JannePeltonen

Yes, it is still a (potential) issue. It is not related to the signal handling issue but about how at fork() all the other threads magically disappear (from child process...

@Bill-Fischofer-Linaro I leave it to others whether this issue should be kept open or not. The potential problem still exists, since AFAIK nothing has been done about it in ODP...

Is this PR still relevant? If yes, it should be rebased and I think the two changes (loop pktio and prio mode overwrite) should be separated into two patches.

What is the purpose of this change? Please provide some context and rationale in the commit message.

> Inline outbound processing would require parsed IP packet to help with the result packet length calculations. The current ODP API does not require parsed IP packet for outbound IPsec...

v2: fixed a couple of typos.

ODP API does not directly support link aggregation but allows the application to implement it in SW on top of the ODP API. In addition, ODP API does not preclude...

I think the meaning of "inner header", "inner L2", "inner L3" may not be sufficiently unambiguously defined. What if there are multiple tunneling encapsulations in the same packet? What if...

v2: fixed a couple of typos.

v3: Split the second commit to two commits so that the addition of ODP_CRYPTO_OP_TYPE_COMBINE is now separate from the addition of the crypto op type. There are no actual code...