OpenFlow 1.1 through 1.4 say that the out_group in a flow_mod, for commands
other than OFPFC_DELETE or OFPFC_DELETE_STRICT, should be ignored. OVS
actually reported an error. This fixes the problem.
Reported-by: chen zhang <3zhangchen9...@gmail.com>
Signed-off-by: Ben Pfaff
---
AUTHORS
On Fri, Dec 06, 2013 at 12:02:59PM +0800, Alexander Wu wrote:
> Some additional test for bitmap performance (10 million loops):
>
> suse-kvm-of13:/test # time ./bit3
>
> real0m0.061s
> user0m0.060s
> sys 0m0.000s
> suse-kvm-of13:/test # time ./bit2
>
> real
On Fri, Dec 06, 2013 at 10:39:21AM +0800, Alexander Wu wrote:
> On 06/12/2013 01:47, Ben Pfaff wrote:
> >On Wed, Dec 04, 2013 at 05:38:29PM +0800, Alexander Wu wrote:
> >>V4:
> >> Add bitmap_count1 function to count all 1 bits.
> >
> >I suggest using the count_1bits() function to speed this up.
>
On Fri, Dec 06, 2013 at 12:29:24PM +0900, Simon Horman wrote:
> On Thu, Dec 05, 2013 at 09:56:25AM -0800, Ben Pfaff wrote:
> > On Thu, Dec 05, 2013 at 11:26:06AM +0900, Simon Horman wrote:
> > > On Wed, Dec 04, 2013 at 05:01:11PM -0800, Ben Pfaff wrote:
> > > > On Thu, Dec 05, 2013 at 09:51:39AM +0
On 06/12/2013 10:39, Alexander Wu wrote:
On 06/12/2013 01:47, Ben Pfaff wrote:
On Wed, Dec 04, 2013 at 05:38:29PM +0800, Alexander Wu wrote:
V4:
Add bitmap_count1 function to count all 1 bits.
I suggest using the count_1bits() function to speed this up.
Thanks,
Ben.
Hi Ben,
Thanks f
Earlier RHEL6 kernel do not have linux/atomic.h header
file. Use types.h to get atomic_t definition.
Signed-off-by: Pravin B Shelar
---
datapath/linux/compat/include/linux/net.h |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/datapath/linux/compat/include/linux/net.h
On Thu, Dec 05, 2013 at 09:56:25AM -0800, Ben Pfaff wrote:
> On Thu, Dec 05, 2013 at 11:26:06AM +0900, Simon Horman wrote:
> > On Wed, Dec 04, 2013 at 05:01:11PM -0800, Ben Pfaff wrote:
> > > On Thu, Dec 05, 2013 at 09:51:39AM +0900, Simon Horman wrote:
> > > > On Wed, Dec 04, 2013 at 04:44:17PM -0
On 06/12/2013 01:47, Ben Pfaff wrote:
On Wed, Dec 04, 2013 at 05:38:29PM +0800, Alexander Wu wrote:
V4:
Add bitmap_count1 function to count all 1 bits.
I suggest using the count_1bits() function to speed this up.
Thanks,
Ben.
Hi Ben,
Thanks for your suggestion! I'll update the functi
Looks fine to me. Thanks.
On Thu, Dec 5, 2013 at 5:02 PM, Ben Pfaff wrote:
> I sent out a revised version:
> http://openvswitch.org/pipermail/dev/2013-December/034644.html
>
> Will you look at it?
>
> On Tue, Dec 03, 2013 at 09:59:19PM -0800, Henry Mai wrote:
>> Sorry, found another violation whe
On Thu, Dec 05, 2013 at 04:27:23PM -0800, Jarno Rajahalme wrote:
> parse_l2_5_onward() previously used out of range 'expected_bit' when
> it was called with no attrs, which happens when parsing a mask with
> zero length.
>
> Signed-off-by: Jarno Rajahalme
Acked-by: Ben Pfaff
___
On 2013/12/4 13:24, Ben Pfaff wrote:
> On Wed, Dec 04, 2013 at 10:24:28AM +0800, jerry wrote:
>> On 2013/12/3 13:08, Ben Pfaff wrote:
>>> On Tue, Dec 03, 2013 at 10:18:43AM +0800, jerry wrote:
It may be easily supported to forward unicast packets. But I'm not sure
how to set OpenFlow rule
I sent out a revised version:
http://openvswitch.org/pipermail/dev/2013-December/034644.html
Will you look at it?
On Tue, Dec 03, 2013 at 09:59:19PM -0800, Henry Mai wrote:
> Sorry, found another violation where I had an extra newline before a function.
>
> On Tue, Dec 3, 2013 at 8:30 PM, Henry
Signed-off-by: Henry Mai
Signed-off-by: Ben Pfaff
---
NEWS |1 +
lib/vlog.c| 97 ++---
lib/vlog.h| 36 +++--
lib/vlog.man |5 +++
utilities/ovs-appctl.8.in |6 +++
Anything inside vlog_init__() that tried to log a message was going to
deadlock, since it would hit pthread_once() recursively and wait for the
previous call to complete. Unfortunately, there was a VLOG_ERR call inside
vlog_init__(), only called in the corner case where the system's clock was
wron
On Thu, Dec 05, 2013 at 04:36:26PM -0800, Jarno Rajahalme wrote:
> Inline, use another well-known algorithm for 64-bit builds, and use
> builtins when they are known to be fast at compile time. A 32-bit
> version of the alternate algorithm is slower than the existing
> implementation, so the old o
On Dec 5, 2013, at 1:02 PM, Ethan Jackson wrote:
> I'm working on a patch to disable wildcard calculation when we're
> doing things like pushing stats and don't actually need to know what
> the mask should be. I had a question for you about your staged match
> patch on this. I notice that you
Inline, use another well-known algorithm for 64-bit builds, and use
builtins when they are known to be fast at compile time. A 32-bit
version of the alternate algorithm is slower than the existing
implementation, so the old one is used for 32-bit builds. Inline
assembler would be a bit faster on
Jarno Rajahalme (4):
lib/odp-util: Only use expected_bit when set.
dpif-netdev: Do not check mask if it doesn't exist.
v5:
- Split dpif_netdev_flow_mask_from_nlattrs() to
dpif_netdev_flow_from_nlattrs() and dpif_netdev_mask_from_nlattrs()
and make the latter to require non-zero mask_key
Ofproto always provides a valid pointer to a mask, but the mask length
is zero when megaflows are disabled. Use match_wc_init() to get an
appropriate exact match mask when there is no mask.
Signed-off-by: Jarno Rajahalme
---
v5: - Split dpif_netdev_flow_mask_from_nlattrs() to
dpif_netdev_f
parse_l2_5_onward() previously used out of range 'expected_bit' when
it was called with no attrs, which happens when parsing a mask with
zero length.
Signed-off-by: Jarno Rajahalme
---
lib/odp-util.c |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/odp-util.c b/lib/odp-
Add the missing code for generating IPv6 packets for testing purposes.
Also make flow_compose() set the l4 and l7 pointers more consistently
with flow_extract().
Signed-off-by: Jarno Rajahalme
---
v5: Set flow label and tclass directly, factor out transport layer
code for both IPv4 and IPv6.
Pushed to master.
Thanks.
On Thu, Dec 5, 2013 at 3:55 PM, Jesse Gross wrote:
> On Thu, Dec 5, 2013 at 3:50 PM, Pravin B Shelar wrote:
>> Use percpu allocator for stats due to objection to stats array.
>> But percpu allocator is not designed for high churn allocation/
>> deallcation. so we need
Thanks,
I pushed this patch to master and 2.0.
On Thu, Dec 5, 2013 at 1:57 PM, Jesse Gross wrote:
> On Wed, Dec 4, 2013 at 9:24 PM, Pravin B Shelar wrote:
>> Add missing return statement.
>>
>> Signed-off-by: Pravin B Shelar
>
> Acked-by: Jesse Gross
___
Thanks,
I pushed these patches to master.
On Wed, Dec 4, 2013 at 1:29 PM, Jesse Gross wrote:
> On Wed, Dec 4, 2013 at 11:26 AM, Pravin B Shelar wrote:
>> Following patch improves rxhash calculation, It is taken from
>> upstream Linux kernel code.
>> From kernel 3.8, skb_get_rxhash() can handle h
On Thu, Dec 5, 2013 at 3:50 PM, Pravin B Shelar wrote:
> Use percpu allocator for stats due to objection to stats array.
> But percpu allocator is not designed for high churn allocation/
> deallcation. so we need to avoid allocating percpu flow for
> short lived flows. One cheaper way to detect fl
Use percpu allocator for stats due to objection to stats array.
But percpu allocator is not designed for high churn allocation/
deallcation. so we need to avoid allocating percpu flow for
short lived flows. One cheaper way to detect flow is by checking
if 5-tuple used in RSS are masked or not. if a
Signed-off-by: Joe Stringer
---
lib/ovs-thread.c| 10 ++
lib/ovs-thread.h|4
ofproto/ofproto.c |2 +-
vswitchd/system-stats.c |2 +-
4 files changed, 16 insertions(+), 2 deletions(-)
diff --git a/lib/ovs-thread.c b/lib/ovs-thread.c
index b3a87bb..
On systems that provide /proc/cpuinfo similar to Linux on x86, this
should allow us to choose a better default value for the number of
upcall handler threads---in particular, it avoids counting hyper-thread
cores. If /proc/cpuinfo cannot be parsed for any reason, fall back to
using sysconf().
Sign
On Thu, Dec 5, 2013 at 11:42 AM, Pravin B Shelar wrote:
> diff --git a/datapath/datapath.c b/datapath/datapath.c
> index 1a5bffb..48a2955 100644
> --- a/datapath/datapath.c
> +++ b/datapath/datapath.c
> @@ -766,6 +760,7 @@ static int ovs_flow_cmd_new_or_set(struct sk_buff *skb,
> struct genl_info
On Thu, Dec 05, 2013 at 11:08:31PM +0100, Thomas Graf wrote:
> On 12/04/2013 11:20 PM, Jesse Gross wrote:
> >If enabling rings on demand can be done cleanly that might be best
> >solution. To me, it seems difficult to generalize the upcall
> >characteristics based on port type.
>
> It would requir
On 12/04/2013 11:20 PM, Jesse Gross wrote:
If enabling rings on demand can be done cleanly that might be best
solution. To me, it seems difficult to generalize the upcall
characteristics based on port type.
It would require to reopen sockets but I don't see that as a major
obstacle.
16K ports
On Wed, Dec 4, 2013 at 9:24 PM, Pravin B Shelar wrote:
> Add missing return statement.
>
> Signed-off-by: Pravin B Shelar
Acked-by: Jesse Gross
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
On Thu, Dec 05, 2013 at 02:53:20PM -0700, Anu Mercian wrote:
> I am trying to send a OF Packet-out message from the controller with
> data, different from the packet Data of the OF Packet In message,
> assuming that my buffer-id is not set, so the entire packet data is
> sent to the controller. i.e
Hi everyone,
I am trying to send a OF Packet-out message from the controller with data,
different from the packet Data of the OF Packet In message, assuming that my
buffer-id is not set, so the entire packet data is sent to the controller. i.e,
create a string, convert it into byte format and se
On 12/03/13 at 09:43pm, Jesse Gross wrote:
Thanks for merging a first set of patches
> On Sat, Nov 30, 2013 at 4:21 AM, Thomas Graf wrote:
> > diff --git a/net/openvswitch/datapath.c b/net/openvswitch/datapath.c
> > index 8eaa39a..867edf1 100644
> > --- a/net/openvswitch/datapath.c
> > +++ b/net
I'm working on a patch to disable wildcard calculation when we're
doing things like pushing stats and don't actually need to know what
the mask should be. I had a question for you about your staged match
patch on this. I notice that you disable the staged matching when
wildcards aren't passed in.
Use percpu allocator for stats due to objection to stats array.
But percpu allocator is not designed for high churn allocation/
deallcation. so we need to avoid allocating percpu flow for
short lived flows. One cheaper way to detect flow is by checking
if 5-tuple used in RSS are masked or not. if a
Thanks, I'll apply this soon.
On Wed, Dec 04, 2013 at 06:18:43PM -0800, Ethan Jackson wrote:
> Acked-by: Ethan Jackson
>
>
> On Wed, Dec 4, 2013 at 4:47 PM, Ben Pfaff wrote:
> > This should behave the same as before but the code reads more naturally to
> > me this way.
> >
> > Signed-off-by: B
On Thu, Dec 5, 2013 at 7:40 PM, Ben Pfaff wrote:
> On Thu, Dec 05, 2013 at 05:14:40PM +0200, Alexandru Copot wrote:
>> I want to announce that I want to work on the bundles feature for
>> OpenFlow 1.4. This way we can avoid duplicating the effort.
>>
>> Most of the bundle features are easy to imp
On Thu, Dec 05, 2013 at 11:26:06AM +0900, Simon Horman wrote:
> On Wed, Dec 04, 2013 at 05:01:11PM -0800, Ben Pfaff wrote:
> > On Thu, Dec 05, 2013 at 09:51:39AM +0900, Simon Horman wrote:
> > > On Wed, Dec 04, 2013 at 04:44:17PM -0800, Ben Pfaff wrote:
> > > > On Thu, Dec 05, 2013 at 08:58:49AM +0
On Thu, Dec 05, 2013 at 09:46:04AM -0800, Gurucharan Shetty wrote:
> On Thu, Dec 5, 2013 at 9:28 AM, Ben Pfaff wrote:
> > Found by inspection.
> >
> > Signed-off-by: Ben Pfaff
> > ---
> > m4/openvswitch.m4 |2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/m4/openv
On Wed, Dec 04, 2013 at 05:38:29PM +0800, Alexander Wu wrote:
> V4:
> Add bitmap_count1 function to count all 1 bits.
I suggest using the count_1bits() function to speed this up.
Thanks,
Ben.
___
dev mailing list
dev@openvswitch.org
http://openvswitc
On Thu, Dec 5, 2013 at 9:28 AM, Ben Pfaff wrote:
> Found by inspection.
>
> Signed-off-by: Ben Pfaff
> ---
> m4/openvswitch.m4 |2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/m4/openvswitch.m4 b/m4/openvswitch.m4
> index 7c37985..5def3d2 100644
> --- a/m4/openvswitch.
On Thu, Dec 05, 2013 at 05:14:40PM +0200, Alexandru Copot wrote:
> I want to announce that I want to work on the bundles feature for
> OpenFlow 1.4. This way we can avoid duplicating the effort.
>
> Most of the bundle features are easy to implement so this is what I am going
> to
> do first. How
On Thu, Dec 5, 2013 at 9:26 AM, Ben Pfaff wrote:
> On Thu, Dec 05, 2013 at 05:24:51PM +, James Page wrote:
> > -BEGIN PGP SIGNED MESSAGE-
> > Hash: SHA256
> >
> > On 05/12/13 17:07, Ben Pfaff wrote:
> > > On Thu, Dec 05, 2013 at 04:16:39PM +, James Page wrote:
> > >> Later version
On Thu, Dec 05, 2013 at 05:29:05PM +, James Page wrote:
> Later versions of gcc on some architectures push atomic functions
> out into a separate atomic library; add a check to see when this
> is required and add it to LIBS if need be.
>
> Signed-off-by: James Page
Applied, thanks. I change
Later versions of gcc on some architectures push atomic functions
out into a separate atomic library; add a check to see when this
is required and add it to LIBS if need be.
Signed-off-by: James Page
---
configure.ac | 1 +
m4/openvswitch.m4 | 6 ++
2 files changed, 7 insertions(+)
dif
Found by inspection.
Signed-off-by: Ben Pfaff
---
m4/openvswitch.m4 |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/m4/openvswitch.m4 b/m4/openvswitch.m4
index 7c37985..5def3d2 100644
--- a/m4/openvswitch.m4
+++ b/m4/openvswitch.m4
@@ -118,7 +118,7 @@ OpenFlow connections
On Thu, Dec 05, 2013 at 05:24:51PM +, James Page wrote:
> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA256
>
> On 05/12/13 17:07, Ben Pfaff wrote:
> > On Thu, Dec 05, 2013 at 04:16:39PM +, James Page wrote:
> >> Later versions of gcc on some architectures push atomic
> >> functions out in
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA256
On 05/12/13 17:07, Ben Pfaff wrote:
> On Thu, Dec 05, 2013 at 04:16:39PM +, James Page wrote:
>> Later versions of gcc on some architectures push atomic
>> functions out into a separate atomic library; add a check to see
>> when this is required
Reported-by: Anton Matsiuk
Signed-off-by: Ben Pfaff
---
AUTHORS |1 +
ofproto/connmgr.c |9 +++--
2 files changed, 8 insertions(+), 2 deletions(-)
diff --git a/AUTHORS b/AUTHORS
index 1c2d9ea..cb72ac8 100644
--- a/AUTHORS
+++ b/AUTHORS
@@ -132,6 +132,7 @@ Amey Bhide
On Thu, Dec 05, 2013 at 04:16:39PM +, James Page wrote:
> Later versions of gcc on some architectures push atomic functions
> out into a separate atomic library; add a check to see when this
> is required and add it to LIBS if need be.
>
> Signed-off-by: James Page
May I assume that you have
Later versions of gcc on some architectures push atomic functions
out into a separate atomic library; add a check to see when this
is required and add it to LIBS if need be.
Signed-off-by: James Page
---
configure.ac | 1 +
m4/openvswitch.m4 | 6 ++
2 files changed, 7 insertions(+)
dif
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA256
On 05/12/13 16:10, Ben Pfaff wrote:
>> Later versions of gcc on some architectures push atomic
>> functions
>>> out into a separate atomic library; add a check to see when
>>> this is required.
>>>
>>> Signed-off-by: James Page
> Is ATOMIC_LIBS nec
On Thu, Dec 05, 2013 at 10:43:09AM +, James Page wrote:
> Later versions of gcc on some architectures push atomic functions
> out into a separate atomic library; add a check to see when this
> is required.
>
> Signed-off-by: James Page
Is ATOMIC_LIBS necessary or useful? AC_SEARCH_LIBS adds
> systemctl is part of systemd. If you have systemctl installed, then the
> system should be ready to use systemd service units just fine.
My point is that Centos and RHEL RPMS do not have the systemd files as
part of the rpms.
So if systemctl exists and the
usr_lib_systemd_system_openvswitch.serv
On Wed, Dec 04, 2013 at 01:06:35PM -0800, Gurucharan Shetty wrote:
> One more thing. There is no BRCOMPAT in master. So that will have to go.
Ok, I will remove it.
> On Wed, Dec 4, 2013 at 1:04 PM, Gurucharan Shetty wrote:
> > On Wed, Dec 4, 2013 at 12:12 PM, Flavio Leitner wrote:
> >> On Wed,
Hello,
I want to announce that I want to work on the bundles feature for OpenFlow 1.4.
This way we can avoid duplicating the effort.
Most of the bundle features are easy to implement so this is what I am going to
do first. However, the actual commit of the messages will be a challenge. I am
think
Later versions of gcc on some architectures push atomic functions
out into a separate atomic library; add a check to see when this
is required.
Signed-off-by: James Page
---
configure.ac | 1 +
m4/openvswitch.m4 | 6 ++
tests/automake.mk | 2 +-
3 files changed, 8 insertions(+), 1 delet
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA256
Hi Flavio
On 29/10/13 12:51, Flavio Leitner wrote:
> While updating the packages to 2.0.0 I found that the build fails
> on PPC.
>
> I am not an expert on PPC arch, but I think it doesn't support
> atomic operations of 8 bytes (64bit) on 32bit syst
60 matches
Mail list logo
