It compiles ok but causes a boot loop. Need an uart bootlog.
If someone wants to test performance difference in their environment here are 2 patches related to L2 cache freq management
https://github.com/dissent1/r7800/commit/448070cf17ef1d288b4f1149c1c683c3de0058ec
@hnyman
update: nvm about 2nd patch (deleted), I've missed one function in the driver so it's ok.
Hi guys, I have a query, I have been downloading the images in development for several days to solve the problems of the wifi, although it seems that the patches are still applied to the problems in the 5GHZ band. Does anyone know how it goes? only the version works well @hnyman. Thank you
what do you mean? what problem with 5ghz?
@Ansuel
I have 50 mb symmetric, the 2.4ghz band me my mobile devices receive 50 mb of download and 50 mb of upload. But in the 5ghz band, if it reaches 1 mb, the router crashes and restarts itself, gives errors in the file's file to the internet, etc. Basically the 5G wifi network falls within seconds of any device connecting.
Runnig it now against latest master, including all the pull requests you've send to NBD.
The router seems to run faultless with the previous ATH firmware and your pull requests, but with this one dmesg is flooded with:
[21952.327921] ath10k_pci 0001:01:00.0: Invalid VHT mcs 15 peer stats
[21952.416374] ath10k_pci 0001:01:00.0: Invalid VHT mcs 15 peer stats
[21954.873996] ath10k_warn: 5 callbacks suppressed
[21954.874012] ath10k_pci 0001:01:00.0: Invalid VHT mcs 15 peer stats
[21955.084142] ath10k_pci 0001:01:00.0: Invalid VHT mcs 15 peer stats
[21955.183536] ath10k_pci 0001:01:00.0: Invalid VHT mcs 15 peer stats
[21958.360060] ath10k_pci 0001:01:00.0: Invalid VHT mcs 15 peer stats
[21959.690105] ath10k_pci 0001:01:00.0: Invalid VHT mcs 15 peer statsThere is a discussion about the same error here: https://www.spinics.net/lists/linux-wireless/msg167522.html
but i can't really make head or tails from it yet. Seems we backport this 4.14 ath10k driver to 4.9?
anyway. ideas to fix this error are appreciated.
According to this msg it is fixed in a firmware, probably in the one that was released a week ago.
https://www.spinics.net/lists/linux-wireless/msg167875.html
Well, iām running that one and I get the error...
Snip from dmesg
[ 22.241382] ath10k_pci 0000:01:00.0: firmware ver 10.4-3.4-00104 api 5 features no-p2p,mfp,peer-flow-ctrl,btcoex-param,allows-mesh-bcast,no-ps crc32 4317698a
[ 31.109029] ath10k_pci 0001:01:00.0: firmware ver 10.4-3.4-00104 api 5 features no-p2p,mfp,peer-flow-ctrl,btcoex-param,allows-mesh-bcast,no-ps crc32 4317698aSo I wondered if we were missing some driver patch?
Try to add this patch into mac80211 patches folder
https://source.codeaurora.org/quic/qsdk/oss/system/feeds/wlan-open/tree/mac80211/patches/a00-0600-0001-mac80211-Ignore-VHT-IE-from-peer-with-wrong-rx_mcs_m.patch?h=coconut
Meanwhile I've enhanced l2 frequency scaling. Not sure if it affects anything in real life, but at least it gets in hand with gpl tarball.
https://github.com/dissent1/r7800/commit/b2351d3c540af443624824bf06f06701b8d5da52
Interesting. Will give both a spin.
Thanks!
tha last firmware include a fix for a bug that we had before the rework of 071 patch
(the firmware crash when you restart it)
Then we continue the same?
that bug i think was caused by the wrong patch.... so... that's not the cause of this problem
Hard to be 100% sure, but it seems this change makes the cpu run about one degree cooler in allmost all thermal zones. For the rest it runs stable since a day now.
Tried it, but still thousands of these errors each day.
I've narrowed it down to 1 device. an Apple Ipad Air 2, with a Murata 339S02541 wifi module, which in turn is based on the Broadcom BCM4334 chip. There are many (recent) fruity devices in this house, using similar chipsets as far as i'm aware, so i'm a bit confused why the ipad seems to trigger the issue.
I've also done a bit of reading here -> http://www.radiotap.org/fields/VHT.html
and i'm reading the following interesting tidbit:
This means that a MCS of 15 is actually perfectly possible, even though it is not part of the normal standard.
I've got k4.14 working
If you want to test k4.14 just apply patches from this post
It was an error in my environment that had led to a boot loop.
1 Like
do you know any difference between using dsa and swconfig?
always wanted to insert a menuconfig option to select between using dsa or swconfig...
1 Like
You can use either or. Both do not coexist. You can try dsa applying this patch
https://github.com/dissent1/r7800/commit/026ae921121b496fb753a656d6d3412f744657b1
i know i already have dsa on a wrt3200acm
so no cons and i have cpu offload?