On Tuesday 11 December 2012 20:07:50 H. Peter Anvin wrote: > On 12/11/2012 10:59 AM, Ondrej Zary wrote: > > As I said before, the BIOS probably breaks on resume before any Linux > > code is executed. So any fix must be done before suspending. > > Well, that code should be independent of A20, being in low memory, but > if the BIOS itself crashes, then that's... yeah. > > > I hate quirks too. A general solution would be to always enable A20 using > > KBC (if KBC is present) but that's probably not acceptable. > > I don't see why not. If so we could just do it as part of the > initialization of the i8042 driver. Something like this? (It works.) diff --git a/drivers/input/serio/i8042-x86ia64io.h b/drivers/input/serio/i8042-x86ia64io.h index d6cc77a..0807ac7 100644 --- a/drivers/input/serio/i8042-x86ia64io.h +++ b/drivers/input/serio/i8042-x86ia64io.h @@ -921,6 +921,7 @@ static int __init i8042_platform_init(void) int retval; #ifdef CONFIG_X86 + u8 a20_on = 0xdf; /* Just return if pre-detection shows no i8042 controller exist */ if (!x86_platform.i8042_detect()) return -ENODEV; @@ -960,6 +961,13 @@ static int __init i8042_platform_init(void) if (dmi_check_system(i8042_dmi_dritek_table)) i8042_dritek = true; + + /* + * A20 was already enabled during early kernel init. But some buggy + * BIOSes (in MSI Laptops) require A20 to be enabled using 8042 to + * resume from S3. So we do it here and hope that nothing breaks. + */ + i8042_command(&a20_on, 0x10d1); #endif /* CONFIG_X86 */ return retval; -- Ondrej Zary -- To unsubscribe from this list: send the line "unsubscribe linux-acpi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
