Signed-off-by: ZhengLingyun
---
lib/vlandev.c |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/vlandev.c b/lib/vlandev.c
index 382487c..d2f5a37 100644
--- a/lib/vlandev.c
+++ b/lib/vlandev.c
@@ -374,7 +374,7 @@ static int
vlandev_del__(const char *vlan_dev
'ovs_datapath'
does not exist. The Open vSwitch kernel module is probably not loaded.
2013-11-22T03:40:26Z|5|bridge|INFO|bridge br0: using datapath ID
c644fe67d34d
2013-11-22T03:40:26Z|6|connmgr|INFO|br0: added service controller
"punix:/home/zhenglingyun/var/run/openv
I set OVS connect to a non-existent controller 10.1.1.1, the controller will
fall into fail-open mode.
Then I delete the controller, the OVS will abort:
2013-11-22T01:25:40Z|00010|connmgr|INFO|br0: added primary controller
"tcp:10.1.1.1:6632"
2013-11-22T01:25:40Z|00011|rconn|INFO|br0<->tcp:1
Starting ovs-controller with '-H' option will lead to a segment fault problem.
Add a check, and adjust the indentation of the following code.
Signed-off-by: ZhengLingyun
---
lib/learning-switch.c | 26 +++---
1 file changed, 15 insertions(+), 11 deletions(-)
di
Hi, all
I'm using a userspace OVS. I find that there is PROBABILITY the 'br0'
opened as type 'internal' before it is opened as type 'tap'. This will
cause the 'br0' inserted into 'netdev_shash' as internal class. When it
is created as type 'tap' later, error occurs. 'br0' then can't be created
in
It is ok now.
Another problem:
netdev_open does shash_add() but no shash_delete() when error
occurs, such as device not exists. It will cause a SIGSEGV signal.
At 2013-08-10 21:49:30,"Ben Pfaff" wrote:
>On Sat, Aug 10, 2013 at 06:33:35PM +0800, ZhengLingyun wrote:
>> I j
Hi, Ben
I just update the code. It can't run (just in userspace?).
There is a recursive dead lock in netdev_linux_set_etheraddr and
netdev_linux_update_flags.
#0 0xb7fdd424 in __kernel_vsyscall ()
#1 0xb7dbe5a2 in __lll_lock_wait () from /lib/i386-linux-gnu/libpthread.so.0
#2 0xb7db9ead in _L
ot; greater than the "size".
Change the sequence of comparison to avoid negative return value cast to
a positive and become greater then a non-negative value.
Signed-off-by: ZhengLingyun
---
lib/netdev-linux.c |6 ++
1 file changed, 2 insertions(+), 4 deletions(-)
diff --
It seems I don't get the purpose of introducing hindex into lib. Now I get it.
Thanks.
At 2013-07-17 01:08:16,"Ben Pfaff" wrote:
>On Tue, Jul 16, 2013 at 06:46:22PM +0800, ZhengLingyun wrote:
>> As you have said in hindex.h, hindex need a high-quality hash function to
&
these two functions seems confusing.
Thank you.
ZhengLingyun
At 2013-07-16 08:16:09,"Ben Pfaff" wrote:
>On Mon, Jul 15, 2013 at 08:21:04AM +0800, ZhengLingyun wrote:
>> hindex_next() shouldn't jump over different hashs in the same bucket.
>> tests/test-hindex
hindex_next() shouldn't jump over different hashs in the same bucket.
tests/test-hindex did not detect it, add a multipart_hash() method to detect it.
Signed-off-by: ZhengLingyun
---
lib/hindex.c| 21 ++---
tests/test-hindex.c |7 +++
2 files change
In function bitmap_allocate1(),
the last "unsigned long" in bitmap will be set to all 0-bits when
(n_bits % BITMAP_ULONG_BITS == 0),
it actually should be set to all 1-bits.
Signed-off-by: ZhengLingyun
---
lib/bitmap.c | 5 -
1 file changed, 4 insertions(+), 1 deletion(-)
diff -
/* Ensure that the last "unsigned long" in the bitmap only has as many
* 1-bits as there actually should be. */
bitmap[n_longs - 1] = (1UL << (n_bits % BITMAP_ULONG_BITS)) - 1;
from bitmap_allocate1(),
It will fail when n_bits % BITMAP_ULONG_BITS == 0, does it matter?
unsigned long *
bitmap_a
13 matches
Mail list logo
