- Alsa-devel - mailman.alsa-project.org

[alsa-devel] Buffering issues 2
by Lennart Poettering 16 May '08

16 May '08

Heya! In addition to the buffering issues pointed out on http://mailman.alsa-project.org/pipermail/alsa-devel/2008-April/007354.html I now have some more issues with the way buffering works in ALSA. This time I am pretty sure it's a bug in the HDA drivers. PulseAudio basically does this (again, as part of the glitch-free playback model, see http://0pointer.de/blog/projects/pulse-glitch-free.html): for (;;) { for (;;) { snd_pcm_hwsync(); n = snd_pcm_avail_update(); if (!n) break; fill_up(n); } usleep(buffer_time - 20ms); } The buffer time is 370ms, so we usually sleep for 350ms. It's a lot more complex actually, due to deviating timing and some code that we don't enter a busy loop if the CPU is slower then the sound card and so on and so on. However, basically it's this algorithm. This works fine on USB. However on HDA it sometimes happens that snd_pcm_avail_update() is not properly updated at the right time. Instead it seems to return what was current one iteration earlier. I.e. we fill up the hw buffer until _avail_update() returns 0. Then assuming that it is fully filled up we go to sleep for 350ms. When we wake up we query _avail_update() again and it will immediately return 0. We thus don't write anything to the device. Instead we go to sleep for another 350ms. Of course, when we wake up again the buffer will already have underrun (since 700 ms passed since the last write to the buffer and th buffer is only 370ms long) and we have a problem. In short: _avail_update() told us that everything was alright and we trusted it and it lied to us. The happens to be configured to 2 periods. So even if _avail_update() has some kind of period granularity (?) it should always have told us after we come back from the sleeping that at least *one* period is free. But it said 0 frames. Nada. Rien. Nichts. Kaputt. I *am* calling snd_pcm_hwsync(). So I'd assume that _avail_update() would be as up to date as it gets. But it isn't. :-( This happens regardless if I open the device as "front:0" (i.e. with softvol in line) or as "hw:0" (no plugins). This doesn't always happen. It seems to depend on the machine how often this happens. On one machine I can reliably reproduce this after 30 iterations. Lennart -- Lennart Poettering Red Hat, Inc. lennart [at] poettering [dot] net ICQ# 11060553 http://0pointer.net/lennart/ GnuPG 0x1A015CC4

3 5

Re: [alsa-devel] Intel hda ICH7 not worked anymore after using wine (SOLVED!). Problems again
by Xavi de Blas 16 May '08

16 May '08

Hello, we have been talking on this thread: http://mailman.alsa-project.org/pipermail/alsa-devel/2007-July/001939.html because i had a problem with my hda that Takashi said it's really specific to my hardware: http://mailman.alsa-project.org/pipermail/alsa-devel/2007-July/001963.html Now with kernel 2.6.24-16 and alsa 1.0.16, i have no audio again: i tried now using the patch Tobin made: http://mailman.alsa-project.org/pipermail/alsa-devel/2007-July/001934.html and it' doesn't work. In fact, i haven't used the patch, i only added the line: { .id = 0x51d41981, .name = "AD1981", .patch = patch_ad1981 }, after the line: { .id = 0x11d41981, .name = "AD1981", .patch = patch_ad1981 }, on: /usr/local/src/alsa-driver-1.0.16/alsa-kernel/pci/hda/patch_analog.c ./configure, make, sudo make install on reboot, the error is the same as before: dmesg|grep alsa [ 41.284537] ALSA /build/buildd/linux-ubuntu-modules-2.6.24-2.6.24/debian/build/build-generic/sound/alsa-driver/pci/hda/hda_codec.c:230: hda_codec: invalid dep_range_val 53:1a here is my /proc/asound/card0/codec#0 http://pastebin.ca/1018596 Thanks a lot 2007/7/9 Xavi de Blas <xaviblas(a)gmail.com>: > It worked!!! > > Thanks a lot > > I thought my hardware was broken! > > 2007/7/9, Tobin Davis <tdavis(a)dsl-only.net>: >> >> Patch needs to be applied as follows: >> >> cd alsa-driver-1.0.14/alsa-kernel >> patch -p1 < LenovoX60S.patch >> cd .. >> >> Then you can rerun configure && make && make install. >> >> Tobin >> >> On Mon, 2007-07-09 at 17:25 +0200, Xavi de Blas wrote: >> >> 2007/7/9, Tobin Davis <tdavis(a)dsl-only.net>: >> > Sorry for missing this thread. I just woke up. >> > >> > I looked at your first pastebin output, and the actual audio codec shows >> > up as 0x51d41981, when it should be 0x11d41981. This is why the Analog >> > driver patch isn't detecting it. >> > >> > Can you try this patch against alsa-driver-1.0.14? It should detect >> > this codec. Still not sure why it is misconfigured, unless there is a >> > bios issue. >> > >> >> I'm afraid, probably when i tried the Linuxant codecs for the modem it >> broke the >> bios. Not sure, but sound worked, and continued working some reboots >> later. But when i tried a few things with wine and rebooted, sound >> never worked again >> >> When i aply the patch (probably i don't do it good) there are problems: >> >> >> xavier@corall:/usr/local/src/alsa-src/alsa-driver-1.0.14$ cat >> LenovoX60S.patch | patch -p1 >> patching file pci/hda/patch_analog.c >> Hunk #1 FAILED at 3402. >> 1 out of 1 hunk FAILED -- saving rejects to file >> pci/hda/patch_analog.c.rej >> >> >> xavier@corall:/usr/local/src/alsa-src/alsa-driver-1.0.14/pci/hda$ cat >> patch_analog.c.rej >> *************** struct hda_codec_preset snd_hda_preset_a >> *** 3402,3407 **** >> { .id = 0x11d41882, .name = "AD1882", .patch = patch_ad1882 }, >> { .id = 0x11d41884, .name = "AD1884", .patch = patch_ad1884 }, >> { .id = 0x11d41981, .name = "AD1981", .patch = patch_ad1981 }, >> { .id = 0x11d41983, .name = "AD1983", .patch = patch_ad1983 }, >> { .id = 0x11d41984, .name = "AD1984", .patch = patch_ad1984 }, >> { .id = 0x11d41986, .name = "AD1986A", .patch = patch_ad1986a }, >> --- 3402,3408 ---- >> { .id = 0x11d41882, .name = "AD1882", .patch = patch_ad1882 }, >> { .id = 0x11d41884, .name = "AD1884", .patch = patch_ad1884 }, >> { .id = 0x11d41981, .name = "AD1981", .patch = patch_ad1981 }, >> + { .id = 0x51d41981, .name = "AD1981", .patch = patch_ad1981 }, >> { .id = 0x11d41983, .name = "AD1983", .patch = patch_ad1983 }, >> { .id = 0x11d41984, .name = "AD1984", .patch = patch_ad1984 }, >> { .id = 0x11d41986, .name = "AD1986A", .patch = patch_ad1986a }, >> >> >> The patch is ok? >> >> Thanks >> >> (Tobin sorry for sending messge two times) >> >> -- >> Tobin Davis <tdavis(a)dsl-only.net> >

2 3

[alsa-devel] [PATCH] PCI168 snd-azt3328 Linux driver: another huge update
by Andreas Mohr 16 May '08

16 May '08

Hello all (and especially my fellow driver users!), ok, color me crazy, but somehow I really felt like doing another round of Undocumented Hacking (tm) again, with my nice sound card, after an overly long total inactivity. - figured out "Digital(ly) Enhanced Game Port" functionality, implemented support for it (eliminating gameport polling overhead) - removed optional joystick activation, gameport now enabled unconditionally, since we now support it via the PCI I/O space, not via conflict-prone legacy I/O (which I was thus able to DISABLE now)! - fix playback bug (a muted wave output would get unmuted upon start of playback, of course this is not what we want, thus remember mute state) - implement partial power management: when idle, lower clock rate and disable codec (reduced noise!), and disable gameport circuit when unused - instantiate OPL3 timer, too - much better implementation of snd_azf3328_mixer_write_volume_gradually() - slightly optimized interrupt handling - lots of cleanup This time, I also found a way to verify proper OPL3 operation via MIDI file playback (emulation via synth hardware). Patch done and hardware-tested against 2.6.26-rc1, which should hopefully match against current ALSA. Everything checked against checkpatch.pl (lots of corrections there as well). Oh, and please don't ask me to split up everything again, lest I'll kill myself - way too many changes this time, and frankly I don't care whether it's debuggable or not since probably not too many people use this driver anyway and I'll fix any problems if/when they turn up (I'm still quite actively enjoying this card myself, thank you very much ;). Note that the new version has an object size increase of merely about 500 bytes, despite all these improvements! Given these additions, the azt3328 driver should now easily have managed to make the PCI168 the best-supported fully undocumented soundcard in the Linux universe ;) (one could say that this card is "a wee bit too old" already, but it's giving me a very nice peace of mind to know that it's almost fully supported now, and that it's now a prime example of really good Linux hardware support) Signed-off-by: Andreas Mohr <andi(a)lisas.de> --- linux-2.6.26-rc1.orig/sound/pci/azt3328.c 2008-05-03 20:59:44.000000000 +0200 +++ linux-2.6.26-rc1/sound/pci/azt3328.c 2008-05-14 22:47:15.000000000 +0200 @@ -1,6 +1,6 @@ /* * azt3328.c - driver for Aztech AZF3328 based soundcards (e.g. PCI168). - * Copyright (C) 2002, 2005, 2006, 2007 by Andreas Mohr <andi AT lisas.de> + * Copyright (C) 2002, 2005 - 2008 by Andreas Mohr <andi AT lisas.de> * * Framework borrowed from Bart Hartgers's als4000.c. * Driver developed on PCI168 AP(W) version (PCI rev. 10, subsystem ID 1801), @@ -35,9 +35,20 @@ * (3 weeks' worth of evenings filled with driver work). * (and no, I did NOT go the easy way: to pick up a SB PCI128 for 9 Euros) * + * It is quite likely that the AZF3328 chip is the PCI cousin of the + * AZF3318 ("azt1020 pnp", "MM Pro 16") ISA chip, given very similar specs. + * * The AZF3328 chip (note: AZF3328, *not* AZT3328, that's just the driver name - * for compatibility reasons) has the following features: + * for compatibility reasons) from Azfin (joint-venture of Aztech and Fincitec, + * Fincitec acquired by National Semiconductor in 2002, together with the + * Fincitec-related company ARSmikro) has the following features: * + * - compatibility & compliance: + * - Microsoft PC 97 ("PC 97 Hardware Design Guide", + * http://www.microsoft.com/whdc/archive/pcguides.mspx) + * - Microsoft PC 98 Baseline Audio + * - MPU401 UART + * - Sound Blaster Emulation (DOS Box) * - builtin AC97 conformant codec (SNR over 80dB) * Note that "conformant" != "compliant"!! this chip's mixer register layout * *differs* from the standard AC97 layout: @@ -48,21 +59,28 @@ * addresses illegally. So far unfortunately it looks like the very flexible * ALSA AC97 support is still not enough to easily compensate for such a * grave layout violation despite all tweaks and quirks mechanisms it offers. - * - builtin genuine OPL3 + * - builtin genuine OPL3 - verified to work fine, 20080506 * - full duplex 16bit playback/record at independent sampling rate - * - MPU401 (+ legacy address support) FIXME: how to enable legacy addr?? + * - MPU401 (+ legacy address support, claimed by one official spec sheet) + * FIXME: how to enable legacy addr?? * - game port (legacy address support) - * - builtin 3D enhancement (said to be YAMAHA Ymersion) * - builtin DirectInput support, helps reduce CPU overhead (interrupt-driven - * features supported) + * features supported). - See common term "Digital Enhanced Game Port"... + * (probably DirectInput 3.0 spec - confirm) + * - builtin 3D enhancement (said to be YAMAHA Ymersion) * - built-in General DirectX timer having a 20 bits counter * with 1us resolution (see below!) - * - I2S serial port for external DAC + * - I2S serial output port for external DAC * - supports 33MHz PCI spec 2.1, PCI power management 1.0, compliant with ACPI * - supports hardware volume control * - single chip low cost solution (128 pin QFP) * - supports programmable Sub-vendor and Sub-system ID * required for Microsoft's logo compliance (FIXME: where?) + * At least the Trident 4D Wave DX has one bit somewhere + * to enable writes to PCI subsystem VID registers, that should be it. + * This might easily be in extended PCI reg space, since PCI168 also has + * some custom data starting at 0x80. What kind of config settings + * are located in our extended PCI space anyway?? * - PCI168 AP(W) card: power amplifier with 4 Watts/channel at 4 Ohms * * Note that this driver now is actually *better* than the Windows driver, @@ -74,6 +92,23 @@ * - "timidity -iAv -B2,8 -Os -EFreverb=0" * - "pmidi -p 128:0 jazz.mid" * + * OPL3 hardware playback testing, try something like: + * cat /proc/asound/hwdep + * and + * aconnect -o + * Then use + * sbiload -Dhw:x,y --opl3 /usr/share/sounds/opl3/std.o3 ......./drums.o3 + * where x,y is the xx-yy number as given in hwdep. + * Then try + * pmidi -p 21:0 jazz.mid + * Oh, and make sure to unmute the FM mixer control (doh!) + * NOTE: power use during OPL3 playback is _VERY_ high (70W --> 90W!) + * despite no CPU activity, possibly due to hindering ACPI idling somehow. + * Shouldn't be a problem of the AZF3328 chip itself, I'd hope. + * Higher PCM / FM mixer levels seem to conflict (causes crackling), + * at least sometimes. Maybe even use with hardware sequencer timer above :) + * adplay/adplug-utils might soon offer hardware-based OPL3 playback, too. + * * Certain PCI versions of this card are susceptible to DMA traffic underruns * in some systems (resulting in sound crackling/clicking/popping), * probably because they don't have a DMA FIFO buffer or so. @@ -87,6 +122,8 @@ * better than a VIA, yet ironically I still get crackling, like many other * people with the same chipset. * Possible remedies: + * - use speaker (amplifier) output instead of headphone output + * (in case crackling is due to overloaded output clipping) * - plug card into a different PCI slot, preferrably one that isn't shared * too much (this helps a lot, but not completely!) * - get rid of PCI VGA card, use AGP instead @@ -94,18 +131,23 @@ * - fiddle with PCI latency settings (setpci -v -s BUSID latency_timer=XX) * Not too helpful. * - Disable ACPI/power management/"Auto Detect RAM/PCI Clk" in BIOS - * + * * BUGS * - full-duplex might *still* be problematic, not fully tested recently * - (non-bug) "Bass/Treble or 3D settings don't work" - they do get evaluated * if you set PCM output switch to "pre 3D" instead of "post 3D". * If this can't be set, then get a mixer application that Isn't Stupid (tm) * (e.g. kmix, gamix) - unfortunately several are!! - * + * - locking is not entirely clean, especially the audio stream activity + * ints --> may be racy + * - an _unconnected_ secondary joystick at the gameport will be reported + * to be "active" (floating values, not precisely -1) due to the way we need + * to read the Digital Enhanced Game Port. Not sure whether it is fixable. + * * TODO * - test MPU401 MIDI playback etc. - * - add some power micro-management (disable various units of the card - * as long as they're unused). However this requires I/O ports which I + * - add more power micro-management (disable various units of the card + * as long as they're unused). However this requires more I/O ports which I * haven't figured out yet and which thus might not even exist... * The standard suspend/resume functionality could probably make use of * some improvement, too... @@ -113,6 +155,7 @@ * - figure out some cleverly evil scheme to possibly make ALSA AC97 code * fully accept our quite incompatible ""AC97"" mixer and thus save some * code (but I'm not too optimistic that doing this is possible at all) + * - use MMIO (memory-mapped I/O)? Slightly faster access, e.g. for gameport. */ #include <asm/io.h> @@ -138,7 +181,7 @@ MODULE_SUPPORTED_DEVICE("{{Aztech,AZF3328}}"); #if defined(CONFIG_GAMEPORT) || (defined(MODULE) && defined(CONFIG_GAMEPORT_MODULE)) -#define SUPPORT_JOYSTICK 1 +#define SUPPORT_GAMEPORT 1 #endif #define DEBUG_MISC 0 @@ -147,13 +190,14 @@ #define DEBUG_PLAY_REC 0 #define DEBUG_IO 0 #define DEBUG_TIMER 0 +#define DEBUG_GAME 0 #define MIXER_TESTING 0 #if DEBUG_MISC #define snd_azf3328_dbgmisc(format, args...) printk(KERN_ERR format, ##args) #else #define snd_azf3328_dbgmisc(format, args...) -#endif +#endif #if DEBUG_CALLS #define snd_azf3328_dbgcalls(format, args...) printk(format, ##args) @@ -163,25 +207,31 @@ #define snd_azf3328_dbgcalls(format, args...) #define snd_azf3328_dbgcallenter() #define snd_azf3328_dbgcallleave() -#endif +#endif #if DEBUG_MIXER #define snd_azf3328_dbgmixer(format, args...) printk(format, ##args) #else #define snd_azf3328_dbgmixer(format, args...) -#endif +#endif #if DEBUG_PLAY_REC #define snd_azf3328_dbgplay(format, args...) printk(KERN_ERR format, ##args) #else #define snd_azf3328_dbgplay(format, args...) -#endif +#endif #if DEBUG_MISC #define snd_azf3328_dbgtimer(format, args...) printk(KERN_ERR format, ##args) #else #define snd_azf3328_dbgtimer(format, args...) -#endif +#endif + +#if DEBUG_GAME +#define snd_azf3328_dbggame(format, args...) printk(KERN_ERR format, ##args) +#else +#define snd_azf3328_dbggame(format, args...) +#endif static int index[SNDRV_CARDS] = SNDRV_DEFAULT_IDX; /* Index 0-MAX */ module_param_array(index, int, NULL, 0444); @@ -195,39 +245,44 @@ module_param_array(enable, bool, NULL, 0444); MODULE_PARM_DESC(enable, "Enable AZF3328 soundcard."); -#ifdef SUPPORT_JOYSTICK -static int joystick[SNDRV_CARDS]; -module_param_array(joystick, bool, NULL, 0444); -MODULE_PARM_DESC(joystick, "Enable joystick for AZF3328 soundcard."); -#endif - static int seqtimer_scaling = 128; module_param(seqtimer_scaling, int, 0444); MODULE_PARM_DESC(seqtimer_scaling, "Set 1024000Hz sequencer timer scale factor (lockup danger!). Default 128."); +typedef struct { + struct snd_pcm_substream *substream; + int enabled; + int running; + unsigned long portbase; +} snd_azf3328_audio_stream_t; + +typedef enum { + AZF_PLAYBACK = 0, + AZF_CAPTURE = 1, +} snd_azf3328_stream_index_t; + struct snd_azf3328 { /* often-used fields towards beginning, then grouped */ - unsigned long codec_port; - unsigned long io2_port; - unsigned long mpu_port; - unsigned long synth_port; - unsigned long mixer_port; + + unsigned long codec_io; /* usually 0xb000, size 128 */ + unsigned long game_io; /* usually 0xb400, size 8 */ + unsigned long mpu_io; /* usually 0xb800, size 4 */ + unsigned long opl3_io; /* usually 0xbc00, size 8 */ + unsigned long mixer_io; /* usually 0xc000, size 64 */ spinlock_t reg_lock; struct snd_timer *timer; - + struct snd_pcm *pcm; - struct snd_pcm_substream *playback_substream; - struct snd_pcm_substream *capture_substream; - unsigned int is_playing; - unsigned int is_recording; + snd_azf3328_audio_stream_t audio_stream[2]; struct snd_card *card; struct snd_rawmidi *rmidi; -#ifdef SUPPORT_JOYSTICK +#ifdef SUPPORT_GAMEPORT struct gameport *gameport; + int axes[4]; #endif struct pci_dev *pci; @@ -237,9 +292,9 @@ /* register value containers for power management * Note: not always full I/O range preserved (just like Win driver!) */ u16 saved_regs_codec [AZF_IO_SIZE_CODEC_PM / 2]; - u16 saved_regs_io2 [AZF_IO_SIZE_IO2_PM / 2]; + u16 saved_regs_game [AZF_IO_SIZE_GAME_PM / 2]; u16 saved_regs_mpu [AZF_IO_SIZE_MPU_PM / 2]; - u16 saved_regs_synth[AZF_IO_SIZE_SYNTH_PM / 2]; + u16 saved_regs_opl3[AZF_IO_SIZE_OPL3_PM / 2]; u16 saved_regs_mixer[AZF_IO_SIZE_MIXER_PM / 2]; #endif }; @@ -252,126 +307,181 @@ MODULE_DEVICE_TABLE(pci, snd_azf3328_ids); + +static int +snd_azf3328_io_reg_setb(unsigned reg, u8 mask, int do_set) +{ + u8 prev = inb(reg), new; + + new = (do_set) ? prev|mask : prev & ~mask; + /* we need to always write the new value no matter whether it differs or not, + * since some register bits don't indicate their setting */ + outb(new, reg); + if (new != prev) + return 1; + + return 0; +} + +static int +snd_azf3328_io_reg_setw(unsigned reg, u16 mask, int do_set) +{ + u16 prev = inw(reg), new; + + new = (do_set) ? prev|mask : prev & ~mask; + /* we need to always write the new value no matter whether it differs or not, + * since some register bits don't indicate their setting */ + outw(new, reg); + if (new != prev) + return 1; + + return 0; +} + static inline void -snd_azf3328_codec_outb(const struct snd_azf3328 *chip, int reg, u8 value) +snd_azf3328_codec_outb(const struct snd_azf3328 *chip, unsigned reg, u8 value) { - outb(value, chip->codec_port + reg); + outb(value, chip->codec_io + reg); } static inline u8 -snd_azf3328_codec_inb(const struct snd_azf3328 *chip, int reg) +snd_azf3328_codec_inb(const struct snd_azf3328 *chip, unsigned reg) { - return inb(chip->codec_port + reg); + return inb(chip->codec_io + reg); } static inline void -snd_azf3328_codec_outw(const struct snd_azf3328 *chip, int reg, u16 value) +snd_azf3328_codec_outw(const struct snd_azf3328 *chip, unsigned reg, u16 value) { - outw(value, chip->codec_port + reg); + outw(value, chip->codec_io + reg); } static inline u16 -snd_azf3328_codec_inw(const struct snd_azf3328 *chip, int reg) +snd_azf3328_codec_inw(const struct snd_azf3328 *chip, unsigned reg) +{ + return inw(chip->codec_io + reg); +} + +static inline void +snd_azf3328_codec_outl(const struct snd_azf3328 *chip, unsigned reg, u32 value) { - return inw(chip->codec_port + reg); + outl(value, chip->codec_io + reg); +} + +static inline u32 +snd_azf3328_codec_inl(const struct snd_azf3328 *chip, unsigned reg) +{ + return inl(chip->codec_io + reg); } static inline void -snd_azf3328_codec_outl(const struct snd_azf3328 *chip, int reg, u32 value) +snd_azf3328_game_outb(const struct snd_azf3328 *chip, unsigned reg, u8 value) { - outl(value, chip->codec_port + reg); + outb(value, chip->game_io + reg); } static inline void -snd_azf3328_io2_outb(const struct snd_azf3328 *chip, int reg, u8 value) +snd_azf3328_game_outw(const struct snd_azf3328 *chip, unsigned reg, u16 value) { - outb(value, chip->io2_port + reg); + outw(value, chip->game_io + reg); } static inline u8 -snd_azf3328_io2_inb(const struct snd_azf3328 *chip, int reg) +snd_azf3328_game_inb(const struct snd_azf3328 *chip, unsigned reg) +{ + return inb(chip->game_io + reg); +} + +static inline u16 +snd_azf3328_game_inw(const struct snd_azf3328 *chip, unsigned reg) { - return inb(chip->io2_port + reg); + return inw(chip->game_io + reg); } static inline void -snd_azf3328_mixer_outw(const struct snd_azf3328 *chip, int reg, u16 value) +snd_azf3328_mixer_outw(const struct snd_azf3328 *chip, unsigned reg, u16 value) { - outw(value, chip->mixer_port + reg); + outw(value, chip->mixer_io + reg); } static inline u16 -snd_azf3328_mixer_inw(const struct snd_azf3328 *chip, int reg) +snd_azf3328_mixer_inw(const struct snd_azf3328 *chip, unsigned reg) { - return inw(chip->mixer_port + reg); + return inw(chip->mixer_io + reg); } -static void -snd_azf3328_mixer_set_mute(const struct snd_azf3328 *chip, int reg, int do_mute) +#define AZF_MUTE_BIT 0x80 + +static int +snd_azf3328_mixer_set_mute(const struct snd_azf3328 *chip, + unsigned reg, int do_mute +) { - unsigned long portbase = chip->mixer_port + reg + 1; - unsigned char oldval; + unsigned long portbase = chip->mixer_io + reg + 1; + int updated; /* the mute bit is on the *second* (i.e. right) register of a * left/right channel setting */ - oldval = inb(portbase); - if (do_mute) - oldval |= 0x80; - else - oldval &= ~0x80; - outb(oldval, portbase); + updated = snd_azf3328_io_reg_setb(portbase, AZF_MUTE_BIT, do_mute); + + /* indicate whether it was muted before */ + return (do_mute) ? !updated : updated; } static void -snd_azf3328_mixer_write_volume_gradually(const struct snd_azf3328 *chip, int reg, unsigned char dst_vol_left, unsigned char dst_vol_right, int chan_sel, int delay) +snd_azf3328_mixer_write_volume_gradually(const struct snd_azf3328 *chip, + unsigned reg, + unsigned char dst_vol_left, + unsigned char dst_vol_right, + int chan_sel, int delay +) { - unsigned long portbase = chip->mixer_port + reg; + unsigned long portbase = chip->mixer_io + reg; unsigned char curr_vol_left = 0, curr_vol_right = 0; - int left_done = 0, right_done = 0; - + int left_change = 0, right_change = 0; + snd_azf3328_dbgcallenter(); - if (chan_sel & SET_CHAN_LEFT) + + if (chan_sel & SET_CHAN_LEFT) { curr_vol_left = inb(portbase + 1); - else - left_done = 1; - if (chan_sel & SET_CHAN_RIGHT) + + /* take care of muting flag contained in left channel */ + if (curr_vol_left & AZF_MUTE_BIT) + dst_vol_left |= AZF_MUTE_BIT; + else + dst_vol_left &= ~AZF_MUTE_BIT; + + left_change = (curr_vol_left > dst_vol_left) ? -1 : 1; + } + + if (chan_sel & SET_CHAN_RIGHT) { curr_vol_right = inb(portbase + 0); - else - right_done = 1; - - /* take care of muting flag (0x80) contained in left channel */ - if (curr_vol_left & 0x80) - dst_vol_left |= 0x80; - else - dst_vol_left &= ~0x80; + + right_change = (curr_vol_right > dst_vol_right) ? -1 : 1; + } do { - if (!left_done) { - if (curr_vol_left > dst_vol_left) - curr_vol_left--; - else - if (curr_vol_left < dst_vol_left) - curr_vol_left++; - else - left_done = 1; - outb(curr_vol_left, portbase + 1); + if (left_change) { + if (curr_vol_left != dst_vol_left) { + curr_vol_left += left_change; + outb(curr_vol_left, portbase + 1); + } else + left_change = 0; } - if (!right_done) { - if (curr_vol_right > dst_vol_right) - curr_vol_right--; - else - if (curr_vol_right < dst_vol_right) - curr_vol_right++; - else - right_done = 1; + if (right_change) { + if (curr_vol_right != dst_vol_right) { + curr_vol_right += right_change; + /* during volume change, the right channel is crackling * somewhat more than the left channel, unfortunately. * This seems to be a hardware issue. */ - outb(curr_vol_right, portbase + 0); + outb(curr_vol_right, portbase + 0); + } else + right_change = 0; } if (delay) mdelay(delay); - } while ((!left_done) || (!right_done)); + } while ((left_change) || (right_change)); snd_azf3328_dbgcallleave(); } @@ -379,7 +489,7 @@ * general mixer element */ struct azf3328_mixer_reg { - unsigned int reg; + unsigned reg; unsigned int lchan_shift, rchan_shift; unsigned int mask; unsigned int invert: 1; @@ -544,13 +654,14 @@ "Mix", "Mic" }; static const char * const texts3[] = { - "Mic", "CD", "Video", "Aux", + "Mic", "CD", "Video", "Aux", "Line", "Mix", "Mix Mono", "Phone" }; static const char * const texts4[] = { "pre 3D", "post 3D" }; struct azf3328_mixer_reg reg; + const char *p = NULL; snd_azf3328_mixer_reg_decode(&reg, kcontrol->private_value); uinfo->type = SNDRV_CTL_ELEM_TYPE_ENUMERATED; @@ -561,18 +672,20 @@ if (reg.reg == IDX_MIXER_ADVCTL2) { switch(reg.lchan_shift) { case 8: /* modem out sel */ - strcpy(uinfo->value.enumerated.name, texts1[uinfo->value.enumerated.item]); + p = texts1[uinfo->value.enumerated.item]; break; case 9: /* mono sel source */ - strcpy(uinfo->value.enumerated.name, texts2[uinfo->value.enumerated.item]); + p = texts2[uinfo->value.enumerated.item]; break; case 15: /* PCM Out Path */ - strcpy(uinfo->value.enumerated.name, texts4[uinfo->value.enumerated.item]); + p = texts4[uinfo->value.enumerated.item]; break; } } else - strcpy(uinfo->value.enumerated.name, texts3[uinfo->value.enumerated.item] -); + if (reg.reg == IDX_MIXER_REC_SELECT) + p = texts3[uinfo->value.enumerated.item]; + + strcpy(uinfo->value.enumerated.name, p); return 0; } @@ -583,7 +696,7 @@ struct snd_azf3328 *chip = snd_kcontrol_chip(kcontrol); struct azf3328_mixer_reg reg; unsigned short val; - + snd_azf3328_mixer_reg_decode(&reg, kcontrol->private_value); val = snd_azf3328_mixer_inw(chip, reg.reg); if (reg.reg == IDX_MIXER_REC_SELECT) { @@ -605,7 +718,7 @@ struct snd_azf3328 *chip = snd_kcontrol_chip(kcontrol); struct azf3328_mixer_reg reg; unsigned int oreg, nreg, val; - + snd_azf3328_mixer_reg_decode(&reg, kcontrol->private_value); oreg = snd_azf3328_mixer_inw(chip, reg.reg); val = oreg; @@ -717,15 +830,16 @@ snd_azf3328_mixer_outw(chip, IDX_MIXER_RESET, 0x0000); /* mute and zero volume channels */ - for (idx = 0; idx < ARRAY_SIZE(snd_azf3328_init_values); idx++) { + for (idx = 0; idx < ARRAY_SIZE(snd_azf3328_init_values); ++idx) { snd_azf3328_mixer_outw(chip, snd_azf3328_init_values[idx][0], snd_azf3328_init_values[idx][1]); } - + /* add mixer controls */ sw = snd_azf3328_mixer_controls; - for (idx = 0; idx < ARRAY_SIZE(snd_azf3328_mixer_controls); idx++, sw++) { + for (idx = 0; idx < ARRAY_SIZE(snd_azf3328_mixer_controls); + ++idx, ++sw) { if ((err = snd_ctl_add(chip->card, snd_ctl_new1(sw, chip))) < 0) return err; } @@ -757,8 +871,8 @@ } static void -snd_azf3328_setfmt(struct snd_azf3328 *chip, - unsigned int reg, +snd_azf3328_codec_setfmt(struct snd_azf3328 *chip, + unsigned reg, unsigned int bitrate, unsigned int format_width, unsigned int channels @@ -769,24 +883,25 @@ snd_azf3328_dbgcallenter(); switch (bitrate) { - case 4000: val |= SOUNDFORMAT_FREQ_SUSPECTED_4000; break; - case 4800: val |= SOUNDFORMAT_FREQ_SUSPECTED_4800; break; - case 5512: val |= SOUNDFORMAT_FREQ_5510; break; /* the AZF3328 names it "5510" for some strange reason */ - case 6620: val |= SOUNDFORMAT_FREQ_6620; break; - case 8000: val |= SOUNDFORMAT_FREQ_8000; break; - case 9600: val |= SOUNDFORMAT_FREQ_9600; break; - case 11025: val |= SOUNDFORMAT_FREQ_11025; break; - case 13240: val |= SOUNDFORMAT_FREQ_SUSPECTED_13240; break; - case 16000: val |= SOUNDFORMAT_FREQ_16000; break; - case 22050: val |= SOUNDFORMAT_FREQ_22050; break; - case 32000: val |= SOUNDFORMAT_FREQ_32000; break; - case 44100: val |= SOUNDFORMAT_FREQ_44100; break; - case 48000: val |= SOUNDFORMAT_FREQ_48000; break; - case 66200: val |= SOUNDFORMAT_FREQ_SUSPECTED_66200; break; + case AZF_FREQ_4000: val |= SOUNDFORMAT_FREQ_SUSPECTED_4000; break; + case AZF_FREQ_4800: val |= SOUNDFORMAT_FREQ_SUSPECTED_4800; break; + case AZF_FREQ_5512: + /* the AZF3328 names it "5510" for some strange reason */ + val |= SOUNDFORMAT_FREQ_5510; break; + case AZF_FREQ_6620: val |= SOUNDFORMAT_FREQ_6620; break; + case AZF_FREQ_8000: val |= SOUNDFORMAT_FREQ_8000; break; + case AZF_FREQ_9600: val |= SOUNDFORMAT_FREQ_9600; break; + case AZF_FREQ_11025: val |= SOUNDFORMAT_FREQ_11025; break; + case AZF_FREQ_13240: val |= SOUNDFORMAT_FREQ_SUSPECTED_13240; break; + case AZF_FREQ_16000: val |= SOUNDFORMAT_FREQ_16000; break; + case AZF_FREQ_22050: val |= SOUNDFORMAT_FREQ_22050; break; + case AZF_FREQ_32000: val |= SOUNDFORMAT_FREQ_32000; break; default: snd_printk(KERN_WARNING "unknown bitrate %d, assuming 44.1kHz!\n", bitrate); - val |= SOUNDFORMAT_FREQ_44100; - break; + /* fall-through */ + case AZF_FREQ_44100: val |= SOUNDFORMAT_FREQ_44100; break; + case AZF_FREQ_48000: val |= SOUNDFORMAT_FREQ_48000; break; + case AZF_FREQ_66200: val |= SOUNDFORMAT_FREQ_SUSPECTED_66200; break; } /* val = 0xff07; 3m27.993s (65301Hz; -> 64000Hz???) hmm, 66120, 65967, 66123 */ /* val = 0xff09; 17m15.098s (13123,478Hz; -> 12000Hz???) hmm, 13237.2Hz? */ @@ -805,10 +920,10 @@ val |= SOUNDFORMAT_FLAG_16BIT; spin_lock_irqsave(&chip->reg_lock, flags); - + /* set bitrate/format */ snd_azf3328_codec_outw(chip, reg, val); - + /* changing the bitrate/format settings switches off the * audio output with an annoying click in case of 8/16bit format change * (maybe shutting down DAC/ADC?), thus immediately @@ -830,31 +945,81 @@ snd_azf3328_dbgcallleave(); } +static inline void +snd_azf3328_codec_setfmt_lowpower(struct snd_azf3328 *chip, + unsigned reg +) +{ + /* choose lowest frequency for low power consumption. + * While this will cause louder noise due to rather coarse frequency, + * it should never matter since output should always + * get disabled properly when idle anyway. */ + snd_azf3328_codec_setfmt(chip, reg, AZF_FREQ_4000, 8, 1); +} + +static inline void +snd_azf3328_codec_enable(struct snd_azf3328 *chip, int enable) +{ + /* no idea what exactly is being done here, but I strongly assume it's + * PM related */ + snd_azf3328_io_reg_setw( + chip->codec_io+IDX_IO_6AH, + IO_6A_PAUSE_PLAYBACK_BIT8, + !enable + ); +} + +static void +snd_azf3328_codec_activity(struct snd_azf3328 *chip, + snd_azf3328_stream_index_t stream_type, + int enable +) +{ + int need_change = (chip->audio_stream[stream_type].running != enable); + + snd_azf3328_dbgplay( + "codec_activity: type %d, enable %d, need_change %d\n", + stream_type, enable, need_change + ); + if (need_change) { + snd_azf3328_stream_index_t other = + (stream_type == AZF_PLAYBACK) ? + AZF_CAPTURE : AZF_PLAYBACK; + /* small check to prevent shutting down the other party + * in case it's active */ + if (!((!enable) && (chip->audio_stream[other].running))) + snd_azf3328_codec_enable(chip, enable); + + /* ...and adjust clock, too + * (reduce noise and power consumption) */ + if (!enable) + snd_azf3328_codec_setfmt_lowpower( + chip, + chip->audio_stream[stream_type].portbase + + IDX_IO_PLAY_SOUNDFORMAT + ); + } + chip->audio_stream[stream_type].running = enable; +} + static void snd_azf3328_setdmaa(struct snd_azf3328 *chip, long unsigned int addr, unsigned int count, unsigned int size, - int do_recording) + snd_azf3328_stream_index_t stream_type +) { - unsigned long flags, portbase; - unsigned int is_running; - snd_azf3328_dbgcallenter(); - if (do_recording) { - /* access capture registers, i.e. skip playback reg section */ - portbase = chip->codec_port + 0x20; - is_running = chip->is_recording; - } else { - /* access the playback register section */ - portbase = chip->codec_port + 0x00; - is_running = chip->is_playing; - } + if (!chip->audio_stream[stream_type].running) { + /* AZF3328 uses a two buffer pointer DMA playback approach */ + + unsigned long flags, portbase, addr_area2; + + /* width 32bit (prevent overflow): */ + unsigned long count_areas, count_tmp; - /* AZF3328 uses a two buffer pointer DMA playback approach */ - if (!is_running) { - unsigned long addr_area2; - unsigned long count_areas, count_tmp; /* width 32bit -- overflow!! */ + portbase = chip->audio_stream[stream_type].portbase; count_areas = size/2; addr_area2 = addr+count_areas; count_areas--; /* max. index */ @@ -884,11 +1049,11 @@ snd_azf3328_dbgcallenter(); #if 0 - snd_azf3328_setfmt(chip, IDX_IO_PLAY_SOUNDFORMAT, + snd_azf3328_codec_setfmt(chip, IDX_IO_PLAY_SOUNDFORMAT, runtime->rate, snd_pcm_format_width(runtime->format), runtime->channels); - snd_azf3328_setdmaa(chip, runtime->dma_addr, count, size, 0); + snd_azf3328_setdmaa(chip, runtime->dma_addr, count, size, AZF_PLAYBACK); #endif snd_azf3328_dbgcallleave(); return 0; @@ -906,11 +1071,11 @@ snd_azf3328_dbgcallenter(); #if 0 - snd_azf3328_setfmt(chip, IDX_IO_REC_SOUNDFORMAT, + snd_azf3328_codec_setfmt(chip, IDX_IO_REC_SOUNDFORMAT, runtime->rate, snd_pcm_format_width(runtime->format), runtime->channels); - snd_azf3328_setdmaa(chip, runtime->dma_addr, count, size, 1); + snd_azf3328_setdmaa(chip, runtime->dma_addr, count, size, AZF_CAPTURE); #endif snd_azf3328_dbgcallleave(); return 0; @@ -923,6 +1088,7 @@ struct snd_pcm_runtime *runtime = substream->runtime; int result = 0; unsigned int status1; + int previously_muted; snd_azf3328_dbgcalls("snd_azf3328_playback_trigger cmd %d\n", cmd); @@ -930,20 +1096,22 @@ case SNDRV_PCM_TRIGGER_START: snd_azf3328_dbgplay("START PLAYBACK\n"); - /* mute WaveOut */ - snd_azf3328_mixer_set_mute(chip, IDX_MIXER_WAVEOUT, 1); + /* mute WaveOut (avoid clicking during setup) */ + previously_muted = + snd_azf3328_mixer_set_mute(chip, IDX_MIXER_WAVEOUT, 1); - snd_azf3328_setfmt(chip, IDX_IO_PLAY_SOUNDFORMAT, + snd_azf3328_codec_setfmt(chip, IDX_IO_PLAY_SOUNDFORMAT, runtime->rate, snd_pcm_format_width(runtime->format), runtime->channels); spin_lock(&chip->reg_lock); - /* stop playback */ status1 = snd_azf3328_codec_inw(chip, IDX_IO_PLAY_FLAGS); + + /* stop playback */ status1 &= ~DMA_RESUME; snd_azf3328_codec_outw(chip, IDX_IO_PLAY_FLAGS, status1); - + /* FIXME: clear interrupts or what??? */ snd_azf3328_codec_outw(chip, IDX_IO_PLAY_IRQTYPE, 0xffff); spin_unlock(&chip->reg_lock); @@ -951,7 +1119,7 @@ snd_azf3328_setdmaa(chip, runtime->dma_addr, snd_pcm_lib_period_bytes(substream), snd_pcm_lib_buffer_bytes(substream), - 0); + AZF_PLAYBACK); spin_lock(&chip->reg_lock); #ifdef WIN9X @@ -978,30 +1146,34 @@ DMA_SOMETHING_ELSE); #endif spin_unlock(&chip->reg_lock); + snd_azf3328_codec_activity(chip, AZF_PLAYBACK, 1); /* now unmute WaveOut */ - snd_azf3328_mixer_set_mute(chip, IDX_MIXER_WAVEOUT, 0); + if (!previously_muted) + snd_azf3328_mixer_set_mute(chip, IDX_MIXER_WAVEOUT, 0); - chip->is_playing = 1; snd_azf3328_dbgplay("STARTED PLAYBACK\n"); break; case SNDRV_PCM_TRIGGER_RESUME: snd_azf3328_dbgplay("RESUME PLAYBACK\n"); /* resume playback if we were active */ - if (chip->is_playing) + spin_lock(&chip->reg_lock); + if (chip->audio_stream[AZF_PLAYBACK].running) snd_azf3328_codec_outw(chip, IDX_IO_PLAY_FLAGS, snd_azf3328_codec_inw(chip, IDX_IO_PLAY_FLAGS) | DMA_RESUME); + spin_unlock(&chip->reg_lock); break; case SNDRV_PCM_TRIGGER_STOP: snd_azf3328_dbgplay("STOP PLAYBACK\n"); - /* mute WaveOut */ - snd_azf3328_mixer_set_mute(chip, IDX_MIXER_WAVEOUT, 1); + /* mute WaveOut (avoid clicking during setup) */ + previously_muted = + snd_azf3328_mixer_set_mute(chip, IDX_MIXER_WAVEOUT, 1); spin_lock(&chip->reg_lock); - /* stop playback */ status1 = snd_azf3328_codec_inw(chip, IDX_IO_PLAY_FLAGS); + /* stop playback */ status1 &= ~DMA_RESUME; snd_azf3328_codec_outw(chip, IDX_IO_PLAY_FLAGS, status1); @@ -1013,10 +1185,12 @@ status1 &= ~DMA_PLAY_SOMETHING1; snd_azf3328_codec_outw(chip, IDX_IO_PLAY_FLAGS, status1); spin_unlock(&chip->reg_lock); - + snd_azf3328_codec_activity(chip, AZF_PLAYBACK, 0); + /* now unmute WaveOut */ - snd_azf3328_mixer_set_mute(chip, IDX_MIXER_WAVEOUT, 0); - chip->is_playing = 0; + if (!previously_muted) + snd_azf3328_mixer_set_mute(chip, IDX_MIXER_WAVEOUT, 0); + snd_azf3328_dbgplay("STOPPED PLAYBACK\n"); break; case SNDRV_PCM_TRIGGER_SUSPEND: @@ -1035,7 +1209,7 @@ printk(KERN_ERR "FIXME: unknown trigger mode!\n"); return -EINVAL; } - + snd_azf3328_dbgcallleave(); return result; } @@ -1057,7 +1231,7 @@ snd_azf3328_dbgplay("START CAPTURE\n"); - snd_azf3328_setfmt(chip, IDX_IO_REC_SOUNDFORMAT, + snd_azf3328_codec_setfmt(chip, IDX_IO_REC_SOUNDFORMAT, runtime->rate, snd_pcm_format_width(runtime->format), runtime->channels); @@ -1067,7 +1241,7 @@ status1 = snd_azf3328_codec_inw(chip, IDX_IO_REC_FLAGS); status1 &= ~DMA_RESUME; snd_azf3328_codec_outw(chip, IDX_IO_REC_FLAGS, status1); - + /* FIXME: clear interrupts or what??? */ snd_azf3328_codec_outw(chip, IDX_IO_REC_IRQTYPE, 0xffff); spin_unlock(&chip->reg_lock); @@ -1075,7 +1249,7 @@ snd_azf3328_setdmaa(chip, runtime->dma_addr, snd_pcm_lib_period_bytes(substream), snd_pcm_lib_buffer_bytes(substream), - 1); + AZF_CAPTURE); spin_lock(&chip->reg_lock); #ifdef WIN9X @@ -1102,16 +1276,18 @@ DMA_SOMETHING_ELSE); #endif spin_unlock(&chip->reg_lock); + snd_azf3328_codec_activity(chip, AZF_CAPTURE, 1); - chip->is_recording = 1; snd_azf3328_dbgplay("STARTED CAPTURE\n"); break; case SNDRV_PCM_TRIGGER_RESUME: snd_azf3328_dbgplay("RESUME CAPTURE\n"); /* resume recording if we were active */ - if (chip->is_recording) + spin_lock(&chip->reg_lock); + if (chip->audio_stream[AZF_CAPTURE].running) snd_azf3328_codec_outw(chip, IDX_IO_REC_FLAGS, snd_azf3328_codec_inw(chip, IDX_IO_REC_FLAGS) | DMA_RESUME); + spin_unlock(&chip->reg_lock); break; case SNDRV_PCM_TRIGGER_STOP: snd_azf3328_dbgplay("STOP CAPTURE\n"); @@ -1129,8 +1305,8 @@ status1 &= ~DMA_PLAY_SOMETHING1; snd_azf3328_codec_outw(chip, IDX_IO_REC_FLAGS, status1); spin_unlock(&chip->reg_lock); - - chip->is_recording = 0; + snd_azf3328_codec_activity(chip, AZF_CAPTURE, 0); + snd_azf3328_dbgplay("STOPPED CAPTURE\n"); break; case SNDRV_PCM_TRIGGER_SUSPEND: @@ -1149,7 +1325,7 @@ printk(KERN_ERR "FIXME: unknown trigger mode!\n"); return -EINVAL; } - + snd_azf3328_dbgcallleave(); return result; } @@ -1162,11 +1338,11 @@ snd_pcm_uframes_t frmres; #ifdef QUERY_HARDWARE - bufptr = inl(chip->codec_port+IDX_IO_PLAY_DMA_START_1); + bufptr = snd_azf3328_codec_inl(chip, IDX_IO_PLAY_DMA_START_1); #else bufptr = substream->runtime->dma_addr; #endif - result = inl(chip->codec_port+IDX_IO_PLAY_DMA_CURRPOS); + result = snd_azf3328_codec_inl(chip, IDX_IO_PLAY_DMA_CURRPOS); /* calculate offset */ result -= bufptr; @@ -1183,11 +1359,11 @@ snd_pcm_uframes_t frmres; #ifdef QUERY_HARDWARE - bufptr = inl(chip->codec_port+IDX_IO_REC_DMA_START_1); + bufptr = snd_azf3328_codec_inl(chip, IDX_IO_REC_DMA_START_1); #else bufptr = substream->runtime->dma_addr; #endif - result = inl(chip->codec_port+IDX_IO_REC_DMA_CURRPOS); + result = snd_azf3328_codec_inl(chip, IDX_IO_REC_DMA_CURRPOS); /* calculate offset */ result -= bufptr; @@ -1196,27 +1372,232 @@ return frmres; } +/******************************************************************/ + +#ifdef SUPPORT_GAMEPORT +static inline void +snd_azf3328_gameport_irq_enable(struct snd_azf3328 *chip, int enable) +{ + snd_azf3328_io_reg_setb( + chip->game_io+IDX_GAME_HWCONFIG, + GAME_HWCFG_IRQ_ENABLE, + enable + ); +} + +static inline void +snd_azf3328_gameport_legacy_address_enable(struct snd_azf3328 *chip, int enable) +{ + snd_azf3328_io_reg_setb( + chip->game_io+IDX_GAME_HWCONFIG, + GAME_HWCFG_LEGACY_ADDRESS_ENABLE, + enable + ); +} + +static inline void +snd_azf3328_gameport_axis_circuit_enable(struct snd_azf3328 *chip, int enable) +{ + snd_azf3328_io_reg_setw( + chip->codec_io+IDX_IO_6AH, + IO_6A_SOMETHING2_GAMEPORT, + !enable + ); +} + +static inline void +snd_azf3328_gameport_interrupt(struct snd_azf3328 *chip) +{ + /* + * skeleton handler only + * (we do not want axis reading in interrupt handler - too much load!) + */ + snd_azf3328_dbggame("gameport irq\n"); + + /* this should ACK the gameport IRQ properly, hopefully. */ + snd_azf3328_game_inw(chip, IDX_GAME_AXIS_VALUE); +} + +static int +snd_azf3328_gameport_open(struct gameport *gameport, int mode) +{ + struct snd_azf3328 *chip = gameport_get_port_data(gameport); + int res; + + snd_azf3328_dbggame("gameport_open, mode %d\n", mode); + switch (mode) { + case GAMEPORT_MODE_COOKED: + case GAMEPORT_MODE_RAW: + res = 0; + break; + default: + res = -1; + break; + } + + snd_azf3328_gameport_axis_circuit_enable(chip, (res == 0)); + + return res; +} + +static void +snd_azf3328_gameport_close(struct gameport *gameport) +{ + struct snd_azf3328 *chip = gameport_get_port_data(gameport); + + snd_azf3328_dbggame("gameport_close\n"); + snd_azf3328_gameport_axis_circuit_enable(chip, 0); +} + +static int +snd_azf3328_gameport_cooked_read(struct gameport *gameport, + int *axes, + int *buttons +) +{ + struct snd_azf3328 *chip = gameport_get_port_data(gameport); + int i; + u8 val; + + snd_assert(chip, return 0); + + spin_lock(&chip->reg_lock); + val = snd_azf3328_game_inb(chip, IDX_GAME_LEGACY_COMPATIBLE); + *buttons = (~(val) >> 4) & 0xf; + + /* ok, this one is a bit dirty: cooked_read is being polled by a timer, + * thus we're atomic and cannot actively wait in here + * (which would be useful for us since it probably would be better + * to trigger a measurement in here, then wait a short amount of + * time until it's finished, then read values of _this_ measurement). + * + * Thus we simply resort to reading values if they're available already + * and trigger the next measurement. + */ + + val = snd_azf3328_game_inb(chip, IDX_GAME_AXES_CONFIG); + if (val & GAME_AXES_SAMPLING_READY) { + for (i = 0; i < 4; ++i) { + /* configure the axis to read */ + val = (i << 4) | 0x0f; + snd_azf3328_game_outb(chip, IDX_GAME_AXES_CONFIG, val); + + chip->axes[i] = snd_azf3328_game_inw( + chip, IDX_GAME_AXIS_VALUE + ); + } + } + + /* trigger next axes sampling, to be evaluated the next time we + * enter this function */ + + /* for some very, very strange reason we cannot enable + * Measurement Ready monitoring for all axes here, + * at least not when only one joystick connected */ + val = 0x03; /* we're able to monitor axes 1 and 2 only */ + snd_azf3328_game_outb(chip, IDX_GAME_AXES_CONFIG, val); + + snd_azf3328_game_outw(chip, IDX_GAME_AXIS_VALUE, 0xffff); + spin_unlock(&chip->reg_lock); + + for (i = 0; i < 4; i++) { + axes[i] = chip->axes[i]; + if (axes[i] == 0xffff) + axes[i] = -1; + } + + snd_azf3328_dbggame("cooked_read: axes %d %d %d %d buttons %d\n", + axes[0], axes[1], axes[2], axes[3], *buttons + ); + + return 0; +} + +static int __devinit +snd_azf3328_gameport(struct snd_azf3328 *chip, int dev) +{ + struct gameport *gp; + + int io_port = chip->game_io; + + chip->gameport = gp = gameport_allocate_port(); + if (!gp) { + printk(KERN_ERR "azt3328: cannot alloc memory for gameport\n"); + return -ENOMEM; + } + + gameport_set_name(gp, "AZF3328 Gameport"); + gameport_set_phys(gp, "pci%s/gameport0", pci_name(chip->pci)); + gameport_set_dev_parent(gp, &chip->pci->dev); + gp->io = io_port; + gameport_set_port_data(gp, chip); + + gp->open = snd_azf3328_gameport_open; + gp->close = snd_azf3328_gameport_close; + gp->fuzz = 16; /* seems ok */ + gp->cooked_read = snd_azf3328_gameport_cooked_read; + + /* DISABLE legacy address: we don't need it! */ + snd_azf3328_gameport_legacy_address_enable(chip, 0); + + snd_azf3328_gameport_axis_circuit_enable(chip, 0); + + gameport_register_port(chip->gameport); + + return 0; +} + +static void +snd_azf3328_gameport_free(struct snd_azf3328 *chip) +{ + if (chip->gameport) { + gameport_unregister_port(chip->gameport); + chip->gameport = NULL; + } + snd_azf3328_gameport_irq_enable(chip, 0); +} +#else +static inline int +snd_azf3328_gameport(struct snd_azf3328 *chip, int dev) { return -ENOSYS; } +static inline void +snd_azf3328_gameport_free(struct snd_azf3328 *chip) { } +static inline void +snd_azf3328_gameport_interrupt(struct snd_azf3328 *chip) +{ + printk(KERN_WARNING "huh, game port IRQ occurred!?\n"); +} +#endif /* SUPPORT_GAMEPORT */ + +/******************************************************************/ + static irqreturn_t snd_azf3328_interrupt(int irq, void *dev_id) { struct snd_azf3328 *chip = dev_id; u8 status, which; +#if DEBUG_PLAY_REC static unsigned long irq_count; +#endif status = snd_azf3328_codec_inb(chip, IDX_IO_IRQSTATUS); /* fast path out, to ease interrupt sharing */ - if (!(status & (IRQ_PLAYBACK|IRQ_RECORDING|IRQ_MPU401|IRQ_TIMER))) + if (!(status & + (IRQ_PLAYBACK|IRQ_RECORDING|IRQ_GAMEPORT|IRQ_MPU401|IRQ_TIMER) + )) return IRQ_NONE; /* must be interrupt for another device */ snd_azf3328_dbgplay("Interrupt %ld!\nIDX_IO_PLAY_FLAGS %04x, IDX_IO_PLAY_IRQTYPE %04x, IDX_IO_IRQSTATUS %04x\n", - irq_count, + irq_count++, snd_azf3328_codec_inw(chip, IDX_IO_PLAY_FLAGS), snd_azf3328_codec_inw(chip, IDX_IO_PLAY_IRQTYPE), status); - + if (status & IRQ_TIMER) { - /* snd_azf3328_dbgplay("timer %ld\n", inl(chip->codec_port+IDX_IO_TIMER_VALUE) & TIMER_VALUE_MASK); */ + /* snd_azf3328_dbgplay("timer %ld\n", + snd_azf3328_codec_inl(chip, IDX_IO_TIMER_VALUE) + & TIMER_VALUE_MASK + ); */ if (chip->timer) snd_timer_interrupt(chip->timer, chip->timer->sticks); /* ACK timer */ @@ -1232,11 +1613,16 @@ snd_azf3328_codec_outb(chip, IDX_IO_PLAY_IRQTYPE, which); spin_unlock(&chip->reg_lock); - if (chip->pcm && chip->playback_substream) { - snd_pcm_period_elapsed(chip->playback_substream); + if (chip->pcm && chip->audio_stream[AZF_PLAYBACK].substream) { + snd_pcm_period_elapsed( + chip->audio_stream[AZF_PLAYBACK].substream + ); snd_azf3328_dbgplay("PLAY period done (#%x), @ %x\n", which, - inl(chip->codec_port+IDX_IO_PLAY_DMA_CURRPOS)); + snd_azf3328_codec_inl( + chip, IDX_IO_PLAY_DMA_CURRPOS + ) + ); } else snd_azf3328_dbgplay("azt3328: ouch, irq handler problem!\n"); if (which & IRQ_PLAY_SOMETHING) @@ -1249,16 +1635,23 @@ snd_azf3328_codec_outb(chip, IDX_IO_REC_IRQTYPE, which); spin_unlock(&chip->reg_lock); - if (chip->pcm && chip->capture_substream) { - snd_pcm_period_elapsed(chip->capture_substream); + if (chip->pcm && chip->audio_stream[AZF_CAPTURE].substream) { + snd_pcm_period_elapsed( + chip->audio_stream[AZF_CAPTURE].substream + ); snd_azf3328_dbgplay("REC period done (#%x), @ %x\n", which, - inl(chip->codec_port+IDX_IO_REC_DMA_CURRPOS)); + snd_azf3328_codec_inl( + chip, IDX_IO_REC_DMA_CURRPOS + ) + ); } else snd_azf3328_dbgplay("azt3328: ouch, irq handler problem!\n"); if (which & IRQ_REC_SOMETHING) snd_azf3328_dbgplay("azt3328: unknown rec IRQ type occurred, please report!\n"); } + if (status & IRQ_GAMEPORT) + snd_azf3328_gameport_interrupt(chip); /* MPU401 has less critical IRQ requirements * than timer and playback/recording, right? */ if (status & IRQ_MPU401) { @@ -1268,7 +1661,6 @@ * If so, then I don't know how... */ snd_azf3328_dbgplay("azt3328: MPU401 IRQ\n"); } - irq_count++; return IRQ_HANDLED; } @@ -1287,8 +1679,8 @@ .rates = SNDRV_PCM_RATE_5512 | SNDRV_PCM_RATE_8000_48000 | SNDRV_PCM_RATE_KNOT, - .rate_min = 4000, - .rate_max = 66200, + .rate_min = AZF_FREQ_4000, + .rate_max = AZF_FREQ_66200, .channels_min = 1, .channels_max = 2, .buffer_bytes_max = 65536, @@ -1315,8 +1707,8 @@ .rates = SNDRV_PCM_RATE_5512 | SNDRV_PCM_RATE_8000_48000 | SNDRV_PCM_RATE_KNOT, - .rate_min = 4000, - .rate_max = 66200, + .rate_min = AZF_FREQ_4000, + .rate_max = AZF_FREQ_66200, .channels_min = 1, .channels_max = 2, .buffer_bytes_max = 65536, @@ -1329,10 +1721,24 @@ static unsigned int snd_azf3328_fixed_rates[] = { - 4000, 4800, 5512, 6620, 8000, 9600, 11025, 13240, 16000, 22050, 32000, - 44100, 48000, 66200 }; + AZF_FREQ_4000, + AZF_FREQ_4800, + AZF_FREQ_5512, + AZF_FREQ_6620, + AZF_FREQ_8000, + AZF_FREQ_9600, + AZF_FREQ_11025, + AZF_FREQ_13240, + AZF_FREQ_16000, + AZF_FREQ_22050, + AZF_FREQ_32000, + AZF_FREQ_44100, + AZF_FREQ_48000, + AZF_FREQ_66200 +}; + static struct snd_pcm_hw_constraint_list snd_azf3328_hw_constraints_rates = { - .count = ARRAY_SIZE(snd_azf3328_fixed_rates), + .count = ARRAY_SIZE(snd_azf3328_fixed_rates), .list = snd_azf3328_fixed_rates, .mask = 0, }; @@ -1346,7 +1752,7 @@ struct snd_pcm_runtime *runtime = substream->runtime; snd_azf3328_dbgcallenter(); - chip->playback_substream = substream; + chip->audio_stream[AZF_PLAYBACK].substream = substream; runtime->hw = snd_azf3328_playback; snd_pcm_hw_constraint_list(runtime, 0, SNDRV_PCM_HW_PARAM_RATE, &snd_azf3328_hw_constraints_rates); @@ -1361,7 +1767,7 @@ struct snd_pcm_runtime *runtime = substream->runtime; snd_azf3328_dbgcallenter(); - chip->capture_substream = substream; + chip->audio_stream[AZF_CAPTURE].substream = substream; runtime->hw = snd_azf3328_capture; snd_pcm_hw_constraint_list(runtime, 0, SNDRV_PCM_HW_PARAM_RATE, &snd_azf3328_hw_constraints_rates); @@ -1375,7 +1781,7 @@ struct snd_azf3328 *chip = snd_pcm_substream_chip(substream); snd_azf3328_dbgcallenter(); - chip->playback_substream = NULL; + chip->audio_stream[AZF_PLAYBACK].substream = NULL; snd_azf3328_dbgcallleave(); return 0; } @@ -1386,7 +1792,7 @@ struct snd_azf3328 *chip = snd_pcm_substream_chip(substream); snd_azf3328_dbgcallenter(); - chip->capture_substream = NULL; + chip->audio_stream[AZF_CAPTURE].substream = NULL; snd_azf3328_dbgcallleave(); return 0; } @@ -1441,102 +1847,8 @@ /******************************************************************/ -#ifdef SUPPORT_JOYSTICK -static int __devinit -snd_azf3328_config_joystick(struct snd_azf3328 *chip, int dev) -{ - struct gameport *gp; - struct resource *r; - - if (!joystick[dev]) - return -ENODEV; - - if (!(r = request_region(0x200, 8, "AZF3328 gameport"))) { - printk(KERN_WARNING "azt3328: cannot reserve joystick ports\n"); - return -EBUSY; - } - - chip->gameport = gp = gameport_allocate_port(); - if (!gp) { - printk(KERN_ERR "azt3328: cannot allocate memory for gameport\n"); - release_and_free_resource(r); - return -ENOMEM; - } - - gameport_set_name(gp, "AZF3328 Gameport"); - gameport_set_phys(gp, "pci%s/gameport0", pci_name(chip->pci)); - gameport_set_dev_parent(gp, &chip->pci->dev); - gp->io = 0x200; - gameport_set_port_data(gp, r); - - snd_azf3328_io2_outb(chip, IDX_IO2_LEGACY_ADDR, - snd_azf3328_io2_inb(chip, IDX_IO2_LEGACY_ADDR) | LEGACY_JOY); - - gameport_register_port(chip->gameport); - - return 0; -} - -static void -snd_azf3328_free_joystick(struct snd_azf3328 *chip) -{ - if (chip->gameport) { - struct resource *r = gameport_get_port_data(chip->gameport); - - gameport_unregister_port(chip->gameport); - chip->gameport = NULL; - /* disable gameport */ - snd_azf3328_io2_outb(chip, IDX_IO2_LEGACY_ADDR, - snd_azf3328_io2_inb(chip, IDX_IO2_LEGACY_ADDR) & ~LEGACY_JOY); - release_and_free_resource(r); - } -} -#else -static inline int -snd_azf3328_config_joystick(struct snd_azf3328 *chip, int dev) { return -ENOSYS; } -static inline void -snd_azf3328_free_joystick(struct snd_azf3328 *chip) { } -#endif - -/******************************************************************/ - -static int -snd_azf3328_free(struct snd_azf3328 *chip) -{ - if (chip->irq < 0) - goto __end_hw; - - /* reset (close) mixer */ - snd_azf3328_mixer_set_mute(chip, IDX_MIXER_PLAY_MASTER, 1); /* first mute master volume */ - snd_azf3328_mixer_outw(chip, IDX_MIXER_RESET, 0x0000); - - /* interrupt setup - mask everything (FIXME!) */ - /* well, at least we know how to disable the timer IRQ */ - snd_azf3328_codec_outb(chip, IDX_IO_TIMER_VALUE + 3, 0x00); - - if (chip->irq >= 0) - synchronize_irq(chip->irq); -__end_hw: - snd_azf3328_free_joystick(chip); - if (chip->irq >= 0) - free_irq(chip->irq, chip); - pci_release_regions(chip->pci); - pci_disable_device(chip->pci); - - kfree(chip); - return 0; -} - -static int -snd_azf3328_dev_free(struct snd_device *device) -{ - struct snd_azf3328 *chip = device->device_data; - return snd_azf3328_free(chip); -} - -/******************************************************************/ - -/*** NOTE: the physical timer resolution actually is 1024000 ticks per second, +/*** NOTE: the physical timer resolution actually is 1024000 ticks per second + *** (probably derived from main crystal via a divider of 24), *** but announcing those attributes to user-space would make programs *** configure the timer to a 1 tick value, resulting in an absolutely fatal *** timer IRQ storm. @@ -1564,7 +1876,7 @@ delay = 49; /* minimum time is 49 ticks */ } snd_azf3328_dbgtimer("setting timer countdown value %d, add COUNTDOWN|IRQ\n", delay); - delay |= TIMER_ENABLE_COUNTDOWN | TIMER_ENABLE_IRQ; + delay |= TIMER_COUNTDOWN_ENABLE | TIMER_IRQ_ENABLE; spin_lock_irqsave(&chip->reg_lock, flags); snd_azf3328_codec_outl(chip, IDX_IO_TIMER_VALUE, delay); spin_unlock_irqrestore(&chip->reg_lock, flags); @@ -1582,7 +1894,7 @@ chip = snd_timer_chip(timer); spin_lock_irqsave(&chip->reg_lock, flags); /* disable timer countdown and interrupt */ - /* FIXME: should we write TIMER_ACK_IRQ here? */ + /* FIXME: should we write TIMER_IRQ_ACK here? */ snd_azf3328_codec_outb(chip, IDX_IO_TIMER_VALUE + 3, 0); spin_unlock_irqrestore(&chip->reg_lock, flags); snd_azf3328_dbgcallleave(); @@ -1626,9 +1938,10 @@ snd_azf3328_timer_hw.resolution *= seqtimer_scaling; snd_azf3328_timer_hw.ticks /= seqtimer_scaling; - if ((err = snd_timer_new(chip->card, "AZF3328", &tid, &timer)) < 0) { + + err = snd_timer_new(chip->card, "AZF3328", &tid, &timer); + if (err < 0) goto out; - } strcpy(timer->name, "AZF3328 timer"); timer->private_data = chip; @@ -1636,6 +1949,8 @@ chip->timer = timer; + snd_azf3328_timer_stop(timer); + err = 0; out: @@ -1645,10 +1960,44 @@ /******************************************************************/ +static int +snd_azf3328_free(struct snd_azf3328 *chip) +{ + if (chip->irq < 0) + goto __end_hw; + + /* reset (close) mixer: + * first mute master volume, then reset + */ + snd_azf3328_mixer_set_mute(chip, IDX_MIXER_PLAY_MASTER, 1); + snd_azf3328_mixer_outw(chip, IDX_MIXER_RESET, 0x0000); + + snd_azf3328_timer_stop(chip->timer); + snd_azf3328_gameport_free(chip); + + if (chip->irq >= 0) + synchronize_irq(chip->irq); +__end_hw: + if (chip->irq >= 0) + free_irq(chip->irq, chip); + pci_release_regions(chip->pci); + pci_disable_device(chip->pci); + + kfree(chip); + return 0; +} + +static int +snd_azf3328_dev_free(struct snd_device *device) +{ + struct snd_azf3328 *chip = device->device_data; + return snd_azf3328_free(chip); +} + #if 0 /* check whether a bit can be modified */ static void -snd_azf3328_test_bit(unsigned int reg, int bit) +snd_azf3328_test_bit(unsigned unsigned reg, int bit) { unsigned char val, valoff, valon; @@ -1659,42 +2008,74 @@ outb(val|(1 << bit), reg); valon = inb(reg); - + outb(val, reg); - printk(KERN_ERR "reg %04x bit %d: %02x %02x %02x\n", reg, bit, val, valoff, valon); + printk(KERN_ERR "reg %04x bit %d: %02x %02x %02x\n", + reg, bit, val, valoff, valon + ); } #endif -#if DEBUG_MISC -static void +static inline void snd_azf3328_debug_show_ports(const struct snd_azf3328 *chip) { +#if DEBUG_MISC u16 tmp; - snd_azf3328_dbgmisc("codec_port 0x%lx, io2_port 0x%lx, mpu_port 0x%lx, synth_port 0x%lx, mixer_port 0x%lx, irq %d\n", chip->codec_port, chip->io2_port, chip->mpu_port, chip->synth_port, chip->mixer_port, chip->irq); - - snd_azf3328_dbgmisc("io2 %02x %02x %02x %02x %02x %02x\n", snd_azf3328_io2_inb(chip, 0), snd_azf3328_io2_inb(chip, 1), snd_azf3328_io2_inb(chip, 2), snd_azf3328_io2_inb(chip, 3), snd_azf3328_io2_inb(chip, 4), snd_azf3328_io2_inb(chip, 5)); - - for (tmp=0; tmp <= 0x01; tmp += 1) - snd_azf3328_dbgmisc("0x%02x: opl 0x%04x, mpu300 0x%04x, mpu310 0x%04x, mpu320 0x%04x, mpu330 0x%04x\n", tmp, inb(0x388 + tmp), inb(0x300 + tmp), inb(0x310 + tmp), inb(0x320 + tmp), inb(0x330 + tmp)); + snd_azf3328_dbgmisc( + "codec_io 0x%lx, game_io 0x%lx, mpu_io 0x%lx, " + "opl3_io 0x%lx, mixer_io 0x%lx, irq %d\n", + chip->codec_io, chip->game_io, chip->mpu_io, + chip->opl3_io, chip->mixer_io, chip->irq + ); + + snd_azf3328_dbgmisc("game %02x %02x %02x %02x %02x %02x\n", + snd_azf3328_game_inb(chip, 0), + snd_azf3328_game_inb(chip, 1), + snd_azf3328_game_inb(chip, 2), + snd_azf3328_game_inb(chip, 3), + snd_azf3328_game_inb(chip, 4), + snd_azf3328_game_inb(chip, 5) + ); + + for (tmp = 0; tmp < 0x07; tmp += 1) + snd_azf3328_dbgmisc("mpu_io 0x%04x\n", inb(chip->mpu_io + tmp)); + + for (tmp = 0; tmp <= 0x07; tmp += 1) + snd_azf3328_dbgmisc("0x%02x: game200 0x%04x, game208 0x%04x\n", + tmp, inb(0x200 + tmp), inb(0x208 + tmp)); + + for (tmp = 0; tmp <= 0x01; tmp += 1) + snd_azf3328_dbgmisc( + "0x%02x: mpu300 0x%04x, mpu310 0x%04x, mpu320 0x%04x, " + "mpu330 0x%04x opl388 0x%04x opl38c 0x%04x\n", + tmp, + inb(0x300 + tmp), + inb(0x310 + tmp), + inb(0x320 + tmp), + inb(0x330 + tmp), + inb(0x388 + tmp), + inb(0x38c + tmp) + ); for (tmp = 0; tmp < AZF_IO_SIZE_CODEC; tmp += 2) - snd_azf3328_dbgmisc("codec 0x%02x: 0x%04x\n", tmp, snd_azf3328_codec_inw(chip, tmp)); + snd_azf3328_dbgmisc("codec 0x%02x: 0x%04x\n", + tmp, snd_azf3328_codec_inw(chip, tmp) + ); for (tmp = 0; tmp < AZF_IO_SIZE_MIXER; tmp += 2) - snd_azf3328_dbgmisc("mixer 0x%02x: 0x%04x\n", tmp, snd_azf3328_mixer_inw(chip, tmp)); + snd_azf3328_dbgmisc("mixer 0x%02x: 0x%04x\n", + tmp, snd_azf3328_mixer_inw(chip, tmp) + ); +#endif /* DEBUG_MISC */ } -#else -static inline void -snd_azf3328_debug_show_ports(const struct snd_azf3328 *chip) {} -#endif static int __devinit snd_azf3328_create(struct snd_card *card, - struct pci_dev *pci, - unsigned long device_type, - struct snd_azf3328 ** rchip) + struct pci_dev *pci, + unsigned long device_type, + struct snd_azf3328 **rchip) { struct snd_azf3328 *chip; int err; @@ -1705,7 +2086,8 @@ *rchip = NULL; - if ((err = pci_enable_device(pci)) < 0) + err = pci_enable_device(pci); + if (err < 0) return err; chip = kzalloc(sizeof(*chip), GFP_KERNEL); @@ -1721,20 +2103,25 @@ /* check if we can restrict PCI DMA transfers to 24 bits */ if (pci_set_dma_mask(pci, DMA_24BIT_MASK) < 0 || pci_set_consistent_dma_mask(pci, DMA_24BIT_MASK) < 0) { - snd_printk(KERN_ERR "architecture does not support 24bit PCI busmaster DMA\n"); + snd_printk(KERN_ERR "architecture does not support " + "24bit PCI busmaster DMA\n" + ); err = -ENXIO; goto out_err; } - if ((err = pci_request_regions(pci, "Aztech AZF3328")) < 0) { + err = pci_request_regions(pci, "Aztech AZF3328"); + if (err < 0) goto out_err; - } - chip->codec_port = pci_resource_start(pci, 0); - chip->io2_port = pci_resource_start(pci, 1); - chip->mpu_port = pci_resource_start(pci, 2); - chip->synth_port = pci_resource_start(pci, 3); - chip->mixer_port = pci_resource_start(pci, 4); + chip->codec_io = pci_resource_start(pci, 0); + chip->game_io = pci_resource_start(pci, 1); + chip->mpu_io = pci_resource_start(pci, 2); + chip->opl3_io = pci_resource_start(pci, 3); + chip->mixer_io = pci_resource_start(pci, 4); + + chip->audio_stream[AZF_PLAYBACK].portbase = chip->codec_io + 0x00; + chip->audio_stream[AZF_CAPTURE].portbase = chip->codec_io + 0x20; if (request_irq(pci->irq, snd_azf3328_interrupt, IRQF_SHARED, card->shortname, chip)) { @@ -1747,29 +2134,29 @@ synchronize_irq(chip->irq); snd_azf3328_debug_show_ports(chip); - - if ((err = snd_device_new(card, SNDRV_DEV_LOWLEVEL, chip, &ops)) < 0) { + + err = snd_device_new(card, SNDRV_DEV_LOWLEVEL, chip, &ops); + if (err < 0) goto out_err; - } /* create mixer interface & switches */ - if ((err = snd_azf3328_mixer_new(chip)) < 0) + err = snd_azf3328_mixer_new(chip); + if (err < 0) goto out_err; -#if 0 - /* set very low bitrate to reduce noise and power consumption? */ - snd_azf3328_setfmt(chip, IDX_IO_PLAY_SOUNDFORMAT, 5512, 8, 1); -#endif + /* shutdown codecs to save power */ + /* have snd_azf3328_codec_activity() act properly */ + chip->audio_stream[AZF_PLAYBACK].running = 1; + snd_azf3328_codec_activity(chip, AZF_PLAYBACK, 0); /* standard chip init stuff */ - /* default IRQ init value */ + /* default IRQ init value */ tmp = DMA_PLAY_SOMETHING2|DMA_EPILOGUE_SOMETHING|DMA_SOMETHING_ELSE; spin_lock_irq(&chip->reg_lock); snd_azf3328_codec_outb(chip, IDX_IO_PLAY_FLAGS, tmp); snd_azf3328_codec_outb(chip, IDX_IO_REC_FLAGS, tmp); snd_azf3328_codec_outb(chip, IDX_IO_SOMETHING_FLAGS, tmp); - snd_azf3328_codec_outb(chip, IDX_IO_TIMER_VALUE + 3, 0x00); /* disable timer */ spin_unlock_irq(&chip->reg_lock); snd_card_set_dev(card, &pci->dev); @@ -1805,52 +2192,61 @@ return -ENOENT; } - card = snd_card_new(index[dev], id[dev], THIS_MODULE, 0 ); + card = snd_card_new(index[dev], id[dev], THIS_MODULE, 0); if (card == NULL) return -ENOMEM; strcpy(card->driver, "AZF3328"); strcpy(card->shortname, "Aztech AZF3328 (PCI168)"); - if ((err = snd_azf3328_create(card, pci, pci_id->driver_data, &chip)) < 0) { + err = snd_azf3328_create(card, pci, pci_id->driver_data, &chip); + if (err < 0) goto out_err; - } card->private_data = chip; - if ((err = snd_mpu401_uart_new( card, 0, MPU401_HW_MPU401, - chip->mpu_port, MPU401_INFO_INTEGRATED, - pci->irq, 0, &chip->rmidi)) < 0) { - snd_printk(KERN_ERR "azf3328: no MPU-401 device at 0x%lx?\n", chip->mpu_port); + err = snd_mpu401_uart_new( + card, 0, MPU401_HW_MPU401, chip->mpu_io, MPU401_INFO_INTEGRATED, + pci->irq, 0, &chip->rmidi + ); + if (err < 0) { + snd_printk(KERN_ERR "azf3328: no MPU-401 device at 0x%lx?\n", + chip->mpu_io + ); goto out_err; } - if ((err = snd_azf3328_timer(chip, 0)) < 0) { + err = snd_azf3328_timer(chip, 0); + if (err < 0) goto out_err; - } - if ((err = snd_azf3328_pcm(chip, 0)) < 0) { + err = snd_azf3328_pcm(chip, 0); + if (err < 0) goto out_err; - } - if (snd_opl3_create(card, chip->synth_port, chip->synth_port+2, + if (snd_opl3_create(card, chip->opl3_io, chip->opl3_io+2, OPL3_HW_AUTO, 1, &opl3) < 0) { snd_printk(KERN_ERR "azf3328: no OPL3 device at 0x%lx-0x%lx?\n", - chip->synth_port, chip->synth_port+2 ); + chip->opl3_io, chip->opl3_io+2 + ); } else { - if ((err = snd_opl3_hwdep_new(opl3, 0, 1, NULL)) < 0) { + /* need to use IDs 1, 2 since ID 0 is snd_azf3328_timer above */ + err = snd_opl3_timer_new(opl3, 1, 2); + if (err < 0) + goto out_err; + err = snd_opl3_hwdep_new(opl3, 0, 1, NULL); + if (err < 0) goto out_err; - } } opl3->private_data = chip; sprintf(card->longname, "%s at 0x%lx, irq %i", - card->shortname, chip->codec_port, chip->irq); + card->shortname, chip->codec_io, chip->irq); - if ((err = snd_card_register(card)) < 0) { + err = snd_card_register(card); + if (err < 0) goto out_err; - } #ifdef MODULE printk( @@ -1861,19 +2257,18 @@ 1024000 / seqtimer_scaling, seqtimer_scaling); #endif - if (snd_azf3328_config_joystick(chip, dev) < 0) - snd_azf3328_io2_outb(chip, IDX_IO2_LEGACY_ADDR, - snd_azf3328_io2_inb(chip, IDX_IO2_LEGACY_ADDR) & ~LEGACY_JOY); + snd_azf3328_gameport(chip, dev); pci_set_drvdata(pci, card); dev++; err = 0; goto out; - + out_err: + snd_printk(KERN_ERR "azf3328: something failed, exiting\n"); snd_card_free(card); - + out: snd_azf3328_dbgcallleave(); return err; @@ -1894,27 +2289,27 @@ { struct snd_card *card = pci_get_drvdata(pci); struct snd_azf3328 *chip = card->private_data; - int reg; + unsigned reg; snd_power_change_state(card, SNDRV_CTL_POWER_D3hot); - + snd_pcm_suspend_all(chip->pcm); - for (reg = 0; reg < AZF_IO_SIZE_MIXER_PM / 2; reg++) - chip->saved_regs_mixer[reg] = inw(chip->mixer_port + reg * 2); + for (reg = 0; reg < AZF_IO_SIZE_MIXER_PM / 2; ++reg) + chip->saved_regs_mixer[reg] = inw(chip->mixer_io + reg * 2); /* make sure to disable master volume etc. to prevent looping sound */ snd_azf3328_mixer_set_mute(chip, IDX_MIXER_PLAY_MASTER, 1); snd_azf3328_mixer_set_mute(chip, IDX_MIXER_WAVEOUT, 1); - - for (reg = 0; reg < AZF_IO_SIZE_CODEC_PM / 2; reg++) - chip->saved_regs_codec[reg] = inw(chip->codec_port + reg * 2); - for (reg = 0; reg < AZF_IO_SIZE_IO2_PM / 2; reg++) - chip->saved_regs_io2[reg] = inw(chip->io2_port + reg * 2); - for (reg = 0; reg < AZF_IO_SIZE_MPU_PM / 2; reg++) - chip->saved_regs_mpu[reg] = inw(chip->mpu_port + reg * 2); - for (reg = 0; reg < AZF_IO_SIZE_SYNTH_PM / 2; reg++) - chip->saved_regs_synth[reg] = inw(chip->synth_port + reg * 2); + + for (reg = 0; reg < AZF_IO_SIZE_CODEC_PM / 2; ++reg) + chip->saved_regs_codec[reg] = inw(chip->codec_io + reg * 2); + for (reg = 0; reg < AZF_IO_SIZE_GAME_PM / 2; ++reg) + chip->saved_regs_game[reg] = inw(chip->game_io + reg * 2); + for (reg = 0; reg < AZF_IO_SIZE_MPU_PM / 2; ++reg) + chip->saved_regs_mpu[reg] = inw(chip->mpu_io + reg * 2); + for (reg = 0; reg < AZF_IO_SIZE_OPL3_PM / 2; ++reg) + chip->saved_regs_opl3[reg] = inw(chip->opl3_io + reg * 2); pci_disable_device(pci); pci_save_state(pci); @@ -1927,7 +2322,7 @@ { struct snd_card *card = pci_get_drvdata(pci); struct snd_azf3328 *chip = card->private_data; - int reg; + unsigned reg; pci_set_power_state(pci, PCI_D0); pci_restore_state(pci); @@ -1939,23 +2334,21 @@ } pci_set_master(pci); - for (reg = 0; reg < AZF_IO_SIZE_IO2_PM / 2; reg++) - outw(chip->saved_regs_io2[reg], chip->io2_port + reg * 2); - for (reg = 0; reg < AZF_IO_SIZE_MPU_PM / 2; reg++) - outw(chip->saved_regs_mpu[reg], chip->mpu_port + reg * 2); - for (reg = 0; reg < AZF_IO_SIZE_SYNTH_PM / 2; reg++) - outw(chip->saved_regs_synth[reg], chip->synth_port + reg * 2); - for (reg = 0; reg < AZF_IO_SIZE_MIXER_PM / 2; reg++) - outw(chip->saved_regs_mixer[reg], chip->mixer_port + reg * 2); - for (reg = 0; reg < AZF_IO_SIZE_CODEC_PM / 2; reg++) - outw(chip->saved_regs_codec[reg], chip->codec_port + reg * 2); + for (reg = 0; reg < AZF_IO_SIZE_GAME_PM / 2; ++reg) + outw(chip->saved_regs_game[reg], chip->game_io + reg * 2); + for (reg = 0; reg < AZF_IO_SIZE_MPU_PM / 2; ++reg) + outw(chip->saved_regs_mpu[reg], chip->mpu_io + reg * 2); + for (reg = 0; reg < AZF_IO_SIZE_OPL3_PM / 2; ++reg) + outw(chip->saved_regs_opl3[reg], chip->opl3_io + reg * 2); + for (reg = 0; reg < AZF_IO_SIZE_MIXER_PM / 2; ++reg) + outw(chip->saved_regs_mixer[reg], chip->mixer_io + reg * 2); + for (reg = 0; reg < AZF_IO_SIZE_CODEC_PM / 2; ++reg) + outw(chip->saved_regs_codec[reg], chip->codec_io + reg * 2); snd_power_change_state(card, SNDRV_CTL_POWER_D0); return 0; } -#endif - - +#endif /* CONFIG_PM */ static struct pci_driver driver = { --- linux-2.6.26-rc1.orig/sound/pci/azt3328.h 2008-05-03 20:59:44.000000000 +0200 +++ linux-2.6.26-rc1/sound/pci/azt3328.h 2008-05-14 22:47:12.000000000 +0200 @@ -54,7 +54,10 @@ #define SOUNDFORMAT_XTAL1 0x00 #define SOUNDFORMAT_XTAL2 0x01 /* all _SUSPECTED_ values are not used by Windows drivers, so we don't - * have any hard facts, only rough measurements */ + * have any hard facts, only rough measurements. + * All we know is that the crystal used on the board has 24.576MHz, + * like many soundcards (which results in the frequencies below when + * using certain divider values selected by the values below) */ #define SOUNDFORMAT_FREQ_SUSPECTED_4000 0x0c | SOUNDFORMAT_XTAL1 #define SOUNDFORMAT_FREQ_SUSPECTED_4800 0x0a | SOUNDFORMAT_XTAL1 #define SOUNDFORMAT_FREQ_5510 0x0c | SOUNDFORMAT_XTAL2 @@ -72,6 +75,26 @@ #define SOUNDFORMAT_FLAG_16BIT 0x0010 #define SOUNDFORMAT_FLAG_2CHANNELS 0x0020 +/* define frequency helpers, for maximum value safety */ +enum { +#define AZF_FREQ(rate) AZF_FREQ_##rate = rate + AZF_FREQ(4000), + AZF_FREQ(4800), + AZF_FREQ(5512), + AZF_FREQ(6620), + AZF_FREQ(8000), + AZF_FREQ(9600), + AZF_FREQ(11025), + AZF_FREQ(13240), + AZF_FREQ(16000), + AZF_FREQ(22050), + AZF_FREQ(32000), + AZF_FREQ(44100), + AZF_FREQ(48000), + AZF_FREQ(66200), +#undef AZF_FREQ +} AZF_FREQUENCIES; + /** recording area (see also: playback bit flag definitions) **/ #define IDX_IO_REC_FLAGS 0x20 /* ??, PU:0x0000 */ #define IDX_IO_REC_IRQTYPE 0x22 /* ??, PU:0x0000 */ @@ -97,40 +120,164 @@ /** DirectX timer, main interrupt area (FIXME: and something else?) **/ #define IDX_IO_TIMER_VALUE 0x60 /* found this timer area by pure luck :-) */ - #define TIMER_VALUE_MASK 0x000fffffUL /* timer countdown value; triggers IRQ when timer is finished */ - #define TIMER_ENABLE_COUNTDOWN 0x01000000UL /* activate the timer countdown */ - #define TIMER_ENABLE_IRQ 0x02000000UL /* trigger timer IRQ on zero transition */ - #define TIMER_ACK_IRQ 0x04000000UL /* being set in IRQ handler in case port 0x00 (hmm, not port 0x64!?!?) had 0x0020 set upon IRQ handler */ + /* timer countdown value; triggers IRQ when timer is finished */ + #define TIMER_VALUE_MASK 0x000fffffUL + /* activate timer countdown */ + #define TIMER_COUNTDOWN_ENABLE 0x01000000UL + /* trigger timer IRQ on zero transition */ + #define TIMER_IRQ_ENABLE 0x02000000UL + /* being set in IRQ handler in case port 0x00 (hmm, not port 0x64!?!?) + * had 0x0020 set upon IRQ handler */ + #define TIMER_IRQ_ACK 0x04000000UL #define IDX_IO_IRQSTATUS 0x64 - #define IRQ_PLAYBACK 0x0001 - #define IRQ_RECORDING 0x0002 - #define IRQ_MPU401 0x0010 - #define IRQ_TIMER 0x0020 /* DirectX timer */ - #define IRQ_UNKNOWN1 0x0040 /* probably unused, or possibly I2S port? or gameport IRQ? */ - #define IRQ_UNKNOWN2 0x0080 /* probably unused, or possibly I2S port? or gameport IRQ? */ + /* some IRQ bit in here might also be used to signal a power-management timer + * timeout, to request shutdown of the chip (e.g. AD1815JS has such a thing). + * Some OPL3 hardware (e.g. in LM4560) has some special timer hardware which + * can trigger an OPL3 timer IRQ, so maybe there's such a thing as well... */ + + #define IRQ_PLAYBACK 0x0001 + #define IRQ_RECORDING 0x0002 + #define IRQ_UNKNOWN1 0x0004 /* most probably I2S port */ + #define IRQ_GAMEPORT 0x0008 /* Interrupt of Digital(ly) Enhanced Game Port */ + #define IRQ_MPU401 0x0010 + #define IRQ_TIMER 0x0020 /* DirectX timer */ + #define IRQ_UNKNOWN2 0x0040 /* probably unused, or possibly I2S port? */ + #define IRQ_UNKNOWN3 0x0080 /* probably unused, or possibly I2S port? */ #define IDX_IO_66H 0x66 /* writing 0xffff returns 0x0000 */ -#define IDX_IO_SOME_VALUE 0x68 /* this is set to e.g. 0x3ff or 0x300, and writable; maybe some buffer limit, but I couldn't find out more, PU:0x00ff */ -#define IDX_IO_6AH 0x6A /* this WORD can be set to have bits 0x0028 activated (FIXME: correct??); actually inhibits PCM playback!!! maybe power management?? */ - #define IO_6A_PAUSE_PLAYBACK 0x0200 /* bit 9; sure, this pauses playback, but what the heck is this really about?? */ -#define IDX_IO_6CH 0x6C -#define IDX_IO_6EH 0x6E /* writing 0xffff returns 0x83fe */ -/* further I/O indices not saved/restored, so probably not used */ + /* this is set to e.g. 0x3ff or 0x300, and writable; + * maybe some buffer limit, but I couldn't find out more, PU:0x00ff: */ +#define IDX_IO_SOME_VALUE 0x68 + #define IO_68_RANDOM_TOGGLE1 0x0100 /* toggles randomly */ + #define IO_68_RANDOM_TOGGLE2 0x0200 /* toggles randomly */ + /* umm, nope, behaviour of these bits changes depending on what we wrote + * to 0x6b!! */ + +/* this WORD can be set to have bits 0x0028 activated (FIXME: correct??); + * actually inhibits PCM playback!!! maybe power management??: */ +#define IDX_IO_6AH 0x6A + /* bit 5: enabling this will activate permanent counting of bytes 2/3 + * at gameport I/O (0xb402/3) (equal values each) and cause + * gameport legacy I/O at 0x0200 to be _DISABLED_! + * Is this Digital Enhanced Game Port Enable??? Or maybe it's Testmode + * for Enhanced Digital Gameport (see 4D Wave DX card): */ + #define IO_6A_SOMETHING1_GAMEPORT 0x0020 + /* bit 8; sure, this _pauses_ playback (later resumes at same spot!), + * but what the heck is this really about??: */ + #define IO_6A_PAUSE_PLAYBACK_BIT8 0x0100 + /* bit 9; sure, this _pauses_ playback (later resumes at same spot!), + * but what the heck is this really about??: */ + #define IO_6A_PAUSE_PLAYBACK_BIT9 0x0200 + /* BIT8 and BIT9 are _NOT_ able to affect OPL3 MIDI playback, + * thus it suggests influence on PCM only!! + * However OTOH there seems to be no bit anywhere around here + * which is able to disable OPL3... */ + /* bit 10: enabling this actually changes values at legacy gameport + * I/O address (0x200); is this enabling of the Digital Enhanced Game Port??? + * Or maybe this simply switches off the NE558 circuit, since enabling this + * still lets us evaluate button states, but not axis states */ + #define IO_6A_SOMETHING2_GAMEPORT 0x0400 + /* writing 0x0300: causes quite some crackling during + * PC activity such as switching windows (PCI traffic?? + * --> FIFO/timing settings???) */ + /* writing 0x0100 plus/or 0x0200 inhibits playback */ + /* since the Windows .INF file has Flag_Enable_JoyStick and + * Flag_Enable_SB_DOS_Emulation directly together, it stands to reason + * that some other bit in this same register might be responsible + * for SB DOS Emulation activation (note that the file did NOT define + * a switch for OPL3!) */ +#define IDX_IO_6CH 0x6C /* unknown; fully read-writable */ +#define IDX_IO_6EH 0x6E + /* writing 0xffff returns 0x83fe (or 0x03fe only). + * writing 0x83 (and only 0x83!!) to 0x6f will cause 0x6c to switch + * from 0000 to ffff. */ + +/* further I/O indices not saved/restored and not readable after writing, + * so probably not used */ -/*** I/O 2 area port indices ***/ +/*** Gameport area port indices ***/ /* (only 0x06 of 0x08 bytes saved/restored by Windows driver) */ -#define AZF_IO_SIZE_IO2 0x08 -#define AZF_IO_SIZE_IO2_PM 0x06 +#define AZF_IO_SIZE_GAME 0x08 +#define AZF_IO_SIZE_GAME_PM 0x06 -#define IDX_IO2_LEGACY_ADDR 0x04 - #define LEGACY_SOMETHING 0x01 /* OPL3?? */ - #define LEGACY_JOY 0x08 +enum { + AZF_GAME_LEGACY_IO_PORT = 0x200 +} AZF_GAME_CONFIGS; + +#define IDX_GAME_LEGACY_COMPATIBLE 0x00 + /* in some operation mode, writing anything to this port + * triggers an interrupt: + * yup, that's in case IDX_GAME_01H has one of the + * axis measurement bits enabled + * (and of course one needs to have GAME_HWCFG_IRQ_ENABLE, too) */ + +#define IDX_GAME_AXES_CONFIG 0x01 + /* NOTE: layout of this register awfully similar (read: "identical??") + * to AD1815JS.pdf (p.29) */ + + /* enables axis 1 (X axis) measurement: */ + #define GAME_AXES_ENABLE_1 0x01 + /* enables axis 2 (Y axis) measurement: */ + #define GAME_AXES_ENABLE_2 0x02 + /* enables axis 3 (X axis) measurement: */ + #define GAME_AXES_ENABLE_3 0x04 + /* enables axis 4 (Y axis) measurement: */ + #define GAME_AXES_ENABLE_4 0x08 + /* selects the current axis to read the measured value of + * (at IDX_GAME_AXIS_VALUE): + * 00 = axis 1, 01 = axis 2, 10 = axis 3, 11 = axis 4: */ + #define GAME_AXES_READ_MASK 0x30 + /* enable to have the latch continuously accept ADC values + * (and continuously cause interrupts in case interrupts are enabled); + * AD1815JS.pdf says it's ~16ms interval there: */ + #define GAME_AXES_LATCH_ENABLE 0x40 + /* joystick data (measured axes) ready for reading: */ + #define GAME_AXES_SAMPLING_READY 0x80 + + /* NOTE: other card specs (SiS960 and others!) state that the + * game position latches should be frozen when reading and be freed + * (== reset?) after reading!!! + * Freezing most likely means disabling 0x40 (GAME_AXES_LATCH_ENABLE), + * but how to free the value? */ + /* An internet search for "gameport latch ADC" should provide some insight + * into how to program such a gameport system. */ + + /* writing 0xf0 to 01H once reset both counters to 0, in some special mode!? + * yup, in case 6AH 0x20 is not enabled + * (and 0x40 is sufficient, 0xf0 is not needed) */ + +#define IDX_GAME_AXIS_VALUE 0x02 + /* R: value of currently configured axis (word value!); + * W: trigger axis measurement */ + +#define IDX_GAME_HWCONFIG 0x04 + /* note: bits 4 to 7 are never set (== 0) when reading! + * --> reserved bits? */ + /* enables IRQ notification upon axes measurement ready: */ + #define GAME_HWCFG_IRQ_ENABLE 0x01 + /* these bits choose a different frequency for the + * internal ADC counter increment. + * hmm, seems to be a combo of bits: + * 00 --> standard frequency + * 10 --> 1/2 + * 01 --> 1/20 + * 11 --> 1/200: */ + #define GAME_HWCFG_ADC_COUNTER_FREQ_MASK 0x06 + + /* enable gameport legacy I/O address (0x200) + * I was unable to locate any configurability for a different address: */ + #define GAME_HWCFG_LEGACY_ADDRESS_ENABLE 0x08 +/*** MPU401 ***/ #define AZF_IO_SIZE_MPU 0x04 #define AZF_IO_SIZE_MPU_PM 0x04 -#define AZF_IO_SIZE_SYNTH 0x08 -#define AZF_IO_SIZE_SYNTH_PM 0x06 +/*** OPL3 synth ***/ +#define AZF_IO_SIZE_OPL3 0x08 +#define AZF_IO_SIZE_OPL3_PM 0x06 +/* hmm, given that a standard OPL3 has 4 registers only, + * there might be some enhanced functionality lurking at the end + * (especially since register 0x04 has a "non-empty" value 0xfe) */ /*** mixer I/O area port indices ***/ /* (only 0x22 of 0x40 bytes saved/restored by Windows driver)

2 3

Re: [alsa-devel] ioctl request code for specific alsa control
by Clemens Ladisch 16 May '08

16 May '08

Peter Wurmsdobler wrote: > What ioctl request code would I have to pass in order to reach this > control's put/get methods? snd_ctl_elem_write() (alsa-lib/src/control/control.c) snd_ctl_hw_elem_write() (alsa-lib/src/control/control_hw.c) ioctl(SNDRV_CTL_IOCTL_ELEM_WRITE) snd_ctl_ioctl() (linux/sound/core/control.c) snd_ctl_elem_write_user() snd_ctl_elem_write() kctl->put() HTH Clemens

2 1

[alsa-devel] Alsa: Get Playback Volume fails if set by outside source
by Ashwin Chandra 16 May '08

16 May '08

Hello, I am having an issue with the Alsa mixer. Currenty I have a function that gets the volume via snd_mixer_selem_get_playback_volume. This function works fine to get the last-set volume if my application is setting the volume. However if I change the volume using an outside source, such as alsamixer or gnome-volume-control and then I run the above function, the old volume that my application had set last is returned. Now I edited amixer's source code to put a while(1) around the same functions (the snd_mixer_selem_get_playback_volume), and it too does not get the correct volume if set from an outside source. Alsamixer on the other hand seems to work find, but I am not quite sure how they do it. I thought maybe I had to reinitialize the alsa interface (ie. snd_mixer_close...and then open, attach, register, etc), but it doesn't seem that alsamixer does that. Does anyone have any suggestions about what I can do to get the correct volume if set by an outside source?? Thanks!

2 1

[alsa-devel] pcm output timing question
by Paul Adolph 15 May '08

15 May '08

ALSA newbie warning... I have a funny application that does av sync by using the timing of a callback to set the audio timebase. In other words, I need to invoke that callback "as close as possible" to the time a previously-sent buffer of samples is presented, i.e. exits the speakers. The application uses the clock time of the arrival of this callback and does some math to figure out whether to repeat or skip video frames. I do not like this arrangement, but unfortunately I am stuck with it. My job is to hook this app to ALSA audio output. My current design uses a producer-consumer model with a thread that writes fixed-size application buffers to the ALSA ring buffer, driven by a message queue fed by the application. This is a little redundant I think, but I do not own the buffers sent by the app so this is what I came up with. The application needs a regular callback with the timing described above each time one of their buffers is played, and it is assumed they are played in sequence. I tried setting the ALSA callback to trigger when there are an application-buffer-size amount of frames available, using snd_pcm_sw_params_set_avail_min(). My idea was that if I primed the ring buffer to be full before it starts transfers with snd_pcm_sw_params_set_start_threshold(), I would get the ALSA callback each time that number of frames exited the ring buffer. But this did not work: I seem to be getting callbacks much more frequently that I wanted, probably on ALSA period boundaries. So this means I do not understand how snd_pcm_sw_params_set_start_threshold() is supposed to work I guess. Will somebody please comment on the viability of and/or problems with my current design, or suggest an alternative way to configure ALSA that will give me the output timing I need? Currently I am testing on an x86 ubuntu box, kernel 2.6.22-14, ALSA 1.0.14, gcc 4.1.3, glibc 2.6.1, but will also cross-compile for several embedded platforms in the future. Thanks in advance.

1 0

[alsa-devel] Oh, BTW... (OPL3 PM issue)
by Andreas Mohr 15 May '08

15 May '08

Hi, unfortunately I forgot to mention this: when I tested OPL3 functionality of my driver, I of course also tested this to do proper power management, and... well... it didn't! The result of doing an S2D was that the formerly perfectly playing emulated MIDI synth ended up with distorted (sort of Micky-Mouse'ish) voices directly after resume. I really don't think that it's my driver's power management which is at fault here, and I'd strongly guesstimate that ALSA does its own generic OPL3 power management handling for standard OPL3 hardware registration. So??? (aka "what the...!?") Thanks, Andreas Mohr

2 1

[alsa-devel] [PATCH asoc-dev (take #2)] ASoC: Convert N810 machine driver to use gpiolib
by Jarkko Nikula 15 May '08

15 May '08

Use gpiolib since it is now available for OMAPs. Change also references to HW version RX44 to product name N810. Signed-off-by: Jarkko Nikula <jarkko.nikula(a)nokia.com> --- sound/soc/omap/n810.c | 25 ++++++++++++++----------- 1 files changed, 14 insertions(+), 11 deletions(-) diff --git a/sound/soc/omap/n810.c b/sound/soc/omap/n810.c index c2afffb..c32487b 100644 --- a/sound/soc/omap/n810.c +++ b/sound/soc/omap/n810.c @@ -30,15 +30,15 @@ #include <asm/mach-types.h> #include <asm/arch/hardware.h> -#include <asm/arch/gpio.h> +#include <linux/gpio.h> #include <asm/arch/mcbsp.h> #include "omap-mcbsp.h" #include "omap-pcm.h" #include "../codecs/tlv320aic3x.h" -#define RX44_HEADSET_AMP_GPIO 10 -#define RX44_SPEAKER_AMP_GPIO 101 +#define N810_HEADSET_AMP_GPIO 10 +#define N810_SPEAKER_AMP_GPIO 101 static struct clk *sys_clkout2; static struct clk *sys_clkout2_src; @@ -154,9 +154,9 @@ static int n810_spk_event(struct snd_soc_dapm_widget *w, struct snd_kcontrol *k, int event) { if (SND_SOC_DAPM_EVENT_ON(event)) - omap_set_gpio_dataout(RX44_SPEAKER_AMP_GPIO, 1); + gpio_set_value(N810_SPEAKER_AMP_GPIO, 1); else - omap_set_gpio_dataout(RX44_SPEAKER_AMP_GPIO, 0); + gpio_set_value(N810_SPEAKER_AMP_GPIO, 0); return 0; } @@ -165,9 +165,9 @@ static int n810_jack_event(struct snd_soc_dapm_widget *w, struct snd_kcontrol *k, int event) { if (SND_SOC_DAPM_EVENT_ON(event)) - omap_set_gpio_dataout(RX44_HEADSET_AMP_GPIO, 1); + gpio_set_value(N810_HEADSET_AMP_GPIO, 1); else - omap_set_gpio_dataout(RX44_HEADSET_AMP_GPIO, 0); + gpio_set_value(N810_HEADSET_AMP_GPIO, 0); return 0; } @@ -303,12 +303,12 @@ static int __init n810_soc_init(void) clk_set_parent(sys_clkout2_src, func96m_clk); clk_set_rate(sys_clkout2, 12000000); - if (omap_request_gpio(RX44_HEADSET_AMP_GPIO) < 0) + if (gpio_request(N810_HEADSET_AMP_GPIO, "hs_amp") < 0) BUG(); - if (omap_request_gpio(RX44_SPEAKER_AMP_GPIO) < 0) + if (gpio_request(N810_SPEAKER_AMP_GPIO, "spk_amp") < 0) BUG(); - omap_set_gpio_direction(RX44_HEADSET_AMP_GPIO, 0); - omap_set_gpio_direction(RX44_SPEAKER_AMP_GPIO, 0); + gpio_direction_output(N810_HEADSET_AMP_GPIO, 0); + gpio_direction_output(N810_SPEAKER_AMP_GPIO, 0); return 0; err2: @@ -323,6 +323,9 @@ err1: static void __exit n810_soc_exit(void) { + gpio_free(N810_SPEAKER_AMP_GPIO); + gpio_free(N810_HEADSET_AMP_GPIO); + platform_device_unregister(n810_snd_device); } -- 1.5.5.1

2 1

[alsa-devel] [PATCH asoc-dev] ASoC: Convert N810 machine driver to use gpiolib
by Jarkko Nikula 15 May '08

15 May '08

Use gpiolib since it is now available for OMAPs. Change also references to HW version RX44 to product name N810. Signed-off-by: Jarkko Nikula <jarkko.nikula(a)nokia.com> --- sound/soc/omap/n810.c | 25 ++++++++++++++----------- 1 files changed, 14 insertions(+), 11 deletions(-) diff --git a/sound/soc/omap/n810.c b/sound/soc/omap/n810.c index c2afffb..2d982b1 100644 --- a/sound/soc/omap/n810.c +++ b/sound/soc/omap/n810.c @@ -30,15 +30,15 @@ #include <asm/mach-types.h> #include <asm/arch/hardware.h> -#include <asm/arch/gpio.h> +#include <asm/gpio.h> #include <asm/arch/mcbsp.h> #include "omap-mcbsp.h" #include "omap-pcm.h" #include "../codecs/tlv320aic3x.h" -#define RX44_HEADSET_AMP_GPIO 10 -#define RX44_SPEAKER_AMP_GPIO 101 +#define N810_HEADSET_AMP_GPIO 10 +#define N810_SPEAKER_AMP_GPIO 101 static struct clk *sys_clkout2; static struct clk *sys_clkout2_src; @@ -154,9 +154,9 @@ static int n810_spk_event(struct snd_soc_dapm_widget *w, struct snd_kcontrol *k, int event) { if (SND_SOC_DAPM_EVENT_ON(event)) - omap_set_gpio_dataout(RX44_SPEAKER_AMP_GPIO, 1); + gpio_set_value(N810_SPEAKER_AMP_GPIO, 1); else - omap_set_gpio_dataout(RX44_SPEAKER_AMP_GPIO, 0); + gpio_set_value(N810_SPEAKER_AMP_GPIO, 0); return 0; } @@ -165,9 +165,9 @@ static int n810_jack_event(struct snd_soc_dapm_widget *w, struct snd_kcontrol *k, int event) { if (SND_SOC_DAPM_EVENT_ON(event)) - omap_set_gpio_dataout(RX44_HEADSET_AMP_GPIO, 1); + gpio_set_value(N810_HEADSET_AMP_GPIO, 1); else - omap_set_gpio_dataout(RX44_HEADSET_AMP_GPIO, 0); + gpio_set_value(N810_HEADSET_AMP_GPIO, 0); return 0; } @@ -303,12 +303,12 @@ static int __init n810_soc_init(void) clk_set_parent(sys_clkout2_src, func96m_clk); clk_set_rate(sys_clkout2, 12000000); - if (omap_request_gpio(RX44_HEADSET_AMP_GPIO) < 0) + if (gpio_request(N810_HEADSET_AMP_GPIO, "hs_amp") < 0) BUG(); - if (omap_request_gpio(RX44_SPEAKER_AMP_GPIO) < 0) + if (gpio_request(N810_SPEAKER_AMP_GPIO, "spk_amp") < 0) BUG(); - omap_set_gpio_direction(RX44_HEADSET_AMP_GPIO, 0); - omap_set_gpio_direction(RX44_SPEAKER_AMP_GPIO, 0); + gpio_direction_output(N810_HEADSET_AMP_GPIO, 0); + gpio_direction_output(N810_SPEAKER_AMP_GPIO, 0); return 0; err2: @@ -323,6 +323,9 @@ err1: static void __exit n810_soc_exit(void) { + gpio_free(N810_SPEAKER_AMP_GPIO); + gpio_free(N810_HEADSET_AMP_GPIO); + platform_device_unregister(n810_snd_device); } -- 1.5.5.1

2 2

[alsa-devel] pcm output timing question
by Paul Adolph 14 May '08

14 May '08

ALSA newbie warning... I have a funny application that does av sync by using the timing of a callback to set the audio timebase. In other words, I need to invoke that callback "as close as possible" to the time a previously-sent buffer of samples is presented, i.e. exits the speakers. The application uses the clock time of the arrival of this callback and does some math to figure out whether to repeat or skip video frames. I do not like this arrangement, but unfortunately I am stuck with it. My job is to hook this app to ALSA audio output. My current design uses a producer-consumer model with a thread that writes fixed-size application buffers to the ALSA ring buffer, driven by a message queue fed by the application. This is a little redundant I think, but I do not own the buffers sent by the app so this is what I came up with. The application needs a regular callback with the timing described above each time one of their buffers is played, and it is assumed they are played in sequence. I tried setting the ALSA callback to trigger when there are an application-buffer-size amount of frames available, using snd_pcm_sw_params_set_avail_min(). My idea was that if I primed the ring buffer to be full before it starts transfers with snd_pcm_sw_params_set_start_threshold(), I would get the ALSA callback each time that number of frames exited the ring buffer. But this did not work: I seem to be getting callbacks much more frequently that I wanted, probably on ALSA period boundaries. So this means I do not understand how snd_pcm_sw_params_set_start_threshold() is supposed to work I guess. Will somebody please comment on the viability of and/or problems with my current design, or suggest an alternative way to configure ALSA that will give me the output timing I need? Currently I am testing on an x86 ubuntu box, kernel 2.6.22-14, ALSA 1.0.14, gcc 4.1.3, glibc 2.6.1, but will also cross-compile for several embedded platforms in the future. Thanks in advance.

1 0