Something is off, cpu0: SoC ID 323 is not part of IPQ8074 family and then DalChipInfo.c:289 Chip Major Version mismatch.
Are you sure it's the RAX120v2 you got?
Sat Aug 5 05:09:00 2023 kern.err kernel: [ 1.375854] cpu cpu0: SoC ID 323 is not part of IPQ8074 family, limiting to 1.4GHz!
...
Sat Aug 5 05:09:00 2023 kern.err kernel: [ 10.419167] qcom-q6v5-wcss-pil cd00000.q6v5_wcss: fatal error received:
Sat Aug 5 05:09:00 2023 kern.err kernel: [ 10.419167] QC Image Version: QC_IMAGE_VERSION_STRING=WLAN.HK.2.9.0.1-01862-QCAHKSWPL_SILICONZ-1
Sat Aug 5 05:09:00 2023 kern.err kernel: [ 10.419167] Image Variant : IMAGE_VARIANT_STRING=8074.wlanfw.eval_v2Q
Sat Aug 5 05:09:00 2023 kern.err kernel: [ 10.419167]
Sat Aug 5 05:09:00 2023 kern.err kernel: [ 10.419167] DalChipInfo.c:289 Chip Major Version mismatch between Hardware:0x1 and Software:0x2
Sat Aug 5 05:09:00 2023 kern.err kernel: [ 10.419167] param0 :1, param1 :2, param2 :zero.
Addition: To check the revision you got (you want ... 02):
To clarify: the image works well for me, I just use lan5 as WAN to use on a 10G switch, Wifi works. The above log is the system log from @nschimme's post, not mine.
Are you using your own build or @patient0 's build? Maybe the build is mis-configured.
try my build, see if it works for you(only need to flash the systemupgrade.bin ) . I have flashed and confirmed this image several times, wifi is working on my side.
when you do systemupgrade, uncheck Keep settings and retain the current configuration. otherwise your old wrong configuration might be still there.
===UPDATE===
when you do systemupgrade, uncheck Keep settings and retain the current configuration. otherwise your old wrong configuration might be still there.
I am not 100% sure if this is sufficient.
for an even cleaner flash:
flash official image, do a factory reset
flash ramfs image via the official upgrade page
ssh into the router, run mtd erase rootfs, reboot
flash the sysupgrade.bin via the luci page
(wait long enough after each step to fully finish)
Hello, since the topic came to life, so will I. The question is, are there any reasons why my wifi results are so embarrassing? Still near the router (20cm) I have about 500mbps, and through the wall I don't exceed 100mbps. Maybe my router is faulty? Does anyone have a relatively new device with wifi6 and can test it? Of course, I did all the additions, i.e. clean flash. Testing on full wpad-mbetls package. For comparison, in my area on gl-inet mt3000 I get about 1.6gbps over wifi on MBP14 M2.
It's the same for me too, not very fast at all (around 550 to 600mps 20 to 30cm away). I did hope that once it got official support maybe it could be worked on to improve.
Edit: It does not mention V2 in the stock UI and the sticker on the bottom doesn't mention V2 either. I have a V1.
I was able to revert to stock using nmrpflash and going back to the factory firmware. I used these instructions on my Mac but you might need to replace the ethernet device with your own equivalent.
sudo ./nmrpflash -i en9 -f RAX120-V1.2.8.40.img -A 10.0.0.2 -a 10.0.0.1
Waiting for Ethernet connection (Ctrl-C to skip).
Advertising NMRP server on en9 ... |
Received configuration request from 38:94:ed:a5:f7:2e.
Sending configuration: 10.0.0.1/24.
Received upload request: filename 'firmware'.
Uploading RAX120-V1.2.8.40.img ... Warning: TFTP block rollover. Upload might fail!
OK (63047809 b)
Waiting for remote to respond.
Remote finished. Closing connection.
Reboot your device now.
In that case, your best bet might be to revert to the stock OS.
It will have proprietary wifi drivers, so that you get almost full/official 802.11ax support, which you will never get with the opensource drivers here. (But the 8074 chip is probably not quite fully "final ax" compatible, still.)
Hi! Im not fluent in English, so writing from my heart.
I'm trying this build https://github.com/boretom/openwrt-fork/releases/tag/2023-08-05 and find some strange things. First, there is low 5Ghz transmit power, around -60 dBm at 0.5m. I guess at stock firmware 1 5Ghz network uses both radios because the guest and main network always work at the same channels and the same HT(with -30dBm at 0.5m both).
Another problem is SQM packets and other modules, AdguardHome works fine, but luci-app-sqm requests another core version, how can I fix that?
Buildbot build are quite normal "official" OpenWrt builds. Just daily images built from the current code.
Buildbot is currently crunching a build, so the images should be downloadable in some 2-3 hours:
If you want a release build, you might need to wait until mid-2024. Rax120 was added to the main/master now, so it is not part of the 23.05 that was branched off in May. (Rax120 might get backported to 23.05 but that is not to be taken for granted)
First, a random freezes while testing upload speed through wan interface. Speedtest or any bufferbloat tests on upload phase causes freeze and requires router restart.
Then, as I wrote a weak ago, extremely low 5G transmit power, check this screenshots with comparation between stock and latest OWRT.
And another one bug, If you try to update software "base-files", as result there is error message and then fully broken statistics on all ports and impossibility to save any changes (hit apply button, see applying process and as result still have unsaved changes). Other modules installs and updates fine.
What kind of tests you use, while testing speed? I tried to use RAX as iperf3 server and mac with 1x1 MU-MIMO and as result 800 MB/s with drops to 600. And channel utilization 98%, I guess this is maximum for 80 Mhz wide)
I use openspeedtest which is placed on docker on my server, the server is connected by 10gbps link to the switch and the 10gbps switch to the AP to the 5gbps port. So generally RAX is used as a DumbAP because the routing router is also on the server. However, from my tests, as I wrote above, the router is useless, my apartment is not large, only 30m2, and RAX does not cover even half of it with fast wifi, one wall is an insurmountable obstacle. Now, after marge, developers should focus on wifi.
