greearb / ath10k-ct

Stand-alone ath10k driver based on Candela Technologies Linux kernel.
111 stars 41 forks source link

bad ping time on ubuntu vs openwrt #194

Closed gsustek closed 2 years ago

gsustek commented 2 years ago

These are the steps how to build kernel for x86 machine with 9984 wifi in AP mode card. OpenWRT and DDWRT has better performance then regular ubuntu 20.04 LTS distro. The one who has full linux distro also as router, can you check your ping/packet loss?

git clone git://git.launchpad.net/~ubuntu-kernel-test/ubuntu/+source/linux/+git/mainline-crack v5.15.5v git reset --hard b10c942b042c318e4d7783cd529f0fd5f468eed6v patch git apply -v ../ath10k-ct-5.15/patches/0001-wireless-Relax-beacon_int_min_gcd-and-ADHOC-check.patch git apply -v ../0404-ath-reg-Relax-application-of-default-world-regd.patch git apply -v ../0404-ath-reg-Relax-application-of-default-world-regd.patch build kernel with those connfig (CONFIG_ATH_REG_DYNAMIC_USER_REG_HINTS=y CONFIG_ATH_REG_DYNAMIC_USER_CERT_TESTING=y CONFIG_ATH10K_DFS_CERTIFIED=y CONFIG_CFG80211_CERTIFICATION_ONUS=y) It seems that is working for now! but ping is fairly high.

64 bytes from 192.168.168.1: icmp_seq=662 ttl=64 time=12.231 ms 64 bytes from 192.168.168.1: icmp_seq=663 ttl=64 time=9.781 ms 64 bytes from 192.168.168.1: icmp_seq=664 ttl=64 time=2.943 ms 64 bytes from 192.168.168.1: icmp_seq=665 ttl=64 time=8.656 ms 64 bytes from 192.168.168.1: icmp_seq=666 ttl=64 time=8.329 ms 64 bytes from 192.168.168.1: icmp_seq=667 ttl=64 time=3.117 ms Request timeout for icmp_seq 668 64 bytes from 192.168.168.1: icmp_seq=669 ttl=64 time=2.529 ms 64 bytes from 192.168.168.1: icmp_seq=670 ttl=64 time=1.881 ms 64 bytes from 192.168.168.1: icmp_seq=671 ttl=64 time=2.522 ms 64 bytes from 192.168.168.1: icmp_seq=672 ttl=64 time=1.523 ms 64 bytes from 192.168.168.1: icmp_seq=673 ttl=64 time=9.509 ms 64 bytes from 192.168.168.1: icmp_seq=674 ttl=64 time=9.758 ms 64 bytes from 192.168.168.1: icmp_seq=675 ttl=64 time=3.261 ms 64 bytes from 192.168.168.1: icmp_seq=676 ttl=64 time=9.246 ms 64 bytes from 192.168.168.1: icmp_seq=677 ttl=64 time=15.283 ms 64 bytes from 192.168.168.1: icmp_seq=678 ttl=64 time=71.470 ms 64 bytes from 192.168.168.1: icmp_seq=679 ttl=64 time=3.129 ms 64 bytes from 192.168.168.1: icmp_seq=680 ttl=64 time=6.796 ms

greearb commented 2 years ago

that could be perfectly legit if station is in power save mode, and anyway, unless you are confident this is some bug in my ath10k, then not sure it is worth a bug?

gsustek commented 2 years ago

i will close it and it should stay for documentation purposes.:-) do not forget to switch to "performance" cpu governor.

for i in {0..7}; do echo performance > /sys/devices/system/cpu/cpufreq/policy$i/scaling_governor; done for i in {0..7}; do cat /sys/devices/system/cpu/cpu$i/cpufreq/scaling_governor; done