qf/quakeforge
0
0
Fork 0
mirror of https://git.code.sf.net/p/quake/quakeforge synced 2024-12-11 21:31:30 +00:00
Code Issues Projects Releases Packages Wiki Activity
quakeforge/libs/audio/targets/snd_alsa.c

768 lines
18 KiB
C
Raw Normal View History

Move sound targets into a library. Still doesn't pick one at runtime though.
2001-04-10 09:40:09 +00:00
/*
drop alsa 0.5 support and rename snd_alsa_0_9.c to snd_alsa.c. also rename the alsa0_9 plugin to alsa.
2003-09-09 18:43:38 +00:00
snd_alsa.c
Move sound targets into a library. Still doesn't pick one at runtime though.
2001-04-10 09:40:09 +00:00
Check return conditions during ALSA setup. Error checking during setup is a good thing...
2004-01-17 22:03:58 +00:00
Support for the ALSA 1.0.1 sound driver
Move sound targets into a library. Still doesn't pick one at runtime though.
2001-04-10 09:40:09 +00:00
Copyright (C) 1999,2000 contributors of the QuakeForge project
Please see the file "AUTHORS" for a list of contributors
This program is free software; you can redistribute it and/or
modify it under the terms of the GNU General Public License
as published by the Free Software Foundation; either version 2
of the License, or (at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to:
Free Software Foundation, Inc.
59 Temple Place - Suite 330
Boston, MA 02111-1307, USA
*/
#ifdef HAVE_CONFIG_H
# include "config.h"
#endif
#include <stdio.h>
the "two birds with one stone" commit 1) when /not/ building static plugins, the plugin info name is, once again, PluginInfo, but is still type_name_PluginInfo for static plugins. This allows plugins to be symlinked (highly desirable for the debian packages, and likely others). This also requires plugins to /not/ be loaded with RTLD_GLOBAL. 2) because of 1, snd_alsa_0_9 has to explicitly load libasound itself. This just happens to fix my segfault on shutdown in RedHat 6.2.
2002-04-19 20:08:54 +00:00
#include <dlfcn.h>
Remove ugly ifdef (and coincidentally support for early ALSA 0.9).
2001-11-24 08:05:44 +00:00
#include <alsa/asoundlib.h>
Move sound targets into a library. Still doesn't pick one at runtime though.
2001-04-10 09:40:09 +00:00
Lots of whitespace.
2001-08-22 20:26:25 +00:00
#include "QF/cvar.h"
#include "QF/qargs.h"
total nukage of Con_*Printf calls from libs (except for video). Sys_*Printf (Sys_DPrintf is new) is now used exclusively for all lib printing. Con_Init sets the sys printf recirection to Con_Print (which has been revamped appropriatly) and the server sets it to SV_Print (which was SV_Printf and the new SV_Printf calls /it/). This should fix the rcon print redirection issues.
2001-09-21 04:22:46 +00:00
#include "QF/sys.h"
Move sound targets into a library. Still doesn't pick one at runtime though.
2001-04-10 09:40:09 +00:00
Rename some headers to fix doxygen warnings.
2011-07-23 06:56:22 +00:00
#include "snd_internal.h"
a bunch of docs for snd_render.h and a some cleanup of namepace and old cruft
2007-03-10 04:21:32 +00:00
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
typedef struct {
const snd_pcm_channel_area_t *areas;
snd_pcm_uframes_t offset;
snd_pcm_uframes_t nframes;
} alsa_pkt_t;
Check return conditions during ALSA setup. Error checking during setup is a good thing...
2004-01-17 22:03:58 +00:00
static int snd_inited;
static int snd_blocked = 0;
static snd_pcm_uframes_t buffer_size;
get alsa0_9 working with latest cvs, we need to dlopen the plugin with RTDL_GLOBAL. To use RTDL_GLOBAL, we need to make as much static as possible and PluginInfo needs a unique name (%s_%s_PluginInfo, type, name).
2001-08-31 00:05:58 +00:00
Check return conditions during ALSA setup. Error checking during setup is a good thing...
2004-01-17 22:03:58 +00:00
static void *alsa_handle;
static snd_pcm_t *pcm;
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
static snd_async_handler_t *async_handler;
get alsa0_9 working with latest cvs, we need to dlopen the plugin with RTDL_GLOBAL. To use RTDL_GLOBAL, we need to make as much static as possible and PluginInfo needs a unique name (%s_%s_PluginInfo, type, name).
2001-08-31 00:05:58 +00:00
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
static int snd_bits;
static cvar_t snd_bits_cvar = {
.name = "snd_bits",
.description =
"sound sample depth. 0 is system default",
.default_value = "0",
.flags = CVAR_ROM,
.value = { .type = &cexpr_int, .value = &snd_bits },
};
static char *snd_device;
static cvar_t snd_device_cvar = {
.name = "snd_device",
.description =
"sound device. \"\" is system default",
.default_value = "",
.flags = CVAR_ROM,
.value = { .type = 0, .value = &snd_device },
};
static int snd_rate;
static cvar_t snd_rate_cvar = {
.name = "snd_rate",
.description =
"sound playback rate. 0 is system default",
.default_value = "0",
.flags = CVAR_ROM,
.value = { .type = &cexpr_int, .value = &snd_rate },
};
static int snd_stereo;
static cvar_t snd_stereo_cvar = {
.name = "snd_stereo",
.description =
"sound stereo output",
.default_value = "1",
.flags = CVAR_ROM,
.value = { .type = &cexpr_int, .value = &snd_stereo },
};
Lots of whitespace, header cleanups, and a few tiny optimizations.
2001-09-09 21:41:35 +00:00
[audio] Allow output plugins to specify model Output plugins can use either a push model (synchronous) or a pull model (asynchronous). The ALSA plugin now uses the pull model. This paves the way for making jack output a simple output plugin rather than the combined render/output plugin it currently is (for #16) as now snd_dma works with both models.
2021-06-25 02:41:42 +00:00
//FIXME xfer probably should not be touching this (such data should probably
//come through snd_t)
static snd_output_data_t plugin_info_snd_output_data;
the "two birds with one stone" commit 1) when /not/ building static plugins, the plugin info name is, once again, PluginInfo, but is still type_name_PluginInfo for static plugins. This allows plugins to be symlinked (highly desirable for the debian packages, and likely others). This also requires plugins to /not/ be loaded with RTLD_GLOBAL. 2) because of 1, snd_alsa_0_9 has to explicitly load libasound itself. This just happens to fix my segfault on shutdown in RedHat 6.2.
2002-04-19 20:08:54 +00:00
#define QF_ALSA_NEED(ret, func, params) \
static ret (*qf##func) params;
#include "alsa_funcs_list.h"
#undef QF_ALSA_NEED
Nuke qboolean from orbit I never liked it, but with C2x coming out, it's best to handle bools properly. I haven't gone through all the uses of int as bool (I'll leave that for fixing when I encounter them), but this gets QF working with both c2x (really, gnu2x because of raw strings).
2023-06-13 09:06:11 +00:00
static bool
the "two birds with one stone" commit 1) when /not/ building static plugins, the plugin info name is, once again, PluginInfo, but is still type_name_PluginInfo for static plugins. This allows plugins to be symlinked (highly desirable for the debian packages, and likely others). This also requires plugins to /not/ be loaded with RTLD_GLOBAL. 2) because of 1, snd_alsa_0_9 has to explicitly load libasound itself. This just happens to fix my segfault on shutdown in RedHat 6.2.
2002-04-19 20:08:54 +00:00
load_libasound (void)
{
if (!(alsa_handle = dlopen ("libasound.so.2", RTLD_GLOBAL | RTLD_NOW))) {
Update ALSA sound target to support 1.0. Breaks 0.9 support, but 1.0 is in 2.6.0. So either update, or revert to older snd_alsa.c & alsa_funcs_list.h.
2003-12-22 01:07:49 +00:00
Sys_Printf ("Couldn't load libasound.so.2: %s\n", dlerror ());
the "two birds with one stone" commit 1) when /not/ building static plugins, the plugin info name is, once again, PluginInfo, but is still type_name_PluginInfo for static plugins. This allows plugins to be symlinked (highly desirable for the debian packages, and likely others). This also requires plugins to /not/ be loaded with RTLD_GLOBAL. 2) because of 1, snd_alsa_0_9 has to explicitly load libasound itself. This just happens to fix my segfault on shutdown in RedHat 6.2.
2002-04-19 20:08:54 +00:00
return false;
}
#define QF_ALSA_NEED(ret, func, params) \
if (!(qf##func = dlsym (alsa_handle, #func))) { \
Sys_Printf ("Couldn't load ALSA function %s\n", #func); \
dlclose (alsa_handle); \
alsa_handle = 0; \
return false; \
}
#include "alsa_funcs_list.h"
#undef QF_ALSA_NEED
return true;
}
#define snd_pcm_hw_params_sizeof qfsnd_pcm_hw_params_sizeof
#define snd_pcm_sw_params_sizeof qfsnd_pcm_sw_params_sizeof
static void
SNDDMA_Init_Cvars (void)
Put the sound renderer in it's proper place: as a full-fledged plugin. Bugs expected. Please report them :)
2001-08-23 04:01:46 +00:00
{
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
Cvar_Register (&snd_stereo_cvar, 0, 0);
Cvar_Register (&snd_rate_cvar, 0, 0);
Cvar_Register (&snd_device_cvar, 0, 0);
Cvar_Register (&snd_bits_cvar, 0, 0);
Put the sound renderer in it's proper place: as a full-fledged plugin. Bugs expected. Please report them :)
2001-08-23 04:01:46 +00:00
}
whitespace and comment cleanup.
2001-05-24 17:00:08 +00:00
Do a pure/const/noreturn/format attribute pass. I always wanted these, but as gcc now provides warnings for functions that could do with such attributes, finding all the functions is much easier.
2018-10-09 03:35:01 +00:00
static __attribute__((const)) snd_pcm_uframes_t
attempt to fix choppy audio in alsa caused by non-power-of 2 buffer size
2005-06-14 05:37:26 +00:00
round_buffer_size (snd_pcm_uframes_t sz)
{
snd_pcm_uframes_t mask = ~0;
while (sz & mask) {
sz &= mask;
mask <<= 1;
}
return sz;
}
Support non-interleaved mmap sound. Unfortuanately, I can't test this properly as I don't have any such hardware, but as the code is mosly an edited copy of the interleaved code, any errors should be easy to fix.
2011-09-07 08:00:57 +00:00
static inline int
clamp_16 (int val)
{
if (val > 0x7fff)
val = 0x7fff;
else if (val < -0x8000)
val = -0x8000;
return val;
}
static inline int
clamp_8 (int val)
{
if (val > 0x7f)
val = 0x7f;
else if (val < -0x80)
val = -0x80;
return val;
}
static void
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
alsa_ni_xfer (snd_t *snd, portable_samplepair_t *paintbuffer, int count,
float volume)
Support non-interleaved mmap sound. Unfortuanately, I can't test this properly as I don't have any such hardware, but as the code is mosly an edited copy of the interleaved code, any errors should be easy to fix.
2011-09-07 08:00:57 +00:00
{
const snd_pcm_channel_area_t *areas;
Fix some undefined symbols in snd_alsa.c. Serves me right for testing with only my debug build :P
2011-09-08 09:10:09 +00:00
int out_idx, out_max;
Support non-interleaved mmap sound. Unfortuanately, I can't test this properly as I don't have any such hardware, but as the code is mosly an edited copy of the interleaved code, any errors should be easy to fix.
2011-09-07 08:00:57 +00:00
float *p;
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
areas = snd->xfer_data;
Support non-interleaved mmap sound. Unfortuanately, I can't test this properly as I don't have any such hardware, but as the code is mosly an edited copy of the interleaved code, any errors should be easy to fix.
2011-09-07 08:00:57 +00:00
Fix some undefined symbols in snd_alsa.c. Serves me right for testing with only my debug build :P
2011-09-08 09:10:09 +00:00
p = (float *) paintbuffer;
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
out_max = snd->frames - 1;
Fix the ghastly sound quality. Due to quake's original sound engine using a push model, the actual place to which the sound data should be written is not necessarily where the "hardware" dma cursor is, but rather where the last write finished off. Thus, the correct output location is indicated by snd_paintedtime rather than snd_shm->framepos.
2011-12-08 02:02:19 +00:00
out_idx = *plugin_info_snd_output_data.paintedtime;
Support non-interleaved mmap sound. Unfortuanately, I can't test this properly as I don't have any such hardware, but as the code is mosly an edited copy of the interleaved code, any errors should be easy to fix.
2011-09-07 08:00:57 +00:00
while (out_idx > out_max)
out_idx -= out_max + 1;
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
if (snd->samplebits == 16) {
Support non-interleaved mmap sound. Unfortuanately, I can't test this properly as I don't have any such hardware, but as the code is mosly an edited copy of the interleaved code, any errors should be easy to fix.
2011-09-07 08:00:57 +00:00
short *out_0 = (short *) areas[0].addr;
short *out_1 = (short *) areas[1].addr;
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
if (snd->channels == 2) {
Support non-interleaved mmap sound. Unfortuanately, I can't test this properly as I don't have any such hardware, but as the code is mosly an edited copy of the interleaved code, any errors should be easy to fix.
2011-09-07 08:00:57 +00:00
while (count--) {
Fix some undefined symbols in snd_alsa.c. Serves me right for testing with only my debug build :P
2011-09-08 09:10:09 +00:00
out_0[out_idx] = clamp_16 ((*p++ * volume) * 0x8000);
out_1[out_idx] = clamp_16 ((*p++ * volume) * 0x8000);
Support non-interleaved mmap sound. Unfortuanately, I can't test this properly as I don't have any such hardware, but as the code is mosly an edited copy of the interleaved code, any errors should be easy to fix.
2011-09-07 08:00:57 +00:00
if (out_idx++ > out_max)
out_idx = 0;
}
} else {
while (count--) {
Fix some undefined symbols in snd_alsa.c. Serves me right for testing with only my debug build :P
2011-09-08 09:10:09 +00:00
out_0[out_idx] = clamp_16 ((*p++ * volume) * 0x8000);
Support non-interleaved mmap sound. Unfortuanately, I can't test this properly as I don't have any such hardware, but as the code is mosly an edited copy of the interleaved code, any errors should be easy to fix.
2011-09-07 08:00:57 +00:00
p++; // skip right channel
if (out_idx++ > out_max)
out_idx = 0;
}
}
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
} else if (snd->samplebits == 8) {
Support non-interleaved mmap sound. Unfortuanately, I can't test this properly as I don't have any such hardware, but as the code is mosly an edited copy of the interleaved code, any errors should be easy to fix.
2011-09-07 08:00:57 +00:00
byte *out_0 = (byte *) areas[0].addr;
byte *out_1 = (byte *) areas[1].addr;
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
if (snd->channels == 2) {
Support non-interleaved mmap sound. Unfortuanately, I can't test this properly as I don't have any such hardware, but as the code is mosly an edited copy of the interleaved code, any errors should be easy to fix.
2011-09-07 08:00:57 +00:00
while (count--) {
Fix some undefined symbols in snd_alsa.c. Serves me right for testing with only my debug build :P
2011-09-08 09:10:09 +00:00
out_0[out_idx] = clamp_8 ((*p++ * volume) * 0x80);
out_1[out_idx] = clamp_8 ((*p++ * volume) * 0x80);
Support non-interleaved mmap sound. Unfortuanately, I can't test this properly as I don't have any such hardware, but as the code is mosly an edited copy of the interleaved code, any errors should be easy to fix.
2011-09-07 08:00:57 +00:00
if (out_idx++ > out_max)
out_idx = 0;
}
} else {
while (count--) {
Fix some undefined symbols in snd_alsa.c. Serves me right for testing with only my debug build :P
2011-09-08 09:10:09 +00:00
out_0[out_idx] = clamp_8 ((*p++ * volume) * 0x8000);
Support non-interleaved mmap sound. Unfortuanately, I can't test this properly as I don't have any such hardware, but as the code is mosly an edited copy of the interleaved code, any errors should be easy to fix.
2011-09-07 08:00:57 +00:00
p++; // skip right channel
if (out_idx++ > out_max)
out_idx = 0;
}
}
}
}
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
static void
alsa_xfer (snd_t *snd, portable_samplepair_t *paintbuffer, int count,
float volume)
{
int out_idx, out_max, step, val;
float *p;
[audio] Allow output plugins to specify model Output plugins can use either a push model (synchronous) or a pull model (asynchronous). The ALSA plugin now uses the pull model. This paves the way for making jack output a simple output plugin rather than the combined render/output plugin it currently is (for #16) as now snd_dma works with both models.
2021-06-25 02:41:42 +00:00
alsa_pkt_t *packet = snd->xfer_data;;
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
p = (float *) paintbuffer;
count *= snd->channels;
out_max = (snd->frames * snd->channels) - 1;
[audio] Clean up jack and alsa dependencies I had forgotten to test with shared libs and it turns out jack and alsa were directly accessing symbols in the renderer (and in jack's case, linking in a duplicate of the renderer). Fixes #16.
2021-06-26 07:18:18 +00:00
out_idx = snd->paintedtime * snd->channels;
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
while (out_idx > out_max)
out_idx -= out_max + 1;
step = 3 - snd->channels;
if (snd->samplebits == 16) {
[audio] Allow output plugins to specify model Output plugins can use either a push model (synchronous) or a pull model (asynchronous). The ALSA plugin now uses the pull model. This paves the way for making jack output a simple output plugin rather than the combined render/output plugin it currently is (for #16) as now snd_dma works with both models.
2021-06-25 02:41:42 +00:00
short *out = (short *) packet->areas[0].addr;
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
while (count--) {
val = (*p * volume) * 0x8000;
p += step;
if (val > 0x7fff)
val = 0x7fff;
else if (val < -0x8000)
val = -0x8000;
out[out_idx++] = val;
if (out_idx > out_max)
out_idx = 0;
}
} else if (snd->samplebits == 8) {
[audio] Allow output plugins to specify model Output plugins can use either a push model (synchronous) or a pull model (asynchronous). The ALSA plugin now uses the pull model. This paves the way for making jack output a simple output plugin rather than the combined render/output plugin it currently is (for #16) as now snd_dma works with both models.
2021-06-25 02:41:42 +00:00
unsigned char *out = (unsigned char *) packet->areas[0].addr;
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
while (count--) {
val = (*p * volume) * 128;
p += step;
if (val > 0x7f)
val = 0x7f;
else if (val < -0x80)
val = -0x80;
out[out_idx++] = val + 0x80;
if (out_idx > out_max)
out_idx = 0;
}
}
}
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
static int
alsa_recover (snd_pcm_t *pcm, int err)
{
if (err == -EPIPE) {
Sys_Printf ("snd_alsa: xrun\n");
// xrun
if ((err = qfsnd_pcm_prepare (pcm)) < 0) {
Sys_MaskPrintf (SYS_snd, "snd_alsa: recover from xrun failed: %s\n",
qfsnd_strerror (err));
return err;
}
return 0;
} else if (err == -ESTRPIPE) {
Sys_Printf ("snd_alsa: suspend\n");
// suspend
while ((err = qfsnd_pcm_resume(pcm)) == -EAGAIN) {
usleep (20 * 1000);
}
if (err < 0 && (err = qfsnd_pcm_prepare (pcm)) < 0) {
Sys_MaskPrintf (SYS_snd,
"snd_alsa: recover from suspend failed: %s\n",
qfsnd_strerror (err));
return err;
}
return 0;
}
return err;
}
static int
alsa_process (snd_pcm_t *pcm, snd_t *snd)
{
alsa_pkt_t packet;
int res;
int ret = 1;
snd_pcm_uframes_t size = snd->submission_chunk;
snd->xfer_data = &packet;
while (size > 0) {
packet.nframes = size;
if ((res = qfsnd_pcm_mmap_begin (pcm, &packet.areas, &packet.offset,
&packet.nframes)) < 0) {
if ((res = alsa_recover (pcm, -EPIPE)) < 0) {
Sys_Printf ("snd_alsa: XRUN recovery failed: %s\n",
qfsnd_strerror (res));
Fix a pile of warnings for gcc 12 Most were pretty easy and fairly logical, but gib's regex was a bit of a pain until I figured out the real problem was the conditional assignments. However, libs/gamecode/test/test-conv4 fails when optimizing due to gcc using vcvttps2dq (which is nice, actually) for vector forms, but not the single equivalent other times. I haven't decided what to do with the test (I might abandon it as it does seem to be UD).
2022-07-26 12:14:28 +00:00
snd->xfer_data = 0;
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
return res;
}
ret = 0;
}
[audio] Allow output plugins to specify model Output plugins can use either a push model (synchronous) or a pull model (asynchronous). The ALSA plugin now uses the pull model. This paves the way for making jack output a simple output plugin rather than the combined render/output plugin it currently is (for #16) as now snd_dma works with both models.
2021-06-25 02:41:42 +00:00
snd->buffer = packet.areas[0].addr;
[audio] Clean up jack and alsa dependencies I had forgotten to test with shared libs and it turns out jack and alsa were directly accessing symbols in the renderer (and in jack's case, linking in a duplicate of the renderer). Fixes #16.
2021-06-26 07:18:18 +00:00
snd->paint_channels (snd, snd->paintedtime + packet.nframes);
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
if ((res = qfsnd_pcm_mmap_commit (pcm, packet.offset,
packet.nframes)) < 0
|| (snd_pcm_uframes_t) res != packet.nframes) {
if ((res = alsa_recover (pcm, res >= 0 ? -EPIPE : res)) < 0) {
Sys_Printf ("snd_alsa: XRUN recovery failed: %s\n",
qfsnd_strerror (res));
Fix a pile of warnings for gcc 12 Most were pretty easy and fairly logical, but gib's regex was a bit of a pain until I figured out the real problem was the conditional assignments. However, libs/gamecode/test/test-conv4 fails when optimizing due to gcc using vcvttps2dq (which is nice, actually) for vector forms, but not the single equivalent other times. I haven't decided what to do with the test (I might abandon it as it does seem to be UD).
2022-07-26 12:14:28 +00:00
snd->xfer_data = 0;
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
return res;
}
ret = 0;
}
size -= packet.nframes;
}
Fix a pile of warnings for gcc 12 Most were pretty easy and fairly logical, but gib's regex was a bit of a pain until I figured out the real problem was the conditional assignments. However, libs/gamecode/test/test-conv4 fails when optimizing due to gcc using vcvttps2dq (which is nice, actually) for vector forms, but not the single equivalent other times. I haven't decided what to do with the test (I might abandon it as it does seem to be UD).
2022-07-26 12:14:28 +00:00
snd->xfer_data = 0;
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
return ret;
}
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
static void
alsa_callback (snd_async_handler_t *handler)
{
snd_pcm_t *pcm = qfsnd_async_handler_get_pcm (handler);
snd_t *snd = qfsnd_async_handler_get_callback_private (handler);
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
int res;
int avail;
int first = 0;
while (1) {
snd_pcm_state_t state = qfsnd_pcm_state (pcm);
if (state == SND_PCM_STATE_XRUN) {
if ((res = alsa_recover (pcm, -EPIPE)) < 0) {
Sys_Printf ("snd_alsa: XRUN recovery failed: %s\n",
qfsnd_strerror (res));
//FIXME disable/restart sound
return;
}
} else if (state == SND_PCM_STATE_SUSPENDED) {
if ((res = alsa_recover (pcm, -EPIPE)) < 0) {
Sys_Printf ("snd_alsa: suspend recovery failed: %s\n",
qfsnd_strerror (res));
//FIXME disable/restart sound
return;
}
}
if ((avail = qfsnd_pcm_avail_update (pcm)) < 0) {
if ((res = alsa_recover (pcm, -EPIPE)) < 0) {
Sys_Printf ("snd_alsa: avail update failed: %s\n",
qfsnd_strerror (res));
//FIXME disable/restart sound
return;
}
first = 1;
continue;
}
if (avail < snd->submission_chunk) {
if (first) {
first = 0;
if ((res = qfsnd_pcm_start (pcm)) < 0) {
Sys_Printf ("snd_alsa: start failed: %s\n",
qfsnd_strerror (res));
return;
}
continue;
}
break;
}
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
if ((res = alsa_process (pcm, snd))) {
if (res < 0) {
//FIXME disable/restart sound
return;
}
break;
}
first = 1;
}
}
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
static int
alsa_open_playback (snd_t *snd, const char *device)
{
if (!*device) {
device = "default";
}
Sys_Printf ("Using PCM %s.\n", device);
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
int res = qfsnd_pcm_open (&pcm, device, SND_PCM_STREAM_PLAYBACK,
SND_PCM_NONBLOCK);
if (res < 0) {
Sys_Printf ("snd_alsa: audio open error: %s\n", qfsnd_strerror (res));
return 0;
}
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
return 1;
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
}
static int
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
alsa_playback_set_mmap (snd_t *snd, snd_pcm_hw_params_t *hw)
Move sound targets into a library. Still doesn't pick one at runtime though.
2001-04-10 09:40:09 +00:00
{
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
int res;
Move sound targets into a library. Still doesn't pick one at runtime though.
2001-04-10 09:40:09 +00:00
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
res = qfsnd_pcm_hw_params_set_access (pcm, hw,
SND_PCM_ACCESS_MMAP_INTERLEAVED);
if (res == 0) {
snd->xfer = alsa_xfer;
return 1;
}
Sys_MaskPrintf (SYS_snd, "snd_alsa: Failure to set interleaved PCM "
"access. (%d) %s\n", res, qfsnd_strerror (res));
res = qfsnd_pcm_hw_params_set_access (pcm, hw,
SND_PCM_ACCESS_MMAP_NONINTERLEAVED);
if (res == 0) {
snd->xfer = alsa_ni_xfer;
return 1;
}
Sys_MaskPrintf (SYS_snd, "snd_alsa: Failure to set noninterleaved PCM "
"access. (%d) %s\n", res, qfsnd_strerror (res));
Sys_Printf ("snd_alsa: could not set mmap access\n");
the "two birds with one stone" commit 1) when /not/ building static plugins, the plugin info name is, once again, PluginInfo, but is still type_name_PluginInfo for static plugins. This allows plugins to be symlinked (highly desirable for the debian packages, and likely others). This also requires plugins to /not/ be loaded with RTLD_GLOBAL. 2) because of 1, snd_alsa_0_9 has to explicitly load libasound itself. This just happens to fix my segfault on shutdown in RedHat 6.2.
2002-04-19 20:08:54 +00:00
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
return 0;
}
Move sound targets into a library. Still doesn't pick one at runtime though.
2001-04-10 09:40:09 +00:00
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
static int
alsa_playback_set_bps (snd_t *snd, snd_pcm_hw_params_t *hw)
{
int res;
int bps = 0;
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
if (snd_bits == 16) {
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
bps = SND_PCM_FORMAT_S16_LE;
snd->samplebits = 16;
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
} else if (snd_bits == 8) {
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
bps = SND_PCM_FORMAT_U8;
snd->samplebits = 8;
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
} else if (snd_bits) {
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
Sys_Printf ("snd_alsa: invalid sample bits: %d\n", bps);
return 0;
}
if (bps) {
if ((res = qfsnd_pcm_hw_params_set_format (pcm, hw, bps)) == 0) {
return 1;
}
} else {
bps = SND_PCM_FORMAT_S16_LE;
if ((res = qfsnd_pcm_hw_params_set_format (pcm, hw, bps)) == 0) {
snd->samplebits = 16;
return 1;
Move sound targets into a library. Still doesn't pick one at runtime though.
2001-04-10 09:40:09 +00:00
}
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
bps = SND_PCM_FORMAT_U8;
if ((res = qfsnd_pcm_hw_params_set_format (pcm, hw, bps)) == 0) {
snd->samplebits = 8;
return 1;
}
Sys_Printf ("snd_alsa: no usable formats. %s\n", qfsnd_strerror (res));
Move sound targets into a library. Still doesn't pick one at runtime though.
2001-04-10 09:40:09 +00:00
}
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
snd->samplebits = -1;
Sys_Printf ("snd_alsa: desired format not supported\n");
return 0;
}
static int
alsa_playback_set_channels (snd_t *snd, snd_pcm_hw_params_t *hw)
{
int res;
int channels = 1;
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
if (snd_stereo) {
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
channels = 2;
}
if ((res = qfsnd_pcm_hw_params_set_channels (pcm, hw, channels)) == 0) {
snd->channels = channels;
return 1;
Move sound targets into a library. Still doesn't pick one at runtime though.
2001-04-10 09:40:09 +00:00
}
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
Sys_Printf ("snd_alsa: desired channels not supported\n");
return 0;
}
static int
alsa_playback_set_rate (snd_t *snd, snd_pcm_hw_params_t *hw)
{
int res;
unsigned rate = 0;
static int default_rates[] = { 48000, 44100, 22050, 11025, 0 };
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
if (snd_rate) {
rate = snd_rate;
Check return conditions during ALSA setup. Error checking during setup is a good thing...
2004-01-17 22:03:58 +00:00
}
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
if (rate) {
if ((res = qfsnd_pcm_hw_params_set_rate (pcm, hw, rate, 0)) == 0) {
snd->speed = rate;
return 1;
}
Sys_Printf ("snd_alsa: desired rate %i not supported. %s\n", rate,
qfsnd_strerror (res));
} else {
// use default rate
int dir = 0;
for (int *def_rate = default_rates; *def_rate; def_rate++) {
rate = *def_rate;
res = qfsnd_pcm_hw_params_set_rate_near (pcm, hw, &rate, &dir);
if (res == 0) {
snd->speed = rate;
return 1;
Beg for mmap access. If the default sound device does not support mmap access, retry with plughw. However, assume the user knows best and do not retry if snd_device has been set to anything, including "default". QF alsa support now works out of the box with pulseaudio.
2011-12-08 02:32:33 +00:00
}
Support non-interleaved mmap sound. Unfortuanately, I can't test this properly as I don't have any such hardware, but as the code is mosly an edited copy of the interleaved code, any errors should be easy to fix.
2011-09-07 08:00:57 +00:00
}
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
Sys_Printf ("snd_alsa: no usable rate\n");
Move sound targets into a library. Still doesn't pick one at runtime though.
2001-04-10 09:40:09 +00:00
}
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
return 0;
}
Support non-interleaved mmap sound. Unfortuanately, I can't test this properly as I don't have any such hardware, but as the code is mosly an edited copy of the interleaved code, any errors should be easy to fix.
2011-09-07 08:00:57 +00:00
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
static int
alsa_playback_set_period_size (snd_t *snd, snd_pcm_hw_params_t *hw)
{
int res;
snd_pcm_uframes_t period_size;
// works out to about 5.5ms (5.3 for 48k, 5.8 for 44k1) but consistent for
// different sample rates give or take rounding
period_size = 64 * (snd->speed / 11025);
res = qfsnd_pcm_hw_params_set_period_size_near (pcm, hw, &period_size, 0);
if (res == 0) {
// don't mix less than this in frames:
res = qfsnd_pcm_hw_params_get_period_size (hw, &period_size, 0);
if (res == 0) {
snd->submission_chunk = period_size;
return 1;
}
Sys_Printf ("snd_alsa: unable to get period size. %s\n",
qfsnd_strerror (res));
} else {
Sys_Printf ("snd_alsa: unable to set period size near %i. %s\n",
(int) period_size, qfsnd_strerror (res));
Move sound targets into a library. Still doesn't pick one at runtime though.
2001-04-10 09:40:09 +00:00
}
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
return 0;
}
Move sound targets into a library. Still doesn't pick one at runtime though.
2001-04-10 09:40:09 +00:00
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
static int
SNDDMA_Init (snd_t *snd)
{
int res;
[cvar] Make cvars properly typed This is an extremely extensive patch as it hits every cvar, and every usage of the cvars. Cvars no longer store the value they control, instead, they use a cexpr value object to reference the value and specify the value's type (currently, a null type is used for strings). Non-string cvars are passed through cexpr, allowing expressions in the cvars' settings. Also, cvars have returned to an enhanced version of the original (id quake) registration scheme. As a minor benefit, relevant code having direct access to the cvar-controlled variables is probably a slight optimization as it removed a pointer dereference, and the variables can be located for data locality. The static cvar descriptors are made private as an additional safety layer, though there's nothing stopping external modification via Cvar_FindVar (which is needed for adding listeners). While not used yet (partly due to working out the design), cvars can have a validation function. Registering a cvar allows a primary listener (and its data) to be specified: it will always be called first when the cvar is modified. The combination of proper listeners and direct access to the controlled variable greatly simplifies the more complex cvar interactions as much less null checking is required, and there's no need for one cvar's callback to call another's. nq-x11 is known to work at least well enough for the demos. More testing will come.
2022-04-23 03:22:45 +00:00
const char *device = snd_device;
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
snd_pcm_hw_params_t *hw;
snd_pcm_sw_params_t *sw;
if (!load_libasound ())
return false;
snd_pcm_hw_params_alloca (&hw);
snd_pcm_sw_params_alloca (&sw);
while (1) {
if (!alsa_open_playback (snd, device)) {
return 0;
}
if ((res = qfsnd_pcm_hw_params_any (pcm, hw)) < 0) {
Sys_Printf ("snd_alsa: error setting hw_params_any. %s\n",
qfsnd_strerror (res));
goto error;
}
if (alsa_playback_set_mmap (snd, hw)) {
Check return conditions during ALSA setup. Error checking during setup is a good thing...
2004-01-17 22:03:58 +00:00
break;
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
}
if (*device) {
Move sound targets into a library. Still doesn't pick one at runtime though.
2001-04-10 09:40:09 +00:00
goto error;
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
}
qfsnd_pcm_close (pcm);
device = "plughw";
Move sound targets into a library. Still doesn't pick one at runtime though.
2001-04-10 09:40:09 +00:00
}
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
if (!alsa_playback_set_bps (snd, hw)) {
goto error;
Update ALSA sound target to support 1.0. Breaks 0.9 support, but 1.0 is in 2.6.0. So either update, or revert to older snd_alsa.c & alsa_funcs_list.h.
2003-12-22 01:07:49 +00:00
}
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
if (!alsa_playback_set_channels (snd, hw)) {
Check return conditions during ALSA setup. Error checking during setup is a good thing...
2004-01-17 22:03:58 +00:00
goto error;
}
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
if (!alsa_playback_set_rate (snd, hw)) {
Check return conditions during ALSA setup. Error checking during setup is a good thing...
2004-01-17 22:03:58 +00:00
goto error;
}
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
if (!alsa_playback_set_period_size (snd, hw)) {
Check return conditions during ALSA setup. Error checking during setup is a good thing...
2004-01-17 22:03:58 +00:00
goto error;
}
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
if ((res = qfsnd_pcm_hw_params (pcm, hw)) < 0) {
Sys_Printf ("snd_alsa: unable to install hw params: %s\n",
qfsnd_strerror (res));
Check return conditions during ALSA setup. Error checking during setup is a good thing...
2004-01-17 22:03:58 +00:00
goto error;
}
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
if ((res = qfsnd_pcm_sw_params_current (pcm, sw)) < 0) {
Sys_Printf ("snd_alsa: unable to determine current sw params. %s\n",
qfsnd_strerror (res));
Move sound targets into a library. Still doesn't pick one at runtime though.
2001-04-10 09:40:09 +00:00
goto error;
}
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
if ((res = qfsnd_pcm_sw_params_set_start_threshold (pcm, sw, ~0U)) < 0) {
Sys_Printf ("snd_alsa: unable to set playback threshold. %s\n",
qfsnd_strerror (res));
Move sound targets into a library. Still doesn't pick one at runtime though.
2001-04-10 09:40:09 +00:00
goto error;
}
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
if ((res = qfsnd_pcm_sw_params_set_stop_threshold (pcm, sw, ~0U)) < 0) {
Sys_Printf ("snd_alsa: unable to set playback stop threshold. %s\n",
qfsnd_strerror (res));
goto error;
}
if ((res = qfsnd_pcm_sw_params (pcm, sw)) < 0) {
Sys_Printf ("snd_alsa: unable to install sw params. %s\n",
qfsnd_strerror (res));
Check return conditions during ALSA setup. Error checking during setup is a good thing...
2004-01-17 22:03:58 +00:00
goto error;
}
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
snd->framepos = 0;
Check return conditions during ALSA setup. Error checking during setup is a good thing...
2004-01-17 22:03:58 +00:00
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
if ((res = qfsnd_pcm_hw_params_get_buffer_size (hw, &buffer_size)) < 0) {
Sys_Printf ("snd_alsa: unable to get buffer size. %s\n",
qfsnd_strerror (res));
Check return conditions during ALSA setup. Error checking during setup is a good thing...
2004-01-17 22:03:58 +00:00
goto error;
}
attempt to fix choppy audio in alsa caused by non-power-of 2 buffer size
2005-06-14 05:37:26 +00:00
if (buffer_size != round_buffer_size (buffer_size)) {
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
Sys_Printf ("snd_alsa: WARNING: non-power of 2 buffer size. sound may be unsatisfactory\n");
can't adjust dmix's buffer size, so print a message when it seems it's likely to be an issue
2005-08-12 02:39:37 +00:00
Sys_Printf ("recommend using either the plughw, or hw devices or adjusting dmix\n");
Sys_Printf ("to have a power of 2 buffer size\n");
attempt to fix choppy audio in alsa caused by non-power-of 2 buffer size
2005-06-14 05:37:26 +00:00
}
Check return conditions during ALSA setup. Error checking during setup is a good thing...
2004-01-17 22:03:58 +00:00
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
if ((res = qfsnd_async_add_pcm_handler (&async_handler, pcm,
alsa_callback, snd)) < 0) {
Sys_Printf ("snd_alsa: unable to register async handler: %s",
qfsnd_strerror (res));
goto error;
}
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
snd->frames = buffer_size;
[audio] Allow output plugins to specify model Output plugins can use either a push model (synchronous) or a pull model (asynchronous). The ALSA plugin now uses the pull model. This paves the way for making jack output a simple output plugin rather than the combined render/output plugin it currently is (for #16) as now snd_dma works with both models.
2021-06-25 02:41:42 +00:00
[sound] Add a threaded field to snd_t Care needs to be taken when freeing channels as doing so while an asynchronous mixer is using them is unlikely to end well. However, whether the mixer is asynchronous depends on the output driver. This lets the driver inform the rest of the system that the output and mixer are running asynchronously.
2022-06-05 09:08:49 +00:00
snd->threaded = 1;//XXX FIXME double check whether it's always true
[audio] Allow output plugins to specify model Output plugins can use either a push model (synchronous) or a pull model (asynchronous). The ALSA plugin now uses the pull model. This paves the way for making jack output a simple output plugin rather than the combined render/output plugin it currently is (for #16) as now snd_dma works with both models.
2021-06-25 02:41:42 +00:00
// send the first period to fill the buffer
// also sets snd->buffer
if (alsa_process (pcm, snd) < 0) {
goto error;
}
qfsnd_pcm_start (pcm);
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
Sys_Printf ("%5d channels %sinterleaved\n", snd->channels,
snd->xfer ? "non-" : "");
Sys_Printf ("%5d samples (%.1fms)\n", snd->frames,
1000.0 * snd->frames / snd->speed);
Sys_Printf ("%5d samplepos\n", snd->framepos);
Sys_Printf ("%5d samplebits\n", snd->samplebits);
Sys_Printf ("%5d submission_chunk (%.1fms)\n", snd->submission_chunk,
1000.0 * snd->submission_chunk / snd->speed);
Sys_Printf ("%5d speed\n", snd->speed);
Sys_Printf ("0x%lx dma buffer\n", (long) snd->buffer);
Move sound targets into a library. Still doesn't pick one at runtime though.
2001-04-10 09:40:09 +00:00
snd_inited = 1;
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
return 1;
Lots of whitespace.
2001-08-22 20:26:25 +00:00
error:
the "two birds with one stone" commit 1) when /not/ building static plugins, the plugin info name is, once again, PluginInfo, but is still type_name_PluginInfo for static plugins. This allows plugins to be symlinked (highly desirable for the debian packages, and likely others). This also requires plugins to /not/ be loaded with RTLD_GLOBAL. 2) because of 1, snd_alsa_0_9 has to explicitly load libasound itself. This just happens to fix my segfault on shutdown in RedHat 6.2.
2002-04-19 20:08:54 +00:00
qfsnd_pcm_close (pcm);
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
snd->channels = 0;
snd->frames = 0;
snd->samplebits = 0;
snd->submission_chunk = 0;
snd->speed = 0;
Move sound targets into a library. Still doesn't pick one at runtime though.
2001-04-10 09:40:09 +00:00
return 0;
}
get alsa0_9 working with latest cvs, we need to dlopen the plugin with RTDL_GLOBAL. To use RTDL_GLOBAL, we need to make as much static as possible and PluginInfo needs a unique name (%s_%s_PluginInfo, type, name).
2001-08-31 00:05:58 +00:00
static void
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
SNDDMA_shutdown (snd_t *snd)
Move sound targets into a library. Still doesn't pick one at runtime though.
2001-04-10 09:40:09 +00:00
{
if (snd_inited) {
[audio] Clean up alsa init and add error checking This gets the alsa target working nicely for mmapped outout. I'm not certain, but I think it will even deal with NPOT buffer sizes (I copied the code from libasound's sample pcm.c, thus the uncertainty). Non-mmapped output isn't supported yet, but the alsa target now works nicely for pull rendering. However, some work still needs to be done for recovery failure: either disable the sound system, or restart the driver entirely (preferable).
2021-06-25 00:50:31 +00:00
qfsnd_async_del_handler (async_handler);
async_handler = 0;
the "two birds with one stone" commit 1) when /not/ building static plugins, the plugin info name is, once again, PluginInfo, but is still type_name_PluginInfo for static plugins. This allows plugins to be symlinked (highly desirable for the debian packages, and likely others). This also requires plugins to /not/ be loaded with RTLD_GLOBAL. 2) because of 1, snd_alsa_0_9 has to explicitly load libasound itself. This just happens to fix my segfault on shutdown in RedHat 6.2.
2002-04-19 20:08:54 +00:00
qfsnd_pcm_close (pcm);
Move sound targets into a library. Still doesn't pick one at runtime though.
2001-04-10 09:40:09 +00:00
snd_inited = 0;
}
}
get alsa0_9 working with latest cvs, we need to dlopen the plugin with RTDL_GLOBAL. To use RTDL_GLOBAL, we need to make as much static as possible and PluginInfo needs a unique name (%s_%s_PluginInfo, type, name).
2001-08-31 00:05:58 +00:00
static void
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
SNDDMA_BlockSound (snd_t *snd)
Put the sound renderer in it's proper place: as a full-fledged plugin. Bugs expected. Please report them :)
2001-08-23 04:01:46 +00:00
{
fix the crash on focus loss when using in_snd_block and -nosound
2002-09-30 03:59:55 +00:00
if (snd_inited && ++snd_blocked == 1)
the "two birds with one stone" commit 1) when /not/ building static plugins, the plugin info name is, once again, PluginInfo, but is still type_name_PluginInfo for static plugins. This allows plugins to be symlinked (highly desirable for the debian packages, and likely others). This also requires plugins to /not/ be loaded with RTLD_GLOBAL. 2) because of 1, snd_alsa_0_9 has to explicitly load libasound itself. This just happens to fix my segfault on shutdown in RedHat 6.2.
2002-04-19 20:08:54 +00:00
qfsnd_pcm_pause (pcm, 1);
Put the sound renderer in it's proper place: as a full-fledged plugin. Bugs expected. Please report them :)
2001-08-23 04:01:46 +00:00
}
get alsa0_9 working with latest cvs, we need to dlopen the plugin with RTDL_GLOBAL. To use RTDL_GLOBAL, we need to make as much static as possible and PluginInfo needs a unique name (%s_%s_PluginInfo, type, name).
2001-08-31 00:05:58 +00:00
static void
[audio] Rework alsa to use a pull model This brings the alsa driver in line with the jack render (progress towards #16), but breaks most of the other drivers (for now: one step at a time). The idea is that once the pull model is working for at least one other target, the jack renderer can become just another target like it should have been in the first place (but I needed to get the pull model working first, then forgot about it). Correct state checking is not done yet, but testsound does produce what seems to be fairly good sound when it starts up correctly (part of the state checking (or lack thereof), I imagine).
2021-06-23 15:04:02 +00:00
SNDDMA_UnblockSound (snd_t *snd)
Put the sound renderer in it's proper place: as a full-fledged plugin. Bugs expected. Please report them :)
2001-08-23 04:01:46 +00:00
{
fix the crash on focus loss when using in_snd_block and -nosound
2002-09-30 03:59:55 +00:00
if (!snd_inited || !snd_blocked)
Put the sound renderer in it's proper place: as a full-fledged plugin. Bugs expected. Please report them :)
2001-08-23 04:01:46 +00:00
return;
if (!--snd_blocked)
the "two birds with one stone" commit 1) when /not/ building static plugins, the plugin info name is, once again, PluginInfo, but is still type_name_PluginInfo for static plugins. This allows plugins to be symlinked (highly desirable for the debian packages, and likely others). This also requires plugins to /not/ be loaded with RTLD_GLOBAL. 2) because of 1, snd_alsa_0_9 has to explicitly load libasound itself. This just happens to fix my segfault on shutdown in RedHat 6.2.
2002-04-19 20:08:54 +00:00
qfsnd_pcm_pause (pcm, 0);
Put the sound renderer in it's proper place: as a full-fledged plugin. Bugs expected. Please report them :)
2001-08-23 04:01:46 +00:00
}
[audio] Allow output plugins to specify model Output plugins can use either a push model (synchronous) or a pull model (asynchronous). The ALSA plugin now uses the pull model. This paves the way for making jack output a simple output plugin rather than the combined render/output plugin it currently is (for #16) as now snd_dma works with both models.
2021-06-25 02:41:42 +00:00
static general_data_t plugin_info_general_data = {
};
static general_funcs_t plugin_info_general_funcs = {
.init = SNDDMA_Init_Cvars,
.shutdown = NULL,
};
static snd_output_data_t plugin_info_snd_output_data = {
.model = som_pull,
};
static snd_output_funcs_t plugin_info_snd_output_funcs = {
.init = SNDDMA_Init,
.shutdown = SNDDMA_shutdown,
.block_sound = SNDDMA_BlockSound,
.unblock_sound = SNDDMA_UnblockSound,
};
static plugin_data_t plugin_info_data = {
.general = &plugin_info_general_data,
.snd_output = &plugin_info_snd_output_data,
};
static plugin_funcs_t plugin_info_funcs = {
.general = &plugin_info_general_funcs,
.snd_output = &plugin_info_snd_output_funcs,
};
static plugin_t plugin_info = {
.type = qfp_snd_output,
.api_version = QFPLUGIN_VERSION,
.plugin_version = "0.1",
.description = "ALSA digital output",
.copyright = "Copyright (C) 1996-1997 id Software, Inc.\n"
Lots of whitespace, header cleanups, and a few tiny optimizations.
2001-09-09 21:41:35 +00:00
"Copyright (C) 1999,2000,2001 contributors of the QuakeForge "
"project\n"
[audio] Allow output plugins to specify model Output plugins can use either a push model (synchronous) or a pull model (asynchronous). The ALSA plugin now uses the pull model. This paves the way for making jack output a simple output plugin rather than the combined render/output plugin it currently is (for #16) as now snd_dma works with both models.
2021-06-25 02:41:42 +00:00
"Please see the file \"AUTHORS\" for a list of contributors",
.functions = &plugin_info_funcs,
.data = &plugin_info_data,
};
Sound now uses runtime plugins. Much thanks to Deek for doing all the thinking :) set fs_pluginpath to point to the right dir, and set snd_plugin to pick a sound plugin. Current issues: - alsa 0.5 won't build properly, dunno why - segfault on exit. I think I know the cause of this, and how to fix it - alsa 0.9, gus, sgi, sun, and win32 havn't been tested
2001-05-19 00:05:35 +00:00
[audio] Allow output plugins to specify model Output plugins can use either a push model (synchronous) or a pull model (asynchronous). The ALSA plugin now uses the pull model. This paves the way for making jack output a simple output plugin rather than the combined render/output plugin it currently is (for #16) as now snd_dma works with both models.
2021-06-25 02:41:42 +00:00
PLUGIN_INFO(snd_output, alsa)
{
Lots of whitespace, header cleanups, and a few tiny optimizations.
2001-09-09 21:41:35 +00:00
return &plugin_info;
Sound now uses runtime plugins. Much thanks to Deek for doing all the thinking :) set fs_pluginpath to point to the right dir, and set snd_plugin to pick a sound plugin. Current issues: - alsa 0.5 won't build properly, dunno why - segfault on exit. I think I know the cause of this, and how to fix it - alsa 0.9, gus, sgi, sun, and win32 havn't been tested
2001-05-19 00:05:35 +00:00
}
Reference in a new issue Copy permalink