Re: [PATCH v3 0/4] virtio/virtio_test

[Guo Zhi]

----- Original Message -----
> From: "Guo Zhi" <qtxuning1999@xxxxxxxxxxx>
> To: "jasowang" <jasowang@xxxxxxxxxx>, "Michael Tsirkin" <mst@xxxxxxxxxx>
> Cc: "eperezma" <eperezma@xxxxxxxxxx>, "virtualization" <virtualization@xxxxxxxxxxxxxxxxxxxxxxxxxx>, "linux-kernel"
> <linux-kernel@xxxxxxxxxxxxxxx>, "sgarzare" <sgarzare@xxxxxxxxxx>
> Sent: Thursday, July 21, 2022 7:55:31 PM
> Subject: Re: [PATCH v3 0/4] virtio/virtio_test

> On 2022/7/9 10:27, Guo Zhi wrote:
>> Original virtio_test only use add one descriptor for each io event, thus code of
>> descriptor chain and indirection have not been tested(one descriptor will not
>> use indirect feature even indirect feature has been specified). In fact it
>> would have not been possible for vhost_test to access to the indirect
>> descriptor table, because it's impossible for virtio_ring.c to allocate it.
>>
>> This series using descriptor chain and enable indirection feature. And through
>> gcov we find the code coverage has been improved(not high for virtio_ring.c
>> because virtio_test only test split virtqueue):
>>
>> +------------+-------------+-------------+
>> |            |virtio_test.c|virtio_ring.c|
>> +------------+-------------+-------------+
>> | original   |   72.32%    |   24.71%    |
>> +------------+-------------+-------------+
>> | current    |    75%      |   28.05%    |
>> +------------+-------------+-------------+
>>
>> Guo Zhi (4):
>>    virtio_test: kick vhost for a batch of descriptors
>>    virtio_test: use random length scatterlists to test descriptor chain
>>    virtio_test: enable indirection feature
>>    virtio_test: pregenerate random numbers
>>
>>   tools/virtio/virtio_test.c | 85 ++++++++++++++++++++++++++++++--------
>>   1 file changed, 68 insertions(+), 17 deletions(-)
>>
> Hi, the new version patch of virtio_test is waiting for review:)
> 
> Thanks.

Friendly ping for the patches:)
Thanks