Guess i am running another build tonight then. Thanks.
With my mt7615 router, I've noticed iPhone 6S+ traffic getting stuck. Flipping wifi on/off fixes it. I haven't been following the thread closely, but is this the issue?
1 Like
Could be anything including iOS version. Broadcom's drivers for iOS are totally closed.
Yes, it is.
Hrm that implies I'm having the issue as well. I'll spin up a build soon to see if master fixes that issue.
The problem is triggered by tearing down and re-establishing a-mpdu sessions. This depends on traffic patterns and client behavior. The first agg session is likely to work, since it will have a low seq number and the driver's seqno will be mostly in sync with mac80211's seqno. Subsequent sessions will trigger this more severely.
5 Likes
Although I am not a good candidate (does not use Apple products) I updated mac80211 and cfg8011 on WRT32X and WRT1900ACS, both are working correctly.
1 Like
This patch is a success for the wireless cutouts.
I had to trigger the wireless cutouts as a part of bisecting OpenWrt commits with @adworacz and then bisecting mac80211 commits with @cotequeiroz.
After triggering hundreds of wireless cutouts, near the end of it all, I had developed such a good technique and repetition that I was consistently able to trigger a wireless cutout within 45 seconds to 90 seconds.
I am still in some sort of shock-like surprise about this patch, but I have not been able to trigger a single wireless cutout since this patch came out. And I have been using my iPhone XR (trigger device) non-stop since then.
Congratulations everyone. I think many of us questioned whether or not we would ever get to the root cause of this issue. Seriously.
The perseverance of @adworacz, @cotequeiroz, @arinc9, and the community here in general all made this happen. Many great tips and suggestions along the way from @slh and @hnyman as well.
And special thanks to @nbd for patching master and 21.02 branches and upstreaming the patch promptly.
Cheers!
14 Likes
This truly is awesome, so much work you guys put into this! I was already researching replacement devices, but I really like these particular devices and now have 4 of them. So this already in the nightly snapshot build of 21.02 for these devices may be in 21.02.2 when it comes out ?? or can I install 21.02.1 and update via the opkg method you mention in an earlier comment??
1 Like
Yes.
If you want it right now into 21.02, you could install the semi-hidden 21.02 snapshot firmware (current 21.02 HEAD. Later than 21.02.1 but not yet officially 21.02.2)
https://downloads.openwrt.org/releases/21.02-SNAPSHOT/targets/mvebu/cortexa9/
Yes.
1 Like
Thank you for the awesome work in finding the issue, made a build last night r18188 Snapshot and the wireless now is fast at connecting on devices.
Although i wasn't getting disconnects like others it however was slow at times to make the connection.
1 Like
Awesome! I had downloaded the SNAPSHOT version from the device page (https://firmware-selector.openwrt.org/?version=SNAPSHOT&target=mvebu%2Fcortexa9&id=linksys_wrt32x), which may or may not contain the patch??
If the image is compiled November 25 or newer, it will contain the updated kmods. You can even open the sysupgrade image with a zip program and check date of the kmods.
2 Likes
That is the master snapshot, not 21.02.
The 21.02 branch has its own daily snapshots, which are not available in the firmware selectors. Those are not linked to the download link pages, but can be found if you know the address (which is visible in the 21.02 branch sources).
But yeah, both master and 21.02 builds since 25 Nov do include the patch.
1 Like
Getting back into town, happy to see the end result being a fix merged into master and 21.02!!
Big thanks to @WildByDesign for the reliable testing, which was incredibly helpful for quickly narrowing down commits. Shout out to @arinc9 for starting this thread and getting the conversation going. And special thanks to @cotequeiroz for doing the final mac80211 bisect + patch list build out + final patch.
And of course thanks to @nbd for merging this all in.
I'm going to update all of my notes + posts to point to the true fix.
Nice job, team.
9 Likes
I also put together a kernel with a new patch for McDebian and I also want to thank all the active contributors for solving this problem.
Thank you very much!
1 Like
Can anybody confirm that this IS or should be an issue starting with version 19.07.8 on a WRT3200ACM? I do see the following at the top.
I had no issues with OpenWrt 19.07.7.
Looks like some folks have had luck reverting to 19.07.8 while others have not?
Context:
I upgraded from a much older version (unfortunately unknown) yesterday to 21.02.1 without knowing about this issue and hit this very problem. Because this was disruptive at home I reverted to 19.07.8 where I am pretty sure I was seeing the same issue as well (I initially missed the version where this may have started). I have temporarily reverted to stock firmware for now, I will be back for 21.02.2 though. Wanted to confirm that I am not missing something else that might be going on.
I also wanted to also say thank you to everybody on this thread.
This issue should not be present I 19.07.8 as that uses mac80211 based on kernel 4.19, where the culprit commit is not present.
The year 2020 commit has not been backported into https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/log/net/mac80211/agg-tx.c?h=v4.19.218
Note that there are various problems with mwlwifi, partially depending on the WiFi chip and its firmware. This was not the only bug there...
Note that there are various problems with mwlwifi, partially depending on the WiFi chip and its firmware. This was not the only bug there...
OK thank you, I may very well have found/experienced something else here then. Its only been about 10h or so, but stock is working for me so something was going on here after my OpenWRT < 19.07.8 upgrade to 21.02.1 after a year or more of stability. Is there something else generally tracking the other issues specifically with WRT3200ACM or related devices? I am going to try again at 21.02.2 but unsure if I should be tracking anything else ATM. (and sorry, little light on specific / debug info)
I can't believe that this is finally going to get fixed. I can't tell you how many times I thought I fixed this by using a particular setting only to have another cut out later that week. Thank you for starting this thread and getting the ball rolling on this.
2 Likes