Re: [PATCH v2] ACPI: sysfs: Change ACPI_MASKABLE_GPE_MAX to 0x100

[Rafael J. Wysocki]

On Mon, Nov 18, 2019 at 8:44 PM Dexuan-Linux Cui <dexuan.linux@xxxxxxxxx> wrote:
>
> On Mon, Nov 18, 2019 at 1:04 AM Rafael J. Wysocki <rafael@xxxxxxxxxx> wrote:
> >
> > On Sat, Nov 16, 2019 at 9:06 PM Dan Williams <dan.j.williams@xxxxxxxxx> wrote:
> > >
> > > On Wed, Nov 13, 2019 at 11:17 PM Yunfeng Ye <yeyunfeng@xxxxxxxxxx> wrote:
> > > >
> > > > The commit 0f27cff8597d ("ACPI: sysfs: Make ACPI GPE mask kernel
> > > > parameter cover all GPEs") says:
> > > >   "Use a bitmap of size 0xFF instead of a u64 for the GPE mask so 256
> > > >    GPEs can be masked"
> > > >
> > > > But the masking of GPE 0xFF it not supported and the check condition
> > > > "gpe > ACPI_MASKABLE_GPE_MAX" is not valid because the type of gpe is
> > > > u8.
> > > >
> > > > So modify the macro ACPI_MASKABLE_GPE_MAX to 0x100, and drop the "gpe >
> > > > ACPI_MASKABLE_GPE_MAX" check. In addition, update the docs "Format" for
> > > > acpi_mask_gpe parameter.
> > > >
> > > > Fixes: 0f27cff8597d ("ACPI: sysfs: Make ACPI GPE mask kernel parameter cover all GPEs")
> > > > Signed-off-by: Yunfeng Ye <yeyunfeng@xxxxxxxxxx>
> > > > ---
> > > > v1 -> v2:
> > > >  - drop the "gpe > ACPI_MASKABLE_GPE_MAX" check
> > > >  - update the docs "Format" from <int> to <byte> for acpi_mask_gpe parameter
> > > >  - update the commit comment
> > > >
> > > >  Documentation/admin-guide/kernel-parameters.txt | 2 +-
> > > >  drivers/acpi/sysfs.c                            | 4 ++--
> > > >  2 files changed, 3 insertions(+), 3 deletions(-)
> > >
> > > Bisect flags commit eb09878e1301 "ACPI: sysfs: Change
> > > ACPI_MASKABLE_GPE_MAX to 0x100" in -next as the reason for a boot
> > > regression in my qemu-kvm test environment. It spews:
> > >
> > > [    1.456728] ACPI: Masking GPE 0x0.
> > > ...
> > > [  161.721420] ACPI: Masking GPE 0x0.
> > >
> > > ...and then hangs.
> > >
> > > A straight revert gets the configuration back on its feet.
> > >
> > > qemu-system-x86_64 --version
> > > QEMU emulator version 4.0.50 (v4.0.0-928-g49c6c6ac0cd8)
> >
> > OK, I'll drop it then, thanks!
>
> We're seeing the same issue wtih 5.4.0-rc7-next-20191118 on a Linux VM
> running on Hyper-V :
>
> [    0.133029] ACPI: 1 ACPI AML tables successfully acquired and loaded
> [    0.144023] ACPI: Interpreter enabled
> [    0.145023] ACPI: (supports S0 S5)
> [    0.146023] ACPI: Using IOAPIC for interrupt routing
> [    0.147024] PCI: Using host bridge windows from ACPI; if necessary,
> use "pci=nocrs" and report a bug
> [    0.148031] ACPI: Masking GPE 0x0.
> ...
> [  774.839023] ACPI: Masking GPE 0x0.
> [  774.840023] ACPI: Masking GPE 0x0.
>
> I guess the patch is only tested on a physical machine and not on a VM...

It looks like the patch hasn't been tested at all.

Please try to change the data type of gpe in
acpi_gpe_apply_masked_gpes() to u16 and see if that helps.