Very good. Is that possible to use mptcp with openvswitch in top of it?
For example openvswitch able to aggregate two eth links to bond
interface with lacp : active-passive, but not able to create 802.3ad
aggregation.
2013/7/12 Christoph Paasch :
> Hello,
>
> there are two weeks left until the nex
When "other-config:enable-vlan-splinters=true" is set, the existing
vlans with ip address must be retained. The bug actually does the
opposite and retains the vlans without ip address. This commit fixes
it.
Reported-by: Roman Sokolkov
Signed-off-by: Alex Wang
---
vswitchd/bridge.c |4 ++--
--
Oferim creditele acordate persoanelor fizice, firme și să coopereze
organelor de 2%
Rata de retrage o sumă minimă puteți împrumuta este
$ 2,000.00 dolari SUA pentru a oferi un maxim de $ 10.Million.Please
în urma:
Http {MR., Pr.., MS., DR, etc.}
1) Numele tău ...
2) țara dumnea
This commit fixes the warning issued by 'clang' when pointer is casted
to one with greater alignment.
Signed-off-by: Alex Wang
---
v1 -> v2:
1. use ofpbuf_at() instead of creating ofpbuf_forward_ref()
2. remove the unnecessary parenthesis in argument to ALIGNED_CAST()
---
lib/hash.c
On Mon, Jul 22, 2013 at 3:08 PM, Andy Zhou wrote:
> After a mask is assigned to a flow, it will not change for the life of
> the flow. Since flow access is protected by RCU lock, access to
> flow->mask after getting a flow is always safe.
>
> Suggested-by: Jesse Gross
> Reported-by: Ben Pfaff
>
Done.
On Mon, Jul 22, 2013 at 2:53 PM, Jesse Gross wrote:
> OK, although the patch also needs a signed-off-by from you and I think
> that we can also remove an ugly cast from ovs_flow_free(). Do you just
> want to send out a new version?
>
> On Mon, Jul 22, 2013 at 2:35 PM, Andy Zhou wrote:
>
After a mask is assigned to a flow, it will not change for the life of
the flow. Since flow access is protected by RCU lock, access to
flow->mask after getting a flow is always safe.
Suggested-by: Jesse Gross
Reported-by: Ben Pfaff
Signed-off-by: Andy Zhou
V1 - V2: remove cast from ovs_fl
OK, although the patch also needs a signed-off-by from you and I think
that we can also remove an ugly cast from ovs_flow_free(). Do you just
want to send out a new version?
On Mon, Jul 22, 2013 at 2:35 PM, Andy Zhou wrote:
> Sorry forgot to mention Ben reported this issue. Jesse, would you pleas
Sorry forgot to mention Ben reported this issue. Jesse, would you please
add it before committing?
On Mon, Jul 22, 2013 at 2:30 PM, Ben Pfaff wrote:
> On Mon, Jul 22, 2013 at 02:28:04PM -0700, Andy Zhou wrote:
> > After a mask is assigned to a flow, it will not change for the life of
> > the fl
On Mon, Jul 22, 2013 at 02:28:04PM -0700, Andy Zhou wrote:
> After a mask is assigned to a flow, it will not change for the life of
> the flow. Since flow access is protected by RCU lock, access to
> flow->mask after getting a flow is always safe.
>
> Suggested-by: Jesse Gross
This patch fixes t
After a mask is assigned to a flow, it will not change for the life of
the flow. Since flow access is protected by RCU lock, access to
flow->mask after getting a flow is always safe.
Suggested-by: Jesse Gross
---
datapath/datapath.c |6 ++
datapath/flow.c |5 ++---
datapath/flow.
Sure, will send out a patch soon.
On Mon, Jul 22, 2013 at 12:44 PM, Jesse Gross wrote:
> On Mon, Jul 22, 2013 at 10:58 AM, Ben Pfaff wrote:
> > I'm getting the following warning on current master (commit
> > dc0d542d5254c1e6 "datapath: Conditionally define skb_unclone in
> > datapath compat co
Thanks, I applied this version.
On Mon, Jul 22, 2013 at 01:55:34PM -0700, Andy Zhou wrote:
> Looks good. It works now in my local build. Thanks.
>
>
> On Mon, Jul 22, 2013 at 1:31 PM, Ben Pfaff wrote:
>
> > Reported-by: Andy Zhou
> > Signed-off-by: Ben Pfaff
> > ---
> > v1->v2: Avoid conflic
Looks good. It works now in my local build. Thanks.
On Mon, Jul 22, 2013 at 1:31 PM, Ben Pfaff wrote:
> Reported-by: Andy Zhou
> Signed-off-by: Ben Pfaff
> ---
> v1->v2: Avoid conflict with ovs-thread.h definitions.
>
> include/sparse/pthread.h | 11 ++-
> 1 files changed, 10 inser
Applied, thank you.
On Fri, Jul 19, 2013 at 03:52:01PM -0700, Ethan Jackson wrote:
> Acked-by: Ethan Jackson
>
>
> On Fri, Jul 19, 2013 at 10:57 AM, Ben Pfaff wrote:
> > The build is clean against -Wextra at least with GCC 4.4.
> >
> > (-Wextra includes -Woverride-init.)
> >
> > Suggested-by:
Thanks Ben, for the review,
> Signed-off-by: Alex Wang
>
> Could you use the existing ofpbuf_at() instead of a new
> ofpbuf_forward_ref()? If there's no appropriate 'size' in a given
> call then you can use 0.
>
> It shouldn't be necessary to cast the return value of
> ofpbuf_forward_ref() or o
Reported-by: Andy Zhou
Signed-off-by: Ben Pfaff
---
v1->v2: Avoid conflict with ovs-thread.h definitions.
include/sparse/pthread.h | 11 ++-
1 files changed, 10 insertions(+), 1 deletions(-)
diff --git a/include/sparse/pthread.h b/include/sparse/pthread.h
index 723c351..6cdf5c8 10064
On Mon, Jul 22, 2013 at 09:19:58AM -0700, Alex Wang wrote:
> This commit fixes the warning issued by 'clang' when pointer is casted
> to one with greater alignment.
>
> Signed-off-by: Alex Wang
Could you use the existing ofpbuf_at() instead of a new
ofpbuf_forward_ref()? If there's no appropria
Thanks, that makes sense.
On Mon, Jul 22, 2013 at 12:51 PM, Andy Zhou wrote:
> The actual case I found this bug is more complicated to explain, but the
> following made up example can illustrate the problem. Consider we have 2
> kernel mega flows:
>
> 1) in_port(1/0x), *
> 2) in_port(2/0xf
Signed-off-by: Ethan Jackson
---
ofproto/ofproto-dpif-xlate.c |5 -
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/ofproto/ofproto-dpif-xlate.c b/ofproto/ofproto-dpif-xlate.c
index eb4ed69..e555603 100644
--- a/ofproto/ofproto-dpif-xlate.c
+++ b/ofproto/ofproto-dpif-xlate.c
The actual case I found this bug is more complicated to explain, but the
following made up example can illustrate the problem. Consider we have 2
kernel mega flows:
1) in_port(1/0x), *
2) in_port(2/0x), src_ip(1.1.1.1/255.255.255.255)
They are not overlapping flows, however, when the
On Mon, Jul 22, 2013 at 12:44:47PM -0700, Alex Wang wrote:
> I folded your changes in, and there is no warning/error,
> when compiling with 'clang' and 'gcc'.
>
> Unit tests all passed,
>
> Thanks for your review, do I still need to post a v2 patch?
No, that's fine, I'll fold it in myself.
Than
ovs-appctl bond/show is being run through the plugin by
calling the script ovs-bugtool-bond-show.
So remove the redundant code.
Signed-off-by: Gurucharan Shetty
---
utilities/bugtool/ovs-bugtool.in | 22 --
1 file changed, 22 deletions(-)
diff --git a/utilities/bugtool/ov
Currently the openvswitch database is being collected with
CAP_NETWORK_CONFIG which has a max size of 50 KB. This is
quite low as the database can easily be larger than 50 KB.
Move database collection to CAP_NETWORK_STATUS which does
not have a max size. Also, compact the database before picking
i
Current size feels very low when we are collecting o/p of
'dmidecode' and 'lspci -vv'
Signed-off-by: Gurucharan Shetty
---
utilities/bugtool/ovs-bugtool.in |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/utilities/bugtool/ovs-bugtool.in b/utilities/bugtool/ovs-bugtool.in
i
Reported-by: Andy Zhou
Signed-off-by: Ben Pfaff
---
include/sparse/pthread.h |8 +++-
1 files changed, 7 insertions(+), 1 deletions(-)
diff --git a/include/sparse/pthread.h b/include/sparse/pthread.h
index 723c351..1a025bf 100644
--- a/include/sparse/pthread.h
+++ b/include/sparse/pthre
On Mon, Jul 22, 2013 at 10:58 AM, Ben Pfaff wrote:
> I'm getting the following warning on current master (commit
> dc0d542d5254c1e6 "datapath: Conditionally define skb_unclone in
> datapath compat code"):
>
> datapath/flow.c:1051:32: error: incompatible types in comparison
> expression (di
I folded your changes in, and there is no warning/error,
when compiling with 'clang' and 'gcc'.
Unit tests all passed,
Thanks for your review, do I still need to post a v2 patch?
On Mon, Jul 22, 2013 at 12:36 PM, Alex Wang wrote:
> I see, I didn't compile use sparse.
>
> My gcc is: gcc (Ubunt
The previously defined config files were never included in
the debug bundle. This will include them.
Also increase the max size for CAP_NETWORK_CONFIG to 5 MB.
A pre-compressed size of 5 MB does not amount to much after
compression for config files.
Signed-off-by: Gurucharan Shetty
---
utilitie
Current situation is that CAP_NETWORK_STATUS has a max size of 50 MB.
When we have around 100,000 openflow flows, we over-run that size
by just running the "ovs-ofctl dump-flows" command. All the openvswitch
commands run through the plugin scripts in this repo won't have its
data stored in the debu
On Mon, Jul 22, 2013 at 09:10:52AM -0700, Alex Wang wrote:
> Looks good to me,
>
> Want to ask question about "ovsthread_once_start()",
> 1. why do we have "ovsthread_once_start()" defined in header file?
The "fast path" for using ovsthread_once_start() is just a test and
conditional jump. The a
I see, I didn't compile use sparse.
My gcc is: gcc (Ubuntu/Linaro 4.6.3-1ubuntu5) 4.6.3
And it didn't issue any error, when sparse is not enabled.
I should always do that,.
On Mon, Jul 22, 2013 at 12:22 PM, Ben Pfaff wrote:
> I didn't do anything special to get the warning. With GCC 4.4.5 an
On Fri, Jul 19, 2013 at 11:11 AM, Andy Zhou wrote:
> A mega flow matches when the masked key matches and the mask applied
> is the same as the mask used to create the mega flow.
>
> This patch adds the implementation of the second match condition
> mentioned above. Without this fix, mega flow look
I didn't do anything special to get the warning. With GCC 4.4.5 and
this patch applied, I see:
../ofproto/ofproto.c:5037:5: error: incompatible types in conditional
expression (different base types)
../ofproto/ofproto.c:5040:9: error: incompatible types in conditional
expression (differ
It seems that 'clang' compiler applies strict protection on pointer
dereference. And it causes unexpected execution in macro functions
like "HMAP_FOR_EACH()" and unit test failures. This commit fixes
this issue and pass all unit tests.
Co-authored-by: Ethan Jackson
Signed-off-by: Alex Wang
---
This commit makes macro function "ASSIGN_CONTAINER()" evaluates
to "(void)0". This is to avoid the 'clang' warning: "expression
result unused", since most of time, the final evaluated value
is not used.
Signed-off-by: Alex Wang
---
lib/classifier.h |2 +-
lib/hindex.h |2 +-
lib/hmap
Hey Ben,
Also want to ask how do you get the gcc warning? I compiled with '-O3' and
didn't see it.
Thanks,
On Mon, Jul 22, 2013 at 11:35 AM, Alex Wang wrote:
> Thanks Ben,
>
> I'll recheck that.
>
>
> On Mon, Jul 22, 2013 at 11:30 AM, Ben Pfaff wrote:
>
>> On Mon, Jul 22, 2013 at 09:19:57AM -
Thanks Ben,
I'll recheck that.
On Mon, Jul 22, 2013 at 11:30 AM, Ben Pfaff wrote:
> On Mon, Jul 22, 2013 at 09:19:57AM -0700, Alex Wang wrote:
> > This commit makes macro function "ASSIGN_CONTAINER()" evaluates
> > to "(void)0". This is to avoid the 'clang' warning: "expression
> > result unus
On Mon, Jul 22, 2013 at 11:13 AM, Ben Pfaff wrote:
> Jesse, I'm leaving this series alone until you give some feedback on
> the kernel side of things. Then I'm happy to review (and commit when
> appropriate) user space patches.
OK, I'll look at the last patch first then.
X-CudaMail-Whitelist-To:
On Mon, Jul 22, 2013 at 09:19:56AM -0700, Alex Wang wrote:
> It seems that 'clang' compiler applies strict protection on pointer
> dereference. And it causes unexpected execution in macro functions
> like "HMAP_FOR_EACH()" and unit test failures. This commit fixes
> this issue and pass all unit tes
On Mon, Jul 22, 2013 at 09:19:57AM -0700, Alex Wang wrote:
> This commit makes macro function "ASSIGN_CONTAINER()" evaluates
> to "(void)0". This is to avoid the 'clang' warning: "expression
> result unused", since most of time, the final evaluated value
> is not used.
>
> Signed-off-by: Alex Wang
Recent versions of Fedora have skb_unclone included in their
kernels. This patch adds a conditional check into the compat directory so as
not to error out by defining it twice. This allows the latest OVS kernel
module to build on Fedora 19.
Signed-off-by: Kyle Mestery
---
acinclude.m4
Jesse, I'm leaving this series alone until you give some feedback on
the kernel side of things. Then I'm happy to review (and commit when
appropriate) user space patches.
Thanks,
Ben.
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mail
I'm getting the following warning on current master (commit
dc0d542d5254c1e6 "datapath: Conditionally define skb_unclone in
datapath compat code"):
datapath/flow.c:1051:32: error: incompatible types in comparison
expression (different address spaces)
That's in ovs_masked_flow_lookup():
Thanks for the reviews. I applied these to master.
On Mon, Jul 22, 2013 at 10:51:43AM -0700, Alex Wang wrote:
> Looks good to me, thanks,
>
> Review for patch 1/23~10/23 is complete.
>
>
> On Thu, Jul 18, 2013 at 4:15 PM, Ben Pfaff wrote:
>
> > Signed-off-by: Ben Pfaff
> > ---
> > lib/ovsd
Looks good to me, thanks,
Review for patch 1/23~10/23 is complete.
On Thu, Jul 18, 2013 at 4:15 PM, Ben Pfaff wrote:
> Signed-off-by: Ben Pfaff
> ---
> lib/ovsdb-data.c |7 ---
> 1 files changed, 4 insertions(+), 3 deletions(-)
>
> diff --git a/lib/ovsdb-data.c b/lib/ovsdb-data.c
> i
Looks good to me, thanks,
On Thu, Jul 18, 2013 at 4:15 PM, Ben Pfaff wrote:
> Signed-off-by: Ben Pfaff
> ---
> lib/dpif-linux.c |7 +--
> lib/dpif.c |6 +++---
> 2 files changed, 8 insertions(+), 5 deletions(-)
>
> diff --git a/lib/dpif-linux.c b/lib/dpif-linux.c
> index 958
Oh! How could I not see the "inline". Sorry for the nonsense,
On Mon, Jul 22, 2013 at 10:39 AM, Ben Pfaff wrote:
> On Mon, Jul 22, 2013 at 10:35:27AM -0700, Alex Wang wrote:
> > On Mon, Jul 22, 2013 at 10:23 AM, Ben Pfaff wrote:
> >
> > > On Mon, Jul 22, 2013 at 09:10:52AM -0700, Alex Wang wro
Yes.
On Mon, Jul 22, 2013 at 10:25:09AM -0700, Alex Wang wrote:
> Seems I understand now,
>
> Is that for directing all error messages to corresponding lines in
> "lib/dirs.c.in"?
>
>
> On Mon, Jul 22, 2013 at 10:20 AM, Alex Wang wrote:
>
> > Looks good to me,
> >
> > One question, why do we
On Mon, Jul 22, 2013 at 10:35:27AM -0700, Alex Wang wrote:
> On Mon, Jul 22, 2013 at 10:23 AM, Ben Pfaff wrote:
>
> > On Mon, Jul 22, 2013 at 09:10:52AM -0700, Alex Wang wrote:
> > > Looks good to me,
> > >
> > > Want to ask question about "ovsthread_once_start()",
> > > 1. why do we have "ovsthr
Thanks Ben for the explanation,
On Mon, Jul 22, 2013 at 10:37 AM, Ben Pfaff wrote:
> Yes.
>
> On Mon, Jul 22, 2013 at 10:25:09AM -0700, Alex Wang wrote:
> > Seems I understand now,
> >
> > Is that for directing all error messages to corresponding lines in
> > "lib/dirs.c.in"?
> >
> >
> > On Mon
On Mon, Jul 22, 2013 at 10:20:54AM -0700, Alex Wang wrote:
> One question, why do we have the "#line directive" in "lib/dirs.c.in"?
> We have already created "dirs.c" in "lib/automake.mk". What is the
> use of "#line directive" here?
dirs.c gets generated by search-and-replace on dirs.c.in. If th
Thanks Ben for the explanation,
On Mon, Jul 22, 2013 at 10:23 AM, Ben Pfaff wrote:
> On Mon, Jul 22, 2013 at 09:10:52AM -0700, Alex Wang wrote:
> > Looks good to me,
> >
> > Want to ask question about "ovsthread_once_start()",
> > 1. why do we have "ovsthread_once_start()" defined in header fil
On Mon, Jul 22, 2013 at 7:38 AM, Kyle Mestery wrote:
> Recent versions of Fedora have skb_unclone included in their
> kernels. This patch adds a conditional check into the compat directory so as
> not to error out by defining it twice. This allows the latest OVS kernel
> module to build on Fedora
Thanks, it makes sense,
On Mon, Jul 22, 2013 at 10:26 AM, Ben Pfaff wrote:
> On Mon, Jul 22, 2013 at 09:17:37AM -0700, Alex Wang wrote:
> > Also, for confirmation, we do not need to guarantee the thread safety
> > of coverage counters, since they are not meant to be perfectly accurate,
> > righ
On Mon, Jul 22, 2013 at 09:17:37AM -0700, Alex Wang wrote:
> Also, for confirmation, we do not need to guarantee the thread safety
> of coverage counters, since they are not meant to be perfectly accurate,
> right?
The coverage counters should probably become per-thread. I haven't
done anything a
Seems I understand now,
Is that for directing all error messages to corresponding lines in
"lib/dirs.c.in"?
On Mon, Jul 22, 2013 at 10:20 AM, Alex Wang wrote:
> Looks good to me,
>
> One question, why do we have the "#line directive" in "lib/dirs.c.in"?
> We have already created "dirs.c" in "l
Looks good to me,
One question, why do we have the "#line directive" in "lib/dirs.c.in"?
We have already created "dirs.c" in "lib/automake.mk". What is the
use of "#line directive" here?
Thanks,
On Thu, Jul 18, 2013 at 4:15 PM, Ben Pfaff wrote:
> Signed-off-by: Ben Pfaff
> ---
> lib/dirs.c
On Sat, Jul 20, 2013 at 2:14 AM, msj wrote:
> Hi,
> It is my honour to communicate with you .
> I want to ask you one question : I notice that the patch for MPLS does not
> change too much in part of include/openflow , and it only contains some change
> in nicira-ext.h .
> well ,does that mean
Looks good to me, thanks,
On Thu, Jul 18, 2013 at 4:15 PM, Ben Pfaff wrote:
> Signed-off-by: Ben Pfaff
> ---
> lib/meta-flow.c | 41 +++--
> 1 files changed, 23 insertions(+), 18 deletions(-)
>
> diff --git a/lib/meta-flow.c b/lib/meta-flow.c
> index 6b3
Looks good to me,
On Thu, Jul 18, 2013 at 4:15 PM, Ben Pfaff wrote:
> Signed-off-by: Ben Pfaff
> ---
> lib/jsonrpc.c | 12 ++--
> 1 files changed, 10 insertions(+), 2 deletions(-)
>
> diff --git a/lib/jsonrpc.c b/lib/jsonrpc.c
> index b4bbc84..6c482c2 100644
> --- a/lib/jsonrpc.c
>
Looks good to me,
Also, for confirmation, we do not need to guarantee the thread safety
of coverage counters, since they are not meant to be perfectly accurate,
right?
Thanks,
On Thu, Jul 18, 2013 at 4:15 PM, Ben Pfaff wrote:
> Signed-off-by: Ben Pfaff
> ---
> lib/lockfile.c | 11
Looks good to me,
Want to ask question about "ovsthread_once_start()",
1. why do we have "ovsthread_once_start()" defined in header file?
2. if __check__ is defined, the macro function will override the
previously defined "ovsthread_once_start()" function. Right?
3. could you explain why you use
Looks good to me, thanks,
On Thu, Jul 18, 2013 at 4:15 PM, Ben Pfaff wrote:
> Signed-off-by: Ben Pfaff
> ---
> lib/memory.c |5 -
> 1 files changed, 4 insertions(+), 1 deletions(-)
>
> diff --git a/lib/memory.c b/lib/memory.c
> index 1137390..6c97e19 100644
> --- a/lib/memory.c
> +++
Looks good to me, thanks,
On Thu, Jul 18, 2013 at 4:15 PM, Ben Pfaff wrote:
> Signed-off-by: Ben Pfaff
> ---
> vswitchd/xenserver.c | 23 +++
> 1 files changed, 11 insertions(+), 12 deletions(-)
>
> diff --git a/vswitchd/xenserver.c b/vswitchd/xenserver.c
> index 1f26e0d
This commit fixes the warning issued by 'clang' when pointer is casted
to one with greater alignment.
Signed-off-by: Alex Wang
---
lib/hash.c |6 +++---
lib/jhash.c| 10 +-
lib/mac-learning.c |4 ++--
lib/netdev-linux.c |3 ++-
lib/netlink.c
Looks good to me, thanks,
On Thu, Jul 18, 2013 at 4:15 PM, Ben Pfaff wrote:
> This ensures that an external lock around a dpif_netdev will allow
> thread-safe access to it.
>
> Signed-off-by: Ben Pfaff
> ---
> lib/dpif-netdev.c | 12 ++--
> 1 files changed, 6 insertions(+), 6 deleti
67 matches
Mail list logo
