* [PATCH net-next] hyperv: Increase the buffer length for netvsc_channel_cb()
@ 2014-08-01 15:55 Haiyang Zhang
2014-08-05 20:03 ` David Miller
0 siblings, 1 reply; 6+ messages in thread
From: Haiyang Zhang @ 2014-08-01 15:55 UTC (permalink / raw)
To: davem, netdev; +Cc: olaf, jasowang, driverdev-devel, linux-kernel, haiyangz
When the buffer is too small for a packet from VMBus, a bigger buffer will be
allocated in netvsc_channel_cb() and retry reading the packet from VMBus.
Increasing this buffer size will reduce the retry overhead.
Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com>
Reviewed-by: Dexuan Cui <decui@microsoft.com>
---
drivers/net/hyperv/hyperv_net.h | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/net/hyperv/hyperv_net.h b/drivers/net/hyperv/hyperv_net.h
index 24441ae..3960acd 100644
--- a/drivers/net/hyperv/hyperv_net.h
+++ b/drivers/net/hyperv/hyperv_net.h
@@ -591,7 +591,7 @@ struct nvsp_message {
#define NETVSC_RECEIVE_BUFFER_ID 0xcafe
-#define NETVSC_PACKET_SIZE 2048
+#define NETVSC_PACKET_SIZE 4096
#define VRSS_SEND_TAB_SIZE 16
--
1.7.1
^ permalink raw reply related [flat|nested] 6+ messages in thread
* [PATCH net-next] hyperv: Increase the buffer length for netvsc_channel_cb()
@ 2014-08-04 19:33 Haiyang Zhang
2014-08-04 22:08 ` David Miller
0 siblings, 1 reply; 6+ messages in thread
From: Haiyang Zhang @ 2014-08-04 19:33 UTC (permalink / raw)
To: davem, netdev; +Cc: olaf, jasowang, driverdev-devel, linux-kernel, haiyangz
When the buffer is too small for a packet from VMBus, a bigger buffer will be
allocated in netvsc_channel_cb() and retry reading the packet from VMBus.
Increasing this buffer size will reduce the retry overhead.
Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com>
Reviewed-by: Dexuan Cui <decui@microsoft.com>
---
drivers/net/hyperv/hyperv_net.h | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/net/hyperv/hyperv_net.h b/drivers/net/hyperv/hyperv_net.h
index 24441ae..3960acd 100644
--- a/drivers/net/hyperv/hyperv_net.h
+++ b/drivers/net/hyperv/hyperv_net.h
@@ -591,7 +591,7 @@ struct nvsp_message {
#define NETVSC_RECEIVE_BUFFER_ID 0xcafe
-#define NETVSC_PACKET_SIZE 2048
+#define NETVSC_PACKET_SIZE 4096
#define VRSS_SEND_TAB_SIZE 16
--
1.7.1
^ permalink raw reply related [flat|nested] 6+ messages in thread
* Re: [PATCH net-next] hyperv: Increase the buffer length for netvsc_channel_cb()
2014-08-04 19:33 Haiyang Zhang
@ 2014-08-04 22:08 ` David Miller
2014-08-04 22:34 ` Haiyang Zhang
0 siblings, 1 reply; 6+ messages in thread
From: David Miller @ 2014-08-04 22:08 UTC (permalink / raw)
To: haiyangz; +Cc: olaf, netdev, jasowang, driverdev-devel, linux-kernel
From: Haiyang Zhang <haiyangz@microsoft.com>
Date: Mon, 4 Aug 2014 12:33:45 -0700
> When the buffer is too small for a packet from VMBus, a bigger buffer will be
> allocated in netvsc_channel_cb() and retry reading the packet from VMBus.
> Increasing this buffer size will reduce the retry overhead.
>
> Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com>
> Reviewed-by: Dexuan Cui <decui@microsoft.com>
Now netvsc_device is significantly larger than a page, you should probably
therefore allocate netvsc_device->cb_buffer indirectly instead of inlining
it into the structure.
^ permalink raw reply [flat|nested] 6+ messages in thread
* RE: [PATCH net-next] hyperv: Increase the buffer length for netvsc_channel_cb()
2014-08-04 22:08 ` David Miller
@ 2014-08-04 22:34 ` Haiyang Zhang
0 siblings, 0 replies; 6+ messages in thread
From: Haiyang Zhang @ 2014-08-04 22:34 UTC (permalink / raw)
To: David Miller
Cc: netdev@vger.kernel.org, KY Srinivasan, olaf@aepfle.de,
jasowang@redhat.com, linux-kernel@vger.kernel.org,
driverdev-devel@linuxdriverproject.org
> -----Original Message-----
> From: David Miller [mailto:davem@davemloft.net]
> Sent: Monday, August 4, 2014 6:09 PM
> To: Haiyang Zhang
> Cc: netdev@vger.kernel.org; KY Srinivasan; olaf@aepfle.de;
> jasowang@redhat.com; linux-kernel@vger.kernel.org; driverdev-
> devel@linuxdriverproject.org
> Subject: Re: [PATCH net-next] hyperv: Increase the buffer length for
> netvsc_channel_cb()
>
> From: Haiyang Zhang <haiyangz@microsoft.com>
> Date: Mon, 4 Aug 2014 12:33:45 -0700
>
> > When the buffer is too small for a packet from VMBus, a bigger buffer
> > will be allocated in netvsc_channel_cb() and retry reading the packet from
> VMBus.
> > Increasing this buffer size will reduce the retry overhead.
> >
> > Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com>
> > Reviewed-by: Dexuan Cui <decui@microsoft.com>
>
> Now netvsc_device is significantly larger than a page, you should probably
> therefore allocate netvsc_device->cb_buffer indirectly instead of inlining it into
> the structure.
I will make this update.
Thanks,
- Haiyang
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH net-next] hyperv: Increase the buffer length for netvsc_channel_cb()
2014-08-01 15:55 [PATCH net-next] hyperv: Increase the buffer length for netvsc_channel_cb() Haiyang Zhang
@ 2014-08-05 20:03 ` David Miller
2014-08-05 20:12 ` Haiyang Zhang
0 siblings, 1 reply; 6+ messages in thread
From: David Miller @ 2014-08-05 20:03 UTC (permalink / raw)
To: thomas1206
Cc: olaf, netdev, jasowang, driverdev-devel, linux-kernel, haiyangz
From: Haiyang Zhang <thomas1206@gmail.com>
Date: Fri, 1 Aug 2014 15:55:30 +0000
> When the buffer is too small for a packet from VMBus, a bigger buffer will be
> allocated in netvsc_channel_cb() and retry reading the packet from VMBus.
> Increasing this buffer size will reduce the retry overhead.
>
> Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com>
> Reviewed-by: Dexuan Cui <decui@microsoft.com>
I'm not applying this, sorry.
I said that because the structure that contains this buffer is now
larger than a page, you should allocate the CB buffer externally
via another kmalloc() or via the page allocator.
^ permalink raw reply [flat|nested] 6+ messages in thread
* RE: [PATCH net-next] hyperv: Increase the buffer length for netvsc_channel_cb()
2014-08-05 20:03 ` David Miller
@ 2014-08-05 20:12 ` Haiyang Zhang
0 siblings, 0 replies; 6+ messages in thread
From: Haiyang Zhang @ 2014-08-05 20:12 UTC (permalink / raw)
To: David Miller, thomas1206@gmail.com
Cc: netdev@vger.kernel.org, KY Srinivasan, olaf@aepfle.de,
jasowang@redhat.com, linux-kernel@vger.kernel.org,
driverdev-devel@linuxdriverproject.org
> -----Original Message-----
> From: David Miller [mailto:davem@davemloft.net]
> Sent: Tuesday, August 5, 2014 4:04 PM
> To: thomas1206@gmail.com
> Cc: netdev@vger.kernel.org; Haiyang Zhang; KY Srinivasan; olaf@aepfle.de;
> jasowang@redhat.com; linux-kernel@vger.kernel.org; driverdev-
> devel@linuxdriverproject.org
> Subject: Re: [PATCH net-next] hyperv: Increase the buffer length for
> netvsc_channel_cb()
>
> From: Haiyang Zhang <thomas1206@gmail.com>
> Date: Fri, 1 Aug 2014 15:55:30 +0000
>
> > When the buffer is too small for a packet from VMBus, a bigger buffer
> will be
> > allocated in netvsc_channel_cb() and retry reading the packet from
> VMBus.
> > Increasing this buffer size will reduce the retry overhead.
> >
> > Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com>
> > Reviewed-by: Dexuan Cui <decui@microsoft.com>
>
> I'm not applying this, sorry.
>
> I said that because the structure that contains this buffer is now
> larger than a page, you should allocate the CB buffer externally
> via another kmalloc() or via the page allocator.
This must have been re-sent accidentally when we are updating our
Email server... I will update the patch and re-submit it soon.
Thanks,
- Haiyang
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2014-08-05 20:12 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2014-08-01 15:55 [PATCH net-next] hyperv: Increase the buffer length for netvsc_channel_cb() Haiyang Zhang
2014-08-05 20:03 ` David Miller
2014-08-05 20:12 ` Haiyang Zhang
-- strict thread matches above, loose matches on Subject: below --
2014-08-04 19:33 Haiyang Zhang
2014-08-04 22:08 ` David Miller
2014-08-04 22:34 ` Haiyang Zhang
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).