Hi sorry for late reply, didn't get this email.
On 03.12.2019 16:16, Ilya Maximets wrote:
> From: Ilya Maximets
> On 03.12.2019 14:45, Roi Dayan wrote:
> > From: Paul Blakey
> >
> > Each recirculation id will create a tc chain, and we translate
> > the recirculation action to a tc goto
On 12/10/2019 7:51 PM, Marcelo Ricardo Leitner wrote:
> On Mon, Dec 09, 2019 at 03:09:37PM +, Paul Blakey wrote:
>> On 12/3/2019 3:45 PM, Roi Dayan wrote:
>>> @@ -1428,38 +1419,35 @@ netdev_tc_flow_put(struct netdev *netdev, struct
>>> match *match,
>>>}
>>>}
>>>
>>>
On Sun, Dec 08, 2019 at 08:11:19AM +, Roni Bar Yanai wrote:
>
Hi Roni,
Thanks for your feedback!
> GTP-U header size is not constant, you *must* take into account the flags,
> mainly
> the sequence. The use of sequence in GTP-U is optional but some devices do
> use
> it. see from 3GPP
On 12/10/19 5:20 PM, David Ahern wrote:
> On 12/10/19 3:09 PM, Jason Baron wrote:
>> Hi David,
>>
>> The idea is that we try and queue new work to 'idle' threads in an
>> attempt to distribute a workload. Thus, once we find an 'idle' thread we
>> stop waking up other threads. While we are
On 12/10/19 4:00 PM, David Ahern wrote:
> [ adding Jason as author of the patch that added the epoll exclusive flag ]
>
> On 12/10/19 12:37 PM, Matteo Croce wrote:
>> On Tue, Dec 10, 2019 at 8:13 PM David Ahern wrote:
>>>
>>> Hi Matteo:
>>>
>>> On a hypervisor running a 4.14.91 kernel and OVS
On 12/10/19 3:09 PM, Jason Baron wrote:
> Hi David,
>
> The idea is that we try and queue new work to 'idle' threads in an
> attempt to distribute a workload. Thus, once we find an 'idle' thread we
> stop waking up other threads. While we are searching the wakeup list for
> idle threads, we do
On 12/10/19 2:20 PM, Matteo Croce wrote:
>
> Before this patch (which unfortunately is needed to avoid -EMFILE
> errors with many ports), how many sockets are awakened when an ARP is
> received?
>
on systems using 2.7.3 I see only a single handler thread awakened on
upcalls.
Yes, I saw the
On Tue, Dec 10, 2019 at 10:00 PM David Ahern wrote:
>
> [ adding Jason as author of the patch that added the epoll exclusive flag ]
>
> On 12/10/19 12:37 PM, Matteo Croce wrote:
> > On Tue, Dec 10, 2019 at 8:13 PM David Ahern wrote:
> >>
> >> Hi Matteo:
> >>
> >> On a hypervisor running a
[ adding Jason as author of the patch that added the epoll exclusive flag ]
On 12/10/19 12:37 PM, Matteo Croce wrote:
> On Tue, Dec 10, 2019 at 8:13 PM David Ahern wrote:
>>
>> Hi Matteo:
>>
>> On a hypervisor running a 4.14.91 kernel and OVS 2.11 I am seeing a
>> thundering herd wake up
On 08.12.2019 14:23, Eli Britstein wrote:
> Signed-off-by: Eli Britstein
> Reviewed-by: Oz Shlomo
> ---
> Documentation/howto/dpdk.rst | 2 ++
> NEWS | 4 ++--
> lib/netdev-dpdk.c| 14 +
> lib/netdev-offload-dpdk.c| 48
>
On Tue, Dec 10, 2019 at 8:41 PM David Ahern wrote:
>
> On 12/10/19 12:37 PM, Matteo Croce wrote:
> > On Tue, Dec 10, 2019 at 8:13 PM David Ahern wrote:
> >>
> >> Hi Matteo:
> >>
> >> On a hypervisor running a 4.14.91 kernel and OVS 2.11 I am seeing a
> >> thundering herd wake up problem. Every
Liebste Geliebte,
Ich bin David Raymon. Ein portugiesischer Staatsbürger. Ich habe gesucht und
Ihre E-Mail gesehen. Also habe ich beschlossen, Ihnen zu schreiben, ob Ihre
E-Mail echt ist. Bei mir wurde Speiseröhrenkrebs diagnostiziert. Es hat alle
Formen der medizinischen Behandlung
On 12/10/19 12:37 PM, Matteo Croce wrote:
> On Tue, Dec 10, 2019 at 8:13 PM David Ahern wrote:
>>
>> Hi Matteo:
>>
>> On a hypervisor running a 4.14.91 kernel and OVS 2.11 I am seeing a
>> thundering herd wake up problem. Every packet punted to userspace wakes
>> up every one of the handler
On Tue, Dec 10, 2019 at 8:13 PM David Ahern wrote:
>
> Hi Matteo:
>
> On a hypervisor running a 4.14.91 kernel and OVS 2.11 I am seeing a
> thundering herd wake up problem. Every packet punted to userspace wakes
> up every one of the handler threads. On a box with 96 cpus, there are 71
> handler
On 08.12.2019 14:23, Eli Britstein wrote:
> Signed-off-by: Eli Britstein
> Reviewed-by: Oz Shlomo
> ---
> Documentation/howto/dpdk.rst | 1 +
> NEWS | 3 +-
> lib/netdev-dpdk.c| 15 ++
> lib/netdev-offload-dpdk.c| 107
>
Hi Matteo:
On a hypervisor running a 4.14.91 kernel and OVS 2.11 I am seeing a
thundering herd wake up problem. Every packet punted to userspace wakes
up every one of the handler threads. On a box with 96 cpus, there are 71
handler threads which means 71 process wakeups for every packet punted.
On 08.12.2019 14:23, Eli Britstein wrote:
> Signed-off-by: Eli Britstein
> Reviewed-by: Oz Shlomo
> ---
> Documentation/howto/dpdk.rst | 1 +
> NEWS | 2 +-
> lib/netdev-dpdk.c| 2 ++
> lib/netdev-offload-dpdk.c| 4 +---
> 4 files changed, 5
On 08.12.2019 14:22, Eli Britstein wrote:
> Support offload of output action, also configuring count action for
> allowing query statistics of HW offloaded flows.
>
> Signed-off-by: Eli Britstein
> Reviewed-by: Oz Shlomo
> ---
> Documentation/howto/dpdk.rst | 17 ---
> NEWS
On 08.12.2019 10:38, Eli Britstein wrote:
>
> On 12/5/2019 6:46 PM, Sriharsha Basavapatna wrote:
>> On Wed, Dec 4, 2019 at 8:55 PM Eli Britstein wrote:
>>>
>>> On 12/3/2019 5:19 PM, Sriharsha Basavapatna wrote:
On Wed, Nov 20, 2019 at 9:07 PM Eli Britstein wrote:
> Signed-off-by: Eli
On 08.12.2019 14:22, Eli Britstein wrote:
> From: Ophir Munk
>
> Use netdev flow get API in order to update the statistics of fully
> offloaded flows.
>
> Co-authored-by: Eli Britstein
> Signed-off-by: Ophir Munk
> Reviewed-by: Oz Shlomo
> Signed-off-by: Eli Britstein
> ---
>
On Mon, Dec 09, 2019 at 03:09:37PM +, Paul Blakey wrote:
>
> On 12/3/2019 3:45 PM, Roi Dayan wrote:
> > @@ -1428,38 +1419,35 @@ netdev_tc_flow_put(struct netdev *netdev, struct
> > match *match,
> > }
> > }
> >
> > -block_id = get_block_id_from_netdev(netdev);
>
>
On 08.12.2019 14:22, Eli Britstein wrote:
> Implement the flow get method for DPDK, to get the statistics of the
> provided ufid, towards reading statistics of fully offloaded flows.
>
> Signed-off-by: Eli Britstein
> Reviewed-by: Oz Shlomo
> ---
> lib/netdev-offload-dpdk.c | 40
On Tue, Dec 10, 2019 at 7:56 PM Eli Britstein wrote:
>
>
> On 12/10/2019 2:53 PM, Sathya Perla wrote:
> > On Tue, Dec 10, 2019 at 5:25 PM Eli Britstein wrote:
> >>
> >> On 12/10/2019 12:09 PM, Sathya Perla wrote:
> >>> On Tue, Dec 10, 2019 at 1:21 PM Eli Britstein wrote:
> On 12/10/2019
On 10.12.2019 15:44, Eli Britstein wrote:
>
> On 12/10/2019 4:38 PM, Ilya Maximets wrote:
>> On 08.12.2019 14:22, Eli Britstein wrote:
>>> Add debug prints when creating and destroying rte flows, with all the
>>> flow details (attributes, patterns, actions).
>>>
>>> Signed-off-by: Eli Britstein
On 12/10/2019 12:17 AM, Martin Varghese wrote:
From: Martin Varghese
The existing PUSH MPLS & POP MPLS actions inserts & removes MPLS header
between ethernet header and the IP header. Though this behaviour is fine
for L3 VPN where an IP packet is encapsulated inside a MPLS tunnel, it
does
On 12/10/2019 3:01 PM, Ilya Maximets wrote:
> We changed datapath port lookup to netdev-offload API usage, but
> forgot that port mutex was there not only to protect datapath
> port hash map. It was there also as a workaround solution for
> complete unsafety of netdev-offload-dpdk functions.
>
On 12/10/2019 5:03 PM, Ilya Maximets wrote:
> On 10.12.2019 15:56, Eli Britstein wrote:
>> On 12/10/2019 4:53 PM, Ilya Maximets wrote:
>>> On 08.12.2019 14:22, Eli Britstein wrote:
Flow deletion and dumping for statistics collection are called from
different threads. As a pre-step
On 10.12.2019 15:56, Eli Britstein wrote:
>
> On 12/10/2019 4:53 PM, Ilya Maximets wrote:
>> On 08.12.2019 14:22, Eli Britstein wrote:
>>> Flow deletion and dumping for statistics collection are called from
>>> different threads. As a pre-step towards collecting HW statistics,
>>> protect the
On 10.12.2019 15:53, Eli Britstein wrote:
>
> On 12/10/2019 4:50 PM, Ilya Maximets wrote:
>> On 08.12.2019 14:22, Eli Britstein wrote:
>>> Introduce a rte flow query function as a pre-step towards reading HW
>>> statistics of fully offloaded flows.
>>>
>>> Signed-off-by: Eli Britstein
>>>
On 12/10/2019 4:53 PM, Ilya Maximets wrote:
> On 08.12.2019 14:22, Eli Britstein wrote:
>> Flow deletion and dumping for statistics collection are called from
>> different threads. As a pre-step towards collecting HW statistics,
>> protect the UFID map by mutex to make it thread safe.
>>
>>
On 08.12.2019 14:22, Eli Britstein wrote:
> Flow deletion and dumping for statistics collection are called from
> different threads. As a pre-step towards collecting HW statistics,
> protect the UFID map by mutex to make it thread safe.
>
> Signed-off-by: Eli Britstein
> ---
>
On 12/10/2019 4:50 PM, Ilya Maximets wrote:
> On 08.12.2019 14:22, Eli Britstein wrote:
>> Introduce a rte flow query function as a pre-step towards reading HW
>> statistics of fully offloaded flows.
>>
>> Signed-off-by: Eli Britstein
>> Reviewed-by: Oz Shlomo
>> ---
>> lib/netdev-dpdk.c |
On 08.12.2019 14:22, Eli Britstein wrote:
> Introduce a rte flow query function as a pre-step towards reading HW
> statistics of fully offloaded flows.
>
> Signed-off-by: Eli Britstein
> Reviewed-by: Oz Shlomo
> ---
> lib/netdev-dpdk.c | 25 +
> lib/netdev-dpdk.h | 6
On 12/10/2019 4:22 PM, Ilya Maximets wrote:
> On 08.12.2019 14:22, Eli Britstein wrote:
>> Fixes: e8a2b5bf92bb ("netdev-dpdk: implement flow offload with rte flow")
>> Signed-off-by: Eli Britstein
>> Reviewed-by: Oz Shlomo
>> ---
>> lib/netdev-offload-dpdk.c | 2 +-
>> 1 file changed, 1
On 12/10/2019 4:38 PM, Ilya Maximets wrote:
> On 08.12.2019 14:22, Eli Britstein wrote:
>> Add debug prints when creating and destroying rte flows, with all the
>> flow details (attributes, patterns, actions).
>>
>> Signed-off-by: Eli Britstein
>> Reviewed-by: Oz Shlomo
>> ---
>>
On 08.12.2019 14:22, Eli Britstein wrote:
> Add debug prints when creating and destroying rte flows, with all the
> flow details (attributes, patterns, actions).
>
> Signed-off-by: Eli Britstein
> Reviewed-by: Oz Shlomo
> ---
> lib/netdev-dpdk.c | 260
>
On 12/10/2019 2:53 PM, Sathya Perla wrote:
> On Tue, Dec 10, 2019 at 5:25 PM Eli Britstein wrote:
>>
>> On 12/10/2019 12:09 PM, Sathya Perla wrote:
>>> On Tue, Dec 10, 2019 at 1:21 PM Eli Britstein wrote:
On 12/10/2019 8:52 AM, Sathya Perla wrote:
> On Sun, Dec 8, 2019 at 6:53 PM Eli
On 08.12.2019 14:22, Eli Britstein wrote:
> Fixes: e8a2b5bf92bb ("netdev-dpdk: implement flow offload with rte flow")
> Signed-off-by: Eli Britstein
> Reviewed-by: Oz Shlomo
> ---
> lib/netdev-offload-dpdk.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git
On 12/10/2019 4:09 PM, Ilya Maximets wrote:
> On 10.12.2019 15:06, Eli Britstein wrote:
>> On 12/10/2019 3:43 PM, Ilya Maximets wrote:
>>> On 08.12.2019 14:22, Eli Britstein wrote:
Action item data structures are pointed by rte_flow_action items.
Refactor the code to free the data
On 12/8/19 4:23 PM, Eli Britstein wrote:
> Signed-off-by: Eli Britstein
> Reviewed-by: Oz Shlomo
> ---
> Documentation/howto/dpdk.rst | 1 +
> NEWS | 3 +-
> lib/netdev-dpdk.c| 15 ++
> lib/netdev-offload-dpdk.c| 107
>
On 12/10/19 5:14 PM, Eli Britstein wrote:
>
> On 12/10/2019 4:12 PM, Andrew Rybchenko wrote:
>> On 12/8/19 4:23 PM, Eli Britstein wrote:
>>> Signed-off-by: Eli Britstein
>>> Reviewed-by: Oz Shlomo
>>> ---
>>> Documentation/howto/dpdk.rst | 1 +
>>> NEWS | 3 +-
>>>
On 12/10/2019 4:12 PM, Andrew Rybchenko wrote:
> On 12/8/19 4:23 PM, Eli Britstein wrote:
>> Signed-off-by: Eli Britstein
>> Reviewed-by: Oz Shlomo
>> ---
>> Documentation/howto/dpdk.rst | 1 +
>> NEWS | 3 +-
>> lib/netdev-dpdk.c| 15 ++
>>
On 10.12.2019 15:06, Eli Britstein wrote:
>
> On 12/10/2019 3:43 PM, Ilya Maximets wrote:
>> On 08.12.2019 14:22, Eli Britstein wrote:
>>> Action item data structures are pointed by rte_flow_action items.
>>> Refactor the code to free the data structures when freeing the
>>> rte_flow_action
On 12/10/2019 3:43 PM, Ilya Maximets wrote:
> On 08.12.2019 14:22, Eli Britstein wrote:
>> Action item data structures are pointed by rte_flow_action items.
>> Refactor the code to free the data structures when freeing the
>> rte_flow_action items, allowing simpler future actions simpler to add
On Tue, Dec 10, 2019 at 5:25 PM Eli Britstein wrote:
>
>
> On 12/10/2019 12:09 PM, Sathya Perla wrote:
> > On Tue, Dec 10, 2019 at 1:21 PM Eli Britstein wrote:
> >>
> >> On 12/10/2019 8:52 AM, Sathya Perla wrote:
> >>> On Sun, Dec 8, 2019 at 6:53 PM Eli Britstein wrote:
> >>> ...
> +static
On 08.12.2019 14:22, Eli Britstein wrote:
> Action item data structures are pointed by rte_flow_action items.
> Refactor the code to free the data structures when freeing the
> rte_flow_action items, allowing simpler future actions simpler to add to
> the code.
>
> Signed-off-by: Eli Britstein
>
On 12/10/2019 3:27 PM, Ilya Maximets wrote:
> On 08.12.2019 14:22, Eli Britstein wrote:
>> Refactor the flow patterns code to a helper function for better
>> readability and towards supporting more matches.
>>
>> Signed-off-by: Eli Britstein
>> Reviewed-by: Oz Shlomo
>> ---
> Although I'm not
On 08.12.2019 14:22, Eli Britstein wrote:
> Refactor the flow patterns code to a helper function for better
> readability and towards supporting more matches.
>
> Signed-off-by: Eli Britstein
> Reviewed-by: Oz Shlomo
> ---
Although I'm not quite sure that we need this change, a few comments
We changed datapath port lookup to netdev-offload API usage, but
forgot that port mutex was there not only to protect datapath
port hash map. It was there also as a workaround solution for
complete unsafety of netdev-offload-dpdk functions.
Turning it back to fix the behaviour and adding a
On 12/10/2019 12:09 PM, Sathya Perla wrote:
> On Tue, Dec 10, 2019 at 1:21 PM Eli Britstein wrote:
>>
>> On 12/10/2019 8:52 AM, Sathya Perla wrote:
>>> On Sun, Dec 8, 2019 at 6:53 PM Eli Britstein wrote:
>>> ...
+static int
+parse_clone_actions(struct netdev *netdev,
+
On Tue, Dec 10, 2019 at 1:21 PM Eli Britstein wrote:
>
>
> On 12/10/2019 8:52 AM, Sathya Perla wrote:
> > On Sun, Dec 8, 2019 at 6:53 PM Eli Britstein wrote:
> > ...
> >> +static int
> >> +parse_clone_actions(struct netdev *netdev,
> >> +struct flow_actions *actions,
> >> +
Most DPDK components make the assumption that rte_lcore_id() returns
a valid lcore_id in [0..RTE_MAX_LCORE[ range (with the exception of
the LCORE_ID_ANY special value).
OVS does not currently check which value is set in
RTE_PER_LCORE(_lcore_id) which exposes us to potential crashes on DPDK
side.
On 12/8/2019 3:23 PM, Eli Britstein wrote:
> Signed-off-by: Eli Britstein
> Reviewed-by: Oz Shlomo
> ---
> Documentation/howto/dpdk.rst | 1 +
> NEWS | 3 +-
> lib/netdev-dpdk.c| 15 ++
> lib/netdev-offload-dpdk.c| 107
>
From: Martin Varghese
The existing PUSH MPLS & POP MPLS actions inserts & removes MPLS header
between ethernet header and the IP header. Though this behaviour is fine
for L3 VPN where an IP packet is encapsulated inside a MPLS tunnel, it
does not suffice the L2 VPN requirements. In L2 VPN the
54 matches
Mail list logo