[PATCH v2 0/2] Fix coherence for VMbus and PCI pass-thru devices in Hyper-V VM

[PATCH v2 0/2] Fix coherence for VMbus and PCI pass-thru devices in Hyper-V VM

[Michael Kelley]

Hyper-V VMs have VMbus synthetic devices and PCI pass-thru devices that are added
dynamically via the VMbus protocol and are not represented in the ACPI DSDT. Only
the top level VMbus node exists in the DSDT. As such, on ARM64 these devices don't
pick up coherence information and default to not hardware coherent.  This results
in extra software coherence management overhead since the synthetic devices are
always hardware coherent. PCI pass-thru devices are also hardware coherent in all
current usage scenarios.

Fix this by propagating coherence information from the top level VMbus node in
the DSDT to all VMbus synthetic devices and PCI pass-thru devices. While smaller
granularity of control would be better, basing on the VMbus node in the DSDT
gives as escape path if a future scenario arises with devices that are not
hardware coherent.

Robin Murphy -- I'm not ignoring your feedback about pci_dma_configure(), but
I wanted to try this alternate approach where pci_dma_configure() works as is.
If reviewers prefer modifying pci_dma_configure() to handle the Hyper-V
specifics, I can go back to that.

Changes since v1:
* Use device_get_dma_attr() instead of acpi_get_dma_attr(), eliminating the
  need to export acpi_get_dma_attr() [Robin Murphy]
* Use arch_setup_dma_ops() to set device coherence [Robin Murphy]
* Move handling of missing _CCA to vmbus_acpi_add() so it is only done once
* Rework handling of PCI devices so existing code in pci_dma_configure()
  just works

Michael Kelley (2):
  Drivers: hv: vmbus: Propagate VMbus coherence to each VMbus device
  PCI: hv: Propagate coherence from VMbus device to PCI device

 drivers/hv/hv_common.c              | 11 +++++++++++
 drivers/hv/vmbus_drv.c              | 23 +++++++++++++++++++++++
 drivers/pci/controller/pci-hyperv.c |  9 +++++++++
 include/asm-generic/mshyperv.h      |  1 +
 4 files changed, 44 insertions(+)

-- 
1.8.3.1

[PATCH v2 1/2] Drivers: hv: vmbus: Propagate VMbus coherence to each VMbus device

[Michael Kelley]

VMbus synthetic devices are not represented in the ACPI DSDT -- only
the top level VMbus device is represented. As a result, on ARM64
coherence information in the _CCA method is not specified for
synthetic devices, so they default to not hardware coherent.
Drivers for some of these synthetic devices have been recently
updated to use the standard DMA APIs, and they are incurring extra
overhead of unneeded software coherence management.

Fix this by propagating coherence information from the VMbus node
in ACPI to the individual synthetic devices. There's no effect on
x86/x64 where devices are always hardware coherent.

Signed-off-by: Michael Kelley <mikelley@microsoft.com>
---
 drivers/hv/hv_common.c         | 11 +++++++++++
 drivers/hv/vmbus_drv.c         | 23 +++++++++++++++++++++++
 include/asm-generic/mshyperv.h |  1 +
 3 files changed, 35 insertions(+)

diff --git a/drivers/hv/hv_common.c b/drivers/hv/hv_common.c
index 181d16b..820e814 100644
--- a/drivers/hv/hv_common.c
+++ b/drivers/hv/hv_common.c
@@ -20,6 +20,7 @@
 #include <linux/panic_notifier.h>
 #include <linux/ptrace.h>
 #include <linux/slab.h>
+#include <linux/dma-map-ops.h>
 #include <asm/hyperv-tlfs.h>
 #include <asm/mshyperv.h>
 
@@ -216,6 +217,16 @@ bool hv_query_ext_cap(u64 cap_query)
 }
 EXPORT_SYMBOL_GPL(hv_query_ext_cap);
 
+void hv_setup_dma_ops(struct device *dev, bool coherent)
+{
+	/*
+	 * Hyper-V does not offer a vIOMMU in the guest
+	 * VM, so pass 0/NULL for the IOMMU settings
+	 */
+	arch_setup_dma_ops(dev, 0, 0, NULL, coherent);
+}
+EXPORT_SYMBOL_GPL(hv_setup_dma_ops);
+
 bool hv_is_hibernation_supported(void)
 {
 	return !hv_root_partition && acpi_sleep_state_supported(ACPI_STATE_S4);
diff --git a/drivers/hv/vmbus_drv.c b/drivers/hv/vmbus_drv.c
index 12a2b37..2d2c54c 100644
--- a/drivers/hv/vmbus_drv.c
+++ b/drivers/hv/vmbus_drv.c
@@ -905,6 +905,14 @@ static int vmbus_probe(struct device *child_device)
 	struct hv_device *dev = device_to_hv_device(child_device);
 	const struct hv_vmbus_device_id *dev_id;
 
+	/*
+	 * On ARM64, propagate the DMA coherence setting from the top level
+	 * VMbus ACPI device to the child VMbus device being added here.
+	 * 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);
+
 	dev_id = hv_vmbus_get_id(drv, dev);
 	if (drv->probe) {
 		ret = drv->probe(dev, dev_id);
@@ -2428,6 +2436,21 @@ static int vmbus_acpi_add(struct acpi_device *device)
 
 	hv_acpi_dev = device;
 
+	/*
+	 * Older versions of Hyper-V for ARM64 fail to include the _CCA
+	 * method on the top level VMbus device in the DSDT. But devices
+	 * are hardware coherent in all current Hyper-V use cases, so fix
+	 * up the ACPI device to behave as if _CCA is present and indicates
+	 * hardware coherence.
+	 */
+	ACPI_COMPANION_SET(&device->dev, device);
+	if (IS_ENABLED(CONFIG_ACPI_CCA_REQUIRED) &&
+	    device_get_dma_attr(&device->dev) == DEV_DMA_NOT_SUPPORTED) {
+		pr_info("No ACPI _CCA found; assuming coherent device I/O\n");
+		device->flags.cca_seen = true;
+		device->flags.coherent_dma = true;
+	}
+
 	result = acpi_walk_resources(device->handle, METHOD_NAME__CRS,
 					vmbus_walk_resources, NULL);
 
diff --git a/include/asm-generic/mshyperv.h b/include/asm-generic/mshyperv.h
index c08758b..c05d2ce 100644
--- a/include/asm-generic/mshyperv.h
+++ b/include/asm-generic/mshyperv.h
@@ -269,6 +269,7 @@ static inline int cpumask_to_vpset_noself(struct hv_vpset *vpset,
 u64 hv_ghcb_hypercall(u64 control, void *input, void *output, u32 input_size);
 void hyperv_cleanup(void);
 bool hv_query_ext_cap(u64 cap_query);
+void hv_setup_dma_ops(struct device *dev, bool coherent);
 void *hv_map_memory(void *addr, unsigned long size);
 void hv_unmap_memory(void *addr);
 #else /* CONFIG_HYPERV */
-- 
1.8.3.1

[PATCH v2 2/2] PCI: hv: Propagate coherence from VMbus device to PCI device

[Michael Kelley]

PCI pass-thru devices in a Hyper-V VM are represented as a VMBus
device and as a PCI device.  The coherence of the VMbus device is
set based on the VMbus node in ACPI, but the PCI device has no
ACPI node and defaults to not hardware coherent.  This results
in extra software coherence management overhead on ARM64 when
devices are hardware coherent.

Fix this by setting up the PCI host bus so that normal
PCI mechanisms will propagate the coherence of the VMbus
device to the PCI device. There's no effect on x86/x64 where
devices are always hardware coherent.

Signed-off-by: Michael Kelley <mikelley@microsoft.com>
---
 drivers/pci/controller/pci-hyperv.c | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/drivers/pci/controller/pci-hyperv.c b/drivers/pci/controller/pci-hyperv.c
index ae0bc2f..88b3b56 100644
--- a/drivers/pci/controller/pci-hyperv.c
+++ b/drivers/pci/controller/pci-hyperv.c
@@ -3404,6 +3404,15 @@ static int hv_pci_probe(struct hv_device *hdev,
 	hbus->bridge->domain_nr = dom;
 #ifdef CONFIG_X86
 	hbus->sysdata.domain = dom;
+#elif defined(CONFIG_ARM64)
+	/*
+	 * Set the PCI bus parent to be the corresponding VMbus
+	 * device. Then the VMbus device will be assigned as the
+	 * ACPI companion in pcibios_root_bridge_prepare() and
+	 * pci_dma_configure() will propagate device coherence
+	 * information to devices created on the bus.
+	 */
+	hbus->sysdata.parent = hdev->device.parent;
 #endif
 
 	hbus->hdev = hdev;
-- 
1.8.3.1

Re: [PATCH v2 2/2] PCI: hv: Propagate coherence from VMbus device to PCI device

[Boqun Feng]

On Wed, Mar 23, 2022 at 01:31:12PM -0700, Michael Kelley wrote:
> PCI pass-thru devices in a Hyper-V VM are represented as a VMBus
> device and as a PCI device.  The coherence of the VMbus device is
> set based on the VMbus node in ACPI, but the PCI device has no
> ACPI node and defaults to not hardware coherent.  This results
> in extra software coherence management overhead on ARM64 when
> devices are hardware coherent.
> 
> Fix this by setting up the PCI host bus so that normal
> PCI mechanisms will propagate the coherence of the VMbus
> device to the PCI device. There's no effect on x86/x64 where
> devices are always hardware coherent.
> 
> Signed-off-by: Michael Kelley <mikelley@microsoft.com>

Acked-by: Boqun Feng <boqun.feng@gmail.com>

Regards,
Boqun

> ---
>  drivers/pci/controller/pci-hyperv.c | 9 +++++++++
>  1 file changed, 9 insertions(+)
> 
> diff --git a/drivers/pci/controller/pci-hyperv.c b/drivers/pci/controller/pci-hyperv.c
> index ae0bc2f..88b3b56 100644
> --- a/drivers/pci/controller/pci-hyperv.c
> +++ b/drivers/pci/controller/pci-hyperv.c
> @@ -3404,6 +3404,15 @@ static int hv_pci_probe(struct hv_device *hdev,
>  	hbus->bridge->domain_nr = dom;
>  #ifdef CONFIG_X86
>  	hbus->sysdata.domain = dom;
> +#elif defined(CONFIG_ARM64)
> +	/*
> +	 * Set the PCI bus parent to be the corresponding VMbus
> +	 * device. Then the VMbus device will be assigned as the
> +	 * ACPI companion in pcibios_root_bridge_prepare() and
> +	 * pci_dma_configure() will propagate device coherence
> +	 * information to devices created on the bus.
> +	 */
> +	hbus->sysdata.parent = hdev->device.parent;
>  #endif
>  
>  	hbus->hdev = hdev;
> -- 
> 1.8.3.1
> 
> 

Re: [PATCH v2 1/2] Drivers: hv: vmbus: Propagate VMbus coherence to each VMbus device

[Robin Murphy]

On 2022-03-23 20:31, Michael Kelley wrote:
> VMbus synthetic devices are not represented in the ACPI DSDT -- only
> the top level VMbus device is represented. As a result, on ARM64
> coherence information in the _CCA method is not specified for
> synthetic devices, so they default to not hardware coherent.
> Drivers for some of these synthetic devices have been recently
> updated to use the standard DMA APIs, and they are incurring extra
> overhead of unneeded software coherence management.
> 
> Fix this by propagating coherence information from the VMbus node
> in ACPI to the individual synthetic devices. There's no effect on
> x86/x64 where devices are always hardware coherent.
> 
> Signed-off-by: Michael Kelley <mikelley@microsoft.com>
> ---
>   drivers/hv/hv_common.c         | 11 +++++++++++
>   drivers/hv/vmbus_drv.c         | 23 +++++++++++++++++++++++
>   include/asm-generic/mshyperv.h |  1 +
>   3 files changed, 35 insertions(+)
> 
> diff --git a/drivers/hv/hv_common.c b/drivers/hv/hv_common.c
> index 181d16b..820e814 100644
> --- a/drivers/hv/hv_common.c
> +++ b/drivers/hv/hv_common.c
> @@ -20,6 +20,7 @@
>   #include <linux/panic_notifier.h>
>   #include <linux/ptrace.h>
>   #include <linux/slab.h>
> +#include <linux/dma-map-ops.h>
>   #include <asm/hyperv-tlfs.h>
>   #include <asm/mshyperv.h>
>   
> @@ -216,6 +217,16 @@ bool hv_query_ext_cap(u64 cap_query)
>   }
>   EXPORT_SYMBOL_GPL(hv_query_ext_cap);
>   
> +void hv_setup_dma_ops(struct device *dev, bool coherent)
> +{
> +	/*
> +	 * Hyper-V does not offer a vIOMMU in the guest
> +	 * VM, so pass 0/NULL for the IOMMU settings
> +	 */
> +	arch_setup_dma_ops(dev, 0, 0, NULL, coherent);
> +}
> +EXPORT_SYMBOL_GPL(hv_setup_dma_ops);
> +
>   bool hv_is_hibernation_supported(void)
>   {
>   	return !hv_root_partition && acpi_sleep_state_supported(ACPI_STATE_S4);
> diff --git a/drivers/hv/vmbus_drv.c b/drivers/hv/vmbus_drv.c
> index 12a2b37..2d2c54c 100644
> --- a/drivers/hv/vmbus_drv.c
> +++ b/drivers/hv/vmbus_drv.c
> @@ -905,6 +905,14 @@ static int vmbus_probe(struct device *child_device)
>   	struct hv_device *dev = device_to_hv_device(child_device);
>   	const struct hv_vmbus_device_id *dev_id;
>   
> +	/*
> +	 * On ARM64, propagate the DMA coherence setting from the top level
> +	 * VMbus ACPI device to the child VMbus device being added here.
> +	 * 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);

Would you mind hooking up the hv_bus.dma_configure method to do this? 
Feel free to fold hv_setup_dma_ops entirely into that if you're not 
likely to need to call it from anywhere else.

> +
>   	dev_id = hv_vmbus_get_id(drv, dev);
>   	if (drv->probe) {
>   		ret = drv->probe(dev, dev_id);
> @@ -2428,6 +2436,21 @@ static int vmbus_acpi_add(struct acpi_device *device)
>   
>   	hv_acpi_dev = device;
>   
> +	/*
> +	 * Older versions of Hyper-V for ARM64 fail to include the _CCA
> +	 * method on the top level VMbus device in the DSDT. But devices
> +	 * are hardware coherent in all current Hyper-V use cases, so fix
> +	 * up the ACPI device to behave as if _CCA is present and indicates
> +	 * hardware coherence.
> +	 */
> +	ACPI_COMPANION_SET(&device->dev, device);
> +	if (IS_ENABLED(CONFIG_ACPI_CCA_REQUIRED) &&
> +	    device_get_dma_attr(&device->dev) == DEV_DMA_NOT_SUPPORTED) {
> +		pr_info("No ACPI _CCA found; assuming coherent device I/O\n");
> +		device->flags.cca_seen = true;
> +		device->flags.coherent_dma = true;
> +	}

I'm not the biggest fan of this, especially since I'm not convinced that 
there are any out-of-support deployments of ARM64 Hyper-V that can't be 
updated. However I suppose it's not "real" firmware, and one Hyper-V 
component is at liberty to hack another Hyper-V component's data if it 
really wants to...

If you can hook up .dma_configure, or clarify if it wouldn't work,

Acked-by: Robin Murphy <robin.murphy@arm.com>

Cheers,
Robin.

> +
>   	result = acpi_walk_resources(device->handle, METHOD_NAME__CRS,
>   					vmbus_walk_resources, NULL);
>   
> diff --git a/include/asm-generic/mshyperv.h b/include/asm-generic/mshyperv.h
> index c08758b..c05d2ce 100644
> --- a/include/asm-generic/mshyperv.h
> +++ b/include/asm-generic/mshyperv.h
> @@ -269,6 +269,7 @@ static inline int cpumask_to_vpset_noself(struct hv_vpset *vpset,
>   u64 hv_ghcb_hypercall(u64 control, void *input, void *output, u32 input_size);
>   void hyperv_cleanup(void);
>   bool hv_query_ext_cap(u64 cap_query);
> +void hv_setup_dma_ops(struct device *dev, bool coherent);
>   void *hv_map_memory(void *addr, unsigned long size);
>   void hv_unmap_memory(void *addr);
>   #else /* CONFIG_HYPERV */

Re: [PATCH v2 2/2] PCI: hv: Propagate coherence from VMbus device to PCI device

[Robin Murphy]

On 2022-03-23 20:31, Michael Kelley wrote:
> PCI pass-thru devices in a Hyper-V VM are represented as a VMBus
> device and as a PCI device.  The coherence of the VMbus device is
> set based on the VMbus node in ACPI, but the PCI device has no
> ACPI node and defaults to not hardware coherent.  This results
> in extra software coherence management overhead on ARM64 when
> devices are hardware coherent.
> 
> Fix this by setting up the PCI host bus so that normal
> PCI mechanisms will propagate the coherence of the VMbus
> device to the PCI device. There's no effect on x86/x64 where
> devices are always hardware coherent.

Honestly, I don't hate this :)

It seems conceptually accurate, as far as I understand, and in 
functional terms I'm starting to think it might even be the most correct 
approach anyway. In the physical world we might be surprised to find the 
PCI side of a host bridge behind anything other than some platform/ACPI 
device representing the other side of a physical host bridge or root 
complex, but who's to say that a paravirtual world can't present a more 
abstract topology? Either way, a one-line way of tying in to the 
standard flow is hard to turn down.

Acked-by: Robin Murphy <robin.murphy@arm.com>

> Signed-off-by: Michael Kelley <mikelley@microsoft.com>
> ---
>   drivers/pci/controller/pci-hyperv.c | 9 +++++++++
>   1 file changed, 9 insertions(+)
> 
> diff --git a/drivers/pci/controller/pci-hyperv.c b/drivers/pci/controller/pci-hyperv.c
> index ae0bc2f..88b3b56 100644
> --- a/drivers/pci/controller/pci-hyperv.c
> +++ b/drivers/pci/controller/pci-hyperv.c
> @@ -3404,6 +3404,15 @@ static int hv_pci_probe(struct hv_device *hdev,
>   	hbus->bridge->domain_nr = dom;
>   #ifdef CONFIG_X86
>   	hbus->sysdata.domain = dom;
> +#elif defined(CONFIG_ARM64)
> +	/*
> +	 * Set the PCI bus parent to be the corresponding VMbus
> +	 * device. Then the VMbus device will be assigned as the
> +	 * ACPI companion in pcibios_root_bridge_prepare() and
> +	 * pci_dma_configure() will propagate device coherence
> +	 * information to devices created on the bus.
> +	 */
> +	hbus->sysdata.parent = hdev->device.parent;
>   #endif
>   
>   	hbus->hdev = hdev;

Re: [PATCH v2 2/2] PCI: hv: Propagate coherence from VMbus device to PCI device

[Robin Murphy]

On 2022-03-24 12:23, Robin Murphy wrote:
> On 2022-03-23 20:31, Michael Kelley wrote:
>> PCI pass-thru devices in a Hyper-V VM are represented as a VMBus
>> device and as a PCI device.  The coherence of the VMbus device is
>> set based on the VMbus node in ACPI, but the PCI device has no
>> ACPI node and defaults to not hardware coherent.  This results
>> in extra software coherence management overhead on ARM64 when
>> devices are hardware coherent.
>>
>> Fix this by setting up the PCI host bus so that normal
>> PCI mechanisms will propagate the coherence of the VMbus
>> device to the PCI device. There's no effect on x86/x64 where
>> devices are always hardware coherent.
> 
> Honestly, I don't hate this :)
> 
> It seems conceptually accurate, as far as I understand, and in 
> functional terms I'm starting to think it might even be the most correct 
> approach anyway. In the physical world we might be surprised to find the 
> PCI side of a host bridge

And of course by "the PCI side of a host bridge" I think I actually mean 
"a PCI root bus", because in my sloppy terminology I'm thinking about 
hardware bridging from PCI(e) to some SoC-internal protocol, which does 
not have to imply an actual PCI-visible Host Bridge device...

Robin.

> behind anything other than some platform/ACPI 
> device representing the other side of a physical host bridge or root 
> complex, but who's to say that a paravirtual world can't present a more 
> abstract topology? Either way, a one-line way of tying in to the 
> standard flow is hard to turn down.
> 
> Acked-by: Robin Murphy <robin.murphy@arm.com>
> 
>> Signed-off-by: Michael Kelley <mikelley@microsoft.com>
>> ---
>>   drivers/pci/controller/pci-hyperv.c | 9 +++++++++
>>   1 file changed, 9 insertions(+)
>>
>> diff --git a/drivers/pci/controller/pci-hyperv.c 
>> b/drivers/pci/controller/pci-hyperv.c
>> index ae0bc2f..88b3b56 100644
>> --- a/drivers/pci/controller/pci-hyperv.c
>> +++ b/drivers/pci/controller/pci-hyperv.c
>> @@ -3404,6 +3404,15 @@ static int hv_pci_probe(struct hv_device *hdev,
>>       hbus->bridge->domain_nr = dom;
>>   #ifdef CONFIG_X86
>>       hbus->sysdata.domain = dom;
>> +#elif defined(CONFIG_ARM64)
>> +    /*
>> +     * Set the PCI bus parent to be the corresponding VMbus
>> +     * device. Then the VMbus device will be assigned as the
>> +     * ACPI companion in pcibios_root_bridge_prepare() and
>> +     * pci_dma_configure() will propagate device coherence
>> +     * information to devices created on the bus.
>> +     */
>> +    hbus->sysdata.parent = hdev->device.parent;
>>   #endif
>>       hbus->hdev = hdev;
> _______________________________________________
> iommu mailing list
> iommu@lists.linux-foundation.org
> https://lists.linuxfoundation.org/mailman/listinfo/iommu

RE: [PATCH v2 1/2] Drivers: hv: vmbus: Propagate VMbus coherence to each VMbus device

[Michael Kelley (LINUX)]

From: Robin Murphy <robin.murphy@arm.com> Sent: Thursday, March 24, 2022 4:59 AM
> 
> On 2022-03-23 20:31, Michael Kelley wrote:
> > VMbus synthetic devices are not represented in the ACPI DSDT -- only
> > the top level VMbus device is represented. As a result, on ARM64
> > coherence information in the _CCA method is not specified for
> > synthetic devices, so they default to not hardware coherent.
> > Drivers for some of these synthetic devices have been recently
> > updated to use the standard DMA APIs, and they are incurring extra
> > overhead of unneeded software coherence management.
> >
> > Fix this by propagating coherence information from the VMbus node
> > in ACPI to the individual synthetic devices. There's no effect on
> > x86/x64 where devices are always hardware coherent.
> >
> > Signed-off-by: Michael Kelley <mikelley@microsoft.com>
> > ---
> >   drivers/hv/hv_common.c         | 11 +++++++++++
> >   drivers/hv/vmbus_drv.c         | 23 +++++++++++++++++++++++
> >   include/asm-generic/mshyperv.h |  1 +
> >   3 files changed, 35 insertions(+)
> >
> > diff --git a/drivers/hv/hv_common.c b/drivers/hv/hv_common.c
> > index 181d16b..820e814 100644
> > --- a/drivers/hv/hv_common.c
> > +++ b/drivers/hv/hv_common.c
> > @@ -20,6 +20,7 @@
> >   #include <linux/panic_notifier.h>
> >   #include <linux/ptrace.h>
> >   #include <linux/slab.h>
> > +#include <linux/dma-map-ops.h>
> >   #include <asm/hyperv-tlfs.h>
> >   #include <asm/mshyperv.h>
> >
> > @@ -216,6 +217,16 @@ bool hv_query_ext_cap(u64 cap_query)
> >   }
> >   EXPORT_SYMBOL_GPL(hv_query_ext_cap);
> >
> > +void hv_setup_dma_ops(struct device *dev, bool coherent)
> > +{
> > +	/*
> > +	 * Hyper-V does not offer a vIOMMU in the guest
> > +	 * VM, so pass 0/NULL for the IOMMU settings
> > +	 */
> > +	arch_setup_dma_ops(dev, 0, 0, NULL, coherent);
> > +}
> > +EXPORT_SYMBOL_GPL(hv_setup_dma_ops);
> > +
> >   bool hv_is_hibernation_supported(void)
> >   {
> >   	return !hv_root_partition && acpi_sleep_state_supported(ACPI_STATE_S4);
> > diff --git a/drivers/hv/vmbus_drv.c b/drivers/hv/vmbus_drv.c
> > index 12a2b37..2d2c54c 100644
> > --- a/drivers/hv/vmbus_drv.c
> > +++ b/drivers/hv/vmbus_drv.c
> > @@ -905,6 +905,14 @@ static int vmbus_probe(struct device *child_device)
> >   	struct hv_device *dev = device_to_hv_device(child_device);
> >   	const struct hv_vmbus_device_id *dev_id;
> >
> > +	/*
> > +	 * On ARM64, propagate the DMA coherence setting from the top level
> > +	 * VMbus ACPI device to the child VMbus device being added here.
> > +	 * 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);
> 
> Would you mind hooking up the hv_bus.dma_configure method to do this?
> Feel free to fold hv_setup_dma_ops entirely into that if you're not
> likely to need to call it from anywhere else.

I'm pretty sure using hv_bus.dma_configure() is doable.  A separate
hv_setup_dma_ops() is still needed because arch_setup_dma_ops() isn't
exported and this VMbus driver can be built as a module.

> 
> > +
> >   	dev_id = hv_vmbus_get_id(drv, dev);
> >   	if (drv->probe) {
> >   		ret = drv->probe(dev, dev_id);
> > @@ -2428,6 +2436,21 @@ static int vmbus_acpi_add(struct acpi_device *device)
> >
> >   	hv_acpi_dev = device;
> >
> > +	/*
> > +	 * Older versions of Hyper-V for ARM64 fail to include the _CCA
> > +	 * method on the top level VMbus device in the DSDT. But devices
> > +	 * are hardware coherent in all current Hyper-V use cases, so fix
> > +	 * up the ACPI device to behave as if _CCA is present and indicates
> > +	 * hardware coherence.
> > +	 */
> > +	ACPI_COMPANION_SET(&device->dev, device);
> > +	if (IS_ENABLED(CONFIG_ACPI_CCA_REQUIRED) &&
> > +	    device_get_dma_attr(&device->dev) == DEV_DMA_NOT_SUPPORTED) {
> > +		pr_info("No ACPI _CCA found; assuming coherent device I/O\n");
> > +		device->flags.cca_seen = true;
> > +		device->flags.coherent_dma = true;
> > +	}
> 
> I'm not the biggest fan of this, especially since I'm not convinced that
> there are any out-of-support deployments of ARM64 Hyper-V that can't be
> updated. However I suppose it's not "real" firmware, and one Hyper-V
> component is at liberty to hack another Hyper-V component's data if it
> really wants to...

Agreed, it's a hack.  But Hyper-V instances are out there as part of
Windows 10/11 on ARM64 PCs, and they run ARM64 VMs for the
Windows Subsystem for Linux.  Microsoft gets pilloried for breaking
stuff, and this removes the potential for that happening if someone
runs a new Linux kernel version in that VM.

Michael

> 
> If you can hook up .dma_configure, or clarify if it wouldn't work,
> 
> Acked-by: Robin Murphy <robin.murphy@arm.com>
> 
> Cheers,
> Robin.
> 
> > +
> >   	result = acpi_walk_resources(device->handle, METHOD_NAME__CRS,
> >   					vmbus_walk_resources, NULL);
> >
> > diff --git a/include/asm-generic/mshyperv.h b/include/asm-generic/mshyperv.h
> > index c08758b..c05d2ce 100644
> > --- a/include/asm-generic/mshyperv.h
> > +++ b/include/asm-generic/mshyperv.h
> > @@ -269,6 +269,7 @@ static inline int cpumask_to_vpset_noself(struct hv_vpset
> *vpset,
> >   u64 hv_ghcb_hypercall(u64 control, void *input, void *output, u32 input_size);
> >   void hyperv_cleanup(void);
> >   bool hv_query_ext_cap(u64 cap_query);
> > +void hv_setup_dma_ops(struct device *dev, bool coherent);
> >   void *hv_map_memory(void *addr, unsigned long size);
> >   void hv_unmap_memory(void *addr);
> >   #else /* CONFIG_HYPERV */

Re: [PATCH v2 1/2] Drivers: hv: vmbus: Propagate VMbus coherence to each VMbus device

[Robin Murphy]

On 2022-03-24 13:18, Michael Kelley (LINUX) wrote:
> From: Robin Murphy <robin.murphy@arm.com> Sent: Thursday, March 24, 2022 4:59 AM
>>
>> On 2022-03-23 20:31, Michael Kelley wrote:
>>> VMbus synthetic devices are not represented in the ACPI DSDT -- only
>>> the top level VMbus device is represented. As a result, on ARM64
>>> coherence information in the _CCA method is not specified for
>>> synthetic devices, so they default to not hardware coherent.
>>> Drivers for some of these synthetic devices have been recently
>>> updated to use the standard DMA APIs, and they are incurring extra
>>> overhead of unneeded software coherence management.
>>>
>>> Fix this by propagating coherence information from the VMbus node
>>> in ACPI to the individual synthetic devices. There's no effect on
>>> x86/x64 where devices are always hardware coherent.
>>>
>>> Signed-off-by: Michael Kelley <mikelley@microsoft.com>
>>> ---
>>>    drivers/hv/hv_common.c         | 11 +++++++++++
>>>    drivers/hv/vmbus_drv.c         | 23 +++++++++++++++++++++++
>>>    include/asm-generic/mshyperv.h |  1 +
>>>    3 files changed, 35 insertions(+)
>>>
>>> diff --git a/drivers/hv/hv_common.c b/drivers/hv/hv_common.c
>>> index 181d16b..820e814 100644
>>> --- a/drivers/hv/hv_common.c
>>> +++ b/drivers/hv/hv_common.c
>>> @@ -20,6 +20,7 @@
>>>    #include <linux/panic_notifier.h>
>>>    #include <linux/ptrace.h>
>>>    #include <linux/slab.h>
>>> +#include <linux/dma-map-ops.h>
>>>    #include <asm/hyperv-tlfs.h>
>>>    #include <asm/mshyperv.h>
>>>
>>> @@ -216,6 +217,16 @@ bool hv_query_ext_cap(u64 cap_query)
>>>    }
>>>    EXPORT_SYMBOL_GPL(hv_query_ext_cap);
>>>
>>> +void hv_setup_dma_ops(struct device *dev, bool coherent)
>>> +{
>>> +	/*
>>> +	 * Hyper-V does not offer a vIOMMU in the guest
>>> +	 * VM, so pass 0/NULL for the IOMMU settings
>>> +	 */
>>> +	arch_setup_dma_ops(dev, 0, 0, NULL, coherent);
>>> +}
>>> +EXPORT_SYMBOL_GPL(hv_setup_dma_ops);
>>> +
>>>    bool hv_is_hibernation_supported(void)
>>>    {
>>>    	return !hv_root_partition && acpi_sleep_state_supported(ACPI_STATE_S4);
>>> diff --git a/drivers/hv/vmbus_drv.c b/drivers/hv/vmbus_drv.c
>>> index 12a2b37..2d2c54c 100644
>>> --- a/drivers/hv/vmbus_drv.c
>>> +++ b/drivers/hv/vmbus_drv.c
>>> @@ -905,6 +905,14 @@ static int vmbus_probe(struct device *child_device)
>>>    	struct hv_device *dev = device_to_hv_device(child_device);
>>>    	const struct hv_vmbus_device_id *dev_id;
>>>
>>> +	/*
>>> +	 * On ARM64, propagate the DMA coherence setting from the top level
>>> +	 * VMbus ACPI device to the child VMbus device being added here.
>>> +	 * 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);
>>
>> Would you mind hooking up the hv_bus.dma_configure method to do this?
>> Feel free to fold hv_setup_dma_ops entirely into that if you're not
>> likely to need to call it from anywhere else.
> 
> I'm pretty sure using hv_bus.dma_configure() is doable.  A separate
> hv_setup_dma_ops() is still needed because arch_setup_dma_ops() isn't
> exported and this VMbus driver can be built as a module.

Ah, right you are, I keep forgetting that.

>>> +
>>>    	dev_id = hv_vmbus_get_id(drv, dev);
>>>    	if (drv->probe) {
>>>    		ret = drv->probe(dev, dev_id);
>>> @@ -2428,6 +2436,21 @@ static int vmbus_acpi_add(struct acpi_device *device)
>>>
>>>    	hv_acpi_dev = device;
>>>
>>> +	/*
>>> +	 * Older versions of Hyper-V for ARM64 fail to include the _CCA
>>> +	 * method on the top level VMbus device in the DSDT. But devices
>>> +	 * are hardware coherent in all current Hyper-V use cases, so fix
>>> +	 * up the ACPI device to behave as if _CCA is present and indicates
>>> +	 * hardware coherence.
>>> +	 */
>>> +	ACPI_COMPANION_SET(&device->dev, device);
>>> +	if (IS_ENABLED(CONFIG_ACPI_CCA_REQUIRED) &&
>>> +	    device_get_dma_attr(&device->dev) == DEV_DMA_NOT_SUPPORTED) {
>>> +		pr_info("No ACPI _CCA found; assuming coherent device I/O\n");
>>> +		device->flags.cca_seen = true;
>>> +		device->flags.coherent_dma = true;
>>> +	}
>>
>> I'm not the biggest fan of this, especially since I'm not convinced that
>> there are any out-of-support deployments of ARM64 Hyper-V that can't be
>> updated. However I suppose it's not "real" firmware, and one Hyper-V
>> component is at liberty to hack another Hyper-V component's data if it
>> really wants to...
> 
> Agreed, it's a hack.  But Hyper-V instances are out there as part of
> Windows 10/11 on ARM64 PCs, and they run ARM64 VMs for the
> Windows Subsystem for Linux.  Microsoft gets pilloried for breaking
> stuff, and this removes the potential for that happening if someone
> runs a new Linux kernel version in that VM.

And actually that one's on me as well - for some reason I was thinking 
that this had never worked, and therefore you could likely get a Hyper-V 
update pushed out long before users get this patch through distros, but 
of course it only becomes an issue now because previously there was no 
connection to any ACPI node at all. As I said, personally I'm happy to 
consider this a Hyper-V internal workaround, but if anyone else objects 
to poking at the ACPI flags, I suppose you've also got the fallback 
option of flipping it around and making the ACPI_COMPANION_SET() 
conditional, so that the behaviour for older versions remains entirely 
unchanged. If it happens, feel free to keep my ack for that approach too.

Cheers,
Robin.

> 
> Michael
> 
>>
>> If you can hook up .dma_configure, or clarify if it wouldn't work,
>>
>> Acked-by: Robin Murphy <robin.murphy@arm.com>
>>
>> Cheers,
>> Robin.
>>
>>> +
>>>    	result = acpi_walk_resources(device->handle, METHOD_NAME__CRS,
>>>    					vmbus_walk_resources, NULL);
>>>
>>> diff --git a/include/asm-generic/mshyperv.h b/include/asm-generic/mshyperv.h
>>> index c08758b..c05d2ce 100644
>>> --- a/include/asm-generic/mshyperv.h
>>> +++ b/include/asm-generic/mshyperv.h
>>> @@ -269,6 +269,7 @@ static inline int cpumask_to_vpset_noself(struct hv_vpset
>> *vpset,
>>>    u64 hv_ghcb_hypercall(u64 control, void *input, void *output, u32 input_size);
>>>    void hyperv_cleanup(void);
>>>    bool hv_query_ext_cap(u64 cap_query);
>>> +void hv_setup_dma_ops(struct device *dev, bool coherent);
>>>    void *hv_map_memory(void *addr, unsigned long size);
>>>    void hv_unmap_memory(void *addr);
>>>    #else /* CONFIG_HYPERV */