VIA Northing workaround /causing/ problems
I recently upgraded my kernel from 2.4.10-pre6 to 2.4.18-pre2. After doing so, X acted extremely weird; whenever just about anything happened, lines would appear across the screen, almost like static.
After playing around with a few config options that I'd changed, with no results, I noticed the message about the VIA northbridge bug in dmesg. I commented out the line listing this chipset in pci-pc.c, recompiled, and sure enough that fixed the problem!
This board is based on the KT33 chipset. If anyone would like more information, email me.
Signature · 6 lines
-Steven In a time of universal deceit, telling the truth is a revolutionary act. -- George Orwell He's alive. He's alive! Oh, that fellow at RadioShack said I was mad! Well, who's mad now? -- Montgomery C. Burns
Quoted 11 lines · Steven Walter
I recently upgraded my kernel from 2.4.10-pre6 to 2.4.18-pre2. After doing so, X acted extremely weird; whenever just about anything happened, lines would appear across the screen, almost like static.
After playing around with a few config options that I'd changed, with no results, I noticed the message about the VIA northbridge bug in dmesg. I commented out the line listing this chipset in
pci-pc.c, recompiled, and sure enough that fixed the problem!This board is based on the KT33 chipset. If anyone would like more information, email me.
Can you play with it a bit more?
Go to that file, uncomment it back, fiddle with pci_fixup_via_northbridge_bug(): try to clear only bit 7,
then only 7 and 6 and see which cause it...
Make it print reg#, old, new contents:
...printk("Trying to stomp on VIA Northbridge bug: [%02x] %02x->%02x\n", where, v, v & 0x1f);
...
etc.
Original function for your reference:
static void __init pci_fixup_via_northbridge_bug(struct pci_dev *d)
{
u8 v;
int where = 0x55;
if (d->device == PCI_DEVICE_ID_VIA_8367_0) {
where = 0x95; /* the memory write queue timer register is
different for the kt266x's: 0x95 not 0x55 */
}
pci_read_config_byte(d, where, &v);
if (v & 0xe0) {
printk("Trying to stomp on VIA Northbridge bug...\n");
v &= 0x1f; /* clear bits 5, 6, 7 */
pci_write_config_byte(d, where, v);
}
}--
vda
Quoted 10 lines · Denis Vlasenko
Quoted 9 lines · Steven Walter
I recently upgraded my kernel from 2.4.10-pre6 to 2.4.18-pre2. After doing so, X acted extremely weird; whenever just about anything happened, lines would appear across the screen, almost like static.
After playing around with a few config options that I'd changed, with no results, I noticed the message about the VIA northbridge bug in dmesg. I commented out the line listing this chipset in
pci-pc.c, recompiled, and sure enough that fixed the problem!
Quoted 8 lines
Quoted 2 lines
This board is based on the KT33 chipset. If anyone would like more information, email me.
Can you play with it a bit more?
Go to that file, uncomment it back, fiddle withpci_fixup_via_northbridge_bug(): try to clear only bit 7, then only 7 and 6 and see which cause it...
Make it print reg#, old, new contents:
Yes, this is a good point. The via 'specs' (really they may not even be from via.. but that's a different story) say that bits 5,6,7 are for the MWQ Timer, but some people have reported really good results just clearing bit 5 and leaving 6 and 7 alone (actually this was the original MWQ timer patch). I would be interested to see what bits are on for this guy by default from his bios and which one causes problems when touched...
-Calin
Quoted 29 lines
...
printk("Trying to stomp on VIA Northbridge bug: [%02x] %02x->%02x\n", where, v, v & 0x1f);
...
etc.
Original function for your reference:static void
__initpci_fixup_via_northbridge_bug(struct pci_dev *d){ u8 v; int where = 0x55; if (d->device == PCI_DEVICE_ID_VIA_8367_0) { where = 0x95; /* the memory write queue timer register is different for the kt266x's: 0x95 not 0x55 */ } pci_read_config_byte(d, where, &v); if (v & 0xe0) { printk("Trying to stomp on VIA Northbridge bug...\n"); v &= 0x1f; /* clear bits 5, 6, 7 */ pci_write_config_byte(d, where, v); } }--
vda
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/