Hi,
On Wed, Sep 25, 2013 at 01:38:58PM +0900, YAMAMOTO Takashi wrote:
> if a queue length is long (ie. non-0), the consumer thread should
> already be busy working on the queue. there's no need to wake it
> up repeatedly.
...
> @@ -530,9 +532,13 @@ recv_upcalls(struct udpif *udpif)
> ovs
On Wed, Sep 25, 2013 at 08:55:21AM -0700, Ben Pfaff wrote:
> On Wed, Sep 25, 2013 at 01:38:58PM +0900, YAMAMOTO Takashi wrote:
> > if a queue length is long (ie. non-0), the consumer thread should
> > already be busy working on the queue. there's no need to wake it
> > up repeatedly.
> >
> > Sign
On Wed, Sep 25, 2013 at 01:38:58PM +0900, YAMAMOTO Takashi wrote:
> if a queue length is long (ie. non-0), the consumer thread should
> already be busy working on the queue. there's no need to wake it
> up repeatedly.
>
> Signed-off-by: YAMAMOTO Takashi
Looks good, I'll apply this after Ethan's
if a queue length is long (ie. non-0), the consumer thread should
already be busy working on the queue. there's no need to wake it
up repeatedly.
Signed-off-by: YAMAMOTO Takashi
---
ofproto/ofproto-dpif-upcall.c | 29 +++--
1 file changed, 19 insertions(+), 10 deletions(
if a queue length is long (ie. non-0), the consumer thread should
already be busy working on the queue. there's no need to wake it
up repeatedly.
Signed-off-by: YAMAMOTO Takashi
---
ofproto/ofproto-dpif-upcall.c | 32 +---
1 file changed, 21 insertions(+), 11 deletio
