On 05.04.2019 06:26, Toshiaki Makita wrote:
On 2019/04/05 5:22, Rafał Miłecki wrote:
On 04.04.2019 17:17, Toshiaki Makita wrote:
On 19/04/04 (木) 21:57:15, Rafał Miłecki wrote:
I'd like to report a regression that goes back to the 2015. I know
it's damn
late, but the good thing is, the regression is still easy to
reproduce, verify &
revert.
Long story short, starting with the commit 66e5133f19e9 ("vlan: Add
GRO support
for non hardware accelerated vlan") - which first hit kernel 4.2 - NAT
performance of my router dropped by 30% - 40%.
My hardware is BCM47094 SoC (dual core ARM) with integrated network
controller
and external BCM53012 switch.
Relevant setup:
* SoC network controller is wired to the hardware switch
* Switch passes 802.1q frames with VID 1 to four LAN ports
* Switch passes 802.1q frames with VID 2 to WAN port
* Linux does NAT for LAN (eth0.1) to WAN (eth0.2)
* Linux uses pfifo and "echo 2 > rps_cpus"
* Ryzen 5 PRO 2500U (x86_64) laptop connected to a LAN port
* Intel i7-2670QM laptop connected to a WAN port
* Speed of LAN to WAN measured using iperf & TCP over 10 minutes
1) 5.1.0-rc3
[ 6] 0.0-600.0 sec 39.9 GBytes 572 Mbits/sec
2) 5.1.0-rc3 + rtcache patch
[ 6] 0.0-600.0 sec 40.0 GBytes 572 Mbits/sec
3) 5.1.0-rc3 + disable GRO support
[ 6] 0.0-300.4 sec 27.5 GBytes 786 Mbits/sec
4) 5.1.0-rc3 + rtcache patch + disable GRO support
[ 6] 0.0-600.0 sec 65.6 GBytes 939 Mbits/sec
Did you test it with disabling GRO by ethtool -K?
Oh, I didn't know about such possibility! I just tested:
1) Kernel with GRO support left in place (no local patch disabling it)
2) ethtool -K eth0 gro off
and it bumped my NAT performance from 576 Mb/s to 939 Mb/s. I can reliably
break/fix NAT performance by just calling ethtool -K eth0 gro on/off.
Is this the result with your reverting patch?
Previous results were coming from kernel with patched
vlan_offload_init() - see
diff at the end of my first e-mail.
It's late night in Japan so I think I will try to reproduce it tomorrow.
My test results:
Receiving packets from eth0.10, forwarding them to eth0.20 and applying
MASQUERADE on eth0.20, using i40e 25G NIC on kernel 4.20.13.
Disabled rxvlan by ethtool -K to exercise vlan_gro_receive().
Measured TCP throughput by netperf.
GRO on : 17 Gbps
GRO off: 5 Gbps
So I failed to reproduce your problem.
:( Thanks for trying & checking that!
Would you check the CPU usage by "mpstat -P ALL" or similar (like "sar
-u ALL -P ALL") to check if the traffic is able to consume 100% CPU on
your machine?
1) ethtool -K eth0 gro on + iperf running (577 Mb/s)
root@OpenWrt:/# mpstat -P ALL 10 3
Linux 5.1.0-rc3+ (OpenWrt) 03/27/19 _armv7l_ (2 CPU)
16:33:40 CPU %usr %nice %sys %iowait %irq %soft %steal
%guest %idle
16:33:50 all 0.00 0.00 0.00 0.00 0.00 58.79 0.00
0.00 41.21
16:33:50 0 0.00 0.00 0.00 0.00 0.00 100.00 0.00
0.00 0.00
16:33:50 1 0.00 0.00 0.00 0.00 0.00 17.58 0.00
0.00 82.42
16:33:50 CPU %usr %nice %sys %iowait %irq %soft %steal
%guest %idle
16:34:00 all 0.00 0.00 0.05 0.00 0.00 59.44 0.00
0.00 40.51
16:34:00 0 0.00 0.00 0.10 0.00 0.00 99.90 0.00
0.00 0.00
16:34:00 1 0.00 0.00 0.00 0.00 0.00 18.98 0.00
0.00 81.02
16:34:00 CPU %usr %nice %sys %iowait %irq %soft %steal
%guest %idle
16:34:10 all 0.00 0.00 0.00 0.00 0.00 59.59 0.00
0.00 40.41
16:34:10 0 0.00 0.00 0.00 0.00 0.00 100.00 0.00
0.00 0.00
16:34:10 1 0.00 0.00 0.00 0.00 0.00 19.18 0.00
0.00 80.82
Average: CPU %usr %nice %sys %iowait %irq %soft %steal
%guest %idle
Average: all 0.00 0.00 0.02 0.00 0.00 59.27 0.00
0.00 40.71
Average: 0 0.00 0.00 0.03 0.00 0.00 99.97 0.00
0.00 0.00
Average: 1 0.00 0.00 0.00 0.00 0.00 18.58 0.00
0.00 81.42
2) ethtool -K eth0 gro off + iperf running (941 Mb/s)
root@OpenWrt:/# mpstat -P ALL 10 3
Linux 5.1.0-rc3+ (OpenWrt) 03/27/19 _armv7l_ (2 CPU)
16:34:39 CPU %usr %nice %sys %iowait %irq %soft %steal
%guest %idle
16:34:49 all 0.00 0.00 0.05 0.00 0.00 86.91 0.00
0.00 13.04
16:34:49 0 0.00 0.00 0.10 0.00 0.00 78.22 0.00
0.00 21.68
16:34:49 1 0.00 0.00 0.00 0.00 0.00 95.60 0.00
0.00 4.40
16:34:49 CPU %usr %nice %sys %iowait %irq %soft %steal
%guest %idle
16:34:59 all 0.00 0.00 0.10 0.00 0.00 87.06 0.00
0.00 12.84
16:34:59 0 0.00 0.00 0.20 0.00 0.00 79.72 0.00
0.00 20.08
16:34:59 1 0.00 0.00 0.00 0.00 0.00 94.41 0.00
0.00 5.59
16:34:59 CPU %usr %nice %sys %iowait %irq %soft %steal
%guest %idle
16:35:09 all 0.00 0.00 0.05 0.00 0.00 85.71 0.00
0.00 14.24
16:35:09 0 0.00 0.00 0.10 0.00 0.00 79.42 0.00
0.00 20.48
16:35:09 1 0.00 0.00 0.00 0.00 0.00 92.01 0.00
0.00 7.99
Average: CPU %usr %nice %sys %iowait %irq %soft %steal
%guest %idle
Average: all 0.00 0.00 0.07 0.00 0.00 86.56 0.00
0.00 13.37
Average: 0 0.00 0.00 0.13 0.00 0.00 79.12 0.00
0.00 20.75
Average: 1 0.00 0.00 0.00 0.00 0.00 94.01 0.00
0.00 5.99
3) System idle (no iperf)
root@OpenWrt:/# mpstat -P ALL 10 1
Linux 5.1.0-rc3+ (OpenWrt) 03/27/19 _armv7l_ (2 CPU)
16:35:31 CPU %usr %nice %sys %iowait %irq %soft %steal
%guest %idle
16:35:41 all 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 100.00
16:35:41 0 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 100.00
16:35:41 1 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 100.00
Average: CPU %usr %nice %sys %iowait %irq %soft %steal
%guest %idle
Average: all 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 100.00
Average: 0 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 100.00
Average: 1 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 100.00
If CPU is 100%, perf may help us analyze your problem. If it's
available, try running below while testing:
# perf record -a -g -- sleep 5
And then run this after testing:
# perf report --no-child
I can see my CPU 0 is fully loaded when using "gro on". I'll try perf now.
