A NOTE has been added to this issue. ====================================================================== <https://bugtrack.alsa-project.org/alsa-bug/view.php?id=1389> ====================================================================== Reported By: wood Assigned To: ====================================================================== Project: ALSA - driver Issue ID: 1389 Category: PCI - ens1371 Reproducibility: always Severity: block Priority: normal Status: new Distribution: Modified Redhat 9.0 Kernel Version: 2.4.24 self compiled ====================================================================== Date Submitted: 09-01-2005 20:32 CEST Last Modified: 06-13-2006 07:05 CEST ====================================================================== Summary: snd-ens1371 / es1938 Creative Ectiva silent playout + new line-in capture failure Description: Soundcard: Creative Soundblaster PCI (Ectiva) Model:CT4730 Chip Label: Creative EV1938 Platform: Stable 2.4.24 Linux kernel on VIA MiniITX ME6000/M10000 w/ multiple soundcards. ALSA compiled from source works with VIA motherboard Audio and some other Creative PCI soundcards. Need a stable ALSA driver for this soundcard. Problem: The snd-ens1371 module loads correctly as 3rd sound device (First audio device is VIA 8235 from motherboard, 2nd is PCSP). Alsamixer -c 2 shows a reasonable set of controls. Playout either with OSS emulation (play) or ALSA player (aplay) plays WAV files but only silence for both 1.0.9 and 1.0.10rc1 releases. On ALSA 1.0.9 codebase,the Master Volume does not always show in alsamixer. In 1.0.10rc1 it does always show. ALSO: On ALSA 1.0.9 code, the Line-In capture works under both OSS Emulation and with ALSA arecord. On the latest development release ALSA 1.0.10rc1, the Line-In capture no longer works. So a new problem has been introduced from .9 to .10. Other info: Native OSS driver es1371 works fine with this card. But I want to switch to ALSA. Earlier model Creative PCI128, Model CT4750 loads snd-ens1371 and works fine with both tested ALSA releases using same platform configuration. Audigy 2 Value also works fine using snd-emu10k1 with same configuration (I included this card info to rule out basic config problems). ====================================================================== ---------------------------------------------------------------------- wood - 06-13-06 06:43 ---------------------------------------------------------------------- IMPORTANT: I got this card to work perfectly in March under 1.0.11rc3 by adding some detection code in ac97_codec.c for this cards device ID. For this specific card I bypass tests that were added in 1.0.8rc2 from 1.0.8rc1. I confirmed that this card worked in 1.0.8rc1 but failed in 1.0.8rc2. The snd_ac97_try_bit() test was added in snd_ac97_cmix_new() as a condition before calling snd_ac97_cmute_new_stereo() with the control named "Switch". There were also some other cases where detection for this card failed and thus controls were never added. I have not tested 1.0.11 but from cursory glance I doubt the problem has been fixed. Here are the two changed functions I made in file ../pci/ac97/ac97_codec.c: ------------------------------------------ static int snd_ac97_cmix_new_stereo(struct snd_card *card, const char *pfx, int reg, int check_stereo, struct snd_ac97 *ac97) { int err; char name[44]; unsigned char lo_max, hi_max; snd_printk(KERN_DEBUG "snd_ac97_cmix_new %s : Enter\n",pfx); if (! snd_ac97_valid_reg(ac97, reg)) { //snd_printk(KERN_DEBUG "snd_ac97_cmix_new %s : Invalid reg 0x%x\n",pfx,reg); return 0; } /*******ADDED THIS CODE *************/ int special_card = (((ac97->id == 0x76058384) || (ac97->id == 0x8a058384)) && (strstr(pfx,"Playback"))); /********************/ if (snd_ac97_try_bit(ac97, reg, 15) /*******ADDED THIS CODE *************/ || special_card) /********************/ { sprintf(name, "%s Switch", pfx); if ((err = snd_ac97_cmute_new_stereo(card, name, reg, check_stereo, ac97)) < 0) return err; } check_volume_resolution(ac97, reg, &lo_max, &hi_max); if (lo_max) { sprintf(name, "%s Volume", pfx); if (special_card && (lo_max > 31)) { lo_max = 31; if (hi_max) hi_max = 31; } if ((err = snd_ac97_cvol_new(card, name, reg, lo_max, hi_max, ac97)) < 0) return err; } else snd_printk(KERN_DEBUG "snd_ac97_cmix_new : %s Volume control omitted, no lo_max val after check_volume_resolution call, lo=%d,hi=%d\n",pfx,lo_max,hi_max); snd_printk(KERN_DEBUG "snd_ac97_cmix_new %s : Done\n",pfx); return 0; } and in function: snd_ac97_mixer_build(): static int snd_ac97_mixer_build(struct snd_ac97 * ac97) { struct snd_card *card = ac97->bus->card; struct snd_kcontrol *kctl; int err; unsigned int idx; unsigned char max; snd_printk(KERN_DEBUG "snd_ac97_mixer_build : Enter ID=0x%x\n",ac97->id); /*******ADDED THIS CODE *************/ int special_card = (((ac97->id == 0x76058384) || (ac97->id == 0x8a058384))); /********************/ then, code the same ..... until: /* build MIC Capture controls */ if (!special_card && s snd_ac97_try_volume_mix(ac97, AC97_REC_GAIN_MIC)) { for (idx = 0; idx < 2; idx++) { if ((err = snd_ctl_add(card, snd_ac97_cnew(&snd_ac97_controls_mic_capture[idx], ac97))) < 0) return err; //snd_printk(KERN_DEBUG "snd_ac97_mixer_build : build MIC Capture [%d]\n",idx); } snd_ac97_write_cache(ac97, AC97_REC_GAIN_MIC, 0x0000); } ... ... ... My fix is a hack, relying on special handling of a specific device. But it works, so someone who knows more about these arcane device differences stands a better chance of making this work correctly. I can send the original file and the changed one outside of this forum to anybody who sends me an email. Raymond, here are the requested posts, but I do not think they shed light: With unmodified 1.0.9 =, sound card not working: # > cat /proc/asound/card2/codec97#0/ac97#0-0 0-0/0: 0x8a057605 ?�F Capabilities : -dedicated MIC PCM IN channel- -bass & treble- DAC resolution : 16-bit ADC resolution : 20-bit 3D enhancement : Reserved 29 Current setup Mic gain : +20dB [+20dB] POP path : post 3D Sim. stereo : off 3D enhancement : off Loudness : off Mono output : Mic Mic select : Mic2 ADC/DAC loopback : on --------------------------------------- With modified 1.0.11rc3 that works with this card: # > cat /proc/asound/card2/codec97#0/ac97#0-0 0-0/0: 0x76058384 F�S PCI Subsys Vendor: 0x0000 PCI Subsys Device: 0x0000 Capabilities : -dedicated MIC PCM IN channel- -bass & treble- DAC resolution : 16-bit ADC resolution : 20-bit 3D enhancement : Reserved 29 Current setup Mic gain : +0dB [+20dB] POP path : post 3D Sim. stereo : off 3D enhancement : off Loudness : off Mono output : MIX Mic select : Mic1 ADC/DAC loopback : off ---------------------------------------------------------------------- Raymond - 06-13-06 07:05 ---------------------------------------------------------------------- Refer to https://bugtrack.alsa-project.org/alsa-bug/view.php?id=2132, it seem that snd_es1371_codec_read() only work on first read and subsequent read return incorrect value. Please post the output of cat /proc/asound/card2/codec97#0/ac97#0-0+regs Issue History Date Modified Username Field Change ====================================================================== 09-01-05 20:32 wood New Issue 09-01-05 20:32 wood Distribution => Modified Redhat 9.0 09-01-05 20:32 wood Kernel Version => 2.4.24 self compiled 06-13-06 05:06 Raymond Note Added: 0010186 06-13-06 06:43 wood Note Added: 0010188 06-13-06 07:05 Raymond Note Added: 0010189 ====================================================================== _______________________________________________ Alsa-devel mailing list Alsa-devel@xxxxxxxxxxxxxxxxxxxxx https://lists.sourceforge.net/lists/listinfo/alsa-devel
