On Fri, Mar 22, 2013 at 09:48:18AM +0200, Gleb Natapov wrote:
On Thu, Mar 21, 2013 at 05:17:11PM +0800, Hu Tao wrote:
On Thu, Mar 21, 2013 at 10:11:54AM +0100, Paolo Bonzini wrote:
Il 21/03/2013 10:08, Hu Tao ha scritto:
Signed-off-by: Hu Tao <hutao@cn.fujitsu.com> --- src/acpi-dsdt-isa.dsl | 14 ++++++++++++-- src/acpi.c | 9 +++++++++ 2 files changed, 21 insertions(+), 2 deletions(-)
diff --git a/src/acpi-dsdt-isa.dsl b/src/acpi-dsdt-isa.dsl index 87a31b9..43fe719 100644 --- a/src/acpi-dsdt-isa.dsl +++ b/src/acpi-dsdt-isa.dsl @@ -102,7 +102,9 @@ Scope(\_SB.PCI0.ISA) {
Device(PEVT) { Name(_HID, "QEMU0001") - OperationRegion(PEOR, SystemIO, 0x0505, 0x01) + ACPI_EXTRACT_NAME_WORD_CONST dsdt_isa_pest + Name(PEST, 0x505) + OperationRegion(PEOR, SystemIO, PEST, 0x01) Field(PEOR, ByteAcc, NoLock, Preserve) { PEPT, 8, } @@ -126,7 +128,15 @@ Scope(\_SB.PCI0.ISA) { }
Name(_CRS, ResourceTemplate() { - IO(Decode16, 0x0505, 0x0505, 0x01, 0x01) + IO(Decode16, 0x0505, 0x0505, 0x01, 0x01, IO) }) + + CreateWordField(_CRS, IO._MIN, IOMN) + CreateWordField(_CRS, IO._MAX, IOMX) + + Method(_INI, 0, NotSerialized) { + Store(PEST, IOMN) + Store(PEST, IOMX) + } } } diff --git a/src/acpi.c b/src/acpi.c index 119d1c1..42fa06e 100644 --- a/src/acpi.c +++ b/src/acpi.c @@ -286,11 +286,20 @@ static const struct pci_device_id fadt_init_tbl[] = { PCI_DEVICE_END };
+static void patch_dsdt(void *dsdt) +{ + u8 *dsdt_ptr = dsdt; + int pvpanic_port = romfile_loadint("etc/pvpanic-port", 0x505);
The default must be 0.
Also, here:
+ + Method(_STA, 0, NotSerialized) { + Store(PEPT, Local0) + If (LEqual(Local0, Zero)) { + Return (0x00) + } Else { + Return (0x0F) + } + } +
You must change it to look at PEST instead of PEPT (i.e. do not probe, just see if you have a meaningful address).
Just squash the patches, it's simpler that way probably.
I forgot to add RFC to the title. This patch doesn't work for q35 with custom ioport.
Why doesn't it work with q35?
ACPI_EXTRACT_NAME_WORD_CONST dsdt_isa_pest generates two dsdt_isa_pest, one in out/acpi-dsdt.hex the other in out/q35-acpi-dsdt.hex. We should use the one in q35-acpi-dsdt.hex to patch DSDT for q35. But including q35-acpi-dsdt.hex, along with acpi-dsdt.hex, will result in redefinition of dsdt_isa_pest. I didn't see a simple way to solve this. -- Regards, Hu Tao