-
Notifications
You must be signed in to change notification settings - Fork 33
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
TP-Link CPE210 Ver: 2.0/3.0 wifi not working #722
Comments
Does disabling the Mesh-link (intern-ch13.freifunk.net) fixes it? |
would suggest bug title: TP-Link CPE210 Ver: 2.0/3.0 wifi not working see: https://bugs.openwrt.org/index.php?do=details&task_id=2333 openwrt/openwrt#1951 (comment) 2-1 week ago searched for wifi bug - other devices using similar hardware seems affected too https://lists.berlin.freifunk.net/pipermail/berlin/2019-July/039539.html -> wifi bug hint: seen somebody mentioned in CPE210 bug threads different irqs wifi not working: 2019.08.19_upstream_master_freifunk-berlin-dev-daily-fdfcd4a-tplink_cpe210-v2-sysupgrade.bin wifi working: 2019.08.19_freifunk-berlin-development-1cfbb77-cpe210-v2-sysupgrade.bin |
doesn't appear to be a problem, at least according to bootlog openwrt/openwrt#2468 (comment) AR9344 and AR9342 work fine with this irq setup |
seems to be fixed: mac80211 ath9k: force QCA953x clock to 25MHz
did not test it yet. |
For the records: |
@SvenRoederer is the openwrt fix openwrt/openwrt@4c6fe32 already in the freifunk firmware? Then it would be easy for me to test it. |
@Mitzlaff the current build of branch "daily/upstream-master" contains all openwrt-commit till 23.Nov. So the mentioned one is also included. Looking forward to your test-results. |
on 21.11.2019 did an OpenWrt trunk test Build: Powered by LuCI Master (git-19.324.23065-4fe16b1) / OpenWrt SNAPSHOT r11562-08d9c95417 wifi working: clients can connect to TP-Link CPE210 Ver: 2.0 and CPE210 is connected via wifi to a Freifunk Box -> internet working |
@everloop2 do you know the exact filename of the image you flashed? the filename also contains the githash, to back-reference to exact code used for this build. But thanks for testing |
unfortuneatly -> openwrt-ath79-generic-tplink_cpe210-v2-squashfs-sysupgrade.bin -> o.O hope/think its right: https://github.com/openwrt/openwrt/commits/317e98a5a2 -> Commits on Nov 21, 2019 Kernel Log
|
Ahh, you used a vanilla OpenWrt for your testing. which will match with our code d98f520 |
today tried to build: daily/upstream-master
think patch/commits 3237490 are missing at master: /patches/packages/luci/0202-split-luci-base_PR2817.patch |
let's agree on commit 39e7543 of branch daily/upstream-master
As you can see, the PR2817 patch splits the current luci-base package into luci-lib-base and luci-base, which depends on luci-lib-base. Without the patch the luci-base package is there, as it's in the original luci-feed. With the patch both packages are there and fulfill all dependencies. |
trying again fresh build (10.12.2019 14:50):
|
|
--> 39e7543 (to browser the directory-tree) you can verify with @everloop2 I assume you are affected by the problem mentioned in #714 (comment) . My approach to split the patches-directory (PR #714), was sometimes causing problems on the feeds. I assume you will find in the logs (using V=s make option) a message indicating that the luci-base Makefile will be found in To check for the fault you can run: |
shown straight at compile start ->
|
disabled PR2817: -> same build error No stress - only wanted to test daily/upstream-master branch Will try test build on master, 25Mhz patch should be included: 81aad75 |
master branch working :)
tested as mesh node: wizard + mesh working -> only "set-password-again-bug" (no major) kernelprotokoll.txt test_firmware: https://wiki.freifunk-cottbus.de/firmware:downloads -> bugs?: box did a self-reboot after ~30min uptime test CPE210_v2: http://monitor.berlin.freifunk.net/detail.php?p=uptime&t=uptime&h=cottbus-lausi36-2&s=604800&x=800&y=350 edit: update monitor link -> will test later -> bugs: unusually high cpu load and self reboots after ~40min. -> tested openwrt trunk yesterday -> cpu load seems ok, box did not self reboot, needs further testing |
so master-branch (openWrt-19.07 based) , right? thanks for testing, keep us informed |
Yep, plain master: https://github.com/freifunk-berlin/firmware/tree/master Deactivated IPv6 everywhere: Olsrd, NetworkDevices, SystemStartup -> does not help. -> lantiq boxes: in past (more half year ago) stumbled over a high CPU load bug -> deactivate IPv6 did help. Vanilla OpenWrt does not seem to have high CPU load and self-reboots: https://downloads.openwrt.org/snapshots/targets/ath79/generic/tplink_cpe210-v2-squashfs-factory.bin |
just made PR #827. Is the node "cottbus-lausi36-2" still a CPE-v3 on ath79? |
Was a tplink_cpe210-v2 (1xLAN). Trying build branch cpe210 fails due gcc10 support check fails (arch linux): https://patchwork.ozlabs.org/project/openwrt/patch/[email protected]/ workaround: trying fresh https://github.com/freifunk-berlin/firmware/commits/openwrt-19.07.3 branch and #827 > seems to compile build stops at an other error related to gcc10, workaround: https://git.openwrt.org/?p=openwrt/openwrt.git;a=commit;h=1bbc1aa884902fd05cc579b53d68b2ba0b18683f Fixes the following build error with GCC 10: |
Yeah, I see your build issues are only with our outdated master-branch. The "mksquashfs"-bug mentioned is already backported to 19.07 (openwrt/openwrt@059db41). - Probably you want to give a build of https://github.com/SvenRoederer/freifunk-berlin-firmware/commits/nightly/upstream-1907 a chance. I can also share you the prebuild imagebuilder, but confirming build on arch linux would be also great. |
canceled building https://github.com/freifunk-berlin/firmware/commits/openwrt-19.07.3 did a build for: https://github.com/SvenRoederer/freifunk-berlin-firmware/commits/nightly/upstream-1907 on Linux 5.7.9-1-MANJARO (x86_64) openwrt gcc 10: internal compiler error: Segmentation fault > 1times in early build process > seems doesnt like multithreaded building (make -j4) make -j1 > build continues > afterwards multithreaded again (make -j4) > build finished :) build: Powered by LuCI branch (git-19.082.34170-bf97eae) / Freifunk Berlin SAm0815-master 0017a2f
bugs:
|
Based on the version-string of your firmware you built my repos master branch (OpenWrt-19.07 of 21st June) and not on "nightly/upstream-1907" branch (recent OpenWrt-19.07). I expect this to be the cause of the build-failure you have seen.
That's great to hear, so its proofed that CPE210v2 is working without wifi-issues. As per https://git.openwrt.org/?p=openwrt/openwrt.git;a=commit;h=141698ce8fbc52d253c9207ef46b3630a0870567 the v2 and v3 are identical I assume v3 is also working.
--> see issue #831
--> are these bugs? :-)
--> this references to issue #769? Edit: Thanks for the testing |
@Mitzlaff @everloop2 shall we close this? |
Close, wifi has been fixed, other bugs > other issues. yep no bugs ;)
|
Bug report
What is the problem?
WLAN does not work.
When using LAN connection I see this:
What is the expected behaviour?
SSID showing up when scanning for configured SSID name (freifunk-backyard)
Firmware Version:
freifunk-berlin-dev-daily-6431131-tplink_cpe210-v3-sysupgrade.bin
Site Configuration:
default, just using Luci
Custom patches:
none
The text was updated successfully, but these errors were encountered: