arch/alpha/kernel/sys_sx164.c
Earlier firmware revisions need MVI fix as well.
arch/alpha/kernel/sys_nautilus.c
On UP1500 firmware reports wrong AGP IRQ (10 instead of 5).
This causes interrupt storm if there is a PCI device that
uses IRQ 5.
Signed-off-by: Ivan Kokshaysky <[EMAIL PROTECTED]>
--- linux.orig/arch/alpha/kernel/sys_sx164.c Thu Jun 1 13:46:21 2006
+++ linux/arch/alpha/kernel/sys_sx164.c Thu Jun 1 13:46:11 2006
@@ -132,7 +132,7 @@ sx164_init_arch(void)
if (amask(AMASK_MAX) != 0
&& alpha_using_srm
- && (cpu->pal_revision & 0xffff) == 0x117) {
+ && (cpu->pal_revision & 0xffff) <= 0x117) {
__asm__ __volatile__(
"lda $16,8($31)\n"
"call_pal 9\n" /* Allow PALRES insns in kernel mode */
--- linux.orig/arch/alpha/kernel/sys_nautilus.c Tue Apr 10 13:59:43 2007
+++ linux/arch/alpha/kernel/sys_nautilus.c Tue Apr 10 14:29:55 2007
@@ -70,6 +70,12 @@ nautilus_map_irq(struct pci_dev *dev, u8
/* Preserve the IRQ set up by the console. */
u8 irq;
+ /* UP1500: AGP INTA is actually routed to IRQ 5, not IRQ 10 as
+ console reports. Check the device id of AGP bridge to distinguish
+ UP1500 from UP1000/1100. Note: 'pin' is 2 due to bridge swizzle. */
+ if (slot == 1 && pin == 2 &&
+ dev->bus->self && dev->bus->self->device == 0x700f)
+ return 5;
pci_read_config_byte(dev, PCI_INTERRUPT_LINE, &irq);
return irq;
}
-
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/
