code.
One thread, the "leader", is responsible for beginning and ending each
flow dump, maintaining the flow_limit, and checking whether the
revalidator threads need to exit. All revalidator threads dump,
revalidate, delete datapath flows and garbage collect ukeys.
Co-authored-by: Jo
code.
One thread, the "leader", is responsible for beginning and ending each
flow dump, maintaining the flow_limit, and checking whether the
revalidator threads need to exit. All revalidator threads dump,
revalidate, delete datapath flows and garbage collect ukeys.
Co-authored-by: Jo
Thanks for patching this. I'll hold off on rebasing until this is pushed.
On 22 April 2014 15:28, Ethan Jackson wrote:
> Acked-by: Ethan Jackson
>
>
> On Mon, Apr 21, 2014 at 8:25 PM, Alex Wang wrote:
> > Commit 1f8675481e (ofproto-dpif-upcall: Fix ovs-vswitchd crash.)
> > directly copied the
code.
One thread, the "leader", is responsible for beginning and ending each
flow dump, maintaining the flow_limit, and checking whether the
revalidator threads need to exit. All revalidator threads dump,
revalidate, delete datapath flows and garbage collect ukeys.
Co-authored-by: Jo
org/pipermail/dev/2014-January/036107.html
If this involves resubmit_stats, then it would also need a new xc_type. The
xlate_group_bucket() code would add an entry to ctx->xin->xcache if it is
nonnull.
This would also need to follow the code in xlate_push_stats() and
xlate_cache_clear
f it has already been handled, we skip
handling the flow again.
We also defer ukey cleanup for flows that fail revalidation, so that the
ukey will still exist if the same flow is dumped twice. This allows the
above logic to work in this case.
Signed-off-by: Joe Stringer
---
ofproto/ofproto-dp
f it has already been handled, we skip
handling the flow again.
We also defer ukey cleanup for flows that fail revalidation, so that the
ukey will still exist if the same flow is dumped twice. This allows the
above logic to work in this case.
Signed-off-by: Joe Stringer
---
v2: Didn't squas
On 24 April 2014 04:56, Alex Wang wrote:
>
> On Tue, Apr 22, 2014 at 10:24 PM, Alex Wang wrote:
>
>> diff --git a/ofproto/ofproto-dpif-upcall.c b/ofproto/ofproto-dpif-upcall.c
>>> index 416cfdc..906bf17 100644
>>> --- a/ofproto/ofproto-dpif-upcall.c
>>> +++ b/ofproto/ofproto-dpif-upcall.c
>>> @@
Pushed to master and branch-2.1.
On 24 April 2014 11:06, Joe Stringer wrote:
> On 24 April 2014 04:56, Alex Wang wrote:
>
>>
>> On Tue, Apr 22, 2014 at 10:24 PM, Alex Wang wrote:
>>
>>> diff --git a/ofproto/ofproto-dpif-upcall.c
>>>> b/ofp
Thanks, I tidied up the comments, added threadsafety annotations for xcache
and pushed.
On 24 April 2014 06:38, Ben Pfaff wrote:
> On Tue, Apr 22, 2014 at 04:54:24PM +1200, Joe Stringer wrote:
> > From: Ethan Jackson
> >
> > Previously, we had a separate flow_dumper thre
Acked-by: Joe Stringer
On 18 April 2014 08:32, Alex Wang wrote:
> This commit adds boolean flag in bfd/cfm module for checking
> status change. If there is no status change, the current
> update to OVS database will skip the bfd/cfm session.
>
> In the experiment with 5K bfd
Good to see this as a separate change.
Acked-by: Joe Stringer
On 18 April 2014 08:32, Alex Wang wrote:
> This commit adds per 'struct ofproto_dpif' sequence number for
> packet-in I/O. Whenever a packet-in is inserted into the 'pins'
> queue, the inserting th
There's a few comments inline.
On 18 April 2014 08:32, Alex Wang wrote:
> This commit removes the 'Instant' stats related logic in bridge.c.
> Instead, the corresponding status is updated immediately after the
> global connectivity sequence number changes.
>
Could you update the commit message
nt
of logs already, then checking for the expected messages.
Signed-off-by: Joe Stringer
---
tests/ofproto-dpif.at |3 +++
1 file changed, 3 insertions(+)
diff --git a/tests/ofproto-dpif.at b/tests/ofproto-dpif.at
index 89c8ad7..77656db 100644
--- a/tests/ofproto-dpif.at
+++ b/tests/ofpro
Looking again, it's probably better to change the lines which look for
"monitor thread created" and "..terminated" to be OVS_WAIT_UNTIL. I plan to
send a fresh version.
On 25 April 2014 12:34, Joe Stringer wrote:
> Occasionally, test #770 "ofproto-dpif - ofp
ose stats
would be lost.
This caused intermittent testsuite failures on "learning action -
self-modifying flow with idle_timeout". Introduced by 698ffe3623f1b630ae
"revalidator: Only revalidate high-throughput flows."
Bug #1238927.
Signed-off-by: Joe Stringer
---
ofproto/ofprot
Thanks for fixing this, Alex. I think there was a bit of mixup the last
time I rebased.
On 26 April 2014 07:33, Alex Wang wrote:
> Applied, thx a lot,
>
> Now, let's fix the build (unittest) issue~~~
>
>
> On Fri, Apr 25, 2014 at 11:31 AM, Gurucharan Shetty wrote:
>
>> On Fri, Apr 25, 2014 at 1
bytes=840, drop=0, errs=0, coll=0
>>
>>
>> On Thu, Apr 24, 2014 at 10:34 PM, Alex Wang wrote:
>>
>>> i'll kick an overnight run on "learning action - self-modifying flow
>>> with idle_timeout".~ (seq 1). if it survives, tmr morning,
>>
r
creation/termination of threads to wait until they see the messages
instead.
Signed-off-by: Joe Stringer
---
v2: Fix multiple races.
---
tests/ofproto-dpif.at | 24 ++--
1 file changed, 6 insertions(+), 18 deletions(-)
diff --git a/tests/ofproto-dpif.at b/tests/ofproto-dpi
bove it too. I think it's useful to have the context that
these are treated separately because we want to keep them as up-to-date as
possible (with the performance tradeoff that this can't be done at large
scale).
Otherwise, looks good.
Acked-by: Joe Stringer
On 19 March 2014 04:50, Ben Pfaff wrote:
> On Tue, Mar 18, 2014 at 03:45:01PM +1300, Joe Stringer wrote:
> > Regarding the testsuite failures, the problem is that we wait a certain
> > amount of time and check the datapath flows to ensure that they are as
> > expected. Patch
This was causing build failures on debian wheezy. Check for the feature
rather than the version.
Signed-off-by: Joe Stringer
---
acinclude.m4 |1 +
datapath/linux/compat/include/linux/skbuff.h |4 ++--
2 files changed, 3 insertions(+), 2 deletions
checked before vswitchd was able to spawn the
> > thread.
> >
> > This patch fixes the race by modifying the commands that check for
> > creation/termination of threads to wait until they see the messages
> > instead.
> >
&g
Thanks, applied.
On 29 April 2014 09:27, Jesse Gross wrote:
> On Sun, Apr 27, 2014 at 7:07 PM, Joe Stringer
> wrote:
> > This was causing build failures on debian wheezy. Check for the feature
> > rather than the version.
> >
> > Signed-off-by: Joe Strin
Will you push a tag for this too?
On 29 April 2014 10:31, Ben Pfaff wrote:
> On Mon, Apr 28, 2014 at 02:57:17PM -0700, Justin Pettit wrote:
> > Signed-off-by: Justin Pettit
>
> For both:
> Acked-by: Ben Pfaff
> ___
> dev mailing list
> dev@openvswit
Applied to branch-2.1 as well.
On 29 April 2014 10:21, Joe Stringer wrote:
> Thanks, applied.
>
>
> On 29 April 2014 09:27, Jesse Gross wrote:
>
>> On Sun, Apr 27, 2014 at 7:07 PM, Joe Stringer
>> wrote:
>> > This was causing build failures on debian wheez
On 30 April 2014 05:29, Alex Wang wrote:
>
> +dnl disable the stats update, the following 'recv' should not be updated.
> +AT_CHECK([ovs-vsctl set O . other_config:stats-update-interval=1000])
> +for i in `seq 0 10`; do ovs-appctl time/warp 1000; done
> +for i in `seq 1 5`; do
> +AT_CHECK(
This configuration option was shifted in 2.0, then removed in 2.1.
Remove the misleading log message.
Signed-off-by: Joe Stringer
---
vswitchd/bridge.c |7 ---
1 file changed, 7 deletions(-)
diff --git a/vswitchd/bridge.c b/vswitchd/bridge.c
index 84e9ab8..12852b4 100644
--- a/vswitchd
Pushed to master and branch-2.2.
On 1 May 2014 09:59, Ethan Jackson wrote:
> Acked-by: Ethan Jackson
>
>
> On Wed, Apr 30, 2014 at 2:57 PM, Joe Stringer
> wrote:
> > This configuration option was shifted in 2.0, then removed in 2.1.
> > Remove the misleading log
netdev_rxq_open() open-codes much of netdev_ref(), so re-use that
function instead.
Signed-off-by: Joe Stringer
---
lib/netdev.c |5 +
1 file changed, 1 insertion(+), 4 deletions(-)
diff --git a/lib/netdev.c b/lib/netdev.c
index 2fc1834..6f997fd 100644
--- a/lib/netdev.c
+++ b/lib
netdev_open() would previously increment a netdev's refcount without
holding a lock for it. This commit shifts the locking to protect it.
Found by inspection.
Signed-off-by: Joe Stringer
---
lib/netdev.c |6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/lib/net
netdev_open() and logging a warning message.
Bug #1198386.
Signed-off-by: Joe Stringer
---
I'm not sure if this case is meant to be possible, but I've observed it,
and this patch makes the error more obvious in the logs.
---
lib/netdev.c |4
1 file changed, 4 insertions(+)
di
of statistics update to database, in milliseconds.
>
Perhaps a bit more natural to say "Interval for updating statistics to the
database, in milliseconds".
+Key-value pairs that report port statistics. The update period
> +is controlled by + key=&qu
> Sorry, this was wrong. I mean this (as you suggested):
>
> +OVS_VSCTL_CHECK_RX_PKT([p1], [1])
> dnl advance the clock by 100K ms, the previous 'recv' should be updated.
> -for i in `seq 0 100`; do ovs-appctl time/warp 1000; done
> +for i in `seq 0 99`; do ovs-appctl time/warp 1000; done
> OVS
attempting to configure the port with the same gre settings as p0,
despite the command specifying the type as internal.
This patch catches this case in netdev_open(), removes the old device
from the netdev shash and ensures that one with the new type is created.
Bug #1198386.
Signed-off-by: Joe Stringer
On 3 May 2014 02:56, Ben Pfaff wrote:
> On Fri, May 02, 2014 at 01:13:43PM +1200, Joe Stringer wrote:
> > Previously, it was possible to open a netdevice as one type, then
> > proceed to open it as a different type without first closing it. The
> > bridge code would expect i
Thanks for review, pushed to master.
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
netdev_open() would previously increment a netdev's refcount without
holding a lock for it. This commit shifts the locking to protect it.
Found by inspection.
Signed-off-by: Joe Stringer
---
lib/netdev.c |6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/lib/net
Thanks for review, pushed to master and branch-2.2.
Backport for 2.1 and 2.0 is here:
http://openvswitch.org/pipermail/dev/2014-May/039761.html
On 3 May 2014 02:52, Ben Pfaff wrote:
> On Fri, May 02, 2014 at 01:13:42PM +1200, Joe Stringer wrote:
> > netdev_open() would previously in
On 5 May 2014 12:19, YAMAMOTO Takashi wrote:
> > netdev_open() would previously increment a netdev's refcount without
> > holding a lock for it. This commit shifts the locking to protect it.
> >
> > Found by inspection.
> >
> > Signed-off-by: Joe
the failure
case of tnl_port_add__(), and ensures that the tnl_port structure is
freed in that case as well.
Bug #1198386.
Signed-off-by: Joe Stringer
---
This is a follow up to "netdev: Open devices as the type specified."
- Fixed root cause instead of avoiding symptoms.
- Removed a t
Thanks for taking a look Ryan. That sounds exactly like the issue I
encountered.
Looking at your assessment made me take a look at it again, and I found a
much better solution. Do you mind taking a look at that one?
http://openvswitch.org/pipermail/dev/2014-May/039790.html
On 6 May 2014 09:18,
Hi Jarno,
Quick question..
On 18 April 2014 13:36, Jarno Rajahalme wrote:
> miniflow_extract() extracts packet headers directly to a miniflow,
> which is a compressed form of the struct flow. This does not require
> a large struct to be cleared to begin with, and accesses less memory.
> These
Thanks Alex, applied to master and backported 2.0-2.2. I can also backport
to 1.10 and 1.11 if necessary.
I'll follow up on the warn patch in the other thread.
On 7 May 2014 19:58, Alex Wang wrote:
> Hey Joe,
>
> Sorry for the delayed review,
>
> Good to see this simpler version and fix of mem
te:
>
>> On Mon, May 05, 2014 at 11:17:23AM +1200, Joe Stringer wrote:
>> > Previously, it was possible to open a netdevice as one type, then
>> > proceed to open it as a different type without first closing it. This
>> > would result in the original devic
This was causing a mismatch between the odp flow key in an upcall
compared to the odp flow key that is serialised upon flow_dump.
Signed-off-by: Joe Stringer
---
lib/dpif-netdev.c | 21 +++--
tests/ofproto-dpif.at | 116 -
2 files
On 10 May 2014 03:14, Ben Pfaff wrote:
> On Fri, May 09, 2014 at 02:35:44PM +1200, Joe Stringer wrote:
>
> This was causing a mismatch between the odp flow key in an upcall
> > compared to the odp flow key that is serialised upon flow_dump.
> >
> > Signed-off-by: Joe
Is there any general principles behind fixing these? I'm wondering if
there's a way I can prevent these in future.
For example, always sleep after sending packets if no flow exists in the
datapath.
___
dev mailing list
dev@openvswitch.org
http://openvswi
On 25 March 2014 14:58, Andy Zhou wrote:
>
> /* Metadata. */
> +if (present_attrs & (UINT64_C(1) << OVS_KEY_ATTR_RECIRC_ID)) {
> +flow->recirc_id = nl_attr_get_u32(attrs[OVS_KEY_ATTR_RECIRC_ID]);
> +expected_attrs |= UINT64_C(1) << OVS_KEY_ATTR_RECIRC_ID;
> +} else if
The comment was very specific to one user of the function, but the
effect of the case is wider.
Signed-off-by: Joe Stringer
---
lib/odp-util.c |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/odp-util.c b/lib/odp-util.c
index af464a0..34ac5cd 100644
--- a/lib/odp
flow key in an upcall compared to the one
that is serialised upon flow_dump.
This patch ensures that dp_hash is always matched for datapaths that
support it, and brings the userspace datapath behaviour back in line
with the kernel datapath.
Signed-off-by: Joe Stringer
---
lib/dpif-netdev.c
Ensure that upcall key matches flow install and flow_dump for userspace
datapath. This was previously assumed, but not tested. This makes the
assumption more explicit.
Signed-off-by: Joe Stringer
---
tests/ofproto-dpif.at | 78 +
1 file changed
AM +1200, Joe Stringer wrote:
> > Ensure that upcall key matches flow install and flow_dump for userspace
> > datapath. This was previously assumed, but not tested. This makes the
> > assumption more explicit.
> >
> > Signed-off-by: Joe Stringer
>
> Do you thi
Thanks for the review, I plan to push these soon (and backport this patch
to 2.2)
On 13 May 2014 13:02, Ben Pfaff wrote:
> On Tue, May 13, 2014 at 11:20:31AM +1200, Joe Stringer wrote:
> > The userspace and kernel datapaths previously differed on their
> > treatment of th
Signed-off-by: Joe Stringer
---
tests/ofproto-dpif.at |3 ---
1 file changed, 3 deletions(-)
diff --git a/tests/ofproto-dpif.at b/tests/ofproto-dpif.at
index 3757318..dd5f4fd 100644
--- a/tests/ofproto-dpif.at
+++ b/tests/ofproto-dpif.at
@@ -2,9 +2,6 @@ AT_BANNER([ofproto-dpif])
# Strips
Ensure that upcall key matches flow install and flow_dump for userspace
datapath. This was previously assumed, but not tested. This makes the
assumption more explicit.
Signed-off-by: Joe Stringer
---
v2: Shift to dpif-netdev.at.
Depends on "dpif-netdev: Always serialise recirculation in u
Shift datapath flow test macros and "ofproto-dpif - dummy interface" out
into a separate file.
Signed-off-by: Joe Stringer
---
tests/automake.mk |1 +
tests/dpif-netdev.at | 53 +
tests/ofproto-dpif
On 13 May 2014 16:44, Andy Zhou wrote:
>
> > diff --git a/ofproto/ofproto-dpif.c b/ofproto/ofproto-dpif.c
> > index 926f3d6..bb16e7b 100644
> > --- a/ofproto/ofproto-dpif.c
> > +++ b/ofproto/ofproto-dpif.c
> > @@ -3211,10 +3211,11 @@ rule_dpif_lookup(struct ofproto_dpif *ofproto,
> struct flow *f
sh, may be we can drop the else if (is_mask) case
> for recirc_id?
>
> On Mon, May 12, 2014 at 4:00 PM, Joe Stringer wrote:
> > On 25 March 2014 14:58, Andy Zhou wrote:
> >>
> >> /* Metadata. */
> >> +if (present_attrs & (UINT64_C
Thanks, applied to master.
On 13 May 2014 13:00, Ben Pfaff wrote:
> On Tue, May 13, 2014 at 11:20:30AM +1200, Joe Stringer wrote:
> > The comment was very specific to one user of the function, but the
> > effect of the case is wider.
>
> Fixes a spelling error too.
&g
On 13 May 2014 17:43, Joe Stringer wrote:
> On 13 May 2014 16:44, Andy Zhou wrote:
>>
>> > diff --git a/ofproto/ofproto-dpif.c b/ofproto/ofproto-dpif.c
>> > index 926f3d6..bb16e7b 100644
>> > --- a/ofproto/ofproto-dpif.c
>> > +++ b/ofp
Signed-off-by: Joe Stringer
---
v3: First post.
---
lib/odp-util.c | 10 +-
1 file changed, 5 insertions(+), 5 deletions(-)
diff --git a/lib/odp-util.c b/lib/odp-util.c
index 34ac5cd..6cff2f1 100644
--- a/lib/odp-util.c
+++ b/lib/odp-util.c
@@ -2643,11 +2643,11
Shift datapath flow test macros and "ofproto-dpif - dummy interface" out
into a separate file.
Signed-off-by: Joe Stringer
---
v3: No change.
---
tests/automake.mk |1 +
tests/dpif-netdev.at | 53 +
tests/ofproto-dpif
Signed-off-by: Joe Stringer
---
v3: No change.
---
tests/ofproto-dpif.at |3 ---
1 file changed, 3 deletions(-)
diff --git a/tests/ofproto-dpif.at b/tests/ofproto-dpif.at
index 0487122..9e387c1 100644
--- a/tests/ofproto-dpif.at
+++ b/tests/ofproto-dpif.at
@@ -2,9 +2,6 @@ AT_BANNER([ofproto
Signed-off-by: Joe Stringer
---
v3: First post.
---
lib/odp-util.c | 41 +
1 file changed, 25 insertions(+), 16 deletions(-)
diff --git a/lib/odp-util.c b/lib/odp-util.c
index 6cff2f1..ff3cb7e 100644
--- a/lib/odp-util.c
+++ b/lib/odp-util.c
@@ -2482,16
Ensure that upcall key matches flow install and flow_dump for userspace
datapath. This was previously assumed, but not tested. This makes the
assumption more explicit.
Signed-off-by: Joe Stringer
---
v3: No change.
v2: Shift to dpif-netdev.at.
---
tests/dpif-netdev.at | 70
Acked-by: Joe Stringer
On 16 May 2014 10:36, Ben Pfaff wrote:
> The loop in revalidate() needs to ensure that any data obtained from
> dpif_flow_dump_next() is used before it is destroyed, as indicated by
> dpif_flow_dump_next_may_destroy_keys(). In the common case, where
>
On 16 May 2014 15:11, Andy Zhou wrote:
> On Wed, May 14, 2014 at 11:57 PM, Joe Stringer
> wrote:
> > Signed-off-by: Joe Stringer
> > ---
> > v3: First post.
> > ---
> > lib/odp-util.c | 41 +
> > 1 fi
On 16 May 2014 15:38, Andy Zhou wrote:
> On Thu, May 15, 2014 at 8:34 PM, Joe Stringer
> wrote:
> > On 16 May 2014 15:11, Andy Zhou wrote:
> >>
> >> On Wed, May 14, 2014 at 11:57 PM, Joe Stringer
> >> wrote:
> >> > Signed-off
On 19 May 2014 13:05, Andy Zhou wrote:
> On Thu, May 15, 2014 at 9:52 PM, Joe Stringer
> wrote:
> > The userspace and kernel datapaths previously differed on their
> > treatment of the recirc_id and dp_hash fields when sending upcalls.
> > While the kernel datapath woul
This looks much tidier, I like it. Particularly given the elimination of 4
goto statements :-). Comments inline.
On 17 May 2014 11:28, Ben Pfaff wrote:
>
> + * Notes
> + * -
> + *
> + * All error reporting is deferred to the call to
> dpif_flow_dump_destroy().
> + *
> + * Previous versions o
Signed-off-by: Joe Stringer
---
utilities/ovs-dev.py | 10 ++
1 file changed, 10 insertions(+)
diff --git a/utilities/ovs-dev.py b/utilities/ovs-dev.py
index 7a4425f..97fab6c 100755
--- a/utilities/ovs-dev.py
+++ b/utilities/ovs-dev.py
@@ -144,6 +144,10 @@ commands.append(make
Signed-off-by: Joe Stringer
---
utilities/ovs-dev.py |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/utilities/ovs-dev.py b/utilities/ovs-dev.py
index 97fab6c..d979afd 100755
--- a/utilities/ovs-dev.py
+++ b/utilities/ovs-dev.py
@@ -144,7 +144,7 @@ commands.append(make
m the
patch.
On 19 May 2014 13:51, Joe Stringer wrote:
> On 19 May 2014 13:05, Andy Zhou wrote:
>
>> On Thu, May 15, 2014 at 9:52 PM, Joe Stringer
>> wrote:
>> > The userspace and kernel datapaths previously differed on their
>> > treatment of the recirc_i
Pushed to master with the following incremental:
diff --git a/lib/odp-util.c b/lib/odp-util.c
index 37ec2be..280e8a6 100644
--- a/lib/odp-util.c
+++ b/lib/odp-util.c
@@ -2483,8 +2483,7 @@ ovs_to_odp_frag_mask(uint8_t nw_frag_mask)
static void
odp_flow_key_from_flow__(struct ofpbuf *buf, const st
Thanks for the review. Applied patches 1, 4, 5, 6 to master.
On 19 May 2014 13:26, Andy Zhou wrote:
> On Wed, May 14, 2014 at 11:57 PM, Joe Stringer
> wrote:
> > Ensure that upcall key matches flow install and flow_dump for userspace
> > datapath. This was previously assum
On 19 May 2014 19:53, Lori Jakab wrote:
> On 5/19/14, 10:13 AM, YAMAMOTO Takashi wrote:
>
>> Signed-off-by: Joe Stringer
>>> ---
>>> utilities/ovs-dev.py |2 +-
>>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>>
>>> diff --gi
Rather than setting and resetting the 'mark' field in the ukey, this
patch introduces a seq to track whether a flow has been seen during the
most recent dump. This tidies the code and simplifies the logic for
detecting when flows are duplicated from the datapath.
Signed-off-by: Jo
Signed-off-by: Joe Stringer
---
v2: Make it configurable.
---
utilities/ovs-dev.py |4
1 file changed, 4 insertions(+)
diff --git a/utilities/ovs-dev.py b/utilities/ovs-dev.py
index 97fab6c..d08ef39 100755
--- a/utilities/ovs-dev.py
+++ b/utilities/ovs-dev.py
@@ -145,6 +145,8
This was previously logging that there was no such function, but passing
the test anyway.
Signed-off-by: Joe Stringer
---
tests/ofproto.at |4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/tests/ofproto.at b/tests/ofproto.at
index dde715e..2f1df2d 100644
--- a/tests
Ah, sorry I should have replied to this series; Andy looked through a
separate posting of them and I applied them.
I'll prepare a patch to remove the "-A" from the grepping in these tests.
On 21 May 2014 07:56, Ben Pfaff wrote:
> On Tue, May 13, 2014 at 03:45:50PM +1200,
On 21 May 2014 06:40, Ben Pfaff wrote:
>
> > I'm not sure whether you would want it in this in a separate patch, but
> it
> > appears that we only try to detect "already_dumped" for flows that
> haven't
> > been used before. Flows which have been used before and are duplicated
> will
> > execute
On 21 May 2014 10:05, Ryan Wilson 76511 wrote:
> Hey Andy,
>
> >At a high level, ofgroup struct current has rwlock that essentially
> >solving
> >the same problem as the ref count proposed in this patch. It would be
> >better
> >it seems confusing if we use both method together.
> >
> >Looking a
On 21 May 2014 10:57, Ben Pfaff wrote:
> On Wed, May 21, 2014 at 09:43:22AM +1200, Joe Stringer wrote:
> > Ah, sorry I should have replied to this series; Andy looked through a
> > separate posting of them and I applied them.
> >
> > I'll prepare a patch to remo
On 21 May 2014 10:33, Joe Stringer wrote:
> On 21 May 2014 10:05, Ryan Wilson 76511 wrote:
>
> Hey Andy,
>>
>> >At a high level, ofgroup struct current has rwlock that essentially
>> >solving
>> >the same problem as the ref count proposed in this
On 21 May 2014 02:52, Ryan Wilson wrote:
> When adding support for Openflow group and bucket stats, a group entry is
> added
> to the xlate cache. If the group is removed from ofproto configuration, we
> need
> to guarantee the group exists in the case where stats are attributed to the
> group by
On 20 May 2014 17:25, Ryan Wilson wrote:
> Ok turns out my Openflow rules weren't totally correct (they were flooding
> all ports like a hub instead of forwarding properly). After adjusting them,
> I achieved equivalent performance with and without my upcall patch (both
> achieved 161-162 trans/s
On 21 May 2014 14:25, Andy Zhou wrote:
> On Tue, May 20, 2014 at 5:27 PM, Joe Stringer
> wrote:
> > On 21 May 2014 10:33, Joe Stringer wrote:
> >>
> >> On 21 May 2014 10:05, Ryan Wilson 76511 wrote:
> >>
> >>> Hey Andy,
> >>>
This no longer applies. I plan to send a fresh version.
On 20 May 2014 09:12, Joe Stringer wrote:
> Rather than setting and resetting the 'mark' field in the ukey, this
> patch introduces a seq to track whether a flow has been seen during the
> most recent dump. This
Rather than setting and resetting the 'mark' field in the ukey, this
patch introduces a seq to track whether a flow has been seen during the
most recent dump. This tidies the code and simplifies the logic for
detecting when flows are duplicated from the datapath.
Signed-off-by: Jo
"failed to flow_get (No such file or directory) skb_priority(0),..."
"failed to flow_del (No such file or directory) skb_priority(0),..."
This patch fixes the issue again.
Signed-off-by: Joe Stringer
---
ofproto/ofproto-dpif-upcall.c |6 --
1 file changed, 4 insert
xlate_cache to attribute stats
when full revalidation is required.
Signed-off-by: Joe Stringer
---
ofproto/ofproto-dpif-upcall.c | 16 ++--
1 file changed, 10 insertions(+), 6 deletions(-)
diff --git a/ofproto/ofproto-dpif-upcall.c b/ofproto/ofproto-dpif-upcall.c
index 3e8dda5
; with HEAD = 5a87054c2d832d0e10b30a1f223707acb8efbeb7. This commit is from
> yesterday, so it includes your fix
> (73a3c4757e596ff156d40f41496a0264373e5bc4).
>
> Ryan
>
> From: Joe Stringer
> Date: Tuesday, May 20, 2014 7:06 PM
> To: Ryan Wilson
> Cc: Ale
On 22 May 2014 06:12, Ben Pfaff wrote:
> On Tue, May 20, 2014 at 12:49:09PM +1200, Joe Stringer wrote:
> > This was previously logging that there was no such function, but passing
> > the test anyway.
> >
> > Signed-off-by: Joe Stringer
>
> Good catch.
>
Thanks, I applied this to master and branch-2.3.
On 21 May 2014 18:14, Alex Wang wrote:
> Tested, and no more flow_del warning,
>
>
> Acked-by: Alex Wang
>
>
> On Tue, May 20, 2014 at 8:45 PM, Joe Stringer wrote:
>
>> Commit 73a3c4757e59 (revalidator: Preven
gt; And this change deserves further discussion~.
>
>
> On Tue, May 20, 2014 at 8:45 PM, Joe Stringer wrote:
>
>> One of the reasons that xlate_cache was introduced was to ensure that
>> statistics were attributed to the correct rules and interfaces according
>> to
So just to make sure I'm understanding correctly,
1) revalidator calls dpif_flow_dump_next_may_destroy_keys(), which
indicates that it will not destroy the buffer.
2) revalidator calls dpif_flow_dump_next(), which dumps a flow without
actions. The most likely case is that this is the last flow in
Acked-by: Joe Stringer
On 22 May 2014 12:27, Alex Wang wrote:
> Commit b256dc525c8 (ofproto-dpif-xlate: Cache xlate_actions() effects.)
> caches the variables needed for refreshing mac-learning table in
> xlate_normal(). Wherein, the cache entry always records reference to
>
ested briefly, didn't check it fixes the original bug.
Acked-by: Joe Stringer
On 22 May 2014 11:26, Ben Pfaff wrote:
> Commit 1340ce0c175 (ofproto-dpif-upcall: Avoid use-after-free in
> revalidate() corner cases.) fixed one use-after-free error in revalidate(),
> but missed one more su
Signed-off-by: Joe Stringer
---
ofproto/ofproto-dpif-xlate.c |2 --
1 file changed, 2 deletions(-)
diff --git a/ofproto/ofproto-dpif-xlate.c b/ofproto/ofproto-dpif-xlate.c
index a87db54..e0b10cb 100644
--- a/ofproto/ofproto-dpif-xlate.c
+++ b/ofproto/ofproto-dpif-xlate.c
@@ -266,7 +266,6
701 - 800 of 2028 matches
Mail list logo
