On Wed, Aug 08, 2012 at 08:57:56PM -0700, Ben Pfaff wrote:
> On Thu, Aug 09, 2012 at 12:21:36PM +0900, Simon Horman wrote:
> > On Tue, Aug 07, 2012 at 11:50:19AM -0700, Ben Pfaff wrote:
> > > The learning-switch implementation needs to know the OpenFlow version in
> > > use to send the initial hand
On Wed, Aug 08, 2012 at 05:52:57PM -0700, Leo Alterman wrote:
> 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
"git am" reported
On Wed, Aug 08, 2012 at 05:52:58PM -0700, Leo Alterman wrote:
> lockfile_lock() accepts a timeout argument but, aside from unit tests
> pertaining to timeout, its value is always 0. Since this feature relies on
> a periodic SIGALRM signal, which is no longer a given after refactoring
> timeval.c, t
On Thu, Aug 09, 2012 at 12:21:36PM +0900, Simon Horman wrote:
> On Tue, Aug 07, 2012 at 11:50:19AM -0700, 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 n
On Tue, Aug 07, 2012 at 11:50:19AM -0700, 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 ha
On Wed, Aug 08, 2012 at 05:04:01PM -0700, Ben Pfaff wrote:
> On Wed, Aug 08, 2012 at 06:49:45AM +0900, Simon Horman wrote:
> > Signed-off-by: Simon Horman
>
> > +bad_in_port_err = OFPERR_OFPBMC_BAD_VALUE;
>
> Oh, I see, I misunderstood that there *was* a reasonable error code in
> other
On Wed, Aug 08, 2012 at 05:07:32PM -0700, Ben Pfaff wrote:
> On Wed, Aug 08, 2012 at 06:49:46AM +0900, Simon Horman wrote:
> > Signed-off-by: Simon Horman
>
> Hmm, ofp12_flow_removed has a 'hard_timeout' member that this doesn't
> populate.
Sorry for missing that. Do you think it should
be set t
On Wed, Aug 08, 2012 at 05:10:18PM -0700, Ben Pfaff wrote:
> On Wed, Aug 08, 2012 at 06:49:47AM +0900, Simon Horman wrote:
> > Signed-off-by: Simon Horman
> > +case OFPUTIL_P_OF12: {
> > +struct ofp11_flow_stats_request *ofsr;
> > +
> > +raw = (fsr->aggregate
> > +
At the end of the NetFlow unit tests we warp time to force any
remaining flows to expire and then immediately kill OVS and the
collector. However, this creates a race where sometimes these
processes are killed before the last records are sent or collected.
It's possible to force OVS to go through
Thanks guys. Submitted revised patch series which checks for Linux, uses
fewer preprocessor conditions, and deals with some of the resulting unit
test fallout.
~Leo
On Wed, Aug 8, 2012 at 12:24 PM, Ben Pfaff wrote:
> On Tue, Aug 07, 2012 at 05:00:59PM -0700, Leo Alterman wrote:
> > 64-bit glibc
lockfile_lock() accepts a timeout argument but, aside from unit tests
pertaining to timeout, its value is always 0. Since this feature relies on
a periodic SIGALRM signal, which is no longer a given after refactoring
timeval.c, the cleanest solution is to just remove it.
Signed-off-by: Leo Alterma
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| 36 +---
lib/timeval.h
On Wed, Aug 08, 2012 at 05:10:18PM -0700, Ben Pfaff wrote:
> On Wed, Aug 08, 2012 at 06:49:47AM +0900, Simon Horman wrote:
> > Signed-off-by: Simon Horman
> > +case OFPUTIL_P_OF12: {
> > +struct ofp11_flow_stats_request *ofsr;
> > +
> > +raw = (fsr->aggregate
> > +
On Wed, Aug 08, 2012 at 06:49:47AM +0900, Simon Horman wrote:
> Signed-off-by: Simon Horman
> +case OFPUTIL_P_OF12: {
> +struct ofp11_flow_stats_request *ofsr;
> +
> +raw = (fsr->aggregate
> + ? OFPRAW_OFPST_AGGREGATE_REQUEST
> + : OFPRAW_OFPST_FLOW_
On Wed, Aug 08, 2012 at 06:49:46AM +0900, Simon Horman wrote:
> Signed-off-by: Simon Horman
Hmm, ofp12_flow_removed has a 'hard_timeout' member that this doesn't
populate.
I queued this up anyway.
___
dev mailing list
dev@openvswitch.org
http://openvsw
On Wed, Aug 08, 2012 at 05:00:02PM -0700, Ben Pfaff wrote:
> On Thu, Aug 09, 2012 at 08:56:10AM +0900, Simon Horman wrote:
> > On Wed, Aug 08, 2012 at 04:42:30PM -0700, Ben Pfaff wrote:
> > > On Wed, Aug 08, 2012 at 06:49:44AM +0900, Simon Horman wrote:
> > > > Signed-off-by: Simon Horman
> > > >
On Wed, Aug 08, 2012 at 06:49:45AM +0900, Simon Horman wrote:
> Signed-off-by: Simon Horman
> +bad_in_port_err = OFPERR_OFPBMC_BAD_VALUE;
Oh, I see, I misunderstood that there *was* a reasonable error code in
other versions of OpenFlow. Sorry about that.
I applied the following increme
On Thu, Aug 09, 2012 at 08:56:10AM +0900, Simon Horman wrote:
> On Wed, Aug 08, 2012 at 04:42:30PM -0700, Ben Pfaff wrote:
> > On Wed, Aug 08, 2012 at 06:49:44AM +0900, Simon Horman wrote:
> > > Signed-off-by: Simon Horman
> > > {
> > > -const struct ofp_packet_out *opo;
> > > -enum ofper
On Wed, Aug 08, 2012 at 04:42:30PM -0700, Ben Pfaff wrote:
> On Wed, Aug 08, 2012 at 06:49:44AM +0900, Simon Horman wrote:
> > Signed-off-by: Simon Horman
> > {
> > -const struct ofp_packet_out *opo;
> > -enum ofperr error;
> > enum ofpraw raw;
> > struct ofpbuf b;
> > +enum
On Wed, Aug 08, 2012 at 04:32:49PM -0700, Ben Pfaff wrote:
> On Wed, Aug 08, 2012 at 12:19:57PM +0900, Simon Horman wrote:
> > Signed-off-by: Simon Horman
>
> > -size_t size;
> > +size_t size = 0;
> >
> > size = po->ofpacts_len;
> > if (po->buffer_id == UINT32_MAX) {
> > -
On Wed, Aug 08, 2012 at 06:49:44AM +0900, Simon Horman wrote:
> Signed-off-by: Simon Horman
> {
> -const struct ofp_packet_out *opo;
> -enum ofperr error;
> enum ofpraw raw;
> struct ofpbuf b;
> +enum ofperr bad_in_port_err = OFPERR_OFPET_BAD_REQUEST;
>
> ofpbuf_use_c
On Wed, Aug 08, 2012 at 06:49:43AM +0900, Simon Horman wrote:
> Signed-off-by: Simon Horman
Thanks. I applied the series up to this point to master.
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
On Wed, Aug 08, 2012 at 12:19:57PM +0900, Simon Horman wrote:
> Signed-off-by: Simon Horman
> -size_t size;
> +size_t size = 0;
>
> size = po->ofpacts_len;
> if (po->buffer_id == UINT32_MAX) {
> -size += po->packet_len;
> +size = po->packet_len;
> }
I kno
Until now, the jsonrpc code has only counted receiving a full JSON-RPC
messages as activity. This could theoretically time out, then, while a
very long message is in transit or if a slow link is involved. This commit
changes this code to count receiving any part of a message as activity.
This is
The problem fixed in commit 612be57384 also exists in the jsonrpc
code, although we haven't actually observed it there. This series
fixes that problem in jsonrpc too, as a preemptive measure.
Ben Pfaff (3):
reconnect: Rename reconnect_received() to reconnect_activity().
jsonrpc: Treat drainin
Until now, the jsonrpc module has used messages received from the
remote peer as the sole means to determine that the JSON-RPC
connection is up. This could in theory interact badly with a
remote peer that stops reading and processing messages from the
receive queue when there is a backlog in the s
Receiving data is not the only reasonable way to verify that a connection
is up. For example, on a TCP connection, receiving an acknowledgment that
the remote side has accepted data that we sent is also a reasonable means.
Therefore, this commit generalizes the naming.
Also, similarly for the Pyt
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/E07082012123134.cfm?WL=969&WS=208833_8707685&WA=417
] aquí
[ http://email.globalbono.com/Go/index.cfm
On Wed, Aug 08, 2012 at 06:48:01PM +, Ed Maste wrote:
> FreeBSD's system compiler is a somewhat old version of GCC that produced
> a spurious warning about a potential unitialized variable use.
>
> Signed-off-by:Ed Maste
Applied to master, thanks.
By the way, somehow a tab keeps sne
On Wed, Aug 08, 2012 at 06:35:13PM +, Ed Maste wrote:
> This fixes a warning on FreeBSD.
>
> Signed-off-by:Ed Maste
Applied to master, thanks.
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
On Wed, Aug 08, 2012 at 06:29:32PM +, Ed Maste wrote:
> Signed-off-by:Ed Maste
Thanks, applied to master.
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
On Tue, Aug 07, 2012 at 05:00:59PM -0700, Leo Alterman wrote:
> 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
I agree with Ed M
FreeBSD's system compiler is a somewhat old version of GCC that produced
a spurious warning about a potential unitialized variable use.
Signed-off-by: Ed Maste
---
lib/netdev-bsd.c |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/netdev-bsd.c b/lib/netdev-bsd.c
index 0
On Aug 8, 2012, at 1:34 PM, Jesse Gross wrote:
> On Wed, Aug 8, 2012 at 11:13 AM, Kyle Mestery (kmestery)
> wrote:
>> On Aug 8, 2012, at 1:08 PM, Jesse Gross wrote:
>>> On Tue, Aug 7, 2012 at 3:48 PM, Kyle Mestery wrote:
Signed-off-by: Kyle Mestery
---
datapath/datapath.c | 4 ++--
This fixes a warning on FreeBSD.
Signed-off-by: Ed Maste
---
lib/socket-util.c |1 +
1 file changed, 1 insertion(+)
diff --git a/lib/socket-util.c b/lib/socket-util.c
index 314761d..7c40ab8 100644
--- a/lib/socket-util.c
+++ b/lib/socket-util.c
@@ -30,6 +30,7 @@
#include
#include
#inc
On Wed, Aug 8, 2012 at 11:13 AM, Kyle Mestery (kmestery)
wrote:
> On Aug 8, 2012, at 1:08 PM, Jesse Gross wrote:
>> On Tue, Aug 7, 2012 at 3:48 PM, Kyle Mestery wrote:
>>> Signed-off-by: Kyle Mestery
>>> ---
>>> datapath/datapath.c | 4 ++--
>>> 1 file changed, 2 insertions(+), 2 deletions(-)
>>
Signed-off-by: Ed Maste
---
lib/route-table-bsd.c |4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/lib/route-table-bsd.c b/lib/route-table-bsd.c
index 1c29071..eb4a168 100644
--- a/lib/route-table-bsd.c
+++ b/lib/route-table-bsd.c
@@ -104,7 +104,7 @@ route_table_get_nam
On Aug 8, 2012, at 1:08 PM, Jesse Gross wrote:
> On Tue, Aug 7, 2012 at 3:48 PM, Kyle Mestery wrote:
>> Signed-off-by: Kyle Mestery
>> ---
>> datapath/datapath.c | 4 ++--
>> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> Applied, thanks Kyle.
Thanks Jesse. Fedora rolled up to this version,
On Tue, Aug 7, 2012 at 3:48 PM, Kyle Mestery wrote:
> Signed-off-by: Kyle Mestery
> ---
> datapath/datapath.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
Applied, thanks Kyle.
___
dev mailing list
dev@openvswitch.org
http://openvswitch
On 8/8/12 4:06 AM, mvp...@iol.pt wrote:
You said, "NICs only trigger an interrupt after the packet has been
completely DMAed into memory, which makes cut-through impossible to
implement on today's servers". The triggers that you speak about is done
at driver level, at kernel level or is somethin
On Tue, Aug 07, 2012 at 05:00:59PM -0700, Leo Alterman wrote:
> 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
I'd heard that th
Qué usted necesita un préstamo?___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
Hello,
First thanks for the tip.
You said, "NICs only trigger an interrupt after the packet has been
completely DMAed into memory, which makes cut-through impossible to
implement on today's servers". The triggers that you speak about is done
at driver level, at kernel level or is something
43 matches
Mail list logo
