Since multiple codec drivers already use the input-jack stuff, let's
make common helper functions to reduce the duplicated codes.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The HDA specification does not allow for a codec to mute itself just
because the volume is reduced, so _of course_ somebody had to go and do
it. This wouldn'\''t hurt too much when the volume is adjusted by hand,
but programs like PA that try to set the volume automatically could
inadvertently mute the output.
To work around this, change the TLV dB information for the Master volume
on all Sigmatel HDA codecs to indicate the the minimal volume setting
actually mutes.
Reported-by: Colin Guthrie <gmane@colin.guthr.ie>
Reported-by: "Alexander E. Patrakov" <patrakov@gmail.com>
Tested-by: Colin Guthrie <cguthrie@mandriva.org>
Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Make the helper function to give the input-pin attribute for jack
connectivity and location. This simplifies checks of input-pin jacks
a bit in some places.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This patch improves the input-source label strings to be generated from
the pin information instead of fixed strings per AUTO_PIN_* type.
This gives more suitable labels, especially for mic and line-in pins.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
We can assign multiple pins to a single role now, let's reduce the
redundant FRONT_MIC and FRONT_LINE. Also, autocfg->input_pins[] is
no longer used, so this is removed as well.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Keep char array in the input_mux item itself instead of pointing to
an external string. This is a preliminary work for improving the
input-mux name based on the pin role.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Added snd_hda_get_input_pin_label() helper function to return the
string that can be used for control or capture-source ids.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Added the new fields to contain all input-pins to struct auto_pin_cfg.
Unlike the existing input_pins[], this array contains all input pins
even if the multiple pins are assigned for a single role (i.e. two
front mics). The former input_pins[] still remains for a while, but
will be removed in near future.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
When a device is plugged over HDMI, it passes some information in ELD
including the supported PCM parameters like formats, rates, channels.
This patch adds the check to PCM open callback of HDMI streams so that
only valid parameters the device supports are used.
When no device is plugged, the parameters the codec supports are used;
it's mostly all parameters the hardware can work. This is for apps
that are started before device plugging and do probing (e.g. a sound
daemon), so that at least, probing would work even before the device
plugging.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
I think this should be automatic pin instead of ping.
(but could be wrong).
Signed-off-by: Justin P. Mattock <justinmattock@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
Fix the wrong implementation of NID <-> kctl mapping for capture mixers
introduced by the ocmmit 5b0cb1d850.
So far, the driver returns an error at probe.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The HDA_SUBDEV_NID_FLAG is duplicate for amplifier control elements. Move
get_amp_nid_() call to the snd_hda_ctl_add() function.
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
The purpose of this changeset is to show information about amplifier
setting in the codec proc file. Something like:
Control: name="Front Playback Volume", index=0, device=0
ControlAmp: chs=3, dir=Out, idx=0, ofs=0
Control: name="Front Playback Switch", index=0, device=0
ControlAmp: chs=3, dir=In, idx=2, ofs=0
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
This set of changes add missing NID values to some static control
elemenents. Also, it handles all "Capture Source" or "Input Source"
controls.
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
This patch adds support for dynamically created controls to proc codec file
(Control: lines).
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This is an initial patch to show universal control<->NID assigment in
proc codec file. The change helps to debug codec related problems.
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Original implementation was keeping registered input device for SND_BEEP
and SND_TONE events all time. This patch changes this behaviour:
If digital PC Beep is turned off using universal control switch,
the input device is unregistered.
Explanation: The kd_mksound() send SND_BEEP and SND_TONE only to last
registered device acceping those events. It means that the HDA Intel
audio driver blocks also the internal PC Speaker device (pcspkr.c
driver) even if the HDA Beep is muted. The user can easy disable
all beeps using 'setterm -blength 0' or 'xset b off' command.
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
CONFIG_SND_HDA_POWER_SAVE is independent from CONFIG_SND_HDA_HWDEP.
Thus snd_hda_hwdep_add_power_sysfs() needs the check of both kconfigs.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Added the power on/off counter and expose via sysfs files.
The sysfs files, power_on_acct and power_off_acct, are created under
each codec hwdep sysfs directory (e.g. /sys/class/sound/hwC0D0).
The files show the msec length of the codec power-on and power-off,
respectively.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The Intel IbexPeak HDMI codec supports 2 converters and 3 pins,
which requires converting the cvt_nid/pin_nid to arrays.
The active pin number (the one connected with a live HDMI monitor/sink)
will be dynamically identified on hotplug events.
It exports two HDMI devices, so that user space can choose the A/V pipe
for sending the audio samples.
It's still undefined behavior when there are two active monitors
connected and routed to the same audio converter.
Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The codec setup call via snd_hda_codec_configure() isn't necessarily
called in snd_hda_codec_new(). For the later added feature, it's better
to change the code flow like:
- create all codec instances
- configure each codec
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Added snd_hda_query_pin_caps() to read and cache pin-cap values
to avoid too frequently issuing the same verbs.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Added snd_hda_get_hint() and snd_hda_get_bool_hint() helper functions
to retrieve a hint value.
Internally, the hint is stored in a pair of two strings, key and val.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Fix this sparse warning:
sound/pci/hda/hda_codec.c:1544:19: warning: incorrect type in assignment (different signedness)
sound/pci/hda/hda_codec.c:1544:19: expected unsigned long *vals
sound/pci/hda/hda_codec.c:1544:19: got long *<noident>
Signed-off-by: Hannes Eder <hannes@hanneseder.net>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Added the pseudo device-locking using card->shutdown flag to avoid
the crash via clear/reconfig during operations.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Added the helper function snd_hda_multi_out_dig_cleanup() to clean up
the digital outputs with multi setup. This call is needed in cases
the codec supports multiple digital outputs as slaves. Otherwise the
slave widgets aren't properly cleaned up.
For a single digital output (e.g. in patch_conexant.c), this call isn't
needed.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Detect multiple digital-out pins in snd_hda_parse_pin_defconfig().
The dig_out_pin and dig_out_type fields become arrays.
The codec parser still doesn't use this multiple pins detection, though.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Add the SPDIF pin as slave digital out to enable concurrent
HDMI/SPDIF outputs for ASUS M3A-H/HDMI with ALC1200 codec.
Tested-by: Thomas Schneider <nailstudio@gmx.net>
Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Add dig_out_type and dig_in_type fields to autocfg struct.
A proper HDA_PCM_TYPE_* value is assigned to these fields according
to the pin-jack location type value.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Added the volume offset to base for the standard volume controls
to handle elements with too big volume scales like -96dB..0dB.
For such elements, you can set the base volume to reduce the range.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Gateway notebooks have their ID inside codec vendor ID, not at PCI ID. Due to
that, model auto-detection were not possible with the standard seek method.
This is what is found at lspci -vnn:
00:14.2 Audio device [0403]: ATI Technologies Inc SB450 HDA Audio [1002:437b] (rev 01)
Subsystem: ATI Technologies Inc SB450 HDA Audio [1002:437b]
Yet, autodetection is possible, since the codec properly reflects the vendor at
the Subsystem ID:
$ cat /proc/asound/card0/codec#0 |head -4
Codec: SigmaTel STAC9250
Address: 0
Vendor Id: 0x83847634
Subsystem Id: 0x107b0367
This patch adds a new autodetection function that seeks for codec subsystem ID.
Signed-off-by: Mauro Carvalho Chehab <mchehab@infradead.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Make the codec re-configuration feature selectable via Kconfig,
CONFIG_SND_HDA_RECONFIG.
Also mark it as experimental (as it really is).
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Introduce a global function snd_print_pcm_bits() and use it in the ELD code.
Signed-off-by: Wu Fengguang <wfg@linux.intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
code refactor: make a global function snd_print_channel_allocation().
Signed-off-by: Wu Fengguang <wfg@linux.intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Create /proc/asound/card<card_no>/eld#<codec_no> to reflect the audio
configurations and capabilities of the attached HDMI sink.
Some notes:
- Shall we show an empty file if the ELD content is not valid?
Well it's not that simple. There could be partially populated ELD,
and there may be malformed ELD provided by buggy drivers/monitors.
So expose ELD as it is.
- The ELD retrieval routines rely on the Intel HDA interface,
others are/could be universal and independent ones.
- How do we name the proc file?
If there are going to be two HDMI pins per codec, then the current naming
scheme (eld#<codec no>) will fail. Luckily the user space dependencies should
be minimal, so it would be trivial to do the rename if that happens.
- The ELD proc file content is designed to be easy for scripts and human reading.
Its lines all have the pattern:
<item_name>\t[\t]*<item_value>
where <item_name> is a keyword in c language, while <item_value> could be any
contents, including white spaces. <item_value> could also be a null value.
Signed-off-by: Wu Fengguang <wfg@linux.intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
ELD handling routines can be shared by all HDMI codecs,
and they are large enough to make a standalone source file.
Signed-off-by: Wu Fengguang <wfg@linux.intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
We want to share some code with print_pcm_rates(),
so extract a common routine snd_print_pcm_rates() from it.
Signed-off-by: Wu Fengguang <wfg@linux.intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Added the sysfs entries to hwdep devices so that the new features
like reconfiguration can be done via sysfs.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Added the reconfiguration feature of any individual codec.
Via the reconfiguration, the old resources are released and
the patch is called again to recreate the PCM and mixers in
addition to the re-initialization.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Manage all kcontrol elements created in the hda-intel driver.
This makes it possible to remove and reconfigure the controls
of each codec.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Explain some of the magic numbers I saw while trying to fix the AD1989 SPDIF
issue. Maybe should just use the expanded form directly in the verbs?
Signed-off-by: Robin H. Johnson <robbat2@gentoo.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Added support for controlling hardware gain amps on input ports
using a volume control mixer with a mux selecting the port being controlled.
Signed-off-by: Matthew Ranostay <mranostay@embeddedalley.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Move the codes for virtual master controls to sound core part so that
not only hda-intel drivers can use it.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Added a new mixer switch to enable/disable the sharing of the default
PCM stream with analog and SPDIF outputs. When "IEC958 Default PCM"
switch is on, the PCM stream is routed both to analog and SPDIF outputs.
This is the behavior in the earlier version.
Turning this switch off has a merit for some codecs, though. Some codec
chips don't support 24bit formats for SPDIF but only for analog outputs.
In this case, you can use 24bit format by disabling this switch.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Allows for dynamically creating mono out mixer controls and well
as mono mux controls.
Signed-off-by: Matthew Ranostay <mranostay@embeddedalley.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Added a mono_out_pin field to autocfg struct, and code to parse
for the mono_out_line.
Signed-off-by: Matthew Ranostay <mranostay@embeddedalley.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Don't query a widget parameter for an invalid NID in get_wcaps() but
rather returns zero (i.e. no attribute).
The read to an non-existing widget may result in a fatal codec
communication error.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Add master controls using vmaster to codecs that have no real hardware
master volume registers.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Added helper functions to implement virtual master volume controls.
The virtual master control is a control element that has multiple
slave controls. The value of master element is equally added to
slave elements.
The functions are written for general purpose, but it's put in the
HD-audio directory as now, since HD-audio driver is the only user.
It should be moved to the common place once after other drivers use
vmaster.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
AD1986A has a hardware problem that it cannot share a stream with
multiple pins properly. The problem occurs e.g. when a volume is changed
during playback.
So far, hda-intel driver unconditionally assigns the stream to multiple
output pins in copy-front mode, and this should be avoided for AD1986A
codec.
The original fix patch was by zhejiang <zhe.jiang@intel.com>.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
dac_nids arrays in each codec support code may have up to 5 items
when assigned from the auto-configurator. Some codec codes have
less numbers than the possible max. This patch defines the constant
and fixes the array definitions.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Fixed the build error from patch_sigmatel.c when built without
CONFIG_SND_HDA_GENERIC by defining a dummy function to return error.
Also, clean up hda_codec.c by removing unneeded ifdefs (the compiler
will optimize out).
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Added CONFIG_SND_HDA_POWER_SAVE kconfig. It's an experimental option
to achieve an aggressive power-saving. With this option, the driver
will turn on/off the power of each codec and controller chip dynamically
on demand.
The patch introduces a new module option 'power_save'. It specifies
the second of time-out for automatic power-down. As default, it's
10 seconds. Setting 0 means to suppress the power-saving feature.
The codec may have analog-input loopbacks, which are usually represented
by mixer elements such as 'Mic Playback Switch' or 'CD Playback Switch'.
When these are on, we cannot turn off the mixer and the codec chip has
to be kept on. For bookkeeping these states, a new codec-callback is
introduced.
For the bus-controller side, a new callback pm_notify is introduced,
which can be used to turn on/off the contoller appropriately.
Note that this power-saving might cause slight click-noise at
power-on/off. Also, it might take some time to wake up the codec, and
might even drop some tones at the very beginning. This seems to be the
side-effect of turning off the controller chip.
This turn-off of the controller can be disabled by undefining
HDA_POWER_SAVE_RESET_CONTOLLER in hda_intel.c.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
Added snd_hda_codec_amp_stereo() function that changes both of stereo
channels with the same mask and value bits. It simplifies most of
amp-handling codes.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
So far, the driver looked the table of snd_kcontrol_new used for creating
mixer elements and forces to call each of its put callbacks in PM resume
code. This is too ugly and hackish.
Now, the resume is simplified using the codec amp and command register
caches. The driver simply restores the values that have been written
in the cache table. With this simplification, most codec support codes
don't require any special resume callback.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
This patch adds the cache for codec command registers.
snd_hda_codec_write_cache() and snd_hda_sequence_write_cache() do
the write operations with caching, which values can be resumed via
snd_hda_codec_resume_cache().
The patch introduces only the framework, and no codec code is using
this cache yet. It'll be implemented in the following patch.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
Added callbacks for a generic bind-control of mixer elements.
This can be used for creating a mixer element controlling multiple
widgets at the same time. Two macros, HDA_BIND_VOL() and HDA_BIND_SW(),
are introduced for creating bind-volume and bind-switch, respectively.
It taks the mixer element name and struct hda_bind_ctls pointer, which
contains the real control callbacks in ops field and long array for
private_value of each bound widget.
All widgets have to be the same type (i.e. the same amp capability).
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
Added a hwdep interface for each codec (enabled per kconfig).
This interface can be used for reading/writing HD-audio verbs
and other purposes as future extensions.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
Fix codes to follow more to the standard kernel coding style.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
STAC922x provides the capture boost level up to 4, but actually it
works only up to 2. Since the range of the mixer is automatically
defined from amp-capability bits, we need to override the value
beforehand. snd_hda_override_amp_caps() is introduced for this
purpose.
The function patch_stac922x() calls this for NID 0x12 (Mux Capture
Volume). This should fix another recording problem on Intel Macs.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
Added line_out_type field to auto_pin_cfg struct to provide the
pin type of default line_outs.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
Fix SPDIF output (at least on Realtek codecs). The DIGI_CONVERT verbs
have to be reset before the PCM stream is set up. Otherwise the digital
setup is screwed up.
Also, check the AMP capability before setting AMP of the digital out
widget.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
Use global workqueue for simplicity.
The unsolicited event frequency isn't so high to have own queue.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
Use snd_pci_quirk_lookup() for looking up a board config table.
The config table is sorted in numerical order of PCI SSIDs.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
Some machines have multiple headpohne pins (usually on the lpatop
and on the docking station) while the current hda-codec driver
assumes a single headphone pin. Now it supports multiple hp pins
(at least for detection).
The sigmatel 92xx code supports this new multiple hp pins.
It detects all hp pins for auto-muting, too.
Also, the driver checks speaker pins in addition. In some cases,
all line-out, speaker and hp-pins coexist.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
Fixed compile errors with older gcc for initialization of a union.
sound/pci/ca0106/ca0106_mixer.c: At top level:
sound/pci/ca0106/ca0106_mixer.c:499: unknown field 'p' specified in initializer
sound/pci/ca0106/ca0106_mixer.c:499: warning: missing braces around initializer
sound/pci/ca0106/ca0106_mixer.c:499: warning: (near initialization for 'snd_ca0106_volume_ctls[0].tlv')
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
Modules: HDA Codec driver,HDA generic driver
- Fix autoconfig speaker/hp detection
Now it allows multiple speaker pins (e.g. Dell laptops have such config)
- Use speaker or hp pins if no line-outs are available
This fixes the silence output on recent Dell laptops with STAC9200
(ALSA bug#1843)
- Fix analog/realtek/sigmatel autoconfig parser
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Modules: HDA Codec driver,HDA generic driver
Add 'fujitsu' model for ALC262 patch to support a FSC laptop.
The internal speaker is turned on/off with jack sensing.
Also fixed alc262 'basic' model.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Modules: HDA Codec driver,HDA generic driver
- Common labels for input pins
- Fix and clean up of Realtek codec parsers
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Modules: HDA Codec driver,HDA generic driver
This patch adds the support of ALC262,ALC883,ALC885,ALC861 to driver
More models and improvements for ALC880, ALC260 and ALC882 codecs, too.
Signed-off-by: Kailang Yang <kailang@realtek.com.tw>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Modules: HDA Codec driver,HDA generic driver
- Fix connection list parsing (with ranged flag).
- Increase the max number of connections
- Introduce widget capabilities cache
- Power up/down widgets at init, suspend and resume
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Modules: HDA Codec driver,HDA generic driver
- Make bound controls global to all patches
- Clean up analog patches (for the upcoming extension to AD1988)
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Documentation,HDA Codec driver,HDA generic driver,HDA Intel driver
- Fix some invalid configurations, typos in the last patch
- Make init_verbs chainable, so that different configs can share the same
init_verbs
- Reorder and clean up the source codes in patch_realtek.c
- Add the pin default configuration parser, used commonly in cmedia
and realtek patch codes.
- Add 'auto' model to ALC880 for auto-configuration from BIOS
Use this model as default, and 3-stack as fallback
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Initial git repository build. I'm not bothering with the full history,
even though we have it. We can create a separate "historical" git
archive of that later if we want to, and in the meantime it's about
3.2GB when imported into git - space that would just make the early
git days unnecessarily complicated, when we don't have a lot of good
infrastructure for it.
Let it rip!
Takashi Iwai