[ovs-dev] [PATCH 3/4 v2] datapath-windows: Use l2 port and tunkey during execute

attributes sent down from userspace. Signed-off-by: Nithin Raju --- datapath-windows/ovsext/User.c | 32 ++-- 1 file changed, 30 insertions(+), 2 deletions(-) diff --git a/datapath-windows/ovsext/User.c b/datapath-windows/ovsext/User.c index 3b3f662..2312940 100644 --- a

[ovs-dev] [PATCH 4/4 v2] datapath-windows: remove extract flow in OvsDoRecirc()

It is not necessary to do a flow extract in OvsDoRecirc(). In fact, doing it would overwrite the tunnel key within 'key'. So, let's remove the call. Signed-off-by: Nithin Raju Signed-off-by: Sairam Venugopal Co-Authored-by: Sairam Venugopal --- datapath-windows/ovsex

Re: [ovs-dev] [PATCH V2] datapath-windows: Validate netlink packets integrity

Hi Paul, Can you point out the particular code that fixed the access violation? I looked at the code, and code such as the following is redundant: +// We need to ensure we have enough data to process +if (NlMsgSize(&ovsMsg->nlMsg) > ovsMsgLength) { +status = STATUS_INVALID_PARAMETE

Re: [ovs-dev] [PATCH V2] datapath-windows: Validate netlink packets integrity

Just a couple of more comments. -Original Message- From: dev mailto:dev-boun...@openvswitch.org>> on behalf of Paul Boca mailto:pb...@cloudbasesolutions.com>> Date: Wednesday, April 27, 2016 at 1:05 AM To: "dev@openvswitch.org" mailto:dev@openvswitch.org>> Su

Re: [ovs-dev] [PATCH] datapath-windows: Sample action support.

Hi Sorin, Thanks for this patch, and thanks for the refactoring. One quick comment is that, we can probably move the new code in Random.h to Util.h if you think we are not going to be adding any more functions to Random.h. I am not sure if 2 more functions warrants a new file. Also, I see a typo

Re: [ovs-dev] [PATCH] Add Geneve support on Windows datapath.

>> >>I am not sure why Windows kernel doesn't compute UDP checksum for all >>tunnels. I added Nithin to the thread to see if he has some idea. My >>wild guess is that it tries to save some computational workload. > >My guess is that it is because VXLAN originally specified that >checksums should no

Re: [ovs-dev] [PATCH] Add Geneve support on Windows datapath.

Yin, Thanks for sending the patch for review. A few things before I take a detailed look: 1. Add a version number to the patch, eg. "[PATCH v2] Add Geneve support Š ³ with each iteration. Also, it would be beneficial if you can annotate what changed in v2 compared to previous versions. 2. I see th

Re: [ovs-dev] [PATCH 2/2] datapath-windows: Fix alignment on Offload.c

Acked-by: Nithin Raju -Original Message- From: dev on behalf of Alin Serdean Date: Wednesday, May 18, 2016 at 9:32 AM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH 2/2] datapath-windows: Fix alignment on Offload.c >Found by inspection. > >Signed-off-by: A

Re: [ovs-dev] [PATCH 1/2] datapath-windows: Add UDP checksum verifications for VXLAN

Thanks Alin for taking this up. A couple of nits but looks great otherwise. Acked-by: Nithin Raju > >@@ -287,17 +289,21 @@ OvsDoEncapVxlan(POVS_VPORT_ENTRY vport, > ipHdr->saddr = fwdInfo->srcIpAddr; > ipHdr->daddr = fwdInfo->dstIpAddr; > >

Re: [ovs-dev] [PATCH V2] datapath-windows: Validate netlink packets integrity

Hi Paul, I looked at the change in detail and it is definitely in the right spirit to harden the kernel datapath code. However, I thought a few things could be simplified a little. I will be sending out a couple of simple reviews on top of your patch (that is already submitted). Pls. take a look.

[ovs-dev] [PATCH 2/2] datapath-windows: o/p buffer must fit NL error message

which is smaller. Since we are ensuring that output buffer can fit OVS_MESSAGE_ERROR at the top of the ioctl function, there's no need to check for that later. Signed-off-by: Nithin Raju --- datapath-windows/ovsext/Datapath.c| 19 -- datapath-windows/ovsext/F

[ovs-dev] [PATCH 1/2] datapath-windows: don't map output buffer in OVS_IOCTL_WRITE

The contract of OVS_IOCTL_WRITE is that write operations will not need the output buffer. Only the input buffer will be used in the IRP. So, better to not map the output buffer at all. Signed-off-by: Nithin Raju --- datapath-windows/ovsext/Datapath.c | 13 - 1 file changed, 13

Re: [ovs-dev] [PATCH 1/3] ovn test: add '-O OpenFlow13' to ovs-ofctl

Yin, Looks like you did a ³git send-email² on a patch that was not yours. It is fine, I¹ll review the remaining 2 in the series. Thanks for breaking it up. -- Nithin -Original Message- From: dev on behalf of Yin Lin Date: Thursday, May 19, 2016 at 2:49 PM To: "dev@openvswitch.org" Cc:

Re: [ovs-dev] [PATCH 2/3] [PATCH v1] Move UDP checksum computation to Offload.c

gt; > if (*newNbl == NULL) { We don¹t need to use tunnelSize + OVS_DEFAULT_COPY_SIZE in the partial copy. Just ŒtunnelSize¹ is sufficient. But, not a big deal. Acked-by: Nithin Raju ___ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev

Re: [ovs-dev] [PATCH 3/3] [PATCH v3] Add Geneve support in Windows datapath

-Original Message- From: dev on behalf of Yin Lin Date: Thursday, May 19, 2016 at 2:49 PM To: "dev@openvswitch.org" Cc: Yin Lin Subject: [ovs-dev] [PATCH 3/3] [PATCH v3] Add Geneve support in Windows datapath >OvsFindTunnelVportByDstPort(POVS_SWITCH_CONTEXT switchContext, >-

Re: [ovs-dev] [PATCH 1/2] [PATCH v2] datapath-windows: Move UDP checksum computation to Offload.c

I am assuming that you¹ve run some tests using VXLAN tunnels. Acked-by: Nithin Raju Thanks for the patch. -Original Message- From: dev on behalf of Yin Lin Date: Friday, May 20, 2016 at 1:57 PM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH 1/2] [PATCH v2] datapa

Re: [ovs-dev] [PATCH 2/2] [PATCH v4] datapath-windows: Add Geneve support

Hi Yin, You also need to update datapath-windows/automake.mk to add the new files you added. -Original Message- From: dev on behalf of Yin Lin Date: Friday, May 20, 2016 at 1:57 PM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH 2/2] [PATCH v4] datapath-windows: Add Geneve support

Re: [ovs-dev] [PATCH 2/2] [PATCH v4] datapath-windows: Add Geneve support

behalf of Nithin Raju Date: Tuesday, May 24, 2016 at 9:32 AM To: Yin Lin , "dev@openvswitch.org" Subject: Re: [ovs-dev] [PATCH 2/2] [PATCH v4] datapath-windows: Add Geneve support >Hi Yin, >You also need to update datapath-windows/automake.mk to add the new files >you added. >

[ovs-dev] [PATCH] list.h: Define OVS_LIST_POISON statically

Looks like part of the patch committed in e32c1f7c got left out while moving header files. Signed-off-by: Nithin Raju Reported-by: Joe Stringer --- include/openvswitch/list.h | 3 --- 1 file changed, 3 deletions(-) diff --git a/include/openvswitch/list.h b/include/openvswitch/list.h index

Re: [ovs-dev] [PATCH v2 1/3] datapath-windows: Add UDP checksum verifications for VXLAN

Acked-by: Nithin Raju -Original Message- From: dev on behalf of Alin Serdean Date: Tuesday, May 24, 2016 at 9:14 AM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH v2 1/3] datapath-windows: Add UDP checksum verifications for VXLAN >Introduce UDP checksum if it was

Re: [ovs-dev] [PATCH v2 3/3] datapath-windows: Add GRE checksum

-Original Message- From: dev on behalf of Alin Serdean Date: Tuesday, May 24, 2016 at 9:14 AM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH v2 3/3] datapath-windows: Add GRE checksum >@@ -369,11 +383,25 @@ OvsDecapGre(POVS_SWITCH_CONTEXT switchContext, > headRoom += 4; >

Re: [ovs-dev] [PATCH 1/2] [PATCH v5] datapath-windows: Move UDP checksum computation to Offload.c

Acked-by: Nithin Raju -Original Message- From: dev on behalf of Yin Lin Date: Tuesday, May 24, 2016 at 4:28 PM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH 1/2] [PATCH v5] datapath-windows: Move UDP checksum computation to Offload.c >UDP checksum computation is s

[ovs-dev] [PATCH] datapath-windows: use ip proto for tunnel port lookup

the vport add code, we make sure that we block tunnel port addition if there's already a tunnel port that uses the same IP protocol type and L4 port number. This is required for Geneve port lookups the references to which can find in the patch. Signed-off-by: Nithin Raju --- datapath-wi

Re: [ovs-dev] [PATCH 2/2] [PATCH v5] datapath-windows: Add Geneve support

Yin, I went ahead and sent out a patch for the updates w.r.t OvsFindTunnelVportByDstPort(). Pls have a look: https://patchwork.ozlabs.org/patch/628483/ I¹ll send out the comments for the remainder of the patch. Thanks, -- Nithin -Original Message- From: dev on behalf of Yin Lin Date:

Re: [ovs-dev] [PATCH] datapath-windows: use ip proto for tunnel port lookup

Thanks for the review. From: Yin Lin mailto:yinli...@gmail.com>> Date: Thursday, June 2, 2016 at 3:00 PM To: Nithin Raju mailto:nit...@vmware.com>> Cc: "dev@openvswitch.org<mailto:dev@openvswitch.org>" mailto:dev@openvswitch.org>> Subject: Re: [ovs-dev] [PATCH]

[ovs-dev] [PATCH v2] datapath-windows: use ip proto for tunnel port lookup

the vport add code, we make sure that we block tunnel port addition if there's already a tunnel port that uses the same IP protocol type and L4 port number. This is required for Geneve port lookups the references to which can find in the patch. Signed-off-by: Nithin Raju --- datapath-wi

Re: [ovs-dev] [PATCH 2/2] [PATCH v5] datapath-windows: Add Geneve support

-Original Message- From: Ben Pfaff Date: Friday, June 3, 2016 at 11:55 AM To: Nithin Raju Cc: Yin Lin , "dev@openvswitch.org" Subject: Re: [ovs-dev] [PATCH 2/2] [PATCH v5] datapath-windows: Add Geneve support >I wasn't sure how to interpret this. Does this mean tha

Re: [ovs-dev] [PATCH v3 1/2] Windows: Add support for handling protocol (netlink family)

Acked-by: Nithin Raju -Original Message- From: Sairam Venugopal Date: Tuesday, July 12, 2016 at 2:41 PM To: Nithin Raju Subject: FW: [PATCH v3 1/2] Windows: Add support for handling protocol (netlink family) > > >On 7/11/16, 2:59 PM, "Sairam Venugopal" wrote: &

Re: [ovs-dev] [PATCH v3 2/2] Windows: Use NETLINK_NETFILTER instead of NETLINK_GENERIC

Acked-by: Nithin Raju One another change that comes to mind is the validation that a socket of type NETLINK_GENERIC does not send a CT command and vice-versa. It should be a small change within Datapath.c and can be done as a incremental patch. -Original Message- From: Sairam Venugopal

Re: [ovs-dev] [PATCH] datapath-windows: Fix various Geneve bugs

-Original Message- From: dev on behalf of Yin Lin Date: Wednesday, July 13, 2016 at 8:21 PM To: "dev@openvswitch.org" Cc: Yin Lin Subject: [ovs-dev] [PATCH] datapath-windows: Fix various Geneve bugs >Signed-off-by: Yin Lin Acked-by

Re: [ovs-dev] [RFC] datapath-windows: Allow validation ports processing

Alin, Do you intend to add them to the vport hash tables? What is the use of doing this? Also, will it not create a collision in the hash table? When the vport gets added from userspace, we should make sure that we don¹t use the validation port as a valid port. -- Nithin -Original Message---

Re: [ovs-dev] [PATCH v3] datapath-windows: Add GRE checksum

> #endif >@@ -299,34 +312,21 @@ OvsDecapGre(POVS_SWITCH_CONTEXT switchContext, > EthHdr *ethHdr; > IPHdr *ipHdr; > GREHdr *greHdr; >-UINT32 tunnelSize = 0, packetLength = 0; >+UINT32 tunnelSize, packetLength; > UINT32 headRoom = 0; > PUINT8 bufferStart; > NDIS_STATU

Re: [ovs-dev] [PATCH v3] datapath-windows: Add GRE checksum

>> >>-Mesaj original- >> De la: Nithin Raju [mailto:nit...@vmware.com] >> Trimis: Wednesday, June 8, 2016 10:10 PM >> Către: Alin Serdean ; >> dev@openvswitch.org >> Subiect: Re: [ovs-dev] [PATCH v3] datapath-windows: Add GRE checksum >>

Re: [ovs-dev] [PATCH v2] datapath-windows: Sample action support.

Acked-by: Nithin Raju -Original Message- From: dev on behalf of Sorin Vinturis Date: Wednesday, June 1, 2016 at 8:50 AM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH v2] datapath-windows: Sample action support. >This patch adds support for sampling to the OVS exte

Re: [ovs-dev] [PATCH] datapath-windows: use ip proto for tunnel port lookup

Hi Alin, Thanks for the review. I¹ve taken care of the extra whitespace and also the dead (placeholder) code. >> --- a/datapath-windows/ovsext/Vport.c > >> +++ b/datapath-windows/ovsext/Vport.c > >> POVS_VPORT_ENTRY > >> +OvsFindTunnelVportByDstPortAndNWProto(POVS_SWITCH_CONTEXT > >> switchConte

[ovs-dev] [PATCH v3] datapath-windows: use ip proto for tunnel port lookup

the vport add code, we make sure that we block tunnel port addition if there's already a tunnel port that uses the same IP protocol type and L4 port number. Signed-off-by: Nithin Raju Acked-by: Sairam Venugopal Acked-by: Yin Lin --- v3: addresed Alin's comments v2: addresed Yin&

Re: [ovs-dev] [PATCH v3] datapath-windows: Add GRE checksum

Glad we were able to resolve this :) -Original Message- From: Alin Serdean Date: Tuesday, June 14, 2016 at 9:12 AM To: Nithin Raju , "dev@openvswitch.org" Subject: RE: [ovs-dev] [PATCH v3] datapath-windows: Add GRE checksum >> >greater than the maximum header size wh

Re: [ovs-dev] [PATCH v6] datapath-windows: Add Geneve support

: "dev@openvswitch.org" , Nithin Raju Subject: [PATCH v6] datapath-windows: Add Geneve support >Signed-off-by: Yin Lin >--- > datapath-windows/automake.mk | 2 + > datapath-windows/ovsext/Actions.c | 74 ++- > datapath-windows/ovsext/Debug.h|

Re: [ovs-dev] [PATCH v6] datapath-windows: Add Geneve support

Hi Jesse, Thanks for the comments. My responses inlined. >>> /* Number of packet attributes required to store OVS tunnel key. */ >>>-#define NUM_PKT_ATTR_REQUIRED 3 >>>+#define NUM_PKT_ATTR_REQUIRED 35 >>>+#define TUN_OPT_MAX_LEN 255 >> >> Wouldn¹t it have been better for alignment purposes to hav

[ovs-dev] [PATCH] datapath-windows: comment cleanup and indentation

Signed-off-by: Nithin Raju --- datapath-windows/ovsext/Vxlan.c | 5 - 1 file changed, 5 deletions(-) diff --git a/datapath-windows/ovsext/Vxlan.c b/datapath-windows/ovsext/Vxlan.c index f6a55f0..ddd8d8e 100644 --- a/datapath-windows/ovsext/Vxlan.c +++ b/datapath-windows/ovsext/Vxlan.c

[ovs-dev] [PATCH] datapath-windows: comment cleanup and indentation

Signed-off-by: Nithin Raju --- datapath-windows/ovsext/Flow.c | 20 ++-- datapath-windows/ovsext/Vxlan.c | 5 - 2 files changed, 10 insertions(+), 15 deletions(-) diff --git a/datapath-windows/ovsext/Flow.c b/datapath-windows/ovsext/Flow.c index 2a91855..595518f 100644

Re: [ovs-dev] [PATCH v3] datapath-windows: use ip proto for tunnel port lookup

>> >>+case IPPROTO_GRE: >[Alin Gabriel Serdean: ] break; >> +default: >[Alin Gabriel Serdean: ] return NULL; >> +break; >> +} >> +return vport; >> +} >> +} >> +return NULL; >> +} >We need to exact match the IP proto. In

[ovs-dev] [PATCH v4] datapath-windows: use ip proto for tunnel port lookup

the vport add code, we make sure that we block tunnel port addition if there's already a tunnel port that uses the same IP protocol type and L4 port number. Signed-off-by: Nithin Raju Acked-by: Sairam Venugopal Acked-by: Yin Lin --- datapath-windows/ovsext/Actions.c

Re: [ovs-dev] [PATCH v6] datapath-windows: Add Geneve support

From: Yin Lin mailto:yinli...@gmail.com>> Date: Friday, June 17, 2016 at 5:10 PM To: "aserd...@cloudbasesolutions.com<mailto:aserd...@cloudbasesolutions.com>" mailto:aserd...@cloudbasesolutions.com>>, Nithin Raju mailto:nit...@vmware.com>> Cc: Yin Li

Re: [ovs-dev] [PATCH 1/2][PATCH v7] datapath-windows: Add Geneve support

Acked-by: Nithin Raju -Original Message- From: dev on behalf of Yin Lin Date: Friday, June 17, 2016 at 5:13 PM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH 1/2][PATCH v7] datapath-windows: Add Geneve support >Signed-off-by: Yin Lin >--- > datapath-

Re: [ovs-dev] [PATCH v4] datapath-windows: Add GRE checksum

ntation cases >v3: Fix access length and address comments >v2: Initial commit Acked-by: Nithin Raju ___ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev

Re: [ovs-dev] [PATCH 2/2] datapath-windows: Address minor alignment issues in Stt code.

Acked-by: Nithin Raju -Original Message- From: dev on behalf of Yin Lin Date: Friday, June 24, 2016 at 2:44 PM To: "dev@openvswitch.org" Cc: Yin Lin Subject: [ovs-dev] [PATCH 2/2] datapath-windows: Address minor alignment issues in Stt code. >Signed-of

Re: [ovs-dev] [PATCH] FAQ: Update support for NAT and Geneve.

Acked-by: Nithin Raju -Original Message- From: dev on behalf of Jesse Gross Date: Friday, June 24, 2016 at 2:58 PM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH] FAQ: Update support for NAT and Geneve. >Signed-off-by: Jesse Gross >--- > FAQ.md | 4 ++--

Re: [ovs-dev] [PATCH] datapath-windows: Conntrack - Fix variable initialization

Sai, There¹s not correctness issue with the existing code. There are 2 fields in pktMdLabel, and they are being set in the function. Do you see any correctness issues? -- Nithin -Original Message- From: dev on behalf of Sairam Venugopal Date: Friday, June 24, 2016 at 6:16 PM To: "dev@op

Re: [ovs-dev] [PATCH] datapath-windows: Conntrack - Fix variable initialization

; >>>Initialize the variable pktMdLabel. >>> >>>Signed-off-by: Sairam Venugopal Never mind my previous oversight. Acked-by: Nithin Raju ___ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev

Re: [ovs-dev] [PATCH v3 1/9] Windows: Add netfilter-conntrack header files to Include folder

Only comment is to add the OVS license on top of each non-empty file. Looks like the original code does not have any licence, but it would still make sense to add a license I think. Looks good otherwise, Acked-by: Nithin Raju Ben or Jesse can confirm what the best practice is. Thanks, -- Nithin

Re: [ovs-dev] [PATCH v3 2/9] datapath-windows: Add support for Netfilter netlink message

A couple of minor comments, but looks good otherwise. Acked-by: Nithin Raju >+/* >+ * >-- >- >+ * Returns size of to nfn

Re: [ovs-dev] [PATCH v3 3/9] datapath-windows: Add support for flushing conntrack entries

CTA_ZONE is not defined yet, and I see it is defined in 4/9. Don¹t you have to return anything upon success in OvsCtDeleteCmdHandler(). The operation type is a OVS_TRANSACTION_DEV_OP. See Flow.c for example: https://github.com/openvswitch/ovs/blob/master/datapath-windows/ovsext/Flow .c#L301 -- N

Re: [ovs-dev] [PATCH v3 1/9] Windows: Add netfilter-conntrack header files to Include folder

>On Mon, Jun 27, 2016 at 11:57 AM, Nithin Raju wrote: >> Only comment is to add the OVS license on top of each non-empty file. >> Looks like the original code does not have any licence, but it would >>still >> make sense to add a license I think. Looks good otherwis

Re: [ovs-dev] [PATCH v3 4/9] datapath-windows: Add support for Conntrack IPCTNL_MSG_CT_DELETE cmd in Datapath.c

Looks good but for a couple of suggestions. >/* Windows kernel datapath extensions to the standard datapath interface. >*/ > > /* Version number of the datapath interface extensions. */ >@@ -65,6 +68,7 @@ > #define OVS_WIN_NL_VPORT_FAMILY_ID (NLMSG_MIN_TYPE + 4) > #define OVS_WIN_NL_FLO

Re: [ovs-dev] [PATCH v3 5/9] Windows: Add conntrack-flush support in userspace

Just one suggestion, but looks good otherwise. Acked-by: Nithin Raju -Original Message- From: dev on behalf of Sairam Venugopal Date: Friday, June 24, 2016 at 12:03 PM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH v3 5/9] Windows: Add conntrack-flush support in

Re: [ovs-dev] [PATCH v3 6/9] datapath-windows: Add new NlFillOvsMsgForNfGenMsg method in Netlink.c

Acked-by: Nithin Raju -Original Message- From: dev on behalf of Sairam Venugopal Date: Friday, June 24, 2016 at 12:03 PM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH v3 6/9] datapath-windows: Add new NlFillOvsMsgForNfGenMsg method in Netlink.c >Create a

Re: [ovs-dev] [PATCH v3 7/9] datapath-windows: Add support for dump-conntrack in datapath

Looks good but for a few comments. Acked-by: Nithin Raju -Original Message- From: dev on behalf of Sairam Venugopal Date: Friday, June 24, 2016 at 12:03 PM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH v3 7/9] datapath-windows: Add support for dump-conntrack i

Re: [ovs-dev] [PATCH v3 8/9] datapath-windows: Conntrack - Handle memory allocation failure

emory for the conntrack entry. > >Signed-off-by: Sairam Venugopal >Acked-by: Paul-Daniel Boca Acked-by: Nithin Raju http://openvswitch.org/mailman/listinfo/dev

Re: [ovs-dev] [PATCH v3 9/9] datapath-windows: Add support for Conntrack IPCTNL_MSG_CT_GET cmd in Datapath.c

-Original Message- From: dev on behalf of Sairam Venugopal Date: Friday, June 24, 2016 at 12:03 PM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH v3 9/9] datapath-windows: Add support for Conntrack IPCTNL_MSG_CT_GET cmd in Datapath.c >This will be used by userspace for dumpin

Re: [ovs-dev] [PATCH] datapath-windows: Cleanup conntrack-tcp.c

Looks good but for a comment. Acked-by: Nithin Raju >On 6/24/16, 6:14 PM, "Sairam Venugopal" wrote: > >>Update the code to use tcp->flags. This keeps the kernel conntrack-tcp.c >>file in sync with userspace version. >> >>This patch also addr

Re: [ovs-dev] [PATCH] datapath-windows: Cleanup conntrack-tcp.c

VALID; >>> } >>> >>>-if ((tcp->syn) && dst->state >= CT_DPIF_TCPS_FIN_WAIT_2 && >>>-src->state >= CT_DPIF_TCPS_FIN_WAIT_2) { >>>+if (((tcp_flags & (TCP_SYN|TCP_ACK)) == TCP_SYN) > >Why do we need to include TCP_ACK in the check? We spoke offline. This is a bug fix. Acked-by: Nithin Raju ___ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev

Re: [ovs-dev] [PATCH v4 01/11] Windows: Add conntrack netfilter netlink definitions to kernel and userspace

Acked-by: Nithin Raju -Original Message- From: dev on behalf of Sairam Venugopal Date: Wednesday, June 29, 2016 at 5:08 PM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH v4 01/11] Windows: Add conntrack netfilter netlink definitions to kernel and userspace >Incl

Re: [ovs-dev] [PATCH v4 02/11] datapath-windows: Add support for Netfilter netlink message

Acked-by: Nithin Raju -Original Message- From: dev on behalf of Sairam Venugopal Date: Wednesday, June 29, 2016 at 5:08 PM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH v4 02/11] datapath-windows: Add support for Netfilter netlink message >Introduce NF_GEN_MS

Re: [ovs-dev] [PATCH v4 03/11] datapath-windows: Add new NlFillOvsMsgForNfGenMsg method in Netlink.c

Acked-by: Nithin Raju -Original Message- From: dev on behalf of Sairam Venugopal Date: Wednesday, June 29, 2016 at 5:08 PM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH v4 03/11] datapath-windows: Add new NlFillOvsMsgForNfGenMsg method in Netlink.c >Create a

Re: [ovs-dev] [PATCH v4 04/11] datapath-windows: Add support for flushing conntrack entries

Looks good but for a few comments. >+static __inline NDIS_STATUS >+OvsCtFlush(UINT16 zone) >+{ >+PLIST_ENTRY link, next; >+POVS_CT_ENTRY entry; >+ >+LOCK_STATE_EX lockState; >+NdisAcquireRWLockWrite(ovsConntrackLockObj, &lockState, 0); >+ >+for (int i = 0; i < CT_HASH_TABLE_SIZ

Re: [ovs-dev] [PATCH v4 05/11] datapath-windows: Add support for Conntrack IPCTNL_MSG_CT_DELETE cmd in Datapath.c

Acked-by: Nithin Raju >+ > /* Windows kernel datapath extensions to the standard datapath >interface. */ > > /* Version number of the datapath interface extensions. */ >@@ -65,6 +68,8 @@ > #define OVS_WIN_NL_VPORT_FAMILY_ID (NLMSG_MIN_TYPE + 4) > #define O

Re: [ovs-dev] [PATCH v4 07/11] datapath-windows: Conntrack - Handle memory allocation failure

Acked-by: Nithin Raju -Original Message- From: dev on behalf of Sairam Venugopal Date: Wednesday, June 29, 2016 at 5:08 PM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH v4 07/11] datapath-windows: Conntrack - Handle memory allocation failure >Return null if Win

Re: [ovs-dev] [PATCH v4 06/11] datapath-windows: Add support for dump-conntrack in datapath

Acked-by: Nithin Raju -Original Message- From: dev on behalf of Sairam Venugopal Date: Wednesday, June 29, 2016 at 5:08 PM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH v4 06/11] datapath-windows: Add support for dump-conntrack in datapath >Create the met

Re: [ovs-dev] [PATCH v4 08/11] datapath-windows: Add support for Conntrack IPCTNL_MSG_CT_GET cmd in Datapath.c

Acked-by: Nithin Raju -Original Message- From: dev on behalf of Sairam Venugopal Date: Wednesday, June 29, 2016 at 5:08 PM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH v4 08/11] datapath-windows: Add support for Conntrack IPCTNL_MSG_CT_GET cmd in Datapath.c >

Re: [ovs-dev] [PATCH v4 09/11] datapath-windows: Track the number of conntrack entries

Acked-by: Nithin Raju -Original Message- From: dev on behalf of Sairam Venugopal Date: Wednesday, June 29, 2016 at 5:08 PM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH v4 09/11] datapath-windows: Track the number of conntrack entries >Add a counter to track

Re: [ovs-dev] [PATCH v4 10/11] Windows: Add conntrack dump and flush support in userspace

Acked-by: Nithin Raju -Original Message- From: dev on behalf of Sairam Venugopal Date: Wednesday, June 29, 2016 at 5:08 PM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH v4 10/11] Windows: Add conntrack dump and flush support in userspace >Modify dpif-netlink.

Re: [ovs-dev] [PATCH v4 11/11] Windows: Ignore the dpif conversions for windows in netlink-conntrack.c

y be more appropriate :) Looks good otherwise. Acked-by: Nithin Raju ___ dev mailing list dev@openvswitch.org http://openvswitch.org/mailman/listinfo/dev

Re: [ovs-dev] [PATCH v5 04/11] datapath-windows: Add support for flushing conntrack entries

Acked-by: Nithin Raju -Original Message- From: dev on behalf of Sairam Venugopal Date: Friday, July 1, 2016 at 1:49 PM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH v5 04/11] datapath-windows: Add support for flushing conntrack entries >Flush out all conntra

Re: [ovs-dev] [PATCH v5 10/11] Windows: Add conntrack dump and flush support in userspace

Acked-by: Nithin Raju -Original Message- From: dev on behalf of Sairam Venugopal Date: Friday, July 1, 2016 at 1:49 PM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH v5 10/11] Windows: Add conntrack dump and flush support in userspace >Modify dpif-netlink.

Re: [ovs-dev] [PATCH v5 00/11] Windows: Add support for debugging conntrack from userspace

orted commands include flushing conntrack entries and >dumping them. > >v2: Addresed review comments from Paul Boca >v3: Added the ACKED-BY for the ones that were acked >v4: Addresed review comments from Nithin Raju >v5: Addresed review comments from Nithin Raju > >Sa

[ovs-dev] [PATCH] datapath-windows: remove unused 'ovsUserTimestampDelta'

Signed-off-by: Nithin Raju --- datapath-windows/ovsext/Flow.h | 1 - 1 file changed, 1 deletion(-) diff --git a/datapath-windows/ovsext/Flow.h b/datapath-windows/ovsext/Flow.h index 0744d30..23d252c 100644 --- a/datapath-windows/ovsext/Flow.h +++ b/datapath-windows/ovsext/Flow.h @@ -44,7 +44,6

Re: [ovs-dev] [PATCH] Windows: Add support for handling protocol (netlink family)

Thanks for doing this. Looks good but for a few cosmetic comments. Also, remember to flip the protocol for nf sockets in userspace to NETLINK_NETFILTER. IIRC, we use NETLINK_GENERIC. Acked-by: Nithin Raju -Original Message- From: dev on behalf of Sairam Venugopal Date: Thursday, July

Re: [ovs-dev] [PATCH] datapath-windows: Pause switch state

Alin, Can you pls. add KeMemoryBarrier(); after the ŒdataFlowState¹ was set to OvsSwitchPaused? What does a ŒNetEventFilterPreDetach¹ event mean in the context of a Hyper-V switch? Will the OvsExtPause() get invoked after that? If so, we¹ll have to update the ASSERT in OvsExtPause() which assumes

Re: [ovs-dev] [PATCH] Build Windows include runtime

Alin, Did you notice any change in the size of the binary? Eg. Ovs-vswitchd.exe? With or without your change, the binary was the same size. I am wondering if the default is /MT? Thanks, -- Nithin -Original Message- From: dev on behalf of Alin Serdean Date: Friday, February 26, 2016 at

[ovs-dev] [PATCH] datapath-windows: Fix a couple of bugs during port enumeration

While enumerating the ports on a switch, if adding one of the ports fails, we are not handling that error gracefully. Instead, we fail adding of all the ports. This patch rectifies that. Signed-off-by: Nithin Raju --- datapath-windows/ovsext/Vport.c | 41

Re: [ovs-dev] [PATCH] datapath-windows: Fix a couple of bugs during port enumeration

: Monday, March 7, 2016 at 3:40 PM To: Nithin Raju , "dev@openvswitch.org" Subject: RE: [ovs-dev] [PATCH] datapath-windows: Fix a couple of bugs during port enumeration >Unless I am reading wrong: OvsAddConfiguredSwitchPorts and >OvsInitConfiguredSwitchNics only fail if we could not

Re: [ovs-dev] [PATCH v6 1/6] datapath-windows: Percpu allocation support function

I had a few comments, but looks good otherwise. Acked-by: Nithin Raju >On 3/18/16, 7:58 AM, "Sorin Vinturis" >wrote: > >>Signed-off-by: Sorin Vinturis >>--- >> datapath-windows/ovsext/Util.c | 14 ++ >> datapath-windows/ovsext/Util

[ovs-dev] [PATCH] list.h: Define OVS_LIST_POISON statically

. Suggested-by: Yin Lin Signed-off-by: Nithin Raju --- lib/list.h | 12 1 file changed, 8 insertions(+), 4 deletions(-) diff --git a/lib/list.h b/lib/list.h index f9c9d85..24c57ba 100644 --- a/lib/list.h +++ b/lib/list.h @@ -24,10 +24,14 @@ #include "openvswitch/list.h"

Re: [ovs-dev] [PATCH v6 2/6] datapath-windows: Added recirculation support.

Pls. disregard this. I hit the ‘send’ too soon. Will send out the full list of comments. -Original Message- From: Nithin Raju Date: Friday, March 18, 2016 at 12:43 PM To: Sorin Vinturis , "dev@openvswitch.org" Subject: Re: [ovs-dev] [PATCH v6 2/6] datapath-windows: Added rec

Re: [ovs-dev] [PATCH v6 1/6] datapath-windows: Percpu allocation support function

Alternately, you could just do combination of KeQueryActiveProcessorCountEx(ALL_PROCESSOR_GROUPS) and make this code future proof. You’ll have to update the code in 2/6 patch also to use GetCurrentProcessorNumberEx(). -- Nitin -Original Message- From: dev on behalf of Nithin Raju

Re: [ovs-dev] [PATCH] list.h: Define OVS_LIST_POISON statically

I am fine with the suggestion and the patch you sent. Thanks Ben. -Original Message- From: Ben Pfaff Date: Friday, March 18, 2016 at 1:47 PM To: Nithin Raju Cc: "dev@openvswitch.org" , Yin Lin Subject: Re: [ovs-dev] [PATCH] list.h: Define OVS_LIST_POISON statically >O

Re: [ovs-dev] [PATCH v6 2/6] datapath-windows: Added recirculation support.

-Original Message- From: dev on behalf of Sorin Vinturis Date: Friday, March 18, 2016 at 7:58 AM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH v6 2/6] datapath-windows: Added recirculation support. >Recirculation support for the OVS extension. > >Tested using PING and iperf wi

[ovs-dev] [PATCH] list.h: Define OVS_LIST_POISON statically

. Suggested-by: Yin Lin Signed-off-by: Nithin Raju --- lib/list.h | 12 1 file changed, 8 insertions(+), 4 deletions(-) diff --git a/lib/list.h b/lib/list.h index f9c9d85..24c57ba 100644 --- a/lib/list.h +++ b/lib/list.h @@ -24,10 +24,14 @@ #include "openvswitch/list.h"

Re: [ovs-dev] [PATCH v6 2/6] datapath-windows: Added recirculation support.

Hi Sorin/Alin, I looked at the patch and had a few comments. Pls. find them inlined. In general, I was a little confused about how we use the recirc ID and the hash. We can discuss more during the IRC meeting. Thanks, -- Nithin -Original Message- From: dev on behalf of Sorin Vinturis

Re: [ovs-dev] [PATCH v6 3/6] datapath-windows: Allow recirculation feature probe.

Acked-by: Nithin Raju This patch is part of a series though and cannot be committed till the previous ones in the series get committed. -Original Message- From: dev on behalf of Sorin Vinturis Date: Friday, March 18, 2016 at 7:58 AM To: "dev@openvswitch.org" Subject

Re: [ovs-dev] [PATCH v6 4/6] datapath-windows: Support for OVS_ACTION_ATTR_HASH attribute

How is key->dphash being used after it gets stored as part of the key. Shouldn¹t it be used for flow lookup? Is it purely for reporting purposes? Thanks, -- Nithin -Original Message- From: dev on behalf of Sorin Vinturis Date: Friday, March 18, 2016 at 7:58 AM To: "dev@openvswitch.org"

Re: [ovs-dev] [PATCH v6 5/6] datapath-windows: Extract flow metadata in execute cmd handler.

Acked-by: Nithin Raju I was wondering if this should be part of the recirc patch itself. -Original Message- From: dev on behalf of Sorin Vinturis Date: Friday, March 18, 2016 at 7:58 AM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH v6 5/6] datapath-windows: Ex

Re: [ovs-dev] [PATCH v6 6/6] datapath-windows: Updated FAQ regarding Hyper-V recirculation.

Acked-by: Nithin Raju Dependent on previous patches in the series getting checked in. -Original Message- From: dev on behalf of Sorin Vinturis Date: Friday, March 18, 2016 at 7:58 AM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH v6 6/6] datapath-windows: Updated FAQ

Re: [ovs-dev] [PATCH v7 2/6] datapath-windows: Added recirculation support.

Thanks for incorporating the review comments. There are a few bugs, but we can fix them in subsequent patches I think. Acked-by: Nithin Raju -Original Message- From: dev on behalf of Sorin Vinturis Date: Thursday, March 24, 2016 at 9:53 AM To: "dev@openvswitch.org" Subject

Re: [ovs-dev] [PATCH v7 4/6] datapath-windows: Support for OVS_ACTION_ATTR_HASH attribute

Sorin, Thanks for the explanation about dpHash. We¹ll see how it gets used. Acked-by: Nithin Raju -Original Message- From: dev on behalf of Sorin Vinturis Date: Thursday, March 24, 2016 at 9:53 AM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH v7 4/6] datapath-window

Re: [ovs-dev] [PATCH] datapath-windows: Revert unrelated change to ovs-atomic.h

Acked-by: Nithin Raju -Original Message- From: dev on behalf of Sorin Vinturis Date: Friday, March 25, 2016 at 8:56 AM To: "dev@openvswitch.org" Subject: [ovs-dev] [PATCH] datapath-windows: Revert unrelated change to ovs-atomic.h >There was an unwanted change to

[ovs-dev] [PATCH] datapath-windows: fix alignment in Flow.c

Cosmetic changes. Acked-by: Nithin Raju --- datapath-windows/ovsext/Flow.c | 56 +- 1 file changed, 28 insertions(+), 28 deletions(-) diff --git a/datapath-windows/ovsext/Flow.c b/datapath-windows/ovsext/Flow.c index f15a3de..02c41b7 100644 --- a

[ovs-dev] [PATCH v2] datapath-windows: fix alignment in Flow.c

Cosmetic changes. Signed-off-by: Nithin Raju --- datapath-windows/ovsext/Flow.c | 56 +- 1 file changed, 28 insertions(+), 28 deletions(-) diff --git a/datapath-windows/ovsext/Flow.c b/datapath-windows/ovsext/Flow.c index f15a3de..02c41b7 100644 --- a

Re: [ovs-dev] [PATCH v2] datapath-windows: Update Recirculation to use the right parameters

Thanks for sending out the fix. I¹m ok with one of the fixes. The other one needs to be reworked to use one of the approaches consistently. Pls. see inlined comments. -- Nithin -Original Message- From: dev on behalf of Sairam Venugopal Date: Friday, March 25, 2016 at 11:06 AM To: "dev@

<    1   2   3   4   5   6   7   8   9   10   >