmers | 2008-03-20 10:58:55 +0200 (Thu, 20 Mar 2008) | 1 line

warning fix
------------------------------------------------------------------------
r963 | ppalmers | 2008-03-19 23:41:56 +0200 (Wed, 19 Mar 2008) | 4 lines

- add clock source control to all panels that don't have it yet
- fix saffire (LE) detection bug

------------------------------------------------------------------------
r962 | ppalmers | 2008-03-19 14:55:31 +0200 (Wed, 19 Mar 2008) | 1 line

apply patch of ticket #82
------------------------------------------------------------------------
r961 | jwoithe | 2008-03-19 01:21:23 +0200 (Wed, 19 Mar 2008) | 4 lines

Nitpick: officially MOTU is all capitals.
Make all sections of readme fall within 80 columns.
Minor grammatical corrections.

------------------------------------------------------------------------
r960 | ppalmers | 2008-03-19 00:45:38 +0200 (Wed, 19 Mar 2008) | 1 line

make dbus server handle bus resets and dynamic add/remove of devices
------------------------------------------------------------------------
r959 | ppalmers | 2008-03-18 23:01:14 +0200 (Tue, 18 Mar 2008) | 1 line

add sconstruct changes for clock source control
------------------------------------------------------------------------
r958 | ppalmers | 2008-03-18 22:52:22 +0200 (Tue, 18 Mar 2008) | 1 line

add clock source control to dbus
------------------------------------------------------------------------
r957 | ppalmers | 2008-03-18 22:50:52 +0200 (Tue, 18 Mar 2008) | 1 line

add adat/spdif channel disable for saffire pro
------------------------------------------------------------------------
r956 | ppalmers | 2008-03-18 22:38:50 +0200 (Tue, 18 Mar 2008) | 1 line

use more kernel-side transmit buffering to reduce effect of scheduling
------------------------------------------------------------------------
r955 | ppalmers | 2008-03-18 22:36:57 +0200 (Tue, 18 Mar 2008) | 1 line

update readme file
------------------------------------------------------------------------
r954 | ppalmers | 2008-03-18 15:15:49 +0200 (Tue, 18 Mar 2008) | 1 line

remove missing controls
------------------------------------------------------------------------
r953 | ppalmers | 2008-03-18 12:40:54 +0200 (Tue, 18 Mar 2008) | 1 line

fixes #76
------------------------------------------------------------------------
r952 | jwoithe | 2008-03-18 04:57:29 +0200 (Tue, 18 Mar 2008) | 2 lines

MOTU: fix a stupid typo.

------------------------------------------------------------------------
r951 | jwoithe | 2008-03-18 04:16:43 +0200 (Tue, 18 Mar 2008) | 2 lines

MOTU: clip float samples to +/- 1.0 to avoid nasty audio wraparound effects.

------------------------------------------------------------------------
r950 | jwoithe | 2008-03-18 03:35:35 +0200 (Tue, 18 Mar 2008) | 2 lines

MOTU: tweak 896HD channel layout.

------------------------------------------------------------------------
r949 | jwoithe | 2008-03-17 04:04:50 +0200 (Mon, 17 Mar 2008) | 3 lines

Add detection support for the 896HD (other functionality is still to be verified).
Fixed typos in some ADAT channel offsets for the Traveler and 828Mk2.

------------------------------------------------------------------------
r948 | jwoithe | 2008-03-17 00:13:45 +0200 (Mon, 17 Mar 2008) | 4 lines

MOTU: Use fast C99 lrint()/lrintf() for float-int casts in MOTU transmit stream if available.  There may be a neater way of detecting the suitability of these in SConstruct but what's there now seems to work.

MOTU: only output the test tone if the debug level is greater than 0.

------------------------------------------------------------------------
r947 | arnonym | 2008-03-16 20:17:58 +0200 (Sun, 16 Mar 2008) | 4 lines

Fix a typo.

And be more verbose at why the mixer isn't built...

------------------------------------------------------------------------
r946 | arnonym | 2008-03-16 20:06:01 +0200 (Sun, 16 Mar 2008) | 4 lines

Sadly I can't get the xdg-tools to respect the DESTDIR. :-(

For now we have to show a message that the packagers have to deal with this their own.

------------------------------------------------------------------------
r945 | ppalmers | 2008-03-16 13:28:07 +0200 (Sun, 16 Mar 2008) | 3 lines

Add mixer application for the Saffire LE (untested)

------------------------------------------------------------------------
r944 | arnonym | 2008-03-13 23:39:55 +0200 (Thu, 13 Mar 2008) | 2 lines

Second half of #70. printMessage is now the full version when debug is enabled and equals printMessaheShort when debug is off...

------------------------------------------------------------------------
r943 | arnonym | 2008-03-13 22:52:28 +0200 (Thu, 13 Mar 2008) | 4 lines

Some Whitespace changes.

And use printMessage instead of printf for the ConfigRomPrint. should fix #70

------------------------------------------------------------------------
r942 | ppalmers | 2008-03-12 21:54:46 +0200 (Wed, 12 Mar 2008) | 1 line

add support for the device spec strings (no more auto-grab-all)
------------------------------------------------------------------------
r941 | ppalmers | 2008-03-12 21:52:52 +0200 (Wed, 12 Mar 2008) | 1 line

make watchdog die faster upon shutdown
------------------------------------------------------------------------
r940 | ppalmers | 2008-03-12 16:05:53 +0200 (Wed, 12 Mar 2008) | 1 line

use RT watchdog for 1394 service threads
------------------------------------------------------------------------
r939 | ppalmers | 2008-03-12 15:26:23 +0200 (Wed, 12 Mar 2008) | 1 line

implement RT thread watchdog
------------------------------------------------------------------------
r938 | ppalmers | 2008-03-12 13:48:37 +0200 (Wed, 12 Mar 2008) | 1 line

implement static iso handler scheduling
------------------------------------------------------------------------
r937 | ppalmers | 2008-03-12 12:06:09 +0200 (Wed, 12 Mar 2008) | 1 line

cleanup some compilation warnings
------------------------------------------------------------------------
r936 | ppalmers | 2008-03-12 11:52:50 +0200 (Wed, 12 Mar 2008) | 1 line

add option to skip IEC61883 bandwidth allocation (bogus anyway)
------------------------------------------------------------------------
r935 | ppalmers | 2008-03-12 11:52:06 +0200 (Wed, 12 Mar 2008) | 1 line

preliminary SSE3 detection in SConstruct
------------------------------------------------------------------------
r934 | ppalmers | 2008-03-11 21:05:58 +0200 (Tue, 11 Mar 2008) | 1 line

add files forgotten in previous commit
------------------------------------------------------------------------
r933 | ppalmers | 2008-03-11 19:32:37 +0200 (Tue, 11 Mar 2008) | 1 line

implement focusrite device naming
------------------------------------------------------------------------
r932 | ppalmers | 2008-03-11 14:46:25 +0200 (Tue, 11 Mar 2008) | 4 lines

- implement bus reset handling for the CTR DLL
- move the functor header to libutil

------------------------------------------------------------------------
r931 | ppalmers | 2008-03-11 13:59:23 +0200 (Tue, 11 Mar 2008) | 1 line

change some defaults; cleanup
------------------------------------------------------------------------
r930 | ppalmers | 2008-03-11 10:37:08 +0200 (Tue, 11 Mar 2008) | 11 lines

Fix latency/reliability issue.
Preliminary tests indicate that this runs fine with
'''jackd -R -P60 -d firewire -v4 -p64 -n3'''
which corresponds to 331 frames of roundtrip latency on the quatafire.

For reference:
'''jackd -R -P60 -d freebob -p64 -n2'''
results in 355 frames for the same device

------------------------------------------------------------------------
r929 | jwoithe | 2008-03-11 01:15:09 +0200 (Tue, 11 Mar 2008) | 5 lines

Add model enumeration for RME devices.
Rename model ID field of supported device structure to unit_version since that's what it's being set to.
Add details of Fireface-400 device.
All this is yet to be compile-tested.

------------------------------------------------------------------------
r928 | jwoithe | 2008-03-11 00:21:26 +0200 (Tue, 11 Mar 2008) | 2 lines

Fixed generation of silent headers for MOTU.  Seems to improve shutdown reliability.  Although this fix makes a slightly incorrect assumption (that every iso packet has audio data) it seems to be sufficient for correct operation.

------------------------------------------------------------------------
r927 | ppalmers | 2008-03-10 16:09:44 +0200 (Mon, 10 Mar 2008) | 1 line

add some new system tests and tools
------------------------------------------------------------------------
r926 | ppalmers | 2008-03-10 16:09:23 +0200 (Mon, 10 Mar 2008) | 1 line

fix some more small bugs
------------------------------------------------------------------------
r925 | ppalmers | 2008-03-10 12:40:19 +0200 (Mon, 10 Mar 2008) | 1 line

small fixes
------------------------------------------------------------------------
r924 | arnonym | 2008-03-09 19:15:35 +0200 (Sun, 09 Mar 2008) | 1 line

ignore me
------------------------------------------------------------------------
r923 | arnonym | 2008-03-09 19:06:12 +0200 (Sun, 09 Mar 2008) | 2 lines

No more ping-pong commits because of the PORT_TO_USE!? This adds a command-option -p (or --port) to set the port to use. And it also adds -/--help to tell you this...

------------------------------------------------------------------------
r922 | ppalmers | 2008-03-09 18:48:26 +0200 (Sun, 09 Mar 2008) | 1 line

add some system test and debug tools
------------------------------------------------------------------------
r921 | ppalmers | 2008-03-09 17:20:19 +0200 (Sun, 09 Mar 2008) | 1 line

fix softirq list
------------------------------------------------------------------------
r920 | ppalmers | 2008-03-09 17:12:55 +0200 (Sun, 09 Mar 2008) | 1 line

extend listing to include softirqs
------------------------------------------------------------------------
r919 | ppalmers | 2008-03-08 22:12:18 +0200 (Sat, 08 Mar 2008) | 1 line

fix concurrency issue in cycle timer updater
------------------------------------------------------------------------
r918 | ppalmers | 2008-03-08 18:08:13 +0200 (Sat, 08 Mar 2008) | 1 line

fix bogus startup timestamp issue
------------------------------------------------------------------------
r917 | ppalmers | 2008-03-08 18:07:42 +0200 (Sat, 08 Mar 2008) | 1 line

update 1394 test
------------------------------------------------------------------------
r916 | ppalmers | 2008-03-08 16:00:14 +0200 (Sat, 08 Mar 2008) | 1 line

forgotten file
------------------------------------------------------------------------
r915 | ppalmers | 2008-03-08 15:42:48 +0200 (Sat, 08 Mar 2008) | 1 line

turn warning into verbose message
------------------------------------------------------------------------
r914 | ppalmers | 2008-03-08 15:29:58 +0200 (Sat, 08 Mar 2008) | 3 lines

- make the ctr read from dll lockless

------------------------------------------------------------------------
r913 | ppalmers | 2008-03-08 14:02:36 +0200 (Sat, 08 Mar 2008) | 1 line

remove bogus line
------------------------------------------------------------------------
r912 | ppalmers | 2008-03-08 13:56:56 +0200 (Sat, 08 Mar 2008) | 3 lines

fix bug in cycle timer dll

------------------------------------------------------------------------
r911 | ppalmers | 2008-03-08 13:56:18 +0200 (Sat, 08 Mar 2008) | 3 lines

Fix compilation with extreme debugging enabled

------------------------------------------------------------------------
r910 | ppalmers | 2008-03-07 22:08:56 +0200 (Fri, 07 Mar 2008) | 1 line

extend 1394 service test
------------------------------------------------------------------------
r909 | ppalmers | 2008-03-07 21:57:14 +0200 (Fri, 07 Mar 2008) | 1 line

fix port out of range bug
------------------------------------------------------------------------
r908 | jwoithe | 2008-03-07 04:54:19 +0200 (Fri, 07 Mar 2008) | 2 lines

Make the skeleton RME device driver compile again.

------------------------------------------------------------------------
r907 | ppalmers | 2008-03-06 21:18:43 +0200 (Thu, 06 Mar 2008) | 1 line

fix concurrency issue
------------------------------------------------------------------------
r906 | ppalmers | 2008-03-06 14:22:38 +0200 (Thu, 06 Mar 2008) | 1 line

implement support for kernel space skipped cycles notification (non-public experimental code)
------------------------------------------------------------------------
r905 | ppalmers | 2008-03-06 11:05:31 +0200 (Thu, 06 Mar 2008) | 1 line

convert some const values into defines
------------------------------------------------------------------------
r904 | ppalmers | 2008-03-06 00:24:49 +0200 (Thu, 06 Mar 2008) | 1 line

simplify threading. Each port now gets two threads: one for transmit and one for receive.
------------------------------------------------------------------------
r903 | ppalmers | 2008-03-05 16:25:21 +0200 (Wed, 05 Mar 2008) | 1 line

init env with shell env
------------------------------------------------------------------------
r902 | jwoithe | 2008-03-05 00:47:41 +0200 (Wed, 05 Mar 2008) | 2 lines

Add a comment to highlight the order sensitivity of DevicesProperty entries.  This should prevent issues like that fixed in rev901 from occuring again when new devices are added.

------------------------------------------------------------------------
r901 | f.ernoult | 2008-03-04 15:53:03 +0200 (Tue, 04 Mar 2008) | 2 lines

[MOTU] fix port initialization bug introduced with the 828mkI preliminary support

------------------------------------------------------------------------
r900 | wagi | 2008-03-02 22:15:55 +0200 (Sun, 02 Mar 2008) | 1 line

Use GUID instead node id for download application. Furthermore, factor out common code from both firmware downloaders
------------------------------------------------------------------------
r899 | wagi | 2008-03-02 22:12:06 +0200 (Sun, 02 Mar 2008) | 1 line

fix typo
------------------------------------------------------------------------
r898 | wagi | 2008-03-02 22:04:52 +0200 (Sun, 02 Mar 2008) | 1 line

add non-debug print function to ConfigRom
------------------------------------------------------------------------
r897 | jwoithe | 2008-02-27 01:47:17 +0200 (Wed, 27 Feb 2008) | 2 lines

Fix a stupid typo in the 828MkI model table entry.

------------------------------------------------------------------------
r896 | jwoithe | 2008-02-27 00:56:53 +0200 (Wed, 27 Feb 2008) | 2 lines

Add untested preliminary support for the 828 MkI.  This has not been tested against an actual device, nor has it been compile tested.  Channel mapping is probably wrong and we don't yet know whether the MkI's protocol is similar enough to the MkII's that things will just work.  This addition may however allow further tests to be done.

------------------------------------------------------------------------
r895 | ppalmers | 2008-02-25 22:05:31 +0200 (Mon, 25 Feb 2008) | 1 line

add irq listing tool
------------------------------------------------------------------------
r894 | ppalmers | 2008-02-24 22:37:04 +0200 (Sun, 24 Feb 2008) | 3 lines

fix initialization issue

------------------------------------------------------------------------
r893 | ppalmers | 2008-02-24 21:24:55 +0200 (Sun, 24 Feb 2008) | 1 line

improve AMDTP receive performance
------------------------------------------------------------------------
r892 | ppalmers | 2008-02-24 21:17:56 +0200 (Sun, 24 Feb 2008) | 1 line

sse clipping fix
------------------------------------------------------------------------
r891 | ppalmers | 2008-02-24 20:22:42 +0200 (Sun, 24 Feb 2008) | 1 line

revert to direct CTR read
------------------------------------------------------------------------
r890 | ppalmers | 2008-02-24 19:03:48 +0200 (Sun, 24 Feb 2008) | 1 line

simplify SSE clipping
------------------------------------------------------------------------
r889 | ppalmers | 2008-02-24 18:02:20 +0200 (Sun, 24 Feb 2008) | 1 line

remove hack
------------------------------------------------------------------------
r888 | ppalmers | 2008-02-24 17:52:15 +0200 (Sun, 24 Feb 2008) | 1 line

reduce verbosity
------------------------------------------------------------------------
r887 | ppalmers | 2008-02-24 17:40:16 +0200 (Sun, 24 Feb 2008) | 1 line

implement clipping for float data
------------------------------------------------------------------------
r886 | ppalmers | 2008-02-24 16:00:43 +0200 (Sun, 24 Feb 2008) | 1 line

fix license, up version
------------------------------------------------------------------------
r885 | ppalmers | 2008-02-24 15:57:51 +0200 (Sun, 24 Feb 2008) | 3 lines

fix debug message

------------------------------------------------------------------------
r884 | ppalmers | 2008-02-24 15:57:26 +0200 (Sun, 24 Feb 2008) | 1 line

switch to CTR read DLL as default
------------------------------------------------------------------------
r883 | ppalmers | 2008-02-20 23:00:50 +0200 (Wed, 20 Feb 2008) | 3 lines

fix initialization bug

------------------------------------------------------------------------
r882 | ppalmers | 2008-02-20 22:33:14 +0200 (Wed, 20 Feb 2008) | 1 line

use more sane value for late wakeup detection
------------------------------------------------------------------------
r881 | ppalmers | 2008-02-20 22:31:31 +0200 (Wed, 20 Feb 2008) | 1 line

remove jitter graph output debug prints
------------------------------------------------------------------------
r880 | ppalmers | 2008-02-20 21:24:27 +0200 (Wed, 20 Feb 2008) | 3 lines

- switch over from a high-bandwidth DLL to a low-bandwidth DLL after initial settling. Prediction performance is better than 100ticks (20 on average).

------------------------------------------------------------------------
r879 | ppalmers | 2008-02-20 20:27:57 +0200 (Wed, 20 Feb 2008) | 7 lines

- improve cycle timer DLL
 * make it robust against bogus CTR reads
 * remove offset introduced by delay between wakeup and CTR read

prediction performance is now around 100 ticks for a 50ms update loop.

------------------------------------------------------------------------
r878 | ppalmers | 2008-02-19 23:26:27 +0200 (Tue, 19 Feb 2008) | 1 line

fix nonmonotonic CTR reads (occurs on NEC card)
------------------------------------------------------------------------
r877 | ppalmers | 2008-02-19 23:25:47 +0200 (Tue, 19 Feb 2008) | 1 line

wait some longer for the dbus server to start
------------------------------------------------------------------------
r876 | jwoithe | 2008-01-28 13:02:29 +0200 (Mon, 28 Jan 2008) | 2 lines

Properly fix the segfault on closedown which was worked around in rev 875.  It turns out that ~Port() calls PortManager::unregister which in turn removes the port from the port vector while ~PortManger() still holds an iterator on the port vector.  This time we simply delete the port in ~PortManager(), leaving the formal unregistering as a side effect of ~Port().  Fix suggested by Pieter.

------------------------------------------------------------------------
r875 | jwoithe | 2008-01-25 13:26:21 +0200 (Fri, 25 Jan 2008) | 2 lines

Apply a fix which appears to prevent the segfault on shutdown, seen particularly with MOTU interfaces.  The fix may not be optimal or even valid - refer to mailing list post for full details.

------------------------------------------------------------------------
r874 | jwoithe | 2008-01-22 23:57:22 +0200 (Tue, 22 Jan 2008) | 2 lines

MAX_XMIT_PACKET_SIZE must be greater than 1160 or else the maximum allowed packet sizes are smaller than the packets required by MOTU interfaces.

------------------------------------------------------------------------
r873 | jwoithe | 2008-01-22 00:01:52 +0200 (Tue, 22 Jan 2008) | 3 lines

MOTU: Reinstate MIDI stream processing code removed during API cleanup.
MOTU: Implement silent audio packets in MOTU transmit stream processor.

------------------------------------------------------------------------
r872 | wagi | 2008-01-21 21:27:51 +0200 (Mon, 21 Jan 2008) | 1 line

let's try to discover the fa-101 with the descriptor discovering code
------------------------------------------------------------------------
r871 | wagi | 2008-01-21 21:03:21 +0200 (Mon, 21 Jan 2008) | 1 line

use debugOutput instead of debugWarning for not supported function blocks
------------------------------------------------------------------------
r870 | ppalmers | 2008-01-20 23:59:43 +0200 (Sun, 20 Jan 2008) | 1 line

try and work around buffer size issues in raw1394
------------------------------------------------------------------------
r869 | arnonym | 2008-01-20 19:37:51 +0200 (Sun, 20 Jan 2008) | 1 line

think I did enough in this script to allow a copyright. Otherwise just remove me again...
------------------------------------------------------------------------
r868 | ppalmers | 2008-01-20 15:05:57 +0200 (Sun, 20 Jan 2008) | 1 line

detect when a handler has died on us
------------------------------------------------------------------------
r867 | arnonym | 2008-01-20 15:04:51 +0200 (Sun, 20 Jan 2008) | 2 lines

Another DESTDIR fix (see #68)

------------------------------------------------------------------------
r866 | ppalmers | 2008-01-20 13:59:17 +0200 (Sun, 20 Jan 2008) | 4 lines

- weed out some unused functions
- introduce 'debugOutputExtreme' allowing to disable debug statements in the speed-sensitive sections. This should reduce the cpu load on a 'normal' debug build significantly.

------------------------------------------------------------------------
r865 | wagi | 2008-01-20 13:48:13 +0200 (Sun, 20 Jan 2008) | 9 lines

BeBoB::SubunitAudio::createFunctionBlock:
	It is no use to add a dummy FunctionBlockProcessing because
	then the function type is not set in FunctionBlockProcessing.
	When we try to discover the plugs attached to this function block
	it will fail. It's better just to skip them.

This fixes bug #15.

------------------------------------------------------------------------
r864 | ppalmers | 2008-01-20 12:29:17 +0200 (Sun, 20 Jan 2008) | 1 line

update license to GPLv2 or GPLv3 instead of GPLv2 or any later version. Update copyrights to reflect the new year
------------------------------------------------------------------------
r863 | ppalmers | 2008-01-20 00:36:47 +0200 (Sun, 20 Jan 2008) | 1 line

fix cycle timer DLL code
------------------------------------------------------------------------
r862 | ppalmers | 2008-01-19 22:09:55 +0200 (Sat, 19 Jan 2008) | 1 line

fix xrun recovery and robustify
------------------------------------------------------------------------
r861 | arnonym | 2008-01-19 21:50:42 +0200 (Sat, 19 Jan 2008) | 2 lines

More on #68: Install the python-stuff to the destdir too.

------------------------------------------------------------------------
r860 | ppalmers | 2008-01-19 20:47:41 +0200 (Sat, 19 Jan 2008) | 1 line

clean up synchronization in streamprocessor
------------------------------------------------------------------------
r859 | ppalmers | 2008-01-19 20:34:45 +0200 (Sat, 19 Jan 2008) | 1 line

enable/disable backlog from config.h.in
------------------------------------------------------------------------
r858 | ppalmers | 2008-01-19 20:33:37 +0200 (Sat, 19 Jan 2008) | 1 line

minor fixes
------------------------------------------------------------------------
r857 | ppalmers | 2008-01-19 14:11:20 +0200 (Sat, 19 Jan 2008) | 4 lines

Introduce distinction between empty and silent packets. When shutting down a transmit SP, we now send silent (= valid, but all audio muted) packets instead of empty (w/o 
payload) packets for a while. This should solve the shutdown issue with the motu's.

------------------------------------------------------------------------
r856 | ppalmers | 2008-01-17 23:55:20 +0200 (Thu, 17 Jan 2008) | 1 line

revert to old structure and implement SSE routines in there. This avoids memory accesses hence speeding things up
------------------------------------------------------------------------
r855 | arnonym | 2008-01-17 23:11:35 +0200 (Thu, 17 Jan 2008) | 1 line

move the xdg-stuff to an own distinct dir, so its easier for packagers to recognize:-P
------------------------------------------------------------------------
r854 | arnonym | 2008-01-17 13:08:17 +0200 (Thu, 17 Jan 2008) | 1 line

check for the qt python module too
------------------------------------------------------------------------
r853 | arnonym | 2008-01-17 13:06:38 +0200 (Thu, 17 Jan 2008) | 2 lines

Try to implement some DESTDIR support (#68). And install the desktop-file if no DESTDIR is set. And have different variables for PATHS to distinguish between use in code (paths in the installed system) and use in scons-scripts...

------------------------------------------------------------------------
r852 | arnonym | 2008-01-17 12:12:19 +0200 (Thu, 17 Jan 2008) | 2 lines

Too much object-oriented-programming for me. I keep on trying to generalize things. Here is another one...

------------------------------------------------------------------------
r851 | arnonym | 2008-01-17 11:30:40 +0200 (Thu, 17 Jan 2008) | 2 lines

Check for a C++-header (and the availability of a c++-compiler) too. see #69

------------------------------------------------------------------------
r850 | arnonym | 2008-01-17 11:24:08 +0200 (Thu, 17 Jan 2008) | 2 lines

Be a bit more verbose if something failed...

------------------------------------------------------------------------
r849 | ppalmers | 2008-01-17 00:19:48 +0200 (Thu, 17 Jan 2008) | 1 line

temp commit of some SSE/memory optimizations
------------------------------------------------------------------------
r848 | jwoithe | 2008-01-15 01:25:47 +0200 (Tue, 15 Jan 2008) | 2 lines

Remove duplicate SPDIF entries in Ultralite port definitions which used the offsets of the main-outs.

------------------------------------------------------------------------
r847 | ppalmers | 2008-01-14 20:59:06 +0200 (Mon, 14 Jan 2008) | 1 line

use SSE2 for byteswaps (20% faster than ntohl())
------------------------------------------------------------------------
r846 | ppalmers | 2008-01-14 20:37:01 +0200 (Mon, 14 Jan 2008) | 1 line

remove AGAIN useage (might cause deadlock)
------------------------------------------------------------------------
r845 | wagi | 2008-01-13 21:15:21 +0200 (Sun, 13 Jan 2008) | 2 lines

add note to enhanced mixer command that is non-functional

------------------------------------------------------------------------
r844 | wagi | 2008-01-13 21:12:15 +0200 (Sun, 13 Jan 2008) | 7 lines

FunctionBlockEnhancedMixer::discover: 
Disable discovering of enhanced mixer because all
device out there do not use, and all implementations
are buggy. So there is no point to use it. 
All 'mixer' functions are implemented with selector function blocks

------------------------------------------------------------------------
r843 | wagi | 2008-01-13 21:04:56 +0200 (Sun, 13 Jan 2008) | 2 lines

enable enhanced mixer test

------------------------------------------------------------------------
r842 | wagi | 2008-01-13 21:02:49 +0200 (Sun, 13 Jan 2008) | 2 lines

add enhanced mixer test

------------------------------------------------------------------------
r841 | ppalmers | 2008-01-13 19:17:58 +0200 (Sun, 13 Jan 2008) | 1 line

fix single ISO thread operation (1394 stack seems to be thread-unsafe)
------------------------------------------------------------------------
r840 | ppalmers | 2008-01-13 17:38:42 +0200 (Sun, 13 Jan 2008) | 1 line

fix uninitialized value bug
------------------------------------------------------------------------
r839 | ppalmers | 2008-01-13 17:21:29 +0200 (Sun, 13 Jan 2008) | 1 line

fix shutdown segfault
------------------------------------------------------------------------
r838 | wagi | 2008-01-12 18:12:19 +0200 (Sat, 12 Jan 2008) | 6 lines

[PATCH] Add cache option also to ffado-dbus-server

reported and fixed by slack.
Ticket #64

------------------------------------------------------------------------
r837 | ppalmers | 2008-01-12 16:46:11 +0200 (Sat, 12 Jan 2008) | 1 line

add generic mackie mixer control
------------------------------------------------------------------------
r836 | ppalmers | 2008-01-12 16:02:14 +0200 (Sat, 12 Jan 2008) | 1 line

fix pkgconfig not updated issue; increase version to allow for jack sync
------------------------------------------------------------------------
r834 | arnonym | 2008-01-11 10:26:17 +0200 (Fri, 11 Jan 2008) | 2 lines

Should fix the return value. see #60

------------------------------------------------------------------------
r833 | ppalmers | 2008-01-10 23:36:13 +0200 (Thu, 10 Jan 2008) | 1 line

merge api-cleanup branch (R808:832) into trunk
------------------------------------------------------------------------
r832 | ppalmers | 2008-01-10 23:21:10 +0200 (Thu, 10 Jan 2008) | 1 line

fix phase24 issue, closes #61
------------------------------------------------------------------------
r828 | wagi | 2008-01-06 22:30:53 +0200 (Sun, 06 Jan 2008) | 1 line

FunctionBlockEnhancedMixer::discover: enable code
------------------------------------------------------------------------
r827 | wagi | 2008-01-06 21:31:59 +0200 (Sun, 06 Jan 2008) | 1 line

untabify
------------------------------------------------------------------------
r826 | wagi | 2008-01-06 21:25:49 +0200 (Sun, 06 Jan 2008) | 2 lines

add feature to disable avc model cache use

------------------------------------------------------------------------
r825 | wagi | 2008-01-06 21:20:05 +0200 (Sun, 06 Jan 2008) | 3 lines

- moved cmd_serialize.h Util:: namespace to Util::Cmd:: in order to avoid ambiguities with serialize.h

------------------------------------------------------------------------
r824 | wagi | 2008-01-06 20:57:59 +0200 (Sun, 06 Jan 2008) | 4 lines

- moved cmd_serialize.h Util:: namespace to Util::Cmd:: in order to avoid ambiguities with serialize.h
- bebob: enhanced mixer stuff added (not finished)

------------------------------------------------------------------------
r823 | wagi | 2008-01-06 17:54:38 +0200 (Sun, 06 Jan 2008) | 1 line

add TAGS for emacs too
------------------------------------------------------------------------
r822 | arnonym | 2008-01-06 17:48:48 +0200 (Sun, 06 Jan 2008) | 1 line

fix again: always rebuild the tags file
------------------------------------------------------------------------
r820 | arnonym | 2008-01-06 17:31:06 +0200 (Sun, 06 Jan 2008) | 1 line

this time it should work for wagi :-)
------------------------------------------------------------------------
r819 | arnonym | 2008-01-06 17:14:12 +0200 (Sun, 06 Jan 2008) | 2 lines

Implement a tags-target for wagi:-)

------------------------------------------------------------------------
r814 | arnonym | 2008-01-05 23:04:45 +0200 (Sat, 05 Jan 2008) | 2 lines

Forgot the property. The ffadomixer should be an executable. Thus the .in should be an executable...

------------------------------------------------------------------------
r810 | wagi | 2008-01-05 19:48:51 +0200 (Sat, 05 Jan 2008) | 1 line

add edirol fa-66 control
------------------------------------------------------------------------
r807 | ppalmers | 2008-01-05 16:07:23 +0200 (Sat, 05 Jan 2008) | 1 line

more reliability things
------------------------------------------------------------------------
r806 | arnonym | 2008-01-05 01:29:45 +0200 (Sat, 05 Jan 2008) | 2 lines

This should be unneeded as the apps needed to create the dbus interfaces are added as dependencies...

------------------------------------------------------------------------
r805 | arnonym | 2008-01-05 01:25:59 +0200 (Sat, 05 Jan 2008) | 6 lines

Install a .desktop-file that adds ffadomixer to the menus of kde and gnome.

Check for xdg-desktop-menu to install the .desktop-file.

And use that test for checking for pyuic too. It seems better than checking for some (developer) python-module that is not really used. Should really fix #48.

------------------------------------------------------------------------
r804 | arnonym | 2008-01-04 22:29:19 +0200 (Fri, 04 Jan 2008) | 6 lines

Try to start ffado-dbus-server if it can't be found on the DBus.

This is probably not the right solution, it would be better to register the server as a dbus-service and request that service so dbus autostarts it.
And the error handling isn't that nice either. if ffado-dbus-server can't be started, it keeps on asking until you press "No".
And probably the time to wait for the next try is to short or to long...

------------------------------------------------------------------------
r803 | ppalmers | 2008-01-03 00:11:58 +0200 (Thu, 03 Jan 2008) | 1 line

more reliable streaming. hackish, but a start for a better implementation
------------------------------------------------------------------------
r802 | arnonym | 2008-01-03 00:04:54 +0200 (Thu, 03 Jan 2008) | 1 line

be more user-friendly and display message-boxes
------------------------------------------------------------------------
r801 | arnonym | 2008-01-02 22:00:52 +0200 (Wed, 02 Jan 2008) | 4 lines

/me is very stupid.

But now the mixers should work and install just fine...

------------------------------------------------------------------------
r800 | arnonym | 2008-01-02 21:52:39 +0200 (Wed, 02 Jan 2008) | 1 line

go away for now...
------------------------------------------------------------------------
r799 | wagi | 2008-01-02 20:26:00 +0200 (Wed, 02 Jan 2008) | 1 line

add BridgeCo Audio5 UI control
------------------------------------------------------------------------
r798 | ppalmers | 2008-01-01 23:27:35 +0200 (Tue, 01 Jan 2008) | 1 line

cleanup and reliability fixes
------------------------------------------------------------------------
r797 | ppalmers | 2008-01-01 22:24:37 +0200 (Tue, 01 Jan 2008) | 1 line

parameters for better latency performance
------------------------------------------------------------------------
r796 | ppalmers | 2008-01-01 21:31:39 +0200 (Tue, 01 Jan 2008) | 4 lines

- move #define constants to config.h.in
- switch receive handler over to packet-per-buffer mode to improve latency performance

------------------------------------------------------------------------
r795 | ppalmers | 2008-01-01 19:27:10 +0200 (Tue, 01 Jan 2008) | 1 line

small bugfixes + debug print changes
------------------------------------------------------------------------
r794 | ppalmers | 2008-01-01 19:09:24 +0200 (Tue, 01 Jan 2008) | 1 line

add options to teststreaming3 (fix)
------------------------------------------------------------------------
r793 | ppalmers | 2008-01-01 19:08:36 +0200 (Tue, 01 Jan 2008) | 1 line

add options to teststreaming3
------------------------------------------------------------------------
r792 | ppalmers | 2008-01-01 18:20:44 +0200 (Tue, 01 Jan 2008) | 1 line

make teststreaming3 a cpp app
------------------------------------------------------------------------
r791 | ppalmers | 2007-12-31 16:41:19 +0200 (Mon, 31 Dec 2007) | 1 line

try and get some better low-latency performance
------------------------------------------------------------------------
r790 | ppalmers | 2007-12-31 15:50:33 +0200 (Mon, 31 Dec 2007) | 1 line

add debugging code
------------------------------------------------------------------------
r789 | ppalmers | 2007-12-31 15:49:32 +0200 (Mon, 31 Dec 2007) | 3 lines

add more testing

------------------------------------------------------------------------
r788 | ppalmers | 2007-12-31 15:47:10 +0200 (Mon, 31 Dec 2007) | 1 line

fix unreliable streaming bug
------------------------------------------------------------------------
r787 | ppalmers | 2007-12-29 14:03:52 +0200 (Sat, 29 Dec 2007) | 1 line

clean up sleeping code
------------------------------------------------------------------------
r786 | ppalmers | 2007-12-29 14:02:04 +0200 (Sat, 29 Dec 2007) | 1 line

add sine output
------------------------------------------------------------------------
r785 | ppalmers | 2007-12-29 13:41:53 +0200 (Sat, 29 Dec 2007) | 1 line

shutdown fix for DICE and MOTU
------------------------------------------------------------------------
r784 | ppalmers | 2007-12-29 13:39:55 +0200 (Sat, 29 Dec 2007) | 1 line

fix shutdown bug
------------------------------------------------------------------------
r783 | ppalmers | 2007-12-29 12:18:54 +0200 (Sat, 29 Dec 2007) | 1 line

cleanup time/wait/sleep code
------------------------------------------------------------------------
r782 | ppalmers | 2007-12-29 12:17:37 +0200 (Sat, 29 Dec 2007) | 1 line

fix mutex macro's; don't run thread in old_style mode
------------------------------------------------------------------------
r781 | ppalmers | 2007-12-29 12:16:19 +0200 (Sat, 29 Dec 2007) | 1 line

add profile build option (no debug messages, optimization and debug symbols)
------------------------------------------------------------------------
r780 | arnonym | 2007-12-27 01:05:22 +0200 (Thu, 27 Dec 2007) | 2 lines

Rebuild config.h (and all it dependants) if options changed.

------------------------------------------------------------------------
r779 | arnonym | 2007-12-27 01:04:18 +0200 (Thu, 27 Dec 2007) | 2 lines

HAVE_SSE is always defined (from config.h). It changes its value from 0 to 1 if sse is supported...

------------------------------------------------------------------------
r778 | ppalmers | 2007-12-27 00:05:57 +0200 (Thu, 27 Dec 2007) | 1 line

some extra optimization flags
------------------------------------------------------------------------
r777 | ppalmers | 2007-12-26 23:59:21 +0200 (Wed, 26 Dec 2007) | 1 line

port freebob SSE optimized event encoding functions (untested)
------------------------------------------------------------------------
r776 | ppalmers | 2007-12-26 17:26:00 +0200 (Wed, 26 Dec 2007) | 1 line

try to fix deadlock / performace issues
------------------------------------------------------------------------
r775 | ppalmers | 2007-12-24 13:21:13 +0200 (Mon, 24 Dec 2007) | 3 lines

- sleep before AGAIN, fixed deadlock

------------------------------------------------------------------------
r774 | arnonym | 2007-12-23 23:50:24 +0200 (Sun, 23 Dec 2007) | 2 lines

Some configure-steps are needed for uninstall too.

------------------------------------------------------------------------
r773 | arnonym | 2007-12-23 23:47:06 +0200 (Sun, 23 Dec 2007) | 2 lines

No need for commented scons-code that is only a copy of the stuff in tests/SConscript

------------------------------------------------------------------------
r772 | arnonym | 2007-12-23 23:42:21 +0200 (Sun, 23 Dec 2007) | 4 lines

Some small things:
 - install the ui-files too. Don't know if they are needed...
 - Be a bit more verbose and userfriendly(?) on the commandline. This should all be done with some nice gui to be really userfriendly.

------------------------------------------------------------------------
r771 | arnonym | 2007-12-23 23:31:14 +0200 (Sun, 23 Dec 2007) | 4 lines

Almost too easy!

Fix the dependency libffado.* -> config.h -> config.h.in -> SConstruct

------------------------------------------------------------------------
r770 | arnonym | 2007-12-23 23:25:31 +0200 (Sun, 23 Dec 2007) | 8 lines

- Add a script to start mixers in a generic way.

- Install the specific mixers with their ui.py-files to $sharedir/python and use them.

- Add another "app" for testing. (/me doesn't have a device with a mixer, so I can't test the real mixers.)

- Ignore ffadomixer

------------------------------------------------------------------------
r769 | arnonym | 2007-12-23 23:21:27 +0200 (Sun, 23 Dec 2007) | 4 lines

Set the rights on the new file the same as the old files rights.

Needed to create python-executables with replaced variables. See next commits...

------------------------------------------------------------------------
r768 | ppalmers | 2007-12-23 17:22:20 +0200 (Sun, 23 Dec 2007) | 2 lines

increment version

------------------------------------------------------------------------
r767 | ppalmers | 2007-12-23 17:21:18 +0200 (Sun, 23 Dec 2007) | 1 line

code refactoring
------------------------------------------------------------------------
r766 | ppalmers | 2007-12-23 17:19:15 +0200 (Sun, 23 Dec 2007) | 1 line

introduce local references to frequently used objects
------------------------------------------------------------------------
r765 | ppalmers | 2007-12-23 16:50:41 +0200 (Sun, 23 Dec 2007) | 1 line

reduce lock contention
------------------------------------------------------------------------
r764 | wagi | 2007-12-23 15:31:32 +0200 (Sun, 23 Dec 2007) | 2 lines

Get the cycle timer ticks in old style. The new one is not stable yet.

------------------------------------------------------------------------
r763 | wagi | 2007-12-23 15:20:30 +0200 (Sun, 23 Dec 2007) | 2 lines

StreamProcessor::getPacket: fixed debug output message

------------------------------------------------------------------------
r762 | wagi | 2007-12-18 14:19:29 +0200 (Tue, 18 Dec 2007) | 7 lines

Bug fix for #53
  Somehow there was a major fuckup. It didn't make any sense at all.
  The deserializing code for the PcrPlug, ExternalPlugs and 
  m_inputConnections/m_outputConnections should now be consistend with 
  the serializing code.

------------------------------------------------------------------------
r761 | arnonym | 2007-12-16 23:25:22 +0200 (Sun, 16 Dec 2007) | 8 lines

Implement a usable(?) configure check to find pyuic. It checks for the module pyqtconfig. At least on gentoo that is installed by the same app/lib as the pyuic app. If we used qt4 the check could also check for other py-modules...

Refactor the config.guess into a config-check so the result gets cached.

And build the mixer-guis again.

Next step: Instaling the mixer apps in a sane way. /me thinks about it...

------------------------------------------------------------------------
r760 | ppalmers | 2007-12-15 21:56:37 +0200 (Sat, 15 Dec 2007) | 1 line

disable thread per handler mode
------------------------------------------------------------------------
r759 | ppalmers | 2007-12-15 21:36:27 +0200 (Sat, 15 Dec 2007) | 1 line

fix streaming problem
------------------------------------------------------------------------
r758 | ppalmers | 2007-12-09 19:49:13 +0200 (Sun, 09 Dec 2007) | 1 line

fix bug
------------------------------------------------------------------------
r757 | ppalmers | 2007-12-09 18:46:12 +0200 (Sun, 09 Dec 2007) | 1 line

fix xml error due to license
------------------------------------------------------------------------
r756 | ppalmers | 2007-12-08 22:26:29 +0200 (Sat, 08 Dec 2007) | 1 line

add optimization flags
------------------------------------------------------------------------
r755 | ppalmers | 2007-12-08 15:44:49 +0200 (Sat, 08 Dec 2007) | 1 line

add a scons option to build static versions of the tools
------------------------------------------------------------------------
r754 | ppalmers | 2007-12-04 00:26:22 +0200 (Tue, 04 Dec 2007) | 4 lines

- simplify IsoHandler
- fix some small issues

------------------------------------------------------------------------
r753 | ppalmers | 2007-12-03 00:52:52 +0200 (Mon, 03 Dec 2007) | 1 line

have separate threads for every handler
------------------------------------------------------------------------
r752 | ppalmers | 2007-12-02 17:49:37 +0200 (Sun, 02 Dec 2007) | 5 lines

- Implement a DLL based mechanism to read the cycle timer. This can potentially be more lightweight for the reader threads since it avoids a the CTR read kernel call. It also has the 
side effect that FFADO now works on older kernels that don't implement the cycle timer read call.

------------------------------------------------------------------------
r751 | ppalmers | 2007-12-01 15:42:30 +0200 (Sat, 01 Dec 2007) | 1 line

implement cache versioning
------------------------------------------------------------------------
r750 | ppalmers | 2007-12-01 00:18:26 +0200 (Sat, 01 Dec 2007) | 3 lines

Code refactoring. Tries to simplify things and tries to put all code where it belongs.

------------------------------------------------------------------------
r749 | jwoithe | 2007-11-30 15:23:37 +0200 (Fri, 30 Nov 2007) | 5 lines

StreamProcessor::transferSilence(): don't assume event size is one quadlet when allocating dummy buffer - not all interfaces work with a frame size of 1 quadlet.
MOTU: restore test tone to facilitate testing.
StreamProcessor::StreamProcessor(): re-order initialisations to prevent compiler warnings.

------------------------------------------------------------------------
r748 | ppalmers | 2007-11-30 00:26:52 +0200 (Fri, 30 Nov 2007) | 1 line

try to reorganize things such that less information is duplicated
------------------------------------------------------------------------
r747 | ppalmers | 2007-11-29 23:56:38 +0200 (Thu, 29 Nov 2007) | 3 lines

- get rid of IsoStream class since it's only overhead and doesn't really do anything specific. Most of it's functionality is going somewhere else anyway.

------------------------------------------------------------------------
r746 | ppalmers | 2007-11-29 23:25:06 +0200 (Thu, 29 Nov 2007) | 3 lines

- fix issue that prevented startup on my NEC based host controller.

------------------------------------------------------------------------
r745 | ppalmers | 2007-11-29 22:24:38 +0200 (Thu, 29 Nov 2007) | 1 line

implement clock source selection for DICE EVM
------------------------------------------------------------------------
r744 | ppalmers | 2007-11-29 14:52:16 +0200 (Thu, 29 Nov 2007) | 1 line

ensure that a mixer is built when devices are loaded from cache
------------------------------------------------------------------------
r743 | ppalmers | 2007-11-29 14:33:24 +0200 (Thu, 29 Nov 2007) | 1 line

add the GPL v3 license text
------------------------------------------------------------------------
r742 | ppalmers | 2007-11-29 14:25:10 +0200 (Thu, 29 Nov 2007) | 10 lines

- Remove some obsolete support files and dirs

- Clean up the license statements in the source files. Everything is
  GPL version 3 now.

- Add license and copyright notices to scons scripts

- Clean up some other text files

------------------------------------------------------------------------
r741 | ppalmers | 2007-11-29 00:40:32 +0200 (Thu, 29 Nov 2007) | 1 line

Fixes #54
------------------------------------------------------------------------
r740 | ppalmers | 2007-11-29 00:06:04 +0200 (Thu, 29 Nov 2007) | 1 line

fix propagation of debug/verbose level
------------------------------------------------------------------------
r739 | ppalmers | 2007-11-28 22:02:08 +0200 (Wed, 28 Nov 2007) | 10 lines

- Adapt the ffado external API (upgrade to v3)
  NEEDS NEW JACK BACKEND
- simplify FFADODevice constructor even more
- implement first framework support for supporting multiple adapters.
  currently all firewire adapters are scanned for supported devices unless specified otherwise
  however attaching devices to separate adapters is not supported. using multiple adapters at
  that are connected together might work.

------------------------------------------------------------------------
r738 | ppalmers | 2007-11-28 17:21:28 +0200 (Wed, 28 Nov 2007) | 1 line

fix bogus use of pointer in descriptor reading
------------------------------------------------------------------------
r737 | ppalmers | 2007-11-28 15:38:57 +0200 (Wed, 28 Nov 2007) | 3 lines

Add Acoustic Reality eAR Master One to device id's

------------------------------------------------------------------------
r736 | ppalmers | 2007-11-28 15:36:19 +0200 (Wed, 28 Nov 2007) | 1 line

fix some stuff that was not merged correctly
------------------------------------------------------------------------
r734 | ppalmers | 2007-11-28 15:03:31 +0200 (Wed, 28 Nov 2007) | 1 line

merge ppalmers-streaming branch
------------------------------------------------------------------------
r728 | wagi | 2007-11-25 23:07:25 +0200 (Sun, 25 Nov 2007) | 2 lines

getConfigurationIdSyncMode: consider also an address as SignalUnitAddress. this should fix a error message for saffire users.

------------------------------------------------------------------------
r726 | wagi | 2007-11-25 22:19:41 +0200 (Sun, 25 Nov 2007) | 2 lines

getConfigurationIdSyncMode: debug message corrected

------------------------------------------------------------------------
r725 | wagi | 2007-11-25 21:55:45 +0200 (Sun, 25 Nov 2007) | 2 lines

AvDevice::loadFromCache: remove debugging code

------------------------------------------------------------------------
r718 | arnonym | 2007-11-18 23:14:51 +0200 (Sun, 18 Nov 2007) | 5 lines

Disable the pyuic-test. its wrong.

And maybe try to create the sharedir. scons should to it automaticly but who knows. at least esox seemed to have problems with that...

------------------------------------------------------------------------
r717 | arnonym | 2007-11-17 17:41:37 +0200 (Sat, 17 Nov 2007) | 2 lines

This should fix redeemans problem with dbus and other libs in a non-standard place.

------------------------------------------------------------------------
r716 | wagi | 2007-11-11 22:03:19 +0200 (Sun, 11 Nov 2007) | 2 lines

cache bebob avc model. lot's of small bug fixed. it should work now.

------------------------------------------------------------------------
r711 | ppalmers | 2007-11-06 18:00:07 +0200 (Tue, 06 Nov 2007) | 1 line

disable unclear check
------------------------------------------------------------------------
r710 | ppalmers | 2007-11-06 17:52:11 +0200 (Tue, 06 Nov 2007) | 1 line

fix for configrom parsing bug (freebob backport)
------------------------------------------------------------------------
r698 | ppalmers | 2007-11-04 13:19:01 +0200 (Sun, 04 Nov 2007) | 4 lines

- make backlog a compile time option
- do some performance optimization on the debugging code

------------------------------------------------------------------------
r693 | ppalmers | 2007-11-03 20:13:53 +0200 (Sat, 03 Nov 2007) | 2 lines

format the message before sending it to the debugModuleManager such that there can't be any interference between messages of different debugModules.

------------------------------------------------------------------------
r692 | ppalmers | 2007-11-03 18:40:05 +0200 (Sat, 03 Nov 2007) | 3 lines

r691, does exactly the opposite of what it should do. fix this.

------------------------------------------------------------------------
r691 | ppalmers | 2007-11-03 18:38:59 +0200 (Sat, 03 Nov 2007) | 3 lines

- fix RT safety

------------------------------------------------------------------------
r690 | ppalmers | 2007-11-03 18:38:39 +0200 (Sat, 03 Nov 2007) | 6 lines

some small changes on the backlog:
- locks the flush_lock such that the backlog is not interrupted by normal messages
- allows to specify a minimum level the messages should have to be put in the backlog
- add another level of debugging

------------------------------------------------------------------------
r689 | ppalmers | 2007-11-03 17:12:39 +0200 (Sat, 03 Nov 2007) | 1 line

more echo firmware work
------------------------------------------------------------------------
r688 | ppalmers | 2007-11-03 14:31:51 +0200 (Sat, 03 Nov 2007) | 7 lines

- Remove bebob sync tool, since this is now generalized into the FFADODevice API.
  the sync source can be set with tests/test-ffado
- introduce a new directory for firmware tools: support/firmware
- Move the bridgeco firmware downloader to the support/firmware directory
- initial code for echo firmware downloading

------------------------------------------------------------------------
r687 | ppalmers | 2007-11-03 01:11:33 +0200 (Sat, 03 Nov 2007) | 1 line

implement Flash routines
------------------------------------------------------------------------
r686 | ppalmers | 2007-11-03 00:21:21 +0200 (Sat, 03 Nov 2007) | 1 line

add device ID's
------------------------------------------------------------------------
r685 | ppalmers | 2007-11-02 17:39:54 +0200 (Fri, 02 Nov 2007) | 1 line

remove device table printout
------------------------------------------------------------------------
r684 | ppalmers | 2007-11-02 14:04:08 +0200 (Fri, 02 Nov 2007) | 3 lines

- extend the saffire pro mixer

------------------------------------------------------------------------
r683 | ppalmers | 2007-11-02 11:46:31 +0200 (Fri, 02 Nov 2007) | 3 lines

- fix bug introduced in r673

------------------------------------------------------------------------
r682 | ppalmers | 2007-11-02 01:14:43 +0200 (Fri, 02 Nov 2007) | 1 line

extend saffire pro mixer
------------------------------------------------------------------------
r681 | ppalmers | 2007-11-01 23:53:23 +0200 (Thu, 01 Nov 2007) | 1 line

move truncated warning to the message itself instead of printf on stdout
------------------------------------------------------------------------
r680 | ppalmers | 2007-11-01 23:35:26 +0200 (Thu, 01 Nov 2007) | 1 line

Fix samplerate selection for the saffire pro
------------------------------------------------------------------------
r679 | ppalmers | 2007-11-01 23:28:35 +0200 (Thu, 01 Nov 2007) | 1 line

workaround for busreset strangeness in libraw1394
------------------------------------------------------------------------
r678 | ppalmers | 2007-11-01 23:27:55 +0200 (Thu, 01 Nov 2007) | 5 lines

Fix bug where the internal data structures of the AVC::Unit were not cleared before discovery. This 
prevented devices from being re-discovered since all old structures remained (e.g. all of a sudden 
there are 2 music subunits)

------------------------------------------------------------------------
r677 | ppalmers | 2007-11-01 23:26:30 +0200 (Thu, 01 Nov 2007) | 3 lines

make generation unambiguous.

------------------------------------------------------------------------
r676 | ppalmers | 2007-11-01 18:07:17 +0200 (Thu, 01 Nov 2007) | 4 lines

- Add direct address space control mechanism
- Add extra device control id's

------------------------------------------------------------------------
r675 | ppalmers | 2007-11-01 18:06:02 +0200 (Thu, 01 Nov 2007) | 1 line

add device id for M-Audio ProFire Lightbridge
------------------------------------------------------------------------
r674 | ppalmers | 2007-11-01 18:04:57 +0200 (Thu, 01 Nov 2007) | 1 line

Introduce bus reset handling
------------------------------------------------------------------------
r673 | ppalmers | 2007-11-01 18:04:02 +0200 (Thu, 01 Nov 2007) | 1 line

fix bus reset behavior
------------------------------------------------------------------------
r672 | ppalmers | 2007-11-01 13:45:14 +0200 (Thu, 01 Nov 2007) | 3 lines

- bail out when not able to lock the device

------------------------------------------------------------------------
r671 | ppalmers | 2007-10-22 23:55:31 +0200 (Mon, 22 Oct 2007) | 4 lines

- make the backlog actually do what it is supposed to do
- make the backlog thread safe

------------------------------------------------------------------------
r670 | ppalmers | 2007-10-22 23:17:17 +0200 (Mon, 22 Oct 2007) | 15 lines

Implement a BackLog. This is a buffer that stores the last debug messages,
regardless of whether their debuglevel is high enough or not. It can be used
to run programs at a low verbosity level, but print highly verbose output when
something goes wrong.

The buffer is a ringbuffer where the oldest message is overwritten when a new
one is added. The readout however is FIFO, oldest messages first.

Use the debugShowBackLog() macro when you want to print the backlog to stderr.

Note that the backlog is not thread safe so sometimes there might be some errors
in the lines, but it otherwise would require the introduction of more locks in
the critical path.

------------------------------------------------------------------------
r669 | ppalmers | 2007-10-21 17:21:02 +0200 (Sun, 21 Oct 2007) | 1 line

fix DICE compilation bug
------------------------------------------------------------------------
r668 | ppalmers | 2007-10-21 17:05:11 +0200 (Sun, 21 Oct 2007) | 1 line

fix small bug in ECHO mixer commands
------------------------------------------------------------------------
r667 | ppalmers | 2007-10-21 16:30:17 +0200 (Sun, 21 Oct 2007) | 1 line

implement basic AF2 control application
------------------------------------------------------------------------
r666 | ppalmers | 2007-10-21 14:17:07 +0200 (Sun, 21 Oct 2007) | 1 line

add omitted files
------------------------------------------------------------------------
r665 | ppalmers | 2007-10-21 14:14:14 +0200 (Sun, 21 Oct 2007) | 1 line

Implement support for Phantom power on the fireworks devices
------------------------------------------------------------------------
r664 | ppalmers | 2007-10-21 13:36:28 +0200 (Sun, 21 Oct 2007) | 4 lines

- Simplify EFC mixer control
- Implement FireWorks generic mixer

------------------------------------------------------------------------
r663 | ppalmers | 2007-10-21 00:13:16 +0200 (Sun, 21 Oct 2007) | 3 lines

- Implement backend for ECHO FireWorks mixer support

------------------------------------------------------------------------
r662 | ppalmers | 2007-10-20 21:23:54 +0200 (Sat, 20 Oct 2007) | 3 lines

- introduce mixer control for the original saffire

------------------------------------------------------------------------
r661 | ppalmers | 2007-10-20 17:32:47 +0200 (Sat, 20 Oct 2007) | 4 lines

- Implement more complete mixer support for the saffire pro
- fix some cleanup issues with control elements

------------------------------------------------------------------------
r660 | jwoithe | 2007-10-19 02:51:38 +0200 (Fri, 19 Oct 2007) | 2 lines

MOTU: allow receive and transmit streams to have different bandwidth requirements, since not all interfaces have about the same number of input and output channels.

------------------------------------------------------------------------
r659 | f.ernoult | 2007-10-19 01:45:57 +0200 (Fri, 19 Oct 2007) | 2 lines

MotU: Centralize all device specific parameters in DeviceProperty

------------------------------------------------------------------------
r658 | f.ernoult | 2007-10-18 21:03:17 +0200 (Thu, 18 Oct 2007) | 2 lines

MotU: Preliminary implementation of UltraLite and 8pre support

------------------------------------------------------------------------
r657 | jwoithe | 2007-10-18 05:55:35 +0200 (Thu, 18 Oct 2007) | 2 lines

Make bandwidth allocation dependent on actual device configuration (and therefore the number of channels actually active).

------------------------------------------------------------------------
r656 | f.ernoult | 2007-10-18 03:58:50 +0200 (Thu, 18 Oct 2007) | 2 lines

MotU: Fix bugs in the ports definition mechanism

------------------------------------------------------------------------
r655 | jwoithe | 2007-10-18 02:15:31 +0200 (Thu, 18 Oct 2007) | 2 lines

Fix up the Traveler's new port definition table to include some of the less obvious modal dependencies.

------------------------------------------------------------------------
r654 | f.ernoult | 2007-10-18 01:04:08 +0200 (Thu, 18 Oct 2007) | 3 lines

MotU: - Rewrite the ports declaration (easier to maintain for new devices)
      - Make MainOut the defaults ports for the 828mkII

------------------------------------------------------------------------
r653 | ppalmers | 2007-10-17 10:16:58 +0200 (Wed, 17 Oct 2007) | 1 line

- a first pass at Focusrite Saffire Pro mixer support. Note: the mixer only works when streaming.
------------------------------------------------------------------------
r650 | ppalmers | 2007-10-14 23:40:59 +0200 (Sun, 14 Oct 2007) | 6 lines

Implement the focusrite saffire pro internal mixers as two connected matrix mixers.
One for the input mix and one for the output mix.

Note: does not check for out-of-boundary yet

------------------------------------------------------------------------
r649 | ppalmers | 2007-10-14 21:26:35 +0200 (Sun, 14 Oct 2007) | 1 line

split up the saffirepro device into a generic focusrite part and a saffirepro part. Add a class for the original saffire.
------------------------------------------------------------------------
r648 | ppalmers | 2007-10-14 21:12:53 +0200 (Sun, 14 Oct 2007) | 1 line

rename saffire pro device implementation file to ease normal saffire implementation
------------------------------------------------------------------------
r647 | ppalmers | 2007-10-14 21:08:00 +0200 (Sun, 14 Oct 2007) | 1 line

Add a matrix mixer control element
------------------------------------------------------------------------
r646 | ppalmers | 2007-10-14 20:23:51 +0200 (Sun, 14 Oct 2007) | 1 line

file cleanup
------------------------------------------------------------------------
r645 | ppalmers | 2007-10-14 20:22:21 +0200 (Sun, 14 Oct 2007) | 1 line

merge all mixers into one application
------------------------------------------------------------------------
r644 | ppalmers | 2007-10-14 19:19:22 +0200 (Sun, 14 Oct 2007) | 2 lines

- Adapt the phase 88 mixer control application to the new DBUS control interface

------------------------------------------------------------------------
r643 | arnonym | 2007-10-13 00:13:12 +0200 (Sat, 13 Oct 2007) | 2 lines

Fix #40: enable_bebob=no not working because of a test-app depending on bebob...

------------------------------------------------------------------------
r642 | arnonym | 2007-10-13 00:11:52 +0200 (Sat, 13 Oct 2007) | 2 lines

fix cleaning

------------------------------------------------------------------------
r641 | ppalmers | 2007-09-26 16:42:31 +0200 (Wed, 26 Sep 2007) | 1 line

- fix 1394 send oops for dm1500 based devices
------------------------------------------------------------------------
r640 | arnonym | 2007-09-15 21:40:49 +0200 (Sat, 15 Sep 2007) | 2 lines

Make the pkgconfig.py a real scons-tool.

------------------------------------------------------------------------
r639 | ppalmers | 2007-09-09 21:04:35 +0200 (Sun, 09 Sep 2007) | 6 lines

- Introduce a generic infrastructure for FFADODevices to present the clock sources they support and their state
- Implement this infrastructure for BeBoB devices
- Implement this infrastructure for ECHO Fireworks devices

------------------------------------------------------------------------
r638 | arnonym | 2007-09-09 18:46:02 +0200 (Sun, 09 Sep 2007) | 6 lines

Check if the alsa-lib has midi-support (its possible to have alsa without midi).

If the check fails, tests/streaming/testmidistreaming1 is not built.

Also add some comments to the configure-checks.

------------------------------------------------------------------------
r637 | wagi | 2007-09-09 14:02:46 +0200 (Sun, 09 Sep 2007) | 4 lines

- some code cleanup (whitespace, unnecessary ';' removed, code indention, ...)
- loadFromCache: problem identified why it doesn't work. The PCPLugs and ExternaPlugs in 
AVC::Unit are not deserialized.

------------------------------------------------------------------------
r636 | wagi | 2007-09-09 13:14:17 +0200 (Sun, 09 Sep 2007) | 2 lines

indention fixed

------------------------------------------------------------------------
r635 | ppalmers | 2007-09-08 19:35:33 +0200 (Sat, 08 Sep 2007) | 1 line

undo tests of trac post-commit hook
------------------------------------------------------------------------
r634 | ppalmers | 2007-09-08 19:34:41 +0200 (Sat, 08 Sep 2007) | 1 line

fixes #26 (test 3 of trac post-commit hook)
------------------------------------------------------------------------
r633 | ppalmers | 2007-09-08 19:09:46 +0200 (Sat, 08 Sep 2007) | 1 line

fixes ticket:26 (test 2 of trac post-commit hook)
------------------------------------------------------------------------
r632 | ppalmers | 2007-09-08 19:08:50 +0200 (Sat, 08 Sep 2007) | 1 line

fixes ticket:26 (test of trac post-commit hook)
------------------------------------------------------------------------
r631 | ppalmers | 2007-09-08 18:59:06 +0200 (Sat, 08 Sep 2007) | 1 line

fixes ticket:26
------------------------------------------------------------------------
r630 | wagi | 2007-09-08 18:19:40 +0200 (Sat, 08 Sep 2007) | 3 lines

Some more fixing of the cashing code. The loading is still disabled because it seems not to work right. But at least it doesn't crash anymore. Some further debugging is 
needed.

------------------------------------------------------------------------
r629 | ppalmers | 2007-09-08 01:09:19 +0200 (Sat, 08 Sep 2007) | 4 lines

- Base infrastructure for the Echo EFC commands
- Some first command implementations

------------------------------------------------------------------------
r628 | ppalmers | 2007-09-07 19:54:12 +0200 (Fri, 07 Sep 2007) | 1 line

some more fooling around with the bebob mixer support
------------------------------------------------------------------------
r627 | ppalmers | 2007-09-07 19:53:17 +0200 (Fri, 07 Sep 2007) | 4 lines

- add test application to send a custom AV/C command to a node
- add test application to examine focusrite vendordep cmds

------------------------------------------------------------------------
r626 | ppalmers | 2007-09-07 19:47:48 +0200 (Fri, 07 Sep 2007) | 1 line

minor things
------------------------------------------------------------------------
r625 | ppalmers | 2007-09-07 19:46:44 +0200 (Fri, 07 Sep 2007) | 1 line

add the saffire (LE) device id
------------------------------------------------------------------------
r624 | ppalmers | 2007-09-07 19:46:04 +0200 (Fri, 07 Sep 2007) | 1 line

some more experiments regarding setting the samplerate on the saffire
------------------------------------------------------------------------
r623 | ppalmers | 2007-09-07 17:12:49 +0200 (Fri, 07 Sep 2007) | 1 line

introduce getSyncState
------------------------------------------------------------------------
r622 | ppalmers | 2007-09-07 17:04:47 +0200 (Fri, 07 Sep 2007) | 1 line

improve samplerate setting for the focusrite
------------------------------------------------------------------------
r621 | arnonym | 2007-09-07 10:48:44 +0200 (Fri, 07 Sep 2007) | 3 lines

Should fix ticket:35, build complaining about missing freebob.h (which is not needed anymore as ffado=freebob2...

------------------------------------------------------------------------
r620 | wagi | 2007-09-06 22:09:35 +0200 (Thu, 06 Sep 2007) | 4 lines

- one tokenize function is enough. currently in serialize.h defined, this might change
- saveCache code workover. The cache is now created in ~/.ffado/cache. It is actived since it should break 
things. Otherwise it needs fixing.

------------------------------------------------------------------------
r619 | ppalmers | 2007-09-05 22:51:06 +0200 (Wed, 05 Sep 2007) | 1 line

add some more device control
------------------------------------------------------------------------
r618 | ppalmers | 2007-09-05 16:07:05 +0200 (Wed, 05 Sep 2007) | 1 line

move serialization routines to libutil such that they can be used for non-AVC stuff too (fireworks EFC)
------------------------------------------------------------------------
r617 | ppalmers | 2007-09-03 22:43:04 +0200 (Mon, 03 Sep 2007) | 1 line

remove bogus code line
------------------------------------------------------------------------
r616 | arnonym | 2007-09-03 20:00:33 +0200 (Mon, 03 Sep 2007) | 2 lines

Fix the pkg-config-file. Thanks to wagi for seeing this...

------------------------------------------------------------------------
r615 | arnonym | 2007-09-03 12:33:17 +0200 (Mon, 03 Sep 2007) | 2 lines

Simplify the build-file.

------------------------------------------------------------------------
r614 | arnonym | 2007-09-03 12:15:46 +0200 (Mon, 03 Sep 2007) | 2 lines

Build and install the bebob-tools only when bebob is enabled.

------------------------------------------------------------------------
r613 | ppalmers | 2007-09-03 11:18:33 +0200 (Mon, 03 Sep 2007) | 1 line

remove last missing reference to bebob functionblock
------------------------------------------------------------------------
r612 | ppalmers | 2007-09-02 20:31:15 +0200 (Sun, 02 Sep 2007) | 4 lines

- Remove dependencies between BeBoB and generic AVC code (functionblocks)
- Clean up the code at some places

------------------------------------------------------------------------
r611 | arnonym | 2007-09-02 17:48:55 +0200 (Sun, 02 Sep 2007) | 1 line

Add myself to the authors. I haven't written any real code until now, but maybe the buildsystem counts.
------------------------------------------------------------------------
r610 | arnonym | 2007-09-02 17:48:17 +0200 (Sun, 02 Sep 2007) | 2 lines

Update the version requirements listed in the readme to match the version wanted by SConstruct. (see #29)

------------------------------------------------------------------------
r609 | arnonym | 2007-08-31 18:24:46 +0200 (Fri, 31 Aug 2007) | 6 lines

Simplify some things.

CACHEDIR is now hardcoded in config.h to be ~/.ffado. Now the sources need code to create the dir if it doesn't exist.

The install target is now depending on bindir, libdir, sharedir and includedir. That way a lot of env.Alias(...) can go away.

------------------------------------------------------------------------
r608 | arnonym | 2007-08-31 17:57:36 +0200 (Fri, 31 Aug 2007) | 2 lines

Fix #30 where scons complains about non-existant directories. Not anymore my friend, if the dir doesn't exist, it gets built (on install-time)...

------------------------------------------------------------------------
r607 | ppalmers | 2007-08-31 10:56:53 +0200 (Fri, 31 Aug 2007) | 11 lines

- Fix file-based discovery. The previous code used a pointer to a
  vendormodel entry. This pointer however pointed to a VendorModel
  object that was created on the stack of the discover function.
  Hence the pointer becomes invalid when the discover function is
  exited.

The changes incorporate using a VendorModelEntry struct instead
  of a pointer to it. So now the FFADODevice has it's own copy of
  this struct, instead of a pointer to it.

------------------------------------------------------------------------
r606 | ppalmers | 2007-08-30 23:09:35 +0200 (Thu, 30 Aug 2007) | 1 line

proposed fix for ticket #30
------------------------------------------------------------------------
r605 | arnonym | 2007-08-30 16:55:24 +0200 (Thu, 30 Aug 2007) | 2 lines

Docs are back!

------------------------------------------------------------------------
r604 | arnonym | 2007-08-30 16:21:14 +0200 (Thu, 30 Aug 2007) | 2 lines

A global include #/ is no longer needed...

------------------------------------------------------------------------
r603 | arnonym | 2007-08-30 16:19:23 +0200 (Thu, 30 Aug 2007) | 1 line

Adopt install-instructions to scons (#25) and replace freebob by ffado/firewire where necessary
------------------------------------------------------------------------
r602 | arnonym | 2007-08-30 15:50:12 +0200 (Thu, 30 Aug 2007) | 1 line

Add options to overwrite bindir,libdir and so on. One TODO less on my list
------------------------------------------------------------------------
r601 | arnonym | 2007-08-30 15:39:58 +0200 (Thu, 30 Aug 2007) | 2 lines

Maybe this time the pkg-flags will get cached and pkg-config isn't called anymore during install / after the first time...

------------------------------------------------------------------------
r600 | arnonym | 2007-08-30 15:38:50 +0200 (Thu, 30 Aug 2007) | 2 lines

Add include guards.

------------------------------------------------------------------------
r599 | wagi | 2007-08-29 23:17:58 +0200 (Wed, 29 Aug 2007) | 6 lines

Instead of static compiled in vendor/model table use configuration files.

Maybe needs some more cleanup but I wanted to check in this baby before
someone else screws me up with some majors changes in the repos :)

------------------------------------------------------------------------
r598 | arnonym | 2007-08-29 19:08:18 +0200 (Wed, 29 Aug 2007) | 2 lines

No auto* means that we can have our own config.h.in without the need to copy...

------------------------------------------------------------------------
r597 | arnonym | 2007-08-29 19:05:19 +0200 (Wed, 29 Aug 2007) | 4 lines

Finally ditch auto*-support.

And bump the version in SConstruct as it was in configure.ac.

------------------------------------------------------------------------
r596 | arnonym | 2007-08-29 19:04:09 +0200 (Wed, 29 Aug 2007) | 1 line

ignore me
------------------------------------------------------------------------
r595 | arnonym | 2007-08-29 19:02:25 +0200 (Wed, 29 Aug 2007) | 2 lines

ignore me

------------------------------------------------------------------------
r594 | arnonym | 2007-08-29 18:51:37 +0200 (Wed, 29 Aug 2007) | 2 lines

Fix cleaning...

------------------------------------------------------------------------
r593 | arnonym | 2007-08-29 18:41:11 +0200 (Wed, 29 Aug 2007) | 4 lines

Clean up the default environment even more.

This makes externalenv redundant.

------------------------------------------------------------------------
r592 | arnonym | 2007-08-29 18:10:16 +0200 (Wed, 29 Aug 2007) | 2 lines

Fix one of my todos and make the env-environment more clean.

------------------------------------------------------------------------
r591 | arnonym | 2007-08-29 17:42:51 +0200 (Wed, 29 Aug 2007) | 1 line

Adopt the pyuic-check to my new-gained knowledge:-)
------------------------------------------------------------------------
r590 | arnonym | 2007-08-29 17:29:05 +0200 (Wed, 29 Aug 2007) | 2 lines

This might(!) fix the issue with pkg-config getting executed during install. But I don't know if the determined flags are cached or only the result (which is [0,1])...

------------------------------------------------------------------------
r589 | ppalmers | 2007-08-29 17:26:18 +0200 (Wed, 29 Aug 2007) | 1 line

small discovery fix
------------------------------------------------------------------------
r588 | arnonym | 2007-08-29 17:18:06 +0200 (Wed, 29 Aug 2007) | 2 lines

refactor the dbus-builders into their own module...

------------------------------------------------------------------------
r587 | ppalmers | 2007-08-29 17:09:34 +0200 (Wed, 29 Aug 2007) | 5 lines

- moved all vendor id's to one include file
- introduced the framework for the ECHO FireWorks platform

------------------------------------------------------------------------
r586 | ppalmers | 2007-08-29 16:06:05 +0200 (Wed, 29 Aug 2007) | 1 line

fix forgotten file
------------------------------------------------------------------------
r585 | ppalmers | 2007-08-29 15:16:32 +0200 (Wed, 29 Aug 2007) | 3 lines

- Implement some device-specific details

------------------------------------------------------------------------
r584 | ppalmers | 2007-08-29 13:44:02 +0200 (Wed, 29 Aug 2007) | 4 lines

- fix bug introduced in previous commit
- focusrite and terratec now have specific bebob devices

------------------------------------------------------------------------
r583 | ppalmers | 2007-08-29 11:19:04 +0200 (Wed, 29 Aug 2007) | 9 lines

- Clean up the FFADODevice constructor to remove the nodeid that
  can be retrieved from the configrom.
- Implement new method of creating devices: a device is supposed
  to implement the createDevice static method that should return
  an instance of itself or a subclass of itself. This should ease
  support for manufacturer specific customizations since we don't 
  have to put them into the device manager class.

------------------------------------------------------------------------
r582 | ppalmers | 2007-08-29 10:28:32 +0200 (Wed, 29 Aug 2007) | 3 lines

see prev commit

------------------------------------------------------------------------
r581 | ppalmers | 2007-08-29 10:28:06 +0200 (Wed, 29 Aug 2007) | 4 lines

- First attempt at mixer control (still disfunctional)
- moved vendor specific code in bebob/*

------------------------------------------------------------------------
r580 | ppalmers | 2007-08-29 10:26:54 +0200 (Wed, 29 Aug 2007) | 1 line

move vendor specific stuff to vendor directory
------------------------------------------------------------------------
r579 | ppalmers | 2007-08-29 10:20:23 +0200 (Wed, 29 Aug 2007) | 1 line

small directory management
------------------------------------------------------------------------
r578 | ppalmers | 2007-08-29 10:19:47 +0200 (Wed, 29 Aug 2007) | 1 line

small directory management
------------------------------------------------------------------------
r577 | jwoithe | 2007-08-28 01:58:41 +0200 (Tue, 28 Aug 2007) | 3 lines

MOTU: more incidental code tidy-ups and reorderings to produce slight increases in efficiency.  No functional changes.
MOTU: make the 1kHz test tone independent of sampling frequency.

------------------------------------------------------------------------
r576 | arnonym | 2007-08-27 23:39:34 +0200 (Mon, 27 Aug 2007) | 6 lines

Some cleanup

I created an externalenv which is a copy of env without all the ffado-defines to be used by the externals. That way there is no need for code-duplication in external/dbus/SConscript.

And the DBUS-Builders should probably go into an own file in admin/...

------------------------------------------------------------------------
r575 | ppalmers | 2007-08-27 23:26:33 +0200 (Mon, 27 Aug 2007) | 24 lines

- Fixed bug in dbus c++ bindings
- First attempt at a decoupled control interface. [WIP]
  
  The src/libcontrol/* elements are the ones that should be
  subclassed to implement control elements on the FFADODevice
  side.

The tests/controlserver.* files contain some code that 
  interfaces the DBus calls to these libcontrol elements. The
  DBus classes allow for introspection and path discovery,
  such that we don't have to care about that anymore.

In the end it should be fairly easy to write another 'backend' 
  to replace the current DBus one, e.g. to implement OSC support
  or MIDI support. (Should we ever want that)

Note that there is no connection between ffado and dbus yet,
  this code is merely to experiment with the Control/DBus infra-
  structure. Once that is sort-of working, connecting ffado to
  this infrastructure is a matter of subclassing the Control::*
  classes, creating them on discovery and putting them into one
  Container::* that is passed on to the DBus handlers.

------------------------------------------------------------------------
r574 | ppalmers | 2007-08-27 17:41:50 +0200 (Mon, 27 Aug 2007) | 6 lines

- scons can now auto-generate the DBUS interface adaptor/proxy classes
- implemented a very basic DBUS client/server pair: 
  test-dbus and test-dbus-server
- some small missing key fixes in the scons files

------------------------------------------------------------------------
r573 | arnonym | 2007-08-26 17:53:12 +0200 (Sun, 26 Aug 2007) | 4 lines

GENERIC_AVC -> GENERICAVC

and added a TODO...

------------------------------------------------------------------------
r571 | arnonym | 2007-08-25 21:14:30 +0200 (Sat, 25 Aug 2007) | 2 lines

Add a first check for pyuic and compile the mixer ui's only if it was successful.

------------------------------------------------------------------------
r570 | arnonym | 2007-08-25 20:55:22 +0200 (Sat, 25 Aug 2007) | 4 lines

Make use of the exported FLAGS from the CheckForPkg.

And I don't think the object-cache needs to be devided as scons is rather good at identifying not be file-descriptor but by a hash of filename and content and compile-flags...

------------------------------------------------------------------------
r569 | arnonym | 2007-08-25 20:53:43 +0200 (Sat, 25 Aug 2007) | 2 lines

Add some help-texts and export the flags to the environment...

------------------------------------------------------------------------
r568 | arnonym | 2007-08-25 20:18:08 +0200 (Sat, 25 Aug 2007) | 1 line

ignore me
------------------------------------------------------------------------
r567 | arnonym | 2007-08-25 17:24:29 +0200 (Sat, 25 Aug 2007) | 2 lines

Simplify the file. After all python is a programming language :-)

------------------------------------------------------------------------
r566 | ppalmers | 2007-08-25 16:09:37 +0200 (Sat, 25 Aug 2007) | 1 line

fix scons problem with PKG_CONFIG_PATH being not set in the calling environment
------------------------------------------------------------------------
r565 | ppalmers | 2007-08-25 15:55:39 +0200 (Sat, 25 Aug 2007) | 1 line

add support for out-of-tree building using 'scons BUILDDIR=target_build_dir'
------------------------------------------------------------------------
r564 | ppalmers | 2007-08-25 14:34:09 +0200 (Sat, 25 Aug 2007) | 1 line

cleaned up the scons file a bit
------------------------------------------------------------------------
r563 | ppalmers | 2007-08-25 13:41:35 +0200 (Sat, 25 Aug 2007) | 1 line

forgotten file
------------------------------------------------------------------------
r562 | ppalmers | 2007-08-25 13:36:32 +0200 (Sat, 25 Aug 2007) | 7 lines

- add DBUS C++ bindings from: 
  http://dev.openwengo.org/svn/openwengo/wengophone-ng/branches/wengophone-dbus-api/libs/dbus

- statically linked because these are not released separately and
  because C++ libs can (apparently) have ABI issues.

------------------------------------------------------------------------
r561 | arnonym | 2007-08-24 19:53:24 +0200 (Fri, 24 Aug 2007) | 13 lines

Merge the scons-branch to trunk.

This basicly adds scons as a buildsystem. This does not mean that auto* is dropped instantly (the added files should not collide) but allows for some testing in the transition-process. The auto*-files are deleted later on...

Due to this, the config.h.in (which gets built by autogen.sh) is copied from config.h.in.scons using scons. When dropping make this file should be moved to config.h.in.

The scons-system is probably missing a lot of things you are used to in make. But at least here it does build a usable libffado. :-)

To use scons, simply do a "scons". If you want to know about configuration, "scons -h" allows you to view and set parameters, which are saved from there on. What do I mean with this? "scons -h PREFIX=$HOME" sets the install-prefix to $HOME and shows the helptext that already incorporates the changed option. Another "scons -h" will show the same help-text and scons remembers the changed prefix. "scons" now builds the libs and apps and "scons install" installs everything into the prefix. "scons -c install" cleans the local dir and uninstalls the package while "scons -c" just cleans the local dir.

BTW: I also set several more ignores, I am just to lazy to split these two commits. ;-)

------------------------------------------------------------------------
r560 | ppalmers | 2007-08-24 17:09:49 +0200 (Fri, 24 Aug 2007) | 8 lines

- Sort the FFADODevice vector on GUID before assigning device id's
  This results in the same device id for identical device setups,
  independent of the way they are connected or the node numbers they
  have been assigned.
- Sanitized debug message reporting a bit
- Cosmetic changes

------------------------------------------------------------------------
r558 | wagi | 2007-08-24 10:33:45 +0200 (Fri, 24 Aug 2007) | 2 lines

tokinize input (WIP)

------------------------------------------------------------------------
r557 | wagi | 2007-08-24 09:22:30 +0200 (Fri, 24 Aug 2007) | 2 lines

moved&renamed configparser from bebob to genericavc

------------------------------------------------------------------------
r556 | ppalmers | 2007-08-22 22:51:46 +0200 (Wed, 22 Aug 2007) | 4 lines

- add forgotten focusrite files from previous commit
- add terratec AV/C commands

------------------------------------------------------------------------
r555 | ppalmers | 2007-08-22 22:27:15 +0200 (Wed, 22 Aug 2007) | 1 line

implement Focusrite specific AV/C commands
------------------------------------------------------------------------
r554 | ppalmers | 2007-08-22 19:15:34 +0200 (Wed, 22 Aug 2007) | 19 lines

Merge echoaudio branch into trunk.

This adds support for the Echo Audiofire devices to FFADO. Possibly also other devices working with the Apple Class Driver will work with this code. It is not fully complete yet, but the main rework is 
done.

First of all the IAvDevice class/interface is renamed to FFADODevice, in order to separate the AV/C code from the FFADO API code. A device supported by FFADO implements a FFADODevice.

The BeBoB device has been split up into three groups:
- libavc/* : all code and commands that are specified by AV/C specs. Note that a lot of the code that used to be in BeBoB::AvDevice now resides in AVC::Unit
- genericavc/* : a FFADODevice that uses AV/C descriptors & commands for discovery and config
- bebob/* : the bebob FFADODevice that inherits from GenericAVC::AvDevice but that uses BridgeCo commands for discovery

Everything has been moved as high as possible in the class hierarchy. If necessary, a subclass that uses device specific commands is introduced (e.g. BeBoB::Plug inherits from AVC::Plug and uses the 
BridgeCo extended plug info command to discover it's properties).

There are some other fixes along the way that have been done too.

------------------------------------------------------------------------
r537 | ppalmers | 2007-08-21 16:57:04 +0200 (Tue, 21 Aug 2007) | 5 lines

- SPDIF out works (tested)
- SPDIF in should work (untested)
- sync from external sync plug works

------------------------------------------------------------------------
r535 | ppalmers | 2007-08-21 16:55:08 +0200 (Tue, 21 Aug 2007) | 1 line

fix bug: returning length without setting it to a specific value sometimes prevents stream startup
------------------------------------------------------------------------
r532 | ppalmers | 2007-08-21 16:53:39 +0200 (Tue, 21 Aug 2007) | 1 line

fix & enhance the streaming test tools
------------------------------------------------------------------------
r530 | wagi | 2007-08-20 22:32:01 +0200 (Mon, 20 Aug 2007) | 2 lines

disable config parser code

------------------------------------------------------------------------
r529 | wagi | 2007-08-20 22:27:46 +0200 (Mon, 20 Aug 2007) | 4 lines

- saveCache and loadFromCache interface added to IAvDevice
- current implenetation of saveCache and loadFromCache moved to bebob
- config parser added for loading vendor and model ids from a file instead hardcoded (not finished yet)

------------------------------------------------------------------------
r525 | jwoithe | 2007-08-20 01:02:18 +0200 (Mon, 20 Aug 2007) | 2 lines

The new device cache code breaks things for non-AVC interfaces such as the MOTU.  Make certain that the cache code (which uses AVC commands) is only used by BeBoB devices.  The solution in this patch may need revisiting but at least it restores functionality for now.

------------------------------------------------------------------------
r523 | ppalmers | 2007-08-17 11:51:29 +0200 (Fri, 17 Aug 2007) | 1 line

fix uninitialized struct params bugs
------------------------------------------------------------------------
r517 | wagi | 2007-08-10 21:13:38 +0200 (Fri, 10 Aug 2007) | 2 lines

remove de/serialize test program

------------------------------------------------------------------------
r516 | wagi | 2007-08-10 21:07:16 +0200 (Fri, 10 Aug 2007) | 6 lines

- some de/serialing bugs found and fixed
- caching enabled for bebob devices

(finally online again! big thanks to my free wireless internet provider. s/he is finally back from 
her/his holiday :)

------------------------------------------------------------------------
r515 | jwoithe | 2007-08-10 01:19:26 +0200 (Fri, 10 Aug 2007) | 2 lines

Cosmetic cleanups of MOTU driver source.  No functional changes.

------------------------------------------------------------------------
r514 | jwoithe | 2007-08-10 01:15:41 +0200 (Fri, 10 Aug 2007) | 2 lines

Port remaining streaming tests to ffado and fix minor compilation errors.

------------------------------------------------------------------------
r512 | jwoithe | 2007-07-30 01:30:54 +0200 (Mon, 30 Jul 2007) | 3 lines

MOTU: more tweaks to improve reliability.  Things are looking pretty good now.
MOTU: Commenced cleanup of MOTU code, removing temporary debug output etc.

------------------------------------------------------------------------
r511 | wagi | 2007-07-29 22:10:05 +0200 (Sun, 29 Jul 2007) | 1 line

Don't use AC_PROG_CXX before CXXFLAGS are defined (AC_PROG_CXX sets '-O2 -g' by default if no CXXFLAGS is defined)
------------------------------------------------------------------------
r506 | ppalmers | 2007-07-28 15:20:25 +0200 (Sat, 28 Jul 2007) | 1 line

compile fix for the slave stream sp
------------------------------------------------------------------------
r498 | jwoithe | 2007-07-27 01:09:57 +0200 (Fri, 27 Jul 2007) | 3 lines

MOTU: more debugging and tweaks.
TimestampedBuffer: ffado_timestamp_t changed to double for now - MOTU needs the extra precision.

------------------------------------------------------------------------
r497 | jwoithe | 2007-07-26 02:01:52 +0200 (Thu, 26 Jul 2007) | 2 lines

Preliminary compile fixes for MOTU following buffer timestamp type change.

------------------------------------------------------------------------
r496 | jwoithe | 2007-07-25 03:45:24 +0200 (Wed, 25 Jul 2007) | 2 lines

Don't update the timestamp DLL if the timestamp isn't moving (as happens during startup).

------------------------------------------------------------------------
r495 | jwoithe | 2007-07-25 02:29:22 +0200 (Wed, 25 Jul 2007) | 3 lines

MOTU: more debugging aids.
TimestampedBuffer: update internal DLL state when changing the tick offset to maintain consistency.  Helps for MOTU, other interfaces need testing.

------------------------------------------------------------------------
r494 | ppalmers | 2007-07-24 18:49:11 +0200 (Tue, 24 Jul 2007) | 9 lines

- switch over to a generic ffado_timestamp_t for the timestamped buffer (currently float)
- implemented some experimental stream phase sync method

- various small things

NOTE: not a very stable commit

------------------------------------------------------------------------
r493 | jwoithe | 2007-07-23 01:55:57 +0200 (Mon, 23 Jul 2007) | 5 lines

StreamProcessorManager.cpp: MOTU requires larger ENABLE_DELAY_CYCLES to successfully start.
TimestampedBuffer.cpp: MOTU needs more exact m_buffer_tail_timestamp.
MOTU: Make startup more reliable and nearly eliminate audio glitches in output.
Note that the changes to StreamProcessorManager.cpp and TimestampedBuffer.cpp are regarded as temporary while a more robust solution is devised following further investigations.

------------------------------------------------------------------------
r492 | jwoithe | 2007-07-16 02:04:05 +0200 (Mon, 16 Jul 2007) | 2 lines

MOTU: more debugging.  There's nothing really profound in these changes.

------------------------------------------------------------------------
r491 | jwoithe | 2007-07-12 02:33:30 +0200 (Thu, 12 Jul 2007) | 2 lines

More changes to restore MOTU functionality.  We're getting closer but it's still not there yet.

------------------------------------------------------------------------
r490 | wagi | 2007-07-09 22:00:22 +0200 (Mon, 09 Jul 2007) | 2 lines

saveCache: create save path "PATH_TO_CACHE + GUID + CONFIGURATION_ID"

------------------------------------------------------------------------
r489 | ppalmers | 2007-07-06 09:51:41 +0200 (Fri, 06 Jul 2007) | 6 lines

Mixer for the Phase24

Phase88 app adapted by andreas@sandberg.pp.se

------------------------------------------------------------------------
r488 | wagi | 2007-07-04 21:35:24 +0200 (Wed, 04 Jul 2007) | 2 lines

Refactoring/cleanup of getConfigurationId function

------------------------------------------------------------------------
r487 | wagi | 2007-06-26 23:28:31 +0200 (Tue, 26 Jun 2007) | 2 lines

implementation of getConfigurationId added (rough/raw version)

------------------------------------------------------------------------
r486 | wagi | 2007-06-18 23:08:54 +0200 (Mon, 18 Jun 2007) | 2 lines

yet another test commit for testing the mail notifaction hook.

------------------------------------------------------------------------
r485 | wagi | 2007-06-18 22:18:11 +0200 (Mon, 18 Jun 2007) | 2 lines

test checkin hook

------------------------------------------------------------------------
r484 | wagi | 2007-06-16 11:42:50 +0200 (Sat, 16 Jun 2007) | 2 lines

test commit hook

------------------------------------------------------------------------
r483 | wagi | 2007-06-16 11:31:25 +0200 (Sat, 16 Jun 2007) | 2 lines

test hook on svn server

------------------------------------------------------------------------
r482 | wagi | 2007-06-16 11:13:34 +0200 (Sat, 16 Jun 2007) | 2 lines

Bogus checkin -> test installed commit hook on svn server

------------------------------------------------------------------------
r481 | jwoithe | 2007-06-13 01:41:42 +0200 (Wed, 13 Jun 2007) | 2 lines

Fix for MOTU transmit stream failure to enable.

------------------------------------------------------------------------
r480 | jwoithe | 2007-06-12 02:53:30 +0200 (Tue, 12 Jun 2007) | 4 lines

Make teststreaming2 work again (needed for MOTU debugging).
Fix a typo in StreamProcessorManager which assigned the value of snoopMode to m_is_slave.
Debugging of MOTU driver.

------------------------------------------------------------------------
r479 | wagi | 2007-06-05 22:04:20 +0200 (Tue, 05 Jun 2007) | 4 lines

- added all missing 'virtual' to function declared in avdevice implementation
- added getConfigurationId function (used for av/c model caching)

------------------------------------------------------------------------
r478 | wagi | 2007-06-02 17:06:28 +0200 (Sat, 02 Jun 2007) | 4 lines

- save cache information to $(localstatedir)/cache/libffade
- refactoring of build and load cache function

------------------------------------------------------------------------
r477 | ppalmers | 2007-05-27 22:03:23 +0200 (Sun, 27 May 2007) | 1 line

Add M-Audio NRV10 device ID
------------------------------------------------------------------------
r476 | ppalmers | 2007-05-27 21:59:14 +0200 (Sun, 27 May 2007) | 1 line

Add device ID for M-Audio Solo
------------------------------------------------------------------------
r472 | ppalmers | 2007-05-21 15:18:54 +0200 (Mon, 21 May 2007) | 1 line

Add focusrite saffire pro10io device id
------------------------------------------------------------------------
r471 | ppalmers | 2007-05-17 23:42:49 +0200 (Thu, 17 May 2007) | 1 line

integrated into jack SVN
------------------------------------------------------------------------
r470 | ppalmers | 2007-04-22 11:39:31 +0200 (Sun, 22 Apr 2007) | 6 lines

At the moment, if there is no C++ compiler installed, ./configure will
abort with an error about libpthread. This change makes the problem
more obvious.

Signed-off-by: Jeremy Kerr <jk@ozlabs.org>

------------------------------------------------------------------------
r467 | ppalmers | 2007-04-17 22:29:25 +0200 (Tue, 17 Apr 2007) | 1 line

- sync to latest jackd SVN
------------------------------------------------------------------------
r466 | ppalmers | 2007-04-17 22:27:39 +0200 (Tue, 17 Apr 2007) | 1 line

- introduce patch with jackd support
------------------------------------------------------------------------
r464 | ppalmers | 2007-04-14 19:39:44 +0200 (Sat, 14 Apr 2007) | 1 line

BeBoB stream start utility
------------------------------------------------------------------------
r462 | ppalmers | 2007-04-09 22:41:33 +0200 (Mon, 09 Apr 2007) | 3 lines

- Rework the OSC space to something more usable

------------------------------------------------------------------------
r461 | ppalmers | 2007-04-09 18:25:30 +0200 (Mon, 09 Apr 2007) | 5 lines

- Osc requests now don't require a perfect path match to work.
  e.g. an OscNode at path '/X/Y' will now handle OscMessages for '/X/Y/Z'
       if 'Z' is not registered as a child OscNode of '/X/Y'

------------------------------------------------------------------------
r460 | ppalmers | 2007-04-09 17:38:42 +0200 (Mon, 09 Apr 2007) | 1 line

- phase88 mixer default window size change
------------------------------------------------------------------------
r459 | ppalmers | 2007-04-09 17:37:40 +0200 (Mon, 09 Apr 2007) | 1 line

- Add basic OSC mixer control for Phase88
------------------------------------------------------------------------
r458 | ppalmers | 2007-04-09 17:36:22 +0200 (Mon, 09 Apr 2007) | 1 line

fix small bug in OSC set volume message parsing
------------------------------------------------------------------------
r457 | ppalmers | 2007-04-09 15:12:19 +0200 (Mon, 09 Apr 2007) | 1 line

add get volume command for the GenericMixer
------------------------------------------------------------------------
r456 | ppalmers | 2007-04-09 13:46:52 +0200 (Mon, 09 Apr 2007) | 1 line

start of an OSC client (non-functional)
------------------------------------------------------------------------
r455 | ppalmers | 2007-04-09 13:25:12 +0200 (Mon, 09 Apr 2007) | 4 lines

- test-mixer volume control works (tested on phase88 only)
- volume control is exposed to the outside

------------------------------------------------------------------------
r454 | ppalmers | 2007-04-09 11:20:36 +0200 (Mon, 09 Apr 2007) | 5 lines

- fix incorrect referencing of ffado.h 
  (used the installed version instead of the source tree version).
- fix makefile

------------------------------------------------------------------------
r453 | ppalmers | 2007-04-09 11:14:19 +0200 (Mon, 09 Apr 2007) | 1 line

correct forgotten files in name change
------------------------------------------------------------------------
r452 | ppalmers | 2007-04-09 11:06:44 +0200 (Mon, 09 Apr 2007) | 5 lines

libosc
- add type conversion functions to convert between the argument
  type and the wanted data type

------------------------------------------------------------------------
r451 | ppalmers | 2007-04-08 17:56:32 +0200 (Sun, 08 Apr 2007) | 13 lines

- First attempt at a OSC controlled mixer. The level of 
  abstraction is very low, meaning that you have to know
  how the function blocks work. It however allows control
  applications to be written and to experiment with them.
- This version only does Selector function blocks.
  The following message switches the phase88 input to the
  * front (or is is back?)
   /devicemanager/dev0/GenericMixer set selector 10 0
  * back (or is it front?)
   /devicemanager/dev0/GenericMixer set selector 10 1

------------------------------------------------------------------------
r450 | ppalmers | 2007-04-08 17:31:44 +0200 (Sun, 08 Apr 2007) | 1 line

add method to retrieve node id of a device
------------------------------------------------------------------------
r449 | ppalmers | 2007-04-08 14:47:44 +0200 (Sun, 08 Apr 2007) | 1 line

- switched osc node name to the device ID instead of the node id
------------------------------------------------------------------------
r448 | ppalmers | 2007-04-08 14:32:08 +0200 (Sun, 08 Apr 2007) | 1 line

fix forgotten FREEBOB to FFADO conversion
------------------------------------------------------------------------
r447 | ppalmers | 2007-04-08 13:28:58 +0200 (Sun, 08 Apr 2007) | 5 lines

- Better display of the function blocks in the graph
- test-mixer can now switch the front/back switch of the phase88
  (function block 10: ./test-mixer PORT NODE 10)

------------------------------------------------------------------------
r446 | ppalmers | 2007-04-07 17:49:26 +0200 (Sat, 07 Apr 2007) | 3 lines

- add some file templates and a script to instantiate them

------------------------------------------------------------------------
r445 | pieterpalmers | 2007-04-02 21:35:17 +0200 (Mon, 02 Apr 2007) | 11 lines

* name change from FreeBoB to FFADO
* replaced tabs by 4 spaces
* got rid of end-of-line spaces
* made all license and copyrights conform
     library becomes LGPL, apps become GPL
     explicitly state LGPL v2.1 and GPL v2 (don't like v3 draft)

copyrights are 2005-2007 Daniel & Pieter
     except for the MotU stuff (C) Jonathan, Pieter

------------------------------------------------------------------------
r444 | pieterpalmers | 2007-04-02 21:29:30 +0200 (Mon, 02 Apr 2007) | 1 line

more filename changes
------------------------------------------------------------------------
r443 | pieterpalmers | 2007-04-02 21:28:35 +0200 (Mon, 02 Apr 2007) | 1 line

another filename change
------------------------------------------------------------------------
r442 | pieterpalmers | 2007-04-02 21:25:05 +0200 (Mon, 02 Apr 2007) | 1 line

rename files & directories
------------------------------------------------------------------------
r441 | pieterpalmers | 2007-04-02 21:19:27 +0200 (Mon, 02 Apr 2007) | 3 lines

- copy over the good revision of strmrework to the trunk

------------------------------------------------------------------------
r437 | pieterpalmers | 2007-03-13 20:22:17 +0200 (Tue, 13 Mar 2007) | 4 lines

- made test-freebob use debugmodule infrastructure
- test-mixer now takes a port argument

------------------------------------------------------------------------
r436 | pieterpalmers | 2007-03-11 21:50:16 +0200 (Sun, 11 Mar 2007) | 4 lines

- fixed verbose level setting
- debugged DICE code, works with EVM

------------------------------------------------------------------------
r435 | pieterpalmers | 2007-03-11 15:20:31 +0200 (Sun, 11 Mar 2007) | 23 lines

src/devicemanager:
- start OSC server for the device manager

src/devicemanager,
src/iavdevice,
src/libieee1394/configrom:
- inherit from OscNode to become Osc'able

src/bounce,
src/libstreaming/AmdtpStreamProcessor,
src/libstreaming/AmdtpSlaveStreamProcessor:
- fixed bounce device implementation, now working

src/bebob:
- fixed midi bug

General:
- removed 'intermediate XML'
- removed obsolete tests
- removed obsolete files
- removed obsolete API calls

------------------------------------------------------------------------
r434 | pieterpalmers | 2007-03-11 15:15:59 +0200 (Sun, 11 Mar 2007) | 1 line

fix debug messages
------------------------------------------------------------------------
r433 | pieterpalmers | 2007-03-11 15:15:41 +0200 (Sun, 11 Mar 2007) | 3 lines

- dice support

------------------------------------------------------------------------
r432 | pieterpalmers | 2007-03-11 15:14:59 +0200 (Sun, 11 Mar 2007) | 3 lines

- extended OSC support

------------------------------------------------------------------------
r430 | jwoithe | 2007-03-06 07:47:17 +0200 (Tue, 06 Mar 2007) | 4 lines

MOTU: more streaming update work.  Rx now appears to be working correctly.
Tx makes a noise but there are regular glitches, indicating that things
aren't quite right yet (the tx timestamps are probably not tight enough).

------------------------------------------------------------------------
r429 | jwoithe | 2007-03-06 07:45:11 +0200 (Tue, 06 Mar 2007) | 6 lines

Fix Ieee1394Service::freeIsoChannel() so it actually deallocates a channel
and its bandwidth.  In short, unregisterIsoChannel() returns false on error
and (most importantly) one can't call unregisterIsoChannel() before
raw1394_bandwidth_modify() and raw1394_channel_modify() because it wipes the
very information one needs for these calls to succeed.

------------------------------------------------------------------------
r428 | jwoithe | 2007-03-02 01:26:52 +0200 (Fri, 02 Mar 2007) | 2 lines

MOTU: more work trying to get MOTU working with the new streaming framework.

------------------------------------------------------------------------
r426 | pieterpalmers | 2007-02-28 15:15:40 +0200 (Wed, 28 Feb 2007) | 6 lines

- changed the IAvDevice from an interface to a base class,
  since there is some code duplication starting to
  appear.

------------------------------------------------------------------------
r425 | jwoithe | 2007-02-28 05:15:25 +0200 (Wed, 28 Feb 2007) | 2 lines

MOTU: more fixes to the new streaming code for MOTU.  Rx shows promise but TX is still broken.

------------------------------------------------------------------------
r424 | pieterpalmers | 2007-02-27 13:31:33 +0200 (Tue, 27 Feb 2007) | 12 lines

- The library can now be started in 'slave mode', creating a BounceSlaveDevice.
  On a discovering node, this slave device is discovered as a BounceDevice. 
  Streaming does not work yet, something wrong with the timestamps.
- Implemented the 'snoop mode', that allows a client to 'snoop' the streams
  between another host and a device. It is only implemented for BeBoB devices.
  The channel numbers and stream configuration are automatically detected.
  Note that it currently relies on a rather hackish support for reading the
  {i,o}PCR plugs by using private functions of libiec61883
- changed jack backend to support these two new features

------------------------------------------------------------------------
r423 | pieterpalmers | 2007-02-25 16:01:07 +0200 (Sun, 25 Feb 2007) | 1 line

commit after rename (2)
------------------------------------------------------------------------
r422 | pieterpalmers | 2007-02-25 16:00:15 +0200 (Sun, 25 Feb 2007) | 1 line

commit after rename
------------------------------------------------------------------------
r421 | pieterpalmers | 2007-02-25 15:58:44 +0200 (Sun, 25 Feb 2007) | 3 lines

- commit before rename

------------------------------------------------------------------------
r420 | pieterpalmers | 2007-02-25 14:51:39 +0200 (Sun, 25 Feb 2007) | 1 line

moved src/libfreebobavc to src/libavc
------------------------------------------------------------------------
r419 | pieterpalmers | 2007-02-25 14:28:00 +0200 (Sun, 25 Feb 2007) | 1 line

namespace simplification
------------------------------------------------------------------------
r418 | pieterpalmers | 2007-02-25 14:14:33 +0200 (Sun, 25 Feb 2007) | 1 line

added serialization support to the OptionContainer
------------------------------------------------------------------------
r417 | pieterpalmers | 2007-02-25 13:27:59 +0200 (Sun, 25 Feb 2007) | 3 lines

- some small segfault fixes in the serialization

------------------------------------------------------------------------
r416 | pieterpalmers | 2007-02-25 12:46:00 +0200 (Sun, 25 Feb 2007) | 11 lines

- Introduced a generic option mechanism. AvDevices now automatically
  inherit from OptionContainer (via IAvDevice) and can specify options 
  (e.g. at construction). These can then be get/set using the container 
  functions from the outside.
- Modified the bebob, bounce, motu, mh, rme AvDevices to make use of the
  option system for their Id value and the 'snoopMode' option.
- Made M-Audio avdevice a subclass of the BeBoB avdevice to avoid code
  duplication.
- Extended the bounce device

------------------------------------------------------------------------
r415 | pieterpalmers | 2007-02-24 19:51:56 +0200 (Sat, 24 Feb 2007) | 22 lines

ieee1394service:
- implemented 64bit compare-swap-lock operation (needed for DICE)
- small name change of (un)registerARMhandler to (un)registerARMHandler

iavdevice.h:
- made the stream start/stop functions return bool instead of int
- updated function documentation for consistency and to reflect changes

BeBoB avdevice:
- replaced the 2 fixed streamprocessor pointers with a 2 vectors of streamprocessors
- implemented the 'snoop mode' (cannot be activated yet)

libstreaming:
- removed unused 'type' attribute from AmdtpPortInfo & children

mh_avdevice, motu_avdevice, rme_avdevice:
- replaced m_1394service with m_p1394service for consistence

maudio_avdevice.cpp:
- removed unused code

------------------------------------------------------------------------
r414 | pieterpalmers | 2007-02-22 20:55:06 +0200 (Thu, 22 Feb 2007) | 1 line

extended ARM handler functionality
------------------------------------------------------------------------
r413 | pieterpalmers | 2007-02-22 19:08:05 +0200 (Thu, 22 Feb 2007) | 9 lines

- moved all generic IEEE1394 classes into libieee1394
    src/libieee1394/ieee1394service.h
    src/libieee1394/csr1212.h
    src/libieee1394/configrom.cpp
    src/libieee1394/configrom.h
    src/libieee1394/ieee1394service.cpp
    src/libieee1394/csr1212.c

------------------------------------------------------------------------
r412 | pieterpalmers | 2007-02-22 18:49:35 +0200 (Thu, 22 Feb 2007) | 8 lines

- added some documentation
- added a lock()/unlock() to IAvDevice (see header)
- reimplemented test-freebob to the new C++ api
- started with support for AddressRangeMapping, i.e. response
  to reads/writes of the 1394 memory space on the host

------------------------------------------------------------------------
r411 | pieterpalmers | 2007-02-22 00:25:53 +0200 (Thu, 22 Feb 2007) | 13 lines

cycletimer.h: 
- some extra operations on Ticks (diffTicks & substractTicks)

StreamProcessor.cpp
AmdtpStreamProcessor.cpp
MotuStreamProcessor.cpp:
- Moved the syncDelay to StreamProcessor::getTimeUntilNextPeriodSignalUsecs(). This delay should be the delay between the actual period boundary and the time it is reported to the SPManager. Therefore it's place is not as a buffer offset, but in the calculation of the signalling time.
This makes that the buffer timestamps correspond to 'real' timestamps. These might have to be manipulated by the transmit or receive handles to account for e.g. iso buffering etc..., but at least the timestamps themselves have a well-defined meaning now.

StreamProcessorManager.cpp:
- The only stream that needs to be running is the sync source stream. It is assumed that the other streams start running in time. 'In time' is currently about 2000 cycles afterwards.

------------------------------------------------------------------------
r410 | jwoithe | 2007-02-21 01:06:00 +0200 (Wed, 21 Feb 2007) | 2 lines

MOTU: rough fixups towards restoring functionality under the new streaming framework.  It is not yet functional.

------------------------------------------------------------------------
r409 | jwoithe | 2007-02-21 00:30:58 +0200 (Wed, 21 Feb 2007) | 2 lines

MOTU: the calculated bandwidth is already a per channel figure.

------------------------------------------------------------------------
r408 | pieterpalmers | 2007-02-20 19:21:31 +0200 (Tue, 20 Feb 2007) | 5 lines

- Implemented a mechanism to allocate and deallocate iso channels in a generic manner, being by cmp or otherwise.

- cleaned all commented out code from RME, as it is getting increasingly outdated.

------------------------------------------------------------------------
r407 | pieterpalmers | 2007-02-20 16:38:52 +0200 (Tue, 20 Feb 2007) | 8 lines

- Changed the way the device class configure options are handled. Now they are handled in the makefiles instead of the source files. The only source file that still contains the #ifdef's is devicemanager.cpp, to conditionally include the device class include files and to conditionally probe the classes that might be supported.
- added a configure option to disable the compilation of the test programs in tests/
- cleaned up the ADMTP transmit streamprocessor. Now it sends silenced packets when in the disabled state, instead of no-data packets
- added a getNodeID() to ieee1394service
- made comments in ieee1394service.h doxygen compliant

------------------------------------------------------------------------
r406 | jwoithe | 2007-02-20 02:25:37 +0200 (Tue, 20 Feb 2007) | 2 lines

Fix uninitialised pointer in debugmodule.

------------------------------------------------------------------------
r405 | pieterpalmers | 2007-02-19 10:02:47 +0200 (Mon, 19 Feb 2007) | 4 lines

- fix some small bugs in ieee1394service and configrom
- modify bounce device's discovery routine

------------------------------------------------------------------------
r404 | pieterpalmers | 2007-02-18 17:09:00 +0200 (Sun, 18 Feb 2007) | 16 lines

- introduce support framework for DICE and Metric Halo
- change probe/discovery code to make adding devices easier
- made conditional compilation effectively work.
  ./configure now has the following switches:
	--enable-bebob          build BeBoB support (default=yes)
	--enable-motu           build Motu support (default=no)
	--enable-dice           build DICE support (default=no)
	--enable-metric-halo    build Metric Halo support (note: completely useless)
							(default=no)
	--enable-rme            build RME support (note: completely useless)
							(default=no)
	--enable-bounce         build Bounce device support (default=no)
	--enable-all-devices    build support for all supported devices (default=no)
  these now turn on/off compilation effectively.

------------------------------------------------------------------------
r403 | pieterpalmers | 2007-02-18 14:47:52 +0200 (Sun, 18 Feb 2007) | 1 line

update of the class diagram in doc
------------------------------------------------------------------------
r402 | pieterpalmers | 2007-02-18 01:05:37 +0200 (Sun, 18 Feb 2007) | 1 line

adapted the Motu SP to the new stream sync framework
------------------------------------------------------------------------
r401 | pieterpalmers | 2007-02-17 18:37:50 +0200 (Sat, 17 Feb 2007) | 1 line

move some code from amdtp SP to SP base class
------------------------------------------------------------------------
r400 | pieterpalmers | 2007-02-17 17:16:02 +0200 (Sat, 17 Feb 2007) | 2 lines

added configure switches for enabling of device class (BeBoB, Motu, ...) support

------------------------------------------------------------------------
r399 | pieterpalmers | 2007-02-17 16:22:15 +0200 (Sat, 17 Feb 2007) | 5 lines

- code cleanup
- introduce sync delay concept to fix latency issues due to intermediate ISO buffering
- made SytMonitor use cycletimer.h functions

------------------------------------------------------------------------
r398 | pieterpalmers | 2007-02-17 11:57:53 +0200 (Sat, 17 Feb 2007) | 1 line

remove cycle timer prediction & DLL code from the IsoHandler, as it is replaced by a raw1394 API call
------------------------------------------------------------------------
r397 | pieterpalmers | 2007-02-17 00:40:14 +0200 (Sat, 17 Feb 2007) | 5 lines

- make timestampedbuffer use floats instead of doubles
- change iso receive back to the efficient case

------------------------------------------------------------------------
r396 | pieterpalmers | 2007-02-16 13:34:07 +0200 (Fri, 16 Feb 2007) | 4 lines

- fixed initialization of buffer timestamps such that xmit-only
  sync generation works

------------------------------------------------------------------------
r395 | pieterpalmers | 2007-02-16 12:35:25 +0200 (Fri, 16 Feb 2007) | 28 lines

debugmodule.cpp:
- removed path from the source file name when printing debug messages

IsoStream.cpp, StreamProcessor.cpp:
- debug message modifications

StreamProcessorManager.cpp:
- removed obsolete debug code

AmdtpStreamProcessor.cpp:
- debug message modifications
- removed DLL that calculates framerate, as it is also calculated in the
  TimestampedBuffer
- converted code to use the new offset feature of the TimestampedBuffer
- converted code to use the new  frame timestamp calculation feature 
  of the TimestampedBuffer
- first try at xmit sync code (unfinished)
- fixed 'unable to start' bug in receive SP

bebob_avdevice.cpp:
- added some #ifdef code to test with xmit SP's only (temporary)

TimestampedBuffer.cpp:
- add offset feature
- add abitrary frame timestamp calculation

------------------------------------------------------------------------
r394 | pieterpalmers | 2007-02-14 11:31:54 +0200 (Wed, 14 Feb 2007) | 3 lines

- fixed SYT timestamp to ticks conversion

------------------------------------------------------------------------
r393 | pieterpalmers | 2007-02-13 19:22:14 +0200 (Tue, 13 Feb 2007) | 7 lines

- fixed some bugs in the timestampedbuffer
- cleaned up the amdtpstreamprocessor
- updated test-sytmonitor and test-cycletimer
  to the new threading structure
- implemented test for timestampedbuffer

------------------------------------------------------------------------
r392 | pieterpalmers | 2007-02-10 14:06:26 +0200 (Sat, 10 Feb 2007) | 5 lines

- document TimestampedBuffer class
- partially ported timestamp handling to TimestampedBuffer
- introduced test for TimestampedBuffer class

------------------------------------------------------------------------
r391 | pieterpalmers | 2007-02-09 10:01:30 +0200 (Fri, 09 Feb 2007) | 7 lines

* Partially finished:
  - Introduce TimestampedBuffer util class
  - replace interal ringbuffer of SP with timed ringbuffer

* Compiles & works

------------------------------------------------------------------------
r390 | pieterpalmers | 2007-02-07 09:59:56 +0200 (Wed, 07 Feb 2007) | 4 lines

* working version of SYT based AMDTP receive and transmit. 
  Still has to be tuned to work with low buffer sizes.

------------------------------------------------------------------------
r386 | pieterpalmers | 2007-02-02 21:03:41 +0200 (Fri, 02 Feb 2007) | 4 lines

- moved files around to the place they belong
- fixed all compile warnings

------------------------------------------------------------------------
r385 | pieterpalmers | 2007-02-02 01:02:24 +0200 (Fri, 02 Feb 2007) | 4 lines

- fixed issues with SYT timestamp processing
- SYT based sync works if syncing to the received stream

------------------------------------------------------------------------
r384 | pieterpalmers | 2007-01-30 23:11:25 +0200 (Tue, 30 Jan 2007) | 11 lines

- temporary commit as backup measure
- rewrote synchronisation code
- receive streaming based on SYT works
- transmit streaming synced to received stream sort of works, still
  have to iron out some issues.

NOTE: all devices but the bebob's are disabled in this code, 
      because they still have to be ported to the new sync
      mechanism.

------------------------------------------------------------------------
r383 | pieterpalmers | 2007-01-28 19:36:11 +0200 (Sun, 28 Jan 2007) | 1 line

mostly whitespace fixes
------------------------------------------------------------------------
r382 | pieterpalmers | 2007-01-28 19:25:28 +0200 (Sun, 28 Jan 2007) | 1 line

bogus file removed
------------------------------------------------------------------------
r381 | pieterpalmers | 2007-01-28 19:24:21 +0200 (Sun, 28 Jan 2007) | 1 line

updated class diagrams
------------------------------------------------------------------------
r380 | pieterpalmers | 2007-01-27 12:05:09 +0200 (Sat, 27 Jan 2007) | 7 lines

- created a temporary branch to rework a part of the streaming code. 
The changes will be rather intrusive and significant, and I'll disable
Motu on this branch so that I don't have to adapt that code (yet).
A good implementation of synchronisation needs another approach than
what has been done up till now.

------------------------------------------------------------------------
r379 | wagi | 2007-01-21 22:22:41 +0200 (Sun, 21 Jan 2007) | 2 lines

AvDevice: all members are de/serialed now

------------------------------------------------------------------------
r378 | wagi | 2007-01-21 14:05:19 +0200 (Sun, 21 Jan 2007) | 2 lines

FunctionBlock: De/Serialize added

------------------------------------------------------------------------
r377 | wagi | 2007-01-20 21:03:10 +0200 (Sat, 20 Jan 2007) | 5 lines

libfreebobavc: output reformated so it looks nice again (debugmodule used instead of 
plain printfs)
debugmodule: use stdout instead of sterr. 
DebugModuleManager: print register and unregister only in higher debug levels

------------------------------------------------------------------------
r376 | wagi | 2007-01-20 20:25:59 +0200 (Sat, 20 Jan 2007) | 5 lines

De/Serialize function added to subunit
Give AvDevice reference to all deserialize functions (instead of
several arguments like config rom etc)

------------------------------------------------------------------------
r375 | wagi | 2007-01-20 17:40:18 +0200 (Sat, 20 Jan 2007) | 5 lines

AVCCommand: All AVC commants take a reference instead a pointer to 1394 service
BeBoB::AvDevice: PreSonus FIREBOX id added
		 De/Serialize all AvPlugs through AvPlugManager
		 AvPlugManager can't be a member because of serialization

------------------------------------------------------------------------
r374 | jwoithe | 2007-01-15 04:55:23 +0200 (Mon, 15 Jan 2007) | 3 lines

More adjustments to make MOTU devices actually work again.
Make MOTU and RME work again now that ieee1394service's read/write functions interpret their length argument in quadlets rather than bytes.

------------------------------------------------------------------------
r373 | jwoithe | 2007-01-10 01:54:40 +0200 (Wed, 10 Jan 2007) | 2 lines

Make detection of RME and MOTU devices work again.

------------------------------------------------------------------------
r372 | wagi | 2007-01-07 17:08:53 +0200 (Sun, 07 Jan 2007) | 5 lines

AvPlug::serialize: finished implementation (untested yet)
AvPlug::deserialize: likewise
AvPlug::deserializeUpdate: new function. second stage of deserializing
IODeserialize::isExisting: new function

------------------------------------------------------------------------
r371 | wagi | 2007-01-06 14:45:31 +0200 (Sat, 06 Jan 2007) | 5 lines

AvDevice::serialize: m_pcrPlugs and m_externalPlug added to the bed processed 
(ongoing)
AvDevice::deserialize: likewise
AvPlug: started with de/serializing functions (ongoing)

------------------------------------------------------------------------
r370 | wagi | 2007-01-06 12:51:12 +0200 (Sat, 06 Jan 2007) | 2 lines

Don't cache node id everywhere, config rom holds an updated version of the node id

------------------------------------------------------------------------
r369 | pieterpalmers | 2007-01-05 13:24:37 +0200 (Fri, 05 Jan 2007) | 3 lines

- fix compilation issue (reported by Klaus Zimmermann)

------------------------------------------------------------------------
r368 | wagi | 2007-01-05 01:31:11 +0200 (Fri, 05 Jan 2007) | 8 lines

ConfigRom::serialize: no prefix needed for member 
ConfigRom::deserialize: ieee1394service argument
	no prefix needed for member
	return 0 if deserializing fails
AvDevice::AvDevice: new ctor
vDevice::serialize: comment removed
AvDevice::deserialize: config rom deserializing code added

------------------------------------------------------------------------
r367 | wagi | 2007-01-04 01:31:25 +0200 (Thu, 04 Jan 2007) | 3 lines

DeviceManager::saveCache: basePath construction fixed for more than one device
DeviceManager::loadCache: likewise

------------------------------------------------------------------------
r366 | wagi | 2007-01-04 01:01:20 +0200 (Thu, 04 Jan 2007) | 1 line

remove 'proof-of-concept' hack definitions (unsed)
------------------------------------------------------------------------
r365 | wagi | 2007-01-04 00:37:50 +0200 (Thu, 04 Jan 2007) | 9 lines

src/libfreebobavc/serialize* moved to src/libfreebob/avc_serialize*, all includes adapted
src/bebob/bebob_serialize* moved to src/libutil/serialize*
src/libutil/serialize: use Glib::ustring instead of std::string.
src/configrom: serialize and deserialize added (not finished)
src/devicemanager: load and save cached functionality added (not finished)

various small whitespace updates (emacs lässt grüssen :))

------------------------------------------------------------------------
r364 | pieterpalmers | 2007-01-02 20:58:18 +0200 (Tue, 02 Jan 2007) | 3 lines

- forgotten in last commit.

------------------------------------------------------------------------
r363 | wagi | 2007-01-02 14:44:46 +0200 (Tue, 02 Jan 2007) | 2 lines

AvDevice::setSamplingFrequencyPlug: do not test format info status when iterating over supported sample rates. same fix was applied onto 1.0 branch earlier.

------------------------------------------------------------------------
r361 | wagi | 2006-12-30 22:27:46 +0200 (Sat, 30 Dec 2006) | 2 lines

serializing interface extended. write and read except a 'string' as argument

------------------------------------------------------------------------
r360 | pieterpalmers | 2006-12-30 21:49:46 +0200 (Sat, 30 Dec 2006) | 22 lines

- temporary commit to backup some work

- Started a framework to synchronize IsoHandlers to
  any generic TimeSource. The idea is to introduce
  one overall time reference, and resynchronize all
  other timed events to this time source.
  This will, on the long run, allow:
  * combining devices on multiple FW busses together,
    as these are not synched by hardware.
  * synchronizing to the system clock
  * synchronizing to any other time source (e.g.
    when implementing a jackd client, i.e. using
    the freebob devices as jackd clients).
- Implemented a realtime safe way to read the cycle 
  timer for an IsoHandler. (+ test application)
- Implemented tests/test-sytmonitor:
  Monitors 2 or more channels and reports the average
  SYT timestamp difference between both.
- Messed around with SYT timestamping for AMDTP. Doesn't
  work (yet).

------------------------------------------------------------------------
r359 | wagi | 2006-12-30 14:04:53 +0200 (Sat, 30 Dec 2006) | 2 lines

serializing interface fixed for 64 bit machines

------------------------------------------------------------------------
r358 | wagi | 2006-12-29 23:25:01 +0200 (Fri, 29 Dec 2006) | 2 lines

basic bebob serializing code added.

------------------------------------------------------------------------
r356 | pieterpalmers | 2006-12-28 14:03:52 +0200 (Thu, 28 Dec 2006) | 3 lines

- move 2.0 to new trunk (take 2)

------------------------------------------------------------------------
r352 | pieterpalmers | 2006-12-28 13:54:10 +0200 (Thu, 28 Dec 2006) | 3 lines

- trunk should become trunk/libfreebob

------------------------------------------------------------------------
r347 | pieterpalmers | 2006-12-28 13:38:54 +0200 (Thu, 28 Dec 2006) | 2 lines

- switch over 2.0 as new trunk

------------------------------------------------------------------------
r341 | pieterpalmers | 2006-12-26 18:21:31 +0200 (Tue, 26 Dec 2006) | 5 lines

- changed bebob avdevice to use debugmodule instead of printf/cout
- fixed some minor merge side-effects
- implement a RT safe mechanism to obtain the cycle counter.

------------------------------------------------------------------------
r340 | pieterpalmers | 2006-12-26 15:52:36 +0200 (Tue, 26 Dec 2006) | 1 line

converted ieee1394service debug messages to messagebuffer
------------------------------------------------------------------------
r339 | pieterpalmers | 2006-12-26 12:49:55 +0200 (Tue, 26 Dec 2006) | 1 line

add fwap.xml (forgotten in last commit)
------------------------------------------------------------------------
r338 | pieterpalmers | 2006-12-26 12:40:33 +0200 (Tue, 26 Dec 2006) | 1 line

add xml file for maudio fw audiophile (experimental)
------------------------------------------------------------------------
r337 | pieterpalmers | 2006-12-25 17:47:17 +0200 (Mon, 25 Dec 2006) | 1 line

- directory maintenance
------------------------------------------------------------------------
r336 | pieterpalmers | 2006-12-25 17:42:25 +0200 (Mon, 25 Dec 2006) | 7 lines

- Merged the developments on trunk since branch-off:
   branch occurred at rev 194
   svn merge -r 194:HEAD https://svn.sourceforge.net/svnroot/freebob/trunk/libfreebob
- Modified libfreebobavc to use the messagebuffer for debug info.
- This should compile and run

------------------------------------------------------------------------
r335 | pieterpalmers | 2006-12-25 17:40:12 +0200 (Mon, 25 Dec 2006) | 3 lines

- another SVN issue solving commit

------------------------------------------------------------------------
r334 | pieterpalmers | 2006-12-25 17:37:39 +0200 (Mon, 25 Dec 2006) | 3 lines

- temporary commit to resove an issue with SVN, another one to follow.

------------------------------------------------------------------------
r333 | pieterpalmers | 2006-12-23 22:56:00 +0200 (Sat, 23 Dec 2006) | 6 lines

- bump version to 1.999.0 such that it stays ahead of 1.x
- fix issue that prevented the code from
  being compiled for --disable-debug

------------------------------------------------------------------------
r332 | jwoithe | 2006-12-04 00:25:37 +0200 (Mon, 04 Dec 2006) | 2 lines

Add the src/rme directory which was inadvertantly omitted from previous commit.

------------------------------------------------------------------------
r331 | jwoithe | 2006-12-04 00:15:46 +0200 (Mon, 04 Dec 2006) | 2 lines

Add detection support for the Fireface-800.  There is no other functionality at this stage.

------------------------------------------------------------------------
r312 | jwoithe | 2006-10-03 01:41:21 +0200 (Tue, 03 Oct 2006) | 5 lines

MOTU: fix xrun recovery glitches inadvertantly introduced in r309.
MOTU: preliminary MIDI port support.  Rate control still needs to be added so it probably won't work in its current form yet.  MIDI has not yet been tested.
MOTU: tweaks to receive/transmit missed cycle detection.
jack driver: during driver shutdown don't try to free NULL ports created for non-audio ports during initialisation to preserve device port index locations.

------------------------------------------------------------------------
r311 | jwoithe | 2006-09-28 01:51:48 +0200 (Thu, 28 Sep 2006) | 3 lines

MOTU: the unknown input channels on the 828MkII are the mic inputs.
MOTU: fix the transposed MainOut/Mic channel labels for the 828MkII.

------------------------------------------------------------------------
r310 | jwoithe | 2006-09-28 01:16:09 +0200 (Thu, 28 Sep 2006) | 2 lines

MOTU: fix optical mode detection and clear up some associated packet sizing issues.

------------------------------------------------------------------------
r309 | jwoithe | 2006-09-25 02:50:17 +0200 (Mon, 25 Sep 2006) | 8 lines

MOTU: Fixed false "missed rx cycle" report following xrun recovery.
Ensure iso rx/tx contexts are deallocated during shutdown/xrun recovery by explicitly deleting IsoHandlers in IsoHandlerManager::pruneHandlers().  If they aren't deleted here they never get deleted because the reference is lost.
IsoHandler destructor should only call stop() if the handle is valid.
IsoXmitHandler's destructor sets the handle NULL to prevent double-free by the inherited IsoHandler destructor.
Don't call raw1394_iso_shutdown() from our code.  libraw1394 1.2.1 has a bug whereby raw1394_new_handle() fails to initialise the iso_packet_infos field.  The bug hits us particularly in IsoRecvHandler::prepare().  It's also not really necessary to call raw1394_iso_shutdown() since raw1394_destroy_handle() will do any cleanups we happen to need.
MOTU: the receive stream no longer falsely complains of buffer problems during device shutdown.
MOTU: fixed a false "missed cycle" detection immediately after the stream was enabled.

------------------------------------------------------------------------
r308 | jwoithe | 2006-09-22 01:18:50 +0200 (Fri, 22 Sep 2006) | 2 lines

Remove another instance of a class name being included in a method declaration which was causing newer gcc versions to trip up.

------------------------------------------------------------------------
r307 | jwoithe | 2006-09-22 01:07:35 +0200 (Fri, 22 Sep 2006) | 2 lines

Minor source code cleanups.  No functional changes.

------------------------------------------------------------------------
r306 | jwoithe | 2006-09-21 01:50:35 +0200 (Thu, 21 Sep 2006) | 2 lines

Add a "/" to the libfreebob path which are needed when compiling in newer environments.

------------------------------------------------------------------------
r305 | jwoithe | 2006-09-21 01:44:29 +0200 (Thu, 21 Sep 2006) | 2 lines

MotuStreamProcessor.h: remove class name in erroneously included in the definition of two transmit stream methods.  Thanks to Francois Ernoult for the bug report.

------------------------------------------------------------------------
r304 | jwoithe | 2006-09-18 09:57:30 +0200 (Mon, 18 Sep 2006) | 3 lines

Whitespace cleanup.
Improve device sample rate configuration.

------------------------------------------------------------------------
r302 | jwoithe | 2006-09-15 04:56:03 +0200 (Fri, 15 Sep 2006) | 4 lines

jack driver: call detach() from finish() if detach() hasn't yet been called.  This works around the fact that jack doesn't call detach() on close at this stage.  Once this is fixed in jack the workaround can be removed.
C streaming API: set sample rate before device initialisation.  This permits jackd to set the sample rate to that requested on the command line when starting jackd.
Motu device: start adding details to setSamplingFrequency() beyond that required for basic functionality.

------------------------------------------------------------------------
r301 | jwoithe | 2006-09-11 09:35:58 +0200 (Mon, 11 Sep 2006) | 3 lines

Port::useExternalBuffer(): don't complain if called after initialisation and the call does not request a change to the external buffer usage.
freebob_driver_attach(): call freebob_streaming_set_capture_stream_buffer() so freebob knows an external buffer will be used.

------------------------------------------------------------------------
r300 | jwoithe | 2006-09-05 02:10:19 +0200 (Tue, 05 Sep 2006) | 4 lines

Major cleanup of MOTU streaming code.  
Fine-tuned the DLL used for iso transmission timestamping.  
Tweaked MOTU shutdown to avoid long delays if streaming isn't operational; this means that xrun recovery nearly always works now.

------------------------------------------------------------------------
r296 | jwoithe | 2006-08-21 03:44:33 +0200 (Mon, 21 Aug 2006) | 7 lines

Code "ticks per frame" DLL inline to the MOTU receive stream processor for
efficiency.  Fine-tune DLL integration coefficient to remove regular audio
glitches.  Set teststreaming2.c to request realtime priority to facilitate
further testing.  Cleanly deal with cycle wraparound in transmit stream
processor.  Improve efficiency of 1 kHz tone generator in transmit stream
processor.

------------------------------------------------------------------------
r294 | jwoithe | 2006-07-19 09:05:56 +0200 (Wed, 19 Jul 2006) | 2 lines

More device sync tests.

------------------------------------------------------------------------
r291 | jwoithe | 2006-07-17 09:44:27 +0200 (Mon, 17 Jul 2006) | 5 lines

Seed the SPH cycle DLL when the receive stream is running but disabled; this
means the DLL has a sensible figure when the transmit stream is enabled.
More testing to work out why device resync (as distinct from rx/tx sync)
doesn't seem to work.

------------------------------------------------------------------------
r290 | jwoithe | 2006-07-14 02:42:27 +0200 (Fri, 14 Jul 2006) | 4 lines

Iso resources (channels, bandwidth) now allocated via the IRM.
Some comments clarified.
Minor code cleanups.

------------------------------------------------------------------------
r289 | jwoithe | 2006-07-12 01:44:36 +0200 (Wed, 12 Jul 2006) | 3 lines

Minor tidy up.  Some comments clarified.
Some test code has been added to the MOTU stream processor to assist in tracking down some sync-related issues.

------------------------------------------------------------------------
r287 | jwoithe | 2006-07-10 01:33:32 +0200 (Mon, 10 Jul 2006) | 4 lines

Some cleanups to the MotuStreamProcessor source.
First pass at a stop preparation method (StreamProcessor::preparedToStop()).
Motu uses preparedToStop to write zero data to device before iso closedown.

------------------------------------------------------------------------
r286 | pieterpalmers | 2006-07-09 00:20:47 +0200 (Sun, 09 Jul 2006) | 3 lines

- fixed small compile error

------------------------------------------------------------------------
r285 | jwoithe | 2006-07-07 09:17:38 +0200 (Fri, 07 Jul 2006) | 3 lines

Connect ports to the Motu playback stream handlers & make playback work.
Make teststreaming2.c copy input ports to output ports for testing.

------------------------------------------------------------------------
r284 | jwoithe | 2006-07-04 01:23:50 +0200 (Tue, 04 Jul 2006) | 2 lines

More MOTU iso transmission work.  Things are now partially operational.

------------------------------------------------------------------------
r283 | jwoithe | 2006-07-03 02:22:25 +0200 (Mon, 03 Jul 2006) | 3 lines

More additions for MOTU playback (transmit) functionality.
Bugfix for MOTU stream stop method.

------------------------------------------------------------------------
r277 | jwoithe | 2006-06-30 01:24:39 +0200 (Fri, 30 Jun 2006) | 3 lines

Start of work leading to MOTU iso transmission.
Fix use of uninitialised local variable in DelayLockedLoop::setIntegrator().

------------------------------------------------------------------------
r275 | pieterpalmers | 2006-06-26 22:36:03 +0200 (Mon, 26 Jun 2006) | 3 lines

- fix bug that prevented RT threads from being created

------------------------------------------------------------------------
r274 | jwoithe | 2006-06-26 03:45:12 +0200 (Mon, 26 Jun 2006) | 3 lines

Fix buffer type configuration in teststreaming{2,3}.c.
Audio data can now be streamed in from a MOTU to teststreaming2.c.

------------------------------------------------------------------------
r272 | jwoithe | 2006-06-23 02:09:24 +0200 (Fri, 23 Jun 2006) | 3 lines

Silence some compiler warnings in teststreaming.c and teststreaming2.c.
Work towards getting iso receive working for the MOTU devices.

------------------------------------------------------------------------
r269 | jwoithe | 2006-06-21 08:46:20 +0200 (Wed, 21 Jun 2006) | 4 lines

More MOTU port infrastructure developments.
Make teststreaming2 compile and maybe work.
Fix some comments in IsoHandlerManager.cpp and freebob_streaming.cpp.

------------------------------------------------------------------------
r268 | jwoithe | 2006-06-20 03:46:29 +0200 (Tue, 20 Jun 2006) | 3 lines

StreamProcessor::m_framecounter is now signed again, since the ability to go
negative is utilised for some devices in certain circumstances.

------------------------------------------------------------------------
r267 | jwoithe | 2006-06-19 02:39:04 +0200 (Mon, 19 Jun 2006) | 5 lines

More Motu streaming infrastructure added.
Added more Motu device configuration details.
tests/teststreaming.c: added call to freebob_streaming_prepare() so this can
naively used to start exercising the Motu functions.

------------------------------------------------------------------------
r266 | pieterpalmers | 2006-06-17 16:15:54 +0200 (Sat, 17 Jun 2006) | 3 lines

- temporary commit

------------------------------------------------------------------------
r265 | jwoithe | 2006-06-16 01:57:46 +0200 (Fri, 16 Jun 2006) | 3 lines

Start adding detail in MOTU port infrastructure.
Clean up some "signed/unsigned comparison" compiler warnings.

------------------------------------------------------------------------
r264 | pieterpalmers | 2006-06-15 18:19:04 +0200 (Thu, 15 Jun 2006) | 3 lines

- fixed small bug introduced by moving the configrom tests

------------------------------------------------------------------------
r263 | pieterpalmers | 2006-06-15 18:07:47 +0200 (Thu, 15 Jun 2006) | 4 lines

- implementation of correct SYT behaviour for transmit streams
  [Work-in-progress: these changes introduce a bug, but I don't
   want to lose this]

------------------------------------------------------------------------
r262 | jwoithe | 2006-06-13 03:31:27 +0200 (Tue, 13 Jun 2006) | 3 lines

Make skeleton MOTU device discovery work.  Proof-of-concept MOTU sample rate
control in place.

------------------------------------------------------------------------
r261 | pieterpalmers | 2006-06-10 15:53:31 +0200 (Sat, 10 Jun 2006) | 4 lines

- enabled verbose debugging for packet handlers in amdtp (PPC issues)
- fixed iec61883_packet struct to work on PPC.

------------------------------------------------------------------------
r255 | pieterpalmers | 2006-06-10 00:33:51 +0200 (Sat, 10 Jun 2006) | 6 lines

- the syt extraction is finally correct, handling all wraparounds correctly
- when receiving a stream the device's samplerate is estimated and this estimate 
  is used to calculate the timestamps for the transmitted packets
  (needs some more work)

------------------------------------------------------------------------
r254 | pieterpalmers | 2006-06-09 22:28:32 +0200 (Fri, 09 Jun 2006) | 1 line

- fix file problems
------------------------------------------------------------------------
r253 | pieterpalmers | 2006-06-09 22:26:17 +0200 (Fri, 09 Jun 2006) | 3 lines

- trying to fix problems with file locations introcuded earlier

------------------------------------------------------------------------
r250 | pieterpalmers | 2006-06-06 17:09:38 +0200 (Tue, 06 Jun 2006) | 6 lines

- removed old streaming code
- removed some obsolete files
- moved utility classes to libutil
- don't do the doxygen processing on make all, explicit make doc in doc/ is now nescessary (speed)

------------------------------------------------------------------------
r249 | pieterpalmers | 2006-06-06 00:12:55 +0200 (Tue, 06 Jun 2006) | 3 lines

- Extensive documentation update.

------------------------------------------------------------------------
r248 | pieterpalmers | 2006-06-05 16:32:01 +0200 (Mon, 05 Jun 2006) | 3 lines

- add API version check

------------------------------------------------------------------------
r247 | pieterpalmers | 2006-06-05 15:44:48 +0200 (Mon, 05 Jun 2006) | 3 lines

- added a modified jackd backend for version 2.0

------------------------------------------------------------------------
r246 | pieterpalmers | 2006-06-05 15:41:55 +0200 (Mon, 05 Jun 2006) | 4 lines

- re-enabled the Motu device and updated it to the IAvDevice changes
- created a framework streamprocessor for the motu devices

------------------------------------------------------------------------
r245 | pieterpalmers | 2006-06-05 12:05:52 +0200 (Mon, 05 Jun 2006) | 5 lines

- added doxygen documentation to the make process
- wrote some preliminary documentation

------------------------------------------------------------------------
r244 | pieterpalmers | 2006-06-05 10:13:34 +0200 (Mon, 05 Jun 2006) | 5 lines

- first try to implement the SYT synchronisation. 
  Not working yet.
- committing to spend some time to get the Motu people going.

------------------------------------------------------------------------
r243 | pieterpalmers | 2006-06-03 22:45:22 +0200 (Sat, 03 Jun 2006) | 1 line

added some statistics gathering to help me figure out internals
------------------------------------------------------------------------
r242 | pieterpalmers | 2006-06-03 22:39:12 +0200 (Sat, 03 Jun 2006) | 3 lines

- made the bounce device actually work

------------------------------------------------------------------------
r235 | pieterpalmers | 2006-05-29 22:53:54 +0200 (Mon, 29 May 2006) | 4 lines

- implemented enable/disable for the ports
- improved the C api somewhat

------------------------------------------------------------------------
r234 | pieterpalmers | 2006-05-29 21:14:23 +0200 (Mon, 29 May 2006) | 3 lines

- ported over the port naming change from the 1.0 branch

------------------------------------------------------------------------
r233 | pieterpalmers | 2006-05-29 20:40:23 +0200 (Mon, 29 May 2006) | 4 lines

- reworked the debug module to use RT safe messaging

------------------------------------------------------------------------
r232 | pieterpalmers | 2006-05-29 19:57:52 +0200 (Mon, 29 May 2006) | 1 line

------------------------------------------------------------------------
r230 | pieterpalmers | 2006-05-28 19:40:49 +0200 (Sun, 28 May 2006) | 3 lines

- xrun handling now works

------------------------------------------------------------------------
r229 | pieterpalmers | 2006-05-28 16:07:03 +0200 (Sun, 28 May 2006) | 1 line

------------------------------------------------------------------------
r228 | pieterpalmers | 2006-05-28 14:43:20 +0200 (Sun, 28 May 2006) | 3 lines

- generic rate control added for ringbuffer based ports.

------------------------------------------------------------------------
r227 | pieterpalmers | 2006-05-27 23:43:40 +0200 (Sat, 27 May 2006) | 5 lines

- another day of good progress comes to and end...
- compiles and runs, only the midi stuff and the xrun handling remain.
  I'll also have to rework the C API somewhat.

------------------------------------------------------------------------
r226 | pieterpalmers | 2006-05-27 17:30:20 +0200 (Sat, 27 May 2006) | 3 lines

- ported the more efficient buffer handling from libfreebob1.0

------------------------------------------------------------------------
r225 | pieterpalmers | 2006-05-27 16:37:53 +0200 (Sat, 27 May 2006) | 5 lines

- reworked the code
- this runs on a bebob device, as long as no xruns occur
- partly documented

------------------------------------------------------------------------
r223 | pieterpalmers | 2006-05-25 22:14:18 +0200 (Thu, 25 May 2006) | 4 lines

- this is the first audio capable commit of the 2.0 code
- xrun handling etc.. is not ready yet

------------------------------------------------------------------------
r222 | pieterpalmers | 2006-05-25 12:03:15 +0200 (Thu, 25 May 2006) | 3 lines

- makefile rework: move headers in noinst, makes life easier with kdevelop

------------------------------------------------------------------------
r221 | pieterpalmers | 2006-05-25 01:14:20 +0200 (Thu, 25 May 2006) | 1 line

------------------------------------------------------------------------
r220 | pieterpalmers | 2006-05-24 15:44:38 +0200 (Wed, 24 May 2006) | 1 line

------------------------------------------------------------------------
r218 | pieterpalmers | 2006-05-24 14:10:08 +0200 (Wed, 24 May 2006) | 3 lines

- documentation start

------------------------------------------------------------------------
r217 | pieterpalmers | 2006-05-22 23:54:50 +0200 (Mon, 22 May 2006) | 3 lines

- implemented receive streaming to MBLA audio ports

------------------------------------------------------------------------
r212 | pieterpalmers | 2006-05-21 22:17:28 +0200 (Sun, 21 May 2006) | 3 lines

- end of day commit

------------------------------------------------------------------------
r211 | pieterpalmers | 2006-05-21 21:41:05 +0200 (Sun, 21 May 2006) | 4 lines

- fixed recursion problem

------------------------------------------------------------------------
r210 | pieterpalmers | 2006-05-21 21:10:55 +0200 (Sun, 21 May 2006) | 1 line

------------------------------------------------------------------------
r209 | pieterpalmers | 2006-05-21 20:53:50 +0200 (Sun, 21 May 2006) | 2 lines

backup commit (2)

------------------------------------------------------------------------
r208 | pieterpalmers | 2006-05-21 20:51:29 +0200 (Sun, 21 May 2006) | 1 line

------------------------------------------------------------------------
r207 | pieterpalmers | 2006-05-21 20:39:59 +0200 (Sun, 21 May 2006) | 3 lines

- temp commit

------------------------------------------------------------------------
r206 | pieterpalmers | 2006-05-20 22:12:11 +0200 (Sat, 20 May 2006) | 3 lines

- end of another day... implemented the 'ports' that represent the demuxed data streams (audio, midi and control)

------------------------------------------------------------------------
r205 | pieterpalmers | 2006-05-19 20:17:12 +0200 (Fri, 19 May 2006) | 3 lines

- test files

------------------------------------------------------------------------
r204 | pieterpalmers | 2006-05-19 20:16:51 +0200 (Fri, 19 May 2006) | 4 lines

- ISO streaming infrastructure starts to look good
- client side infrastructure unexistant

------------------------------------------------------------------------
r203 | pieterpalmers | 2006-05-19 15:31:44 +0200 (Fri, 19 May 2006) | 3 lines

- another backup commit, going to switch the streamhandlermanager to a rx/tx invariant one.

------------------------------------------------------------------------
r202 | pieterpalmers | 2006-05-19 10:37:34 +0200 (Fri, 19 May 2006) | 2 lines

backup commit (2)

------------------------------------------------------------------------
r201 | pieterpalmers | 2006-05-19 10:36:34 +0200 (Fri, 19 May 2006) | 2 lines

backup commit

------------------------------------------------------------------------
r200 | pieterpalmers | 2006-05-18 18:17:03 +0200 (Thu, 18 May 2006) | 3 lines

- temp commit

------------------------------------------------------------------------
r199 | pieterpalmers | 2006-05-17 11:57:45 +0200 (Wed, 17 May 2006) | 2 lines

- start of a new streaming API implementation, C++ based

------------------------------------------------------------------------
r198 | pieterpalmers | 2006-05-16 19:40:07 +0200 (Tue, 16 May 2006) | 3 lines

- branched off to work on post-1.0 code

------------------------------------------------------------------------
r194 | pieterpalmers | 2006-05-12 18:19:08 +0200 (Fri, 12 May 2006) | 5 lines

- code cleanup in the streaming part.
- calculate ISO connection parameters instead of user-supplying them
- SSE code improvement

------------------------------------------------------------------------
r192 | wagi | 2006-05-06 10:40:30 +0200 (Sat, 06 May 2006) | 5 lines

This patch fixes two issues when building:
 * libfreebob include files aren't found when build dir != source dir
 * some libraries aren't found when they're not in a system install dir

Jeremy Kerr
------------------------------------------------------------------------
r191 | pieterpalmers | 2006-05-05 20:49:19 +0200 (Fri, 05 May 2006) | 2 lines

- changelog update

------------------------------------------------------------------------
r190 | pieterpalmers | 2006-05-05 20:47:48 +0200 (Fri, 05 May 2006) | 4 lines

- fixed SSE-accellerated encoders/decoders
- moved SYT CIP_TRANSFER_DELAY to cip.h

------------------------------------------------------------------------
r189 | pieterpalmers | 2006-05-04 22:34:36 +0200 (Thu, 04 May 2006) | 3 lines

- added SSE-accelerated code for decoding the packets, but it's not completely working yet.

------------------------------------------------------------------------
r188 | wagi | 2006-05-04 00:17:38 +0200 (Thu, 04 May 2006) | 8 lines

2006-05-04  Daniel Wagner  <wagi@monom.org>

* configure.ac: Version bumped to 0.8.0

* Discovering code ported stable 0.2.0 version
	to head and activated for default discovering.

------------------------------------------------------------------------
r187 | wagi | 2006-05-04 00:10:41 +0200 (Thu, 04 May 2006) | 7 lines

2006-05-04  Daniel Wagner  <wagi@monom.org>

* configure.ac: Version bumped to 0.7.2

* src/bebob: Some cleanup on naming and
	namespaces.

------------------------------------------------------------------------
r186 | wagi | 2006-05-03 23:05:14 +0200 (Wed, 03 May 2006) | 7 lines

2006-05-03  Daniel Wagner  <wagi@monom.org>

* configure.ac: Version bumped to 0.7.1

* Fixed bug in SignalSource command which
	prevented one to set the samplerate.

------------------------------------------------------------------------
r185 | wagi | 2006-04-24 22:59:21 +0200 (Mon, 24 Apr 2006) | 2 lines

CVS-SVN migration developer public sync patch

------------------------------------------------------------------------
r183 | wagi | 2006-03-30 00:57:53 +0200 (Thu, 30 Mar 2006) | 20 lines

2006-03-30  Daniel Wagner  <wagi@monom.org>

* configure.ac: Version bump to 0.6.0

* src/*: Verbose bool turned int verbose level.

* src/libfreebobavc/*: Moved fire implementation to
	base class (AVCCommand).

* libfreebob/freebob.h: freebob_sleep_after_avc_command
	workaround added.

* tests/test-freebob.c: Verbose bool turned into verbose level.
	Workaround: sleep after avc command added.

* tests/streaming/Makefile.am (*_SOURCES): Added libavc1394 to LDADD.
	* tests/Makefile.am (*_SOURCES): Likewise.

* tests/streaming/teststreaming2.c (fid_in): compiler warning fixed.

------------------------------------------------------------------------
r181 | wagi | 2006-03-21 00:50:25 +0200 (Tue, 21 Mar 2006) | 8 lines

2006-03-20  Daniel Wagner  <wagi@monom.org>

* configure.ac: Version bump to 0.5.2

* src/avplug.cpp (discoverStreamFormat): If no stream format information
	is available or format hierarchy root is not Audio&Music skip plug.
	(discoverSupportedStreamFormats): Added 'support' for IEC60968-3 (SPDIF transparent)

------------------------------------------------------------------------
r175 | wagi | 2006-03-15 01:12:08 +0200 (Wed, 15 Mar 2006) | 12 lines

2006-03-15  Daniel Wagner  <wagi@monom.org>

* configure.ac: Version bump to 0.5.1

* src/libfreebobavc/avc_extended_plug_info.cpp (fire):
	  Handle NOT_IMPLEMENTED.

* src/avplug.h: Input and output plug discovering part
	fixed. Now connection are correctly handled.
	* src/avplug.cpp: Likewise.
	* src/avdevicesubunit.cpp: Likewise.

------------------------------------------------------------------------
r174 | wagi | 2006-03-12 21:20:31 +0200 (Sun, 12 Mar 2006) | 5 lines

2006-03-12  Daniel Wagner  <wagi@monom.org>

* src/libfreebobavc/ieee1394service.cpp (transactionBlock): Set retry
	value to 10.

------------------------------------------------------------------------
r173 | wagi | 2006-03-10 01:40:21 +0200 (Fri, 10 Mar 2006) | 13 lines

2006-03-10  Daniel Wagner  <wagi@monom.org>

* configure.ac: Version bump to 0.5.0

* src/avdevice.cpp: Plug connection discovering
	finished.
	* src/avdevicesubunit.cpp: Likewise.
	* src/avdevicesubunit.h: Likewise.
	* src/avplug.cpp: Likewise.
	* src/avplug.h: Likewise.
	* src/functionblock.cpp: Likewise.
	* src/functionblock.h: Likewise.

------------------------------------------------------------------------
r172 | wagi | 2006-03-10 01:35:47 +0200 (Fri, 10 Mar 2006) | 12 lines

2006-03-10  Daniel Wagner  <wagi@monom.org>

* src/libfreebobavc/Makefile.am: New files added.
	* src/libfreebobavc/avc_extended_subunit_info.cpp: Likewise.
	* src/libfreebobavc/avc_extended_subunit_info.h: Likewise

* src/libfreebobavc/avc_definitions.h: Several small improvements.
	* src/libfreebobavc/avc_extended_cmd_generic.cpp: Likewise.
	* src/libfreebobavc/avc_extended_cmd_generic.h: Likewise.
	* src/libfreebobavc/avc_extended_plug_info.cpp: Likewise.
	* src/libfreebobavc/avc_generic.cpp: Likewise.

------------------------------------------------------------------------
r171 | wagi | 2006-03-05 02:36:53 +0200 (Sun, 05 Mar 2006) | 16 lines

2006-03-05  Daniel Wagner  <wagi@monom.org>

* src/avdevice.cpp: Plug connection discovering also on
	subunits (disabled because missing funciton blocks).
	Sync connections code cleanup.
	* src/avdevice.h: Likewise.
	* src/avdevicesubunit.cpp: Likewise.
	* src/avdevicesubunit.h: Likewise.
	* src/avplug.cpp: Likewise.
	* src/avplug.h: Likewise.

* src/Makefile.am: Added functionblock.cpp and
	functionblock.h
	* src/functionblock.cpp: New file
	* src/functionblock.h: Likewise.

------------------------------------------------------------------------
r170 | wagi | 2006-03-04 19:45:03 +0200 (Sat, 04 Mar 2006) | 7 lines

2006-03-04  Daniel Wagner  <wagi@monom.org>

* src/avdevice.cpp: Sync mode discovering improved.
	* src/avdevice.h: Likewise.
	* src/avplug.cpp: Likewise.
	* src/avplug.h: Likewise.

------------------------------------------------------------------------
r169 | wagi | 2006-03-04 19:42:34 +0200 (Sat, 04 Mar 2006) | 5 lines

2006-03-04  Daniel Wagner  <wagi@monom.org>

* tests/test-freebob.c: Added 'odiscover' command which
	does not print the xml stuff.

------------------------------------------------------------------------
r168 | wagi | 2006-03-04 17:14:28 +0200 (Sat, 04 Mar 2006) | 16 lines

2006-03-04  Daniel Wagner  <wagi@monom.org>

* src/avdevice.cpp: Added 'verbose' argument to constructor
	and set debug level to DEBUG_LEVEL_VERBOSE if 'verbose'
	is true, e.g enables verbose mode with '-v' argument passed
	to test-freebob.
	* src/avdevice.h: Likewsise.
	* src/avdevicesubunit.cpp: Likewsise.
	* src/avdevicesubunit.h: Likewsise.
	* src/avplug.cpp: Likewsise.
	* src/avplug.h: Likewsise.
	* src/devicemanager.cpp: Likewsise.
	* src/debugmodule/debugmodule.cpp: Likewsise.
	* src/debugmodule/debugmodule.h: Likewsise.
	* src/debugmodule/test_debugmodule.cpp: Likewsise.

------------------------------------------------------------------------
r167 | wagi | 2006-03-04 16:44:57 +0200 (Sat, 04 Mar 2006) | 13 lines

2006-03-04  Daniel Wagner  <wagi@monom.org>

* src/avdevice.cpp: eAP_* renamed into eAPA_*.
	EAvPlugDirection and EAvPlugType introduced. All PlugAddress:ePD_ and
	ExtendedPlugInfoPlugTypeSpecificData::eEPIPT_* exchanged with corresponding
	AvPlug definitions.
	* src/avdevice.h: Likewise.
	* src/avdevicesubunit.cpp: Likewise.
	* src/avdevicesubunit.h: Likewise.
	* src/avdevicexml.cpp: Likewise.
	* src/avplug.cpp: Likewise.
	* src/avplug.h: Likewise.

------------------------------------------------------------------------
r166 | wagi | 2006-03-02 23:53:36 +0200 (Thu, 02 Mar 2006) | 2 lines

* src/avplug.cpp (discoverStreamFormat): Sanity check turned into warning.

------------------------------------------------------------------------
r165 | wagi | 2006-03-02 02:41:38 +0200 (Thu, 02 Mar 2006) | 27 lines

2006-03-02  Daniel Wagner  <wagi@monom.org>

* configure.ac: bumb minor version to 0.4

* src/libfreebobavc/avc_extended_cmd_generic.h: plugAddressPlugDirectionToString
	and plugAddressAddressModeToString added.
	* src/libfreebobavc/avc_extended_cmd_generic.cpp: Likewise.

* src/avplug.h: Enum EAvPlugType renamed to EAvPlugAddressType.
	getPlugAddressType added. getPlugType returns now real plug type.
	(class AvPlugManager): getPlugByType added.

* src/avdevice.h: getPlugByType added.
	* src/avdevice.cpp: Likewise.
	(discoverSyncModes): Use getPlugByType instead
	of hardcoded id to find sync plugs.

* src/devicemanager.h: Verbose argument not in constructor
	instead in discover method.
	* src/devicemanager.cpp: Likewise.

* libfreebob/freebob.h (freebob_discover_devices): Verbose argument
	added.
	* src/libfreebobstreaming/freebob_streaming.c: Likewise.
	* src/freebob.cpp: Likewise.
	* tests/test-freebob.c: Likewise.

------------------------------------------------------------------------
r164 | wagi | 2006-03-01 01:29:09 +0200 (Wed, 01 Mar 2006) | 36 lines

* configure.ac: bumb minor version to 0.3

* src/avdevice.cpp: Discover connections between plugs.
* src/avdevice.h: Likewise.
* src/avdevicesubunit.cpp: Likewise.
* src/avdevicesubunit.h: Likewise.
* src/avdevicexml.cpp: Likewise.
* src/avplug.cpp: Likewise.
* src/avplug.h: Likewise.
* src/devicemanager.cpp: Likewise.
* src/devicemanager.h: Likewise.
* src/freebob.cpp: Likewise.

* src/libfreebobavc/avc_extended_cmd_generic.cpp: Give
return value to caller. New undefined plug address
introduced. Handling of response message is more rebust now.
* src/libfreebobavc/avc_extended_cmd_generic.h: Likewise.
* src/libfreebobavc/avc_extended_plug_info.cpp: Likewise.
* src/libfreebobavc/avc_extended_stream_format.cpp: Likewise.
* src/libfreebobavc/avc_extended_stream_format.h: Likewise.
* src/libfreebobavc/avc_plug_info.cpp: Likewise.
* src/libfreebobavc/avc_signal_source.cpp: Likewise.
* src/libfreebobavc/avc_signal_source.h: Likewise.
* src/libfreebobavc/avc_subunit_info.cpp: Likewise.
* src/libfreebobavc/avc_unit_info.cpp: Likewise.
* src/libfreebobavc/avc_extended_plug_info.cpp: Likewise.
* src/libfreebobavc/avc_extended_stream_format.cpp: Likewise.
* src/libfreebobavc/avc_extended_cmd_generic.cpp:
* src/libfreebobavc/avc_extended_cmd_generic.h: Likewise.

* src/libfreebobstreaming/freebob_debug.h: DEBUG is defined through --enable-debug
switch.

* src/libfreebobstreaming/freebob_connections.h: Removed double defined variable
if DEBUG is defined.

------------------------------------------------------------------------
r163 | wagi | 2006-03-01 01:27:15 +0200 (Wed, 01 Mar 2006) | 38 lines

2006-03-01  Daniel Wagner  <wagi@monom.org>

* configure.ac: bumb minor version to 0.3

* src/avdevice.cpp: Discover connections between plugs.
	* src/avdevice.h: Likewise.
	* src/avdevicesubunit.cpp: Likewise.
	* src/avdevicesubunit.h: Likewise.
	* src/avdevicexml.cpp: Likewise.
	* src/avplug.cpp: Likewise.
	* src/avplug.h: Likewise.
	* src/devicemanager.cpp: Likewise.
	* src/devicemanager.h: Likewise.
	* src/freebob.cpp: Likewise.

* src/libfreebobavc/avc_extended_cmd_generic.cpp: Give
	return value to caller. New undefined plug address
	introduced. Handling of response message is more rebust now.
	* src/libfreebobavc/avc_extended_cmd_generic.h: Likewise.
	* src/libfreebobavc/avc_extended_plug_info.cpp: Likewise.
	* src/libfreebobavc/avc_extended_stream_format.cpp: Likewise.
	* src/libfreebobavc/avc_extended_stream_format.h: Likewise.
	* src/libfreebobavc/avc_plug_info.cpp: Likewise.
	* src/libfreebobavc/avc_signal_source.cpp: Likewise.
	* src/libfreebobavc/avc_signal_source.h: Likewise.
	* src/libfreebobavc/avc_subunit_info.cpp: Likewise.
	* src/libfreebobavc/avc_unit_info.cpp: Likewise.
	* src/libfreebobavc/avc_extended_plug_info.cpp: Likewise.
	* src/libfreebobavc/avc_extended_stream_format.cpp: Likewise.
	* src/libfreebobavc/avc_extended_cmd_generic.cpp:
	* src/libfreebobavc/avc_extended_cmd_generic.h: Likewise.

* src/libfreebobstreaming/freebob_debug.h: DEBUG is defined through --enable-debug
	switch.

* src/libfreebobstreaming/freebob_connections.h: Removed double defined variable
	if DEBUG is defined.

------------------------------------------------------------------------
r160 | pieterpalmers | 2006-02-28 16:24:59 +0200 (Tue, 28 Feb 2006) | 2 lines

- fixed some build errors. A fresh checkout should compile.

------------------------------------------------------------------------
r159 | pieterpalmers | 2006-02-28 15:53:37 +0200 (Tue, 28 Feb 2006) | 3 lines

- Added streaming API
- Added test programs for the streaming API

------------------------------------------------------------------------
r158 | wagi | 2006-02-26 18:58:04 +0200 (Sun, 26 Feb 2006) | 12 lines

* src/avdevice.cpp: Some more refactoring (m_isoInputPlugs and m_isoOutputPlugs
unified into m_isoPlugs). Discovering of unit external plugs.
* src/avdevice.h: Likewise.
* src/avdevicexml.cpp: Likewise.
* src/avdevicesubunit.cpp: Likewise.
* src/avdevicesubunit.h: Likewise.
* src/avplug.cpp: Likewise.
* src/avplug.h: Likewise.

* src/libfreebobavc/avc_extended_cmd_generic.h (class UnitPlugAddress):
ePT_Unknown enum added.

------------------------------------------------------------------------
r157 | wagi | 2006-02-22 01:42:19 +0200 (Wed, 22 Feb 2006) | 4 lines

Refactoring: plug discovering is now part of AvPlug instead of AvDevice.
All plugs of music and audio subunit are now also discovered.
libfreebobavc: Copy constructor added.

------------------------------------------------------------------------
r155 | wagi | 2006-01-19 02:21:39 +0200 (Thu, 19 Jan 2006) | 21 lines

2006-01-19  Daniel Wagner  <wagi@monom.org>

* This patch cleans up the external dependency detection for
	libfreebob. At present, building libfreebob is not possible
	if libraries are outside the system include/library paths
	(ie /usr/ and /usr/local). Use the results of pkg-config to
	specify the correct paths to the automake system.

Becuase we're using pkgconfig, there's no real need to use
	AC_CHECK_LIB or AC_CHECK_HEADERS. However, these could
	potentially be reinstated, but need to be fixed to check in
	the locations specified by pkgconfig.

* It is not currently possible to build freebob in a directory
	outside the source dir (ie, builddir != srcdir). This patch
	fixes the include paths to allow this. (it's handy for keeping
	the source tree clean, or when building on multiple architectures
	- I use a build dir of $srcdir/obj-$arch-$os)

Both patches are written by Jeremy Kerr <jk@ozlabs.org>

------------------------------------------------------------------------
r154 | wagi | 2006-01-12 17:04:04 +0200 (Thu, 12 Jan 2006) | 5 lines

2006-01-12  Daniel Wagner  <wagi@monom.org>

* src/avdevice.cpp: Include stdint.h.
	Some compilers seem to need the definitions.

------------------------------------------------------------------------
r153 | wagi | 2006-01-09 00:19:28 +0200 (Mon, 09 Jan 2006) | 9 lines

2006-01-08  Daniel Wagner  <wagi@monom.org>

*  libfreebob/freebob.h
	libfreebob/xmlparser.h
	src/avdevice.cpp
	src/xmlparser.c
	tests/test-freebob.c:
	Add stream format info support.

------------------------------------------------------------------------
r152 | wagi | 2006-01-08 22:13:28 +0200 (Sun, 08 Jan 2006) | 6 lines

2006-01-08  Daniel Wagner  <wagi@monom.org>

* src/avdevice.h
	src/avdevice.cpp:
	Refactoring and code duplication removal.

------------------------------------------------------------------------
r151 | wagi | 2006-01-08 20:44:20 +0200 (Sun, 08 Jan 2006) | 35 lines

2006-01-08  Daniel Wagner  <wagi@monom.org>

* src/avdevice.cpp (getModelName):
	Return model name and not vendor name.
	(getVendorName):
	Return vendor name and not model name.
	(discoverStep10Plug): Implementation added.
	(enumerateSubUnits): max_subunit_id removed.
	(addXmlDescription): Add stream format infos.

* src/avdevice.h: discoverStep10 added.
	Renaming of addPlugToXmlDescription to addXmlDescriptionPlug.
	ESampleRate to ESamplingFrequency changes.

* src/avplug.cpp (getSampleRate):
	Use convertESamplingFrequency instead of
	local implementation.

* src/avplug.h:
	ESampleRate to ESamplingFrequency changes.
	Added FormatInfo struct.

* src/devicemanager.cpp (getXmlDescription):
	Print GUID correctly.

* src/freebob.cpp (parseSampleRate):
	Moved to avc_definitions.cpp

* src/libfreebobavc/avc_definitions.h
	src/libfreebobavc/avc_definitions.cpp:
	Renamed ESampleRate back to ESamlingFrequency.
	Added connvert functions for ESamplingFrequency.

* configure.ac: New micro version.

------------------------------------------------------------------------
r146 | wagi | 2006-01-06 20:04:18 +0200 (Fri, 06 Jan 2006) | 16 lines

2006-01-06  Daniel Wagner  <wagi@monom.org>

* tests/test-freebob.c (main): setsamplerate on all
	devices takes now the samplerate argument.

* src/freebob.cpp (freebob_set_samplerate): After setting
	sampling rate rediscover devices.

* src/devicemanager.cpp (discover): Clear vector as well after
	delete all elements. Cout and cerr replaced with debugModule.

* src/avdevice.cpp (discoverStep8): Skip sync plug.
	(discoverStep9): Sanity check fixed and reenabled.

* configure.ac: New micro version.

------------------------------------------------------------------------
r145 | wagi | 2006-01-06 01:15:36 +0200 (Fri, 06 Jan 2006) | 6 lines

2006-01-06  Daniel Wagner  <wagi@monom.org>

* src/libfreebobavc/avc_definitions.h
	src/libfreebobavc/avc_definitions.cpp: Changed ESampleFrequency
	to ESampleRate.

------------------------------------------------------------------------
r144 | wagi | 2006-01-06 01:08:22 +0200 (Fri, 06 Jan 2006) | 17 lines

2006-01-06  Daniel Wagner  <wagi@monom.org>

* src/avdevice.h
	src/avdevice.cpp (setSampleRate): Implementation added.

* src/devicemanager.cpp
	src/devicemanager.h (getAvDevice): New functions
	(setNodeSampleFrequency): removed.

* src/freebob.cpp (parseSampleRate): New function.
	(freebob_set_samplerate): Use AvDevice directly
	for setting samplerate.

* src/libfreebobavc/avc_definitions.h
	src/libfreebobavc/avc_definitions.cpp: Changed ESampleFrequency
	to ESampleRate.

------------------------------------------------------------------------
r143 | wagi | 2006-01-06 01:03:10 +0200 (Fri, 06 Jan 2006) | 6 lines

2006-01-06  Daniel Wagner  <wagi@monom.org>

* tests/test-freebob.c: Info added about OPERATION arguments.
	Parse sample rate for setsamplerate command.
	Let emacs reindent code. Sorry Pieter :)

------------------------------------------------------------------------
r142 | wagi | 2006-01-04 01:08:13 +0200 (Wed, 04 Jan 2006) | 10 lines

2006-01-04  Daniel Wagner  <wagi@monom.org>

* tests/test-extplugcmd.cpp: Changed test application
	to use PlugType and PlugName command for testing.

* src/libfreebobavc/avc_extended_plug_info.cpp (serialize):
	The string.length() thing did not work as expected. A workaround
	added which converts the C++ string into a C string before strlen
	is used.

------------------------------------------------------------------------
r141 | wagi | 2006-01-04 00:35:09 +0200 (Wed, 04 Jan 2006) | 6 lines

2006-01-03  Daniel Wagner  <wagi@monom.org>

* src/devicemanager.cpp (discover): If config rom initializing
	failes just skip this device. The PHY might be in power safe
	mode and then there is no config rom to be discovered.

------------------------------------------------------------------------
r140 | pieterpalmers | 2006-01-02 23:36:45 +0200 (Mon, 02 Jan 2006) | 2 lines

* some small memory access bugfixes

------------------------------------------------------------------------
r139 | pieterpalmers | 2006-01-02 19:24:29 +0200 (Mon, 02 Jan 2006) | 2 lines

* added support to detect & use multiple devices on the bus

------------------------------------------------------------------------
r138 | pieterpalmers | 2006-01-02 18:10:34 +0200 (Mon, 02 Jan 2006) | 12 lines

* implemented dummy set samplerate functions

* fixed NodeId bug

* commented out sanity check that failed for midi channels

* fixed Dimension bug

* increased all "XXX random length" requests

* made a mess of clean code ;)

------------------------------------------------------------------------
r137 | pieterpalmers | 2006-01-02 18:08:10 +0200 (Mon, 02 Jan 2006) | 13 lines

* added freebob_print_xml_description
   to print an XML description of the bus

* added freebob_node_is_valid_freebob_device()

* added freebob_get_nb_devices_on_bus()

* added freebob_get_device_node_id()

* added freebob_set_samplerate()
    as a dummy (unimplemented)

------------------------------------------------------------------------
r136 | pieterpalmers | 2006-01-02 18:05:46 +0200 (Mon, 02 Jan 2006) | 3 lines

* added support for scanning the whole bus
* added xmldump operation to dump an XML description of the FW bus

------------------------------------------------------------------------
r135 | wagi | 2006-01-02 15:48:15 +0200 (Mon, 02 Jan 2006) | 3 lines

(discoverStep9): Sanity check: 8 logical
midi channel fit into 1 channel.

------------------------------------------------------------------------
r134 | wagi | 2006-01-02 15:46:38 +0200 (Mon, 02 Jan 2006) | 2 lines

cvsignore files added

------------------------------------------------------------------------
r133 | wagi | 2005-12-31 18:50:25 +0200 (Sat, 31 Dec 2005) | 16 lines

* src/debugmodule/debugmodule.h: debugFatal, debugError
and debugWarning declaration always declared.

* src/xmlparser.c (freebob_xmlparse_connection): Parse IsoChannel
node.

* src/freebob.cpp (freebob_print_connection_info): Print all
members from connection_spec (IsoChannel and IsMaster).

* src/avdevice.cpp (addPlugToXmlDescription): Set direction
according library interface definition and not according
avc model.
Add IsoChannel node.

* configure.ac: Version set to 0.0.2

------------------------------------------------------------------------
r132 | wagi | 2005-12-30 18:53:48 +0200 (Fri, 30 Dec 2005) | 8 lines

2005-12-30  Daniel Wagner  <wagi@monom.org>

* src/libfreebobavc/avc_extended_stream_format.cpp (fire): Buffer
	length increased.

* src/avdevice.cpp (discoverStep9): Error message extended.
	More debug info added.

------------------------------------------------------------------------
r131 | wagi | 2005-12-30 16:33:54 +0200 (Fri, 30 Dec 2005) | 7 lines

2005-12-30  Daniel Wagner  <wagi@monom.org>

* src/avdevice.cpp (discoverStep9): Error message extended.

* src/debugmodule/debugmodule.h: Enable debugFatal, debugError and
	also in the release case.

------------------------------------------------------------------------
r130 | wagi | 2005-12-28 17:48:55 +0200 (Wed, 28 Dec 2005) | 14 lines

* tests/test-freebob.c (main): Debug info added.

* src/libfreebobavc/ieee1394service.cpp (initialize): Use
raw1394_new_handle_on_port instead of raw1394_new_handle and
raw1394_set_port.

* src/devicemanager.cpp (initialize): After deleteing 1394 service
clear also pointer.

* libfreebob/freebob.h: freebob_direction enum names changed
members which jack backed freebob needs to freebob_connection added

* configure.ac: Update library version depency.

------------------------------------------------------------------------
r129 | wagi | 2005-12-27 11:22:06 +0200 (Tue, 27 Dec 2005) | 6 lines

2005-12-27  Daniel Wagner  <wagi@monom.org>

* src/libfreebobavc/avc_extended_plug_info.cpp (serialize):
	Do cast of the length field to let the compiler know which version
	of the serializing code it has to take.

------------------------------------------------------------------------
r125 | wagi | 2005-12-26 17:48:15 +0200 (Mon, 26 Dec 2005) | 2 lines

Initial revision

------------------------------------------------------------------------

Places

File ChangeLog of Package ffado (Revision af6b1d434d875699d420d4a174942169)

Places