1031 lines
27 KiB
C
Raw Normal View History

/*-
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
* Copyright (c) 2001 Cameron Grant <cg@FreeBSD.org>
* Copyright (c) 2006 Ariff Abdullah <ariff@FreeBSD.org>
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
/* Almost entirely rewritten to add multi-format/channels mixing support. */
#include <dev/sound/pcm/sound.h>
#include <dev/sound/pcm/vchan.h>
#include "feeder_if.h"
SND_DECLARE_FILE("$FreeBSD$");
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
MALLOC_DEFINE(M_VCHANFEEDER, "vchanfeed", "pcm vchan feeder");
typedef uint32_t (*feed_vchan_mixer)(uint8_t *, uint8_t *, uint32_t);
Whats New: 1. Support wide range sampling rate, as low as 1hz up to int32 max (which is, insane) through new feeder_rate, multiple precisions choice (32/64 bit converter). This is indeed, quite insane, but it does give us more room and flexibility. Plenty sysctl options to adjust resampling characteristics. 2. Support 24/32 bit pcm format conversion through new, much improved, simplified and optimized feeder_fmt. Changes: 1. buffer.c / dsp.c / sound.h * Support for 24/32 AFMT. 2. feeder_rate.c * New implementation of sampling rate conversion with 32/64 bit precision, 1 - int32max hz (which is, ridiculous, yet very addictive). Much improved / smarter buffer management to not cause any missing samples at the end of conversion process * Tunable sysctls for various aspect: hw.snd.feeder_rate_ratemin - minimum allowable sampling rate (default to 4000) hw.snd.feeder_rate_ratemax - maximum allowable sampling rate (default to 1102500) hw.snd.feeder_rate_buffersize - conversion buffer size (default to 8192) hw.snd.feeder_rate_scaling - scaling / conversion method (please refer to the source for explaination). Default to previous implementation type. 3. feeder_fmt.c / sound.h * New implementation, support for 24/32bit conversion, optimized, and simplified. Few routines has been removed (8 to xlaw, 16 to 8). It just doesn't make sense. 4. channel.c * Support for 24/32 AFMT * Fix wrong xruns increment, causing incorrect underruns statistic while using vchans. 5. vchan.c * Support for 24/32 AFMT * Proper speed / rate detection especially for fixed rate ac97. User can override it using kernel hint: hint.pcm.<unit>.vchanrate="xxxx". Notes / Issues: * Virtual Channels (vchans) Enabling vchans can really, really help to solve overrun issues. This is quite understandable, because it operates entirely within its own buffering system without relying on hardware interrupt / state. Even if you don't need vchan, just enable single channel can help much. Few soundcards (notably via8233x, sblive, possibly others) have their own hardware multi channel, and this is unfortunately beyond vchan reachability. * The arrival of 24/32 also come with a price. Applications that can do 24/32bit playback need to be recompiled (notably mplayer). Use (recompiled) mplayer to experiment / test / debug this various format using -af format=fmt. Note that 24bit seeking in mplayer is a little bit broken, sometimes can cause silence or loud static noise. Pausing / seeking few times can solve this problem. You don't have to rebuild world entirely for this. Simply copy /usr/src/sys/sys/soundcard.h to /usr/include/sys/soundcard.h would suffice. Few drivers also need recompilation, and this can be done via /usr/src/sys/modules/sound/. Support for 24bit hardware playback is beyond the scope of this changes. That would require spessific hardware driver changes. * Don't expect playing 9999999999hz is a wise decision. Be reasonable. The new feeder_rate implemention provide flexibility, not insanity. You can easily chew up your CPU with this kind of mind instability. Please use proper mosquito repellent device for this obvious cracked brain attempt. As for testing purposes, you can use (again) mplayer to generate / play with different sampling rate. Use something like "mplayer -af resample=192000:0:0 <files>". Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my> Tested by: multimedia@
2005-07-31 16:16:22 +00:00
struct vchinfo {
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
struct pcm_channel *channel;
struct pcmchan_caps caps;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
uint32_t fmtlist[2];
int trigger;
};
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
/* support everything (mono / stereo), except a-law / mu-law */
static struct afmtstr_table vchan_supported_fmts[] = {
{ "u8", AFMT_U8 }, { "s8", AFMT_S8 },
{ "s16le", AFMT_S16_LE }, { "s16be", AFMT_S16_BE },
{ "u16le", AFMT_U16_LE }, { "u16be", AFMT_U16_BE },
{ "s24le", AFMT_S24_LE }, { "s24be", AFMT_S24_BE },
{ "u24le", AFMT_U24_LE }, { "u24be", AFMT_U24_BE },
{ "s32le", AFMT_S32_LE }, { "s32be", AFMT_S32_BE },
{ "u32le", AFMT_U32_LE }, { "u32be", AFMT_U32_BE },
{ NULL, 0 },
};
/* alias table, shorter. */
static const struct {
char *alias, *fmtstr;
} vchan_fmtstralias[] = {
{ "8", "u8" }, { "16", "s16le" },
{ "24", "s24le" }, { "32", "s32le" },
{ NULL, NULL },
};
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
#define vchan_valid_format(fmt) \
afmt2afmtstr(vchan_supported_fmts, fmt, NULL, 0, 0, \
AFMTSTR_STEREO_RETURN)
#define vchan_valid_strformat(strfmt) \
afmtstr2afmt(vchan_supported_fmts, strfmt, AFMTSTR_STEREO_RETURN);
/*
* Need specialized WRITE macros since 32bit might involved saturation
* if calculation is done within 32bit arithmetic.
*/
#define VCHAN_PCM_WRITE_S8_NE(b8, val) PCM_WRITE_S8(b8, val)
#define VCHAN_PCM_WRITE_S16_LE(b8, val) PCM_WRITE_S16_LE(b8, val)
#define VCHAN_PCM_WRITE_S24_LE(b8, val) PCM_WRITE_S24_LE(b8, val)
#define VCHAN_PCM_WRITE_S32_LE(b8, val) _PCM_WRITE_S32_LE(b8, val)
#define VCHAN_PCM_WRITE_S16_BE(b8, val) PCM_WRITE_S16_BE(b8, val)
#define VCHAN_PCM_WRITE_S24_BE(b8, val) PCM_WRITE_S24_BE(b8, val)
#define VCHAN_PCM_WRITE_S32_BE(b8, val) _PCM_WRITE_S32_BE(b8, val)
#define VCHAN_PCM_WRITE_U8_NE(b8, val) PCM_WRITE_U8(b8, val)
#define VCHAN_PCM_WRITE_U16_LE(b8, val) PCM_WRITE_U16_LE(b8, val)
#define VCHAN_PCM_WRITE_U24_LE(b8, val) PCM_WRITE_U24_LE(b8, val)
#define VCHAN_PCM_WRITE_U32_LE(b8, val) _PCM_WRITE_U32_LE(b8, val)
#define VCHAN_PCM_WRITE_U16_BE(b8, val) PCM_WRITE_U16_BE(b8, val)
#define VCHAN_PCM_WRITE_U24_BE(b8, val) PCM_WRITE_U24_BE(b8, val)
#define VCHAN_PCM_WRITE_U32_BE(b8, val) _PCM_WRITE_U32_BE(b8, val)
#define FEEDER_VCHAN_MIX(FMTBIT, VCHAN_INTCAST, SIGN, SIGNS, ENDIAN, ENDIANS) \
static uint32_t \
feed_vchan_mix_##SIGNS##FMTBIT##ENDIANS(uint8_t *to, uint8_t *tmp, \
uint32_t count) \
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
{ \
int32_t x, y; \
VCHAN_INTCAST z; \
int i; \
\
i = count; \
tmp += i; \
to += i; \
\
do { \
tmp -= PCM_##FMTBIT##_BPS; \
to -= PCM_##FMTBIT##_BPS; \
i -= PCM_##FMTBIT##_BPS; \
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
x = PCM_READ_##SIGN##FMTBIT##_##ENDIAN(tmp); \
y = PCM_READ_##SIGN##FMTBIT##_##ENDIAN(to); \
z = (VCHAN_INTCAST)x + y; \
x = PCM_CLAMP_##SIGN##FMTBIT(z); \
VCHAN_PCM_WRITE_##SIGN##FMTBIT##_##ENDIAN(to, x); \
} while (i != 0); \
\
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (count); \
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
}
FEEDER_VCHAN_MIX(8, int32_t, S, s, NE, ne)
FEEDER_VCHAN_MIX(16, int32_t, S, s, LE, le)
FEEDER_VCHAN_MIX(24, int32_t, S, s, LE, le)
FEEDER_VCHAN_MIX(32, intpcm_t, S, s, LE, le)
FEEDER_VCHAN_MIX(16, int32_t, S, s, BE, be)
FEEDER_VCHAN_MIX(24, int32_t, S, s, BE, be)
FEEDER_VCHAN_MIX(32, intpcm_t, S, s, BE, be)
FEEDER_VCHAN_MIX(8, int32_t, U, u, NE, ne)
FEEDER_VCHAN_MIX(16, int32_t, U, u, LE, le)
FEEDER_VCHAN_MIX(24, int32_t, U, u, LE, le)
FEEDER_VCHAN_MIX(32, intpcm_t, U, u, LE, le)
FEEDER_VCHAN_MIX(16, int32_t, U, u, BE, be)
FEEDER_VCHAN_MIX(24, int32_t, U, u, BE, be)
FEEDER_VCHAN_MIX(32, intpcm_t, U, u, BE, be)
struct feed_vchan_info {
uint32_t format;
int bps;
feed_vchan_mixer mix;
};
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
static struct feed_vchan_info feed_vchan_info_tbl[] = {
{ AFMT_S8, PCM_8_BPS, feed_vchan_mix_s8ne },
{ AFMT_S16_LE, PCM_16_BPS, feed_vchan_mix_s16le },
{ AFMT_S24_LE, PCM_24_BPS, feed_vchan_mix_s24le },
{ AFMT_S32_LE, PCM_32_BPS, feed_vchan_mix_s32le },
{ AFMT_S16_BE, PCM_16_BPS, feed_vchan_mix_s16be },
{ AFMT_S24_BE, PCM_24_BPS, feed_vchan_mix_s24be },
{ AFMT_S32_BE, PCM_32_BPS, feed_vchan_mix_s32be },
{ AFMT_U8, PCM_8_BPS, feed_vchan_mix_u8ne },
{ AFMT_U16_LE, PCM_16_BPS, feed_vchan_mix_u16le },
{ AFMT_U24_LE, PCM_24_BPS, feed_vchan_mix_u24le },
{ AFMT_U32_LE, PCM_32_BPS, feed_vchan_mix_u32le },
{ AFMT_U16_BE, PCM_16_BPS, feed_vchan_mix_u16be },
{ AFMT_U24_BE, PCM_24_BPS, feed_vchan_mix_u24be },
{ AFMT_U32_BE, PCM_32_BPS, feed_vchan_mix_u32be },
};
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
#define FVCHAN_DATA(i, c) ((intptr_t)((((i) & 0x1f) << 4) | ((c) & 0xf)))
#define FVCHAN_INFOIDX(m) (((m) >> 4) & 0x1f)
#define FVCHAN_CHANNELS(m) ((m) & 0xf)
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
static int
feed_vchan_init(struct pcm_feeder *f)
{
int i, channels;
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
if (f->desc->out != f->desc->in)
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (EINVAL);
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
channels = (f->desc->out & AFMT_STEREO) ? 2 : 1;
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
for (i = 0; i < sizeof(feed_vchan_info_tbl) /
sizeof(feed_vchan_info_tbl[0]); i++) {
if ((f->desc->out & ~AFMT_STEREO) ==
feed_vchan_info_tbl[i].format) {
f->data = (void *)FVCHAN_DATA(i, channels);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (0);
}
}
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (-1);
}
static __inline int
feed_vchan_rec(struct pcm_channel *c)
{
struct pcm_channel *ch;
struct snd_dbuf *b, *bs;
int cnt, rdy;
/*
* Reset ready and moving pointer. We're not using bufsoft
* anywhere since its sole purpose is to become the primary
* distributor for the recorded buffer and also as an interrupt
* threshold progress indicator.
*/
b = c->bufsoft;
b->rp = 0;
b->rl = 0;
cnt = sndbuf_getsize(b);
do {
cnt = FEEDER_FEED(c->feeder->source, c, b->tmpbuf, cnt,
c->bufhard);
if (cnt != 0) {
sndbuf_acquire(b, b->tmpbuf, cnt);
cnt = sndbuf_getfree(b);
}
} while (cnt != 0);
/* Not enough data */
if (b->rl < sndbuf_getbps(b)) {
b->rl = 0;
return (0);
}
/*
* Keep track of ready and moving pointer since we will use
* bufsoft over and over again, pretending nothing has happened.
*/
rdy = b->rl;
CHN_FOREACH(ch, c, children.busy) {
CHN_LOCK(ch);
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
if (!(ch->flags & CHN_F_TRIGGERED)) {
CHN_UNLOCK(ch);
continue;
}
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
bs = ch->bufsoft;
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
if (ch->flags & CHN_F_MAPPED)
sndbuf_dispose(bs, NULL, sndbuf_getready(bs));
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
cnt = sndbuf_getfree(bs);
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
if (cnt < sndbuf_getbps(bs)) {
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
CHN_UNLOCK(ch);
continue;
}
do {
cnt = FEEDER_FEED(ch->feeder, ch, bs->tmpbuf, cnt, b);
if (cnt != 0) {
sndbuf_acquire(bs, bs->tmpbuf, cnt);
cnt = sndbuf_getfree(bs);
}
} while (cnt != 0);
/*
* Not entirely flushed out...
*/
if (b->rl != 0)
ch->xruns++;
CHN_UNLOCK(ch);
/*
* Rewind buffer position for next virtual channel.
*/
b->rp = 0;
b->rl = rdy;
}
/*
* Set ready pointer to indicate that our children are ready
* to be woken up, also as an interrupt threshold progress
* indicator.
*/
b->rl = 1;
/*
* Return 0 to bail out early from sndbuf_feed() loop.
* No need to increase feedcount counter since part of this
* feeder chains already include feed_root().
*/
return (0);
}
static int
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
feed_vchan(struct pcm_feeder *f, struct pcm_channel *c, uint8_t *b,
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
uint32_t count, void *source)
{
struct feed_vchan_info *info;
struct snd_dbuf *src = source;
struct pcm_channel *ch;
uint32_t cnt, mcnt, rcnt, sz;
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
uint8_t *tmp;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
if (c->direction == PCMDIR_REC)
return (feed_vchan_rec(c));
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
sz = sndbuf_getsize(src);
if (sz < count)
count = sz;
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
info = &feed_vchan_info_tbl[FVCHAN_INFOIDX((intptr_t)f->data)];
sz = info->bps * FVCHAN_CHANNELS((intptr_t)f->data);
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
count -= count % sz;
if (count < sz)
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (0);
/*
* we are going to use our source as a temporary buffer since it's
* got no other purpose. we obtain our data by traversing the channel
* list of children and calling vchan_mix_* to mix count bytes from
* each into our destination buffer, b
*/
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
tmp = sndbuf_getbuf(src);
rcnt = 0;
mcnt = 0;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
CHN_FOREACH(ch, c, children.busy) {
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
CHN_LOCK(ch);
if (!(ch->flags & CHN_F_TRIGGERED)) {
CHN_UNLOCK(ch);
continue;
}
if ((ch->flags & CHN_F_MAPPED) && !(ch->flags & CHN_F_CLOSING))
sndbuf_acquire(ch->bufsoft, NULL,
sndbuf_getfree(ch->bufsoft));
if (rcnt == 0) {
rcnt = FEEDER_FEED(ch->feeder, ch, b, count,
ch->bufsoft);
rcnt -= rcnt % sz;
mcnt = count - rcnt;
} else {
cnt = FEEDER_FEED(ch->feeder, ch, tmp, count,
ch->bufsoft);
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
cnt -= cnt % sz;
if (cnt != 0) {
if (mcnt != 0) {
memset(b + rcnt,
sndbuf_zerodata(f->desc->out),
mcnt);
mcnt = 0;
}
cnt = info->mix(b, tmp, cnt);
if (cnt > rcnt)
rcnt = cnt;
}
}
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
CHN_UNLOCK(ch);
}
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
if (++c->feedcount == 0)
c->feedcount = 2;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (rcnt);
}
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
static struct pcm_feederdesc feeder_vchan_desc[] = {
{FEEDER_MIXER, AFMT_S8, AFMT_S8, 0},
{FEEDER_MIXER, AFMT_S16_LE, AFMT_S16_LE, 0},
{FEEDER_MIXER, AFMT_S24_LE, AFMT_S24_LE, 0},
{FEEDER_MIXER, AFMT_S32_LE, AFMT_S32_LE, 0},
{FEEDER_MIXER, AFMT_S16_BE, AFMT_S16_BE, 0},
{FEEDER_MIXER, AFMT_S24_BE, AFMT_S24_BE, 0},
{FEEDER_MIXER, AFMT_S32_BE, AFMT_S32_BE, 0},
{FEEDER_MIXER, AFMT_S8 | AFMT_STEREO, AFMT_S8 | AFMT_STEREO, 0},
{FEEDER_MIXER, AFMT_S16_LE | AFMT_STEREO, AFMT_S16_LE | AFMT_STEREO, 0},
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
{FEEDER_MIXER, AFMT_S24_LE | AFMT_STEREO, AFMT_S24_LE | AFMT_STEREO, 0},
{FEEDER_MIXER, AFMT_S32_LE | AFMT_STEREO, AFMT_S32_LE | AFMT_STEREO, 0},
{FEEDER_MIXER, AFMT_S16_BE | AFMT_STEREO, AFMT_S16_BE | AFMT_STEREO, 0},
{FEEDER_MIXER, AFMT_S24_BE | AFMT_STEREO, AFMT_S24_BE | AFMT_STEREO, 0},
{FEEDER_MIXER, AFMT_S32_BE | AFMT_STEREO, AFMT_S32_BE | AFMT_STEREO, 0},
{FEEDER_MIXER, AFMT_U8, AFMT_U8, 0},
{FEEDER_MIXER, AFMT_U16_LE, AFMT_U16_LE, 0},
{FEEDER_MIXER, AFMT_U24_LE, AFMT_U24_LE, 0},
{FEEDER_MIXER, AFMT_U32_LE, AFMT_U32_LE, 0},
{FEEDER_MIXER, AFMT_U16_BE, AFMT_U16_BE, 0},
{FEEDER_MIXER, AFMT_U24_BE, AFMT_U24_BE, 0},
{FEEDER_MIXER, AFMT_U32_BE, AFMT_U32_BE, 0},
{FEEDER_MIXER, AFMT_U8 | AFMT_STEREO, AFMT_U8 | AFMT_STEREO, 0},
{FEEDER_MIXER, AFMT_U16_LE | AFMT_STEREO, AFMT_U16_LE | AFMT_STEREO, 0},
{FEEDER_MIXER, AFMT_U24_LE | AFMT_STEREO, AFMT_U24_LE | AFMT_STEREO, 0},
{FEEDER_MIXER, AFMT_U32_LE | AFMT_STEREO, AFMT_U32_LE | AFMT_STEREO, 0},
{FEEDER_MIXER, AFMT_U16_BE | AFMT_STEREO, AFMT_U16_BE | AFMT_STEREO, 0},
{FEEDER_MIXER, AFMT_U24_BE | AFMT_STEREO, AFMT_U24_BE | AFMT_STEREO, 0},
{FEEDER_MIXER, AFMT_U32_BE | AFMT_STEREO, AFMT_U32_BE | AFMT_STEREO, 0},
{0, 0, 0, 0},
};
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
static kobj_method_t feeder_vchan_methods[] = {
KOBJMETHOD(feeder_init, feed_vchan_init),
KOBJMETHOD(feeder_feed, feed_vchan),
{0, 0}
};
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
FEEDER_DECLARE(feeder_vchan, 2, NULL);
/************************************************************/
static void *
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
vchan_init(kobj_t obj, void *devinfo, struct snd_dbuf *b,
struct pcm_channel *c, int dir)
{
struct vchinfo *ch;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
KASSERT(dir == PCMDIR_PLAY || dir == PCMDIR_REC,
("vchan_init: bad direction"));
KASSERT(c != NULL && c->parentchannel != NULL,
("vchan_init: bad channels"));
ch = malloc(sizeof(*ch), M_DEVBUF, M_WAITOK | M_ZERO);
ch->channel = c;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
ch->trigger = PCMTRIG_STOP;
c->flags |= CHN_F_VIRTUAL;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (ch);
}
static int
vchan_free(kobj_t obj, void *data)
{
free(data, M_DEVBUF);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (0);
}
static int
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
vchan_setformat(kobj_t obj, void *data, uint32_t format)
{
struct vchinfo *ch = data;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
if (fmtvalid(format, ch->fmtlist) == 0)
return (-1);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (0);
}
static int
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
vchan_setspeed(kobj_t obj, void *data, uint32_t speed)
{
struct vchinfo *ch = data;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
struct pcm_channel *p = ch->channel->parentchannel;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (sndbuf_getspd(p->bufsoft));
}
static int
vchan_trigger(kobj_t obj, void *data, int go)
{
struct vchinfo *ch = data;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
struct pcm_channel *c, *p;
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
int err, otrigger;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
if (!PCMTRIG_COMMON(go) || go == ch->trigger)
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (0);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
c = ch->channel;
p = c->parentchannel;
otrigger = ch->trigger;
ch->trigger = go;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
CHN_UNLOCK(c);
CHN_LOCK(p);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
switch (go) {
case PCMTRIG_START:
if (otrigger != PCMTRIG_START) {
CHN_INSERT_HEAD(p, c, children.busy);
}
break;
case PCMTRIG_STOP:
case PCMTRIG_ABORT:
if (otrigger == PCMTRIG_START) {
CHN_REMOVE(p, c, children.busy);
}
break;
default:
break;
}
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
err = chn_notify(p, CHN_N_TRIGGER);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
CHN_UNLOCK(p);
CHN_LOCK(c);
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
return (err);
}
static struct pcmchan_caps *
vchan_getcaps(kobj_t obj, void *data)
{
struct vchinfo *ch = data;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
struct pcm_channel *c, *p;
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
uint32_t fmt;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
c = ch->channel;
p = c->parentchannel;
ch->caps.minspeed = sndbuf_getspd(p->bufsoft);
ch->caps.maxspeed = ch->caps.minspeed;
ch->caps.caps = 0;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
ch->fmtlist[1] = 0;
fmt = sndbuf_getfmt(p->bufsoft);
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
if (fmt != vchan_valid_format(fmt)) {
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
device_printf(c->dev,
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
"%s: WARNING: invalid vchan format! (0x%08x)\n",
__func__, fmt);
fmt = VCHAN_DEFAULT_AFMT;
}
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
ch->fmtlist[0] = fmt;
ch->caps.fmtlist = ch->fmtlist;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (&ch->caps);
}
static kobj_method_t vchan_methods[] = {
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
KOBJMETHOD(channel_init, vchan_init),
KOBJMETHOD(channel_free, vchan_free),
KOBJMETHOD(channel_setformat, vchan_setformat),
KOBJMETHOD(channel_setspeed, vchan_setspeed),
KOBJMETHOD(channel_trigger, vchan_trigger),
KOBJMETHOD(channel_getcaps, vchan_getcaps),
{0, 0}
};
CHANNEL_DECLARE(vchan);
Whats New: 1. Support wide range sampling rate, as low as 1hz up to int32 max (which is, insane) through new feeder_rate, multiple precisions choice (32/64 bit converter). This is indeed, quite insane, but it does give us more room and flexibility. Plenty sysctl options to adjust resampling characteristics. 2. Support 24/32 bit pcm format conversion through new, much improved, simplified and optimized feeder_fmt. Changes: 1. buffer.c / dsp.c / sound.h * Support for 24/32 AFMT. 2. feeder_rate.c * New implementation of sampling rate conversion with 32/64 bit precision, 1 - int32max hz (which is, ridiculous, yet very addictive). Much improved / smarter buffer management to not cause any missing samples at the end of conversion process * Tunable sysctls for various aspect: hw.snd.feeder_rate_ratemin - minimum allowable sampling rate (default to 4000) hw.snd.feeder_rate_ratemax - maximum allowable sampling rate (default to 1102500) hw.snd.feeder_rate_buffersize - conversion buffer size (default to 8192) hw.snd.feeder_rate_scaling - scaling / conversion method (please refer to the source for explaination). Default to previous implementation type. 3. feeder_fmt.c / sound.h * New implementation, support for 24/32bit conversion, optimized, and simplified. Few routines has been removed (8 to xlaw, 16 to 8). It just doesn't make sense. 4. channel.c * Support for 24/32 AFMT * Fix wrong xruns increment, causing incorrect underruns statistic while using vchans. 5. vchan.c * Support for 24/32 AFMT * Proper speed / rate detection especially for fixed rate ac97. User can override it using kernel hint: hint.pcm.<unit>.vchanrate="xxxx". Notes / Issues: * Virtual Channels (vchans) Enabling vchans can really, really help to solve overrun issues. This is quite understandable, because it operates entirely within its own buffering system without relying on hardware interrupt / state. Even if you don't need vchan, just enable single channel can help much. Few soundcards (notably via8233x, sblive, possibly others) have their own hardware multi channel, and this is unfortunately beyond vchan reachability. * The arrival of 24/32 also come with a price. Applications that can do 24/32bit playback need to be recompiled (notably mplayer). Use (recompiled) mplayer to experiment / test / debug this various format using -af format=fmt. Note that 24bit seeking in mplayer is a little bit broken, sometimes can cause silence or loud static noise. Pausing / seeking few times can solve this problem. You don't have to rebuild world entirely for this. Simply copy /usr/src/sys/sys/soundcard.h to /usr/include/sys/soundcard.h would suffice. Few drivers also need recompilation, and this can be done via /usr/src/sys/modules/sound/. Support for 24bit hardware playback is beyond the scope of this changes. That would require spessific hardware driver changes. * Don't expect playing 9999999999hz is a wise decision. Be reasonable. The new feeder_rate implemention provide flexibility, not insanity. You can easily chew up your CPU with this kind of mind instability. Please use proper mosquito repellent device for this obvious cracked brain attempt. As for testing purposes, you can use (again) mplayer to generate / play with different sampling rate. Use something like "mplayer -af resample=192000:0:0 <files>". Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my> Tested by: multimedia@
2005-07-31 16:16:22 +00:00
/*
* On the fly vchan rate settings
*/
#ifdef SND_DYNSYSCTL
static int
sysctl_hw_snd_vchanrate(SYSCTL_HANDLER_ARGS)
{
struct snddev_info *d;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
struct pcm_channel *c, *ch = NULL;
Whats New: 1. Support wide range sampling rate, as low as 1hz up to int32 max (which is, insane) through new feeder_rate, multiple precisions choice (32/64 bit converter). This is indeed, quite insane, but it does give us more room and flexibility. Plenty sysctl options to adjust resampling characteristics. 2. Support 24/32 bit pcm format conversion through new, much improved, simplified and optimized feeder_fmt. Changes: 1. buffer.c / dsp.c / sound.h * Support for 24/32 AFMT. 2. feeder_rate.c * New implementation of sampling rate conversion with 32/64 bit precision, 1 - int32max hz (which is, ridiculous, yet very addictive). Much improved / smarter buffer management to not cause any missing samples at the end of conversion process * Tunable sysctls for various aspect: hw.snd.feeder_rate_ratemin - minimum allowable sampling rate (default to 4000) hw.snd.feeder_rate_ratemax - maximum allowable sampling rate (default to 1102500) hw.snd.feeder_rate_buffersize - conversion buffer size (default to 8192) hw.snd.feeder_rate_scaling - scaling / conversion method (please refer to the source for explaination). Default to previous implementation type. 3. feeder_fmt.c / sound.h * New implementation, support for 24/32bit conversion, optimized, and simplified. Few routines has been removed (8 to xlaw, 16 to 8). It just doesn't make sense. 4. channel.c * Support for 24/32 AFMT * Fix wrong xruns increment, causing incorrect underruns statistic while using vchans. 5. vchan.c * Support for 24/32 AFMT * Proper speed / rate detection especially for fixed rate ac97. User can override it using kernel hint: hint.pcm.<unit>.vchanrate="xxxx". Notes / Issues: * Virtual Channels (vchans) Enabling vchans can really, really help to solve overrun issues. This is quite understandable, because it operates entirely within its own buffering system without relying on hardware interrupt / state. Even if you don't need vchan, just enable single channel can help much. Few soundcards (notably via8233x, sblive, possibly others) have their own hardware multi channel, and this is unfortunately beyond vchan reachability. * The arrival of 24/32 also come with a price. Applications that can do 24/32bit playback need to be recompiled (notably mplayer). Use (recompiled) mplayer to experiment / test / debug this various format using -af format=fmt. Note that 24bit seeking in mplayer is a little bit broken, sometimes can cause silence or loud static noise. Pausing / seeking few times can solve this problem. You don't have to rebuild world entirely for this. Simply copy /usr/src/sys/sys/soundcard.h to /usr/include/sys/soundcard.h would suffice. Few drivers also need recompilation, and this can be done via /usr/src/sys/modules/sound/. Support for 24bit hardware playback is beyond the scope of this changes. That would require spessific hardware driver changes. * Don't expect playing 9999999999hz is a wise decision. Be reasonable. The new feeder_rate implemention provide flexibility, not insanity. You can easily chew up your CPU with this kind of mind instability. Please use proper mosquito repellent device for this obvious cracked brain attempt. As for testing purposes, you can use (again) mplayer to generate / play with different sampling rate. Use something like "mplayer -af resample=192000:0:0 <files>". Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my> Tested by: multimedia@
2005-07-31 16:16:22 +00:00
struct pcmchan_caps *caps;
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
int *vchanrate, vchancount, direction, err, newspd;
Whats New: 1. Support wide range sampling rate, as low as 1hz up to int32 max (which is, insane) through new feeder_rate, multiple precisions choice (32/64 bit converter). This is indeed, quite insane, but it does give us more room and flexibility. Plenty sysctl options to adjust resampling characteristics. 2. Support 24/32 bit pcm format conversion through new, much improved, simplified and optimized feeder_fmt. Changes: 1. buffer.c / dsp.c / sound.h * Support for 24/32 AFMT. 2. feeder_rate.c * New implementation of sampling rate conversion with 32/64 bit precision, 1 - int32max hz (which is, ridiculous, yet very addictive). Much improved / smarter buffer management to not cause any missing samples at the end of conversion process * Tunable sysctls for various aspect: hw.snd.feeder_rate_ratemin - minimum allowable sampling rate (default to 4000) hw.snd.feeder_rate_ratemax - maximum allowable sampling rate (default to 1102500) hw.snd.feeder_rate_buffersize - conversion buffer size (default to 8192) hw.snd.feeder_rate_scaling - scaling / conversion method (please refer to the source for explaination). Default to previous implementation type. 3. feeder_fmt.c / sound.h * New implementation, support for 24/32bit conversion, optimized, and simplified. Few routines has been removed (8 to xlaw, 16 to 8). It just doesn't make sense. 4. channel.c * Support for 24/32 AFMT * Fix wrong xruns increment, causing incorrect underruns statistic while using vchans. 5. vchan.c * Support for 24/32 AFMT * Proper speed / rate detection especially for fixed rate ac97. User can override it using kernel hint: hint.pcm.<unit>.vchanrate="xxxx". Notes / Issues: * Virtual Channels (vchans) Enabling vchans can really, really help to solve overrun issues. This is quite understandable, because it operates entirely within its own buffering system without relying on hardware interrupt / state. Even if you don't need vchan, just enable single channel can help much. Few soundcards (notably via8233x, sblive, possibly others) have their own hardware multi channel, and this is unfortunately beyond vchan reachability. * The arrival of 24/32 also come with a price. Applications that can do 24/32bit playback need to be recompiled (notably mplayer). Use (recompiled) mplayer to experiment / test / debug this various format using -af format=fmt. Note that 24bit seeking in mplayer is a little bit broken, sometimes can cause silence or loud static noise. Pausing / seeking few times can solve this problem. You don't have to rebuild world entirely for this. Simply copy /usr/src/sys/sys/soundcard.h to /usr/include/sys/soundcard.h would suffice. Few drivers also need recompilation, and this can be done via /usr/src/sys/modules/sound/. Support for 24bit hardware playback is beyond the scope of this changes. That would require spessific hardware driver changes. * Don't expect playing 9999999999hz is a wise decision. Be reasonable. The new feeder_rate implemention provide flexibility, not insanity. You can easily chew up your CPU with this kind of mind instability. Please use proper mosquito repellent device for this obvious cracked brain attempt. As for testing purposes, you can use (again) mplayer to generate / play with different sampling rate. Use something like "mplayer -af resample=192000:0:0 <files>". Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my> Tested by: multimedia@
2005-07-31 16:16:22 +00:00
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
d = devclass_get_softc(pcm_devclass, VCHAN_SYSCTL_UNIT(oidp->oid_arg1));
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
if (!PCM_REGISTERED(d) || !(d->flags & SD_F_AUTOVCHAN))
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (EINVAL);
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
pcm_lock(d);
PCM_WAIT(d);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
switch (VCHAN_SYSCTL_DIR(oidp->oid_arg1)) {
case VCHAN_PLAY:
direction = PCMDIR_PLAY;
vchancount = d->pvchancount;
vchanrate = &d->pvchanrate;
break;
case VCHAN_REC:
direction = PCMDIR_REC;
vchancount = d->rvchancount;
vchanrate = &d->rvchanrate;
break;
default:
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
pcm_unlock(d);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (EINVAL);
break;
}
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
if (vchancount < 1) {
pcm_unlock(d);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (EINVAL);
}
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
PCM_ACQUIRE(d);
pcm_unlock(d);
newspd = 0;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
CHN_FOREACH(c, d, channels.pcm) {
Whats New: 1. Support wide range sampling rate, as low as 1hz up to int32 max (which is, insane) through new feeder_rate, multiple precisions choice (32/64 bit converter). This is indeed, quite insane, but it does give us more room and flexibility. Plenty sysctl options to adjust resampling characteristics. 2. Support 24/32 bit pcm format conversion through new, much improved, simplified and optimized feeder_fmt. Changes: 1. buffer.c / dsp.c / sound.h * Support for 24/32 AFMT. 2. feeder_rate.c * New implementation of sampling rate conversion with 32/64 bit precision, 1 - int32max hz (which is, ridiculous, yet very addictive). Much improved / smarter buffer management to not cause any missing samples at the end of conversion process * Tunable sysctls for various aspect: hw.snd.feeder_rate_ratemin - minimum allowable sampling rate (default to 4000) hw.snd.feeder_rate_ratemax - maximum allowable sampling rate (default to 1102500) hw.snd.feeder_rate_buffersize - conversion buffer size (default to 8192) hw.snd.feeder_rate_scaling - scaling / conversion method (please refer to the source for explaination). Default to previous implementation type. 3. feeder_fmt.c / sound.h * New implementation, support for 24/32bit conversion, optimized, and simplified. Few routines has been removed (8 to xlaw, 16 to 8). It just doesn't make sense. 4. channel.c * Support for 24/32 AFMT * Fix wrong xruns increment, causing incorrect underruns statistic while using vchans. 5. vchan.c * Support for 24/32 AFMT * Proper speed / rate detection especially for fixed rate ac97. User can override it using kernel hint: hint.pcm.<unit>.vchanrate="xxxx". Notes / Issues: * Virtual Channels (vchans) Enabling vchans can really, really help to solve overrun issues. This is quite understandable, because it operates entirely within its own buffering system without relying on hardware interrupt / state. Even if you don't need vchan, just enable single channel can help much. Few soundcards (notably via8233x, sblive, possibly others) have their own hardware multi channel, and this is unfortunately beyond vchan reachability. * The arrival of 24/32 also come with a price. Applications that can do 24/32bit playback need to be recompiled (notably mplayer). Use (recompiled) mplayer to experiment / test / debug this various format using -af format=fmt. Note that 24bit seeking in mplayer is a little bit broken, sometimes can cause silence or loud static noise. Pausing / seeking few times can solve this problem. You don't have to rebuild world entirely for this. Simply copy /usr/src/sys/sys/soundcard.h to /usr/include/sys/soundcard.h would suffice. Few drivers also need recompilation, and this can be done via /usr/src/sys/modules/sound/. Support for 24bit hardware playback is beyond the scope of this changes. That would require spessific hardware driver changes. * Don't expect playing 9999999999hz is a wise decision. Be reasonable. The new feeder_rate implemention provide flexibility, not insanity. You can easily chew up your CPU with this kind of mind instability. Please use proper mosquito repellent device for this obvious cracked brain attempt. As for testing purposes, you can use (again) mplayer to generate / play with different sampling rate. Use something like "mplayer -af resample=192000:0:0 <files>". Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my> Tested by: multimedia@
2005-07-31 16:16:22 +00:00
CHN_LOCK(c);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
if (c->direction == direction) {
Whats New: 1. Support wide range sampling rate, as low as 1hz up to int32 max (which is, insane) through new feeder_rate, multiple precisions choice (32/64 bit converter). This is indeed, quite insane, but it does give us more room and flexibility. Plenty sysctl options to adjust resampling characteristics. 2. Support 24/32 bit pcm format conversion through new, much improved, simplified and optimized feeder_fmt. Changes: 1. buffer.c / dsp.c / sound.h * Support for 24/32 AFMT. 2. feeder_rate.c * New implementation of sampling rate conversion with 32/64 bit precision, 1 - int32max hz (which is, ridiculous, yet very addictive). Much improved / smarter buffer management to not cause any missing samples at the end of conversion process * Tunable sysctls for various aspect: hw.snd.feeder_rate_ratemin - minimum allowable sampling rate (default to 4000) hw.snd.feeder_rate_ratemax - maximum allowable sampling rate (default to 1102500) hw.snd.feeder_rate_buffersize - conversion buffer size (default to 8192) hw.snd.feeder_rate_scaling - scaling / conversion method (please refer to the source for explaination). Default to previous implementation type. 3. feeder_fmt.c / sound.h * New implementation, support for 24/32bit conversion, optimized, and simplified. Few routines has been removed (8 to xlaw, 16 to 8). It just doesn't make sense. 4. channel.c * Support for 24/32 AFMT * Fix wrong xruns increment, causing incorrect underruns statistic while using vchans. 5. vchan.c * Support for 24/32 AFMT * Proper speed / rate detection especially for fixed rate ac97. User can override it using kernel hint: hint.pcm.<unit>.vchanrate="xxxx". Notes / Issues: * Virtual Channels (vchans) Enabling vchans can really, really help to solve overrun issues. This is quite understandable, because it operates entirely within its own buffering system without relying on hardware interrupt / state. Even if you don't need vchan, just enable single channel can help much. Few soundcards (notably via8233x, sblive, possibly others) have their own hardware multi channel, and this is unfortunately beyond vchan reachability. * The arrival of 24/32 also come with a price. Applications that can do 24/32bit playback need to be recompiled (notably mplayer). Use (recompiled) mplayer to experiment / test / debug this various format using -af format=fmt. Note that 24bit seeking in mplayer is a little bit broken, sometimes can cause silence or loud static noise. Pausing / seeking few times can solve this problem. You don't have to rebuild world entirely for this. Simply copy /usr/src/sys/sys/soundcard.h to /usr/include/sys/soundcard.h would suffice. Few drivers also need recompilation, and this can be done via /usr/src/sys/modules/sound/. Support for 24bit hardware playback is beyond the scope of this changes. That would require spessific hardware driver changes. * Don't expect playing 9999999999hz is a wise decision. Be reasonable. The new feeder_rate implemention provide flexibility, not insanity. You can easily chew up your CPU with this kind of mind instability. Please use proper mosquito repellent device for this obvious cracked brain attempt. As for testing purposes, you can use (again) mplayer to generate / play with different sampling rate. Use something like "mplayer -af resample=192000:0:0 <files>". Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my> Tested by: multimedia@
2005-07-31 16:16:22 +00:00
if (c->flags & CHN_F_VIRTUAL) {
/* Sanity check */
if (ch != NULL && ch != c->parentchannel) {
CHN_UNLOCK(c);
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
PCM_RELEASE_QUICK(d);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (EINVAL);
}
} else if (c->flags & CHN_F_HAS_VCHAN) {
/* No way!! */
if (ch != NULL) {
CHN_UNLOCK(c);
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
PCM_RELEASE_QUICK(d);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (EINVAL);
}
ch = c;
newspd = ch->speed;
Whats New: 1. Support wide range sampling rate, as low as 1hz up to int32 max (which is, insane) through new feeder_rate, multiple precisions choice (32/64 bit converter). This is indeed, quite insane, but it does give us more room and flexibility. Plenty sysctl options to adjust resampling characteristics. 2. Support 24/32 bit pcm format conversion through new, much improved, simplified and optimized feeder_fmt. Changes: 1. buffer.c / dsp.c / sound.h * Support for 24/32 AFMT. 2. feeder_rate.c * New implementation of sampling rate conversion with 32/64 bit precision, 1 - int32max hz (which is, ridiculous, yet very addictive). Much improved / smarter buffer management to not cause any missing samples at the end of conversion process * Tunable sysctls for various aspect: hw.snd.feeder_rate_ratemin - minimum allowable sampling rate (default to 4000) hw.snd.feeder_rate_ratemax - maximum allowable sampling rate (default to 1102500) hw.snd.feeder_rate_buffersize - conversion buffer size (default to 8192) hw.snd.feeder_rate_scaling - scaling / conversion method (please refer to the source for explaination). Default to previous implementation type. 3. feeder_fmt.c / sound.h * New implementation, support for 24/32bit conversion, optimized, and simplified. Few routines has been removed (8 to xlaw, 16 to 8). It just doesn't make sense. 4. channel.c * Support for 24/32 AFMT * Fix wrong xruns increment, causing incorrect underruns statistic while using vchans. 5. vchan.c * Support for 24/32 AFMT * Proper speed / rate detection especially for fixed rate ac97. User can override it using kernel hint: hint.pcm.<unit>.vchanrate="xxxx". Notes / Issues: * Virtual Channels (vchans) Enabling vchans can really, really help to solve overrun issues. This is quite understandable, because it operates entirely within its own buffering system without relying on hardware interrupt / state. Even if you don't need vchan, just enable single channel can help much. Few soundcards (notably via8233x, sblive, possibly others) have their own hardware multi channel, and this is unfortunately beyond vchan reachability. * The arrival of 24/32 also come with a price. Applications that can do 24/32bit playback need to be recompiled (notably mplayer). Use (recompiled) mplayer to experiment / test / debug this various format using -af format=fmt. Note that 24bit seeking in mplayer is a little bit broken, sometimes can cause silence or loud static noise. Pausing / seeking few times can solve this problem. You don't have to rebuild world entirely for this. Simply copy /usr/src/sys/sys/soundcard.h to /usr/include/sys/soundcard.h would suffice. Few drivers also need recompilation, and this can be done via /usr/src/sys/modules/sound/. Support for 24bit hardware playback is beyond the scope of this changes. That would require spessific hardware driver changes. * Don't expect playing 9999999999hz is a wise decision. Be reasonable. The new feeder_rate implemention provide flexibility, not insanity. You can easily chew up your CPU with this kind of mind instability. Please use proper mosquito repellent device for this obvious cracked brain attempt. As for testing purposes, you can use (again) mplayer to generate / play with different sampling rate. Use something like "mplayer -af resample=192000:0:0 <files>". Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my> Tested by: multimedia@
2005-07-31 16:16:22 +00:00
}
}
CHN_UNLOCK(c);
}
if (ch == NULL) {
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
PCM_RELEASE_QUICK(d);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (EINVAL);
Whats New: 1. Support wide range sampling rate, as low as 1hz up to int32 max (which is, insane) through new feeder_rate, multiple precisions choice (32/64 bit converter). This is indeed, quite insane, but it does give us more room and flexibility. Plenty sysctl options to adjust resampling characteristics. 2. Support 24/32 bit pcm format conversion through new, much improved, simplified and optimized feeder_fmt. Changes: 1. buffer.c / dsp.c / sound.h * Support for 24/32 AFMT. 2. feeder_rate.c * New implementation of sampling rate conversion with 32/64 bit precision, 1 - int32max hz (which is, ridiculous, yet very addictive). Much improved / smarter buffer management to not cause any missing samples at the end of conversion process * Tunable sysctls for various aspect: hw.snd.feeder_rate_ratemin - minimum allowable sampling rate (default to 4000) hw.snd.feeder_rate_ratemax - maximum allowable sampling rate (default to 1102500) hw.snd.feeder_rate_buffersize - conversion buffer size (default to 8192) hw.snd.feeder_rate_scaling - scaling / conversion method (please refer to the source for explaination). Default to previous implementation type. 3. feeder_fmt.c / sound.h * New implementation, support for 24/32bit conversion, optimized, and simplified. Few routines has been removed (8 to xlaw, 16 to 8). It just doesn't make sense. 4. channel.c * Support for 24/32 AFMT * Fix wrong xruns increment, causing incorrect underruns statistic while using vchans. 5. vchan.c * Support for 24/32 AFMT * Proper speed / rate detection especially for fixed rate ac97. User can override it using kernel hint: hint.pcm.<unit>.vchanrate="xxxx". Notes / Issues: * Virtual Channels (vchans) Enabling vchans can really, really help to solve overrun issues. This is quite understandable, because it operates entirely within its own buffering system without relying on hardware interrupt / state. Even if you don't need vchan, just enable single channel can help much. Few soundcards (notably via8233x, sblive, possibly others) have their own hardware multi channel, and this is unfortunately beyond vchan reachability. * The arrival of 24/32 also come with a price. Applications that can do 24/32bit playback need to be recompiled (notably mplayer). Use (recompiled) mplayer to experiment / test / debug this various format using -af format=fmt. Note that 24bit seeking in mplayer is a little bit broken, sometimes can cause silence or loud static noise. Pausing / seeking few times can solve this problem. You don't have to rebuild world entirely for this. Simply copy /usr/src/sys/sys/soundcard.h to /usr/include/sys/soundcard.h would suffice. Few drivers also need recompilation, and this can be done via /usr/src/sys/modules/sound/. Support for 24bit hardware playback is beyond the scope of this changes. That would require spessific hardware driver changes. * Don't expect playing 9999999999hz is a wise decision. Be reasonable. The new feeder_rate implemention provide flexibility, not insanity. You can easily chew up your CPU with this kind of mind instability. Please use proper mosquito repellent device for this obvious cracked brain attempt. As for testing purposes, you can use (again) mplayer to generate / play with different sampling rate. Use something like "mplayer -af resample=192000:0:0 <files>". Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my> Tested by: multimedia@
2005-07-31 16:16:22 +00:00
}
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
err = sysctl_handle_int(oidp, &newspd, 0, req);
Whats New: 1. Support wide range sampling rate, as low as 1hz up to int32 max (which is, insane) through new feeder_rate, multiple precisions choice (32/64 bit converter). This is indeed, quite insane, but it does give us more room and flexibility. Plenty sysctl options to adjust resampling characteristics. 2. Support 24/32 bit pcm format conversion through new, much improved, simplified and optimized feeder_fmt. Changes: 1. buffer.c / dsp.c / sound.h * Support for 24/32 AFMT. 2. feeder_rate.c * New implementation of sampling rate conversion with 32/64 bit precision, 1 - int32max hz (which is, ridiculous, yet very addictive). Much improved / smarter buffer management to not cause any missing samples at the end of conversion process * Tunable sysctls for various aspect: hw.snd.feeder_rate_ratemin - minimum allowable sampling rate (default to 4000) hw.snd.feeder_rate_ratemax - maximum allowable sampling rate (default to 1102500) hw.snd.feeder_rate_buffersize - conversion buffer size (default to 8192) hw.snd.feeder_rate_scaling - scaling / conversion method (please refer to the source for explaination). Default to previous implementation type. 3. feeder_fmt.c / sound.h * New implementation, support for 24/32bit conversion, optimized, and simplified. Few routines has been removed (8 to xlaw, 16 to 8). It just doesn't make sense. 4. channel.c * Support for 24/32 AFMT * Fix wrong xruns increment, causing incorrect underruns statistic while using vchans. 5. vchan.c * Support for 24/32 AFMT * Proper speed / rate detection especially for fixed rate ac97. User can override it using kernel hint: hint.pcm.<unit>.vchanrate="xxxx". Notes / Issues: * Virtual Channels (vchans) Enabling vchans can really, really help to solve overrun issues. This is quite understandable, because it operates entirely within its own buffering system without relying on hardware interrupt / state. Even if you don't need vchan, just enable single channel can help much. Few soundcards (notably via8233x, sblive, possibly others) have their own hardware multi channel, and this is unfortunately beyond vchan reachability. * The arrival of 24/32 also come with a price. Applications that can do 24/32bit playback need to be recompiled (notably mplayer). Use (recompiled) mplayer to experiment / test / debug this various format using -af format=fmt. Note that 24bit seeking in mplayer is a little bit broken, sometimes can cause silence or loud static noise. Pausing / seeking few times can solve this problem. You don't have to rebuild world entirely for this. Simply copy /usr/src/sys/sys/soundcard.h to /usr/include/sys/soundcard.h would suffice. Few drivers also need recompilation, and this can be done via /usr/src/sys/modules/sound/. Support for 24bit hardware playback is beyond the scope of this changes. That would require spessific hardware driver changes. * Don't expect playing 9999999999hz is a wise decision. Be reasonable. The new feeder_rate implemention provide flexibility, not insanity. You can easily chew up your CPU with this kind of mind instability. Please use proper mosquito repellent device for this obvious cracked brain attempt. As for testing purposes, you can use (again) mplayer to generate / play with different sampling rate. Use something like "mplayer -af resample=192000:0:0 <files>". Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my> Tested by: multimedia@
2005-07-31 16:16:22 +00:00
if (err == 0 && req->newptr != NULL) {
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
if (newspd < 1 || newspd < feeder_rate_min ||
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
newspd > feeder_rate_max) {
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
PCM_RELEASE_QUICK(d);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (EINVAL);
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
}
CHN_LOCK(ch);
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
if (feeder_rate_round) {
caps = chn_getcaps(ch);
if (caps == NULL || newspd < caps->minspeed ||
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
newspd > caps->maxspeed) {
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
CHN_UNLOCK(ch);
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
PCM_RELEASE_QUICK(d);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (EINVAL);
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
}
Whats New: 1. Support wide range sampling rate, as low as 1hz up to int32 max (which is, insane) through new feeder_rate, multiple precisions choice (32/64 bit converter). This is indeed, quite insane, but it does give us more room and flexibility. Plenty sysctl options to adjust resampling characteristics. 2. Support 24/32 bit pcm format conversion through new, much improved, simplified and optimized feeder_fmt. Changes: 1. buffer.c / dsp.c / sound.h * Support for 24/32 AFMT. 2. feeder_rate.c * New implementation of sampling rate conversion with 32/64 bit precision, 1 - int32max hz (which is, ridiculous, yet very addictive). Much improved / smarter buffer management to not cause any missing samples at the end of conversion process * Tunable sysctls for various aspect: hw.snd.feeder_rate_ratemin - minimum allowable sampling rate (default to 4000) hw.snd.feeder_rate_ratemax - maximum allowable sampling rate (default to 1102500) hw.snd.feeder_rate_buffersize - conversion buffer size (default to 8192) hw.snd.feeder_rate_scaling - scaling / conversion method (please refer to the source for explaination). Default to previous implementation type. 3. feeder_fmt.c / sound.h * New implementation, support for 24/32bit conversion, optimized, and simplified. Few routines has been removed (8 to xlaw, 16 to 8). It just doesn't make sense. 4. channel.c * Support for 24/32 AFMT * Fix wrong xruns increment, causing incorrect underruns statistic while using vchans. 5. vchan.c * Support for 24/32 AFMT * Proper speed / rate detection especially for fixed rate ac97. User can override it using kernel hint: hint.pcm.<unit>.vchanrate="xxxx". Notes / Issues: * Virtual Channels (vchans) Enabling vchans can really, really help to solve overrun issues. This is quite understandable, because it operates entirely within its own buffering system without relying on hardware interrupt / state. Even if you don't need vchan, just enable single channel can help much. Few soundcards (notably via8233x, sblive, possibly others) have their own hardware multi channel, and this is unfortunately beyond vchan reachability. * The arrival of 24/32 also come with a price. Applications that can do 24/32bit playback need to be recompiled (notably mplayer). Use (recompiled) mplayer to experiment / test / debug this various format using -af format=fmt. Note that 24bit seeking in mplayer is a little bit broken, sometimes can cause silence or loud static noise. Pausing / seeking few times can solve this problem. You don't have to rebuild world entirely for this. Simply copy /usr/src/sys/sys/soundcard.h to /usr/include/sys/soundcard.h would suffice. Few drivers also need recompilation, and this can be done via /usr/src/sys/modules/sound/. Support for 24bit hardware playback is beyond the scope of this changes. That would require spessific hardware driver changes. * Don't expect playing 9999999999hz is a wise decision. Be reasonable. The new feeder_rate implemention provide flexibility, not insanity. You can easily chew up your CPU with this kind of mind instability. Please use proper mosquito repellent device for this obvious cracked brain attempt. As for testing purposes, you can use (again) mplayer to generate / play with different sampling rate. Use something like "mplayer -af resample=192000:0:0 <files>". Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my> Tested by: multimedia@
2005-07-31 16:16:22 +00:00
}
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
if (CHN_STOPPED(ch) && newspd != ch->speed) {
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
err = chn_setspeed(ch, newspd);
/*
* Try to avoid FEEDER_RATE on parent channel if the
* requested value is not supported by the hardware.
*/
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
if (!err && feeder_rate_round &&
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
(ch->feederflags & (1 << FEEDER_RATE))) {
newspd = sndbuf_getspd(ch->bufhard);
err = chn_setspeed(ch, newspd);
}
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
if (err == 0)
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
*vchanrate = newspd;
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
}
CHN_UNLOCK(ch);
Whats New: 1. Support wide range sampling rate, as low as 1hz up to int32 max (which is, insane) through new feeder_rate, multiple precisions choice (32/64 bit converter). This is indeed, quite insane, but it does give us more room and flexibility. Plenty sysctl options to adjust resampling characteristics. 2. Support 24/32 bit pcm format conversion through new, much improved, simplified and optimized feeder_fmt. Changes: 1. buffer.c / dsp.c / sound.h * Support for 24/32 AFMT. 2. feeder_rate.c * New implementation of sampling rate conversion with 32/64 bit precision, 1 - int32max hz (which is, ridiculous, yet very addictive). Much improved / smarter buffer management to not cause any missing samples at the end of conversion process * Tunable sysctls for various aspect: hw.snd.feeder_rate_ratemin - minimum allowable sampling rate (default to 4000) hw.snd.feeder_rate_ratemax - maximum allowable sampling rate (default to 1102500) hw.snd.feeder_rate_buffersize - conversion buffer size (default to 8192) hw.snd.feeder_rate_scaling - scaling / conversion method (please refer to the source for explaination). Default to previous implementation type. 3. feeder_fmt.c / sound.h * New implementation, support for 24/32bit conversion, optimized, and simplified. Few routines has been removed (8 to xlaw, 16 to 8). It just doesn't make sense. 4. channel.c * Support for 24/32 AFMT * Fix wrong xruns increment, causing incorrect underruns statistic while using vchans. 5. vchan.c * Support for 24/32 AFMT * Proper speed / rate detection especially for fixed rate ac97. User can override it using kernel hint: hint.pcm.<unit>.vchanrate="xxxx". Notes / Issues: * Virtual Channels (vchans) Enabling vchans can really, really help to solve overrun issues. This is quite understandable, because it operates entirely within its own buffering system without relying on hardware interrupt / state. Even if you don't need vchan, just enable single channel can help much. Few soundcards (notably via8233x, sblive, possibly others) have their own hardware multi channel, and this is unfortunately beyond vchan reachability. * The arrival of 24/32 also come with a price. Applications that can do 24/32bit playback need to be recompiled (notably mplayer). Use (recompiled) mplayer to experiment / test / debug this various format using -af format=fmt. Note that 24bit seeking in mplayer is a little bit broken, sometimes can cause silence or loud static noise. Pausing / seeking few times can solve this problem. You don't have to rebuild world entirely for this. Simply copy /usr/src/sys/sys/soundcard.h to /usr/include/sys/soundcard.h would suffice. Few drivers also need recompilation, and this can be done via /usr/src/sys/modules/sound/. Support for 24bit hardware playback is beyond the scope of this changes. That would require spessific hardware driver changes. * Don't expect playing 9999999999hz is a wise decision. Be reasonable. The new feeder_rate implemention provide flexibility, not insanity. You can easily chew up your CPU with this kind of mind instability. Please use proper mosquito repellent device for this obvious cracked brain attempt. As for testing purposes, you can use (again) mplayer to generate / play with different sampling rate. Use something like "mplayer -af resample=192000:0:0 <files>". Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my> Tested by: multimedia@
2005-07-31 16:16:22 +00:00
}
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
PCM_RELEASE_QUICK(d);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (err);
Whats New: 1. Support wide range sampling rate, as low as 1hz up to int32 max (which is, insane) through new feeder_rate, multiple precisions choice (32/64 bit converter). This is indeed, quite insane, but it does give us more room and flexibility. Plenty sysctl options to adjust resampling characteristics. 2. Support 24/32 bit pcm format conversion through new, much improved, simplified and optimized feeder_fmt. Changes: 1. buffer.c / dsp.c / sound.h * Support for 24/32 AFMT. 2. feeder_rate.c * New implementation of sampling rate conversion with 32/64 bit precision, 1 - int32max hz (which is, ridiculous, yet very addictive). Much improved / smarter buffer management to not cause any missing samples at the end of conversion process * Tunable sysctls for various aspect: hw.snd.feeder_rate_ratemin - minimum allowable sampling rate (default to 4000) hw.snd.feeder_rate_ratemax - maximum allowable sampling rate (default to 1102500) hw.snd.feeder_rate_buffersize - conversion buffer size (default to 8192) hw.snd.feeder_rate_scaling - scaling / conversion method (please refer to the source for explaination). Default to previous implementation type. 3. feeder_fmt.c / sound.h * New implementation, support for 24/32bit conversion, optimized, and simplified. Few routines has been removed (8 to xlaw, 16 to 8). It just doesn't make sense. 4. channel.c * Support for 24/32 AFMT * Fix wrong xruns increment, causing incorrect underruns statistic while using vchans. 5. vchan.c * Support for 24/32 AFMT * Proper speed / rate detection especially for fixed rate ac97. User can override it using kernel hint: hint.pcm.<unit>.vchanrate="xxxx". Notes / Issues: * Virtual Channels (vchans) Enabling vchans can really, really help to solve overrun issues. This is quite understandable, because it operates entirely within its own buffering system without relying on hardware interrupt / state. Even if you don't need vchan, just enable single channel can help much. Few soundcards (notably via8233x, sblive, possibly others) have their own hardware multi channel, and this is unfortunately beyond vchan reachability. * The arrival of 24/32 also come with a price. Applications that can do 24/32bit playback need to be recompiled (notably mplayer). Use (recompiled) mplayer to experiment / test / debug this various format using -af format=fmt. Note that 24bit seeking in mplayer is a little bit broken, sometimes can cause silence or loud static noise. Pausing / seeking few times can solve this problem. You don't have to rebuild world entirely for this. Simply copy /usr/src/sys/sys/soundcard.h to /usr/include/sys/soundcard.h would suffice. Few drivers also need recompilation, and this can be done via /usr/src/sys/modules/sound/. Support for 24bit hardware playback is beyond the scope of this changes. That would require spessific hardware driver changes. * Don't expect playing 9999999999hz is a wise decision. Be reasonable. The new feeder_rate implemention provide flexibility, not insanity. You can easily chew up your CPU with this kind of mind instability. Please use proper mosquito repellent device for this obvious cracked brain attempt. As for testing purposes, you can use (again) mplayer to generate / play with different sampling rate. Use something like "mplayer -af resample=192000:0:0 <files>". Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my> Tested by: multimedia@
2005-07-31 16:16:22 +00:00
}
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
static int
sysctl_hw_snd_vchanformat(SYSCTL_HANDLER_ARGS)
{
struct snddev_info *d;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
struct pcm_channel *c, *ch = NULL;
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
uint32_t newfmt, spd;
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
int *vchanformat, vchancount, direction, err, i;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
char fmtstr[AFMTSTR_MAXSZ];
d = devclass_get_softc(pcm_devclass, VCHAN_SYSCTL_UNIT(oidp->oid_arg1));
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
if (!PCM_REGISTERED(d) || !(d->flags & SD_F_AUTOVCHAN))
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (EINVAL);
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
pcm_lock(d);
PCM_WAIT(d);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
switch (VCHAN_SYSCTL_DIR(oidp->oid_arg1)) {
case VCHAN_PLAY:
direction = PCMDIR_PLAY;
vchancount = d->pvchancount;
vchanformat = &d->pvchanformat;
break;
case VCHAN_REC:
direction = PCMDIR_REC;
vchancount = d->rvchancount;
vchanformat = &d->rvchanformat;
break;
default:
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
pcm_unlock(d);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (EINVAL);
break;
}
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
if (vchancount < 1) {
pcm_unlock(d);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (EINVAL);
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
}
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
PCM_ACQUIRE(d);
pcm_unlock(d);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
CHN_FOREACH(c, d, channels.pcm) {
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
CHN_LOCK(c);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
if (c->direction == direction) {
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
if (c->flags & CHN_F_VIRTUAL) {
/* Sanity check */
if (ch != NULL && ch != c->parentchannel) {
CHN_UNLOCK(c);
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
PCM_RELEASE_QUICK(d);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (EINVAL);
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
}
} else if (c->flags & CHN_F_HAS_VCHAN) {
/* No way!! */
if (ch != NULL) {
CHN_UNLOCK(c);
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
PCM_RELEASE_QUICK(d);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (EINVAL);
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
}
ch = c;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
if (ch->format !=
afmt2afmtstr(vchan_supported_fmts,
ch->format, fmtstr, sizeof(fmtstr),
AFMTSTR_FULL, AFMTSTR_STEREO_RETURN)) {
strlcpy(fmtstr, VCHAN_DEFAULT_STRFMT,
sizeof(fmtstr));
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
}
}
}
CHN_UNLOCK(c);
}
if (ch == NULL) {
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
PCM_RELEASE_QUICK(d);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (EINVAL);
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
}
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
err = sysctl_handle_string(oidp, fmtstr, sizeof(fmtstr), req);
if (err == 0 && req->newptr != NULL) {
for (i = 0; vchan_fmtstralias[i].alias != NULL; i++) {
if (strcmp(fmtstr, vchan_fmtstralias[i].alias) == 0) {
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
strlcpy(fmtstr, vchan_fmtstralias[i].fmtstr,
sizeof(fmtstr));
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
break;
}
}
newfmt = vchan_valid_strformat(fmtstr);
if (newfmt == 0) {
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
PCM_RELEASE_QUICK(d);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (EINVAL);
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
}
CHN_LOCK(ch);
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
if (CHN_STOPPED(ch) && newfmt != ch->format) {
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
/* Get channel speed, before chn_reset() screw it. */
spd = ch->speed;
err = chn_reset(ch, newfmt);
if (err == 0)
err = chn_setspeed(ch, spd);
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
if (err == 0)
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
*vchanformat = newfmt;
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
}
CHN_UNLOCK(ch);
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
}
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
PCM_RELEASE_QUICK(d);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (err);
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
}
Whats New: 1. Support wide range sampling rate, as low as 1hz up to int32 max (which is, insane) through new feeder_rate, multiple precisions choice (32/64 bit converter). This is indeed, quite insane, but it does give us more room and flexibility. Plenty sysctl options to adjust resampling characteristics. 2. Support 24/32 bit pcm format conversion through new, much improved, simplified and optimized feeder_fmt. Changes: 1. buffer.c / dsp.c / sound.h * Support for 24/32 AFMT. 2. feeder_rate.c * New implementation of sampling rate conversion with 32/64 bit precision, 1 - int32max hz (which is, ridiculous, yet very addictive). Much improved / smarter buffer management to not cause any missing samples at the end of conversion process * Tunable sysctls for various aspect: hw.snd.feeder_rate_ratemin - minimum allowable sampling rate (default to 4000) hw.snd.feeder_rate_ratemax - maximum allowable sampling rate (default to 1102500) hw.snd.feeder_rate_buffersize - conversion buffer size (default to 8192) hw.snd.feeder_rate_scaling - scaling / conversion method (please refer to the source for explaination). Default to previous implementation type. 3. feeder_fmt.c / sound.h * New implementation, support for 24/32bit conversion, optimized, and simplified. Few routines has been removed (8 to xlaw, 16 to 8). It just doesn't make sense. 4. channel.c * Support for 24/32 AFMT * Fix wrong xruns increment, causing incorrect underruns statistic while using vchans. 5. vchan.c * Support for 24/32 AFMT * Proper speed / rate detection especially for fixed rate ac97. User can override it using kernel hint: hint.pcm.<unit>.vchanrate="xxxx". Notes / Issues: * Virtual Channels (vchans) Enabling vchans can really, really help to solve overrun issues. This is quite understandable, because it operates entirely within its own buffering system without relying on hardware interrupt / state. Even if you don't need vchan, just enable single channel can help much. Few soundcards (notably via8233x, sblive, possibly others) have their own hardware multi channel, and this is unfortunately beyond vchan reachability. * The arrival of 24/32 also come with a price. Applications that can do 24/32bit playback need to be recompiled (notably mplayer). Use (recompiled) mplayer to experiment / test / debug this various format using -af format=fmt. Note that 24bit seeking in mplayer is a little bit broken, sometimes can cause silence or loud static noise. Pausing / seeking few times can solve this problem. You don't have to rebuild world entirely for this. Simply copy /usr/src/sys/sys/soundcard.h to /usr/include/sys/soundcard.h would suffice. Few drivers also need recompilation, and this can be done via /usr/src/sys/modules/sound/. Support for 24bit hardware playback is beyond the scope of this changes. That would require spessific hardware driver changes. * Don't expect playing 9999999999hz is a wise decision. Be reasonable. The new feeder_rate implemention provide flexibility, not insanity. You can easily chew up your CPU with this kind of mind instability. Please use proper mosquito repellent device for this obvious cracked brain attempt. As for testing purposes, you can use (again) mplayer to generate / play with different sampling rate. Use something like "mplayer -af resample=192000:0:0 <files>". Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my> Tested by: multimedia@
2005-07-31 16:16:22 +00:00
#endif
/* virtual channel interface */
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
#define VCHAN_FMT_HINT(x) ((x) == PCMDIR_PLAY_VIRTUAL) ? \
"play.vchanformat" : "rec.vchanformat"
#define VCHAN_SPD_HINT(x) ((x) == PCMDIR_PLAY_VIRTUAL) ? \
"play.vchanrate" : "rec.vchanrate"
int
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
vchan_create(struct pcm_channel *parent, int num)
{
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
struct snddev_info *d = parent->parentsnddev;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
struct pcm_channel *ch, *tmp, *after;
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
struct pcmchan_caps *parent_caps;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
uint32_t vchanfmt;
int err, first, speed, r;
int direction;
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
PCM_BUSYASSERT(d);
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
if (!(parent->flags & CHN_F_BUSY))
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (EBUSY);
if (parent->direction == PCMDIR_PLAY) {
direction = PCMDIR_PLAY_VIRTUAL;
vchanfmt = d->pvchanformat;
speed = d->pvchanrate;
} else if (parent->direction == PCMDIR_REC) {
direction = PCMDIR_REC_VIRTUAL;
vchanfmt = d->rvchanformat;
speed = d->rvchanrate;
} else
return (EINVAL);
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
CHN_UNLOCK(parent);
Change KASSERT() in feed_vchan16() into an explicit test and call to panic() so that the buffer overflow just beyond this point is always caught, even when the code is not compiled with INVARIANTS. Change chn_setblocksize() buffer reallocation code to attempt to avoid the feed_vchan16() buffer overflow by attempting to always keep the bufsoft buffer at least as large as the bufhard buffer. Print a diagnositic message Danger! %s bufsoft size increasing from %d to %d after CHANNEL_SETBLOCKSIZE() if our best attempts fail. If feed_vchan16() were to be called by the interrupt handler while locks are dropped in chn_setblocksize() to increase the size bufsoft to match the size of bufhard, the panic() code in feed_vchan16() will be triggered. If the diagnostic message is printed, it is a warning that a panic is possible if the system were to see events in an "unlucky" order. Change the locking code to avoid the need for MTX_RECURSIVE mutexes. Add the MTX_DUPOK option to the channel mutexes and change the locking sequence to always lock the parent channel before its children to avoid the possibility of deadlock. Actually implement locking assertions for the channel mutexes and fix the problems found by the resulting assertion violations. Clean up the locking code in dsp_ioctl(). Allocate the channel buffers using the malloc() M_WAITOK option instead of M_NOWAIT so that buffer allocation won't fail. Drop locks across the malloc() calls. Add/modify KASSERTS() in attempt to detect problems early. Abuse layering by adding a pointer to the snd_dbuf structure that points back to the pcm_channel that owns it. This allows sndbuf_resize() to do proper locking without having to change the its API, which is used by the hardware drivers. Don't dereference a NULL pointer when setting hw.snd.maxautovchans if a hardware driver is not loaded. Noticed by Ryan Sommers <ryans at gamersimpact.com>. Tested by: Stefan Ehmann <shoesoft AT gmx.net> Tested by: matk (Mathew Kanner) Tested by: Gordon Bergling <gbergling AT 0xfce3.net>
2004-01-28 08:02:15 +00:00
/* create a new playback channel */
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
pcm_lock(d);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
ch = pcm_chn_create(d, parent, &vchan_class, direction, num, parent);
if (ch == NULL) {
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
pcm_unlock(d);
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
CHN_LOCK(parent);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (ENODEV);
}
/* add us to our grandparent's channel list */
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
err = pcm_chn_add(d, ch);
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
pcm_unlock(d);
if (err) {
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
pcm_chn_destroy(ch);
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
CHN_LOCK(parent);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (err);
}
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
CHN_LOCK(parent);
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
/* add us to our parent channel's children */
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
first = CHN_EMPTY(parent, children);
after = NULL;
CHN_FOREACH(tmp, parent, children) {
if (CHN_CHAN(tmp) > CHN_CHAN(ch))
after = tmp;
else if (CHN_CHAN(tmp) < CHN_CHAN(ch))
break;
}
if (after != NULL) {
CHN_INSERT_AFTER(after, ch, children);
} else {
CHN_INSERT_HEAD(parent, ch, children);
}
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
parent->flags |= CHN_F_HAS_VCHAN;
Whats New: 1. Support wide range sampling rate, as low as 1hz up to int32 max (which is, insane) through new feeder_rate, multiple precisions choice (32/64 bit converter). This is indeed, quite insane, but it does give us more room and flexibility. Plenty sysctl options to adjust resampling characteristics. 2. Support 24/32 bit pcm format conversion through new, much improved, simplified and optimized feeder_fmt. Changes: 1. buffer.c / dsp.c / sound.h * Support for 24/32 AFMT. 2. feeder_rate.c * New implementation of sampling rate conversion with 32/64 bit precision, 1 - int32max hz (which is, ridiculous, yet very addictive). Much improved / smarter buffer management to not cause any missing samples at the end of conversion process * Tunable sysctls for various aspect: hw.snd.feeder_rate_ratemin - minimum allowable sampling rate (default to 4000) hw.snd.feeder_rate_ratemax - maximum allowable sampling rate (default to 1102500) hw.snd.feeder_rate_buffersize - conversion buffer size (default to 8192) hw.snd.feeder_rate_scaling - scaling / conversion method (please refer to the source for explaination). Default to previous implementation type. 3. feeder_fmt.c / sound.h * New implementation, support for 24/32bit conversion, optimized, and simplified. Few routines has been removed (8 to xlaw, 16 to 8). It just doesn't make sense. 4. channel.c * Support for 24/32 AFMT * Fix wrong xruns increment, causing incorrect underruns statistic while using vchans. 5. vchan.c * Support for 24/32 AFMT * Proper speed / rate detection especially for fixed rate ac97. User can override it using kernel hint: hint.pcm.<unit>.vchanrate="xxxx". Notes / Issues: * Virtual Channels (vchans) Enabling vchans can really, really help to solve overrun issues. This is quite understandable, because it operates entirely within its own buffering system without relying on hardware interrupt / state. Even if you don't need vchan, just enable single channel can help much. Few soundcards (notably via8233x, sblive, possibly others) have their own hardware multi channel, and this is unfortunately beyond vchan reachability. * The arrival of 24/32 also come with a price. Applications that can do 24/32bit playback need to be recompiled (notably mplayer). Use (recompiled) mplayer to experiment / test / debug this various format using -af format=fmt. Note that 24bit seeking in mplayer is a little bit broken, sometimes can cause silence or loud static noise. Pausing / seeking few times can solve this problem. You don't have to rebuild world entirely for this. Simply copy /usr/src/sys/sys/soundcard.h to /usr/include/sys/soundcard.h would suffice. Few drivers also need recompilation, and this can be done via /usr/src/sys/modules/sound/. Support for 24bit hardware playback is beyond the scope of this changes. That would require spessific hardware driver changes. * Don't expect playing 9999999999hz is a wise decision. Be reasonable. The new feeder_rate implemention provide flexibility, not insanity. You can easily chew up your CPU with this kind of mind instability. Please use proper mosquito repellent device for this obvious cracked brain attempt. As for testing purposes, you can use (again) mplayer to generate / play with different sampling rate. Use something like "mplayer -af resample=192000:0:0 <files>". Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my> Tested by: multimedia@
2005-07-31 16:16:22 +00:00
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
if (first) {
Whats New: 1. Support wide range sampling rate, as low as 1hz up to int32 max (which is, insane) through new feeder_rate, multiple precisions choice (32/64 bit converter). This is indeed, quite insane, but it does give us more room and flexibility. Plenty sysctl options to adjust resampling characteristics. 2. Support 24/32 bit pcm format conversion through new, much improved, simplified and optimized feeder_fmt. Changes: 1. buffer.c / dsp.c / sound.h * Support for 24/32 AFMT. 2. feeder_rate.c * New implementation of sampling rate conversion with 32/64 bit precision, 1 - int32max hz (which is, ridiculous, yet very addictive). Much improved / smarter buffer management to not cause any missing samples at the end of conversion process * Tunable sysctls for various aspect: hw.snd.feeder_rate_ratemin - minimum allowable sampling rate (default to 4000) hw.snd.feeder_rate_ratemax - maximum allowable sampling rate (default to 1102500) hw.snd.feeder_rate_buffersize - conversion buffer size (default to 8192) hw.snd.feeder_rate_scaling - scaling / conversion method (please refer to the source for explaination). Default to previous implementation type. 3. feeder_fmt.c / sound.h * New implementation, support for 24/32bit conversion, optimized, and simplified. Few routines has been removed (8 to xlaw, 16 to 8). It just doesn't make sense. 4. channel.c * Support for 24/32 AFMT * Fix wrong xruns increment, causing incorrect underruns statistic while using vchans. 5. vchan.c * Support for 24/32 AFMT * Proper speed / rate detection especially for fixed rate ac97. User can override it using kernel hint: hint.pcm.<unit>.vchanrate="xxxx". Notes / Issues: * Virtual Channels (vchans) Enabling vchans can really, really help to solve overrun issues. This is quite understandable, because it operates entirely within its own buffering system without relying on hardware interrupt / state. Even if you don't need vchan, just enable single channel can help much. Few soundcards (notably via8233x, sblive, possibly others) have their own hardware multi channel, and this is unfortunately beyond vchan reachability. * The arrival of 24/32 also come with a price. Applications that can do 24/32bit playback need to be recompiled (notably mplayer). Use (recompiled) mplayer to experiment / test / debug this various format using -af format=fmt. Note that 24bit seeking in mplayer is a little bit broken, sometimes can cause silence or loud static noise. Pausing / seeking few times can solve this problem. You don't have to rebuild world entirely for this. Simply copy /usr/src/sys/sys/soundcard.h to /usr/include/sys/soundcard.h would suffice. Few drivers also need recompilation, and this can be done via /usr/src/sys/modules/sound/. Support for 24bit hardware playback is beyond the scope of this changes. That would require spessific hardware driver changes. * Don't expect playing 9999999999hz is a wise decision. Be reasonable. The new feeder_rate implemention provide flexibility, not insanity. You can easily chew up your CPU with this kind of mind instability. Please use proper mosquito repellent device for this obvious cracked brain attempt. As for testing purposes, you can use (again) mplayer to generate / play with different sampling rate. Use something like "mplayer -af resample=192000:0:0 <files>". Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my> Tested by: multimedia@
2005-07-31 16:16:22 +00:00
parent_caps = chn_getcaps(parent);
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
if (parent_caps == NULL)
err = EINVAL;
if (!err) {
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
if (vchanfmt == 0) {
const char *vfmt;
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
CHN_UNLOCK(parent);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
r = resource_string_value(
device_get_name(parent->dev),
device_get_unit(parent->dev),
VCHAN_FMT_HINT(direction),
&vfmt);
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
CHN_LOCK(parent);
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
if (r != 0)
vfmt = NULL;
if (vfmt != NULL) {
vchanfmt = vchan_valid_strformat(vfmt);
for (r = 0; vchanfmt == 0 &&
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
vchan_fmtstralias[r].alias != NULL;
r++) {
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
if (strcmp(vfmt, vchan_fmtstralias[r].alias) == 0) {
vchanfmt = vchan_valid_strformat(vchan_fmtstralias[r].fmtstr);
break;
}
}
}
if (vchanfmt == 0)
vchanfmt = VCHAN_DEFAULT_AFMT;
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
}
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
err = chn_reset(parent, vchanfmt);
}
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
if (!err) {
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
/*
* This is very sad. Few soundcards advertised as being
* able to do (insanely) higher/lower speed, but in
* reality, they simply can't. At least, we give user chance
* to set sane value via kernel hints or sysctl.
*/
if (speed < 1) {
CHN_UNLOCK(parent);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
r = resource_int_value(
device_get_name(parent->dev),
device_get_unit(parent->dev),
VCHAN_SPD_HINT(direction),
&speed);
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
CHN_LOCK(parent);
if (r != 0) {
/*
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
* No saved value, no hint, NOTHING.
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
*
* Workaround for sb16 running
* poorly at 45k / 49k.
*/
switch (parent_caps->maxspeed) {
case 45000:
case 49000:
speed = 44100;
break;
default:
speed = VCHAN_DEFAULT_SPEED;
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
if (speed > parent_caps->maxspeed)
speed = parent_caps->maxspeed;
break;
}
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
if (speed < parent_caps->minspeed)
speed = parent_caps->minspeed;
}
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
}
Whats New: 1. Support wide range sampling rate, as low as 1hz up to int32 max (which is, insane) through new feeder_rate, multiple precisions choice (32/64 bit converter). This is indeed, quite insane, but it does give us more room and flexibility. Plenty sysctl options to adjust resampling characteristics. 2. Support 24/32 bit pcm format conversion through new, much improved, simplified and optimized feeder_fmt. Changes: 1. buffer.c / dsp.c / sound.h * Support for 24/32 AFMT. 2. feeder_rate.c * New implementation of sampling rate conversion with 32/64 bit precision, 1 - int32max hz (which is, ridiculous, yet very addictive). Much improved / smarter buffer management to not cause any missing samples at the end of conversion process * Tunable sysctls for various aspect: hw.snd.feeder_rate_ratemin - minimum allowable sampling rate (default to 4000) hw.snd.feeder_rate_ratemax - maximum allowable sampling rate (default to 1102500) hw.snd.feeder_rate_buffersize - conversion buffer size (default to 8192) hw.snd.feeder_rate_scaling - scaling / conversion method (please refer to the source for explaination). Default to previous implementation type. 3. feeder_fmt.c / sound.h * New implementation, support for 24/32bit conversion, optimized, and simplified. Few routines has been removed (8 to xlaw, 16 to 8). It just doesn't make sense. 4. channel.c * Support for 24/32 AFMT * Fix wrong xruns increment, causing incorrect underruns statistic while using vchans. 5. vchan.c * Support for 24/32 AFMT * Proper speed / rate detection especially for fixed rate ac97. User can override it using kernel hint: hint.pcm.<unit>.vchanrate="xxxx". Notes / Issues: * Virtual Channels (vchans) Enabling vchans can really, really help to solve overrun issues. This is quite understandable, because it operates entirely within its own buffering system without relying on hardware interrupt / state. Even if you don't need vchan, just enable single channel can help much. Few soundcards (notably via8233x, sblive, possibly others) have their own hardware multi channel, and this is unfortunately beyond vchan reachability. * The arrival of 24/32 also come with a price. Applications that can do 24/32bit playback need to be recompiled (notably mplayer). Use (recompiled) mplayer to experiment / test / debug this various format using -af format=fmt. Note that 24bit seeking in mplayer is a little bit broken, sometimes can cause silence or loud static noise. Pausing / seeking few times can solve this problem. You don't have to rebuild world entirely for this. Simply copy /usr/src/sys/sys/soundcard.h to /usr/include/sys/soundcard.h would suffice. Few drivers also need recompilation, and this can be done via /usr/src/sys/modules/sound/. Support for 24bit hardware playback is beyond the scope of this changes. That would require spessific hardware driver changes. * Don't expect playing 9999999999hz is a wise decision. Be reasonable. The new feeder_rate implemention provide flexibility, not insanity. You can easily chew up your CPU with this kind of mind instability. Please use proper mosquito repellent device for this obvious cracked brain attempt. As for testing purposes, you can use (again) mplayer to generate / play with different sampling rate. Use something like "mplayer -af resample=192000:0:0 <files>". Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my> Tested by: multimedia@
2005-07-31 16:16:22 +00:00
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
if (feeder_rate_round) {
/*
* Limit speed based on driver caps.
* This is supposed to help fixed rate, non-VRA
* AC97 cards, but.. (see below)
*/
if (speed < parent_caps->minspeed)
speed = parent_caps->minspeed;
if (speed > parent_caps->maxspeed)
speed = parent_caps->maxspeed;
}
Whats New: 1. Support wide range sampling rate, as low as 1hz up to int32 max (which is, insane) through new feeder_rate, multiple precisions choice (32/64 bit converter). This is indeed, quite insane, but it does give us more room and flexibility. Plenty sysctl options to adjust resampling characteristics. 2. Support 24/32 bit pcm format conversion through new, much improved, simplified and optimized feeder_fmt. Changes: 1. buffer.c / dsp.c / sound.h * Support for 24/32 AFMT. 2. feeder_rate.c * New implementation of sampling rate conversion with 32/64 bit precision, 1 - int32max hz (which is, ridiculous, yet very addictive). Much improved / smarter buffer management to not cause any missing samples at the end of conversion process * Tunable sysctls for various aspect: hw.snd.feeder_rate_ratemin - minimum allowable sampling rate (default to 4000) hw.snd.feeder_rate_ratemax - maximum allowable sampling rate (default to 1102500) hw.snd.feeder_rate_buffersize - conversion buffer size (default to 8192) hw.snd.feeder_rate_scaling - scaling / conversion method (please refer to the source for explaination). Default to previous implementation type. 3. feeder_fmt.c / sound.h * New implementation, support for 24/32bit conversion, optimized, and simplified. Few routines has been removed (8 to xlaw, 16 to 8). It just doesn't make sense. 4. channel.c * Support for 24/32 AFMT * Fix wrong xruns increment, causing incorrect underruns statistic while using vchans. 5. vchan.c * Support for 24/32 AFMT * Proper speed / rate detection especially for fixed rate ac97. User can override it using kernel hint: hint.pcm.<unit>.vchanrate="xxxx". Notes / Issues: * Virtual Channels (vchans) Enabling vchans can really, really help to solve overrun issues. This is quite understandable, because it operates entirely within its own buffering system without relying on hardware interrupt / state. Even if you don't need vchan, just enable single channel can help much. Few soundcards (notably via8233x, sblive, possibly others) have their own hardware multi channel, and this is unfortunately beyond vchan reachability. * The arrival of 24/32 also come with a price. Applications that can do 24/32bit playback need to be recompiled (notably mplayer). Use (recompiled) mplayer to experiment / test / debug this various format using -af format=fmt. Note that 24bit seeking in mplayer is a little bit broken, sometimes can cause silence or loud static noise. Pausing / seeking few times can solve this problem. You don't have to rebuild world entirely for this. Simply copy /usr/src/sys/sys/soundcard.h to /usr/include/sys/soundcard.h would suffice. Few drivers also need recompilation, and this can be done via /usr/src/sys/modules/sound/. Support for 24bit hardware playback is beyond the scope of this changes. That would require spessific hardware driver changes. * Don't expect playing 9999999999hz is a wise decision. Be reasonable. The new feeder_rate implemention provide flexibility, not insanity. You can easily chew up your CPU with this kind of mind instability. Please use proper mosquito repellent device for this obvious cracked brain attempt. As for testing purposes, you can use (again) mplayer to generate / play with different sampling rate. Use something like "mplayer -af resample=192000:0:0 <files>". Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my> Tested by: multimedia@
2005-07-31 16:16:22 +00:00
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
/*
* We still need to limit the speed between
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
* feeder_rate_min <-> feeder_rate_max. This is
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
* just an escape goat if all of the above failed
* miserably.
*/
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
if (speed < feeder_rate_min)
speed = feeder_rate_min;
if (speed > feeder_rate_max)
speed = feeder_rate_max;
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
err = chn_setspeed(parent, speed);
/*
* Try to avoid FEEDER_RATE on parent channel if the
* requested value is not supported by the hardware.
*/
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
if (!err && feeder_rate_round &&
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
(parent->feederflags & (1 << FEEDER_RATE))) {
speed = sndbuf_getspd(parent->bufhard);
err = chn_setspeed(parent, speed);
}
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
if (!err) {
Whats New: 1. Support wide range sampling rate, as low as 1hz up to int32 max (which is, insane) through new feeder_rate, multiple precisions choice (32/64 bit converter). This is indeed, quite insane, but it does give us more room and flexibility. Plenty sysctl options to adjust resampling characteristics. 2. Support 24/32 bit pcm format conversion through new, much improved, simplified and optimized feeder_fmt. Changes: 1. buffer.c / dsp.c / sound.h * Support for 24/32 AFMT. 2. feeder_rate.c * New implementation of sampling rate conversion with 32/64 bit precision, 1 - int32max hz (which is, ridiculous, yet very addictive). Much improved / smarter buffer management to not cause any missing samples at the end of conversion process * Tunable sysctls for various aspect: hw.snd.feeder_rate_ratemin - minimum allowable sampling rate (default to 4000) hw.snd.feeder_rate_ratemax - maximum allowable sampling rate (default to 1102500) hw.snd.feeder_rate_buffersize - conversion buffer size (default to 8192) hw.snd.feeder_rate_scaling - scaling / conversion method (please refer to the source for explaination). Default to previous implementation type. 3. feeder_fmt.c / sound.h * New implementation, support for 24/32bit conversion, optimized, and simplified. Few routines has been removed (8 to xlaw, 16 to 8). It just doesn't make sense. 4. channel.c * Support for 24/32 AFMT * Fix wrong xruns increment, causing incorrect underruns statistic while using vchans. 5. vchan.c * Support for 24/32 AFMT * Proper speed / rate detection especially for fixed rate ac97. User can override it using kernel hint: hint.pcm.<unit>.vchanrate="xxxx". Notes / Issues: * Virtual Channels (vchans) Enabling vchans can really, really help to solve overrun issues. This is quite understandable, because it operates entirely within its own buffering system without relying on hardware interrupt / state. Even if you don't need vchan, just enable single channel can help much. Few soundcards (notably via8233x, sblive, possibly others) have their own hardware multi channel, and this is unfortunately beyond vchan reachability. * The arrival of 24/32 also come with a price. Applications that can do 24/32bit playback need to be recompiled (notably mplayer). Use (recompiled) mplayer to experiment / test / debug this various format using -af format=fmt. Note that 24bit seeking in mplayer is a little bit broken, sometimes can cause silence or loud static noise. Pausing / seeking few times can solve this problem. You don't have to rebuild world entirely for this. Simply copy /usr/src/sys/sys/soundcard.h to /usr/include/sys/soundcard.h would suffice. Few drivers also need recompilation, and this can be done via /usr/src/sys/modules/sound/. Support for 24bit hardware playback is beyond the scope of this changes. That would require spessific hardware driver changes. * Don't expect playing 9999999999hz is a wise decision. Be reasonable. The new feeder_rate implemention provide flexibility, not insanity. You can easily chew up your CPU with this kind of mind instability. Please use proper mosquito repellent device for this obvious cracked brain attempt. As for testing purposes, you can use (again) mplayer to generate / play with different sampling rate. Use something like "mplayer -af resample=192000:0:0 <files>". Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my> Tested by: multimedia@
2005-07-31 16:16:22 +00:00
/*
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
* Save new value.
Whats New: 1. Support wide range sampling rate, as low as 1hz up to int32 max (which is, insane) through new feeder_rate, multiple precisions choice (32/64 bit converter). This is indeed, quite insane, but it does give us more room and flexibility. Plenty sysctl options to adjust resampling characteristics. 2. Support 24/32 bit pcm format conversion through new, much improved, simplified and optimized feeder_fmt. Changes: 1. buffer.c / dsp.c / sound.h * Support for 24/32 AFMT. 2. feeder_rate.c * New implementation of sampling rate conversion with 32/64 bit precision, 1 - int32max hz (which is, ridiculous, yet very addictive). Much improved / smarter buffer management to not cause any missing samples at the end of conversion process * Tunable sysctls for various aspect: hw.snd.feeder_rate_ratemin - minimum allowable sampling rate (default to 4000) hw.snd.feeder_rate_ratemax - maximum allowable sampling rate (default to 1102500) hw.snd.feeder_rate_buffersize - conversion buffer size (default to 8192) hw.snd.feeder_rate_scaling - scaling / conversion method (please refer to the source for explaination). Default to previous implementation type. 3. feeder_fmt.c / sound.h * New implementation, support for 24/32bit conversion, optimized, and simplified. Few routines has been removed (8 to xlaw, 16 to 8). It just doesn't make sense. 4. channel.c * Support for 24/32 AFMT * Fix wrong xruns increment, causing incorrect underruns statistic while using vchans. 5. vchan.c * Support for 24/32 AFMT * Proper speed / rate detection especially for fixed rate ac97. User can override it using kernel hint: hint.pcm.<unit>.vchanrate="xxxx". Notes / Issues: * Virtual Channels (vchans) Enabling vchans can really, really help to solve overrun issues. This is quite understandable, because it operates entirely within its own buffering system without relying on hardware interrupt / state. Even if you don't need vchan, just enable single channel can help much. Few soundcards (notably via8233x, sblive, possibly others) have their own hardware multi channel, and this is unfortunately beyond vchan reachability. * The arrival of 24/32 also come with a price. Applications that can do 24/32bit playback need to be recompiled (notably mplayer). Use (recompiled) mplayer to experiment / test / debug this various format using -af format=fmt. Note that 24bit seeking in mplayer is a little bit broken, sometimes can cause silence or loud static noise. Pausing / seeking few times can solve this problem. You don't have to rebuild world entirely for this. Simply copy /usr/src/sys/sys/soundcard.h to /usr/include/sys/soundcard.h would suffice. Few drivers also need recompilation, and this can be done via /usr/src/sys/modules/sound/. Support for 24bit hardware playback is beyond the scope of this changes. That would require spessific hardware driver changes. * Don't expect playing 9999999999hz is a wise decision. Be reasonable. The new feeder_rate implemention provide flexibility, not insanity. You can easily chew up your CPU with this kind of mind instability. Please use proper mosquito repellent device for this obvious cracked brain attempt. As for testing purposes, you can use (again) mplayer to generate / play with different sampling rate. Use something like "mplayer -af resample=192000:0:0 <files>". Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my> Tested by: multimedia@
2005-07-31 16:16:22 +00:00
*/
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
CHN_UNLOCK(parent);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
if (direction == PCMDIR_PLAY_VIRTUAL) {
d->pvchanformat = vchanfmt;
d->pvchanrate = speed;
} else {
d->rvchanformat = vchanfmt;
d->rvchanrate = speed;
}
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
CHN_LOCK(parent);
Whats New: 1. Support wide range sampling rate, as low as 1hz up to int32 max (which is, insane) through new feeder_rate, multiple precisions choice (32/64 bit converter). This is indeed, quite insane, but it does give us more room and flexibility. Plenty sysctl options to adjust resampling characteristics. 2. Support 24/32 bit pcm format conversion through new, much improved, simplified and optimized feeder_fmt. Changes: 1. buffer.c / dsp.c / sound.h * Support for 24/32 AFMT. 2. feeder_rate.c * New implementation of sampling rate conversion with 32/64 bit precision, 1 - int32max hz (which is, ridiculous, yet very addictive). Much improved / smarter buffer management to not cause any missing samples at the end of conversion process * Tunable sysctls for various aspect: hw.snd.feeder_rate_ratemin - minimum allowable sampling rate (default to 4000) hw.snd.feeder_rate_ratemax - maximum allowable sampling rate (default to 1102500) hw.snd.feeder_rate_buffersize - conversion buffer size (default to 8192) hw.snd.feeder_rate_scaling - scaling / conversion method (please refer to the source for explaination). Default to previous implementation type. 3. feeder_fmt.c / sound.h * New implementation, support for 24/32bit conversion, optimized, and simplified. Few routines has been removed (8 to xlaw, 16 to 8). It just doesn't make sense. 4. channel.c * Support for 24/32 AFMT * Fix wrong xruns increment, causing incorrect underruns statistic while using vchans. 5. vchan.c * Support for 24/32 AFMT * Proper speed / rate detection especially for fixed rate ac97. User can override it using kernel hint: hint.pcm.<unit>.vchanrate="xxxx". Notes / Issues: * Virtual Channels (vchans) Enabling vchans can really, really help to solve overrun issues. This is quite understandable, because it operates entirely within its own buffering system without relying on hardware interrupt / state. Even if you don't need vchan, just enable single channel can help much. Few soundcards (notably via8233x, sblive, possibly others) have their own hardware multi channel, and this is unfortunately beyond vchan reachability. * The arrival of 24/32 also come with a price. Applications that can do 24/32bit playback need to be recompiled (notably mplayer). Use (recompiled) mplayer to experiment / test / debug this various format using -af format=fmt. Note that 24bit seeking in mplayer is a little bit broken, sometimes can cause silence or loud static noise. Pausing / seeking few times can solve this problem. You don't have to rebuild world entirely for this. Simply copy /usr/src/sys/sys/soundcard.h to /usr/include/sys/soundcard.h would suffice. Few drivers also need recompilation, and this can be done via /usr/src/sys/modules/sound/. Support for 24bit hardware playback is beyond the scope of this changes. That would require spessific hardware driver changes. * Don't expect playing 9999999999hz is a wise decision. Be reasonable. The new feeder_rate implemention provide flexibility, not insanity. You can easily chew up your CPU with this kind of mind instability. Please use proper mosquito repellent device for this obvious cracked brain attempt. As for testing purposes, you can use (again) mplayer to generate / play with different sampling rate. Use something like "mplayer -af resample=192000:0:0 <files>". Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my> Tested by: multimedia@
2005-07-31 16:16:22 +00:00
}
}
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
if (err) {
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
CHN_REMOVE(parent, ch, children);
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
parent->flags &= ~CHN_F_HAS_VCHAN;
CHN_UNLOCK(parent);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
pcm_lock(d);
if (pcm_chn_remove(d, ch) == 0) {
pcm_unlock(d);
pcm_chn_destroy(ch);
} else
pcm_unlock(d);
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
CHN_LOCK(parent);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (err);
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
}
}
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (0);
}
int
vchan_destroy(struct pcm_channel *c)
{
struct pcm_channel *parent = c->parentchannel;
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
struct snddev_info *d = parent->parentsnddev;
uint32_t spd;
int err;
Last (again ?!?) major commit for RELENG_7, featuring total Giant eradication in/from userland path, countless locking fixes, etc. - General sleep call through msleep(9) has been converted to condvar(9) with better consistencies. - Heavily guard every possible "slow path" entries (open(), close(), few ioctl()s, sysctls), but once it entering "fast path" (io, interrupt started), they are free to fly on their own. - Rearrange locking sequences, resulting better concurrency and serialization. Large part doesn't even need locking at all, and will be removed in future. Less clutter, except in few places due to lock ordering. - Anonymous mixer object creation/deletion to simplify mixer handling beyond typical mixer ioctls. Submitted by: chibis (with modifications) - Add few mix_[get|set|..] functions to avoid calling mixer_ioctl() directly using cryptic arguments. - Locking fixes to avoid possible deadlock with (still under Giant) USB. - Better simplex/duplex device handling. - Recover mmap() functionality for recording, which has been lost since 2.2.x - 3.x (the introduction of newpcm). Full-duplex mmap still doesn't work (due to VM/page design), but people still can mmap both by opening each direction separately. mmaped playback is guarantee to work either way. - New sysctl: "hw.snd.compat_linux_mmap" to allow PROT_EXEC page mapping, due to recent changes in linux compatibility layer which require it. All linux applications that using sound + mmap() (mostly games) require this to be enabled. Disabled by default. - Other goodies.. too many, that will increase releng7 shareholder value and make users of releng6 (and below) cry ;) * This commit should be atomic. If anything goes wrong (not counting problem originated from elsewhere), I will not hesitate to revert everything back within 12 hours. This substantial changes itself not a rocket science and the process has begun for almost 2 years, and lots of incremental changes are already in place during that period of time. * Some issues does occur in snd_emu10kx (note the 'x') due to various internal locking issues and it is currently being worked on by chibis. Tested by: chibis (Yuriy Tsibizov), joel, Alexandre Vieira, many innocent souls...
2007-06-16 03:37:28 +00:00
PCM_BUSYASSERT(d);
CHN_LOCK(parent);
if (!(parent->flags & CHN_F_BUSY)) {
CHN_UNLOCK(parent);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (EBUSY);
}
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
if (CHN_EMPTY(parent, children)) {
CHN_UNLOCK(parent);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (EINVAL);
}
/* remove us from our parent's children list */
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
CHN_REMOVE(parent, c, children);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
if (CHN_EMPTY(parent, children)) {
parent->flags &= ~(CHN_F_BUSY | CHN_F_HAS_VCHAN);
spd = parent->speed;
if (chn_reset(parent, parent->format) == 0)
chn_setspeed(parent, spd);
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
}
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
CHN_UNLOCK(parent);
/* remove us from our grandparent's channel list */
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
pcm_lock(d);
err = pcm_chn_remove(d, c);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
pcm_unlock(d);
/* destroy ourselves */
- channel.h * New definition CHN_F_HAS_VCHAN. - channel.c * Use CHN_F_HAS_VCHAN to mark channel with vchan capability instead of relying on SLIST_EMPTY(&channel->children) == true for better clarification and future possible usages of children (like 'slave' channel). * Various fixes, including blocksize / format bps allignment, better 24bit seeking (mplayer, others). * Improve format chain building, it's now possible to record something to a format non-native to the soundcard through various feeder format converters or to higher sampling rate. This also gains another feature, like doing vchan mixing on non s16le soundcard such as sb8. - sound.c * Increase robustness within various function that handle vchan creation / termination (these function need a total rewrite, but that would cause other major rewrite within various places too!). As far as its robustness can be guaranteed, leave it as is. * Optimize channel ordering, prefer *real* hardware playback channels over virtual channels. cat /dev/sndstat should look better. * Increase sndstat verbosity to include bufsoft/bufhard allocation. - vchan.c * Fix LOR 119. - http://sources.zabbadoz.net/freebsd/lor.html#119 * Reorder / increase robustness of vchan_create() / destroy(). Enforce destroy_dev() during destroy operation, fix possible panic / dangling character device. - http://lists.freebsd.org/pipermail/freebsd-current/2005-May/050308.html * Tolerate a little bit more during mixing process, this should help non s16le soundcards. Note: Recoring in a non-native rate/format may result in overruns. A friendly application is wavrec from audio/wavplay. The problem is under investigation. Submitted by: Ariff Abdullah <skywizard@MyBSD.org.my>
2005-09-10 18:10:31 +00:00
if (!err)
err = pcm_chn_destroy(c);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (err);
}
int
vchan_initsys(device_t dev)
{
#ifdef SND_DYNSYSCTL
struct snddev_info *d;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
int unit;
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
unit = device_get_unit(dev);
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
d = device_get_softc(dev);
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
/* Play */
SYSCTL_ADD_PROC(&d->play_sysctl_ctx,
SYSCTL_CHILDREN(d->play_sysctl_tree),
OID_AUTO, "vchans", CTLTYPE_INT | CTLFLAG_RW,
VCHAN_SYSCTL_DATA(unit, PLAY), VCHAN_SYSCTL_DATA_SIZE,
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
sysctl_hw_snd_vchans, "I", "total allocated virtual channel");
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
SYSCTL_ADD_PROC(&d->play_sysctl_ctx,
SYSCTL_CHILDREN(d->play_sysctl_tree),
OID_AUTO, "vchanrate", CTLTYPE_INT | CTLFLAG_RW,
VCHAN_SYSCTL_DATA(unit, PLAY), VCHAN_SYSCTL_DATA_SIZE,
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
sysctl_hw_snd_vchanrate, "I", "virtual channel mixing speed/rate");
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
SYSCTL_ADD_PROC(&d->play_sysctl_ctx,
SYSCTL_CHILDREN(d->play_sysctl_tree),
OID_AUTO, "vchanformat", CTLTYPE_STRING | CTLFLAG_RW,
VCHAN_SYSCTL_DATA(unit, PLAY), VCHAN_SYSCTL_DATA_SIZE,
sysctl_hw_snd_vchanformat, "A", "virtual channel format");
/* Rec */
SYSCTL_ADD_PROC(&d->rec_sysctl_ctx,
SYSCTL_CHILDREN(d->rec_sysctl_tree),
OID_AUTO, "vchans", CTLTYPE_INT | CTLFLAG_RW,
VCHAN_SYSCTL_DATA(unit, REC), VCHAN_SYSCTL_DATA_SIZE,
sysctl_hw_snd_vchans, "I", "total allocated virtual channel");
SYSCTL_ADD_PROC(&d->rec_sysctl_ctx,
SYSCTL_CHILDREN(d->rec_sysctl_tree),
OID_AUTO, "vchanrate", CTLTYPE_INT | CTLFLAG_RW,
VCHAN_SYSCTL_DATA(unit, REC), VCHAN_SYSCTL_DATA_SIZE,
sysctl_hw_snd_vchanrate, "I", "virtual channel base speed/rate");
SYSCTL_ADD_PROC(&d->rec_sysctl_ctx,
SYSCTL_CHILDREN(d->rec_sysctl_tree),
OID_AUTO, "vchanformat", CTLTYPE_STRING | CTLFLAG_RW,
VCHAN_SYSCTL_DATA(unit, REC), VCHAN_SYSCTL_DATA_SIZE,
Welcome to Once-a-year Sound Mega-Commit. Enjoy numerous updates and fixes in every sense. General ------- - Multichannel safe, endian safe, format safe * Large part of critical pcm filters such as vchan.c, feeder_rate.c, feeder_volume.c, feeder_fmt.c and feeder.c has been rewritten so that using them does not cause the pcm data to be converted to 16bit little endian. * Macrosses for accessing pcm data safely are defined within sound.h in the form of PCM_READ_* / PCM_WRITE_* * Currently, most of them are probably limited for mono/stereo handling, but the future addition of true multichannel will be much easier. - Low latency operation * Well, this require lot more works to do not just within sound driver, but we're heading towards right direction. Buffer/block sizing within channel.c is rewritten to calculate precise allocation for various combination of sample/data/rate size. As a result, applying correct SNDCTL_DSP_POLICY value will achive expected latency behaviour simmilar to what commercial 4front driver do. * Signal handling fix. ctrl+c of "cat /dev/zero > /dev/dsp" does not result long delay. * Eliminate sound truncation if the sound data is too small. DIY: 1) Download / extract http://people.freebsd.org/~ariff/lowlatency/shortfiles.tar.gz 2) Do a comparison between "cat state*.au > /dev/dsp" and "for x in state*.au ; do cat $x > /dev/dsp ; done" - there should be no "perceivable" differences. Double close for PR kern/31445. CAVEAT: Low latency come with (unbearable) price especially for poorly written applications. Applications that trying to act smarter by requesting (wrong) blocksize/blockcount will suffer the most. Fixup samples/patches can be found at: http://people.freebsd.org/~ariff/ports/ - Switch minimum/maximum sampling rate limit to "1" and "2016000" (48k * 42) due to closer compatibility with 4front driver. Discussed with: marcus@ (long time ago?) - All driver specific sysctls in the form of "hw.snd.pcm%d.*" have been moved to their own dev sysctl nodes, notably: hw.snd.pcm%d.vchans -> dev.pcm.%d.vchans Bump __FreeBSD_version. Driver specific --------------- - Ditto for sysctls. - snd_atiixp, snd_es137x, snd_via8233, snd_hda * Numerous cleanups and fixes. * _EXPERIMENTAL_ polling mode support using simple callout_* mechanisme. This was intended for pure debugging and latency measurement, but proven good enough in few unexpected and rare cases (such as problematic shared IRQ with GIANT devices - USB). Polling can be enabled/disabled through dev.pcm.0.polling. Disabled by default. - snd_ich * Fix possible overflow during speed calibration. Delay final initialization (pcm_setstatus) after calibration finished. PR: kern/100169 Tested by: Kevin Overman <oberman@es.net> * Inverted EAPD for few Nec VersaPro. PR: kern/104715 Submitted by: KAWATA Masahiko <kawata@mta.biglobe.ne.jp> Thanks to various people, notably Joel Dahl, Yuriy Tsibizov, Kevin Oberman, those at #freebsd-azalia @ freenode and others for testing. Joel Dahl will do the manpage update.
2006-11-26 12:24:06 +00:00
sysctl_hw_snd_vchanformat, "A", "virtual channel format");
#endif
Last major commit and updates for RELENG_7: - Rework the entire pcm_channel structure: * Remove rarely used link placeholder, instead, make each pcm_channel as head/link of each own/each other. Unlock - Lock sequence due to sleep malloc has been reduced. * Implement "busy" queue which will contain list of busy/active channels. This greatly reduce locking contention for example while servicing interrupt for hardware with many channels or when virtual channels reach its 256 peak channels. - So I heard you like v chan ... O RLY? Welcome to Virtual **Record** Channels (vrec, rec vchans, vchans for recording, Rec-Chan, you decide), the ultimate solutions for your nagging O_RDWR full-duplex wannabe (note: flash plugins) monopolizing single record channel causing EBUSY. Vrec works exactly like Vchans (or, should I rename it to "Vplay" :) , except that it operates on the opposite direction (recording). Up to 256 vrecs (like vchans) are possible. Notes: * Relocate dev.pcm.%d.{vchans,vchanformat,vchanrate} to each of its respective node/direction: dev.pcm.%d.play.* for "play" (cdev = dsp%d.vp%d) dev.pcm.%d.rec.* for "record" (cdev = dsp%d.vr%d) * Don't expect that it will magically give you ability to split "recording source" (eg: 1 channel for cdrom, 1 channel for mic, etc). Just admit that you only have a *single* recording source / channel. Please bug your hardware vendor instead :) - Bump maxautovchans from 4 to 16. For a full-fledged multimedia desktop/workstation with too many soundservers installed (esound, artsd, jackd, pulse/polypaudio, ding-dong pling plong mudkip fuh fuh, etc), 4 seems inadequate. There will be no memory penalty here, since virtual channels are allocate only by demand. - Nuke/Rework the entire statically created cdev entries. Everything is clonable through snd own clone manager which designed to withstand many kind of abusive devfs droids such as: * while : ; do /bin/test -e /dev/dsp ; done * jot 16777216 0 | while read x ; do ls /dev/dsp0.$x ; done * hundreds (could be thousands) concurrent threads/process opening "/dev/dsp" (previously, this might result EBUSY even with just 3 contesting threads/procs). o Reusable clone objects (instead of creating new one like there's no tomorrow) after certain expiration deadline. The clone allocator will decide whether to reuse, share, or creating new clone. o Automatic garbage collector. - Dynamic unit magic allocator. Maximum attached soundcards can be tuned using tunable "hw.snd.maxunit" (Default to 512). Minimum is 16, and maximum is 2048. - ..other fixes, mostly related to concurrency issues. joel@ will do the manpage updates on sound(4). Have fun.
2007-05-31 18:43:33 +00:00
return (0);
}