Re: [ovs-dev] [PATCH] Create a NBL for each NB when required

2014-09-06 Thread Samuel Ghinet
>> Perhaps I should refactor OvsStartNBLIngress in a separate commit. >> I would not like to combine multiple issues in a single commit. Plus, it >> would >> look clearer in the history. > I strongly prefer/advice not to do such refactoring in a separate review. It > wasn't fun reviewing this f

Re: [ovs-dev] [PATCH] Create a NBL for each NB when required

2014-09-06 Thread Samuel Ghinet
Nithin, Thanks a lot for taking the time to review the code! Yes, you were right with the TSO. I'll send a new version of the patch, though I am not sure how it would be easier for you to review. > minor: NET_BUFFER* => PNET_BUFFER, VOID* => PVOID. I am quite confused here, the CodingStyle says

[ovs-dev] datapath-windows: add description to members of OVS_USER_PARAMS_CONTEXT

2014-09-06 Thread Samuel Ghinet
It might be a bit late for this note now, but perhaps a separate patch could add it: > UINT32 devOp;/* Device operation of the userspace call. */ "devOp" is as vague as "Device operation" IMHO. Could you please refer to these flags (defined in Datapath.c): > #define OVS_READ_DEV_OP

[ovs-dev] [PATCH 9/9 v3] datapath-windows: refactor code to setup dump start state

2014-09-06 Thread Samuel Ghinet
Looks good, Sam Date: Fri, 29 Aug 2014 12:05:14 -0700 From: Ankur Sharma To: dev@openvswitch.org Subject: [ovs-dev] [PATCH 9/9 v3] datapath-windows: refactor code to setup dump start state Message-ID: <1409339114-15838-1-git-send-email-ankursha..

Re: [ovs-dev] [PATCH 9/9 v2] datapath-windows: refactor code to setup dump start state

2014-09-06 Thread Samuel Ghinet
Hello guys, I'm sorry I could not reply earlier. What has been pushed in the mean time looks pretty ok. Regarding my "Not acked-by": Generally speaking, I am not very fixed on details. And even if some tiny detail I do not like, if the others like it, then it's ok by me. I simply prefer that so

[ovs-dev] [PATCH] Windows NetLink Socket - Support for asynchronous event notification

2014-09-06 Thread Samuel Ghinet
Looks good, as far as I can tell. One minor style thing though: > +goto done; > +} > +} > +else { > +/* The I/O was completed synchronously */ > +poll_immediate_wake(); > +} I think it should have been: > } else { Regards, Sam _

Re: [ovs-dev] [PATCH 9/9 v3] datapath-windows: refactor code to setup dump start state

2014-09-06 Thread Samuel Ghinet
Acked-by: Samuel Ghinet From: Samuel Ghinet Sent: Sunday, September 07, 2014 2:54 AM To: dev@openvswitch.org Cc: Alin Serdean; nit...@vmware.com; ssaur...@vmware.com; Ankur Sharma Subject: [PATCH 9/9 v3] datapath-windows: refactor code to setup

Re: [ovs-dev] [PATCH v2 2/6] NetlinkBuf.c: Netlink buffer mgmt apis.

2014-09-06 Thread Samuel Ghinet
Hello Ankur, I've got one questions about the buffer management and netlink put functions: Do we have here some netlink put functions to use for nested netlink attributes? Thanks, Samuel Date: Thu, 4 Sep 2014 09:44:45 -0700 From: Ben Pfaff To: Ankur Sha