Current-Users archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: ACPI call for testing

Hi again,

On my shiny new ASUS M4A89TD PRO/USB3, the patch does indeed help a bit,
albeit there are some problems remaining. Without the patch, the
system boots up, but it looses quite a lot of timer interrupts (or so
it looks.) Keyboard repeat is quite intermittent, console beeps last
way too long, and the clock looses quite some time. Booting without acpi
makes the box work quite well. With acpi, using your patch, it seems
that the JMicron SATA-Controller doesn't get any interrupts, but on the SCI
(int 9) I get some form of interrupt storm, i.e. more than 90,000
interrupts per second, which more or less completely locks up one of the
cores on the machine. Note that I had to disable pciide and jmide in order
to get the system to boot with this.

Turns out that the patch is actually the culprit responsible for the
interrupt storm.  The problem is that with the patch, the ACPI-Interrupt
(SCI) is enabled before the MADT with the interrupt redirection information
is analyzed.  On this board, this results in the SCI to be configured in
the IOAPIC with level triggered, active high instead of level triggered,
active low.  The fact that the system seems to run better with the patch
is just a side effect of this interrupt storm.

So it's back to square one, unfortunately ;-(.

--                          Wolfgang Solfrank

Home | Main Index | Thread Index | Old Index