On Thu, Oct 24, 2013 at 12:07:44PM +0900, YAMAMOTO Takashi wrote:
> an option question: what to do for the following messages with the
> same xid? especially for the last message which doesn't have the
> "more" flag set? it would be neater to detect and drop them.
That's a great question. I mig
On Thu, Oct 24, 2013 at 12:07:43PM +0900, YAMAMOTO Takashi wrote:
> the correct error in that case is OFPERR_OFPBRC_BAD_STAT,
> not OFPERR_OFPBRC_BAD_TYPE.
>
> currently, the only example of unsupported stats/multipart request is
> OFPTYPE_TABLE_FEATURES_STATS_REQUEST.
>
> Signed-off-by: YAMAMOTO
On Thu, Oct 24, 2013 at 12:08:18PM +0900, YAMAMOTO Takashi wrote:
> > On Wed, Oct 23, 2013 at 03:04:21PM +0900, YAMAMOTO Takashi wrote:
> >> the correct error in that case is OFPERR_OFPBRC_BAD_STAT,
> >> not OFPERR_OFPBRC_BAD_TYPE.
> >>
> >> Signed-off-by: YAMAMOTO Takashi
> >
> > If there exist
On Thu, Oct 24, 2013 at 06:27:10PM -0700, Ansis Atteka wrote:
> It is more obvious to figure out what exactly test is doing
> if ovs-vsctl commands contain full table names.
>
> Signed-off-by: Ansis Atteka
Acked-by: Ben Pfaff
___
dev mailing list
dev@
It is more obvious to figure out what exactly test is doing
if ovs-vsctl commands contain full table names.
Signed-off-by: Ansis Atteka
---
tests/ovs-vsctl.at | 70 +++---
1 file changed, 35 insertions(+), 35 deletions(-)
diff --git a/tests/ovs-vs
On Thu, Oct 24, 2013 at 3:39 PM, Romain Lenglet wrote:
> - Original Message -
>> From: "Jesse Gross"
>> To: "Romain Lenglet"
>> Cc: dev@openvswitch.org
>> Sent: Tuesday, October 22, 2013 3:46:54 PM
>> Subject: Re: [ovs-dev] [PATCH] sFlow export: include standard tunnel
>> structures (fo
On Thu, Oct 24, 2013 at 9:01 AM, Jarno Rajahalme wrote:
> tcp_flags=flags/mask
> Bitwise match on TCP flags. The flags and mask are 16-bit num‐
> bers written in decimal or in hexadecimal prefixed by 0x. Each
> 1-bit in mask requires that the corresponding bit in
On Thu, Oct 24, 2013 at 9:01 AM, Jarno Rajahalme wrote:
> Widen TCP flags handling from 7 bits to 12 bits. The kernel interface
> remains at 8 bits, which makes no functional difference now, as none
> of the higher bits is currently of interest to the userspace.
>
> Signed-off-by: Jarno Rajahalme
On Thu, Oct 17, 2013 at 02:20:33PM -0700, Ben Pfaff wrote:
> On Thu, Oct 17, 2013 at 09:35:27AM -0700, Ben Pfaff wrote:
> > From: Venkitachalam Gopalakrishnan
>
> Thanks gops.
>
> I'm spending a little bit of time working on this. I'll probably post a
> new version tomorrow or Monday.
I send o
From: Venkitachalam Gopalakrishnan
Open vSwitch has never implemented this request and reply, even though they
have been in OpenFlow since version 1.0. This commit adds an
implementation.
Co-authored-by: Venkitachalam Gopalakrishnan
Signed-off-by: Ben Pfaff
---
include/openflow/openflow-1.0.
Message types are defined in ofp-msgs.h these days.
Signed-off-by: Ben Pfaff
---
include/openflow/openflow-1.2.h |9 -
1 file changed, 9 deletions(-)
diff --git a/include/openflow/openflow-1.2.h b/include/openflow/openflow-1.2.h
index 541b143..249e861 100644
--- a/include/openflow/o
- Original Message -
> From: "Jesse Gross"
> To: "Romain Lenglet"
> Cc: dev@openvswitch.org
> Sent: Tuesday, October 22, 2013 3:46:54 PM
> Subject: Re: [ovs-dev] [PATCH] sFlow export: include standard tunnel
> structures (for GRE, VXLAN etc.)
>
> On Mon, Oct 21, 2013 at 2:33 PM, Romain
This looks good to me.
I verified that the Fedora rpm gets built fine after applying your patch.
I applied it to master.
Thanks.
On Thu, Oct 24, 2013 at 12:35 PM, Kyle Mestery wrote:
> Add the appropriate vtep files into the Fedora RPM build
> to make this work on master again.
>
> Signed-off-b
On Oct 21, 2013, at 3:52 PM, Ben Pfaff wrote:
> Commit e3b5693319c (Fix table checking for goto table instruction.) moved
> action checking into modify_flows__(), for good reason, but as a side
> effect made modify_flows__() abandon and never commit the ofopgroup that it
> started, if action ch
On Thu, Oct 24, 2013 at 12:59 PM, Kyle Mestery (kmestery)
wrote:
> I'm seeing the following error on *one* of my Fedora test VMs.
> It's strange because on the other two, I don't see the GRE errors
> below. Any ideas as to what might be going wrong?
>
> [kmestery@km-dhcp-64-197 ~]$ dmesg|grep gre
Signed-off-by: Jarno Rajahalme
---
lib/ofp-actions.c | 141
lib/ofp-actions.h |3 +-
tests/ovs-ofctl.at | 12 -
3 files changed, 153 insertions(+), 3 deletions(-)
diff --git a/lib/ofp-actions.c b/lib/ofp-actions.c
index 09a98bf..e
OpenFlow 1.1 set vlan actions only modify existing vlan
headers, while OF 1.0 actions push a new vlan header if one
does not exist already.
Signed-off-by: Jarno Rajahalme
---
lib/ofp-actions.c| 91 +-
lib/ofp-actions.h| 20 +
This makes later changes simpler.
Signed-off-by: Jarno Rajahalme
---
lib/ofp-actions.c | 274 +++--
lib/ofp-actions.h | 32 +++---
lib/ofp-util.c| 127 +--
lib/ofp-util.h|1 +
utilities/ovs-ofctl.c |
Signed-off-by: Jarno Rajahalme
---
v3: Fix byte order errors.
include/openflow/nicira-ext.h | 24
lib/ofp-actions.c | 82 +++--
lib/ofp-actions.h | 22 ++-
lib/ofp-parse.c | 46
Output set field actions as standard OF1.0/1.1 set actions or to
reg_load instructions, when a compatible set action(s) do not exist.
Signed-off-by: Jarno Rajahalme
---
v4: Clear the OFPVID_PRESENT bit when translating Set-Field
OXM_OF_VLAN_VID to OpenFlow 1.0/1.1.
lib/ofp-actions.c | 211
Signed-off-by: Jarno Rajahalme
---
lib/ofp-actions.c | 18 ++
1 file changed, 18 insertions(+)
diff --git a/lib/ofp-actions.c b/lib/ofp-actions.c
index eef4c16..4713371 100644
--- a/lib/ofp-actions.c
+++ b/lib/ofp-actions.c
@@ -1105,6 +1105,24 @@ ofpact_from_openflow11(const un
Store the original action code with the strip vlan action,
so that it can be printed back properly.
Signed-off-by: Jarno Rajahalme
---
lib/ofp-actions.c|7 ---
lib/ofp-parse.c |2 +-
tests/ofp-actions.at |2 +-
tests/ovs-ofctl.at |4 ++--
4 files changed, 8 inserti
Signed-off-by: Jarno Rajahalme
---
v4:
- Address review comments by Ben
- Simplify implementation
- Enforce presence of the OFPVID_PRESENT bit for Set-Field on OXM_OF_VLAN_VID.
- Add OFPVID_PRESENT when parsing from string, and omit the bit when printing.
- Explicitly check for header presence bef
Adds OXM inspired aliases for match fields that don't have them
already ("ip_proto", "ip_ecn", "ip_dscp", and "tunnel_id").
"ip_dscp" replaces the earlier undocumented "nw_tos_shifted",
and takes the DSCP value (0-63), which is then shifted
appropriately when applied to an IP packet.
The number of
mf_from_id accesses a static table, so the compiler should be able to
completely optimize it away.
Also use OVS_PACKED_ENUM to waste less space.
Signed-off-by: Jarno Rajahalme
---
lib/meta-flow.c | 12 +++-
lib/meta-flow.h | 18 +-
2 files changed, 16 insertions(+),
This series adds many of the missing OpenFlow 1.1 actions.
Due to the different semantics, the vlan actions are annotated
based on the policy on missing vlan header.
A native OF 1.2 Set Field action with translation to/from
OF 1.0/1.1 actions is added.
Action consistency is enforced for OF 1.1+,
I'm seeing the following error on *one* of my Fedora test VMs.
It's strange because on the other two, I don't see the GRE errors
below. Any ideas as to what might be going wrong?
[kmestery@km-dhcp-64-197 ~]$ dmesg|grep gre
[0.175404] ACPI: Added _OSI(Processor Aggregator Device)
[8.365982]
Add the appropriate vtep files into the Fedora RPM build
to make this work on master again.
Signed-off-by: Kyle Mestery
---
rhel/openvswitch-fedora.spec.in | 7 +++
1 file changed, 7 insertions(+)
diff --git a/rhel/openvswitch-fedora.spec.in b/rhel/openvswitch-fedora.spec.in
index 27a3b03..
Ben,
Looks mostly good (and a helpful improvement on the prior documentation). One
small error:
+ and the rate at which it is willing to transmit them. Open vSwitch
uses a detection multiplier of three, meaning that an endpoint
+ signals a connectivity fault if it fails to r
On Oct 23, 2013, at 9:14 PM, Ben Pfaff wrote:
> On Wed, Oct 23, 2013 at 12:53:01PM -0700, Jarno Rajahalme wrote:
>> Signed-off-by: Jarno Rajahalme
>
> Is MFF_VLAN_VID the only field for which the "match" and "set"
> prerequisites differ? If so, then I'd like to think of some way to
> avoid ha
On Wed, Oct 23, 2013 at 11:44 PM, Takayuki Usui wrote:
> Hi Jesse,
>
> Thank you for your kind review and comments. As you mentioned,
> this experimental work is, more or less, a result of the effort
> to solve a specific problem of ours (quickly) and it's highly
> depends on the use case for sure
Ben,
OpenFlow 1.3.3 is much more specific about handling the VID with
set-field action. It states that the OFPVID_PRESENT bit must be on,
and that the switch behavior when VLAN header does not exist is
implementation-dependent:
"The use of a set-fi eld action assumes that the corresponding
header
On Thu, Oct 24, 2013 at 09:49:33AM -0700, Ben Pfaff wrote:
> This update improves the BFD documentation in a few ways:
>
>- Demand mode is now supported.
>
>- Wordsmithing, spelling, etc.
>
>- Attempt to better explain decay_min_rx, forwarding_if_rx, and
> cpath_down.
>
>-
I don't think you should put in that work yet, I'm thinking it over and
wouldn't want you to spend time on it that might be wasted.
On Thu, Oct 24, 2013 at 09:10:06AM -0700, Jarno Rajahalme wrote:
> Ben,
>
> While I'm reworking on the series, I could add the separate actions for the
> OF 1.1 set
On Thu, Oct 24, 2013 at 09:53:19AM -0700, Ben Pfaff wrote:
> On Thu, Oct 24, 2013 at 08:30:51AM -0700, Bruce Davie wrote:
> > A number of new key-value pairs have been added to the bfd and bfd_status
> > columns of the OVS schema since the VTEP schema was created. To aid
> > interoperability betwee
On Thu, Oct 24, 2013 at 08:30:51AM -0700, Bruce Davie wrote:
> A number of new key-value pairs have been added to the bfd and bfd_status
> columns of the OVS schema since the VTEP schema was created. To aid
> interoperability between OVS instances and VTEPs, this patch brings
> the VTEP schema into
This update improves the BFD documentation in a few ways:
- Demand mode is now supported.
- Wordsmithing, spelling, etc.
- Attempt to better explain decay_min_rx, forwarding_if_rx, and
cpath_down.
- Break into subgroups for configuration and status, to better explain
which
On Oct 23, 2013, at 6:27 PM, Jesse Gross wrote:
> On Wed, Oct 23, 2013 at 10:24 AM, Jarno Rajahalme
> wrote:
>> Widen TCP flags handling from 7 bits (uint8_t) to 12 bits (uint16_t).
>> The kernel interface remains at 8 bits, which makes no functional
>> difference now, as none of the higher bit
Ben,
While I'm reworking on the series, I could add the separate actions for the OF
1.1 set VLAN
VID/PCP actions if you think it will make the code easier to read and maintain.
If we do this, we should think about separating the PUSH MPLS actions as well.
Please let me know,
Jarno
On Oct 23
Widen TCP flags handling from 7 bits to 12 bits. The kernel interface
remains at 8 bits, which makes no functional difference now, as none
of the higher bits is currently of interest to the userspace.
Signed-off-by: Jarno Rajahalme
---
v3-v6: Rebase
v2: Address Ben's review comments:
- datapath
A number of new key-value pairs have been added to the bfd and bfd_status
columns of the OVS schema since the VTEP schema was created. To aid
interoperability between OVS instances and VTEPs, this patch brings
the VTEP schema into line with that of OVS.
Signed-off-by: Bruce Davie
---
vtep/vtep.x
41 matches
Mail list logo
