ath11k high jitter and packet loss when operating in 802.11ax mode

[Andrea Ieri]

On 6.5.0-26-generic, my QCNFA765 cannot really operate in 802.11ax mode.

I have a Thinkpad P16s Gen 2 with a (soldered, sadly) QCNFA765 card, which uses the ath11k driver.
I also have a Ruckus R650 AP, which supports wifi6.

Connecting the two and pinging the AP yields the following:

```
64 bytes from 192.168.16.77: icmp_seq=45 ttl=64 time=22.4 ms
64 bytes from 192.168.16.77: icmp_seq=46 ttl=64 time=29.6 ms
64 bytes from 192.168.16.77: icmp_seq=47 ttl=64 time=26.8 ms
64 bytes from 192.168.16.77: icmp_seq=48 ttl=64 time=2174 ms
64 bytes from 192.168.16.77: icmp_seq=49 ttl=64 time=1158 ms
64 bytes from 192.168.16.77: icmp_seq=50 ttl=64 time=130 ms
64 bytes from 192.168.16.77: icmp_seq=51 ttl=64 time=503 ms
64 bytes from 192.168.16.77: icmp_seq=54 ttl=64 time=1832 ms
64 bytes from 192.168.16.77: icmp_seq=55 ttl=64 time=808 ms
64 bytes from 192.168.16.77: icmp_seq=56 ttl=64 time=9.32 ms
64 bytes from 192.168.16.77: icmp_seq=57 ttl=64 time=2.11 ms
64 bytes from 192.168.16.77: icmp_seq=58 ttl=64 time=25.6 ms
64 bytes from 192.168.16.77: icmp_seq=59 ttl=64 time=30.9 ms
64 bytes from 192.168.16.77: icmp_seq=60 ttl=64 time=95.6 ms
^C
--- 192.168.16.77 ping statistics ---
60 packets transmitted, 55 received, 8.33333% packet loss, time 59273ms
rtt min/avg/max/mdev = 2.105/275.084/2344.792/560.334 ms, pipe 3
```

Those periodic bursts of >1s latency make some webpages fail to load and videochats unusable.

No interference is present in my channel of choice.

Nothing interesting is printed in dmesg while the above occurs.

The above does not happen if:
* I disable wifi 6 in the AP
* I ping a wifi 5 AP (I also have an older R710, which is a 802.11ac AP)
* I ping the R650 in wifi 6 mode from a AX210 on kernel 6.2
* I ping the R650 from a completely different OS and device (e.g. Android)

I'll test with a 24.04 daily image next.

ProblemType: Bug
DistroRelease: Ubuntu 23.10
Package: linux-image-6.5.0-26-generic 6.5.0-26.26
ProcVersionSignature: Ubuntu 6.5.0-26.26-generic 6.5.13
Uname: Linux 6.5.0-26-generic x86_64
NonfreeKernelModules: zfs
ApportVersion: 2.27.0-0ubuntu5
Architecture: amd64
CasperMD5CheckResult: unknown
CurrentDesktop: ubuntu:GNOME
Date: Thu Mar 21 16:58:07 2024
MachineType: {report['dmi.sys.vendor']} {report['dmi.product.name']}
ProcFB: 0 amdgpudrmfb
ProcKernelCmdLine: root=zfs:zroot/ROOT/ubuntu quiet loglevel=4 rtc_cmos.use_acpi_alarm=1 spl.spl_hostid=0x4605c990
RelatedPackageVersions:
 linux-restricted-modules-6.5.0-26-generic N/A
 linux-backports-modules-6.5.0-26-generic N/A
 linux-firmware 20230919.git3672ccab-0ubuntu2.9
SourcePackage: linux
UpgradeStatus: No upgrade log present (probably fresh install)
dmi.bios.date: 01/15/2024
dmi.bios.release: 1.18
dmi.bios.vendor: LENOVO
dmi.bios.version: R2FET38W (1.18 )
dmi.board.asset.tag: Not Available
dmi.board.name: 21K9001NUS
dmi.board.vendor: LENOVO
dmi.board.version: SDK0T76530 WIN
dmi.chassis.asset.tag: No Asset Tag
dmi.chassis.type: 10
dmi.chassis.vendor: LENOVO
dmi.chassis.version: None
dmi.ec.firmware.release: 1.11
dmi.modalias: dmi:bvnLENOVO:bvrR2FET38W(1.18):bd01/15/2024:br1.18:efr1.11:svnLENOVO:pn21K9001NUS:pvrThinkPadP16sGen2:rvnLENOVO:rn21K9001NUS:rvrSDK0T76530WIN:cvnLENOVO:ct10:cvrNone:skuLENOVO_MT_21K9_BU_Think_FM_ThinkPadP16sGen2:
dmi.product.family: ThinkPad P16s Gen 2
dmi.product.name: 21K9001NUS
dmi.product.sku: LENOVO_MT_21K9_BU_Think_FM_ThinkPad P16s Gen 2
dmi.product.version: ThinkPad P16s Gen 2
dmi.sys.vendor: LENOVO

[Andrea Ieri (aieri)]

Well, that was a pleasant surprise! On 6.8.0-11 the problem seems to be completely gone. I guess I'll be upgrading to Noble as soon as possible.

[Andrea Ieri (aieri)]

unfortunately the problem is not actually gone in noble. I have upgraded to the beta and are having significant but intermittent problems. On 6.8.0-31-generic I get periods of normal traffic mixed with 100% packet loss for some seconds. This looks as follows:

64 bytes from 192.168.16.77: icmp_seq=2622 ttl=64 time=2.47 ms
64 bytes from 192.168.16.77: icmp_seq=2623 ttl=64 time=2.80 ms
64 bytes from 192.168.16.77: icmp_seq=2624 ttl=64 time=2.77 ms
From 192.168.16.231 icmp_seq=2637 Destination Host Unreachable
From 192.168.16.231 icmp_seq=2638 Destination Host Unreachable
From 192.168.16.231 icmp_seq=2639 Destination Host Unreachable
From 192.168.16.231 icmp_seq=2640 Destination Host Unreachable
From 192.168.16.231 icmp_seq=2644 Destination Host Unreachable
From 192.168.16.231 icmp_seq=2645 Destination Host Unreachable
64 bytes from 192.168.16.77: icmp_seq=2646 ttl=64 time=162 ms
64 bytes from 192.168.16.77: icmp_seq=2647 ttl=64 time=1.95 ms
64 bytes from 192.168.16.77: icmp_seq=2648 ttl=64 time=44.3 ms
64 bytes from 192.168.16.77: icmp_seq=2649 ttl=64 time=143 ms
64 bytes from 192.168.16.77: icmp_seq=2650 ttl=64 time=2.37 ms
64 bytes from 192.168.16.77: icmp_seq=2651 ttl=64 time=151 ms
64 bytes from 192.168.16.77: icmp_seq=2652 ttl=64 time=2.69 ms
64 bytes from 192.168.16.77: icmp_seq=2653 ttl=64 time=6.66 ms
64 bytes from 192.168.16.77: icmp_seq=2654 ttl=64 time=47.2 ms
64 bytes from 192.168.16.77: icmp_seq=2655 ttl=64 time=40.1 ms
64 bytes from 192.168.16.77: icmp_seq=2656 ttl=64 time=44.9 ms
64 bytes from 192.168.16.77: icmp_seq=2657 ttl=64 time=1583 ms
64 bytes from 192.168.16.77: icmp_seq=2658 ttl=64 time=560 ms
64 bytes from 192.168.16.77: icmp_seq=2659 ttl=64 time=1506 ms
64 bytes from 192.168.16.77: icmp_seq=2660 ttl=64 time=457 ms
64 bytes from 192.168.16.77: icmp_seq=2661 ttl=64 time=809 ms
64 bytes from 192.168.16.77: icmp_seq=2662 ttl=64 time=3014 ms
64 bytes from 192.168.16.77: icmp_seq=2663 ttl=64 time=2005 ms
64 bytes from 192.168.16.77: icmp_seq=2664 ttl=64 time=982 ms
64 bytes from 192.168.16.77: icmp_seq=2665 ttl=64 time=2.47 ms
64 bytes from 192.168.16.77: icmp_seq=2666 ttl=64 time=2471 ms
64 bytes from 192.168.16.77: icmp_seq=2667 ttl=64 time=1464 ms
64 bytes from 192.168.16.77: icmp_seq=2668 ttl=64 time=440 ms
64 bytes from 192.168.16.77: icmp_seq=2669 ttl=64 time=54.2 ms
64 bytes from 192.168.16.77: icmp_seq=2670 ttl=64 time=2.35 ms

If I find some time I'll try rerunning the same test from windows.