qemu-devel
[Top][All Lists]
Advanced

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

Re: [Qemu-devel] [PATCH] Update i440FX/PIIX3 emulation


From: Avi Kivity
Subject: Re: [Qemu-devel] [PATCH] Update i440FX/PIIX3 emulation
Date: Wed, 31 Oct 2007 15:17:04 +0200
User-agent: Thunderbird 2.0.0.5 (X11/20070727)

Michael Hanselmann wrote:

This does not yet remove the workaround introduced by Igor Lvovsky's
patch. However, I'm working on that since it, despite my earlier mail,
seems to help with my ACPI shutdown problem.

So, I found the bug causing this behaviour. It turned out to be a
wrongly named variable in the ACPI DSDT from Bochs. See the patch for
Bochs below. I already sent it to the bochs-developers list[1].
qemu/pc-bios/bios.bin needs to be rebuilt from Bochs' code,
qemu/pc-bios/bios.diff and my patch.

The second patch below reverts the changes made by Igor Lvovsky. After
applying the patch to the BIOS, ACPI IRQs finally reach the system.

Finding this bug took me about the free time of four weeks. However, I
learned a lot about the internals of a PC. :-)

Thanks,
Michael

[1] 
http://sourceforge.net/mailarchive/forum.php?thread_name=20071031000835.GA20915%40hansmi.ch&forum_name=bochs-developers

---
Index: bios/acpi-dsdt.dsl
===================================================================
RCS file: /cvsroot/bochs/bochs/bios/acpi-dsdt.dsl,v
retrieving revision 1.1
diff -u -p -u -p -r1.1 acpi-dsdt.dsl
--- bios/acpi-dsdt.dsl  28 Sep 2006 18:56:20 -0000      1.1
+++ bios/acpi-dsdt.dsl  30 Oct 2007 23:52:22 -0000
@@ -369,7 +369,7 @@ DefinitionBlock (
                 Method (_STA, 0, NotSerialized)
                 {
                     Store (0x0B, Local0)
-                    If (And (0x80, PRQ0, Local1))
+                    If (And (0x80, PRQ0, Local0))
                     {
                          Store (0x09, Local0)
                     }

Can you explain this?

The original code seems to return either 0xb (present, enabled, functional) or 0x9 (present, functional). The new code seems to return either 0x9 (present, functional) or 0 (if PRQ0 had its seventh bit clear).

Am I reading the code incorrectly?

--
Any sufficiently difficult bug is indistinguishable from a feature.





reply via email to

[Prev in Thread] Current Thread [Next in Thread]