The list of remote maintenance points can be used by other daemons
in leader election. This commit makes ovs-vswitchd update the
row of remote maintenance points in ovsdb immediately after it is
changed.
Signed-off-by: Alex Wang
---
lib/cfm.c |2 ++
1 file changed, 2 insertions(+)
diff --g
Reported-by: Mukesh Hira
Signed-off-by: Ben Pfaff
---
lib/socket-util.c |4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/lib/socket-util.c b/lib/socket-util.c
index 4c79cd6..aeda543 100644
--- a/lib/socket-util.c
+++ b/lib/socket-util.c
@@ -53,9 +53,9 @@ VLOG_DEFINE_THI
On 11/03/14 19:41, Zoltan Kiss wrote:
On 07/03/14 17:59, 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
Acked-by: Andy Zhou
On Tue, Mar 11, 2014 at 1:56 PM, Ben Pfaff wrote:
> This should scale better than a single mutex, though still not
> ideally.
>
> Signed-off-by: Ben Pfaff
> ---
> lib/dpif-netdev.c | 99
> +++--
> 1 file changed, 74 inser
Acked-by: Andy Zhou
On Tue, Mar 11, 2014 at 1:56 PM, Ben Pfaff wrote:
> This allows clients to do more than just increment a counter. The
> following commit will make the first use of that feature.
>
> Signed-off-by: Ben Pfaff
> ---
> lib/dpif-netdev.c | 75 +++
log_backtrace() and log_backtrace_msg() logs the back trace into
the log file. It may be most useful when debugging unit tests.
"backtrace.h" documents the usage. It is not being called directly
in the code, but rather as a handy tool available when needed.
---
v1->v2:
address BEN's review fe
Thanks for the input. I will send out a V2 implementing the suggested
changes.
On Fri, Mar 14, 2014 at 1:33 PM, Ben Pfaff wrote:
> On Thu, Mar 13, 2014 at 06:17:14PM -0700, Andy Zhou wrote:
> > LOG_BACKTRACE() logs the back trace into the log file. I find it
> > to be helpful when debugging uni
On Thu, Mar 13, 2014 at 06:17:14PM -0700, Andy Zhou wrote:
> LOG_BACKTRACE() logs the back trace into the log file. I find it
> to be helpful when debugging unit tests. "backtrace.h" documents
> the usage. It is not being called directly in the code, but rather as
> a handy tool available when need
On Mar 14, 2014, at 2:25 AM, Andy Zhou wrote:
> How about using atomic operations to maintain stats instead of per thread
> buckets? It seems to be a win at least within a NUMA node from kernel
> datapath experience.
Atomic modify operations require locking the bus and are thus very expensive
I think that using atomics and using per-thread or per-NUMA node
buckets is orthogonal: one can use either technique without the other.
I am reluctant to use atomics here because I think we need 64-bit
stats (packet counters can definitely roll over past 2**32) and 64-bit
atomics are internally im
On Fri, Mar 14, 2014 at 10:44 AM, Ben Pfaff wrote:
> On Fri, Mar 14, 2014 at 01:36:15AM -0700, Andy Zhou wrote:
> >Acked-by: Andy Zhou
>
> Thanks.
>
> > On Tue, Mar 11, 2014 at 1:56 PM, Ben Pfaff wrote:
> >
> > > RCU allows multiple threads to read objects in parallel without any
> > > perf
On Fri, Mar 14, 2014 at 01:36:15AM -0700, Andy Zhou wrote:
>Acked-by: Andy Zhou
Thanks.
> On Tue, Mar 11, 2014 at 1:56 PM, Ben Pfaff wrote:
>
> > RCU allows multiple threads to read objects in parallel without any
> > performance penalty. The following commit will introduce the first use.
Windows does not have a sys/fcntl.h
Signed-off-by: Gurucharan Shetty
---
utilities/ovs-ofctl.c |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/utilities/ovs-ofctl.c b/utilities/ovs-ofctl.c
index e62e646..8ae7c34 100644
--- a/utilities/ovs-ofctl.c
+++ b/utilities/ovs-ofctl.
advapi32 is needed by multiple functions
So include it in a common place.
Signed-off-by: Gurucharan Shetty
---
include/windows/windefs.h |2 ++
lib/daemon-windows.c |2 --
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/include/windows/windefs.h b/include/windows/wind
We can now compile a few OVS userspace components on Windows.
Signed-off-by: Gurucharan Shetty
---
BUILD.Windows |8 +---
1 file changed, 5 insertions(+), 3 deletions(-)
diff --git a/BUILD.Windows b/BUILD.Windows
index 9bbaa80..119334b 100644
--- a/BUILD.Windows
+++ b/BUILD.Windows
@@ -
Windows does not have a srandom() and random(). But it does
have a srand() and rand(). We use these functions in sflow code.
Signed-off-by: Gurucharan Shetty
---
include/windows/windefs.h |3 +++
1 file changed, 3 insertions(+)
diff --git a/include/windows/windefs.h b/include/windows/windef
Signed-off-by: Gurucharan Shetty
---
include/windows/unistd.h |5 +
1 file changed, 5 insertions(+)
diff --git a/include/windows/unistd.h b/include/windows/unistd.h
index 0f553bb..d9ded5a 100644
--- a/include/windows/unistd.h
+++ b/include/windows/unistd.h
@@ -18,4 +18,9 @@
#define fs
This does not provide us all the functionality that
is available in Linux. But should be a start.
Signed-off-by: Gurucharan Shetty
---
include/windows/automake.mk |1 +
vswitchd/system-stats.c | 17 +
2 files changed, 18 insertions(+)
create mode 100644 include/windows
sflow.h uses u_int32_t that needs windefs.h for compilation
in visual studio.
Signed-off-by: Gurucharan Shetty
---
include/windows/windefs.h |1 +
lib/sflow.h |4
2 files changed, 5 insertions(+)
diff --git a/include/windows/windefs.h b/include/windows/windefs.h
index
Visual studio does not like it. It is mostly a keyword.
Signed-off-by: Gurucharan Shetty
---
ofproto/in-band.c |6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/ofproto/in-band.c b/ofproto/in-band.c
index 265dcb2..416043e 100644
--- a/ofproto/in-band.c
+++ b/ofproto/in
Windows does not have a strsep.
Signed-off-by: Gurucharan Shetty
---
COPYING |1 +
debian/copyright.in |3 +-
include/windows/windefs.h |2 ++
lib/automake.mk |1 +
lib/strsep.c | 75 +
5
Hi!
Search is provided in mininet-discuss and openflow-discuss mailing archives.
But why it has not been provided in ovs-dev and ovs-discuss archives? How can I
search my question in ovs-dev and ovs-discuss?
Regards,Rizwan Jamil ___
Acked-by: Andy Zhou
On Tue, Mar 11, 2014 at 1:56 PM, Ben Pfaff wrote:
> Signed-off-by: Ben Pfaff
> ---
> lib/dpif-netdev.c | 148
> -
> 1 file changed, 43 insertions(+), 105 deletions(-)
>
> diff --git a/lib/dpif-netdev.c b/lib/dpif-netd
How about using atomic operations to maintain stats instead of per thread
buckets? It seems to be a win at least within a NUMA node from kernel
datapath experience.
On Tue, Mar 11, 2014 at 1:56 PM, Ben Pfaff wrote:
> This allows clients to do more than just increment a counter. The
> following
Acked-by: Andy Zhou
On Tue, Mar 11, 2014 at 1:56 PM, Ben Pfaff wrote:
> This factors code out of fat-rwlock, making it easily usable by other code.
>
> Signed-off-by: Ben Pfaff
> ---
> lib/fat-rwlock.c | 29 ++
> lib/util.c | 60
>
Acked-by: Andy Zhou
On Tue, Mar 11, 2014 at 1:56 PM, Ben Pfaff wrote:
> Signed-off-by: Ben Pfaff
>
> Conflicts:
> lib/ovs-rcu.h
> ---
> ofproto/connmgr.c|5 +-
> ofproto/ofproto-dpif-xlate.c |2 -
> ofproto/ofproto-dpif.c |2 -
> ofproto/ofproto-provid
Acked-by: Andy Zhou
On Tue, Mar 11, 2014 at 1:56 PM, Ben Pfaff wrote:
> RCU allows multiple threads to read objects in parallel without any
> performance penalty. The following commit will introduce the first use.
>
> Signed-off-by: Ben Pfaff
> ---
> lib/automake.mk |2 +
When creating a flow in the datapath as the result of an upcall
the match itself is the match supplied in the upcall while
the mask of the match, if supplied, is generated based on the
flow and mask composed during action translation.
In the case of, for example a UDP packet, the match will includ
Hi Pravin, Hi Jesse, Hi All,
I have updated this patch so that:
* The datapath rejects push MPLS actions in the presence of VLAN tags.
I have done this by white-listing the following:
- ETH_P_IP (0x0800)
- ETH_P_ARP (0x0806)
- ETH_P_RARP (0x0835)
- ETH_P_IPV6 (0x86DD)
Previously it
Allow datapath to recognize and extract MPLS labels into flow keys
and execute actions which push, pop, and set labels on packets.
Based heavily on work by Leo Alterman, Ravi K, Isaku Yamahata and Joe Stringer.
Cc: Ravi K
Cc: Leo Alterman
Cc: Isaku Yamahata
Cc: Joe Stringer
Signed-off-by: Sim
30 matches
Mail list logo
