Regression with commit c3bc26d4b4e3

Regression with commit c3bc26d4b4e3

[Jan Kara]

Hello,

commit c3bc26d4b4e3 (ACPICA: ACPI 2.0, Hardware: Add access_width/bit_offset
support in acpi_hw_read()) broke booting of my KVM test guest. When I
revert that commit and commit 66b1ed5aa8d (ACPICA: ACPI 2.0, Hardware: Add
access_width/bit_offset support for acpi_hw_write()) which looks similar,
the vm boots again. The problem manifests itself that the qemu-kvm just
aborts after the kernel prints:

[    0.333707] ACPI: bus type PCI registered
[    0.335494] PCI: Using configuration type 1 for base access
[    0.472716] ACPI: Added _OSI(Module Device)
[    0.475360] ACPI: Added _OSI(Processor Device)
[    0.476012] ACPI: Added _OSI(3.0 _SCP Extensions)
[    0.480035] ACPI: Added _OSI(Processor Aggregator Device)
[    0.494742] ACPI: 2 ACPI AML tables successfully acquired and loaded
[    0.496003] 

For working kernel the messages continue like:

[    0.507616] ACPI: Interpreter enabled
[    0.508018] ACPI: (supports S0 S5)
[    0.509129] ACPI: Using IOAPIC for interrupt routing
[    0.510717] PCI: Using host bridge windows from ACPI; if necessary, use "pci=nocrs" and report a bug
[    0.533308] ACPI: PCI Root Bridge [PCI0] (domain 0000 [bus 00-ff])
[    0.535049] acpi PNP0A03:00: _OSC: OS supports [Segments]
[    0.536111] acpi PNP0A03:00: _OSC failed (AE_NOT_FOUND); disabling ASPM
[    0.538702] PCI host bridge to bus 0000:00

								Honza

-- 
Jan Kara <jack@suse.com>
SUSE Labs, CR

Re: Regression with commit c3bc26d4b4e3

[Rafael J. Wysocki]

On Thu, Jun 2, 2016 at 1:08 PM, Jan Kara <jack@suse.cz> wrote:
> Hello,
>
> commit c3bc26d4b4e3 (ACPICA: ACPI 2.0, Hardware: Add access_width/bit_offset
> support in acpi_hw_read()) broke booting of my KVM test guest. When I
> revert that commit and commit 66b1ed5aa8d (ACPICA: ACPI 2.0, Hardware: Add
> access_width/bit_offset support for acpi_hw_write()) which looks similar,
> the vm boots again. The problem manifests itself that the qemu-kvm just
> aborts after the kernel prints:
>
> [    0.333707] ACPI: bus type PCI registered
> [    0.335494] PCI: Using configuration type 1 for base access
> [    0.472716] ACPI: Added _OSI(Module Device)
> [    0.475360] ACPI: Added _OSI(Processor Device)
> [    0.476012] ACPI: Added _OSI(3.0 _SCP Extensions)
> [    0.480035] ACPI: Added _OSI(Processor Aggregator Device)
> [    0.494742] ACPI: 2 ACPI AML tables successfully acquired and loaded
> [    0.496003]
>
> For working kernel the messages continue like:
>
> [    0.507616] ACPI: Interpreter enabled
> [    0.508018] ACPI: (supports S0 S5)
> [    0.509129] ACPI: Using IOAPIC for interrupt routing
> [    0.510717] PCI: Using host bridge windows from ACPI; if necessary, use "pci=nocrs" and report a bug
> [    0.533308] ACPI: PCI Root Bridge [PCI0] (domain 0000 [bus 00-ff])
> [    0.535049] acpi PNP0A03:00: _OSC: OS supports [Segments]
> [    0.536111] acpi PNP0A03:00: _OSC failed (AE_NOT_FOUND); disabling ASPM
> [    0.538702] PCI host bridge to bus 0000:00

We have a fix in this area,
https://patchwork.kernel.org/patch/9145955/, does it help?

Re: Regression with commit c3bc26d4b4e3

[Jan Kara]

On Thu 02-06-16 14:22:24, Rafael J. Wysocki wrote:
> On Thu, Jun 2, 2016 at 1:08 PM, Jan Kara <jack@suse.cz> wrote:
> > Hello,
> >
> > commit c3bc26d4b4e3 (ACPICA: ACPI 2.0, Hardware: Add access_width/bit_offset
> > support in acpi_hw_read()) broke booting of my KVM test guest. When I
> > revert that commit and commit 66b1ed5aa8d (ACPICA: ACPI 2.0, Hardware: Add
> > access_width/bit_offset support for acpi_hw_write()) which looks similar,
> > the vm boots again. The problem manifests itself that the qemu-kvm just
> > aborts after the kernel prints:
> >
> > [    0.333707] ACPI: bus type PCI registered
> > [    0.335494] PCI: Using configuration type 1 for base access
> > [    0.472716] ACPI: Added _OSI(Module Device)
> > [    0.475360] ACPI: Added _OSI(Processor Device)
> > [    0.476012] ACPI: Added _OSI(3.0 _SCP Extensions)
> > [    0.480035] ACPI: Added _OSI(Processor Aggregator Device)
> > [    0.494742] ACPI: 2 ACPI AML tables successfully acquired and loaded
> > [    0.496003]
> >
> > For working kernel the messages continue like:
> >
> > [    0.507616] ACPI: Interpreter enabled
> > [    0.508018] ACPI: (supports S0 S5)
> > [    0.509129] ACPI: Using IOAPIC for interrupt routing
> > [    0.510717] PCI: Using host bridge windows from ACPI; if necessary, use "pci=nocrs" and report a bug
> > [    0.533308] ACPI: PCI Root Bridge [PCI0] (domain 0000 [bus 00-ff])
> > [    0.535049] acpi PNP0A03:00: _OSC: OS supports [Segments]
> > [    0.536111] acpi PNP0A03:00: _OSC failed (AE_NOT_FOUND); disabling ASPM
> > [    0.538702] PCI host bridge to bus 0000:00
> 
> We have a fix in this area,
> https://patchwork.kernel.org/patch/9145955/, does it help?

Yes, this patch makes my KVM guest boot again. Thanks!

								Honza
-- 
Jan Kara <jack@suse.com>
SUSE Labs, CR