Title: Email Newsletter Templates by DeliverySuccess.com
Mr. Park Kyu-Ho
Chief Financial Officer - (KEP)
167 Samseong-dong
Gangnam-gu
Seoul, 135-791
Korea, Republic of (South)
Letter of
With the foundation laid in previous commits, this commit
removes the 'dispatcher' thread by allowing 'handler'
threads to read upcalls directly from dpif.
This commit significantly simplifies the flow miss handling
code and brings slight improvement to flow setup rate.
Note:
- the flow setup rat
Signed-off-by: Alex Wang
---
V2 -> V3:
- use OVS_DP_ATTR_USER_FEATURES to inform datapath about the type of
OVS_VPORT_ATTR_UPCALL_PID attribute.
- close epoll fd when destroying all channels.
PATCH -> V2:
- rebase.
major changes since RFC:
- free the 'upcall_pids' pointer in dpif_linux_refres
In order to allow handlers directly read upcalls from datapath,
we need to support per-handler netlink socket for each vport in
datapath. This commit makes this happen. Also, it is guaranteed
backward and forward compatibility with previous branch.
Signed-off-by: Alex Wang
---
V2 -> V3:
- use
This commit implements the API functions to allow multiple handler
threads read upcall.
Also, this commit removes the handling priority of DPIF_UC_MISS
over DPIF_UC_ACTION. So, both misses will be put to the same
queue. The decision is based on the fact that a lot has changed
since the age when
I have received many good reviews. I think it is good to apply the
suggestions and rebase the whole series. So everyone can have a good
new start point.
V2 -> V3:
- Rebase the dpif-netdev, dpif-linux based on the change of API.
- Refine the datapath code based on the reviews.
PATCH -> V2:
- Ma
This commit changes the API in 'dpif-provider.h' to allow multiple
handler threads call dpif_recv() simultaneously.
Signed-off-by: Alex Wang
---
V2 -> V3:
- detach the channel refreshing from recv_set().
Note:
the recv_set() still calls the refresh_channels(), the first time receiving
packe
Applied. Thanks.
--Justin
On Mar 5, 2014, at 4:56 PM, Joe Stringer wrote:
> This patch adds a new configuration option, "max-idle" to the
> Open_vSwitch "other-config" column. This sets how long datapath flows
> are cached in the datapath before revalidators expire them.
>
> Signed-off-by: J
Thanks for the review Thomas~ ;D
Please see my reply inline:
It's nice to see that you preserve backwards compatibility but why
> rename the attribute? It doesn't really serve a purpose except that you
> break the API.
>
> Since OVS_VPORT_ATTR_UPCALL_PID was a single u32 it is already forward
> c
On Mar 7, 2014, at 12:51 PM, Pravin Shelar wrote:
> On Fri, Mar 7, 2014 at 11:52 AM, Pritesh Kothari (pritkoth)
> wrote:
>> Hi Pravin,
>>
if (type > OVS_TUNNEL_KEY_ATTR_MAX) {
@@ -388,11 +390,16 @@ static int ipv4_tun_from_nlattr(const struct nlattr
*attr,
>
On Fri, Mar 7, 2014 at 11:52 AM, Pritesh Kothari (pritkoth)
wrote:
> Hi Pravin,
>
>>>
>>>if (type > OVS_TUNNEL_KEY_ATTR_MAX) {
>>> @@ -388,11 +390,16 @@ static int ipv4_tun_from_nlattr(const struct nlattr
>>> *attr,
>>>case OVS_TUNNEL_KEY_ATTR_CSUM:
>>>
Hi Pravin,
>>
>>if (type > OVS_TUNNEL_KEY_ATTR_MAX) {
>> @@ -388,11 +390,16 @@ static int ipv4_tun_from_nlattr(const struct nlattr
>> *attr,
>>case OVS_TUNNEL_KEY_ATTR_CSUM:
>>tun_flags |= TUNNEL_CSUM;
>>break;
>> +
On Fri, Mar 7, 2014 at 10:05 AM, Thomas Graf wrote:
> On 03/07/2014 06:19 PM, Pravin Shelar wrote:
>>
>> On Fri, Mar 7, 2014 at 7:58 AM, Thomas Graf wrote:
>>>
>>> On 03/07/2014 05:46 AM, Pravin Shelar wrote:
But I found bug in datapath user-space queue code. I am not sure how
On Fri, Mar 7, 2014 at 9:59 AM, Thomas Graf wrote:
> On 03/07/2014 06:28 PM, Pravin Shelar wrote:
>>
>> Problem is mapping SKBTX_DEV_ZEROCOPY pages to userspace. skb_zerocopy
>> is not doing that.
>>
>> Unless I missing something, Current netlink code can not handle
>> skb-frags with zero copy. So
On 03/07/2014 06:19 PM, Pravin Shelar wrote:
On Fri, Mar 7, 2014 at 7:58 AM, Thomas Graf wrote:
On 03/07/2014 05:46 AM, Pravin Shelar wrote:
But I found bug in datapath user-space queue code. I am not sure how
this can work with skb fragments and MMAP-netlink socket.
Here is what happens, OVS
On 03/07/2014 06:28 PM, Pravin Shelar wrote:
Problem is mapping SKBTX_DEV_ZEROCOPY pages to userspace. skb_zerocopy
is not doing that.
Unless I missing something, Current netlink code can not handle
skb-frags with zero copy. So we have to copy skb anyways and no need
to orphan-frags here.
If you
On Tue, Feb 25, 2014 at 3:44 PM, Pritesh Kothari
wrote:
> This patch adds support for Network Service Headers (nsh) over VXLAN
> as mentioned in [1]. Here changes are made to datapath to add nsh
> headers whenever a vxlan port with destination port as 6633 (which is
> IANA allocated port for nsh o
On Fri, Mar 7, 2014 at 4:29 AM, Zoltan Kiss wrote:
> On 07/03/14 04:46, Pravin Shelar wrote:
>>
>> On Thu, Mar 6, 2014 at 9:09 AM, Zoltan Kiss
>> wrote:
>>>
>>> Do you have any feedback on this? I'm also adding KVM list as they might
>>> be
>>> interested in this.
>>>
>>> Zoli
>>>
>>>
>>> On 28/0
On Fri, Mar 7, 2014 at 8:23 AM, Thomas Graf wrote:
> On 02/28/2014 08:16 PM, Zoltan Kiss wrote:
>>
>> The kernel datapath now switched to zerocopy Netlink messages, but that
>> also
>> means that the pages on frags array are sent straight to userspace. If
>> those
>> pages came outside the kernel,
On Fri, Mar 7, 2014 at 7:58 AM, Thomas Graf wrote:
> On 03/07/2014 05:46 AM, Pravin Shelar wrote:
>>
>> But I found bug in datapath user-space queue code. I am not sure how
>> this can work with skb fragments and MMAP-netlink socket.
>> Here is what happens, OVS allocates netlink skb and adds frag
On 02/28/2014 08:16 PM, Zoltan Kiss wrote:
The kernel datapath now switched to zerocopy Netlink messages, but that also
means that the pages on frags array are sent straight to userspace. If those
pages came outside the kernel, we have to swap them out with local copies.
Signed-off-by: Zoltan Ki
On 03/07/2014 05:46 AM, Pravin Shelar wrote:
But I found bug in datapath user-space queue code. I am not sure how
this can work with skb fragments and MMAP-netlink socket.
Here is what happens, OVS allocates netlink skb and adds fragments to
skb using skb_zero_copy(), then calls genlmsg_unicast()
On 07/03/14 04:46, Pravin Shelar wrote:
On Thu, Mar 6, 2014 at 9:09 AM, Zoltan Kiss wrote:
Do you have any feedback on this? I'm also adding KVM list as they might be
interested in this.
Zoli
On 28/02/14 19:16, Zoltan Kiss wrote:
The kernel datapath now switched to zerocopy Netlink message
On Mon, Mar 3, 2014 at 11:52 PM, Ben Pfaff wrote:
> On Mon, Mar 03, 2014 at 11:51:48PM +0200, Alexandru Copot wrote:
>> On Mon, Mar 3, 2014 at 11:00 PM, Ben Pfaff wrote:
>> > The above data structure isn't needed; OVS already handles experimenter
>> > error codes through infrastructure in ofp-err
24 matches
Mail list logo