Skip to contentSkip to conversation

Aureal Vortex PATCH [Fwd: via workaround error]

PATCH Manuel Jander 3 participants 5 messages Jan 10, 2004 – Jan 13, 2004

lore
Share & tools
lore
#0 lore raw

Please apply to CVS. Thanks

-----Forwarded Message-----
From: Wilfried Weissmann <[email protected]>
To: [email protected], [email protected]
Subject: via workaround error
Date: Fri, 09 Jan 2004 20:16:01 +0100

i made a stupid error when i wrote the via workaround patch. the pci latency for the vortex was still set to 0xff when using autodetection and no via chipset was found. i discovered this after i exchanged my motherboard from a via kt133 to a sis 735. the fix is attached. the bug should not have caused any troubles but it is just not right...

bye,
wilfried

________________________________________________________________________

pci/au88x0/au88x0.c +4 -0
@@ -77,6 +77,10 @@ static void __devinit snd_vortex_workaro
 		}
 		break;
 	}
+
+		/* do not do anything if autodetection was enabled and found no VIA */
+	if(fix == 255)
+		return;
 	
 	int rc;
 

-------------------------------------------------------
This SF.net email is sponsored by: Perforce Software. Perforce is the Fast Software Configuration Management System offering advanced branching capabilities and atomic changes on 50+ platforms. Free Eval! http://www.perforce.com/perforce/loadprog.html

#1 lore raw

At Fri, 09 Jan 2004 23:04:49 -0400,

Quoted 9 lines · Manuel Jander

Please apply to CVS. Thanks

-----Forwarded Message-----
From: Wilfried Weissmann <[email protected]>
To: [email protected], [email protected]
Subject: via workaround error
Date: Fri, 09 Jan 2004 20:16:01 +0100

Quoted 5 lines

i made a stupid error when i wrote the via workaround patch. the pci latency for the vortex was still set to 0xff when using autodetection and no via chipset was found. i discovered this after i exchanged my motherboard from a via kt133 to a sis 735. the fix is attached. the bug should not have caused any troubles but it is just not right...

i didn't receive the VIA workaround patch itself yet. could you send the whole patch?

thanks,

Takashi

-------------------------------------------------------
This SF.net email is sponsored by: Perforce Software. Perforce is the Fast Software Configuration Management System offering advanced branching capabilities and atomic changes on 50+ platforms. Free Eval! http://www.perforce.com/perforce/loadprog.html

#2 lore raw
Quoted 11 lines · Takashi Iwai

At Fri, 09 Jan 2004 23:04:49 -0400,

Quoted 8 lines · Manuel Jander

Please apply to CVS. Thanks

-----Forwarded Message-----
From: Wilfried Weissmann <[email protected]>
To: [email protected], [email protected]
Subject: via workaround error
Date: Fri, 09 Jan 2004 20:16:01 +0100

Quoted 5 lines
Quoted 1 line

i made a stupid error when i wrote the via workaround patch. the pci latency for the vortex was still set to 0xff when using autodetection and no via chipset was found. i discovered this after i exchanged my motherboard from a via kt133 to a sis 735. the fix is attached. the bug should not have caused any troubles but it is just not right...

i didn't receive the VIA workaround patch itself yet. could you send the whole patch?

here we go...

pci/au88x0/au88x0.c +55 -0
@@ -26,6 +26,7 @@
 static int index[SNDRV_CARDS] = SNDRV_DEFAULT_IDX;
 static char *id[SNDRV_CARDS] = SNDRV_DEFAULT_STR;
 static int enable[SNDRV_CARDS] = SNDRV_DEFAULT_ENABLE_PNP;
+static int pcifix[SNDRV_CARDS] = { [0 ... (SNDRV_CARDS-1)] = 255 };
 
 MODULE_PARM(index, "1-" __MODULE_STRING(SNDRV_CARDS) "i");
 MODULE_PARM_DESC(index, "Index value for " CARD_NAME " soundcard.");
@@ -36,6 +37,9 @@ MODULE_PARM_SYNTAX(id, SNDRV_ID_DESC);
 MODULE_PARM(enable, "1-" __MODULE_STRING(SNDRV_CARDS) "i");
 MODULE_PARM_DESC(enable, "Enable " CARD_NAME " soundcard.");
 MODULE_PARM_SYNTAX(enable, SNDRV_ENABLE_DESC);
+MODULE_PARM(pcifix, "1-255i");
+MODULE_PARM_DESC(pcifix, "Enable VIA-workaround for " CARD_NAME " soundcard.");
+MODULE_PARM_SYNTAX(pcifix, SNDRV_ENABLED ",allows:{{0,Disabled},{1,Latency},{2,Bridge},{3,Both},{255,Auto}},default:4,dialog:check");
 
 MODULE_DESCRIPTION("Aureal vortex");
 MODULE_CLASSES("{sound}");
@@ -61,6 +65,56 @@ __setup("snd-au88x0=", alsa_card_vortex_
 
 MODULE_DEVICE_TABLE(pci, snd_vortex_ids);
 
+static void __devinit snd_vortex_workaround(struct pci_dev *vortex, int fix) {
+	struct pci_dev *via=NULL;
+		/* autodetect if workarounds are required */
+	while( (via = pci_find_device(PCI_VENDOR_ID_VIA,
+				PCI_DEVICE_ID_VIA_8365_1, via)) ) {
+		if(fix == 255) {
+			printk(KERN_INFO CARD_NAME
+				": detected VIA KT133/KM133. activating workaround...\n");
+			fix = 3; // do latency and via bridge workaround
+		}
+		break;
+	}
+	
+	int rc;
+
+		/* fix vortex latency */
+	if(fix & 0x01) {
+		if( !(rc = pci_write_config_byte(vortex, 0x40, 0xff)) ) {
+			printk(KERN_INFO CARD_NAME
+					": vortex latency is 0xff\n");
+		}
+		else {
+			printk(KERN_WARNING CARD_NAME ": could not set vortex latency: pci error 0x%x\n", rc);
+		}
+	}
+
+		/* fix via agp bridge */
+	if(via && (fix & 0x02)) {
+		u8 value;
+
+			/*
+			 * only set the bit (Extend PCI#2 Internal Master for
+			 * Efficient Handling of Dummy Requests) if the can
+			 * read the config and it is not already set
+			 */
+
+		if( !(rc = pci_read_config_byte(via, 0x42, &value)) && (
+			(value & 0x10) ||
+			!(rc=pci_write_config_byte(via, 0x42, value|0x10)) ) ) {
+
+			printk(KERN_INFO CARD_NAME
+					": bridge config is 0x%x\n",
+					value|0x10);
+		}
+		else {
+			printk(KERN_WARNING CARD_NAME ": could not set vortex latency: pci error 0x%x\n", rc);
+		}
+	}
+}
+
 
 // component-destructor
 // (see "Management of Cards and Components")
@@ -198,6 +252,7 @@ snd_vortex_probe(struct pci_dev *pci, co
         snd_card_free(card);
         return err;
     }
+    snd_vortex_workaround(pci, pcifix[dev]);
     // (4) Alloc components.
     // ADB pcm.
     if ((err = snd_vortex_new_pcm(chip, VORTEX_PCM_ADB, NR_ADB)) < 0) {
#3 lore raw

At Mon, 12 Jan 2004 19:08:21 +0100,

Quoted 21 lines · Wilfried Weissmann

[1 <text/plain; us-ascii (7bit)>]

Quoted 18 lines · Takashi Iwai

At Fri, 09 Jan 2004 23:04:49 -0400,

Quoted 13 lines · Manuel Jander

Please apply to CVS. Thanks

-----Forwarded Message-----
From: Wilfried Weissmann <[email protected]>
To: [email protected], [email protected]
Subject: via workaround error
Date: Fri, 09 Jan 2004 20:16:01 +0100

i made a stupid error when i wrote the via workaround patch. the pci latency for the vortex was still set to 0xff when using autodetection and no via chipset was found. i discovered this after i exchanged my motherboard from a via kt133 to a sis 735. the fix is attached. the bug should not have caused any troubles but it is just not right...

Quoted 4 lines
Quoted 2 lines

i didn't receive the VIA workaround patch itself yet. could you send the whole patch?

here we go...

thanks!

pci/au88x0/au88x0.c __setup("snd-au88x0=", alsa_card_vortex_ · line 79

Quoted 12 lines
pci/au88x0/au88x0.c +12 -0
+static void __devinit snd_vortex_workaround(struct pci_dev *vortex, int fix) {
+	struct pci_dev *via=NULL;
+		/* autodetect if workarounds are required */
+	while( (via = pci_find_device(PCI_VENDOR_ID_VIA,
+				PCI_DEVICE_ID_VIA_8365_1, via)) ) {
+		if(fix == 255) {
+			printk(KERN_INFO CARD_NAME
+				": detected VIA KT133/KM133. activating workaround...\n");
+			fix = 3; // do latency and via bridge workaround
+		}
+		break;
+	}

i think "if" is enough in the above.

pci/au88x0/au88x0.c __setup("snd-au88x0=", alsa_card_vortex_ · line 81

Quoted 2 lines
pci/au88x0/au88x0.c +2 -0
+	
+	int rc;

gcc-2.9.x won't accept this.

i applied the fixed version to cvs. please check it later.

could you also provide the patch to INSTALL (document) file?

thanks,

Takashi

-------------------------------------------------------
This SF.net email is sponsored by: Perforce Software. Perforce is the Fast Software Configuration Management System offering advanced branching capabilities and atomic changes on 50+ platforms. Free Eval! http://www.perforce.com/perforce/loadprog.html

#4 lore raw

Takashi Iwai wrote:
[snip]

Quoted 1 line

i applied the fixed version to cvs. please check it later.

the changes are fine.

Quoted 2 lines

could you also provide the patch to INSTALL (document) file?

the patch is attached (against current alsa CVS)...

Quoted 5 lines

thanks,

Takashi

greetings,
wilfried

Index: INSTALL
===================================================================
RCS file: /cvsroot/alsa/alsa-driver/INSTALL,v
retrieving revision 1.154

INSTALL +8 -0
diff -u -r1.154 INSTALL
@@ -261,6 +261,14 @@
            control correctly. If you have problems regarding this, try
            another ALSA compliant mixer (alsamixer works).
 
+    pcifix	- Control PCI workarounds
+		  0 = Disable all workarounds
+		  1 = Force the PCI latency of the Aureal card to 0xff
+		  2 = Force the Extend PCI#2 Internal Master for Efficient
+		      Handling of Dummy Requests on the VIA KT133 AGP Bridge
+		  3 = Force both settings
+		  255 = Autodetect what is required (default)
+
 
 Trouble Shooting
 ================