All three patches look good to me. Guolin, please feel free to do
additional review or testing.
Acked-by: Andy Zhou
On Fri, Jan 10, 2014 at 3:18 PM, Ben Pfaff wrote:
> handle_upcalls() always installed a flow for every packet, as long as
> the datapath didn't already have too many flows, but
Both mega flow mask's reference counter and per flow table mask list
should only be accessed when holding ovs_mutex() lock. However
this is not true with ovs_flow_table_flush(). The patch fixes this bug.
Signed-off-by: Andy Zhou
---
v1->v2:
Atomic ops can protect mask's reference coun
On Fri, Jan 10, 2014 at 03:50:09PM -0800, Jesse Gross wrote:
> On Thu, Jan 9, 2014 at 9:17 AM, Ben Pfaff wrote:
> > On Thu, Jan 02, 2014 at 11:51:15AM -0500, Jesse Gross wrote:
> >> On Sun, Dec 29, 2013 at 2:50 AM, Ben Pfaff wrote:
> >> * I'm not sure that flow_count_mpls_labels() is megaflow-saf
On Thu, Jan 9, 2014 at 9:17 AM, Ben Pfaff wrote:
> On Thu, Jan 02, 2014 at 11:51:15AM -0500, Jesse Gross wrote:
>> On Sun, Dec 29, 2013 at 2:50 AM, Ben Pfaff wrote:
>> > I've been a little frustrated with the current approach to MPLS, because it
>> > seems quite difficult to understand. One part
On Fri, Jan 10, 2014 at 07:12:26PM -0200, Flavio Leitner wrote:
> On Fri, Jan 10, 2014 at 12:30:06PM -0800, Ben Pfaff wrote:
> > ovsdbmonitor was poorly maintained and not widely used.
> >
> > CC: Flavio Leitner
> > Signed-off-by: Ben Pfaff
> > ---
>
> I can build using openvswitch-fedora.spec
On Fri, Jan 10, 2014 at 03:18:24PM -0800, Ben Pfaff wrote:
> handle_upcalls() always installed a flow for every packet, as long as
> the datapath didn't already have too many flows, but there are cases where
> we don't want to do this:
>
> - If we get multiple packets in a single microflow all
The revalidator code in ofproto-dpif-upcall.c(), in revalidate_ukey(),
deletes any datapath flow for which the kernel reports wildcarded bits
that userspace requires to be matched. Until now, a couple of pieces of
code in ofproto-dpif always marked nw_frag (which tracks whether a packet
is an IPv4
handle_upcalls() always installed a flow for every packet, as long as
the datapath didn't already have too many flows, but there are cases where
we don't want to do this:
- If we get multiple packets in a single microflow all in one batch
(perhaps due to GSO breaking up a large TCP packe
The revalidator code in ofproto-dpif-upcall.c(), in revalidate_ukey(),
deletes any datapath flow for which the kernel reports wildcarded bits
that userspace requires to be matched. Until now, tnl_xlate_init() marked
every bit in the tunnel flags as required to be matched. Since most of
those bits
Fixes the following error on distcheck:-
ERROR: files left in build directory after distclean:
./tests/test-controller.8
make[1]: *** [distcleancheck] Error 1
Signed-off-by: Joe Stringer
---
tests/automake.mk |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tests/automake.
On Fri, Jan 10, 2014 at 12:53 PM, Jesse Gross wrote:
> On Thu, Jan 9, 2014 at 5:19 PM, Andy Zhou wrote:
> > diff --git a/datapath/flow_table.c b/datapath/flow_table.c
> > index 4232b82..4170a96 100644
> > --- a/datapath/flow_table.c
> > +++ b/datapath/flow_table.c
> > void ovs_flow_free(struct
On Fri, Jan 10, 2014 at 12:30:06PM -0800, Ben Pfaff wrote:
> ovsdbmonitor was poorly maintained and not widely used.
>
> CC: Flavio Leitner
> Signed-off-by: Ben Pfaff
> ---
I can build using openvswitch-fedora.spec now.
Thanks,
Acked-by: Flavio Leitner
__
On Thu, Jan 9, 2014 at 5:19 PM, Andy Zhou wrote:
> diff --git a/datapath/flow_table.c b/datapath/flow_table.c
> index 4232b82..4170a96 100644
> --- a/datapath/flow_table.c
> +++ b/datapath/flow_table.c
> void ovs_flow_free(struct sw_flow *flow, bool deferred)
> {
> if (!flow)
>
On Wed, Jan 08, 2014 at 10:51:43AM -0800, Ben Pfaff wrote:
> On Wed, Jan 08, 2014 at 09:53:51AM -0200, Flavio Leitner wrote:
> > On Tue, Jan 07, 2014 at 06:17:59PM -0800, Ben Pfaff wrote:
> > > On Tue, Jan 07, 2014 at 11:12:41PM -0200, Flavio Leitner wrote:
> > > > The fedora's rpm policy doesn't a
In an earlier incarnation of dpif, the interface specified that if
flow_dump_next() returned an error code, then the function must not be
called again in the same dump operation. As such, the dpif itself needed
to cache the error status to ensure that it did not make such calls.
Recent changes have
The 'buffer' member in 'struct nl_dump' will be going away, so this
patch ensures that nl_dump_done() no longer uses it to drain the
nl_sock.
Signed-off-by: Joe Stringer
---
lib/netlink-socket.c |7 +--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/lib/netlink-socket.c b/
Previously when dumping flows from the datapath, callers of
dpif_flow_dump_next() could specify whether they wish to fetch the
actions for a flow. Depending on the datapath implementation, these may
be fetched at no performance cost or may require additional resources
to fetch.
The most frequent c
This patch makes it the caller's responsibility to allocate and pass a
buffer down to the dpif_flow_dump_next() implementation, to act as
storage for the next flow. The implementation can expect to be called
from multiple threads with the same 'state' and different 'buffer's.
When flow_dump_next()
This new function allows callers to determine whether the contents of
the given buffer will be modified or reallocated on the next call to
dpif_flow_dump_next(). This will be used in a future commit to allow
batched flow deletion by revalidator threads.
Signed-off-by: Joe Stringer
---
I was a lit
The function comment still referred to a 'msg' variable, which has been
renamed.
Signed-off-by: Joe Stringer
---
lib/netlink-socket.c |6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/lib/netlink-socket.c b/lib/netlink-socket.c
index 4bd6d36..bf8eb32 100644
--- a/lib/n
This patch modifies 'struct nl_dump' and nl_dump_next() to allow
multiple threads to share the same nl_dump. These changes are targeted
around synchronizing buffer status between multiple callers, and
allowing callers to fully process their existing buffers before
determining whether to stop fetchi
This patch makes all of the users of 'struct nl_dump' allocate their own
buffers to pass down to nl_dump_next(). This paves the way for allowing
multithreaded flow dumping.
Signed-off-by: Joe Stringer
---
lib/dpif-linux.c | 16
lib/netdev-linux.c | 42 +
This patch simplifies buffer usage in dpif_linux_flow_dump_next() in
preparation for later changes.
Signed-off-by: Joe Stringer
---
lib/dpif-linux.c | 16 ++--
1 file changed, 10 insertions(+), 6 deletions(-)
diff --git a/lib/dpif-linux.c b/lib/dpif-linux.c
index ed824a6..d985f4b
Currently, we have a single thread (the 'flow_dumper') that fetches the flows
from the datapath and distributes them to revalidator threads. To clean up the
code (and provide a small performance benefit), we are looking at removing this
flow_dumper thread, and have each revalidator fetch flows from
This should help to highlight the different users of nl_dump.buffer.
A later patch will remove this field.
Signed-off-by: Joe Stringer
---
lib/dpif-linux.c | 14 --
lib/netdev-linux.c |6 +++---
lib/netlink-socket.c | 30 --
lib/netlink-socke
Thanks. I applied this to master. I improved the commit message to:
ofproto: Add more thread safety annotations.
These would have found the problem fixed in commit c7be3f559349 (connmgr:
Fix attempt to take mutex recursively when exiting fail-open.).
Signed-off-by: Ben Pfaff
ofproto_check_ofpacts() checks for a valid meter ID but ofpacts_check()
does not.
Reported-by: John Hurley
Signed-off-by: Ben Pfaff
---
It looks like this bug is only present in OVS 2.0. I don't think
later versions have the same problem, and earlier versions didn't
have meters at all.
diff --
Thanks, I applied this to master and branch-2.1.
On Thu, Jan 09, 2014 at 01:38:04PM -0800, Alex Wang wrote:
> Looks good to me,
>
>
> On Tue, Dec 31, 2013 at 2:24 PM, Ben Pfaff wrote:
>
> > Libtool is now required as of commit 38b7a52b61 (openvswitch: Use libtool
> > and allow building shared
Thank you for the review. I applied this to master.
On Thu, Jan 09, 2014 at 01:34:38PM -0800, Alex Wang wrote:
> Looks good to me
> Tested on machine without libtool, worked as expected,
>
>
> On Tue, Dec 31, 2013 at 2:20 PM, Ben Pfaff wrote:
>
> > Otherwise users get an error later like:
> >
NETKAMER.COM - Защита от штрафов с камер ГИБДД на дорогах.
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
30 matches
Mail list logo
