New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
FS#1197 - Netgear ipq806x severe wifi problem #6512
Comments
dissent1: I must add that all those throughput numbers are TX (router wise), while RX (router wise) is completely ok and greatly exceeds those values, so the issue with the router when transmitting data, not receiving. |
Ansuel: I can confirm this too.... and i can also confirm that problems are with rx... tx is all good. |
dissent1: A reason and a fix seem to be found. Waiting for more users to provide feedback to be completely sure. |
zefie: On R7500v2 I've observed the following: Images built with lede image builder for appropriate versions: 17.01.3 (prior to november): Stable Edit: Log from 2018-01-08 snapshot:
If there is any more useful information, it has been flooded out of the logs by this repeating over and over (although CPU can be 0 or 1) |
bouwew: Here is the (not yet included) pull request to LEDE: lede-project/source#1559 |
dissent1:
Netgear R7800 (ipq8065, dual QCA9984)
Netgear R7500v2 (ipq8064, dual QCA9980)
Kernel 4.9 - happens on both 17.01 and trunk with k4.9 but k4.4 is unaffected.
Wireless transmits broken/malformed frames that are not detected/corrected on low layer protocols. It can be seen by:
My observations:
The issue affects only Netgear ipq806x devices and not related to qca9984. According to recent findings Netgear R7500v2 with qca9980 also suffers from the issue
https://bugs.lede-project.org/index.php?do=details&task_id=1173
The bug happens only on k4.9 according to tests done between k4.4 and k4.9 with completely same set of wireless backports and firmware + cal/pre-cal data + board (API 1/2 or GPL) data. I've tried all available firmware, cal/pre-cal data and board data options including GPL - no effect.
It doesn't seem to be related to stmmac, I've tried different buffer options.
Actually I'm starting to think that it's smth within uboot that messes with ram, like it does in the last 2 mibs of memory region.
Additional forum investigations https://forum.lede-project.org/t/netgear-r7800-exploration-ipq8065-qca9984/285/481
But I must say that that workaround didn't work for me.
The issue seem to be floating and probably related to some kind of code/byte padding/alignment in memory?
As fact I have been having this issue only with 2.4ghz before and after some unrelated commits now I have it on 5ghz as well.
The text was updated successfully, but these errors were encountered: