Re: [PATCH v2 4/6] dt-bindings: hypervisor: Rename virtio to hypervisor

Re: [PATCH v2 4/6] dt-bindings: hypervisor: Rename virtio to hypervisor

[Rob Herring]

On Tue, 31 Jan 2023 10:10:07 -0800, Saurabh Sengar wrote:
> Rename virtio folder to more generic hypervisor, so that this can
> accommodate more devices of similar type.
> 
> Signed-off-by: Saurabh Sengar <ssengar@linux.microsoft.com>
> ---
>  .../devicetree/bindings/{virtio => hypervisor}/mmio.yaml        | 2 +-
>  .../devicetree/bindings/{virtio => hypervisor}/pci-iommu.yaml   | 2 +-
>  .../bindings/{virtio => hypervisor}/virtio-device.yaml          | 2 +-
>  MAINTAINERS                                                     | 2 +-
>  4 files changed, 4 insertions(+), 4 deletions(-)
>  rename Documentation/devicetree/bindings/{virtio => hypervisor}/mmio.yaml (95%)
>  rename Documentation/devicetree/bindings/{virtio => hypervisor}/pci-iommu.yaml (98%)
>  rename Documentation/devicetree/bindings/{virtio => hypervisor}/virtio-device.yaml (93%)
> 

My bot found errors running 'make DT_CHECKER_FLAGS=-m dt_binding_check'
on your patch (DT_CHECKER_FLAGS is new in v5.13):

yamllint warnings/errors:

dtschema/dtc warnings/errors:
./Documentation/devicetree/bindings/i2c/i2c-virtio.yaml: Unable to find schema file matching $id: http://devicetree.org/schemas/virtio/virtio-device.yaml
./Documentation/devicetree/bindings/gpio/gpio-virtio.yaml: Unable to find schema file matching $id: http://devicetree.org/schemas/virtio/virtio-device.yaml
/builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/i2c/i2c-virtio.example.dtb: i2c: False schema does not allow {'compatible': ['virtio,device22'], '#address-cells': [[1]], '#size-cells': [[0]], 'light-sensor@20': {'compatible': ['dynaimage,al3320a'], 'reg': [[32]]}, '$nodename': ['i2c']}
	From schema: /builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/i2c/i2c-virtio.yaml
/builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/gpio/gpio-virtio.example.dtb: gpio: False schema does not allow {'compatible': ['virtio,device29'], 'gpio-controller': True, '#gpio-cells': [[2]], 'interrupt-controller': True, '#interrupt-cells': [[2]], '$nodename': ['gpio']}
	From schema: /builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/gpio/gpio-virtio.yaml
/builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/hypervisor/virtio-device.example.dtb: i2c: False schema does not allow {'compatible': ['virtio,device22'], '$nodename': ['i2c']}
	From schema: /builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/i2c/i2c-virtio.yaml

doc reference errors (make refcheckdocs):
MAINTAINERS: Documentation/devicetree/bindings/virtio/

See https://patchwork.ozlabs.org/project/devicetree-bindings/patch/1675188609-20913-5-git-send-email-ssengar@linux.microsoft.com

The base for the series is generally the latest rc1. A different dependency
should be noted in *this* patch.

If you already ran 'make dt_binding_check' and didn't see the above
error(s), then make sure 'yamllint' is installed and dt-schema is up to
date:

pip3 install dtschema --upgrade

Please check and re-submit after running the above command yourself. Note
that DT_SCHEMA_FILES can be set to your schema file to speed up checking
your schema. However, it must be unset to test all examples with your schema.

_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

Re: [PATCH v2 0/6] Device tree support for Hyper-V VMBus driver

[Rob Herring]

On Tue, Jan 31, 2023 at 06:04:49PM -0800, Saurabh Singh Sengar wrote:
> On Tue, Jan 31, 2023 at 02:27:51PM -0600, Rob Herring wrote:
> > On Tue, Jan 31, 2023 at 12:10 PM Saurabh Sengar
> > <ssengar@linux.microsoft.com> wrote:
> > >
> > > This set of patches expands the VMBus driver to include device tree
> > > support.
> > >
> > > The first two patches enable compilation of Hyper-V APIs in a non-ACPI
> > > build.
> > >
> > > The third patch converts the VMBus driver from acpi to more generic
> > > platform driver.
> > >
> > > Further to add device tree documentation for VMBus, it needs to club with
> > > other virtualization driver's documentation. For this rename the virtio
> > > folder to more generic hypervisor, so that all the hypervisor based
> > > devices can co-exist in a single place in device tree documentation. The
> > > fourth patch does this renaming.
> > >
> > > The fifth patch introduces the device tree documentation for VMBus.
> > >
> > > The sixth patch adds device tree support to the VMBus driver. Currently
> > > this is tested only for x86 and it may not work for other archs.
> > 
> > I can read all the patches and see *what* they do. You don't really
> > need to list that here. I'm still wondering *why*. That is what the
> > cover letter and commit messages should answer. Why do you need DT
> > support? How does this even work on x86? FDT is only enabled for
> > CE4100 platform.
> 
> HI Rob,
> 
> Thanks for your comments.
> We are working on a solution where kernel is booted without ACPI tables to keep
> the overall system's memory footprints slim and possibly faster boot time.
> We have tested this by enabling CONFIG_OF for x86.

It's CONFIG_OF_EARLY_FLATTREE which you would need and that's not user 
selectable. At a minimum, you need some kconfig changes. Where are 
those?

Also see my comment on v1 about running DT validation on your dtb. I'm 
sure running it would point out other issues. Such as the root level 
comaptible string(s) need to be documented. You need cpu nodes, 
interrupt controller, timers, etc. Those all have to be documented.

Rob
_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

Re: [PATCH v2 0/6] Device tree support for Hyper-V VMBus driver

[Rob Herring]

On Wed, Feb 1, 2023 at 10:34 AM Saurabh Singh Sengar
<ssengar@linux.microsoft.com> wrote:
>
> On Wed, Feb 01, 2023 at 08:51:46AM -0600, Rob Herring wrote:
> > On Tue, Jan 31, 2023 at 06:04:49PM -0800, Saurabh Singh Sengar wrote:
> > > On Tue, Jan 31, 2023 at 02:27:51PM -0600, Rob Herring wrote:
> > > > On Tue, Jan 31, 2023 at 12:10 PM Saurabh Sengar
> > > > <ssengar@linux.microsoft.com> wrote:
> > > > >
> > > > > This set of patches expands the VMBus driver to include device tree
> > > > > support.
> > > > >
> > > > > The first two patches enable compilation of Hyper-V APIs in a non-ACPI
> > > > > build.
> > > > >
> > > > > The third patch converts the VMBus driver from acpi to more generic
> > > > > platform driver.
> > > > >
> > > > > Further to add device tree documentation for VMBus, it needs to club with
> > > > > other virtualization driver's documentation. For this rename the virtio
> > > > > folder to more generic hypervisor, so that all the hypervisor based
> > > > > devices can co-exist in a single place in device tree documentation. The
> > > > > fourth patch does this renaming.
> > > > >
> > > > > The fifth patch introduces the device tree documentation for VMBus.
> > > > >
> > > > > The sixth patch adds device tree support to the VMBus driver. Currently
> > > > > this is tested only for x86 and it may not work for other archs.
> > > >
> > > > I can read all the patches and see *what* they do. You don't really
> > > > need to list that here. I'm still wondering *why*. That is what the
> > > > cover letter and commit messages should answer. Why do you need DT
> > > > support? How does this even work on x86? FDT is only enabled for
> > > > CE4100 platform.
> > >
> > > HI Rob,
> > >
> > > Thanks for your comments.
> > > We are working on a solution where kernel is booted without ACPI tables to keep
> > > the overall system's memory footprints slim and possibly faster boot time.
> > > We have tested this by enabling CONFIG_OF for x86.
> >
> > It's CONFIG_OF_EARLY_FLATTREE which you would need and that's not user
> > selectable. At a minimum, you need some kconfig changes. Where are
> > those?
>
> You are right we have define a new config flag in Kconfig, and selected CONFIG_OF
> and CONFIG_OF_EARLY_FLATTREE. We are working on upstreaming that patch as well
> however that will be a separate patch series.

Fair enough, but that should come first IMO. Really I just want to see
a complete picture. That can be a reference to a git branch(es) or
other patch series. But again, what I want to see in particular is the
actual DT and validation run on it.

> > Also see my comment on v1 about running DT validation on your dtb. I'm
> > sure running it would point out other issues. Such as the root level
> > comaptible string(s) need to be documented. You need cpu nodes,
> > interrupt controller, timers, etc. Those all have to be documented.
>
> I will be changing the parent node to soc node as suggested by Krzysztof
> in other thread.

Another issue yes, but orthogonal to my comments.

>
> soc {
>         #address-cells = <2>;
>         #size-cells = <2>;

You are missing 'ranges' here. Without it, addresses aren't translatable.

You are also missing 'compatible = "simple-bus";'. This happens to
work on x86 because of legacy reasons, but we don't want new cases
added.

>
>         vmbus@ff0000000 {
>             #address-cells = <2>;
>             #size-cells = <1>;
>             compatible = "Microsoft,vmbus";

'Microsoft' is not a vendor prefix.

>             ranges = <0x00 0x00 0x0f 0xf0000000 0x10000000>;
>         };
> };
>
> This will be sufficient.

All these comments are unnecessary because the tools will now check
these things and we shouldn't have to.

Rob
_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

Re: [PATCH v2 6/6] Driver: VMBus: Add device tree support

[Rob Herring]

On Wed, Feb 01, 2023 at 08:51:33AM -0800, Saurabh Singh Sengar wrote:
> On Tue, Jan 31, 2023 at 02:12:53PM -0600, Rob Herring wrote:
> > On Tue, Jan 31, 2023 at 12:10 PM Saurabh Sengar
> > <ssengar@linux.microsoft.com> wrote:
> > >
> > > Update the driver to support device tree boot as well along with ACPI.
> > > At present the device tree parsing only provides the mmio region info
> > > and is not the exact copy of ACPI parsing. This is sufficient to cater
> > > all the current device tree usecases for VMBus.
> > >
> > > Signed-off-by: Saurabh Sengar <ssengar@linux.microsoft.com>
> > > ---
> > >  drivers/hv/vmbus_drv.c | 75 ++++++++++++++++++++++++++++++++++++++++--
> > >  1 file changed, 73 insertions(+), 2 deletions(-)
> > >
> > > diff --git a/drivers/hv/vmbus_drv.c b/drivers/hv/vmbus_drv.c
> > > index 49030e756b9f..1741f1348f9f 100644
> > > --- a/drivers/hv/vmbus_drv.c
> > > +++ b/drivers/hv/vmbus_drv.c
> > > @@ -2152,7 +2152,7 @@ void vmbus_device_unregister(struct hv_device *device_obj)
> > >         device_unregister(&device_obj->device);
> > >  }
> > >
> > > -
> > > +#ifdef CONFIG_ACPI
> > >  /*
> > >   * VMBUS is an acpi enumerated device. Get the information we
> > >   * need from DSDT.
> > > @@ -2262,6 +2262,7 @@ static acpi_status vmbus_walk_resources(struct acpi_resource *res, void *ctx)
> > >
> > >         return AE_OK;
> > >  }
> > > +#endif
> > >
> > >  static void vmbus_mmio_remove(void)
> > >  {
> > > @@ -2282,7 +2283,7 @@ static void vmbus_mmio_remove(void)
> > >         }
> > >  }
> > >
> > > -static void vmbus_reserve_fb(void)
> > > +static void __maybe_unused vmbus_reserve_fb(void)
> > >  {
> > >         resource_size_t start = 0, size;
> > >         struct pci_dev *pdev;
> > > @@ -2442,6 +2443,7 @@ void vmbus_free_mmio(resource_size_t start, resource_size_t size)
> > >  }
> > >  EXPORT_SYMBOL_GPL(vmbus_free_mmio);
> > >
> > > +#ifdef CONFIG_ACPI
> > 
> > It's better to put C 'if (!IS_ENABLED(CONFIG_ACPI)' code in the
> 
> I wanted to have separate function for ACPI and device tree flow, which
> can be easily maintained with #ifdef. Please let me know if its fine.

Yes, you can have separate functions:

static int vmbus_acpi_add(struct platform_device *pdev)
{
	if (!IS_ENABLED(CONFIG_ACPI))
		return -ENODEV;

	...
}

The compiler will throw away the function in the end if CONFIG_ACPI is 
not enabled.

That is easier for us to maintain because it reduces the combinations to 
build.

> 
> > 
> > >  static int vmbus_acpi_add(struct platform_device *pdev)
> > >  {
> > >         acpi_status result;
> > > @@ -2496,10 +2498,68 @@ static int vmbus_acpi_add(struct platform_device *pdev)
> > >                 vmbus_mmio_remove();
> > >         return ret_val;
> > >  }
> > > +#else
> > > +
> > > +static int vmbus_device_add(struct platform_device *pdev)
> > > +{
> > > +       struct resource **cur_res = &hyperv_mmio;
> > > +       struct device_node *np;
> > > +       u32 *ranges, len;
> > > +       u64 start;
> > > +       int nr_ranges, child_cells = 2, cur_cell = 0, ret = 0;
> > > +
> > > +       hv_dev = pdev;
> > > +       np = pdev->dev.of_node;
> > > +
> > > +       nr_ranges = device_property_count_u32(&pdev->dev, "ranges");
> > 
> > Parsing ranges yourself is a bad sign. It's a standard property and we
> > have functions which handle it. If those don't work, then something is
> > wrong with your DT or they need to be fixed/expanded.
> 
> I find all the  standard functions which parse "ranges" property are doing
> much more then I need. Our requirement is to only pass the mmio memory range
> and size, I couldn't find any standard API doing this.

You can't just change how standard properties work to suit your needs.

We shouldn't even be having this discussion because we have tools to 
check all this now. dtc does some and dtschema does a lot more.

> I see some of the drivers are using these APIs to parse ranges property hence
> I follwed those examples. I will be happy to improve it if I get any better
> alternative.

You can always find bad examples to follow...

> > > +       if (nr_ranges < 0)
> > > +               return nr_ranges;
> > > +       ranges = kcalloc(nr_ranges, sizeof(u32), GFP_KERNEL);
> > > +       if (!ranges)
> > > +               return -ENOMEM;
> > > +
> > > +       if (device_property_read_u32_array(&pdev->dev, "ranges", ranges, nr_ranges)) {
> > > +               ret =  -EINVAL;
> > > +               goto free_ranges;
> > > +       }
> > > +
> > > +       while (cur_cell < nr_ranges) {
> > > +               struct resource *res;
> > > +
> > > +               /* The first u64 in the ranges description isn't used currently. */
> > > +               cur_cell = cur_cell + child_cells;
> > > +               start = ranges[cur_cell++];
> > > +               start = (start << 32) | ranges[cur_cell++];
> > > +               len = ranges[cur_cell++];
> > 
> > To expand my last point, the format of ranges is <child_addr
> > parent_addr length>. That's not what your 'ranges' has. You've also
> > just ignored '#address-cells' and '#size-cells'.
> 
> Got it. However I need to check if there is any standard API which can
> give me these values, otherwise I may have to parse these as well :(

for_each_of_range()

That is not how linux works. When the core code doesn't do what you 
want, you adapt it to your needs. You don't work around it. Read 
this[1].

Rob

[1] https://lwn.net/Articles/443531/
_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

Re: [PATCH v2 6/6] Driver: VMBus: Add device tree support

[Rob Herring]

On Fri, Feb 3, 2023 at 11:36 AM Saurabh Singh Sengar
<ssengar@linux.microsoft.com> wrote:
>
> On Wed, Feb 01, 2023 at 11:46:38AM -0600, Rob Herring wrote:
> > On Wed, Feb 01, 2023 at 08:51:33AM -0800, Saurabh Singh Sengar wrote:
> > > On Tue, Jan 31, 2023 at 02:12:53PM -0600, Rob Herring wrote:
> > > > On Tue, Jan 31, 2023 at 12:10 PM Saurabh Sengar
> > > > <ssengar@linux.microsoft.com> wrote:
> > > > >
> > > > > Update the driver to support device tree boot as well along with ACPI.
> > > > > At present the device tree parsing only provides the mmio region info
> > > > > and is not the exact copy of ACPI parsing. This is sufficient to cater
> > > > > all the current device tree usecases for VMBus.
> > > > >
> > > > > Signed-off-by: Saurabh Sengar <ssengar@linux.microsoft.com>
> > > > > ---
> > > > >  drivers/hv/vmbus_drv.c | 75 ++++++++++++++++++++++++++++++++++++++++--
> > > > >  1 file changed, 73 insertions(+), 2 deletions(-)
> > > > >
> > > > > diff --git a/drivers/hv/vmbus_drv.c b/drivers/hv/vmbus_drv.c
> > > > > index 49030e756b9f..1741f1348f9f 100644
> > > > > --- a/drivers/hv/vmbus_drv.c
> > > > > +++ b/drivers/hv/vmbus_drv.c
> > > > > @@ -2152,7 +2152,7 @@ void vmbus_device_unregister(struct hv_device *device_obj)
> > > > >         device_unregister(&device_obj->device);
> > > > >  }
> (...)
> > > > >         struct pci_dev *pdev;
> > > > > @@ -2442,6 +2443,7 @@ void vmbus_free_mmio(resource_size_t start, resource_size_t size)
> > > > >  }
> > > > >  EXPORT_SYMBOL_GPL(vmbus_free_mmio);
> > > > >
> > > > > +#ifdef CONFIG_ACPI
> > > >
> > > > It's better to put C 'if (!IS_ENABLED(CONFIG_ACPI)' code in the
> > >
> > > I wanted to have separate function for ACPI and device tree flow, which
> > > can be easily maintained with #ifdef. Please let me know if its fine.
> >
> > Yes, you can have separate functions:
> >
> > static int vmbus_acpi_add(struct platform_device *pdev)
> > {
> >       if (!IS_ENABLED(CONFIG_ACPI))
> >               return -ENODEV;
> >
> >       ...
> > }
> >
> > The compiler will throw away the function in the end if CONFIG_ACPI is
> > not enabled.
> >
> > That is easier for us to maintain because it reduces the combinations to
> > build.
> >
>
> I tried removing #ifdef CONFIG_ACPI and use C's if(!IS_ENABLED(CONFIG_ACPI)) but looks
> compiler is not optimizing out the rest of function, it still throwing errors
> for acpi functions. This doesn't look 1:1 replacement to me.
> Please let me know if I have missunderstood any of your suggestion.
>
> drivers/hv/vmbus_drv.c:2175:8: error: implicit declaration of function ‘acpi_dev_resource_interrupt’ [-Werror=implicit-function-

That's a failure of the ACPI headers not having empty function
declarations. The DT functions do...

Also, this is just a broken assumption:

#ifdef CONFIG_ACPI

#else
// Assume DT
#endif

Both ACPI and DT can be enabled at the same time. They may be mutually
exclusive for a platform, but not the kernel. For distro kernels, both
will be enabled typically if the arch supports both. On arm64, DT is
never disabled because the boot interface is always DT.

Furthermore, this makes compile testing your code difficult. The arm64
defconfig, allmodconfig and allyesconfig all will not build the DT
code. The same for x86. This means all the CI builds that happen can't
build test this.

Rob
_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

RE: [PATCH v2 2/6] Drivers: hv: allow non ACPI compilation for hv_is_hibernation_supported

[Michael Kelley (LINUX) via Virtualization]

From: Saurabh Sengar <ssengar@linux.microsoft.com> Sent: Tuesday, January 31, 2023 10:10 AM
> 
> acpi_sleep_state_supported API is only define for CONFIG_ACPI flag and
> thus it can't be used for non-ACPI builds. Initaly there won't be

s/Initaly/Initially/

> hibernate support for non ACPI builds.

s/hibernate/hibernation/

> 
> This change will help adding device tree support in subsequent commits.
> 
> Signed-off-by: Saurabh Sengar <ssengar@linux.microsoft.com>
> ---
>  drivers/hv/hv_common.c | 4 ++++
>  1 file changed, 4 insertions(+)
> 
> diff --git a/drivers/hv/hv_common.c b/drivers/hv/hv_common.c
> index 52a6f89ccdbd..370ec20d1993 100644
> --- a/drivers/hv/hv_common.c
> +++ b/drivers/hv/hv_common.c
> @@ -234,7 +234,11 @@ EXPORT_SYMBOL_GPL(hv_setup_dma_ops);
> 
>  bool hv_is_hibernation_supported(void)
>  {
> +#ifdef CONFIG_ACPI
>  	return !hv_root_partition && acpi_sleep_state_supported(ACPI_STATE_S4);
> +#else
> +	return false;
> +#endif

Is this patch needed?  If CONFIG_ACPI is not set, then per
arch/x86/Kconfig, CONFIG_ACPI_SYSTEM_POWER_STATES_SUPPORT
is not selected.  In that case, the #ifdef in include/acpi/acpi_bus.h
provides a stub for acpi_sleep_state_supported() that returns "false".
So it seems like the existing code should compile and correctly return
"false" when CONFIG_ACPI is not set.

Michael

>  }
>  EXPORT_SYMBOL_GPL(hv_is_hibernation_supported);
> 
> --
> 2.25.1

_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

RE: [PATCH v2 2/6] Drivers: hv: allow non ACPI compilation for hv_is_hibernation_supported

[Michael Kelley (LINUX) via Virtualization]

From: Saurabh Singh Sengar <ssengar@linux.microsoft.com> Sent: Thursday, February 2, 2023 6:49 AM
> 
> On Wed, Feb 01, 2023 at 05:47:44PM +0000, Michael Kelley (LINUX) wrote:
> > From: Saurabh Sengar <ssengar@linux.microsoft.com> Sent: Tuesday, January 31,
> 2023 10:10 AM
> > >
> > > acpi_sleep_state_supported API is only define for CONFIG_ACPI flag and
> > > thus it can't be used for non-ACPI builds. Initaly there won't be
> >
> > s/Initaly/Initially/
> 
> OK
> 
> >
> > > hibernate support for non ACPI builds.
> >
> > s/hibernate/hibernation/
> 
> OK
> 
> >
> > >
> > > This change will help adding device tree support in subsequent commits.
> > >
> > > Signed-off-by: Saurabh Sengar <ssengar@linux.microsoft.com>
> > > ---
> > >  drivers/hv/hv_common.c | 4 ++++
> > >  1 file changed, 4 insertions(+)
> > >
> > > diff --git a/drivers/hv/hv_common.c b/drivers/hv/hv_common.c
> > > index 52a6f89ccdbd..370ec20d1993 100644
> > > --- a/drivers/hv/hv_common.c
> > > +++ b/drivers/hv/hv_common.c
> > > @@ -234,7 +234,11 @@ EXPORT_SYMBOL_GPL(hv_setup_dma_ops);
> > >
> > >  bool hv_is_hibernation_supported(void)
> > >  {
> > > +#ifdef CONFIG_ACPI
> > >  	return !hv_root_partition && acpi_sleep_state_supported(ACPI_STATE_S4);
> > > +#else
> > > +	return false;
> > > +#endif
> >
> > Is this patch needed?  If CONFIG_ACPI is not set, then per
> > arch/x86/Kconfig, CONFIG_ACPI_SYSTEM_POWER_STATES_SUPPORT
> > is not selected.  In that case, the #ifdef in include/acpi/acpi_bus.h
> > provides a stub for acpi_sleep_state_supported() that returns "false".
> > So it seems like the existing code should compile and correctly return
> > "false" when CONFIG_ACPI is not set.
> 
> You are right, if CONFIG_ACPI_SYSTEM_POWER_STATES_SUPPORT is not set
> acpi_sleep_state_supported will return false, but this is applicable only
> when CONFIG_ACPI is enable. If CONFIG_ACPI is not enable both these
> functions are not defined.
> 

Indeed, you are right.  Most of include/acpi/acpi_bus.h is bracketed with
#ifdef CONFIG_ACPI, including the stub acpi_sleep_state_supported().
Oh well. :-(

Michael
_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization

RE: [PATCH v2 3/6] Drivers: hv: vmbus: Convert acpi_device to platform_device

[Michael Kelley (LINUX) via Virtualization]

From: Saurabh Sengar <ssengar@linux.microsoft.com> Sent: Tuesday, January 31, 2023 10:10 AM
> 
> Use more generic platform device instead of acpi device. Also rename the
> function vmbus_acpi_remove to more generic name vmbus_mmio_remove.
> 
> Signed-off-by: Saurabh Sengar <ssengar@linux.microsoft.com>
> ---
>  drivers/hv/vmbus_drv.c | 78 +++++++++++++++++++++++++-----------------
>  1 file changed, 46 insertions(+), 32 deletions(-)
> 
> diff --git a/drivers/hv/vmbus_drv.c b/drivers/hv/vmbus_drv.c
> index d24dd65b33d4..49030e756b9f 100644
> --- a/drivers/hv/vmbus_drv.c
> +++ b/drivers/hv/vmbus_drv.c
> @@ -12,6 +12,7 @@
>  #include <linux/init.h>
>  #include <linux/module.h>
>  #include <linux/device.h>
> +#include <linux/platform_device.h>
>  #include <linux/interrupt.h>
>  #include <linux/sysctl.h>
>  #include <linux/slab.h>
> @@ -44,7 +45,7 @@ struct vmbus_dynid {
>  	struct hv_vmbus_device_id id;
>  };
> 
> -static struct acpi_device  *hv_acpi_dev;
> +static struct platform_device  *hv_dev;
> 
>  static int hyperv_cpuhp_online;
> 
> @@ -143,7 +144,7 @@ static DEFINE_MUTEX(hyperv_mmio_lock);
> 
>  static int vmbus_exists(void)
>  {
> -	if (hv_acpi_dev == NULL)
> +	if (hv_dev == NULL)
>  		return -ENODEV;
> 
>  	return 0;
> @@ -932,7 +933,7 @@ static int vmbus_dma_configure(struct device *child_device)
>  	 * On x86/x64 coherence is assumed and these calls have no effect.
>  	 */
>  	hv_setup_dma_ops(child_device,
> -		device_get_dma_attr(&hv_acpi_dev->dev) == DEV_DMA_COHERENT);
> +		device_get_dma_attr(&hv_dev->dev) == DEV_DMA_COHERENT);
>  	return 0;
>  }
> 
> @@ -2090,7 +2091,7 @@ int vmbus_device_register(struct hv_device
> *child_device_obj)
>  		     &child_device_obj->channel->offermsg.offer.if_instance);
> 
>  	child_device_obj->device.bus = &hv_bus;
> -	child_device_obj->device.parent = &hv_acpi_dev->dev;
> +	child_device_obj->device.parent = &hv_dev->dev;
>  	child_device_obj->device.release = vmbus_device_release;
> 
>  	child_device_obj->device.dma_parms = &child_device_obj->dma_parms;
> @@ -2262,7 +2263,7 @@ static acpi_status vmbus_walk_resources(struct
> acpi_resource *res, void *ctx)
>  	return AE_OK;
>  }
> 
> -static void vmbus_acpi_remove(struct acpi_device *device)
> +static void vmbus_mmio_remove(void)
>  {
>  	struct resource *cur_res;
>  	struct resource *next_res;
> @@ -2441,13 +2442,15 @@ void vmbus_free_mmio(resource_size_t start,
> resource_size_t size)
>  }
>  EXPORT_SYMBOL_GPL(vmbus_free_mmio);
> 
> -static int vmbus_acpi_add(struct acpi_device *device)
> +static int vmbus_acpi_add(struct platform_device *pdev)
>  {
>  	acpi_status result;
>  	int ret_val = -ENODEV;
> -	struct acpi_device *ancestor;
> +	struct platform_device *ancestor;
> +	struct acpi_device *adev = to_acpi_device(&pdev->dev);

This doesn't work.  The argument to vmbus_acpi_add() is a struct
platform_device, which has a struct device embedded in it (not a
pointer).   to_acpi_device() takes a struct device as an argument,
assuming that the struct device is embedded in a struct
acpi_device, which is not the case here.  The resulting local
variable adev is actually pointing to some (perhaps negative)
offset within the struct platform_device, and uses of adev are
getting unknown random data from within (or before) the
struct platform_device.

> 
> -	hv_acpi_dev = device;
> +	hv_dev = pdev;
> +	adev->fwnode.dev = &pdev->dev;
> 
>  	/*
>  	 * Older versions of Hyper-V for ARM64 fail to include the _CCA
> @@ -2456,15 +2459,16 @@ static int vmbus_acpi_add(struct acpi_device *device)
>  	 * up the ACPI device to behave as if _CCA is present and indicates
>  	 * hardware coherence.
>  	 */
> -	ACPI_COMPANION_SET(&device->dev, device);
> +	ACPI_COMPANION_SET(&pdev->dev, ACPI_COMPANION(&pdev->dev));

This statement seems tautological.  If ACPI_COMPANION(&pdev->dev)
returns a valid result,  why would the ACPI companion for &pdev->dev
need to be set?  The original code was setting the ACPI companion for the
embedded struct device to be the struct acpi_device.   I forget why this
wasn't already done for the VMBus device when it was originally parsed
from the ACPI DSDT ... 

>  	if (IS_ENABLED(CONFIG_ACPI_CCA_REQUIRED) &&
> -	    device_get_dma_attr(&device->dev) == DEV_DMA_NOT_SUPPORTED) {
> +	    device_get_dma_attr(&pdev->dev) == DEV_DMA_NOT_SUPPORTED) {
> +		struct acpi_device *adev_node = ACPI_COMPANION(&pdev->dev);

If earlier code in this function can get a correct pointer to the struct acpi_device,
then this statement shouldn't be necessary.  You already have it.

>  		pr_info("No ACPI _CCA found; assuming coherent device I/O\n");
> -		device->flags.cca_seen = true;
> -		device->flags.coherent_dma = true;
> +		adev_node->flags.cca_seen = true;
> +		adev_node->flags.coherent_dma = true;
>  	}
> 
> -	result = acpi_walk_resources(device->handle, METHOD_NAME__CRS,
> +	result = acpi_walk_resources(ACPI_HANDLE(&pdev->dev), METHOD_NAME__CRS,

Again, if you have a correct pointer to the struct acpi_device, then adev->handle
(like the original code) should be simpler than looking it up again with ACPI_HANDLE().  

>  					vmbus_walk_resources, NULL);
> 
>  	if (ACPI_FAILURE(result))
> @@ -2473,9 +2477,9 @@ static int vmbus_acpi_add(struct acpi_device *device)
>  	 * Some ancestor of the vmbus acpi device (Gen1 or Gen2
>  	 * firmware) is the VMOD that has the mmio ranges. Get that.
>  	 */
> -	for (ancestor = acpi_dev_parent(device); ancestor;
> -	     ancestor = acpi_dev_parent(ancestor)) {
> -		result = acpi_walk_resources(ancestor->handle, METHOD_NAME__CRS,
> +	for (ancestor = to_platform_device(pdev->dev.parent); ancestor;
> +	     ancestor = to_platform_device(ancestor->dev.parent)) {
> +		result = acpi_walk_resources(ACPI_HANDLE(&ancestor->dev), METHOD_NAME__CRS,

Similarly, if you get a correct pointer to the struct acpi_device, does the above
code need any changes?  I'm hoping not.

>  					     vmbus_walk_resources, NULL);
> 
>  		if (ACPI_FAILURE(result))
> @@ -2489,10 +2493,21 @@ static int vmbus_acpi_add(struct acpi_device *device)
> 
>  acpi_walk_err:
>  	if (ret_val)
> -		vmbus_acpi_remove(device);
> +		vmbus_mmio_remove();
>  	return ret_val;
>  }
> 
> +static int vmbus_platform_driver_probe(struct platform_device *pdev)
> +{
> +	return vmbus_acpi_add(pdev);
> +}
> +
> +static int vmbus_platform_driver_remove(struct platform_device *pdev)
> +{
> +	vmbus_mmio_remove();
> +	return 0;
> +}
> +
>  #ifdef CONFIG_PM_SLEEP
>  static int vmbus_bus_suspend(struct device *dev)
>  {
> @@ -2658,15 +2673,15 @@ static const struct dev_pm_ops vmbus_bus_pm = {
>  	.restore_noirq	= vmbus_bus_resume
>  };
> 
> -static struct acpi_driver vmbus_acpi_driver = {
> -	.name = "vmbus",
> -	.ids = vmbus_acpi_device_ids,
> -	.ops = {
> -		.add = vmbus_acpi_add,
> -		.remove = vmbus_acpi_remove,
> -	},
> -	.drv.pm = &vmbus_bus_pm,
> -	.drv.probe_type = PROBE_FORCE_SYNCHRONOUS,
> +static struct platform_driver vmbus_platform_driver = {
> +	.probe = vmbus_platform_driver_probe,
> +	.remove = vmbus_platform_driver_remove,
> +	.driver = {
> +		.name = "vmbus",
> +		.acpi_match_table = ACPI_PTR(vmbus_acpi_device_ids),
> +		.pm = &vmbus_bus_pm,
> +		.probe_type = PROBE_FORCE_SYNCHRONOUS,
> +	}
>  };
> 
>  static void hv_kexec_handler(void)
> @@ -2750,12 +2765,11 @@ static int __init hv_acpi_init(void)
>  	/*
>  	 * Get ACPI resources first.
>  	 */
> -	ret = acpi_bus_register_driver(&vmbus_acpi_driver);
> -
> +	ret = platform_driver_register(&vmbus_platform_driver);
>  	if (ret)
>  		return ret;
> 
> -	if (!hv_acpi_dev) {
> +	if (!hv_dev) {
>  		ret = -ENODEV;
>  		goto cleanup;
>  	}
> @@ -2785,8 +2799,8 @@ static int __init hv_acpi_init(void)
>  	return 0;
> 
>  cleanup:
> -	acpi_bus_unregister_driver(&vmbus_acpi_driver);
> -	hv_acpi_dev = NULL;
> +	platform_driver_unregister(&vmbus_platform_driver);
> +	hv_dev = NULL;
>  	return ret;
>  }
> 
> @@ -2839,7 +2853,7 @@ static void __exit vmbus_exit(void)
> 
>  	cpuhp_remove_state(hyperv_cpuhp_online);
>  	hv_synic_free();
> -	acpi_bus_unregister_driver(&vmbus_acpi_driver);
> +	platform_driver_unregister(&vmbus_platform_driver);
>  }
> 
> 
> --
> 2.25.1

_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization