Re: [PATCH] vfio: Remove VFIO_TYPE1_NESTING_IOMMU

From: "zhangfei.gao@foxmail.com" <zhangfei.gao@foxmail.com>
To: Jason Gunthorpe <jgg@nvidia.com>
Cc: Jean-Philippe Brucker <jean-philippe@linaro.org>,
	kvm@vger.kernel.org, Robin Murphy <robin.murphy@arm.com>,
	Cornelia Huck <cohuck@redhat.com>,
	iommu@lists.linux-foundation.org,
	Vivek Kumar Gautam <Vivek.Gautam@arm.com>,
	Alex Williamson <alex.williamson@redhat.com>,
	Zhangfei Gao <zhangfei.gao@linaro.org>,
	Will Deacon <will@kernel.org>,
	linux-arm-kernel@lists.infradead.org
Subject: Re: [PATCH] vfio: Remove VFIO_TYPE1_NESTING_IOMMU
Date: Thu, 12 May 2022 20:07:41 +0800	[thread overview]
Message-ID: <tencent_D40360903A6274DB4A4DA4A93F2DBC01BA05@qq.com> (raw)
In-Reply-To: <20220512115906.GX49344@nvidia.com>

On 2022/5/12 下午7:59, Jason Gunthorpe wrote:
> On Thu, May 12, 2022 at 07:57:26PM +0800, zhangfei.gao@foxmail.com wrote:
>>
>> On 2022/5/12 下午7:32, Jason Gunthorpe via iommu wrote:
>>> On Thu, May 12, 2022 at 03:07:09PM +0800, Zhangfei Gao wrote:
>>>>>> I can't help feeling a little wary about removing this until IOMMUFD
>>>>>> can actually offer a functional replacement - is it in the way of
>>>>>> anything upcoming?
>>>>>    From an upstream perspective if someone has a patched kernel to
>>>>> complete the feature, then they can patch this part in as well, we
>>>>> should not carry dead code like this in the kernel and in the uapi.
>>>>>
>>>>> It is not directly in the way, but this needs to get done at some
>>>>> point, I'd rather just get it out of the way.
>>>> We are using this interface for nested mode.
>>> How are you using it? It doesn't do anything. Do you have out of tree
>>> patches as well?
>> Yes, there are some patches out of tree, since they are pending for almost
>> one year.
>>
>> By the way, I am trying to test nesting mode based on iommufd, still
>> requires iommu_enable_nesting,
>> which is removed in this patch as well.
> This is temporary.
>
>> So can we wait for alternative patch then remove them?
> Do you see a problem with patching this along with all the other
> patches you need?
Not yet.

Currently I am using two workarounds, but it can simply work.

1. Hard code to to enable nesting mode, both in kernel and qemu.
Will consider then.

2. Adding VFIOIOASHwpt *hwpt in VFIOIOMMUFDContainer.
And save hwpt when vfio_device_attach_container.

While currently VFIOIOMMUFDContainer has hwpt_list now.
Have asked Yi in another mail.

Thanks

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