On 23/01/2017 11:56, Yuanhan Liu wrote:
[..]

quoted

quoted

http://dpdk.org/ml/archives/dev/2017-January/054422.html

Yes, it should fix that issue.

Well, few more thoughts: it may fix the crash issue Remy saw, but it
looks like more a workaround to me. Basically, if primary and secondary
shares a same port id, they should point to same device. Otherwise,
primary process may use eth_dev->data for a device A, while the
secondary process may use it for another device, as you said, it
could be a vdev.

In such case, there is no way we could continue safely. That said,
the given patch avoids the total reset of eth_dev->data, while it
continues reset the eth_dev->data->name, which is wrong.

I did wonder whether 7f95f78a8aea ought to be rolled back rather than 
the memset being made process-conditional. You going to be fixing the 
issue in your own patch?

quoted

One question: do Remy or you regularly
run some multiple process test cases (and with vdev both in primary
and secondary process)?

Not aware of there being any multiproc-related unit tests.

..Remy