[v7] VIRTIO-IOMMU: Introduce aw-bits and granule options

[PATCH v7 4/9] qemu-options.hx: Document the virtio-iommu-pci granule option

Posted by Eric Auger 8 months, 3 weeks ago

We are missing an entry for the virtio-iommu-pci device. Add the
information on which machine it is currently supported and document
the new granule option.

Signed-off-by: Eric Auger <eric.auger@redhat.com>
---
 qemu-options.hx | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/qemu-options.hx b/qemu-options.hx
index 9a47385c15..757df3eac0 100644
--- a/qemu-options.hx
+++ b/qemu-options.hx
@@ -1172,6 +1172,14 @@ SRST
     Please also refer to the wiki page for general scenarios of VT-d
     emulation in QEMU: https://wiki.qemu.org/Features/VT-d.
 
+``-device virtio-iommu-pci[,option=...]``
+    This is only supported by ``-machine q35`` and ``-machine virt``.
+    It supports below options:
+
+    ``granule=val`` (possible values are 4k, 8k, 16k, 64k and host; default: host)
+        This decides the default granule to be be exposed by the
+        virtio-iommu. If host, the granule matches the host page size.
+
 ERST
 
 DEF("name", HAS_ARG, QEMU_OPTION_name,
-- 
2.41.0

Re: [PATCH v7 4/9] qemu-options.hx: Document the virtio-iommu-pci granule option

Posted by Philippe Mathieu-Daudé 8 months, 3 weeks ago

On 6/3/24 21:32, Eric Auger wrote:
> We are missing an entry for the virtio-iommu-pci device. Add the
> information on which machine it is currently supported and document
> the new granule option.
> 
> Signed-off-by: Eric Auger <eric.auger@redhat.com>
> ---
>   qemu-options.hx | 8 ++++++++
>   1 file changed, 8 insertions(+)
> 
> diff --git a/qemu-options.hx b/qemu-options.hx
> index 9a47385c15..757df3eac0 100644
> --- a/qemu-options.hx
> +++ b/qemu-options.hx
> @@ -1172,6 +1172,14 @@ SRST
>       Please also refer to the wiki page for general scenarios of VT-d
>       emulation in QEMU: https://wiki.qemu.org/Features/VT-d.
>   
> +``-device virtio-iommu-pci[,option=...]``
> +    This is only supported by ``-machine q35`` and ``-machine virt``.

Don't we need to precise x86 and arm here?

> +    It supports below options:
> +
> +    ``granule=val`` (possible values are 4k, 8k, 16k, 64k and host; default: host)
> +        This decides the default granule to be be exposed by the
> +        virtio-iommu. If host, the granule matches the host page size.
> +
>   ERST
>   
>   DEF("name", HAS_ARG, QEMU_OPTION_name,

Re: [PATCH v7 4/9] qemu-options.hx: Document the virtio-iommu-pci granule option

Posted by Eric Auger 8 months, 3 weeks ago

Hi Philippe,

On 3/6/24 21:44, Philippe Mathieu-Daudé wrote:
> On 6/3/24 21:32, Eric Auger wrote:
>> We are missing an entry for the virtio-iommu-pci device. Add the
>> information on which machine it is currently supported and document
>> the new granule option.
>>
>> Signed-off-by: Eric Auger <eric.auger@redhat.com>
>> ---
>>   qemu-options.hx | 8 ++++++++
>>   1 file changed, 8 insertions(+)
>>
>> diff --git a/qemu-options.hx b/qemu-options.hx
>> index 9a47385c15..757df3eac0 100644
>> --- a/qemu-options.hx
>> +++ b/qemu-options.hx
>> @@ -1172,6 +1172,14 @@ SRST
>>       Please also refer to the wiki page for general scenarios of VT-d
>>       emulation in QEMU: https://wiki.qemu.org/Features/VT-d.
>>   +``-device virtio-iommu-pci[,option=...]``
>> +    This is only supported by ``-machine q35`` and ``-machine virt``.
>
> Don't we need to precise x86 and arm here?
Well I got inspired of intel_iommu doc. I think this is sufficiently
precise, no?

Thanks

Eric
>
>> +    It supports below options:
>> +
>> +    ``granule=val`` (possible values are 4k, 8k, 16k, 64k and host;
>> default: host)
>> +        This decides the default granule to be be exposed by the
>> +        virtio-iommu. If host, the granule matches the host page size.
>> +
>>   ERST
>>     DEF("name", HAS_ARG, QEMU_OPTION_name,
>

Re: [PATCH v7 4/9] qemu-options.hx: Document the virtio-iommu-pci granule option

Posted by Philippe Mathieu-Daudé 8 months, 3 weeks ago

On 7/3/24 08:32, Eric Auger wrote:
> Hi Philippe,
> 
> On 3/6/24 21:44, Philippe Mathieu-Daudé wrote:
>> On 6/3/24 21:32, Eric Auger wrote:
>>> We are missing an entry for the virtio-iommu-pci device. Add the
>>> information on which machine it is currently supported and document
>>> the new granule option.
>>>
>>> Signed-off-by: Eric Auger <eric.auger@redhat.com>
>>> ---
>>>    qemu-options.hx | 8 ++++++++
>>>    1 file changed, 8 insertions(+)
>>>
>>> diff --git a/qemu-options.hx b/qemu-options.hx
>>> index 9a47385c15..757df3eac0 100644
>>> --- a/qemu-options.hx
>>> +++ b/qemu-options.hx
>>> @@ -1172,6 +1172,14 @@ SRST
>>>        Please also refer to the wiki page for general scenarios of VT-d
>>>        emulation in QEMU: https://wiki.qemu.org/Features/VT-d.
>>>    +``-device virtio-iommu-pci[,option=...]``
>>> +    This is only supported by ``-machine q35`` and ``-machine virt``.
>>
>> Don't we need to precise x86 and arm here?
> Well I got inspired of intel_iommu doc. I think this is sufficiently
> precise, no?

Apparently not:

$ git grep -w '"virt"' hw
hw/arm/virt.c:96:            mc->alias = "virt"; \
hw/m68k/virt.c:314:    .name       = MACHINE_TYPE_NAME("virt"),
hw/openrisc/virt.c:37:#define TYPE_VIRT_MACHINE MACHINE_TYPE_NAME("virt")
hw/riscv/virt.c:1754:    .name       = MACHINE_TYPE_NAME("virt"),
hw/xtensa/virt.c:127:DEFINE_MACHINE("virt", xtensa_virt_machine_init)

Re: [PATCH v7 4/9] qemu-options.hx: Document the virtio-iommu-pci granule option

Posted by Eric Auger 8 months, 3 weeks ago


On 3/7/24 11:21, Philippe Mathieu-Daudé wrote:
> On 7/3/24 08:32, Eric Auger wrote:
>> Hi Philippe,
>>
>> On 3/6/24 21:44, Philippe Mathieu-Daudé wrote:
>>> On 6/3/24 21:32, Eric Auger wrote:
>>>> We are missing an entry for the virtio-iommu-pci device. Add the
>>>> information on which machine it is currently supported and document
>>>> the new granule option.
>>>>
>>>> Signed-off-by: Eric Auger <eric.auger@redhat.com>
>>>> ---
>>>>    qemu-options.hx | 8 ++++++++
>>>>    1 file changed, 8 insertions(+)
>>>>
>>>> diff --git a/qemu-options.hx b/qemu-options.hx
>>>> index 9a47385c15..757df3eac0 100644
>>>> --- a/qemu-options.hx
>>>> +++ b/qemu-options.hx
>>>> @@ -1172,6 +1172,14 @@ SRST
>>>>        Please also refer to the wiki page for general scenarios of
>>>> VT-d
>>>>        emulation in QEMU: https://wiki.qemu.org/Features/VT-d.
>>>>    +``-device virtio-iommu-pci[,option=...]``
>>>> +    This is only supported by ``-machine q35`` and ``-machine virt``.
>>>
>>> Don't we need to precise x86 and arm here?
>> Well I got inspired of intel_iommu doc. I think this is sufficiently
>> precise, no?
>
> Apparently not:
>
> $ git grep -w '"virt"' hw
> hw/arm/virt.c:96:            mc->alias = "virt"; \
> hw/m68k/virt.c:314:    .name       = MACHINE_TYPE_NAME("virt"),
> hw/openrisc/virt.c:37:#define TYPE_VIRT_MACHINE MACHINE_TYPE_NAME("virt")
> hw/riscv/virt.c:1754:    .name       = MACHINE_TYPE_NAME("virt"),
> hw/xtensa/virt.c:127:DEFINE_MACHINE("virt", xtensa_virt_machine_init)

ah ok. I will precise then.

Thanks!

Eric