DPDK rings must have a power-of-two size.
Signed-off-by: Daniele Di Proietto
---
lib/netdev-dpdk.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c
index b45e367..c1b644a 100644
--- a/lib/netdev-dpdk.c
+++ b/lib/netdev-dpdk.c
On 7/30/14, 8:29 AM, "Ben Pfaff" wrote:
>On Tue, Jul 29, 2014 at 04:21:32PM -0700, Daniele Di Proietto wrote:
>> DPDK rings must have a power-of-two size.
>>
>> Signed-off-by: Daniele Di Proietto
>> ---
>> lib/netdev-dpdk.c | 5 +++--
>&
DPDK rings must have a power-of-two size.
Signed-off-by: Daniele Di Proietto
---
lib/netdev-dpdk.c | 6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c
index b45e367..a779833 100644
--- a/lib/netdev-dpdk.c
+++ b/lib/netdev-dpdk.c
Thanks for the review and for the fixes!
Daniele
On 8/4/14, 2:31 PM, "Ben Pfaff" wrote:
>On Thu, Jul 17, 2014 at 05:26:00PM -0700, Daniele Di Proietto wrote:
>> This commit introduces multiple appctl commands (dpctl/*)
>>
>> They are needed to interact with u
There's no reason for the local variable 'old_count' to be atomic. In fact, if
it is atomic it triggers a GCC4.9 warning (Wunused-value)
The global variables 'a' and 'paux' could be static, according to sparse.
Signed-off-by: Daniele Di Proietto
---
Ther
,
>
> Jarno
>
>On Aug 6, 2014, at 10:35 AM, Daniele Di Proietto
>wrote:
>
>> There's no reason for the local variable 'old_count' to be atomic. In
>>fact, if
>> it is atomic it triggers a GCC4.9 warning (Wunused-value)
>> The global variab
On 8/6/14, 3:30 PM, "Pravin Shelar" wrote:
>On Wed, Jul 30, 2014 at 8:51 AM, Daniele Di Proietto
> wrote:
>> DPDK rings must have a power-of-two size.
>>
>> Signed-off-by: Daniele Di Proietto
>> ---
>> lib/netdev-dpdk.c | 6 --
&g
This commit introduces the BUILD_MESSAGE() macro. It uses _Pragma("message"),
with compilers that support that, to output a warning-like compile-time message
without blocking the compilation.
Used by next commit.
Signed-off-by: Daniele Di Proietto
---
configure.ac | 1 +
lib/
Signed-off-by: Daniele Di Proietto
---
lib/flow.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/lib/flow.c b/lib/flow.c
index 2e5ca0a..78b132e 100644
--- a/lib/flow.c
+++ b/lib/flow.c
@@ -123,6 +123,9 @@ struct mf_ctx {
#if (FLOW_WC_SEQ != 27)
#define MINIFLOW_ASSERT(X) ovs_assert(X
Signed-off-by: Daniele Di Proietto
---
lib/flow.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/lib/flow.c b/lib/flow.c
index 5e04015..2e5ca0a 100644
--- a/lib/flow.c
+++ b/lib/flow.c
@@ -118,10 +118,10 @@ struct mf_ctx {
/* miniflow_push_* macros allow filling in a
testing of miniflow_extract() with run-time assertions.
Daniele Di Proietto (3):
lib/flow: Update FLOW_WC_SEQ to skip assertions on miniflow_extract()
Add BUILD_MESSAGE() macro
lib/flow: Use BUILD_MESSAGE() to warn if BUILD_SEQ is not updated
configure.ac | 1 +
lib/compiler.h| 10
Hi Pravin,
couple of comments:
- I would remove also the check for match->mask in update_range__().
- I think braces are not necessary for if/else statements in the macros.
Of course I would wait for review by Jesse or Andy, but other than that,
LGTM.
Acked-by: Daniele Di Proietto
On 8/7
halme wrote:
>
>> Acked-by: Jarno Rajahalme
>>
>>> On Aug 7, 2014, at 6:15 PM, Daniele Di Proietto
>>> wrote:
>>>
>>> Signed-off-by: Daniele Di Proietto
>>> ---
>>> lib/flow.c | 4 ++--
>>> 1 file changed, 2 insertions(
-2).
>
>The main issue is that this logic does not handle genuine multiple
>level of nested attributes. For example, encap within tunnel. OVS
>does not support ittoday but Geneve does allow for outer header to
>have a vlan tag.
>
>If the issue is to handle variable len
With this commit we move our DPDK support to 1.7.0.
DPDK binaries (starting with dpdk 1.7.0) should be linked with --whole-archive
to include pmd drivers
Signed-off-by: Daniele Di Proietto
---
INSTALL.DPDK | 12 ++--
acinclude.m4 | 13 -
lib
only if rte_eth_tx_burst
returns 0.
Signed-off-by: Daniele Di Proietto
---
lib/netdev-dpdk.c | 19 +--
1 file changed, 17 insertions(+), 2 deletions(-)
diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c
index f2202b4..97206e6 100644
--- a/lib/netdev-dpdk.c
+++ b/lib/netdev-d
Hi Andy,
Two comments inline
Other than those, LGTM, but review by others might be useful :)
Acked-by: Daniele Di Proietto
On 8/12/14, 1:45 PM, "Andy Zhou" wrote:
>-1 was used to indicate both nested and variable length attributes.
>This patch introduces OVS_KE
On 8/12/14, 5:40 PM, "Pravin Shelar" wrote:
>On Tue, Aug 12, 2014 at 10:43 AM, Daniele Di Proietto
> wrote:
>> rte_eth_tx_burst() _should_ transmit every packet that it is passed
>>unless the
>> queue is full. Nontheless some implementation of rte_eth_tx_
On 8/12/14, 5:41 PM, "Pravin Shelar" wrote:
>On Tue, Jul 22, 2014 at 5:06 PM, Daniele Di Proietto
> wrote:
>> struct dp_netdev_flow used to have a reference counter. It has been
>>replaced by
>> RCU. Unfortunately RCU is not enough if we plan to hold a referenc
On 8/12/14, 5:42 PM, "Pravin Shelar" wrote:
>On Tue, Jul 22, 2014 at 5:06 PM, Daniele Di Proietto
> wrote:
>> ofpbuf_rss() is introduced to get the RSS hash from the NIC. It works
>>only for
>> packets received by DPDK (otherwise it returns 0, which tells the
On 8/12/14, 5:39 PM, "Pravin Shelar" wrote:
>On Tue, Aug 12, 2014 at 10:43 AM, Daniele Di Proietto
> wrote:
>> With this commit we move our DPDK support to 1.7.0.
>> DPDK binaries (starting with dpdk 1.7.0) should be linked with
>>--whole-archive
>&
On 8/13/14, 3:28 PM, "Pravin Shelar" wrote:
>On Wed, Aug 13, 2014 at 8:49 AM, Daniele Di Proietto
> wrote:
>>
>>
>> On 8/12/14, 5:42 PM, "Pravin Shelar" wrote:
>>
>>>On Tue, Jul 22, 2014 at 5:06 PM, Daniele Di Proietto
>>> wro
cache processing
- Use variable length arrays to reduce cache footprint
- Rebased
Daniele Di Proietto (3):
ofpbuf: Add ofpbuf_rss()
dpif-netdev: Add dp_netdev_packet_rss()
dpif-netdev: Exact match cache
lib/dpif-netdev.c | 451 +-
l
Di Proietto
---
lib/netdev-dpdk.c | 3 ++-
lib/ofpbuf.h | 14 ++
2 files changed, 16 insertions(+), 1 deletion(-)
diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c
index a8f041b..23cf410 100644
--- a/lib/netdev-dpdk.c
+++ b/lib/netdev-dpdk.c
@@ -96,7 +96,8 @@ static const
Since lookups in the classifier can be pretty expensive, we introduce this
(thread local) cache which simply compares the miniflows of the packets
Signed-off-by: Daniele Di Proietto
---
lib/dpif-netdev.c | 422 +-
1 file changed, 357
This new function uses ofpbuf_rss(), or hashes the packet 5-tuple, to get the
hash of a packet. Used by OVS_ACTION_ATTR_HASH and future commits.
Signed-off-by: Daniele Di Proietto
---
lib/dpif-netdev.c | 29 ++---
1 file changed, 22 insertions(+), 7 deletions(-)
diff
On 8/15/14, 2:13 PM, "Gurucharan Shetty" wrote:
>On Tue, Aug 12, 2014 at 5:39 PM, Pravin Shelar wrote:
>> On Tue, Aug 12, 2014 at 10:43 AM, Daniele Di Proietto
>> wrote:
>>> With this commit we move our DPDK support to 1.7.0.
>>> DPDK binaries (star
Pfaff
Signed-off-by: Daniele Di Proietto
---
Guru, does this properly address the issue?
We could also put the extra line in vswitchd/automake.mk inside an "if
DPDK_NETDEV" block. What do you guys think?
---
acinclude.m4 | 4 ++--
vswitchd/automake.mk | 1 +
2 files changed, 3
I just sent a patch to the list. Please, let me know if this solves all
the issues.
Thanks,
Daniele
On 8/18/14, 10:31 AM, "Ben Pfaff" wrote:
>On Mon, Aug 18, 2014 at 10:28:44AM -0700, Gurucharan Shetty wrote:
>> On Mon, Aug 18, 2014 at 9:51 AM, Ben Pfaff wrote:
>> > On Mon, Aug 18, 2014 at 07
fat_rwlock_tryrdlock() used to call fat_rwlock_get_slot__() which could block
in the "slow path" case. This commit adds fat_rwlock_try_get_slot__() which
does not block, even in the "slow path" case.
Signed-off-by: Daniele Di Proietto
---
l
On 8/18/14, 11:47 AM, "Ben Pfaff" wrote:
>On Mon, Aug 18, 2014 at 11:26:22AM -0700, Daniele Di Proietto wrote:
>> fat_rwlock_tryrdlock() used to call fat_rwlock_get_slot__() which could
>>block
>> in the "slow path" case. This commit adds fat_rwlock_t
upcall handler (e.g. if it is created with dpctl commands), dp_netdev_input()
hangs if it does not find a packet's flow in the classifier.
Signed-off-by: Daniele Di Proietto
---
lib/fat-rwlock.c | 33 -
1 file changed, 32 insertions(+), 1 deletion(-)
diff -
Sorry about those,
I will send an updated version soon.
Thanks,
Daniele
On 8/18/14, 9:45 AM, "Ben Pfaff" wrote:
>On Fri, Aug 15, 2014 at 12:43:35PM -0700, Daniele Di Proietto wrote:
>> ofpbuf_rss() is introduced to get the RSS hash from the NIC. It works
>>only f
Thanks!
On 8/18/14, 1:23 PM, "Ben Pfaff" wrote:
>On Mon, Aug 18, 2014 at 12:57:42PM -0700, Daniele Di Proietto wrote:
>> fat_rwlock_tryrdlock() used to call fat_rwlock_get_slot__() which could
>>block
>> in the "slow path" case. This commit adds fat_rw
This new function uses ofpbuf_rss(), or hashes the packet 5-tuple, to get the
hash of a packet. Used by OVS_ACTION_ATTR_HASH and future commits.
Signed-off-by: Daniele Di Proietto
---
lib/dpif-netdev.c | 29 ++---
1 file changed, 22 insertions(+), 7 deletions(-)
diff
Di Proietto
---
lib/netdev-dpdk.c | 3 ++-
lib/ofpbuf.h | 14 ++
2 files changed, 16 insertions(+), 1 deletion(-)
diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c
index a8f041b..23cf410 100644
--- a/lib/netdev-dpdk.c
+++ b/lib/netdev-dpdk.c
@@ -96,7 +96,8 @@ static const
Since lookups in the classifier can be pretty expensive, we introduce this
(thread local) cache which simply compares the miniflows of the packets
Signed-off-by: Daniele Di Proietto
---
lib/dpif-netdev.c | 422 +-
1 file changed, 357
Split dp_netdev_input() in two functions to save memory in exact match
cache processing
- Use variable length arrays to reduce cache footprint
- Rebased
Daniele Di Proietto (3):
ofpbuf: Add ofpbuf_rss()
dpif-netdev: Add dp_netdev_packet_rss()
dpif-netdev: Exact match cache
lib/dpif-netd
On 8/18/14, 11:06 PM, "Pravin Shelar" wrote:
>On Mon, Aug 18, 2014 at 1:29 PM, Daniele Di Proietto
> wrote:
>> ofpbuf_rss() is introduced to get the RSS hash from the NIC. It works
>>only for
>> packets received by DPDK (otherwise it returns 0, which tells t
Thanks for the detailed and fast review.
I applied all your suggestions and I'll post a v4 soon
On 8/18/14, 11:06 PM, "Pravin Shelar" wrote:
>On Mon, Aug 18, 2014 at 1:29 PM, Daniele Di Proietto
> wrote:
>> Since lookups in the classifier can be pretty expensive, we
struct dpif_packet's member 'dp_hash' is removed
and 'pkt.hash.rss' from DPDK mbuf is used
This commit also configure DPDK devices to compute RSS hash for UDP and IPv6
packets
Signed-off-by: Daniele Di Proietto
---
lib/dpif-netdev.c | 2 +-
lib/netdev-bsd.c | 1 +
li
s:
+ fixed several coding style issues
v2:
- Implemented Pravin suggestions:
+ dp_netdev_rss()
+ reverse order of batch lookup in dp_netdev_queue_batches()
- Split dp_netdev_input() in two functions to save memory in exact match
cache processing
- Use variable lengt
Since lookups in the classifier can be pretty expensive, we introduce this
(thread local) cache which simply compares the miniflows of the packets
Signed-off-by: Daniele Di Proietto
---
lib/dpif-netdev.c | 431 +-
1 file changed, 358
ow it is needed.
I will post a v5 soon.
Thanks,
Daniele
On 8/29/14, 1:41 PM, "Pravin Shelar" wrote:
>On Tue, Aug 19, 2014 at 5:11 PM, Daniele Di Proietto
> wrote:
>> Since lookups in the classifier can be pretty expensive, we introduce
>>this
>> (thread local) ca
()
- Split dp_netdev_input() in two functions to save memory in exact match
cache processing
- Use variable length arrays to reduce cache footprint
- Rebased
Daniele Di Proietto (2):
packet-dpif: Add dpif_packet_{get,set}_hash()
dpif-netdev: Exact match cache
lib/dpif-n
struct dpif_packet's member 'dp_hash' is removed
and 'pkt.hash.rss' from DPDK mbuf is used
This commit also configure DPDK devices to compute RSS hash for UDP and IPv6
packets
Signed-off-by: Daniele Di Proietto
---
lib/dpif-netdev.c | 2 +-
lib/netdev-bsd.c | 1 +
li
Since lookups in the classifier can be pretty expensive, we introduce this
(thread local) cache which simply compares the miniflows of the packets
Signed-off-by: Daniele Di Proietto
---
lib/dpif-netdev.c | 447 +-
1 file changed, 374
On 8/29/14, 4:33 PM, "Pravin Shelar" wrote:
>On Fri, Aug 29, 2014 at 4:06 PM, Daniele Di Proietto
> wrote:
>> This series introduces in dpif-netdev an exact match cache for
>>frequently
>> used flows. It uses the RSS hash (currently implemented only for
>&g
dp_netdev_free() must free 'dp->upcall_rwlock', but when upcalls are disabled
(if the datapath is being freed upcalls should be disabled) 'dp->upcall_rwlock'
is taken and freeing it causes an assertion to fail.
This commit takes makes sure that the upcalls are disabled and releases
'dp->upcall_rwl
On 9/2/14, 1:39 PM, "Pravin Shelar" wrote:
>On Fri, Aug 29, 2014 at 4:52 PM, Daniele Di Proietto
> wrote:
>> dp_netdev_free() must free 'dp->upcall_rwlock', but when upcalls are
>>disabled
>> (if the datapath is being freed upcalls should be
disabled and releases
'dp->upcall_rwlock' before freeing it. A simple testcase is added to detect the
failure.
Signed-off-by: Daniele Di Proietto
---
lib/dpif-netdev.c| 17 -
tests/dpif-netdev.at | 8
2 files changed, 24 insertions(+), 1 deletion(-)
diff --git
On 9/2/14, 3:28 PM, "Pravin Shelar" wrote:
>On Tue, Sep 2, 2014 at 2:44 PM, Daniele Di Proietto
> wrote:
>> On 9/2/14, 1:39 PM, "Pravin Shelar" wrote:
>>
>>>On Fri, Aug 29, 2014 at 4:52 PM, Daniele Di Proietto
>>> wrote:
>>>&
r the specific CPU
Signed-off-by: Daniele Di Proietto
---
This commit improves 1-flow UDP 64-bytes packets test throughput by 6%
(compiled without -march=native)
---
lib/dpif-netdev.c | 22 ++
1 file changed, 14 insertions(+), 8 deletions(-)
diff --git a/lib/dpif-netdev.c b/lib
(performance related) benefits:
- Because of (1) the functions can be inlined.
- Because of (2) and (3) the netdev_flow_key can be treated as POD.
Specifically, because of (3), we can do comparisons with memcmp, since if the
map is different the miniflow must be different.
Signed-off-by: Daniele Di
Hi Jarno,
Thanks for figuring out this bug.
I erroneously thought that since commit 623540e4617e (dpif-netdev:
Streamline miss handling.) dpif_netdev_execute() could not be called
anymore inside a dp_netdev_input() frame.
I added the exact match cache on the non pmd-thread path for two reasons:
Thanks for the fix, Jarno
Acked-by: Daniele Di Proietto
On 9/9/14, 8:20 AM, "Jarno Rajahalme" wrote:
>Here¹s the updated description:
>
>lib/dpif-netdev: Make emc_mutex recursive.
>
>dpif_netdev_execute may be called while doing upcall processing.
>
should use dpif_packet_get_dp_hash() to get each packet
private hash.
This commit fixes the build with DPDK
Signed-off-by: Daniele Di Proietto
---
lib/odp-execute.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/lib/odp-execute.c b/lib/odp-execute.c
index e4bee18..78b1f
Thanks,
Daniele
On 9/9/14, 2:22 PM, "Jarno Rajahalme" wrote:
>Pushed to master,
>
> Jarno
>
>On Sep 9, 2014, at 2:15 PM, Jarno Rajahalme wrote:
>
>> Acked-by: Jarno Rajahalme
>>
>> On Sep 9, 2014, at 1:32 PM, Daniele Di Proietto
>> wro
Signed-off-by: Daniele Di Proietto
---
I didn't test the fix. If it works "netdev-dpdk.o" should appear in the travis
output (on master it doesn't)
Thanks,
Daniele
---
.travis.yml | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/.travis.yml b/.trav
On 9/9/14, 3:40 PM, "Ben Pfaff" wrote:
>On Tue, Sep 09, 2014 at 03:17:00PM -0700, Daniele Di Proietto wrote:
>> Signed-off-by: Daniele Di Proietto
>> ---
>> I didn't test the fix. If it works "netdev-dpdk.o" should appear in the
>>travis
On 9/10/14, 7:54 AM, "Thomas Graf" wrote:
>On 09/09/14 at 10:45pm, Daniele Di Proietto wrote:
>> I¹ll try to figure out what¹s wrong an post another patch.
>
>Daniele,
>
>I have a patchset to properly enable the DPDK build, fix CFLAGS
>and disable some errors to
Hi Thomas,
I finally managed to care of the issue. With the following incremental on
the top of your series OVS DPDK builds properly on travis:
diff --git a/.travis/build.sh b/.travis/build.sh
index c096eaf..b96181f 100755
--- a/.travis/build.sh
+++ b/.travis/build.sh
@@ -23,9 +23,8 @@ function i
concisely, like this:
ovs-appctl dpctl/add-flow dummy@br0
"in_port(1),eth_type(0x0800),ipv4(frag=no)² 2
Tested-by: Daniele Di Proietto
On 9/11/14, 11:25 AM, "Jarno Rajahalme" wrote:
>
>On Sep 11, 2014, at 11:11 AM, Ben Pfaff wrote:
>
>> On Thu, Sep 11, 2014 at
it aligned.
>
>I¹d like to see if we could generalize the inlining benefits, e.g., by
>defining special ³inline only² functions in lib/flow.h, if necessary.
>
>So, maybe you could take a look and consider if you could build on those
>two in the direction you proposed here?
>
&g
There¹s a minor issue with this patch: not every option is supported by
every compiler (for example, on my system, clang doesn¹t support -Og).
In this case ovs-dev.py fails to generate a Makefile.
I think we can address this in a separate commit. In the meantime:
Acked-by: Daniele Di Proietto
Acked-by: Daniele Di Proietto
On 9/5/14, 2:20 PM, "Ethan Jackson" wrote:
>They have slightly different support characteristics, so it's nice to
>easily switch between them for testing.
>
>Signed-off-by: Ethan Jackson
>---
> utilities/ovs-dev.py | 18 ++
--enable-dummy was useless anyway for ovsdb-server. Now it is an error to pass
it.
Signed-off-by: Daniele Di Proietto
---
utilities/ovs-dev.py | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/utilities/ovs-dev.py b/utilities/ovs-dev.py
index 5cf4771..869d89e 100755
--- a
If 'configure' with clang fails (this can happen for example if it doesn't
support all the cflags), simply skip it
Signed-off-by: Daniele Di Proietto
---
utilities/ovs-dev.py | 9 ++---
1 file changed, 6 insertions(+), 3 deletions(-)
diff --git a/utilities/ovs-dev.py b/utili
Thanks for the fix!
Acked-by: Daniele Di Proietto
On 9/15/14, 1:34 PM, "Alex Wang" wrote:
>With the separation of tx queue and rx queue configuration
>in netdev-dpdk module, the netdev_dpdk_get_config() can no
>longer report 'n_rxq' as tx queue configuration.
&
Acked-by: Daniele Di Proietto
On 9/15/14, 2:18 PM, "Alex Wang" wrote:
>dpdk_eth_dev_init() must be called with dpdk_mutex. However,
>netdev_dpdk_set_multiq() fails to follow this rule. This commit
>fixes this breach.
>
>Found by clang.
>
>Signed-off-by: Alex
Hi Ben,
I tested it on GCC 4.9.1 (debian testing) and GCC 4.8.2 (ubuntu 14.04.1)
and it works.
I think the errors you are experiencing come from sparse.
My guess is that sparse does not support _Pragma(message()), but since GCC
supports it, HAVE_PRAGMA_MESSAGE is 1.
I think the best solution wou
Since sparse doesn't support _Pragma(message()), but uses the same defines as
GCC, the BUILD_MESSAGE macro will cause sparse errors.
This commit simply disables BUILD_MESSAGE if __CHECKER__ is defined
Reported-by: Ben Pfaff
Signed-off-by: Daniele Di Proietto
---
lib/compiler.h | 2 +-
1
On 14/04/2016 14:51, "Ben Pfaff" wrote:
>On Tue, Apr 12, 2016 at 11:44:15AM +0100, Bhanuprakash Bodireddy wrote:
>> When using DPDK datapath, the OVS configure script requires the DPDK
>> build directory passed on --with-dpdk. This can be avoided if DPDK
>> library, headers are in standard comp
Thanks for the patch! I pushed this to master
2016-04-14 9:40 GMT-07:00 mweglicx :
> Following changes are applied:
> - INSTALL.DPDK.md: CONFIG_RTE_BUILD_COMBINE_LIBS step has been
>removed because it is no longer present in DPDK configuration
>(combined library is created by default),
>
Hi Mark,
On 14/04/2016 05:36, "Kavanagh, Mark B" wrote:
>Hi Daniele,
>
>One minor comment inline.
>
>Cheers,
>Mark
>
>>
>>The interface will be more similar to the cmap.
>>
>>Signed-off-by: Daniele Di Proietto
>>---
>> l
t, I removed the reference to ovs_mutex_cond_wait() there.
Thanks for noticing this!
>
>>
>>In system_stats_thread_func() the code relied on ovs_mutex_cond_wait()
>>to introduce a quiescent state, so explicit calls to
>>ovsrcu_quiesce_start() and ovsrcu_quiesce_end() a
;This mechanism can also be used to:
>>* Automatically match the number of rxq with the one provided by qemu
>> via the new_device callback.
>>* Provide a way to change the MTU of dpdk devices at runtime.
>>* Move a DPDK vhost device to the proper NUMA socket.
hing strategy in conntrack_execute() to allow a newly
created connection to be picked up by packets in the same batch.
* Added an ovs-test module to throw pcap files at the connection tracker.
* Added a workaround for the userspace testsuite on new kernels and a tcp
non-conntrack test.
Daniel
This will be used by a future commit.
Signed-off-by: Daniele Di Proietto
---
lib/flow.c | 140 ++---
lib/flow.h | 3 ++
2 files changed, 81 insertions(+), 62 deletions(-)
diff --git a/lib/flow.c b/lib/flow.c
index 560a90f..972a996
The function simply returns the ethernet type of the packet (after
eventually discarding the VLAN tag). It will be used by a following
commit.
Signed-off-by: Daniele Di Proietto
---
lib/flow.c | 14 --
lib/flow.h | 1 +
2 files changed, 13 insertions(+), 2 deletions(-)
diff --git
Signed-off-by: Daniele Di Proietto
---
lib/conntrack.c | 21 +
lib/conntrack.h | 2 ++
2 files changed, 23 insertions(+)
diff --git a/lib/conntrack.c b/lib/conntrack.c
index 840335b..7913e76 100644
--- a/lib/conntrack.c
+++ b/lib/conntrack.c
@@ -849,3 +849,24 @@ delete_conn
Signed-off-by: Daniele Di Proietto
---
lib/dpif-netdev.c | 60 ---
1 file changed, 57 insertions(+), 3 deletions(-)
diff --git a/lib/dpif-netdev.c b/lib/dpif-netdev.c
index 436359a..5ac2bf3 100644
--- a/lib/dpif-netdev.c
+++ b/lib/dpif
Linux and FreeBSD have slightly different names for these constants.
Windows doesn't define them. It is simpler to redefine them from
scratch for OVS. The new names are different than those used in Linux
and FreeBSD.
These definitions will be used by a future commit.
Signed-off-by: Danie
different batching and locking strategies.
E.g. the line:
`./test/ovstest test-conntrack benchmark 1 1488 32`
starts 1 thread that will send 1488 packets to the connection
tracker, 32 at a time. It will print the time taken to process them.
Signed-off-by: Daniele Di Proietto
---
tests
Signed-off-by: Daniele Di Proietto
---
lib/dpif-netdev.c | 10 +-
1 file changed, 9 insertions(+), 1 deletion(-)
diff --git a/lib/dpif-netdev.c b/lib/dpif-netdev.c
index 5ac2bf3..995cbc0 100644
--- a/lib/dpif-netdev.c
+++ b/lib/dpif-netdev.c
@@ -3991,6 +3991,14
This commit implements the OVS_ACTION_ATTR_CT action in dpif-netdev.
To allow ofproto-dpif to detect the conntrack feature, flow_put will not
discard anymore flows with ct_* fields set. We still shouldn't allow
flows with NAT bits set, since there is no support for NAT.
Signed-off-by: Danie
g.
Signed-off-by: Daniele Di Proietto
---
tests/test-conntrack.c | 65 ++
1 file changed, 65 insertions(+)
diff --git a/tests/test-conntrack.c b/tests/test-conntrack.c
index 414d7dc..5292320 100644
--- a/tests/test-conntrack.c
+++ b/tests/test-c
Useful to test the datapath ability to forward tcp packets without the
complexity of connection tracking.
Signed-off-by: Daniele Di Proietto
---
tests/system-traffic.at | 20
1 file changed, 20 insertions(+)
diff --git a/tests/system-traffic.at b/tests/system-traffic.at
While the system testsuite already has connection tracking tests, it
will be still useful to add some to the standard testsuite because:
* They're run more often by developers.
* Some of them are more interesting for the userspace datapath.
Signed-off-by: Daniele Di Proietto
---
tests/of
Signed-off-by: Daniele Di Proietto
---
lib/conntrack-private.h | 3 ++
lib/conntrack-tcp.c | 34 +
lib/conntrack.c | 125
lib/conntrack.h | 16 +++
4 files changed, 178 insertions(+)
diff --git a/lib
a workaround.
Signed-off-by: Daniele Di Proietto
---
tests/system-common-macros.at| 1 +
tests/system-kmod-macros.at | 7 +++
tests/system-userspace-macros.at | 18 ++
3 files changed, 26 insertions(+)
diff --git a/tests/system-common-macros.at b/tests/system-c
ff-by: Daniele Di Proietto
---
tests/system-kmod-macros.at | 28 +++
tests/system-traffic.at | 49 ++--
tests/system-userspace-macros.at | 45 +---
3 files changed, 107 insertions(+), 15 deletions(-)
ding style and to allow the pickup of already
established connections.
Signed-off-by: Daniele Di Proietto
---
COPYING | 1 +
debian/copyright.in | 4 +
lib/automake.mk | 5 +
lib/conntrack-other.c | 91 ++
lib/conntrack-private.h | 77 +
lib/connt
instead of 'unsigned' for 'len', since the value can be
negative.
CC: Sairam Venugopal
Signed-off-by: Daniele Di Proietto
---
I tested a similar fix on the userspace connection tracker, but I didn't
compile this for the windows datapath.
---
datapath-windows/ovsext/Conntrack-
W(199): Line has trailing whitespace
+
This commit suppresses the two warnings for lines with form feeds as the
only character.
Signed-off-by: Daniele Di Proietto
---
utilities/checkpatch.py | 8 ++--
1 file changed, 6 insertions(+), 2 deletions(-)
diff --git a/utilities/checkpatch.py b/uti
ch_size must be between 1 and
>NETDEV_MAX_BURST(%u)",
You're right there are several genuine errors reported by checkpatch.py.
I'll fix them.
>
>patches 7/15 and 8/15, 10/15, 11/15 - as there's no comment into the
>patch, maybe
>they just need an empty line bef
On 18/04/2016 07:50, "Ilya Maximets" wrote:
>On 08.04.2016 06:13, Daniele Di Proietto wrote:
>> Signed-off-by: Daniele Di Proietto
>> ---
>> lib/dpif-netdev.c | 243
>>+++---
>> 1 file changed, 175
On 18/04/2016 14:26, "Ben Pfaff" wrote:
>On Mon, Apr 18, 2016 at 02:02:37PM -0700, Daniele Di Proietto wrote:
>> CodingStyle.md says:
>>
>> "Use form feeds (control+L) to divide long source files into logical
>> pieces. A form feed sho
now accessed
while holding the appropriate mutex.
* Fixed some outdated comments about rx queue configuration.
Daniele Di Proietto (16):
dpif-netdev: Destroy 'port_mutex' in dp_netdev_free().
dpif-netdev: Remove unused 'index' in dp_netdev_pmd_thread.
dpif-netdev: F
Signed-off-by: Daniele Di Proietto
---
lib/dpif-netdev.c | 14 +-
1 file changed, 5 insertions(+), 9 deletions(-)
diff --git a/lib/dpif-netdev.c b/lib/dpif-netdev.c
index 24717cc..060f5e0 100644
--- a/lib/dpif-netdev.c
+++ b/lib/dpif-netdev.c
@@ -430,8 +430,6 @@ struct
1 - 100 of 1317 matches
Mail list logo
