Subject: Re: [Qemu-devel] [virtio-dev] Re: [PATCH v1] virtio-net: enable configurable tx queue size
On 2017年06月13日 11:55, Jason Wang wrote:
On 2017年06月13日 11:51, Wei Wang wrote:On 06/13/2017 11:19 AM, Jason Wang wrote:On 2017年06月13日 11:10, Wei Wang wrote:On 06/13/2017 04:43 AM, Michael S. Tsirkin wrote:On Mon, Jun 12, 2017 at 05:30:46PM +0800, Wei Wang wrote:Ping for comments, thanks.This was only posted a week ago, might be a bit too short for some people.OK, sorry for the push.A couple of weeks is more reasonable before you ping. Also, I sent a bunch of comments on Thu, 8 Jun 2017. You should probably address these.I responded to the comments. The main question is that I'm not sure why we need the vhost backend to support VIRTIO_F_MAX_CHAIN_SIZE. IMHO, that should be a feature proposed to solve the possible issue caused by the QEMU implemented backend.The issue is what if there's a mismatch of max #sgs between qemu and vhost?When the vhost backend is used, QEMU is not involved in the data path. The vhost backenddirectly gets what is offered by the guest from the vq.
FYI, qemu will try to fallback to userspace if there's something wrong with vhost-kernel (e.g the IOMMU support). This doesn't work for vhost-user actually, but it works for vhost-kernel.
Why would there be a mismatch ofmax #sgs between QEMU and vhost, and what is the QEMU side max #sgs used for? Thanks.You need query the backend max #sgs in this case at least. no? If not how do you know the value is supported by the backend?ThanksBest, Wei