Works me. I can now receive at 1.15 Mpps, slightly
faster than my previous patch which generated unnecessary
writes to the signalling socket.
Tested-by: Luigi Rizzo
On Thu, May 31, 2012 at 12:03 PM, Paolo Bonzini wrote:
> Il 31/05/2012 10:23, Jan Kiszka ha scritto:
> >> > @@ -922,6 +923,7 @@ s
Hi Paolo,
I suggest using false / true instead of 0 / 1 for the boolean values,
see my comments below.
Regards,
Stefan
Am 31.05.2012 12:03, schrieb Paolo Bonzini:
[...]
-void qemu_net_queue_flush(NetQueue *queue)
+bool qemu_net_queue_flush(NetQueue *queue)
{
while (!QTAILQ_EMPTY(&qu
Il 31/05/2012 13:43, Stefano Stabellini ha scritto:
>> > No, it's not that. I was talking about feature-rx-notify. But I
>> > remembered wrong, that feature is advertised by the front-end, not the
>> > back-end.
> Yes, that's right.
>
>
>> > IIRC there is only one event channel per Xen network
On Thu, 31 May 2012, Paolo Bonzini wrote:
> Il 31/05/2012 13:06, Stefano Stabellini ha scritto:
> >> > For Xen I think this is not possible at the moment because it doesn't
> >> > implement rx notification.
> >
> > Why do you say that? Xen supports the iothread and CONFIG_EVENTFD.
>
> No, it's n
Il 31/05/2012 13:06, Stefano Stabellini ha scritto:
>> > For Xen I think this is not possible at the moment because it doesn't
>> > implement rx notification.
>
> Why do you say that? Xen supports the iothread and CONFIG_EVENTFD.
No, it's not that. I was talking about feature-rx-notify. But I
On Thu, 31 May 2012, Paolo Bonzini wrote:
> Il 31/05/2012 00:53, Luigi Rizzo ha scritto:
> > The image contains my fast packet generator "pkt-gen" (a stock
> > traffic generator such as netperf etc. is too slow to show the
> > problem). pkt-gen can send about 1Mpps in this configuration using
> > -
Il 31/05/2012 12:40, Jan Kiszka ha scritto:
> On 2012-05-31 12:03, Paolo Bonzini wrote:
>> Il 31/05/2012 10:23, Jan Kiszka ha scritto:
> @@ -922,6 +923,7 @@ set_rdt(E1000State *s, int index, uint32_t val)
> {
> s->check_rxov = 0;
> s->mac_reg[index] = val & 0x;
>
On 2012-05-31 12:03, Paolo Bonzini wrote:
> Il 31/05/2012 10:23, Jan Kiszka ha scritto:
@@ -922,6 +923,7 @@ set_rdt(E1000State *s, int index, uint32_t val)
{
s->check_rxov = 0;
s->mac_reg[index] = val & 0x;
+qemu_notify_event();
>> This still looks like t
Il 31/05/2012 10:23, Jan Kiszka ha scritto:
>> > @@ -922,6 +923,7 @@ set_rdt(E1000State *s, int index, uint32_t val)
>> > {
>> > s->check_rxov = 0;
>> > s->mac_reg[index] = val & 0x;
>> > +qemu_notify_event();
> This still looks like the wrong tool: Packets that can't be delivere
On 2012-05-31 10:32, Luigi Rizzo wrote:
> On Thu, May 31, 2012 at 10:23 AM, Jan Kiszka wrote:
>
>> On 2012-05-31 09:38, Paolo Bonzini wrote
>>
> ...
>
>>
>> This still looks like the wrong tool: Packets that can't be delivered
>> are queued. So we need to flush the queue and clear the blocked de
On Thu, May 31, 2012 at 10:23 AM, Jan Kiszka wrote:
> On 2012-05-31 09:38, Paolo Bonzini wrote
>
...
>
> This still looks like the wrong tool: Packets that can't be delivered
> are queued. So we need to flush the queue and clear the blocked delivery
> there. qemu_flush_queued_packets appears mor
On 2012-05-31 10:23, Luigi Rizzo wrote:
> But i think it would be more robust to make fewer assumptions on
> what the guest does, and send the notify on all register writes
> (those are relatively rare in a driver, and the datapath touches
> exactly the registers we ought to be interested in),
> an
On 2012-05-31 09:38, Paolo Bonzini wrote:
> Il 31/05/2012 00:53, Luigi Rizzo ha scritto:
>> The image contains my fast packet generator "pkt-gen" (a stock
>> traffic generator such as netperf etc. is too slow to show the
>> problem). pkt-gen can send about 1Mpps in this configuration using
>> -net
On Thu, May 31, 2012 at 9:38 AM, Paolo Bonzini wrote:
> Il 31/05/2012 00:53, Luigi Rizzo ha scritto:
> > The image contains my fast packet generator "pkt-gen" (a stock
> > traffic generator such as netperf etc. is too slow to show the
> > problem). pkt-gen can send about 1Mpps in this configurati
Il 31/05/2012 00:53, Luigi Rizzo ha scritto:
> The image contains my fast packet generator "pkt-gen" (a stock
> traffic generator such as netperf etc. is too slow to show the
> problem). pkt-gen can send about 1Mpps in this configuration using
> -net netmap in the backend. The qemu process in this
On Thu, May 31, 2012 at 12:11 AM, Jan Kiszka wrote:
> On 2012-05-30 23:55, Luigi Rizzo wrote:
> > you can take the freebsd image from the netmap page in my link and run it
> > in qemu, and then run the pkt-gen program in the image in either
> > send or receive mode. But this is overkill, as you h
On 2012-05-30 23:55, Luigi Rizzo wrote:
> you can take the freebsd image from the netmap page in my link and run it
> in qemu, and then run the pkt-gen program in the image in either
> send or receive mode. But this is overkill, as you have described the
> problem exactly in your post: when the gue
On Wed, May 30, 2012 at 11:39 PM, Jan Kiszka wrote:
> Please keep CCs.
>
> ok
> On 2012-05-30 23:23, Luigi Rizzo wrote:
> >> On Wed, May 30, 2012 at 10:23:11PM +0200, Luigi Rizzo wrote:
> > ...
> >>> The problem was fixed by the following one-line addition to
> >>> hw/e1000.c :: e1000_mmio_writ
Please keep CCs.
On 2012-05-30 23:23, Luigi Rizzo wrote:
>> On Wed, May 30, 2012 at 10:23:11PM +0200, Luigi Rizzo wrote:
> ...
>>> The problem was fixed by the following one-line addition to
>>> hw/e1000.c :: e1000_mmio_write() , to wakeup the qemu mainloop and
>>> check that some buffers might be
> On Wed, May 30, 2012 at 10:23:11PM +0200, Luigi Rizzo wrote:
...
> > The problem was fixed by the following one-line addition to
> > hw/e1000.c :: e1000_mmio_write() , to wakeup the qemu mainloop and
> > check that some buffers might be available.
> >
> > --- hw/e1000.c.orig 2012-02-17 20:4
On 2012-05-30 22:23, Luigi Rizzo wrote:
> Hi,
> while testing qemu with netmap (see [Note 1] for details) on e1000
> emulation, i noticed that my sender program using a custom backend
> [Note 2] could reach 1 Mpps (million packets per second) but on the
> receive side i was limited to 50 Kpps (and
Hi,
while testing qemu with netmap (see [Note 1] for details) on e1000
emulation, i noticed that my sender program using a custom backend
[Note 2] could reach 1 Mpps (million packets per second) but on the
receive side i was limited to 50 Kpps (and CPU always below 5%).
The problem was fixed by th
22 matches
Mail list logo
