search for: 5600u

...ing. Fix this by calling vhost_vq_avail_empty() instead. This issue could be noticed by doing netperf TCP_RR benchmark as client from guest (but not host). With this fix, TCP_RR from guest to localhost restores from 1375.91 trans per sec to 55235.28 trans per sec on my laptop (Intel(R) Core(TM) i7-5600U CPU @ 2.60GHz). Fixes: 030881372460 ("vhost_net: basic polling support") Signed-off-by: Jason Wang <jasowang at redhat.com> --- - The patch is needed for -stable --- drivers/vhost/net.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/vhost/net.c b/drive...

...ing. Fix this by calling vhost_vq_avail_empty() instead. This issue could be noticed by doing netperf TCP_RR benchmark as client from guest (but not host). With this fix, TCP_RR from guest to localhost restores from 1375.91 trans per sec to 55235.28 trans per sec on my laptop (Intel(R) Core(TM) i7-5600U CPU @ 2.60GHz). Fixes: 030881372460 ("vhost_net: basic polling support") Signed-off-by: Jason Wang <jasowang at redhat.com> --- - The patch is needed for -stable --- drivers/vhost/net.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/vhost/net.c b/drive...

...ing. Fix this by calling vhost_vq_avail_empty() instead. This issue could be noticed by doing netperf TCP_RR benchmark as client from guest (but not host). With this fix, TCP_RR from guest to localhost restores from 1375.91 trans per sec to 55235.28 trans per sec on my laptop (Intel(R) Core(TM) i7-5600U CPU @ 2.60GHz). Fixes: 030881372460 ("vhost_net: basic polling support") Signed-off-by: Jason Wang <jasowang at redhat.com> --- - The patch is needed for -stable - Changes from V1: enable vq notification when needed --- drivers/vhost/net.c | 7 ++++++- 1 file changed, 6 insertions...

...ing. Fix this by calling vhost_vq_avail_empty() instead. This issue could be noticed by doing netperf TCP_RR benchmark as client from guest (but not host). With this fix, TCP_RR from guest to localhost restores from 1375.91 trans per sec to 55235.28 trans per sec on my laptop (Intel(R) Core(TM) i7-5600U CPU @ 2.60GHz). Fixes: 030881372460 ("vhost_net: basic polling support") Signed-off-by: Jason Wang <jasowang at redhat.com> --- - The patch is needed for -stable - Changes from V1: enable vq notification when needed --- drivers/vhost/net.c | 7 ++++++- 1 file changed, 6 insertions...

...d, using all of descriptor buffer might cause a slowdown. Rather you should be able to get about the same speedup, but from skipping checking the used ring in virtio. > Virito-user + vhost_kernel + XDP_DROP gives about ~10% improvement on > TX from 4.8Mpps to 5.3Mpps on Intel(R) Core(TM) i7-5600U CPU @ > 2.60GHz. > > Signed-off-by: Jason Wang <jasowang at redhat.com> > --- > drivers/vhost/vhost.c | 19 ++++++++++++------- > 1 file changed, 12 insertions(+), 7 deletions(-) > > diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c > index 3a5f81a66d34....

Hi: This series implement basic in order feature support for vhost_net. This feature requires both driver and device to use descriptors in order which can simplify the implementation and optimizaton for both side. The series also implement a simple optimization that avoid read available ring. Test shows 10% performance improvement. More optimizations could be done on top. Jason Wang (3):

...vhost_vq_avail_empty() instead. > > This issue could be noticed by doing netperf TCP_RR benchmark as > client from guest (but not host). With this fix, TCP_RR from guest to > localhost restores from 1375.91 trans per sec to 55235.28 trans per > sec on my laptop (Intel(R) Core(TM) i7-5600U CPU @ 2.60GHz). > > Fixes: 030881372460 ("vhost_net: basic polling support") > Signed-off-by: Jason Wang <jasowang at redhat.com> > --- > - The patch is needed for -stable > --- > drivers/vhost/net.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) &...

Device use descriptors table in order, so there's no need to read index from available ring. This eliminate the cache contention on avail ring completely. Virito-user + vhost_kernel + XDP_DROP gives about ~10% improvement on TX from 4.8Mpps to 5.3Mpps on Intel(R) Core(TM) i7-5600U CPU @ 2.60GHz. Signed-off-by: Jason Wang <jasowang at redhat.com> --- drivers/vhost/vhost.c | 19 ++++++++++++------- 1 file changed, 12 insertions(+), 7 deletions(-) diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c index 3a5f81a66d34..c8be151bc897 100644 --- a/drivers/vhost/vhos...

Hi: This series implement batch updating of used ring for TX. This help to reduce the cache contention on used ring. The idea is first split datacopy path from zerocopy, and do only batching for datacopy. This is because zercopy had already supported its own batching. TX PPS was increased 25.8% and Netperf TCP does not show obvious differences. The split of datapath will also be helpful for

Hi: This series implement batch updating of used ring for TX. This help to reduce the cache contention on used ring. The idea is first split datacopy path from zerocopy, and do only batching for datacopy. This is because zercopy had already supported its own batching. TX PPS was increased 25.8% and Netperf TCP does not show obvious differences. The split of datapath will also be helpful for

v1 was here: https://www.redhat.com/archives/libguestfs/2016-March/thread.html#00157 Not running the 'hwclock' command reduces boot times considerably. However I'm not sure if it is safe. See the question I posted on qemu-devel: http://thread.gmane.org/gmane.comp.emulators.qemu/402194 At the moment, about 50% of the time is consumed by SeaBIOS. Of this, about ⅓rd is SGABIOS

...[ 0.031307] smpboot: Max logical packages: 1 [ 0.031963] x2apic enabled nbdkit: debug: starting worker thread nbd.14 [ 0.032006] Switched APIC routing to physical x2apic. [ 0.034000] ..TIMER: vector=0x30 apic1=0 pin1=2 apic2=-1 pin2=-1 [ 0.034000] smpboot: CPU0: Intel(R) Core(TM) i7-5600U nbdkit: debug: starting worker thread nbd.15 CPU @ 2.60GHz (family: 0x6, model: 0x3d, stepping: 0x4) [ 0.034071] Performance Events: Broadwell events, Intel PMU driver. [ 0.035023] ... version: 2 [ 0.035529] ... bit width: 48 [ 0.036007] ... generic registers...

When converting from a single transaction to a linked list, I forgot to free the storage for each member of the list. Reported-by: Richard W.M. Jones <rjones at redhat.com> Fixes: 7f5bb9bf13f041ea7702bda557d9dd668bc3423a Signed-off-by: Eric Blake <eblake at redhat.com> --- I'm still not sure why 'make check' passes while 'make check-valgrind' fails for

Lots of changes since v2, too much to remember or summarize. Please ignore patch 11/11, it's just for my testing. Rich.