Re: [ovs-dev] [ovs-dev, PATCHv6] netdev-afxdp: add new netdev type for AF_XDP.

2019-04-26 Thread Ilya Maximets 
On 25.04.2019 2:47, William Tu wrote:
> diff --git a/lib/netdev.c b/lib/netdev.c
> index 7d7ecf6f0946..c30016b34033 100644
> --- a/lib/netdev.c
> +++ b/lib/netdev.c
> @@ -145,6 +145,7 @@ netdev_initialize(void)
>  netdev_register_provider(&netdev_linux_class);
>  netdev_register_provider(&netdev_internal_class);
>  netdev_register_provider(&netdev_tap_class);
> +netdev_register_provider(&netdev_afxdp_class);

It's better to move this under #ifdef HAVE_AF_XDP to not register
netdev class that will not work. Otherwise it'll be confusing for
users because OVS will report afxdp in a list of supported port
types and will allow port creation.

>  netdev_vport_tunnel_register();
>  #endif
>  #if defined(__FreeBSD__) || defined(__NetBSD__)
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] [PATCH 1/1] dpdk: Use DPDK 18.11.1 release.

2019-04-26 Thread Ilya Maximets 
On 25.04.2019 13:42, Ian Stokes wrote:
> On 4/25/2019 11:09 AM, Ilya Maximets wrote:
>> "On 24.04.2019 20:03, Ian Stokes wrote:
>>> Modify travis linux build script to use the latest
>>> DPDK stable release 18.11.1. Update docs for latest
>>> DPDK stable releases.
>>>
>>> Signed-off-by: Ian Stokes 
>>> ---
>>>   .travis/linux-build.sh   | 2 +-
>>>   Documentation/faq/releases.rst   | 4 ++--
>>>   Documentation/intro/install/dpdk.rst | 8 
>>>   Documentation/topics/dpdk/vhost-user.rst | 6 +++---
>>>   4 files changed, 10 insertions(+), 10 deletions(-)
>>
>> Do we need a NEWS update for this change?
>> 18.11.1 should not have any user-visible changes from the OVS
>> point of view, however upcoming 18.11.2 will have in case we'll
>> merge vhost destroy_connection related patch. The questions are:
>> Are we going to add stable release updates to NEWS?
> We can, if people think it's of use. It's now mentioned in the release notes 
> of 18.11.1 that OVS DPDK was validated, so adding an entry into NEWS for the 
> next OVS dot release could help push users to use the correct DPDK to OVS 
> release mapping. To date we still see users on xx.yy.0 LTS releases which I'd 
> like to encourage to move on.

OK. Sounds good.

> 
>> Do we need to mention stable releases that has no direct impact
>> on OVS itself?
> If it has no impact then we can just add an item into NEWS saying it's been 
> validated and is supported.

OK.

> If there is a known impact then we should detail the nature of that impact, 
> typically NEWS has an entry of 'various bug fixes' or something to that 
> effect, are you thinking we should provide specifics of the bug fix that 
> affects users as part of NEWS from now on? I don't see the harm once we are 
> aware of it.

Yes, we can do that. Another option:
* If there is no visible impact on OVS --> mention that new stable release 
validated.
* If there is some impact (feature depends on a bug fix in a new stable DPDK)
  --> bump the minimal supported DPDK version.

Like this:
NEWS:
 * DPDK:
   - OVS validated with DPDK 18.11.1 which is recommended to use.

 * DPDK:
   - DPDK 18.11.2 is a new minimal supported version.
   - DPDK 18.11.1 and lower is no longer supported.

The key point here is that we relay on bug fixes in a new stable DPDK and able
to merge new functionality/fixes that depends on them in OVS.

Thoughts?

Best regards, Ilya Maximets.
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

[ovs-dev] [PATCH] system-offloads-traffic.at: Fix requesting HW oofloaded flows from veth.

2019-04-26 Thread Ilya Maximets 
veth pair doesn't offload anything to HW. i.e. we should use 'tc' type
while requesting flows. 'offloaded' kept just in case to not update the
test if veths will be HW offloaded someday.

Additionally fixed missed for unknown reason 'ipv4' fields. Also
dropped stripping of the errors from log.

Fixes test:

  2: offloads - ping between two ports - offloads enabled ok

CC: Gavi Teitz 
Fixes: d63ca5329ff9 ("dpctl: Properly reflect a rule's offloaded to HW state")
Signed-off-by: Ilya Maximets 
---
 tests/system-offloads-traffic.at | 13 ++---
 1 file changed, 6 insertions(+), 7 deletions(-)

diff --git a/tests/system-offloads-traffic.at b/tests/system-offloads-traffic.at
index 1f8079344..27997db60 100644
--- a/tests/system-offloads-traffic.at
+++ b/tests/system-offloads-traffic.at
@@ -51,18 +51,17 @@ AT_CHECK([ovs-appctl dpctl/dump-flows], [0], [ignore])
 NS_CHECK_EXEC([at_ns0], [ping -q -c 10 -i 0.1 -w 2 10.1.1.2 | FORMAT_PING], 
[0], [dnl
 10 packets transmitted, 10 received, 0% packet loss, time 0ms
 ])
-
 AT_CHECK([ovs-appctl dpctl/dump-flows | grep "eth_type(0x0800)" | 
DUMP_CLEAN_SORTED], [0], [dnl
-in_port(2),eth(macs),eth_type(0x0800), packets:9, bytes:756, used:0.001s, 
actions:output
-in_port(3),eth(macs),eth_type(0x0800), packets:9, bytes:756, used:0.001s, 
actions:output
+in_port(2),eth(macs),eth_type(0x0800),ipv4(frag=no), packets:9, bytes:756, 
used:0.001s, actions:output
+in_port(3),eth(macs),eth_type(0x0800),ipv4(frag=no), packets:9, bytes:756, 
used:0.001s, actions:output
 ])
 
 AT_CHECK([ovs-appctl dpctl/dump-flows type=ovs | grep "eth_type(0x0800)" | 
DUMP_CLEAN_SORTED], [0], [])
 
-AT_CHECK([ovs-appctl dpctl/dump-flows type=offloaded | grep "eth_type(0x0800)" 
| DUMP_CLEAN_SORTED], [0], [dnl
-in_port(2),eth(macs),eth_type(0x0800), packets:9, bytes:756, used:0.001s, 
actions:output
-in_port(3),eth(macs),eth_type(0x0800), packets:9, bytes:756, used:0.001s, 
actions:output
+AT_CHECK([ovs-appctl dpctl/dump-flows type=tc,offloaded | grep 
"eth_type(0x0800)" | DUMP_CLEAN_SORTED], [0], [dnl
+in_port(2),eth(macs),eth_type(0x0800),ipv4(frag=no), packets:9, bytes:756, 
used:0.001s, actions:output
+in_port(3),eth(macs),eth_type(0x0800),ipv4(frag=no), packets:9, bytes:756, 
used:0.001s, actions:output
 ])
 
-OVS_TRAFFIC_VSWITCHD_STOP('/.*/d')
+OVS_TRAFFIC_VSWITCHD_STOP
 AT_CLEANUP
-- 
2.17.1

___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] [PATCH] system-offloads-traffic.at: Fix requesting HW oofloaded flows from veth.

2019-04-26 Thread Ilya Maximets 
s/oofloaded/offloaded/
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

[ovs-dev] [PATCH v2] system-offloads-traffic.at: Fix requesting HW offloaded flows from veth.

2019-04-26 Thread Ilya Maximets 
veth pair doesn't offload anything to HW. i.e. we should use 'tc' type
while requesting flows. 'offloaded' kept just in case to not update the
test if veths will be HW offloaded someday.

Additionally fixed missed for unknown reason 'ipv4' fields. Also
dropped stripping of the errors from log.

Fixes test:

  2: offloads - ping between two ports - offloads enabled ok

CC: Gavi Teitz 
Fixes: d63ca5329ff9 ("dpctl: Properly reflect a rule's offloaded to HW state")
Signed-off-by: Ilya Maximets 
---

Version 2:

  * Fixed typos.

 tests/system-offloads-traffic.at | 12 ++--
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/tests/system-offloads-traffic.at b/tests/system-offloads-traffic.at
index 1f8079344..379a8a5e9 100644
--- a/tests/system-offloads-traffic.at
+++ b/tests/system-offloads-traffic.at
@@ -53,16 +53,16 @@ NS_CHECK_EXEC([at_ns0], [ping -q -c 10 -i 0.1 -w 2 10.1.1.2 
| FORMAT_PING], [0],
 ])
 
 AT_CHECK([ovs-appctl dpctl/dump-flows | grep "eth_type(0x0800)" | 
DUMP_CLEAN_SORTED], [0], [dnl
-in_port(2),eth(macs),eth_type(0x0800), packets:9, bytes:756, used:0.001s, 
actions:output
-in_port(3),eth(macs),eth_type(0x0800), packets:9, bytes:756, used:0.001s, 
actions:output
+in_port(2),eth(macs),eth_type(0x0800),ipv4(frag=no), packets:9, bytes:756, 
used:0.001s, actions:output
+in_port(3),eth(macs),eth_type(0x0800),ipv4(frag=no), packets:9, bytes:756, 
used:0.001s, actions:output
 ])
 
 AT_CHECK([ovs-appctl dpctl/dump-flows type=ovs | grep "eth_type(0x0800)" | 
DUMP_CLEAN_SORTED], [0], [])
 
-AT_CHECK([ovs-appctl dpctl/dump-flows type=offloaded | grep "eth_type(0x0800)" 
| DUMP_CLEAN_SORTED], [0], [dnl
-in_port(2),eth(macs),eth_type(0x0800), packets:9, bytes:756, used:0.001s, 
actions:output
-in_port(3),eth(macs),eth_type(0x0800), packets:9, bytes:756, used:0.001s, 
actions:output
+AT_CHECK([ovs-appctl dpctl/dump-flows type=tc,offloaded | grep 
"eth_type(0x0800)" | DUMP_CLEAN_SORTED], [0], [dnl
+in_port(2),eth(macs),eth_type(0x0800),ipv4(frag=no), packets:9, bytes:756, 
used:0.001s, actions:output
+in_port(3),eth(macs),eth_type(0x0800),ipv4(frag=no), packets:9, bytes:756, 
used:0.001s, actions:output
 ])
 
-OVS_TRAFFIC_VSWITCHD_STOP('/.*/d')
+OVS_TRAFFIC_VSWITCHD_STOP
 AT_CLEANUP
-- 
2.17.1

___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] OVS/OVN Split initial commits are complete

2019-04-26 Thread Mark Michelson 

On 4/25/19 3:51 PM, Ben Pfaff wrote:

On Thu, Apr 25, 2019 at 08:38:06AM -0400, Mark Michelson wrote:

Yesterday, I once again performed the task of splitting OVN from OVS (I've
gotten pretty good at it now). This time, rather than living in a personal
repo, you can find the ovn repo at https://github.com/ovn-org/ovn.git


I noticed that most or all of your commits lack sign-offs.  Oops?



Oops indeed. I tend to add the sign-off when using `git format-patch` 
when I send an e-mail. In this case since I pushed the commits directly, 
I skipped over that step.

___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] OVS/OVN Split initial commits are complete

2019-04-26 Thread Mark Michelson 

On 4/26/19 1:44 AM, Numan Siddique wrote:



On Fri, Apr 26, 2019 at 1:56 AM Ben Pfaff > wrote:


On Thu, Apr 25, 2019 at 12:51:29PM -0700, Ben Pfaff wrote:
 > On Thu, Apr 25, 2019 at 08:38:06AM -0400, Mark Michelson wrote:
 > > Yesterday, I once again performed the task of splitting OVN
from OVS (I've
 > > gotten pretty good at it now). This time, rather than living in
a personal
 > > repo, you can find the ovn repo at
https://github.com/ovn-org/ovn.git
 >
 > I noticed that most or all of your commits lack sign-offs.  Oops?

By the way, I guess that you should be the main person pushing to this
repo for now, so I'll review some patches but I'll plan to mostly ack
them rather than applying them to the new repo.



Thanks Mark for all the work.
Shall we consider that all the OVN patches will now be applied to the 
new repo ?


OVN is buildable and testable in the new repo, so developers should be 
able to do their work against the new repo.



If so, then all the present OVN patches under review should be resubmitted ?


Yes, unfortunately. The structure of the new repo should allow for 
patches to be applied fairly easily, as long as the patches are only for 
ovn. The ovn/ subdirectory has been promoted to the top level. So it 
should be possible to apply a patch using `git apply -p2 ` or 
`patch -p2 < `


If your patch touches both OVS and OVN, then the patch will need to be 
split so that the OVS parts are submitted to OVS and the OVN parts are 
submitted to OVN. Note that a policy for doing this sort of development 
is in the TODO_SPLIT.rst file.




Thanks
Numan

___
dev mailing list
d...@openvswitch.org 
https://mail.openvswitch.org/mailman/listinfo/ovs-dev



___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] [PATCH] OVN: Fix failures after OVS/OVN split

2019-04-26 Thread Mark Michelson 

On 4/25/19 4:45 PM, Ben Pfaff wrote:

On Thu, Apr 25, 2019 at 04:33:03PM -0400, Brian Haley wrote:

On 4/25/19 4:26 PM, Ben Pfaff wrote:

On Thu, Apr 25, 2019 at 04:08:15PM -0400, Brian Haley wrote:

Added TODO_SPLIT.rst to Makefile.am to avoid an error
during build.

Removed a section from config-h-check section of Makefile.am
that was no longer relevant.

Signed-off-by: Brian Haley 
---
   Makefile.am | 8 +---
   1 file changed, 1 insertion(+), 7 deletions(-)

diff --git a/Makefile.am b/Makefile.am
index 1b492a314..2bcdb0a76 100644
--- a/Makefile.am
+++ b/Makefile.am
@@ -106,6 +106,7 @@ EXTRA_DIST = \
Vagrantfile-FreeBSD \
.mailmap \
TODO.rst \
+   TODO_SPLIT.rst \
ovn-architecture.7.xml \
ovn-nb.ovsschema \
ovn-nb.xml \
@@ -263,13 +264,6 @@ config-h-check:
  echo "See above for list of violations of the rule that"; \
  echo "every C source file must #include ."; \
  exit 1; \
-   fi; \
-   if grep '#include' include/openvswitch/*.h | \
-   grep -vE '(<.*>)|("openvswitch)|("openflow)'; \
-   then \
- echo "See above for list of violations of the rule that"; \
- echo "public openvswitch header file should not include internal 
library."; \
- exit 1; \
fi
   .PHONY: config-h-check


I wonder whether there should be a similar rule for include/ovn, though?


The top-level include/ only had a single directory, openflow, which itself
was empty.  I now see there is an ovs/include directory, so maybe that
second part needs to change instead of being deleted?  It was only
generating a non-fatal warning, the TODO_SPLIT.rst was causing a
dist-hook-git error.


There might be more than one thing going on here; the principle is just
that this is trying to check that public headers don't include
non-public headers.  Probably Mark has an idea of what should really be
done.


Currently, there's not really a concept of public vs. non-public headers 
in OVN. Common headers used amongst the different apps are in lib/ . For 
the time, it's fine to just delete the section.



___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev



___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] [ovs-dev, PATCHv6] netdev-afxdp: add new netdev type for AF_XDP.

2019-04-26 Thread William Tu 
On Fri, Apr 26, 2019 at 2:46 AM Ilya Maximets 
wrote:

> On 25.04.2019 2:47, William Tu wrote:
> > diff --git a/lib/netdev.c b/lib/netdev.c
> > index 7d7ecf6f0946..c30016b34033 100644
> > --- a/lib/netdev.c
> > +++ b/lib/netdev.c
> > @@ -145,6 +145,7 @@ netdev_initialize(void)
> >  netdev_register_provider(&netdev_linux_class);
> >  netdev_register_provider(&netdev_internal_class);
> >  netdev_register_provider(&netdev_tap_class);
> > +netdev_register_provider(&netdev_afxdp_class);
>
> It's better to move this under #ifdef HAVE_AF_XDP to not register
> netdev class that will not work. Otherwise it'll be confusing for
> users because OVS will report afxdp in a list of supported port
> types and will allow port creation.
>

Yes, thanks.
Will do it next version.
William

>
> >  netdev_vport_tunnel_register();
> >  #endif
> >  #if defined(__FreeBSD__) || defined(__NetBSD__)
>
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] OVS/OVN Split initial commits are complete

2019-04-26 Thread Mark Michelson 

On 4/26/19 1:47 AM, Numan Siddique wrote:



On Fri, Apr 26, 2019 at 11:14 AM Numan Siddique > wrote:




On Fri, Apr 26, 2019 at 1:56 AM Ben Pfaff mailto:b...@ovn.org>> wrote:

On Thu, Apr 25, 2019 at 12:51:29PM -0700, Ben Pfaff wrote:
 > On Thu, Apr 25, 2019 at 08:38:06AM -0400, Mark Michelson wrote:
 > > Yesterday, I once again performed the task of splitting OVN
from OVS (I've
 > > gotten pretty good at it now). This time, rather than
living in a personal
 > > repo, you can find the ovn repo at
https://github.com/ovn-org/ovn.git
 >
 > I noticed that most or all of your commits lack sign-offs.  Oops?

By the way, I guess that you should be the main person pushing
to this
repo for now, so I'll review some patches but I'll plan to
mostly ack
them rather than applying them to the new repo.



Thanks Mark for all the work.
Shall we consider that all the OVN patches will now be applied to
the new repo ?
If so, then all the present OVN patches under review should be
resubmitted ?

Thanks
Numan


I found an issue when compiling if I configure like below

$mkdir _gcc
$ cd _gcc
$.../configure --enable-Werror
$make -C _gcc -j5
...
libtool: link: gcc -Wstrict-prototypes -Wall -Wextra -Wno-sign-compare 
-Wpointer-arith -Wformat -Wformat-security -Wswitch-enum 
-Wunused-parameter -Wbad-function-cast -Wcast-align -Wstrict-prototypes 
-Wold-style-definition -Wmissing-prototypes -Wmissing-field-initializers 
-fno-strict-aliasing -Wswitch-bool -Wlogical-not-parentheses 
-Wsizeof-array-argument -Wbool-compare -Wshift-negative-value 
-Wduplicated-cond -Wshadow -Wmultistatement-macros -Wcast-align=strict 
-Werror -Werror -g -O2 -o vswitchd/ovs-vswitchd vswitchd/bridge.o 
vswitchd/ovs-vswitchd.o vswitchd/system-stats.o vswitchd/xenserver.o  
ofproto/.libs/libofproto.a 
/home/nusiddiq/workspace_cpp/ovn-org/ovn/_gcc/ovs/lib/.libs/libsflow.a 
lib/.libs/libsflow.a lib/.libs/libopenvswitch.a -lssl -lcrypto -lcap-ng 
-latomic -lpthread -lrt -lm -lunbound

touch -c manpage-check
make[4]: Leaving directory 
'/home/nusiddiq/workspace_cpp/ovn-org/ovn/_gcc/ovs'
make[3]: Leaving directory 
'/home/nusiddiq/workspace_cpp/ovn-org/ovn/_gcc/ovs'
make[2]: Leaving directory 
'/home/nusiddiq/workspace_cpp/ovn-org/ovn/_gcc/ovs'

make[2]: Entering directory '/home/nusiddiq/workspace_cpp/ovn-org/ovn/_gcc'
make[2]: *** No rule to make target '../ovs/ovsdb/libovsdb.la 
', needed by 'utilities/ovn-nbctl'.  Stop.

make[2]: *** Waiting for unfinished jobs
make[2]: Leaving directory '/home/nusiddiq/workspace_cpp/ovn-org/ovn/_gcc'
make[1]: *** [Makefile:3346: all-recursive] Error 1
make[1]: Leaving directory '/home/nusiddiq/workspace_cpp/ovn-org/ovn/_gcc'
make: *** [Makefile:2216: all] Error 2
make: Leaving directory '/home/nusiddiq/workspace_cpp/ovn-org/ovn/_gcc'
...

We need to fix this.

Thanks
Numan


I tried this and actually got a slightly different build error:

/bin/sh ./libtool  --tag=CC   --mode=link gcc -Wstrict-prototypes -Wall 
-Wextra -Wno-sign-compare -Wpointer-arith -Wformat -Wformat-security 
-Wswitch-enum -Wunused-parameter -Wbad-function-cast -Wcast-align 
-Wstrict-prototypes -Wold-style-definition -Wmissing-prototypes 
-Wmissing-field-initializers -fno-strict-aliasing -Wswitch-bool 
-Wlogical-not-parentheses -Wsizeof-array-argument -Wbool-compare 
-Wshift-negative-value -Wduplicated-cond -Wshadow-g -O2 -o 
utilities/ovn-nbctl utilities/ovn-nbctl.o lib/libovn.la 
../ovs/ovsdb/libovsdb.la ../ovs/lib/libopenvswitch.la -lpthread -lrt -lm 
 -lunbound
libtool:   error: cannot find the library '../ovs/ovsdb/libovsdb.la' or 
unhandled argument '../ovs/ovsdb/libovsdb.la'


In both cases though, it appears the issue is that it is trying to 
either build or find ovs libraries from the top-level source directory, 
but they have been built in the _gcc subdirectory. I'm looking into this 
to try to fix it.

___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

[ovs-dev] [PATCH] OVN: Fix build when using a separate build directory.

2019-04-26 Thread Mark Michelson 
The linker was searching for built libraries in the source directory
rather than in the build directory. This fixes the problem.

Reported-by: Numan Siddique 
Signed-off-by: Mark Michelson 
---
 configure.ac | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/configure.ac b/configure.ac
index e23df2647..0033d7fdc 100644
--- a/configure.ac
+++ b/configure.ac
@@ -77,8 +77,8 @@ AC_SUBST([LT_AGE])
 
 AC_SUBST([ovs_srcdir], ['${top_srcdir}/ovs'])
 AC_SUBST([ovs_builddir], ['${top_builddir}/ovs'])
-AC_SUBST([ovs_libdir], ['${top_srcdir}/ovs/lib'])
-AC_SUBST([ovsdb_libdir], ['${top_srcdir}/ovs/ovsdb'])
+AC_SUBST([ovs_libdir], ['${top_builddir}/ovs/lib'])
+AC_SUBST([ovsdb_libdir], ['${top_builddir}/ovs/ovsdb'])
 AC_SUBST([ovs_mandir], ['${top_srcdir}/ovs/'])
 
 AC_SEARCH_LIBS([pow], [m])
-- 
2.14.5

___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] [PATCH] OVN: Fix build when using a separate build directory.

2019-04-26 Thread Numan Siddique 
On Fri, Apr 26, 2019, 9:55 PM Mark Michelson  wrote:

> The linker was searching for built libraries in the source directory
> rather than in the build directory. This fixes the problem.
>
> Reported-by: Numan Siddique 
>

Hi Mark

Can you plz correct the email id - nusid...@redhat.com :)

Thanks
Numan

Signed-off-by: Mark Michelson 
> ---
>  configure.ac | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/configure.ac b/configure.ac
> index e23df2647..0033d7fdc 100644
> --- a/configure.ac
> +++ b/configure.ac
> @@ -77,8 +77,8 @@ AC_SUBST([LT_AGE])
>
>  AC_SUBST([ovs_srcdir], ['${top_srcdir}/ovs'])
>  AC_SUBST([ovs_builddir], ['${top_builddir}/ovs'])
> -AC_SUBST([ovs_libdir], ['${top_srcdir}/ovs/lib'])
> -AC_SUBST([ovsdb_libdir], ['${top_srcdir}/ovs/ovsdb'])
> +AC_SUBST([ovs_libdir], ['${top_builddir}/ovs/lib'])
> +AC_SUBST([ovsdb_libdir], ['${top_builddir}/ovs/ovsdb'])
>  AC_SUBST([ovs_mandir], ['${top_srcdir}/ovs/'])
>
>  AC_SEARCH_LIBS([pow], [m])
> --
> 2.14.5
>
> ___
> dev mailing list
> d...@openvswitch.org
> https://mail.openvswitch.org/mailman/listinfo/ovs-dev
>
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] OVN: Fix build when using a separate build directory.

2019-04-26 Thread 0-day Robot 
Bleep bloop.  Greetings Mark Michelson, I am a robot and I have tried out your 
patch.
Thanks for your contribution.

I encountered some error that I wasn't expecting.  See the details below.


git-am:
fatal: sha1 information is lacking or useless (configure.ac).
Repository lacks necessary blobs to fall back on 3-way merge.
Cannot fall back to three-way merge.
Patch failed at 0001 OVN: Fix build when using a separate build directory.
The copy of the patch that failed is found in:
   
/var/lib/jenkins/jobs/upstream_build_from_pw/workspace/.git/rebase-apply/patch
When you have resolved this problem, run "git am --resolved".
If you prefer to skip this patch, run "git am --skip" instead.
To restore the original branch and stop patching, run "git am --abort".


Please check this out.  If you feel there has been an error, please email 
acon...@bytheb.org

Thanks,
0-day Robot
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

[ovs-dev] [PATCH] MAINTAINERS: Add Ilya Maximets.

2019-04-26 Thread Ben Pfaff 
Ilya was elected by the Open vSwitch committers on Thursday.  Welcome to
the team, Ilya!

CC: Ilya Maximets 
Signed-off-by: Ben Pfaff 
---
 MAINTAINERS.rst | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/MAINTAINERS.rst b/MAINTAINERS.rst
index e5028400f72d..760f5c9736f0 100644
--- a/MAINTAINERS.rst
+++ b/MAINTAINERS.rst
@@ -57,6 +57,8 @@ This is the current list of active Open vSwitch committers:
  - g...@ovn.org
* - Ian Stokes
  - isto...@ovn.org
+   * - Ilya Maximets
+ - i.maxim...@samsung.com
* - Jarno Rajahalme
  - ja...@ovn.org
* - Jesse Gross
-- 
2.20.1

___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

[ovs-dev] [PATCH] datapath: Fix compiling error for latest Linux 4.14 kernel

2019-04-26 Thread Yifeng Sun 
CC: Greg Rose 
Fixes: f72469405eec9 ("datapath: meter: Use struct_size() in kzalloc()")
Signed-off-by: Yifeng Sun 
---
 acinclude.m4   | 2 ++
 datapath/linux/compat/include/linux/overflow.h | 2 +-
 datapath/meter.c   | 1 +
 3 files changed, 4 insertions(+), 1 deletion(-)

diff --git a/acinclude.m4 b/acinclude.m4
index 1607d5f4b1d9..548cbe9ef100 100644
--- a/acinclude.m4
+++ b/acinclude.m4
@@ -920,6 +920,8 @@ AC_DEFUN([OVS_CHECK_LINUX_COMPAT], [
 [OVS_DEFINE([HAVE_INET_FRAGS_RND])])
   OVS_GREP_IFELSE([$KSRC/include/linux/overflow.h], [__LINUX_OVERFLOW_H],
   [OVS_DEFINE([HAVE_OVERFLOW_H])])
+  OVS_GREP_IFELSE([$KSRC/include/linux/overflow.h], [struct_size],
+  [OVS_DEFINE([HAVE_STRUCT_SIZE])])
   OVS_GREP_IFELSE([$KSRC/include/linux/mm.h], [kvmalloc_array],
   [OVS_DEFINE([HAVE_KVMALLOC_ARRAY])])
   OVS_GREP_IFELSE([$KSRC/include/linux/mm.h], [kvmalloc_node],
diff --git a/datapath/linux/compat/include/linux/overflow.h 
b/datapath/linux/compat/include/linux/overflow.h
index ff84356db16c..8f1b5b005491 100644
--- a/datapath/linux/compat/include/linux/overflow.h
+++ b/datapath/linux/compat/include/linux/overflow.h
@@ -1,5 +1,5 @@
 /* SPDX-License-Identifier: GPL-2.0 OR MIT */
-#ifdef HAVE_OVERFLOW_H
+#if defined(HAVE_OVERFLOW_H) && defined(HAVE_STRUCT_SIZE)
 #include_next 
 #else
 #ifndef __LINUX_OVERFLOW_H
diff --git a/datapath/meter.c b/datapath/meter.c
index 51ec1495033a..eda14682fb96 100644
--- a/datapath/meter.c
+++ b/datapath/meter.c
@@ -13,6 +13,7 @@
 #include 
 #include 
 #include 
+#include 
 #include 
 #include 
 
-- 
2.7.4

___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] [PATCH] OVN: Fix build when using a separate build directory.

2019-04-26 Thread Ben Pfaff 
On Fri, Apr 26, 2019 at 12:14:06PM -0400, Mark Michelson wrote:
> The linker was searching for built libraries in the source directory
> rather than in the build directory. This fixes the problem.
> 
> Reported-by: Numan Siddique 
> Signed-off-by: Mark Michelson 

I did not test this but it looks perfectly reasonable.

Acked-by: Ben Pfaff 
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] OVS/OVN Split initial commits are complete

2019-04-26 Thread Ben Pfaff 
On Fri, Apr 26, 2019 at 11:14:23AM +0530, Numan Siddique wrote:
> On Fri, Apr 26, 2019 at 1:56 AM Ben Pfaff  wrote:
> 
> > On Thu, Apr 25, 2019 at 12:51:29PM -0700, Ben Pfaff wrote:
> > > On Thu, Apr 25, 2019 at 08:38:06AM -0400, Mark Michelson wrote:
> > > > Yesterday, I once again performed the task of splitting OVN from OVS
> > (I've
> > > > gotten pretty good at it now). This time, rather than living in a
> > personal
> > > > repo, you can find the ovn repo at https://github.com/ovn-org/ovn.git
> > >
> > > I noticed that most or all of your commits lack sign-offs.  Oops?
> >
> > By the way, I guess that you should be the main person pushing to this
> > repo for now, so I'll review some patches but I'll plan to mostly ack
> > them rather than applying them to the new repo.
> >
> 
> 
> Thanks Mark for all the work.
> Shall we consider that all the OVN patches will now be applied to the new
> repo ?
> If so, then all the present OVN patches under review should be resubmitted ?

We need to pick a date (or a trigger) for freezing the OVN parts of OVS.
I don't know what that date should be.  Thoughts?
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] [PATCH] MAINTAINERS: Add Ilya Maximets.

2019-04-26 Thread Justin Pettit 
Acked-by: Justin Pettit 

--Justin


> On Apr 26, 2019, at 10:52 AM, Ben Pfaff  wrote:
> 
> Ilya was elected by the Open vSwitch committers on Thursday.  Welcome to
> the team, Ilya!
> 
> CC: Ilya Maximets 
> Signed-off-by: Ben Pfaff 
> ---
> MAINTAINERS.rst | 2 ++
> 1 file changed, 2 insertions(+)
> 
> diff --git a/MAINTAINERS.rst b/MAINTAINERS.rst
> index e5028400f72d..760f5c9736f0 100644
> --- a/MAINTAINERS.rst
> +++ b/MAINTAINERS.rst
> @@ -57,6 +57,8 @@ This is the current list of active Open vSwitch committers:
>  - g...@ovn.org
>* - Ian Stokes
>  - isto...@ovn.org
> +   * - Ilya Maximets
> + - i.maxim...@samsung.com
>* - Jarno Rajahalme
>  - ja...@ovn.org
>* - Jesse Gross
> -- 
> 2.20.1
> 
> ___
> dev mailing list
> d...@openvswitch.org
> https://mail.openvswitch.org/mailman/listinfo/ovs-dev

___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] [PATCH] MAINTAINERS: Add Ilya Maximets.

2019-04-26 Thread Ben Pfaff 
Thanks!  Applied to master.

On Fri, Apr 26, 2019 at 12:49:11PM -0700, Justin Pettit wrote:
> Acked-by: Justin Pettit 
> 
> --Justin
> 
> 
> > On Apr 26, 2019, at 10:52 AM, Ben Pfaff  wrote:
> > 
> > Ilya was elected by the Open vSwitch committers on Thursday.  Welcome to
> > the team, Ilya!
> > 
> > CC: Ilya Maximets 
> > Signed-off-by: Ben Pfaff 
> > ---
> > MAINTAINERS.rst | 2 ++
> > 1 file changed, 2 insertions(+)
> > 
> > diff --git a/MAINTAINERS.rst b/MAINTAINERS.rst
> > index e5028400f72d..760f5c9736f0 100644
> > --- a/MAINTAINERS.rst
> > +++ b/MAINTAINERS.rst
> > @@ -57,6 +57,8 @@ This is the current list of active Open vSwitch 
> > committers:
> >  - g...@ovn.org
> >* - Ian Stokes
> >  - isto...@ovn.org
> > +   * - Ilya Maximets
> > + - i.maxim...@samsung.com
> >* - Jarno Rajahalme
> >  - ja...@ovn.org
> >* - Jesse Gross
> > -- 
> > 2.20.1
> > 
> > ___
> > dev mailing list
> > d...@openvswitch.org
> > https://mail.openvswitch.org/mailman/listinfo/ovs-dev
> 
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] [PATCH] datapath: Fix compiling error for latest Linux 4.14 kernel

2019-04-26 Thread Gregory Rose 


On 4/26/2019 11:07 AM, Yifeng Sun wrote:

CC: Greg Rose 
Fixes: f72469405eec9 ("datapath: meter: Use struct_size() in kzalloc()")
Signed-off-by: Yifeng Sun 
---
  acinclude.m4   | 2 ++
  datapath/linux/compat/include/linux/overflow.h | 2 +-
  datapath/meter.c   | 1 +
  3 files changed, 4 insertions(+), 1 deletion(-)

diff --git a/acinclude.m4 b/acinclude.m4
index 1607d5f4b1d9..548cbe9ef100 100644
--- a/acinclude.m4
+++ b/acinclude.m4
@@ -920,6 +920,8 @@ AC_DEFUN([OVS_CHECK_LINUX_COMPAT], [
  [OVS_DEFINE([HAVE_INET_FRAGS_RND])])
OVS_GREP_IFELSE([$KSRC/include/linux/overflow.h], [__LINUX_OVERFLOW_H],
[OVS_DEFINE([HAVE_OVERFLOW_H])])
+  OVS_GREP_IFELSE([$KSRC/include/linux/overflow.h], [struct_size],
+  [OVS_DEFINE([HAVE_STRUCT_SIZE])])
OVS_GREP_IFELSE([$KSRC/include/linux/mm.h], [kvmalloc_array],
[OVS_DEFINE([HAVE_KVMALLOC_ARRAY])])
OVS_GREP_IFELSE([$KSRC/include/linux/mm.h], [kvmalloc_node],
diff --git a/datapath/linux/compat/include/linux/overflow.h 
b/datapath/linux/compat/include/linux/overflow.h
index ff84356db16c..8f1b5b005491 100644
--- a/datapath/linux/compat/include/linux/overflow.h
+++ b/datapath/linux/compat/include/linux/overflow.h
@@ -1,5 +1,5 @@
  /* SPDX-License-Identifier: GPL-2.0 OR MIT */
-#ifdef HAVE_OVERFLOW_H
+#if defined(HAVE_OVERFLOW_H) && defined(HAVE_STRUCT_SIZE)
  #include_next 
  #else
  #ifndef __LINUX_OVERFLOW_H
diff --git a/datapath/meter.c b/datapath/meter.c
index 51ec1495033a..eda14682fb96 100644
--- a/datapath/meter.c
+++ b/datapath/meter.c
@@ -13,6 +13,7 @@
  #include 
  #include 
  #include 
+#include 
  #include 
  #include 
  


Thanks for the patch Yifeng, it looks good but I'm curious because when 
I posted the original patch for this it was passing travis build 
checks.  Or I thought so at the time.  It bugs me if I missed something.


In any case I'll give it a quick check and regression test.

- Greg
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] [PATCH] datapath: Fix compiling error for latest Linux 4.14 kernel

2019-04-26 Thread Yifeng Sun 
Hi Greg,

I saw the compilation error for kernel 4.14.111 because `struct_size`
isn't defined in kernel source tree. I didn't dig into it though.

Thanks,
Yifeng

On Fri, Apr 26, 2019 at 1:13 PM Gregory Rose  wrote:
>
>
> On 4/26/2019 11:07 AM, Yifeng Sun wrote:
> > CC: Greg Rose 
> > Fixes: f72469405eec9 ("datapath: meter: Use struct_size() in kzalloc()")
> > Signed-off-by: Yifeng Sun 
> > ---
> >   acinclude.m4   | 2 ++
> >   datapath/linux/compat/include/linux/overflow.h | 2 +-
> >   datapath/meter.c   | 1 +
> >   3 files changed, 4 insertions(+), 1 deletion(-)
> >
> > diff --git a/acinclude.m4 b/acinclude.m4
> > index 1607d5f4b1d9..548cbe9ef100 100644
> > --- a/acinclude.m4
> > +++ b/acinclude.m4
> > @@ -920,6 +920,8 @@ AC_DEFUN([OVS_CHECK_LINUX_COMPAT], [
> >   [OVS_DEFINE([HAVE_INET_FRAGS_RND])])
> > OVS_GREP_IFELSE([$KSRC/include/linux/overflow.h], [__LINUX_OVERFLOW_H],
> > [OVS_DEFINE([HAVE_OVERFLOW_H])])
> > +  OVS_GREP_IFELSE([$KSRC/include/linux/overflow.h], [struct_size],
> > +  [OVS_DEFINE([HAVE_STRUCT_SIZE])])
> > OVS_GREP_IFELSE([$KSRC/include/linux/mm.h], [kvmalloc_array],
> > [OVS_DEFINE([HAVE_KVMALLOC_ARRAY])])
> > OVS_GREP_IFELSE([$KSRC/include/linux/mm.h], [kvmalloc_node],
> > diff --git a/datapath/linux/compat/include/linux/overflow.h 
> > b/datapath/linux/compat/include/linux/overflow.h
> > index ff84356db16c..8f1b5b005491 100644
> > --- a/datapath/linux/compat/include/linux/overflow.h
> > +++ b/datapath/linux/compat/include/linux/overflow.h
> > @@ -1,5 +1,5 @@
> >   /* SPDX-License-Identifier: GPL-2.0 OR MIT */
> > -#ifdef HAVE_OVERFLOW_H
> > +#if defined(HAVE_OVERFLOW_H) && defined(HAVE_STRUCT_SIZE)
> >   #include_next 
> >   #else
> >   #ifndef __LINUX_OVERFLOW_H
> > diff --git a/datapath/meter.c b/datapath/meter.c
> > index 51ec1495033a..eda14682fb96 100644
> > --- a/datapath/meter.c
> > +++ b/datapath/meter.c
> > @@ -13,6 +13,7 @@
> >   #include 
> >   #include 
> >   #include 
> > +#include 
> >   #include 
> >   #include 
> >
>
> Thanks for the patch Yifeng, it looks good but I'm curious because when
> I posted the original patch for this it was passing travis build
> checks.  Or I thought so at the time.  It bugs me if I missed something.
>
> In any case I'll give it a quick check and regression test.
>
> - Greg
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] [PATCH] datapath: Fix compiling error for latest Linux 4.14 kernel

2019-04-26 Thread Gregory Rose 



On 4/26/2019 1:18 PM, Yifeng Sun wrote:

Hi Greg,

I saw the compilation error for kernel 4.14.111 because `struct_size`
isn't defined in kernel source tree. I didn't dig into it though.


I see the problem then.  The .travis.yml file tests 4.14.63.  Please 
update the patch to include an update
for travis to set the 4.14 kernel being tested to 4.14.111.  That way 
we'll get travis to verify the build as

well.

Thanks!

- Greg




Thanks,
Yifeng

On Fri, Apr 26, 2019 at 1:13 PM Gregory Rose  wrote:


On 4/26/2019 11:07 AM, Yifeng Sun wrote:

CC: Greg Rose 
Fixes: f72469405eec9 ("datapath: meter: Use struct_size() in kzalloc()")
Signed-off-by: Yifeng Sun 
---
   acinclude.m4   | 2 ++
   datapath/linux/compat/include/linux/overflow.h | 2 +-
   datapath/meter.c   | 1 +
   3 files changed, 4 insertions(+), 1 deletion(-)

diff --git a/acinclude.m4 b/acinclude.m4
index 1607d5f4b1d9..548cbe9ef100 100644
--- a/acinclude.m4
+++ b/acinclude.m4
@@ -920,6 +920,8 @@ AC_DEFUN([OVS_CHECK_LINUX_COMPAT], [
   [OVS_DEFINE([HAVE_INET_FRAGS_RND])])
 OVS_GREP_IFELSE([$KSRC/include/linux/overflow.h], [__LINUX_OVERFLOW_H],
 [OVS_DEFINE([HAVE_OVERFLOW_H])])
+  OVS_GREP_IFELSE([$KSRC/include/linux/overflow.h], [struct_size],
+  [OVS_DEFINE([HAVE_STRUCT_SIZE])])
 OVS_GREP_IFELSE([$KSRC/include/linux/mm.h], [kvmalloc_array],
 [OVS_DEFINE([HAVE_KVMALLOC_ARRAY])])
 OVS_GREP_IFELSE([$KSRC/include/linux/mm.h], [kvmalloc_node],
diff --git a/datapath/linux/compat/include/linux/overflow.h 
b/datapath/linux/compat/include/linux/overflow.h
index ff84356db16c..8f1b5b005491 100644
--- a/datapath/linux/compat/include/linux/overflow.h
+++ b/datapath/linux/compat/include/linux/overflow.h
@@ -1,5 +1,5 @@
   /* SPDX-License-Identifier: GPL-2.0 OR MIT */
-#ifdef HAVE_OVERFLOW_H
+#if defined(HAVE_OVERFLOW_H) && defined(HAVE_STRUCT_SIZE)
   #include_next 
   #else
   #ifndef __LINUX_OVERFLOW_H
diff --git a/datapath/meter.c b/datapath/meter.c
index 51ec1495033a..eda14682fb96 100644
--- a/datapath/meter.c
+++ b/datapath/meter.c
@@ -13,6 +13,7 @@
   #include 
   #include 
   #include 
+#include 
   #include 
   #include 


Thanks for the patch Yifeng, it looks good but I'm curious because when
I posted the original patch for this it was passing travis build
checks.  Or I thought so at the time.  It bugs me if I missed something.

In any case I'll give it a quick check and regression test.

- Greg


___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] [PATCH] datapath: Fix compiling error for latest Linux 4.14 kernel

2019-04-26 Thread Yifeng Sun 
Sure, I will do that.
Thanks!
Yifeng

On Fri, Apr 26, 2019 at 1:31 PM Gregory Rose  wrote:
>
>
>
> On 4/26/2019 1:18 PM, Yifeng Sun wrote:
> > Hi Greg,
> >
> > I saw the compilation error for kernel 4.14.111 because `struct_size`
> > isn't defined in kernel source tree. I didn't dig into it though.
>
> I see the problem then.  The .travis.yml file tests 4.14.63.  Please
> update the patch to include an update
> for travis to set the 4.14 kernel being tested to 4.14.111.  That way
> we'll get travis to verify the build as
> well.
>
> Thanks!
>
> - Greg
>
>
> >
> > Thanks,
> > Yifeng
> >
> > On Fri, Apr 26, 2019 at 1:13 PM Gregory Rose  wrote:
> >>
> >> On 4/26/2019 11:07 AM, Yifeng Sun wrote:
> >>> CC: Greg Rose 
> >>> Fixes: f72469405eec9 ("datapath: meter: Use struct_size() in kzalloc()")
> >>> Signed-off-by: Yifeng Sun 
> >>> ---
> >>>acinclude.m4   | 2 ++
> >>>datapath/linux/compat/include/linux/overflow.h | 2 +-
> >>>datapath/meter.c   | 1 +
> >>>3 files changed, 4 insertions(+), 1 deletion(-)
> >>>
> >>> diff --git a/acinclude.m4 b/acinclude.m4
> >>> index 1607d5f4b1d9..548cbe9ef100 100644
> >>> --- a/acinclude.m4
> >>> +++ b/acinclude.m4
> >>> @@ -920,6 +920,8 @@ AC_DEFUN([OVS_CHECK_LINUX_COMPAT], [
> >>>[OVS_DEFINE([HAVE_INET_FRAGS_RND])])
> >>>  OVS_GREP_IFELSE([$KSRC/include/linux/overflow.h], 
> >>> [__LINUX_OVERFLOW_H],
> >>>  [OVS_DEFINE([HAVE_OVERFLOW_H])])
> >>> +  OVS_GREP_IFELSE([$KSRC/include/linux/overflow.h], [struct_size],
> >>> +  [OVS_DEFINE([HAVE_STRUCT_SIZE])])
> >>>  OVS_GREP_IFELSE([$KSRC/include/linux/mm.h], [kvmalloc_array],
> >>>  [OVS_DEFINE([HAVE_KVMALLOC_ARRAY])])
> >>>  OVS_GREP_IFELSE([$KSRC/include/linux/mm.h], [kvmalloc_node],
> >>> diff --git a/datapath/linux/compat/include/linux/overflow.h 
> >>> b/datapath/linux/compat/include/linux/overflow.h
> >>> index ff84356db16c..8f1b5b005491 100644
> >>> --- a/datapath/linux/compat/include/linux/overflow.h
> >>> +++ b/datapath/linux/compat/include/linux/overflow.h
> >>> @@ -1,5 +1,5 @@
> >>>/* SPDX-License-Identifier: GPL-2.0 OR MIT */
> >>> -#ifdef HAVE_OVERFLOW_H
> >>> +#if defined(HAVE_OVERFLOW_H) && defined(HAVE_STRUCT_SIZE)
> >>>#include_next 
> >>>#else
> >>>#ifndef __LINUX_OVERFLOW_H
> >>> diff --git a/datapath/meter.c b/datapath/meter.c
> >>> index 51ec1495033a..eda14682fb96 100644
> >>> --- a/datapath/meter.c
> >>> +++ b/datapath/meter.c
> >>> @@ -13,6 +13,7 @@
> >>>#include 
> >>>#include 
> >>>#include 
> >>> +#include 
> >>>#include 
> >>>#include 
> >>>
> >> Thanks for the patch Yifeng, it looks good but I'm curious because when
> >> I posted the original patch for this it was passing travis build
> >> checks.  Or I thought so at the time.  It bugs me if I missed something.
> >>
> >> In any case I'll give it a quick check and regression test.
> >>
> >> - Greg
>
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

[ovs-dev] [PATCHv2] datapath: Fix compiling error for latest Linux 4.14 kernel

2019-04-26 Thread Yifeng Sun 
CC: Greg Rose 
Fixes: f72469405eec9 ("datapath: meter: Use struct_size() in kzalloc()")
Signed-off-by: Yifeng Sun 
---
v1->v2: update 4.14.111 in travis by Greg's suggestion, thanks!
 .travis.yml| 2 +-
 acinclude.m4   | 2 ++
 datapath/linux/compat/include/linux/overflow.h | 2 +-
 datapath/meter.c   | 1 +
 4 files changed, 5 insertions(+), 2 deletions(-)

diff --git a/.travis.yml b/.travis.yml
index 32d5f1918495..468c39ecd505 100644
--- a/.travis.yml
+++ b/.travis.yml
@@ -39,7 +39,7 @@ env:
   - KERNEL=4.17.19
   - KERNEL=4.16.18
   - KERNEL=4.15.18
-  - KERNEL=4.14.63
+  - KERNEL=4.14.111
   - KERNEL=4.9.149
   - KERNEL=4.4.148
   - KERNEL=3.19.8
diff --git a/acinclude.m4 b/acinclude.m4
index 301aeb70d82a..b532a4579266 100644
--- a/acinclude.m4
+++ b/acinclude.m4
@@ -920,6 +920,8 @@ AC_DEFUN([OVS_CHECK_LINUX_COMPAT], [
 [OVS_DEFINE([HAVE_INET_FRAGS_RND])])
   OVS_GREP_IFELSE([$KSRC/include/linux/overflow.h], [__LINUX_OVERFLOW_H],
   [OVS_DEFINE([HAVE_OVERFLOW_H])])
+  OVS_GREP_IFELSE([$KSRC/include/linux/overflow.h], [struct_size],
+  [OVS_DEFINE([HAVE_STRUCT_SIZE])])
   OVS_GREP_IFELSE([$KSRC/include/linux/mm.h], [kvmalloc_array],
   [OVS_DEFINE([HAVE_KVMALLOC_ARRAY])])
   OVS_GREP_IFELSE([$KSRC/include/linux/mm.h], [kvmalloc_node],
diff --git a/datapath/linux/compat/include/linux/overflow.h 
b/datapath/linux/compat/include/linux/overflow.h
index ff84356db16c..8f1b5b005491 100644
--- a/datapath/linux/compat/include/linux/overflow.h
+++ b/datapath/linux/compat/include/linux/overflow.h
@@ -1,5 +1,5 @@
 /* SPDX-License-Identifier: GPL-2.0 OR MIT */
-#ifdef HAVE_OVERFLOW_H
+#if defined(HAVE_OVERFLOW_H) && defined(HAVE_STRUCT_SIZE)
 #include_next 
 #else
 #ifndef __LINUX_OVERFLOW_H
diff --git a/datapath/meter.c b/datapath/meter.c
index 51ec1495033a..eda14682fb96 100644
--- a/datapath/meter.c
+++ b/datapath/meter.c
@@ -13,6 +13,7 @@
 #include 
 #include 
 #include 
+#include 
 #include 
 #include 
 
-- 
2.7.4

___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] [PATCHv2] datapath: Fix compiling error for latest Linux 4.14 kernel

2019-04-26 Thread Yi-Hung Wei 
Thanks for the patch. It fixes an issue where overflow.h is there but
the struct_size marco is not available.

Some minor comment.  For the patch title, instead of having "Fix
compiling error for latest Linux 4.14 kernel", maybe we can specify
the minor version number say "4.14.111+ kernel) so that it is more
clear on which kernel version this commit fix?

In ./datapath/linux/compat/include/linux/overflow.h, there is a paring
#endif. Should we update the comment over there as well?
#endif /* HAVE_OVERFLOW_H */

Thanks,

-Yi-Hung
On Fri, Apr 26, 2019 at 1:39 PM Yifeng Sun  wrote:
>
> CC: Greg Rose 
> Fixes: f72469405eec9 ("datapath: meter: Use struct_size() in kzalloc()")
> Signed-off-by: Yifeng Sun 
> ---
> v1->v2: update 4.14.111 in travis by Greg's suggestion, thanks!
>  .travis.yml| 2 +-
>  acinclude.m4   | 2 ++
>  datapath/linux/compat/include/linux/overflow.h | 2 +-
>  datapath/meter.c   | 1 +
>  4 files changed, 5 insertions(+), 2 deletions(-)
>
> diff --git a/.travis.yml b/.travis.yml
> index 32d5f1918495..468c39ecd505 100644
> --- a/.travis.yml
> +++ b/.travis.yml
> @@ -39,7 +39,7 @@ env:
>- KERNEL=4.17.19
>- KERNEL=4.16.18
>- KERNEL=4.15.18
> -  - KERNEL=4.14.63
> +  - KERNEL=4.14.111
>- KERNEL=4.9.149
>- KERNEL=4.4.148
>- KERNEL=3.19.8
> diff --git a/acinclude.m4 b/acinclude.m4
> index 301aeb70d82a..b532a4579266 100644
> --- a/acinclude.m4
> +++ b/acinclude.m4
> @@ -920,6 +920,8 @@ AC_DEFUN([OVS_CHECK_LINUX_COMPAT], [
>  [OVS_DEFINE([HAVE_INET_FRAGS_RND])])
>OVS_GREP_IFELSE([$KSRC/include/linux/overflow.h], [__LINUX_OVERFLOW_H],
>[OVS_DEFINE([HAVE_OVERFLOW_H])])
> +  OVS_GREP_IFELSE([$KSRC/include/linux/overflow.h], [struct_size],
> +  [OVS_DEFINE([HAVE_STRUCT_SIZE])])
>OVS_GREP_IFELSE([$KSRC/include/linux/mm.h], [kvmalloc_array],
>[OVS_DEFINE([HAVE_KVMALLOC_ARRAY])])
>OVS_GREP_IFELSE([$KSRC/include/linux/mm.h], [kvmalloc_node],
> diff --git a/datapath/linux/compat/include/linux/overflow.h 
> b/datapath/linux/compat/include/linux/overflow.h
> index ff84356db16c..8f1b5b005491 100644
> --- a/datapath/linux/compat/include/linux/overflow.h
> +++ b/datapath/linux/compat/include/linux/overflow.h
> @@ -1,5 +1,5 @@
>  /* SPDX-License-Identifier: GPL-2.0 OR MIT */
> -#ifdef HAVE_OVERFLOW_H
> +#if defined(HAVE_OVERFLOW_H) && defined(HAVE_STRUCT_SIZE)
>  #include_next 
>  #else
>  #ifndef __LINUX_OVERFLOW_H
> diff --git a/datapath/meter.c b/datapath/meter.c
> index 51ec1495033a..eda14682fb96 100644
> --- a/datapath/meter.c
> +++ b/datapath/meter.c
> @@ -13,6 +13,7 @@
>  #include 
>  #include 
>  #include 
> +#include 
>  #include 
>  #include 
>
> --
> 2.7.4
>
> ___
> dev mailing list
> d...@openvswitch.org
> https://mail.openvswitch.org/mailman/listinfo/ovs-dev
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] [PATCHv2] datapath: Fix compiling error for latest Linux 4.14 kernel

2019-04-26 Thread Yifeng Sun 
Thanks Yi-Hung for the review, I will submit a new version.
Yifeng

On Fri, Apr 26, 2019 at 2:28 PM Yi-Hung Wei  wrote:
>
> Thanks for the patch. It fixes an issue where overflow.h is there but
> the struct_size marco is not available.
>
> Some minor comment.  For the patch title, instead of having "Fix
> compiling error for latest Linux 4.14 kernel", maybe we can specify
> the minor version number say "4.14.111+ kernel) so that it is more
> clear on which kernel version this commit fix?
>
> In ./datapath/linux/compat/include/linux/overflow.h, there is a paring
> #endif. Should we update the comment over there as well?
> #endif /* HAVE_OVERFLOW_H */
>
> Thanks,
>
> -Yi-Hung
> On Fri, Apr 26, 2019 at 1:39 PM Yifeng Sun  wrote:
> >
> > CC: Greg Rose 
> > Fixes: f72469405eec9 ("datapath: meter: Use struct_size() in kzalloc()")
> > Signed-off-by: Yifeng Sun 
> > ---
> > v1->v2: update 4.14.111 in travis by Greg's suggestion, thanks!
> >  .travis.yml| 2 +-
> >  acinclude.m4   | 2 ++
> >  datapath/linux/compat/include/linux/overflow.h | 2 +-
> >  datapath/meter.c   | 1 +
> >  4 files changed, 5 insertions(+), 2 deletions(-)
> >
> > diff --git a/.travis.yml b/.travis.yml
> > index 32d5f1918495..468c39ecd505 100644
> > --- a/.travis.yml
> > +++ b/.travis.yml
> > @@ -39,7 +39,7 @@ env:
> >- KERNEL=4.17.19
> >- KERNEL=4.16.18
> >- KERNEL=4.15.18
> > -  - KERNEL=4.14.63
> > +  - KERNEL=4.14.111
> >- KERNEL=4.9.149
> >- KERNEL=4.4.148
> >- KERNEL=3.19.8
> > diff --git a/acinclude.m4 b/acinclude.m4
> > index 301aeb70d82a..b532a4579266 100644
> > --- a/acinclude.m4
> > +++ b/acinclude.m4
> > @@ -920,6 +920,8 @@ AC_DEFUN([OVS_CHECK_LINUX_COMPAT], [
> >  [OVS_DEFINE([HAVE_INET_FRAGS_RND])])
> >OVS_GREP_IFELSE([$KSRC/include/linux/overflow.h], [__LINUX_OVERFLOW_H],
> >[OVS_DEFINE([HAVE_OVERFLOW_H])])
> > +  OVS_GREP_IFELSE([$KSRC/include/linux/overflow.h], [struct_size],
> > +  [OVS_DEFINE([HAVE_STRUCT_SIZE])])
> >OVS_GREP_IFELSE([$KSRC/include/linux/mm.h], [kvmalloc_array],
> >[OVS_DEFINE([HAVE_KVMALLOC_ARRAY])])
> >OVS_GREP_IFELSE([$KSRC/include/linux/mm.h], [kvmalloc_node],
> > diff --git a/datapath/linux/compat/include/linux/overflow.h 
> > b/datapath/linux/compat/include/linux/overflow.h
> > index ff84356db16c..8f1b5b005491 100644
> > --- a/datapath/linux/compat/include/linux/overflow.h
> > +++ b/datapath/linux/compat/include/linux/overflow.h
> > @@ -1,5 +1,5 @@
> >  /* SPDX-License-Identifier: GPL-2.0 OR MIT */
> > -#ifdef HAVE_OVERFLOW_H
> > +#if defined(HAVE_OVERFLOW_H) && defined(HAVE_STRUCT_SIZE)
> >  #include_next 
> >  #else
> >  #ifndef __LINUX_OVERFLOW_H
> > diff --git a/datapath/meter.c b/datapath/meter.c
> > index 51ec1495033a..eda14682fb96 100644
> > --- a/datapath/meter.c
> > +++ b/datapath/meter.c
> > @@ -13,6 +13,7 @@
> >  #include 
> >  #include 
> >  #include 
> > +#include 
> >  #include 
> >  #include 
> >
> > --
> > 2.7.4
> >
> > ___
> > dev mailing list
> > d...@openvswitch.org
> > https://mail.openvswitch.org/mailman/listinfo/ovs-dev
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

[ovs-dev] [PATCHv3] datapath: Fix compiling error for 4.14.111+ kernel

2019-04-26 Thread Yifeng Sun 
CC: Greg Rose 
CC: Yi-Hung Wei 
Fixes: f72469405eec9 ("datapath: meter: Use struct_size() in kzalloc()")
Signed-off-by: Yifeng Sun 
---
v1->v2: Update 4.14.111 in travis by Greg's suggestion, thanks!
v2->v3: Fix patch title and comment of endif, thanks Yi-Hung!
 .travis.yml| 2 +-
 acinclude.m4   | 2 ++
 datapath/linux/compat/include/linux/overflow.h | 4 ++--
 datapath/meter.c   | 1 +
 4 files changed, 6 insertions(+), 3 deletions(-)

diff --git a/.travis.yml b/.travis.yml
index 32d5f1918495..468c39ecd505 100644
--- a/.travis.yml
+++ b/.travis.yml
@@ -39,7 +39,7 @@ env:
   - KERNEL=4.17.19
   - KERNEL=4.16.18
   - KERNEL=4.15.18
-  - KERNEL=4.14.63
+  - KERNEL=4.14.111
   - KERNEL=4.9.149
   - KERNEL=4.4.148
   - KERNEL=3.19.8
diff --git a/acinclude.m4 b/acinclude.m4
index 301aeb70d82a..b532a4579266 100644
--- a/acinclude.m4
+++ b/acinclude.m4
@@ -920,6 +920,8 @@ AC_DEFUN([OVS_CHECK_LINUX_COMPAT], [
 [OVS_DEFINE([HAVE_INET_FRAGS_RND])])
   OVS_GREP_IFELSE([$KSRC/include/linux/overflow.h], [__LINUX_OVERFLOW_H],
   [OVS_DEFINE([HAVE_OVERFLOW_H])])
+  OVS_GREP_IFELSE([$KSRC/include/linux/overflow.h], [struct_size],
+  [OVS_DEFINE([HAVE_STRUCT_SIZE])])
   OVS_GREP_IFELSE([$KSRC/include/linux/mm.h], [kvmalloc_array],
   [OVS_DEFINE([HAVE_KVMALLOC_ARRAY])])
   OVS_GREP_IFELSE([$KSRC/include/linux/mm.h], [kvmalloc_node],
diff --git a/datapath/linux/compat/include/linux/overflow.h 
b/datapath/linux/compat/include/linux/overflow.h
index ff84356db16c..13ae6cf6a94a 100644
--- a/datapath/linux/compat/include/linux/overflow.h
+++ b/datapath/linux/compat/include/linux/overflow.h
@@ -1,5 +1,5 @@
 /* SPDX-License-Identifier: GPL-2.0 OR MIT */
-#ifdef HAVE_OVERFLOW_H
+#if defined(HAVE_OVERFLOW_H) && defined(HAVE_STRUCT_SIZE)
 #include_next 
 #else
 #ifndef __LINUX_OVERFLOW_H
@@ -310,4 +310,4 @@ static inline __must_check size_t __ab_c_size(size_t n, 
size_t size, size_t c)
sizeof(*(p)))
 
 #endif /* __LINUX_OVERFLOW_H */
-#endif /* HAVE_OVERFLOW_H */
+#endif /* defined(HAVE_OVERFLOW_H) && defined(HAVE_STRUCT_SIZE) */
diff --git a/datapath/meter.c b/datapath/meter.c
index 51ec1495033a..eda14682fb96 100644
--- a/datapath/meter.c
+++ b/datapath/meter.c
@@ -13,6 +13,7 @@
 #include 
 #include 
 #include 
+#include 
 #include 
 #include 
 
-- 
2.7.4

___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] [PATCH] OVN: Fix build when using a separate build directory.

2019-04-26 Thread Mark Michelson 

On 4/26/19 2:14 PM, Ben Pfaff wrote:

On Fri, Apr 26, 2019 at 12:14:06PM -0400, Mark Michelson wrote:

The linker was searching for built libraries in the source directory
rather than in the build directory. This fixes the problem.

Reported-by: Numan Siddique 
Signed-off-by: Mark Michelson 


I did not test this but it looks perfectly reasonable.

Acked-by: Ben Pfaff 



Thanks, I pushed this to master.
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] [PATCH] OVN: Fix failures after OVS/OVN split

2019-04-26 Thread Mark Michelson 

Thanks for the correction. I've pushed it to master.

On 4/25/19 4:08 PM, Brian Haley wrote:

Added TODO_SPLIT.rst to Makefile.am to avoid an error
during build.

Removed a section from config-h-check section of Makefile.am
that was no longer relevant.

Signed-off-by: Brian Haley 
---
  Makefile.am | 8 +---
  1 file changed, 1 insertion(+), 7 deletions(-)

diff --git a/Makefile.am b/Makefile.am
index 1b492a314..2bcdb0a76 100644
--- a/Makefile.am
+++ b/Makefile.am
@@ -106,6 +106,7 @@ EXTRA_DIST = \
Vagrantfile-FreeBSD \
.mailmap \
TODO.rst \
+   TODO_SPLIT.rst \
ovn-architecture.7.xml \
ovn-nb.ovsschema \
ovn-nb.xml \
@@ -263,13 +264,6 @@ config-h-check:
  echo "See above for list of violations of the rule that"; \
  echo "every C source file must #include ."; \
  exit 1; \
-   fi; \
-   if grep '#include' include/openvswitch/*.h | \
-   grep -vE '(<.*>)|("openvswitch)|("openflow)'; \
-   then \
- echo "See above for list of violations of the rule that"; \
- echo "public openvswitch header file should not include internal 
library."; \
- exit 1; \
fi
  .PHONY: config-h-check
  



___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] OVS/OVN Split initial commits are complete

2019-04-26 Thread Mark Michelson 

On 4/26/19 3:42 PM, Ben Pfaff wrote:

On Fri, Apr 26, 2019 at 11:14:23AM +0530, Numan Siddique wrote:

On Fri, Apr 26, 2019 at 1:56 AM Ben Pfaff  wrote:


On Thu, Apr 25, 2019 at 12:51:29PM -0700, Ben Pfaff wrote:

On Thu, Apr 25, 2019 at 08:38:06AM -0400, Mark Michelson wrote:

Yesterday, I once again performed the task of splitting OVN from OVS

(I've

gotten pretty good at it now). This time, rather than living in a

personal

repo, you can find the ovn repo at https://github.com/ovn-org/ovn.git


I noticed that most or all of your commits lack sign-offs.  Oops?


By the way, I guess that you should be the main person pushing to this
repo for now, so I'll review some patches but I'll plan to mostly ack
them rather than applying them to the new repo.




Thanks Mark for all the work.
Shall we consider that all the OVN patches will now be applied to the new
repo ?
If so, then all the present OVN patches under review should be resubmitted ?


We need to pick a date (or a trigger) for freezing the OVN parts of OVS.
I don't know what that date should be.  Thoughts?



I'm of the opinion that it should be ASAP. With the new repo created, 
it's best for developers to start using the new repo for their 
day-to-day development. Adding more commits to the OVS repo at this 
point will potentially result in lost work if a maintainer neglects to 
get it moved over.

___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] [PATCHv3] datapath: Fix compiling error for 4.14.111+ kernel

2019-04-26 Thread Gregory Rose 



On 4/26/2019 2:42 PM, Yifeng Sun wrote:

CC: Greg Rose 
CC: Yi-Hung Wei 
Fixes: f72469405eec9 ("datapath: meter: Use struct_size() in kzalloc()")
Signed-off-by: Yifeng Sun 
---
v1->v2: Update 4.14.111 in travis by Greg's suggestion, thanks!
v2->v3: Fix patch title and comment of endif, thanks Yi-Hung!
  .travis.yml| 2 +-
  acinclude.m4   | 2 ++
  datapath/linux/compat/include/linux/overflow.h | 4 ++--
  datapath/meter.c   | 1 +
  4 files changed, 6 insertions(+), 3 deletions(-)

diff --git a/.travis.yml b/.travis.yml
index 32d5f1918495..468c39ecd505 100644
--- a/.travis.yml
+++ b/.travis.yml
@@ -39,7 +39,7 @@ env:
- KERNEL=4.17.19
- KERNEL=4.16.18
- KERNEL=4.15.18
-  - KERNEL=4.14.63
+  - KERNEL=4.14.111
- KERNEL=4.9.149
- KERNEL=4.4.148
- KERNEL=3.19.8
diff --git a/acinclude.m4 b/acinclude.m4
index 301aeb70d82a..b532a4579266 100644
--- a/acinclude.m4
+++ b/acinclude.m4
@@ -920,6 +920,8 @@ AC_DEFUN([OVS_CHECK_LINUX_COMPAT], [
  [OVS_DEFINE([HAVE_INET_FRAGS_RND])])
OVS_GREP_IFELSE([$KSRC/include/linux/overflow.h], [__LINUX_OVERFLOW_H],
[OVS_DEFINE([HAVE_OVERFLOW_H])])
+  OVS_GREP_IFELSE([$KSRC/include/linux/overflow.h], [struct_size],
+  [OVS_DEFINE([HAVE_STRUCT_SIZE])])
OVS_GREP_IFELSE([$KSRC/include/linux/mm.h], [kvmalloc_array],
[OVS_DEFINE([HAVE_KVMALLOC_ARRAY])])
OVS_GREP_IFELSE([$KSRC/include/linux/mm.h], [kvmalloc_node],
diff --git a/datapath/linux/compat/include/linux/overflow.h 
b/datapath/linux/compat/include/linux/overflow.h
index ff84356db16c..13ae6cf6a94a 100644
--- a/datapath/linux/compat/include/linux/overflow.h
+++ b/datapath/linux/compat/include/linux/overflow.h
@@ -1,5 +1,5 @@
  /* SPDX-License-Identifier: GPL-2.0 OR MIT */
-#ifdef HAVE_OVERFLOW_H
+#if defined(HAVE_OVERFLOW_H) && defined(HAVE_STRUCT_SIZE)
  #include_next 
  #else
  #ifndef __LINUX_OVERFLOW_H
@@ -310,4 +310,4 @@ static inline __must_check size_t __ab_c_size(size_t n, 
size_t size, size_t c)
sizeof(*(p)))
  
  #endif /* __LINUX_OVERFLOW_H */

-#endif /* HAVE_OVERFLOW_H */
+#endif /* defined(HAVE_OVERFLOW_H) && defined(HAVE_STRUCT_SIZE) */
diff --git a/datapath/meter.c b/datapath/meter.c
index 51ec1495033a..eda14682fb96 100644
--- a/datapath/meter.c
+++ b/datapath/meter.c
@@ -13,6 +13,7 @@
  #include 
  #include 
  #include 
+#include 
  #include 
  #include 
  

LGTM - passes travis, no check-kmod regressions...

Tested-by: Greg Rose 
Reviewed-by: Greg Rose 

___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] [PATCHv3] datapath: Fix compiling error for 4.14.111+ kernel

2019-04-26 Thread Yi-Hung Wei 
On Fri, Apr 26, 2019 at 2:57 PM Gregory Rose  wrote:
>
>
> On 4/26/2019 2:42 PM, Yifeng Sun wrote:
> > CC: Greg Rose 
> > CC: Yi-Hung Wei 
> > Fixes: f72469405eec9 ("datapath: meter: Use struct_size() in kzalloc()")
> > Signed-off-by: Yifeng Sun 
> > ---
> > v1->v2: Update 4.14.111 in travis by Greg's suggestion, thanks!
> > v2->v3: Fix patch title and comment of endif, thanks Yi-Hung!
> >   .travis.yml| 2 +-
> >   acinclude.m4   | 2 ++
> >   datapath/linux/compat/include/linux/overflow.h | 4 ++--
> >   datapath/meter.c   | 1 +
> >   4 files changed, 6 insertions(+), 3 deletions(-)
> >
> > diff --git a/.travis.yml b/.travis.yml
> > index 32d5f1918495..468c39ecd505 100644
> > --- a/.travis.yml
> > +++ b/.travis.yml
> > @@ -39,7 +39,7 @@ env:
> > - KERNEL=4.17.19
> > - KERNEL=4.16.18
> > - KERNEL=4.15.18
> > -  - KERNEL=4.14.63
> > +  - KERNEL=4.14.111
> > - KERNEL=4.9.149
> > - KERNEL=4.4.148
> > - KERNEL=3.19.8
> > diff --git a/acinclude.m4 b/acinclude.m4
> > index 301aeb70d82a..b532a4579266 100644
> > --- a/acinclude.m4
> > +++ b/acinclude.m4
> > @@ -920,6 +920,8 @@ AC_DEFUN([OVS_CHECK_LINUX_COMPAT], [
> >   [OVS_DEFINE([HAVE_INET_FRAGS_RND])])
> > OVS_GREP_IFELSE([$KSRC/include/linux/overflow.h], [__LINUX_OVERFLOW_H],
> > [OVS_DEFINE([HAVE_OVERFLOW_H])])
> > +  OVS_GREP_IFELSE([$KSRC/include/linux/overflow.h], [struct_size],
> > +  [OVS_DEFINE([HAVE_STRUCT_SIZE])])
> > OVS_GREP_IFELSE([$KSRC/include/linux/mm.h], [kvmalloc_array],
> > [OVS_DEFINE([HAVE_KVMALLOC_ARRAY])])
> > OVS_GREP_IFELSE([$KSRC/include/linux/mm.h], [kvmalloc_node],
> > diff --git a/datapath/linux/compat/include/linux/overflow.h 
> > b/datapath/linux/compat/include/linux/overflow.h
> > index ff84356db16c..13ae6cf6a94a 100644
> > --- a/datapath/linux/compat/include/linux/overflow.h
> > +++ b/datapath/linux/compat/include/linux/overflow.h
> > @@ -1,5 +1,5 @@
> >   /* SPDX-License-Identifier: GPL-2.0 OR MIT */
> > -#ifdef HAVE_OVERFLOW_H
> > +#if defined(HAVE_OVERFLOW_H) && defined(HAVE_STRUCT_SIZE)
> >   #include_next 
> >   #else
> >   #ifndef __LINUX_OVERFLOW_H
> > @@ -310,4 +310,4 @@ static inline __must_check size_t __ab_c_size(size_t n, 
> > size_t size, size_t c)
> >   sizeof(*(p)))
> >
> >   #endif /* __LINUX_OVERFLOW_H */
> > -#endif /* HAVE_OVERFLOW_H */
> > +#endif /* defined(HAVE_OVERFLOW_H) && defined(HAVE_STRUCT_SIZE) */
> > diff --git a/datapath/meter.c b/datapath/meter.c
> > index 51ec1495033a..eda14682fb96 100644
> > --- a/datapath/meter.c
> > +++ b/datapath/meter.c
> > @@ -13,6 +13,7 @@
> >   #include 
> >   #include 
> >   #include 
> > +#include 
> >   #include 
> >   #include 
> >
> LGTM - passes travis, no check-kmod regressions...
>
> Tested-by: Greg Rose 
> Reviewed-by: Greg Rose 
>
Acked-by: Yi-Hung Wei 
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Re: [ovs-dev] OVS/OVN Split initial commits are complete

2019-04-26 Thread Justin Pettit 


> On Apr 26, 2019, at 2:50 PM, Mark Michelson  wrote:
> 
> On 4/26/19 3:42 PM, Ben Pfaff wrote:
> 
>> We need to pick a date (or a trigger) for freezing the OVN parts of OVS.
>> I don't know what that date should be.  Thoughts?
> 
> I'm of the opinion that it should be ASAP. With the new repo created, it's 
> best for developers to start using the new repo for their day-to-day 
> development. Adding more commits to the OVS repo at this point will 
> potentially result in lost work if a maintainer neglects to get it moved over.

I think we should continue to do development in the OVS repo until we're 
certain that the approach being taken is the right one and that we've got all 
the pieces in place to fully do OVN development on its own (e.g., separate 
mailing list, CI/CD infrastructure, etc).  Once we have all those pieces in 
place, then we should delete the "ovn" branch from the OVS repo.  Having the 
place where it has traditionally lived continue on in a zombie state will be 
confusing, and possibly problematic if we decide the path we're currently going 
down is wrong for some reason.  Those concerns combined with your imminent 
paternity lean me towards waiting.

Let's see how things go for the next couple months, and then revisit if it's 
time to stop development in the "ovn" branch.  If it goes as I expect, I'll 
enthusiastically support that change.

Thanks,

--Justin


___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

[ovs-dev] Is there kernel datapath support for OVS_HASH_ALG_SYM_L4

2019-04-26 Thread Sa A 
Hello,

Below commits have been added for supporting  OVS_HASH_ALG_SYM_L4 in
userspace datapath. Is there plan to add support in kernel datapath?
Currently, while probing for max_dp_hash_alg, I get the error as
-22(Invalid argument) in netlink message reply, and it falls back to
OVS_HASH_ALG_L4.
Commit 1. ofproto-dpif: Use dp_hash as default selection method:
https://github.com/openvswitch/ovs/commit/06db81ccfe6d4c779de2ca73033abd7020db419b
Commit 2. userspace datapath: Add OVS_HASH_L4_SYMMETRIC dp_hash algorithm:
https://github.com/openvswitch/ovs/commit/6a0b0d3be8573c42d750eb22e3445a3c759c8e8f

Thank you for help.
___
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev