Hello Ferruh,

On Friday 24 February 2017 03:28 PM, Ferruh Yigit wrote:

[snip]

quoted

Now, we have these possibility:
1. Have a shared library with non rte_* symbols
2. We have shared library with rte_* symbols
3. We have non-net devices (crypto, eventdev, ..) depend on net for
these hardware interfaces

(2) is hitting performance significantly.
(3) it not a clean solution, having driver/crypto depend on driver/net.
When new devices are there, more dependencies will occur.

In crux, probably we need to have a discussion on (1) and how strongly
we feel about that (specially in context of drivers).

Insight of above information, I would be OK with (1).

Great. Thank you for understanding.

We can go with option (1) now, since these are not real APIs to user
application, it can be possible to change them if better solution found.

Do you think is it good idea to have different naming syntax for those
libraries to clarify they are for PMD internal usage?

Indeed. Current name is librte_common_dpaa2_*.
Do you think librte_drvlib_dpaa2 or librte_drvlib_dpaa2_pmd is better?