search for: busy_poll

...07/16/2014 04:38 PM, Varka Bhadram wrote: >> On 07/16/2014 11:51 AM, Jason Wang wrote: >>> Add basic support for rx busy polling. >>> >>> Test was done between a kvm guest and an external host. Two hosts were >>> connected through 40gb mlx4 cards. With both busy_poll and busy_read >>> are set to 50 in guest, 1 byte netperf tcp_rr shows 116% improvement: >>> transaction rate was increased from 9151.94 to 19787.37. >>> >>> Cc: Rusty Russell <rusty at rustcorp.com.au> >>> Cc: Michael S. Tsirkin <mst at redhat.c...

...07/16/2014 04:38 PM, Varka Bhadram wrote: >> On 07/16/2014 11:51 AM, Jason Wang wrote: >>> Add basic support for rx busy polling. >>> >>> Test was done between a kvm guest and an external host. Two hosts were >>> connected through 40gb mlx4 cards. With both busy_poll and busy_read >>> are set to 50 in guest, 1 byte netperf tcp_rr shows 116% improvement: >>> transaction rate was increased from 9151.94 to 19787.37. >>> >>> Cc: Rusty Russell <rusty at rustcorp.com.au> >>> Cc: Michael S. Tsirkin <mst at redhat.c...

On 07/16/2014 11:51 AM, Jason Wang wrote: > Add basic support for rx busy polling. > > Test was done between a kvm guest and an external host. Two hosts were > connected through 40gb mlx4 cards. With both busy_poll and busy_read > are set to 50 in guest, 1 byte netperf tcp_rr shows 116% improvement: > transaction rate was increased from 9151.94 to 19787.37. > > Cc: Rusty Russell <rusty at rustcorp.com.au> > Cc: Michael S. Tsirkin <mst at redhat.com> > Cc: Vlad Yasevich <vyasev...

On 07/16/2014 11:51 AM, Jason Wang wrote: > Add basic support for rx busy polling. > > Test was done between a kvm guest and an external host. Two hosts were > connected through 40gb mlx4 cards. With both busy_poll and busy_read > are set to 50 in guest, 1 byte netperf tcp_rr shows 116% improvement: > transaction rate was increased from 9151.94 to 19787.37. > > Cc: Rusty Russell <rusty at rustcorp.com.au> > Cc: Michael S. Tsirkin <mst at redhat.com> > Cc: Vlad Yasevich <vyasev...

...nlocks, this series re-uses NAPI state to synchonrize between NAPI and busy polling. This grealy simplified the codes and reduce the overheads of spinlocks for normal NAPI fast path. Test was done between a kvm guest and an external host. Two hosts were connected through 40gb mlx4 cards. With both busy_poll and busy_read are set to 50 in guest, 1 byte netperf tcp_rr shows 127% improvement: transaction rate was increased from 8353.33 to 18966.87. Changes from V2: - Avoid introducing new states and spinlocks by reusuing the NAPI state - Fix the budget calculation in virtnet_poll() - Drop patch 1/3 f...

...nlocks, this series re-uses NAPI state to synchonrize between NAPI and busy polling. This grealy simplified the codes and reduce the overheads of spinlocks for normal NAPI fast path. Test was done between a kvm guest and an external host. Two hosts were connected through 40gb mlx4 cards. With both busy_poll and busy_read are set to 50 in guest, 1 byte netperf tcp_rr shows 127% improvement: transaction rate was increased from 8353.33 to 18966.87. Changes from V2: - Avoid introducing new states and spinlocks by reusuing the NAPI state - Fix the budget calculation in virtnet_poll() - Drop patch 1/3 f...

...rx busy polling support. This was useful for reduing the latency for a kvm guest. Patch 1-2 introduces helpers which is used for rx busy polling. Patch 3 implement the main function. Test was done between a kvm guest and an external host. Two hosts were connected through 40gb mlx4 cards. With both busy_poll and busy_read are set to 50 in guest, 1 byte netperf tcp_rr shows 116% improvement: transaction rate was increased from 9151.94 to 19787.37. Changes from V1: - split the patch info smaller ones - add more details about test setup/configuration Please review. Jason Wang (3): virtio-net: introdu...

...rx busy polling support. This was useful for reduing the latency for a kvm guest. Patch 1-2 introduces helpers which is used for rx busy polling. Patch 3 implement the main function. Test was done between a kvm guest and an external host. Two hosts were connected through 40gb mlx4 cards. With both busy_poll and busy_read are set to 50 in guest, 1 byte netperf tcp_rr shows 116% improvement: transaction rate was increased from 9151.94 to 19787.37. Changes from V1: - split the patch info smaller ones - add more details about test setup/configuration Please review. Jason Wang (3): virtio-net: introdu...

On 07/16/2014 04:38 PM, Varka Bhadram wrote: > On 07/16/2014 11:51 AM, Jason Wang wrote: >> Add basic support for rx busy polling. >> >> Test was done between a kvm guest and an external host. Two hosts were >> connected through 40gb mlx4 cards. With both busy_poll and busy_read >> are set to 50 in guest, 1 byte netperf tcp_rr shows 116% improvement: >> transaction rate was increased from 9151.94 to 19787.37. >> >> Cc: Rusty Russell <rusty at rustcorp.com.au> >> Cc: Michael S. Tsirkin <mst at redhat.com> >> Cc:...

...Varka Bhadram wrote: >>> On 07/16/2014 11:51 AM, Jason Wang wrote: >>>> Add basic support for rx busy polling. >>>> >>>> Test was done between a kvm guest and an external host. Two hosts were >>>> connected through 40gb mlx4 cards. With both busy_poll and busy_read >>>> are set to 50 in guest, 1 byte netperf tcp_rr shows 116% improvement: >>>> transaction rate was increased from 9151.94 to 19787.37. >>>> >>>> Cc: Rusty Russell <rusty at rustcorp.com.au> >>>> Cc: Michael S. Tsirkin...

...receive and transmit performance. > On the handle_tx side, we poll the sock receive queue at the > same time. handle_rx do that in the same way. > > We set the poll-us=100us and use the iperf3 to test Where/how do you configure poll-us=100us ? Are you talking about /proc/sys/net/core/busy_poll ? p.s. Nice performance boost! :-) > its bandwidth, use the netperf to test throughput and mean > latency. When running the tests, the vhost-net kthread of > that VM, is alway 100% CPU. The commands are shown as below. > > iperf3 -s -D > iperf3 -c IP -i 1 -P 1 -t 20 -M 1400...

On Wed, Jul 16, 2014 at 02:21:47PM +0800, Jason Wang wrote: > Add basic support for rx busy polling. > > Test was done between a kvm guest and an external host. Two hosts were > connected through 40gb mlx4 cards. With both busy_poll and busy_read > are set to 50 in guest, 1 byte netperf tcp_rr shows 116% improvement: > transaction rate was increased from 9151.94 to 19787.37. Pls include data about non polling tests: any effect on cpu utilization there? There could be as we are adding locking. > > Cc: Rusty Russe...

On Wed, Jul 16, 2014 at 02:21:47PM +0800, Jason Wang wrote: > Add basic support for rx busy polling. > > Test was done between a kvm guest and an external host. Two hosts were > connected through 40gb mlx4 cards. With both busy_poll and busy_read > are set to 50 in guest, 1 byte netperf tcp_rr shows 116% improvement: > transaction rate was increased from 9151.94 to 19787.37. Pls include data about non polling tests: any effect on cpu utilization there? There could be as we are adding locking. > > Cc: Rusty Russe...

Add basic support for rx busy polling. Test was done between a kvm guest and an external host. Two hosts were connected through 40gb mlx4 cards. With both busy_poll and busy_read are set to 50 in guest, 1 byte netperf tcp_rr shows 116% improvement: transaction rate was increased from 9151.94 to 19787.37. Cc: Rusty Russell <rusty at rustcorp.com.au> Cc: Michael S. Tsirkin <mst at redhat.com> Cc: Vlad Yasevich <vyasevic at redhat.com> Cc: Eric...

...intel.com> --- drivers/net/virtio_net.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c index e0638e5..39815e4 100644 --- a/drivers/net/virtio_net.c +++ b/drivers/net/virtio_net.c @@ -30,11 +30,11 @@ #include <net/busy_poll.h> static int napi_weight = NAPI_POLL_WEIGHT; -module_param(napi_weight, int, 0444); +module_param(napi_weight, int, S_IRUSR | S_IRGRP | S_IROTH); static bool csum = true, gso = true; -module_param(csum, bool, 0444); -module_param(gso, bool, 0444); +module_param(csum, bool, S_IRUSR | S_IRG...

...intel.com> --- drivers/net/virtio_net.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c index e0638e5..39815e4 100644 --- a/drivers/net/virtio_net.c +++ b/drivers/net/virtio_net.c @@ -30,11 +30,11 @@ #include <net/busy_poll.h> static int napi_weight = NAPI_POLL_WEIGHT; -module_param(napi_weight, int, 0444); +module_param(napi_weight, int, S_IRUSR | S_IRGRP | S_IROTH); static bool csum = true, gso = true; -module_param(csum, bool, 0444); -module_param(gso, bool, 0444); +module_param(csum, bool, S_IRUSR | S_IRG...

...> was useful for reduing the latency for a kvm guest. Patch 1-2 > introduces helpers which is used for rx busy polling. Patch 3 > implement the main function. > > Test was done between a kvm guest and an external host. Two hosts were > connected through 40gb mlx4 cards. With both busy_poll and busy_read are > set to 50 in guest, 1 byte netperf tcp_rr shows 116% improvement: > transaction rate was increased from 9151.94 to 19787.37. > > Changes from V1: > - split the patch info smaller ones > - add more details about test setup/configuration > > Please review...

...> was useful for reduing the latency for a kvm guest. Patch 1-2 > introduces helpers which is used for rx busy polling. Patch 3 > implement the main function. > > Test was done between a kvm guest and an external host. Two hosts were > connected through 40gb mlx4 cards. With both busy_poll and busy_read are > set to 50 in guest, 1 byte netperf tcp_rr shows 116% improvement: > transaction rate was increased from 9151.94 to 19787.37. > > Changes from V1: > - split the patch info smaller ones > - add more details about test setup/configuration > > Please review...

...changed, 241 insertions(+), 1 deletion(-) diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c index 59caa06..f83f6e4 100644 --- a/drivers/net/virtio_net.c +++ b/drivers/net/virtio_net.c @@ -28,6 +28,7 @@ #include <linux/cpu.h> #include <linux/average.h> #include <net/busy_poll.h> +#include <linux/pci.h> static int napi_weight = NAPI_POLL_WEIGHT; module_param(napi_weight, int, 0444); @@ -51,6 +52,23 @@ module_param(gso, bool, 0444); #define MERGEABLE_BUFFER_ALIGN max(L1_CACHE_BYTES, 256) #define VIRTNET_DRIVER_VERSION "1.0.0" +#define __VIRTNET_...

...changed, 241 insertions(+), 1 deletion(-) diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c index 59caa06..f83f6e4 100644 --- a/drivers/net/virtio_net.c +++ b/drivers/net/virtio_net.c @@ -28,6 +28,7 @@ #include <linux/cpu.h> #include <linux/average.h> #include <net/busy_poll.h> +#include <linux/pci.h> static int napi_weight = NAPI_POLL_WEIGHT; module_param(napi_weight, int, 0444); @@ -51,6 +52,23 @@ module_param(gso, bool, 0444); #define MERGEABLE_BUFFER_ALIGN max(L1_CACHE_BYTES, 256) #define VIRTNET_DRIVER_VERSION "1.0.0" +#define __VIRTNET_...