Re: Linux-2.4.20-pre4-ac1: i810_audio broken

[Juergen Sawinski <juergen.sawinski@mpimf-heidelberg.mpg.de>]

[-- Attachment #1: Type: text/plain, Size: 4335 bytes --]

I can't see any reason, why my i845 updates should touch dma stuff... 

The only changes I could think of causing troubles are 
1. dma reset (line 790, 842): no check if dma reset is complete
2. dma reset: reset occurs before latent interrupt bits are cleared
3. channel support (lines 1951 ff): more bits are cleared then should
   be? (look for 0xcfffff, shouldn't that read 0xffcfffff; compare
   lines 2829 and 2487)
4. At least on ICH4, OFF_CIV is a read only location, but writes
   occur quite frequently.

Don't know if that helps.

George

P.S.: The attached patch goes to Alan seperately. Give it a try. (You
can also remove the comment tags from line 795:
	while( inb(card->iobase + PI_CR) & 0x02 ) ;
and line 849
	while( inb(card->iobase + PO_CR) & 0x02 ) ;
They are  not necessary on my box, but...
)
	
On Tue, 2002-08-27 at 20:46, Doug Ledford wrote:
> On Tue, Aug 27, 2002 at 12:53:12PM +0300, Andris Pavenis wrote:
> > Found that i810_audio has been broken in kernel 2.4.20-pre4-ac1. It was Ok with 
> > 2.4.20-pre1-ac1 I used before.
> > 
> > With 2.4.20-pre4-ac1 I'm only getting garbled sound and kernel messages (see below).
> > Didn't have time yet to study mire detailed which change breaks driver.
> 
> The important part of my change is just two lines.  There is the line that 
> prints out the message "Defaulting to base 2 channel mode." and the line 
> after it where we mask off a couple bits in the global control register.  
> Comment those two lines out and let me know if it makes a difference on 
> your machine.
> 
> > In Alan's changelog I see:
> > 
> > 2.4.20-pre2-ac5: Further i810_audio updates for 845 (Juergen Sawinski) 
> > 2.4.20-pre1-ac3: Tidy up error paths on i810_audio init (Alan) 
> > 2.4.20-pre1-ac2: First set of i810 audio updates (Doug Ledford) 
> > 
> > Andris
> > 
> > ------ at startup -----------------
> > Intel 810 + AC97 Audio, version 0.22, 11:18:00 Aug 26 2002
> > PCI: Found IRQ 5 for device 00:1f.5
> > PCI: Sharing IRQ 5 with 00:1f.3
> > PCI: Setting latency timer of device 00:1f.5 to 64
> > i810: Intel ICH 82801AA found at IO 0xe100 and 0xe000, MEM 0x0000 and 0x0000, IRQ 5
> > i810_audio: Audio Controller supports 2 channels.
> > i810_audio: Defaulting to base 2 channel mode.
> > i810_audio: resetting hw channel 0
> > ac97_codec: AC97 Audio codec, id: 0x4144:0x5348 (Analog Devices AD1881A)
> > i810_audio: AC'97 codec 0 Unable to map surround DAC's (or DAC's not present), to
> > ------ error messages  ------------
> > i810_audio: DMA overrun on write
> > i810_audio: CIV 0, LVI 27, hwptr 253a, count -13626
> > i810_audio: DMA overrun on write
> > i810_audio: CIV 0, LVI 27, hwptr 2662, count -296
> > i810_audio: DMA overrun on write
> > i810_audio: CIV 0, LVI 27, hwptr 2662, count -296
> > i810_audio: DMA overrun on write
> > i810_audio: CIV 0, LVI 27, hwptr 266a, count -8
> > i810_audio: DMA overrun on write
> > i810_audio: CIV 1, LVI 31, hwptr 2924, count -10526
> > i810_audio: DMA overrun on write
> > i810_audio: CIV 1, LVI 31, hwptr 2924, count -10526
> > i810_audio: DMA overrun on write
> > i810_audio: CIV 0, LVI 3, hwptr 253a, count -5434
> > i810_audio: DMA overrun on write
> > i810_audio: CIV 0, LVI 3, hwptr 2562, count -40
> > ......
> > 
> > ---------  error message from artsd (KDE-3.1 beta1) -------
> > Sound server fatal error:
> > AudioSubSystem::handleIO: write failed
> > len = 3228, can_write = 4096, errno = 17 (File exists)
> > This might be a sound hardware/driver specific problem (see aRts FAQ)
> > 
> > -------------------------------------------------------------------- 
> > Kernel was compiled with gcc-3.1 (like earlier kernels where i810_audio worked
> > Ok)
> 
> -- 
>   Doug Ledford <dledford@redhat.com>     919-754-3700 x44233
>          Red Hat, Inc. 
>          1801 Varsity Dr.
>          Raleigh, NC 27606
>   
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
-- 
Juergen "George" Sawinski
Max-Planck Institute for Medical Research
Dept. of Biomedical Optics
Jahnstr. 29
D-69120 Heidelberg
Germany

Phone:  +49-6221-486-308
Fax:    +49-6221-486-325

priv.
Phone:  +49-6221-418 858
Mobile: +49-171-532 5302


[-- Attachment #2: diff-pre4-ac2-jsaw --]
[-- Type: text/plain, Size: 7296 bytes --]

--- linux-2.4.20-pre4-ac2/drivers/sound/i810_audio.c	Tue Aug 27 21:44:25 2002
+++ jsaw/drivers/sound/i810_audio.c	Wed Aug 28 01:26:28 2002
@@ -69,12 +69,12 @@
  *
  *  ICH 4 caveats
  *
- *      The ICH4 has the feature, that the codec ID may not be congruent 
- *      with the AC-link channel.
+ *      The ICH4 has the feature, that the codec ID doesn't have to be 
+ *      congruent with the IO connection.
  * 
- *      Right now, the codec ID is not the real codec ID but the AC-link
- *      channel. A ID <-> AC-link mapping has still to be implemented.
- *      
+ *      Therefore, from driver version 0.23 on, there is a "codec ID" <->
+ *      "IO register base offset" mapping (card->ac97_id_map) field.
+ *   
  *      Juergen "George" Sawinski (jsaw) 
  */
  
@@ -207,7 +207,7 @@
 	CAS	 = 	0x34			/* Codec Write Semaphore Register */
 };
 
-ENUM_ENGINE(MC2,4);     /* Mic. 2 */
+ENUM_ENGINE(MC2,4);     /* Mic In 2 */
 ENUM_ENGINE(PI2,5);     /* PCM In 2 */
 ENUM_ENGINE(SP,6);      /* S/PDIF */
 
@@ -234,8 +234,7 @@
 #define INT_GPI		(1<<0)
 #define INT_MASK (INT_SEC|INT_PRI|INT_MC|INT_PO|INT_PI|INT_MO|INT_NI|INT_GPI)
 
-
-#define DRIVER_VERSION "0.22"
+#define DRIVER_VERSION "0.23"
 
 /* magic numbers to protect our data structures */
 #define I810_CARD_MAGIC		0x5072696E /* "Prin" */
@@ -295,7 +294,7 @@
 	/*@FIXME to be verified*/	{  2, 0x0000 }, /* SI7012 */
 	/*@FIXME to be verified*/	{  2, 0x0000 }, /* NVIDIA_NFORCE */
 	/*@FIXME to be verified*/	{  2, 0x0000 }, /* AMD768 */
-	/*@FIXME to be verified*/	{  2, 0x0000 }, /* AMD8111 */
+	/*@FIXME to be verified*/	{  3, 0x0001 }, /* AMD8111 */
 };
 
 static struct pci_device_id i810_pci_tbl [] __initdata = {
@@ -417,6 +416,7 @@
 	int dev_audio;
 
 	/* structures for abstraction of hardware facilities, codecs, banks and channels*/
+	u16    ac97_id_map[NR_AC97];
 	struct ac97_codec *ac97_codec[NR_AC97];
 	struct i810_state *states[NR_HW_CH];
 	struct i810_channel *channel;	/* 1:1 to states[] but diff. lifetime */
@@ -449,6 +449,9 @@
 	int initializing;
 };
 
+/* extract register offset from codec struct */
+#define IO_REG_OFF(codec) (((struct i810_card *) codec->private_data)->ac97_id_map[codec->id])
+
 static struct i810_card *devs = NULL;
 
 static int i810_open_mixdev(struct inode *inode, struct file *file);
@@ -1948,13 +1955,13 @@
 
 		switch ( val ) {
 			case 2: /* 2 channels is always supported */
-				outl(i_glob_cnt & 0xcfffff,
+				outl(i_glob_cnt & 0xffcfffff,
 				     state->card->iobase + GLOB_CNT);
 				/* Do we need to change mixer settings????  */
 				break;
 			case 4: /* Supported on some chipsets, better check first */
 				if ( state->card->channels >= 4 ) {
-					outl((i_glob_cnt & 0xcfffff) | 0x100000,
+					outl((i_glob_cnt & 0xffcfffff) | 0x100000,
 					      state->card->iobase + GLOB_CNT);
 					/* Do we need to change mixer settings??? */
 				} else {
@@ -1963,7 +1970,7 @@
 				break;
 			case 6: /* Supported on some chipsets, better check first */
 				if ( state->card->channels >= 6 ) {
-					outl((i_glob_cnt & 0xcfffff) | 0x200000,
+					outl((i_glob_cnt & 0xffcfffff) | 0x200000,
 					      state->card->iobase + GLOB_CNT);
 					/* Do we need to change mixer settings??? */
 				} else {
@@ -2553,8 +2560,7 @@
 {
 	struct i810_card *card = dev->private_data;
 	int count = 100;
-	u16 reg_set = ((u16) reg) & 0x7f;
-	reg_set |= ((u16) dev->id) << 7;
+	u16 reg_set = IO_REG_OFF(dev) | (reg&0x7f);
 	
 	while(count-- && (readb(card->iobase_mmio + CAS) & 1)) 
 		udelay(1);
@@ -2574,7 +2580,7 @@
 {
 	struct i810_card *card = dev->private_data;
 	int count = 100;
-	u8 reg_set = ((dev->id)?((reg&0x7f)|0x80):(reg&0x7f));
+	u16 reg_set = IO_REG_OFF(dev) | (reg&0x7f);
 	
 	while(count-- && (inb(card->iobase + CAS) & 1)) 
 		udelay(1);
@@ -2586,8 +2592,7 @@
 {
 	struct i810_card *card = dev->private_data;
 	int count = 100;
-	u16 reg_set = ((u16) reg) & 0x7f;
-	reg_set |= ((u16) dev->id) << 7;
+	u16 reg_set = IO_REG_OFF(dev) | (reg&0x7f);
 	
 	while(count-- && (readb(card->iobase_mmio + CAS) & 1)) 
 		udelay(1);
@@ -2603,7 +2608,7 @@
 {
 	struct i810_card *card = dev->private_data;
 	int count = 100;
-	u8 reg_set = ((dev->id)?((reg&0x7f)|0x80):(reg&0x7f));
+	u16 reg_set = IO_REG_OFF(dev) | (reg&0x7f);
 	
 	while(count-- && (inb(card->iobase + CAS) & 1)) 
 		udelay(1);
@@ -2779,7 +2784,7 @@
 	if ((card->pci_id == PCI_DEVICE_ID_INTEL_ICH4)
 	    && (card->use_mmio)) {
 		primary_codec_id = (int) readl(card->iobase_mmio + SDM) & 0x3;
-		printk(KERN_INFO "i810_audio: primary codec id %d\n",
+		printk(KERN_INFO "i810_audio: Primary codec has ID %d\n",
 		       primary_codec_id);
 	}
 
@@ -2803,6 +2808,7 @@
 	int num_ac97 = 0;
 	int ac97_id;
 	int total_channels = 0;
+	int nr_ac97_max = card_cap[card->pci_id_internal].nr_ac97;
 	struct ac97_codec *codec;
 	u16 eid;
 	u32 reg;
@@ -2828,13 +2834,15 @@
 	reg = inl(card->iobase + GLOB_CNT);
 	outl(reg & 0xffcfffff, card->iobase + GLOB_CNT);
 		
-	for (num_ac97 = 0; num_ac97 < card_cap[card->pci_id_internal].nr_ac97; num_ac97++) {
+	for (num_ac97 = 0; num_ac97 < NR_AC97; num_ac97++) 
 		card->ac97_codec[num_ac97] = NULL;
-	}
 
-	for (num_ac97 = 0; num_ac97 < card_cap[card->pci_id_internal].nr_ac97; num_ac97++) {
+	/*@FIXME I don't know, if I'm playing to safe here... (jsaw) */
+	if ((nr_ac97_max > 2) && !card->use_mmio) nr_ac97_max = 2;
+
+	for (num_ac97 = 0; num_ac97 < nr_ac97_max; num_ac97++) {
 		/* codec reset */
-		printk(KERN_INFO "i810_audio: resetting hw channel %d\n", num_ac97);
+		printk(KERN_INFO "i810_audio: Resetting connection %d\n", num_ac97);
 		if (card->use_mmio) readw(card->ac97base_mmio + 0x80*num_ac97);
 		else inw(card->ac97base + 0x80*num_ac97);
 
@@ -2846,7 +2854,7 @@
 		if ((card->pci_id == PCI_DEVICE_ID_INTEL_ICH4)
 		    && (card->use_mmio)) {
 			ac97_id = (int) readl(card->iobase_mmio + SDM) & 0x3;
-			printk(KERN_INFO "i810_audio: hw channel %d, codec id %d\n",
+			printk(KERN_INFO "i810_audio: Connection %d with codec id %d\n",
 			       num_ac97, ac97_id);
 		}
 		else {
@@ -2869,9 +2877,8 @@
 		/* initialize some basic codec information, other fields will be filled
 		   in ac97_probe_codec */
 		codec->private_data = card;
-
-		/*@FIXME this will lead to problems!!! id=2 <-> io offset=0*/
-		codec->id = num_ac97;
+		codec->id = ac97_id;
+		card->ac97_id_map[ac97_id] = num_ac97 * 0x80;
 
 		if (card->use_mmio) {	
 			codec->codec_read = i810_ac97_get_mmio;
@@ -2883,7 +2890,7 @@
 		}
 	
 		if(!i810_ac97_probe_and_powerup(card,codec)) {
-			printk(KERN_ERR "i810_audio: timed out waiting for codec %d analog ready.\n", num_ac97);
+			printk(KERN_ERR "i810_audio: timed out waiting for codec %d analog ready.\n", ac97_id);
 			kfree(codec);
 			break;	/* it didn't work */
 		}
@@ -2903,7 +2910,7 @@
 		codec->codec_write(codec, AC97_EXTENDED_MODEM_ID, 0L);
 		if(codec->codec_read(codec, AC97_EXTENDED_MODEM_ID))
 		{
-			printk(KERN_WARNING "i810_audio: codec %d is a softmodem - skipping.\n", num_ac97);
+			printk(KERN_WARNING "i810_audio: codec %d is a softmodem - skipping.\n", ac97_id);
 			kfree(codec);
 			continue;
 		}