On 17-04-15 10:12 AM, Jamal Hadi Salim wrote:
On 17-04-12 09:22 PM, Cong Wang wrote:
Sorry - didnt get the full thread (the OP managed to post to CC linux kernel list but not me); trying to catch up. So the issue if i understood: Policer was created but failed and we the error code mistakenly did a hash_release? I do observe an issue if the above was to happen. I think we need two variables instead of one. Will dig through the list.
I found them. The patches look sane. Please resend v2 (the second version of patch 2 where you moved module check up) and add my Acked-by. Also - please dont cross post to linux kernel list (and Cc me). I think Roman has slightly different use case but he triggers a related issue your patch seems to address. cheers, jamal
