Re: [PATCH RFC 00/12] vdpa: generalize vdpa simulator and add block device

[Stefano Garzarella <sgarzare@redhat.com>]

On Tue, Dec 22, 2020 at 08:29:20PM +0800, Jason Wang wrote:
>
>On 2020/12/22 下午6:57, Stefano Garzarella wrote:
>>On Tue, Dec 22, 2020 at 10:44:48AM +0800, Jason Wang wrote:
>>>
>>>On 2020/12/21 下午7:14, Stefano Garzarella wrote:
>>>>On Mon, Dec 21, 2020 at 11:16:54AM +0800, Jason Wang wrote:
>>>>>
>>>>>On 2020/12/18 下午7:38, Stefano Garzarella wrote:
>>>>>>On Mon, Nov 16, 2020 at 11:37:48AM +0800, Jason Wang wrote:
>>>>>>>
>>>>>>>On 2020/11/13 下午9:47, Stefano Garzarella wrote:
>>>>>>>>Thanks to Max that started this work!
>>>>>>>>I took his patches, and extended the block simulator a bit.
>>>>>>>>
>>>>>>>>This series moves the network device simulator in a new module
>>>>>>>>(vdpa_sim_net) and leaves the generic functions in the 
>>>>>>>>vdpa_sim core
>>>>>>>>module, allowing the possibility to add new vDPA device simulators.
>>>>>>>>Then we added a new vdpa_sim_blk module to simulate a block device.
>>>>>>>>
>>>>>>>>I'm not sure about patch 11 ("vringh: allow 
>>>>>>>>vringh_iov_xfer() to skip
>>>>>>>>bytes when ptr is NULL"), maybe we can add a new 
>>>>>>>>functions instead of
>>>>>>>>modify vringh_iov_xfer().
>>>>>>>>
>>>>>>>>As Max reported, I'm also seeing errors with 
>>>>>>>>vdpa_sim_blk related to
>>>>>>>>iotlb and vringh when there is high load, these are some 
>>>>>>>>of the error
>>>>>>>>messages I can see randomly:
>>>>>>>>
>>>>>>>>  vringh: Failed to access avail idx at 00000000e8deb2cc
>>>>>>>>  vringh: Failed to read head: idx 6289 address 00000000e1ad1d50
>>>>>>>>  vringh: Failed to get flags at 000000006635d7a3
>>>>>>>>
>>>>>>>>  virtio_vdpa vdpa0: vringh_iov_push_iotlb() error: -14 
>>>>>>>>offset:   0x2840000 len: 0x20000
>>>>>>>>  virtio_vdpa vdpa0: vringh_iov_pull_iotlb() error: -14 
>>>>>>>>offset:   0x58ee000 len: 0x3000
>>>>>>>>
>>>>>>>>These errors should all be related to the fact that 
>>>>>>>>iotlb_translate()
>>>>>>>>fails with -EINVAL, so it seems that we miss some mapping.
>>>>>>>
>>>>>>>
>>>>>>>Is this only reproducible when there's multiple co-current 
>>>>>>>accessing of IOTLB? If yes, it's probably a hint that some 
>>>>>>>kind of synchronization is still missed somewhere.
>>>>>>>
>>>>>>>It might be useful to log the dma_map/unmp in both 
>>>>>>>virtio_ring and vringh to see who is missing the map.
>>>>>>>
>>>>>>
>>>>>>Just an update about these issues with vdpa-sim-blk.
>>>>>>I've been focusing a little bit on these failures over the 
>>>>>>last few days and have found two issues related to the 
>>>>>>IOTLB/IOMMU:
>>>>>>
>>>>>>1. Some requests coming from the block layer fills the SG 
>>>>>>list with multiple buffers that had the same physical 
>>>>>>address. This happens for example while using 'mkfs', at 
>>>>>>some points multiple sectors are zeroed so multiple SG 
>>>>>>elements point to the same physical page that is zeroed.
>>>>>>Since we are using vhost_iotlb_del_range() in the 
>>>>>>vdpasim_unmap_page(), this removes all the overlapped 
>>>>>>ranges. I fixed removing a single map in 
>>>>>>vdpasim_unmap_page(), but has an alternative we can 
>>>>>>implement some kind of reference counts.
>>>>>
>>>>>
>>>>>I think we need to do what hardware do. So using refcount is 
>>>>>probably not a good ida.
>>>>
>>>>Okay, so since we are using for simplicity an identical mapping, 
>>>>we are assigning the same dma_addr to multiple pages.
>>>
>>>
>>>I think I get you now. That's the root cause for the failure.
>>
>>Yes, sorry, I didn't explain well previously.
>>
>>>
>>>Then I think we need an simple iova allocator for vdpa simulator, 
>>>and it might be useful for VDUSE as well.
>>
>>Okay, I'll work on it.
>>If you have an example to follow or some pointers, they are welcome :-)
>
>
>Kernel had implemented one in iova.c but I'm not sure we need the 
>complexity like that. Or we can just use rbtree or idr to implement a 
>simpler one.

Yeah, I found it and I started to integrate it in the simulator.
Also if it appears complicated, it seams to me that it should be simple 
to integrate.

I'll give it a try, and if it is too complicate, I'll switch to a simple 
rbtree.

Thanks,
Stefano