[ofa-general] [RFC] ipoib: avoid using stale ipoib_neigh* in ipoib_neigh_cleanup()

[Or Gerlitz]

Eli Cohen wrote:
> ipoib_cm_handle_tx_wc() called ipoib_neigh_free() from the first commit. 

Okay, thanks for pointing this out. Looking on the code, I'm not sure why the non sucess/flush path of ipoib_cm_handle_tx_wc() must access the neighbour while ipoib_ib_handle_tx_wc
can get a way with only a warning print... do we agree that accessing the neigbour from the cm tx completion flow is buggy?

Or.