Hello,
For what I understand the openvSwitch works in store-and-forward mode
using linux kernel.
Can someone tell me where is the buffer and the files that are connected
and do or treats the store-and-forward on openvswitch sources. Where is
this done?
Can someone tell me if is in mind an cut
On Tue, Aug 07, 2012 at 02:21:20PM +0100, mvp...@iol.pt wrote:
> For what I understand the openvSwitch works in store-and-forward mode
> using linux kernel.
> Can someone tell me where is the buffer and the files that are connected
> and do or treats the store-and-forward on openvswitch sources.
Hello,
but in finction: netdev_linux_listen(struct netdev *netdev_) on
netdev_linux.c file inside lib, they have sockets open to the interfaces.
Where it receive a packet and copy to memory. It reads the memory content
only when packet reach the exact length of the packet, for example 1500
byt
OK, it sounds like you answered your own questions. Fine.
On Tue, Aug 07, 2012 at 04:01:05PM +0100, mvp...@iol.pt wrote:
> Hello,
>
>but in finction: netdev_linux_listen(struct netdev *netdev_) on
> netdev_linux.c file inside lib, they have sockets open to the interfaces.
> Where it receive
On Mon, Aug 06, 2012 at 11:55:15PM -0700, Mehak Mahajan wrote:
> OVS provides a utility to create ICMP packets for the purpose of
> testing using ovs-appctl netdev-dummy/receive. These packets created
> by flow_compose() earlier did not have the ICMP checksum in them.
> With this commit, the check
It's documented to return -1 if the version isn't yet known, but in fact
it returned 0.
Signed-off-by: Ben Pfaff
---
lib/vconn.c |4 ++--
lib/vconn.h |3 ++-
2 files changed, 4 insertions(+), 3 deletions(-)
diff --git a/lib/vconn.c b/lib/vconn.c
index eb848ba..8b53b06 100644
--- a/lib/v
The learning-switch implementation needs to know the OpenFlow version in
use to send the initial handshake messages (e.g. the feature request), but
the version is not always available at the time that the code currently
sends the handshake. This can cause an assertion failure later when
ofputil_en
Until now, it seems that all vconn users have immediately started reading
messages from the connection. Today, however, I added a new user that
only wants to read packets after the OpenFlow version is negotiated, so
it never called vconn_recv() before that happened. It turns out that if
you do th
OVS has all kinds of odd fields, e.g. registers, and it doesn't make sense
to try to match on all of them. This commit changes learning-switch to
only try to match on the fields defined by OpenFlow 1.0. That's still not
minimal, but it's more reasonable.
This commit should not have an immediatel
Until now, ovs-controller and the learning-switch code split responsibility
for the OpenFlow connection. This commit moves all the responsibility into
the learning-switch code.
The rationale here is twofold. First, the split itself seems odd; I think
there must have been a reason for it at one t
This suppresses a long-standing warning from ovs-vswitchd about
non-normalized flows.
Signed-off-by: Ben Pfaff
---
lib/learning-switch.c |1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/lib/learning-switch.c b/lib/learning-switch.c
index e4287c4..c1cd909 100644
--- a/lib/
This fixes a crashing bug in ovs-controller reported by Simon Horman,
as well as a long-standing annoyance with lack of flow normalization.
Ben Pfaff (6):
vconn: Ensure that vconn_run() is enough to complete a connection.
vconn: Fix vconn_get_version().
learning-switch: Make lswitch own its
On 8/7/12 8:21 AM, mvp...@iol.pt wrote:
Can someone tell me if is in mind an cut-through implementation for
openvSwitch? I was thinking in do something like that.
AFAIK, NICs only trigger an interrupt after the packet has been
completely DMAed into memory, which makes cut-through impossible t
On Aug 7, 2012, at 1:50 PM, Ben Pfaff wrote:
> It's documented to return -1 if the version isn't yet known, but in fact
> it returned 0.
>
> Signed-off-by: Ben Pfaff
Looks good to me.
Thanks,
Kyle
___
dev mailing list
dev@openvswitch.org
http://openv
On Aug 7, 2012, at 1:50 PM, Ben Pfaff wrote:
> Until now, it seems that all vconn users have immediately started reading
> messages from the connection. Today, however, I added a new user that
> only wants to read packets after the OpenFlow version is negotiated, so
> it never called vconn_recv()
On Aug 7, 2012, at 1:50 PM, Ben Pfaff wrote:
> Until now, ovs-controller and the learning-switch code split responsibility
> for the OpenFlow connection. This commit moves all the responsibility into
> the learning-switch code.
>
> The rationale here is twofold. First, the split itself seems odd
On Aug 7, 2012, at 1:50 PM, Ben Pfaff wrote:
> The learning-switch implementation needs to know the OpenFlow version in
> use to send the initial handshake messages (e.g. the feature request), but
> the version is not always available at the time that the code currently
> sends the handshake. This
On Aug 7, 2012, at 1:50 PM, Ben Pfaff wrote:
> OVS has all kinds of odd fields, e.g. registers, and it doesn't make sense
> to try to match on all of them. This commit changes learning-switch to
> only try to match on the fields defined by OpenFlow 1.0. That's still not
> minimal, but it's more
On Aug 7, 2012, at 1:50 PM, Ben Pfaff wrote:
> This suppresses a long-standing warning from ovs-vswitchd about
> non-normalized flows.
>
> Signed-off-by: Ben Pfaff
Looks good to me.
Thanks,
Kyle
___
dev mailing list
dev@openvswitch.org
http://openvsw
I think you and I might have slightly different models in mind. The
model that I'm after at the moment is that the ofproto generic layer
checks that a set of actions is valid (which can be done without
knowing anything about the hardware) and then the implementation
checks that it can actually imp
On Tue, Aug 07, 2012 at 07:34:19PM +, Kyle Mestery (kmestery) wrote:
> On Aug 7, 2012, at 1:50 PM, Ben Pfaff wrote:
> > This suppresses a long-standing warning from ovs-vswitchd about
> > non-normalized flows.
> >
> > Signed-off-by: Ben Pfaff
>
>
> Looks good to me.
Thanks for the reviews.
I don't feel particularly strongly about it. The current
implementation is fine with me.
Ethan
On Tue, Aug 7, 2012 at 12:40 PM, Ben Pfaff wrote:
> I think you and I might have slightly different models in mind. The
> model that I'm after at the moment is that the ofproto generic layer
> checks
On Wed, Jul 25, 2012 at 09:05:58AM +0900, Simon Horman wrote:
> On Tue, Jul 24, 2012 at 04:03:59PM -0700, Ben Pfaff wrote:
> > On Mon, Jul 23, 2012 at 03:16:40PM +0900, Simon Horman wrote:
> > > This is a rather unpleasant hack but I am unsure of how to rework things
> > > such that ovs-controller
OK, thanks.
Perhaps when (if?) someone submits another ofproto provider to master,
we can figure out what best suits the set of providers as a whole.
On Tue, Aug 07, 2012 at 12:44:52PM -0700, Ethan Jackson wrote:
> I don't feel particularly strongly about it. The current
> implementation is fine
> Perhaps when (if?) someone submits another ofproto provider to master,
> we can figure out what best suits the set of providers as a whole.
Yah it's hard to get the abstraction right when we only have deep
knowledge of one implementation. Hopefully a significant open source
secondary implementa
On Tue, Aug 07, 2012 at 12:53:52PM -0700, Ethan Jackson wrote:
> > Perhaps when (if?) someone submits another ofproto provider to master,
> > we can figure out what best suits the set of providers as a whole.
>
> Yah it's hard to get the abstraction right when we only have deep
> knowledge of one
I think you might be right about which way is a little prettier, but
all this code gets deleted a few commits later so I don't think I'm
going to tweak the style.
Thanks,
Ben.
On Mon, Jul 30, 2012 at 03:12:37PM -0700, Ethan Jackson wrote:
> In flow_wildcards_combine(), does it make sense to move
On Mon, Jul 30, 2012 at 06:10:22PM -0700, Ethan Jackson wrote:
> In flow_wildcards_init_catchall() and flow_wildcards_init_exact() why
> not just memset? Perhaps the appropriate time to make that change
> would have been when in_port was changed to a mask come to think of
> it. It's fine to leave
Good point, thanks, I've now removed the functions that are generic.
On Mon, Jul 30, 2012 at 06:43:20PM -0700, Ethan Jackson wrote:
> Oh forgot to mention. We should probably remove the FLOW_WC_SEQ build
> assertions from a lot of these functions as they no longer need to
> change when new fields
Sounds good.
Ethan
On Tue, Aug 7, 2012 at 1:31 PM, Ben Pfaff wrote:
> I think you might be right about which way is a little prettier, but
> all this code gets deleted a few commits later so I don't think I'm
> going to tweak the style.
>
> Thanks,
>
> Ben.
>
> On Mon, Jul 30, 2012 at 03:12:37PM
Sounds reasonable.
Ethan
On Tue, Aug 7, 2012 at 1:41 PM, Ben Pfaff wrote:
> On Mon, Jul 30, 2012 at 06:10:22PM -0700, Ethan Jackson wrote:
>> In flow_wildcards_init_catchall() and flow_wildcards_init_exact() why
>> not just memset? Perhaps the appropriate time to make that change
>> would have
One more thought: it's pretty easy to get rid of flow_wildcards later
if we don't want it; it's harder to reintroduce it.
On Tue, Aug 07, 2012 at 01:45:04PM -0700, Ethan Jackson wrote:
> Sounds reasonable.
>
> Ethan
>
> On Tue, Aug 7, 2012 at 1:41 PM, Ben Pfaff wrote:
> > On Mon, Jul 30, 2012 a
> One more thought: it's pretty easy to get rid of flow_wildcards later
> if we don't want it; it's harder to reintroduce it.
Yep Yep, don't care that much either way, just wondered if you thought
about it. You obviously have, so lets leave it for now.
Ethan
>
> On Tue, Aug 07, 2012 at 01:45:04
On Tue, Jul 31, 2012 at 10:40:13AM -0700, Ethan Jackson wrote:
> Oh one more thing:
>
> This seems like the kind of thing that's both easy to get wrong, and
> easy to unit test. Does it make sense to add one?
Yes, I folded this in, thanks:
diff --git a/tests/library.at b/tests/library.at
index
Awesome, thanks!
Ethan
On Tue, Aug 7, 2012 at 2:06 PM, Ben Pfaff wrote:
> On Tue, Jul 31, 2012 at 10:40:13AM -0700, Ethan Jackson wrote:
>> Oh one more thing:
>>
>> This seems like the kind of thing that's both easy to get wrong, and
>> easy to unit test. Does it make sense to add one?
>
> Yes,
On Mon, Aug 06, 2012 at 02:29:30PM -0700, Ethan Jackson wrote:
> lib/match.c:
> s/clr_match/match
Thanks, fixed.
> match_set_metadata_masked()'s prototype is too long by a character.
Oops, I've broken the line now.
> lib/ofp-util.c:
> It's strange to me that ofputil_match_from_ofp10_m
> I dropped the parameter and return value from
> ofputil_match_from_ofp10_match(). After some thought, it seemed that
> I only needed the priority calculation in one place (in
> ofputil_decode_flow_mod()) so I inlined it there.
The other caller didn't need to calculate it?
Ethan
>
>> The inde
Modify ofpacts_put_openflow11_actions() to return the length of actions
appended. This will be used when encoding Packet Out messages for
Open Flow 1.1 and 1.2. The motivation for this is to avoid open coding
the size calculation which may end up being needed elsewhere too.
Signed-off-by: Simon Ho
Signed-off-by: Simon Horman
---
v10
* Remove non-test portions as they have already been merged.
Ben, did you intentionally omit the tests? If so, feel free
to drop this patch.
* Rename from "ofp-util: Allow encoding of Open Flow 1.2 Flow Mod Messages"
to "ofp-print: Open Flow 1.2 Flow Mod
Signed-off-by: Simon Horman
---
v10
* No change
v9
* No change
v8
* Manual Rebase
v7
* Manual Rebase
* Use struct ofp12_packet_in instead of struct ofp11_packet_in
v6
* No change
v5
* No change
v4
* No change
v3
* Add protocol parameter to ofputil_encode_packet_in().
This allows packet_
* Allow OXM matches which specified in OpenFlow 1.2.
Also allow them for OpenFlow 1.1 as there seems little reason not to.
* Pass padded_match_len parameter which if on NULL will be set to
the padded match len. This will be used when decoding flow statistics
response messages.
Signed-off-by
Hi Ben, Hi All,
this series adds infrastructure and message encoding and decoding
to allow OpenFlow 1.2 to be used to the extent that ovs-controller
works as a learning-switch and similar logic implemented using Ryu[1].
Note that some previously posted hacks are required in order for Open
vSwtich
Signed-off-by: Simon Horman
---
v10
* No change
v9
* No change
v8
* No change
v7
* Manual Rebase
* Only use OFPERR_NXBRC_BAD_IN_PORT in the case of OFPRAW_OFPT10_PACKET_OUT
v6
* No change
v5
* No change
v4
* No change
v3
* Differentiate versions using oh->version rather than relying
on
This is sufficient to allow encoding and decoding of
OpenFlow 1.2 Set Config messages as the format is the same
as OpenFlow 1.0 and OpenFlow 1.2.
Signed-off-by: Simon Horman
---
v10
* No change
v9
* No change
v8
* No change
v7
* Initial Post
---
lib/ofp-msgs.h | 2 +-
1 file changed, 1 inse
Signed-off-by: Simon Horman
---
v10
* No change
v9
* No change
v8
* Manual Rebase
* ofpacts_put_openflow11_actions() now returns the length of encoded
actions, make use of this.
v7
* Manual Rebase
v6
* No change
v5
* No change
v4
* Manual rebase
v3
* Correct title: this patch relates to
Signed-off-by: Simon Horman
---
v10
* No change
v9
* No change
v8
* Manual rebase
v7
* Omitted
v6
* No change
v5
* No change
v4
* Manual rebase
v3
* No change
v2
* Use ofputil_put_match()
Fix ofputil_encode_flow_stats_request
---
lib/ofp-util.c | 18 +-
1 file changed,
This enables the use of the OF12 format, the prerequisites of which were
added by "Add OFPUTIL_P_OF12 and NXFF_OPENFLOW12" and patches to fill out
other functions that use OFPUTIL_P_OF12 directly.
Signed-off-by: Simon Horman
---
v10
* No change
v9
* No change
v8
* Initial post
---
lib/ofp-ut
Signed-off-by: Simon Horman
---
v10
* No change
v9
* No change
v8
* Manual Rebase
v7
* Omitted
v6
* No change
v5
* No change
v4
* Manual rebase
v3
* No change
v2
* Use ofputil_put_match()
---
lib/ofp-msgs.h | 3 +++
lib/ofp-util.c | 21 -
2 files changed, 23 inserti
Signed-off-by: Simon Horman
---
v10
* No change
v9
* No change
v8
* Return error value returned by ofputil_port_from_ofp11()
rather than OFPERR_OFPBMC_BAD_VALUE.
- This fix, from v3, disappeared somewhere along the way
* Add ofp-print test for Open Flow 1.2
- I'm not sure how to generate
Signed-off-by: Simon Horman
---
v10
* No change
v9
* No change
v8
* Manual Rebase
v7
* Omitted
v6
* No change
v5
* No change
v4
* Manual rebase
v3
* No change
v2
* No change
---
lib/ofp-util.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/lib/ofp-util.c b/lib/o
Signed-off-by: Simon Horman
---
v10
* No change
v9
* No change
v8
* Manual rebase
* Add ofp-print test
v7
* Manual rebase
* Use struct ofp12_packet_in instead of struct ofp11_packet_in
v6
* No change
v5
* Manual rebase
v4
* No change
v3
* Add OF1.2 entry to to ofputil_msg_types
* Make OF1
Signed-off-by: Simon Horman
---
v10
* No change
v9
* Resolve breakage caused by reordering patches in v8
v8
* Make use of enum ofp_version
v7
* Manual Rebase
v6
* No change
v5
* No change
v4
* Manual rebase
v3
* Add protocol variable to do_probe().
Previously this was added by a separat
Signed-off-by: Simon Horman
---
v10
* No change
v9
* No change
v8
* Initial post
---
lib/ofp-msgs.h | 2 +-
tests/ofp-print.at | 14 ++
2 files changed, 15 insertions(+), 1 deletion(-)
diff --git a/lib/ofp-msgs.h b/lib/ofp-msgs.h
index 05ef334..0683bc5 100644
--- a/lib/ofp-m
Signed-off-by: Simon Horman
---
v10
* No change
v9
* No change
v8
* Initial post
---
lib/ofp-util.c | 35 +++
1 file changed, 27 insertions(+), 8 deletions(-)
diff --git a/lib/ofp-util.c b/lib/ofp-util.c
index 878d307..d5e8bbf 100644
--- a/lib/ofp-util.c
+++ b
Signed-off-by: Simon Horman
---
v10
* No change
v9
* No change
v8
* Manual Rebase
v7
* Omitted
v6
* No change
v5
* No change
v4
* Do not manually add header pad, it is handled by
ofputil_postappend_stats_reply()
v3
* Initial post
fix
---
lib/ofp-util.c | 23 ++-
1
This is intended for use with OFPRAW_OFPST_TABLE_REQUEST
in order for it to be symmetric with OpenFlow 1.0, 1.1 and 1.2
versions of OFPRAW_OFPST1TABLE_REPLY.
OpenFlow 1.3 introduces yet another format for OFPRAW_OFPST1TABLE_REPLY.
Signed-off-by: Simon Horman
---
v10
* No change
v9
* No change
Split OFPRAW_OFPST_TABLE_REPLY into OpenFlow 1.0, 1.1 and 1.2 versions.
This is preparation for allowing encoding and decoding of Open Flow 1.1
and 1.2 Table Stats Reply messages.
Signed-off-by: Simon Horman
---
v10
* No change
v9
* No change
v8
* Initial post
---
lib/ofp-msgs.h | 12 +
Allow decoding of Open Flow 1.1 and 1.2 flow and aggregate flow statistics
request messages.
Signed-off-by: Simon Horman
---
v10
* No change
v10
* Manual rebase
v9
* No change
v8
* Manual rebase
* Add ofp-print test
v7
* Omitted
v6
* No change
v5
* Manual rebase
v4
* Use OFPG11_ANY in pl
Split OFPRAW_OFPST_FLOW_{REQUEST,REPLY} into OpenFlow 1.0 and 1.1+
versions.
This is in preparation for adding encoding and decoding of
Open Flow 1.1 & 1.2 messages.
Signed-off-by: Simon Horman
---
v10
* No change
v9
* No change
v8
* Initial post
---
lib/ofp-msgs.h | 14 ++
lib/
Signed-off-by: Simon Horman
---
v10
* Manual rebase
v9
* No change
v8
* Manual rebase
* Add ofp-print test
v7
* Omitted
v6
* No change
v5
* Manual rebase
* Use padded_match_len parameter of ofputil_pull_ofp12_match()
to obtain the length of a match. Using the remaining
length of the mes
Split OFPRAW_OFPST_AGGREGATE_REQUEST into OpenFlow 1.0 and 1.1+
versions.
This should be sufficient to allow adding encoding and decoding of
Open Flow 1.1 and 1.2 Aggregate Stats Request messages.
Encoding and decoding of Open Flow 1.1 and 1.2 Aggregate Stats Response
messages works using the exi
Signed-off-by: Simon Horman
---
v10
* No change
v9
* Set wildcards, match, write_setfields and apply_setfields based
on a bitmap of (1 << OFPXMT_*)
v8
* Manual rebase
* Make use of enum ofp_version
* Add ofp-tests
v7
* Omitted
v6
* No change
v5
* Manual rebase
* Add OFPST_TABLE entry for
Signed-off-by: Simon Horman
---
v10
* No change
v9
* No change
v8
* Initial post
---
lib/ofp-msgs.h | 4 ++--
tests/ofp-print.at | 32 ++--
2 files changed, 32 insertions(+), 4 deletions(-)
diff --git a/lib/ofp-msgs.h b/lib/ofp-msgs.h
index 7ef7c5a..752d12c 1
Pass vconn to fetch_port_by_features() and callers.
In some cases this will reduce the number of connections
that ovs-ofputil sets up.
It should not alter the behaviour of ovs-ofputil.
Signed-off-by: Simon Horman
---
v10
* No change
v9
* Manual rebase
v8
* Omitted
v7
* Omitted
v6
* No chan
In order to form a stats message for the prevailing OpenFlow version
of a vconn the vconn needs to be open at the time the request is
encoded. Thus there is no longer a case where it makes sense to
use dump_stats_transaction() without a vconn already being open and
available to pass as a parameter.
Split OFPRAW_OFPST_PORT_{REQUEST,REPLY} into OpenFlow 1.0 and 1.1+
versions.
This prepares for encoding and decoding Open Flow 1.1 and 1.2 Port Stats
Request and Reply messages.
Signed-off-by: Simon Horman
---
v10
* No change
v9
* Update description - this only prepares for both encoding and
Open Flow 1.1 and 1.2 make use of 32 bit ports,
however Open vSwtich maps them to 16 bits. Make
ovs-ofputl aware of this.
Also, only accept ports that fit into 16 bits
for Open Flow 1.0.
Signed-off-by: Simon Horman
---
v10
* No change
v9
* Manual Rebase
v8
* Omitted
v7
* Omitted
v6
* No ch
Signed-off-by: Simon Horman
---
v10
* Generate OFPST_TABLE reply - OF1.2 test to save
rather a lot of lines in tests/ofp-print.at.
v9
* No change
v8
* Manual rebase
v8
* Manual rebase
* Make use of enum ofp_version
* Add ofp-print test
- Currently there is no Open Flow 1.1 test as I am un
Signed-off-by: Simon Horman
---
v10
* No change
v9
* Move print test to "ofp-print: Allow printing of Open Flow 1.1 & 1.2 Port
Reply Messages"
v8
* Manual rebase
* Make use of enum ofp_version
* Add ofp-print test
v7
* Omitted
v6
* No change
v5
* No change
v4
* Initial post
fix port tes
Signed-off-by: Simon Horman
---
v10
* No change
v9
* Manual Rebase
* Make use of enum ofp_version
v8
* Omitted
v7
* Omitted
v6
* No change
v5
* Initial Post
---
ofproto/ofproto.c | 27 ---
1 file changed, 24 insertions(+), 3 deletions(-)
diff --git a/ofproto/ofprot
Signed-off-by: Simon Horman
---
v10
* No change
v9
* Make use of enum ofp_version
* Add ofp-print test
* Correct printing of port number for Open Flow 1.0
* Replace UINT64_MAX with htonll(UINT64_MAX) in ofp_print_port_stat()
to correct byte-order problem detected by sparse
v8
* Omitted
v7
*
Signed-off-by: Simon Horman
---
v10
* No change
v9
* Initial post
---
utilities/ovs-ofctl.c | 20 +---
1 file changed, 17 insertions(+), 3 deletions(-)
diff --git a/utilities/ovs-ofctl.c b/utilities/ovs-ofctl.c
index 606a3f2..81df40c 100644
--- a/utilities/ovs-ofctl.c
+++ b/ut
Signed-off-by: Simon Horman
---
v10
* No change
v9
* Manual Rebase
* Make use of enum ofp_version
* Change title from "ofp-print: Allow display of Open Flow 1.1 & 1.2 Port Status
Request Messages" to "ofp-print: Allow printing of Open Flow 1.1 & 1.2 Port
Status
Request Messages"
* Add ofp-
Signed-off-by: Simon Horman
---
v10
* Manual rebase
* Make use of enum ofp_version
v9
* Omitted
v8
* Omitted
v7
* Omitted
v6
* No change
v5
* Initial Post
---
ofproto/ofproto.c | 40 ++--
1 file changed, 38 insertions(+), 2 deletions(-)
diff --git a/ofp
Split OFPRAW_OFPST_QUEUE_{REQUEST,REPLY} into OpenFlow 1.0 and 1.1+
versions.
This prepares for encoding and decoding Open Flow 1.1 and 1.2 Queue Stats
Request and Reply messages.
Signed-off-by: Simon Horman
---
v10
* No change
v10
* Initial Post
---
lib/ofp-msgs.h| 14 ++
Signed-off-by: Simon Horman
---
v10
* Make use of enum ofp_version
v9
* Omitted
v8
* Omitted
v7
* Omitted
v6
* No change
v5
* Initial Post
---
ofproto/ofproto.c | 33 +
1 file changed, 29 insertions(+), 4 deletions(-)
diff --git a/ofproto/ofproto.c b/ofprot
Signed-off-by: Simon Horman
---
v10
* Manual rebase
* Make use of enum ofp_version
v9
* Omitted
v8
* Omitted
v7
* Omitted
v6
* The queue_id variable in do_dump_aggregate() should be uint32_t not
uint16_t.
v5
* Initial Post
---
utilities/ovs-ofctl.c | 43 ++
Signed-off-by: Simon Horman
---
v10
* Make use of enum ofp_version
* Add ofp-print test
v9
* Omitted
v8
* Omitted
v7
* Omitted
v6
* No change
v5
* Initial Post
---
lib/ofp-print.c| 63 --
tests/ofp-print.at | 56 ++
Signed-off-by: Simon Horman
---
v10
* Manual rebase
* Make use of enum ofp_version
* Change subject name from "ovs-print: Enable display of Open Flow 1.1 &
1.2 Queue Stats Response" to "ovs-print: Enable display of Open Flow 1.1
& 1.2 Queue Stats Request"
* Add ofp-print test
* Use ofputil_p
On Tue, Aug 07, 2012 at 02:49:19PM -0700, Ethan Jackson wrote:
> > I dropped the parameter and return value from
> > ofputil_match_from_ofp10_match(). After some thought, it seemed that
> > I only needed the priority calculation in one place (in
> > ofputil_decode_flow_mod()) so I inlined it there
Makes sense. Thanks.
Ethan (iPhone)
On Aug 7, 2012, at 14:53, Ben Pfaff wrote:
> On Tue, Aug 07, 2012 at 02:49:19PM -0700, Ethan Jackson wrote:
>>> I dropped the parameter and return value from
>>> ofputil_match_from_ofp10_match(). After some thought, it seemed that
>>> I only needed the prior
Signed-off-by: Justin Pettit
---
lib/vconn.c |4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/lib/vconn.c b/lib/vconn.c
index cbe3c77..da4d9ca 100644
--- a/lib/vconn.c
+++ b/lib/vconn.c
@@ -138,10 +138,10 @@ vconn_usage(bool active, bool passive, bool bootstrap
OVS_UNU
On Tue, Aug 07, 2012 at 03:19:33PM -0700, Justin Pettit wrote:
> Signed-off-by: Justin Pettit
Looks good.
I didn't compare the output, but I'll assume you did.
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
Cabestan. Layout Doble
Asegúrate de no perderte ninguna oferta, añade ofer...@globalbono.com a tu
lista de contactos.
Si no ves correctamente las imágenes, pulsa [
http://email.globalbono.com/E07082012110854.cfm?WL=955&WS=208833_8707685&WA=399
] aquí
[ http://email.globalbono.com/Go/index.cfm
On Aug 7, 2012, at 3:21 PM, Ben Pfaff wrote:
> On Tue, Aug 07, 2012 at 03:19:33PM -0700, Justin Pettit wrote:
>> Signed-off-by: Justin Pettit
>
> Looks good.
>
> I didn't compare the output, but I'll assume you did.
Yep. Thanks. I pushed this.
--Justin
__
It's a good idea.
There's one case, in collect_rules_strict(), where it's nice to avoid
the repeated overhead of converting the same match to a cls_rule
repeatedly, so I just added a wrapper classifier_find_match_exactly()
and used that in the other cases.
Thanks,
Ben.
On Mon, Aug 06, 2012 at 0
Thank you. I applied this to master.
On Mon, Jul 30, 2012 at 02:16:37PM -0700, Mehak Mahajan wrote:
> Hey Ben,
>
> Looks good to me.
>
> thanx!
> mehak
>
> On Fri, Jul 27, 2012 at 1:23 PM, Ben Pfaff wrote:
>
> > Would someone review this please? It should not be hard.
> >
> > Thanks,
> >
>
Signed-off-by: Kyle Mestery
---
datapath/datapath.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/datapath/datapath.c b/datapath/datapath.c
index 7a7dc4c..e98c84b 100644
--- a/datapath/datapath.c
+++ b/datapath/datapath.c
@@ -61,8 +61,8 @@
#include "vport-internal_dev.h
A couple of calls to write() would generate warnings when the
"-Wunused-result" compiler option is enabled. This change ignores the
return value, since we can't do anything about it in logging code.
Signed-off-by: Justin Pettit
---
lib/vlog.c |4 ++--
1 files changed, 2 insertions(+), 2 del
On Tue, Aug 07, 2012 at 03:49:00PM -0700, Justin Pettit wrote:
> A couple of calls to write() would generate warnings when the
> "-Wunused-result" compiler option is enabled. This change ignores the
> return value, since we can't do anything about it in logging code.
>
> Signed-off-by: Justin Pet
On Aug 7, 2012, at 4:08 PM, Ben Pfaff wrote:
> On Tue, Aug 07, 2012 at 03:49:00PM -0700, Justin Pettit wrote:
>> A couple of calls to write() would generate warnings when the
>> "-Wunused-result" compiler option is enabled. This change ignores the
>> return value, since we can't do anything abou
64-bit glibc appears to avoid syscalls for clock_gettime(), so we can get
higher resolution timing and avoid having a timer firing off SIGALRM
without introducing extra overhead.
Signed-off-by: Leo Alterman
---
lib/timeval.c | 46 --
1 file changed,
>
> +/* On Linux IA64 systems, glibc avoids using syscalls for clock_gettime().
>
I assume you mean x86-64 or one of the equivalent names here, not IA64.
> +#ifndef __LP64__
> +#define CACHE_TIME 1
> +#endif
>
This will disable the syscall mitigation for 64-bit FreeBSD as well, which
is not (ye
On Tue, Aug 07, 2012 at 12:46:12PM -0700, Ben Pfaff wrote:
> On Wed, Jul 25, 2012 at 09:05:58AM +0900, Simon Horman wrote:
> > On Tue, Jul 24, 2012 at 04:03:59PM -0700, Ben Pfaff wrote:
> > > On Mon, Jul 23, 2012 at 03:16:40PM +0900, Simon Horman wrote:
> > > > This is a rather unpleasant hack but
On Wed, Aug 08, 2012 at 06:49:35AM +0900, Simon Horman wrote:
> Hi Ben, Hi All,
>
> this series adds infrastructure and message encoding and decoding
> to allow OpenFlow 1.2 to be used to the extent that ovs-controller
> works as a learning-switch and similar logic implemented using Ryu[1].
> Note
Signed-off-by: Simon Horman
---
v11
* Manual rebase
v10
* No change
v9
* Resolve breakage caused by reordering patches in v8
v8
* Make use of enum ofp_version
v7
* Manual Rebase
v6
* No change
v5
* No change
v4
* Manual rebase
v3
* Add protocol variable to do_probe().
Previously this w
The original developer of the BSD port (Gaetano Catalli) also implemented a
threaded version of the userspace datapath to increase performance. As
with the original port, the threading work was done against Open vSwitch
1.1.0pre2. Giuseppe Lettieri and I have ported it forward, and it's
functiona
97 matches
Mail list logo
