Posted: Sun Jun 13, 2021 14:56 Post subject: What was the breaking build for iOS/Apple?
So I'm still using build 44048 since I need to support iOS/Apple devices. After research, I've seen that some build after 44048 caused instability in iOS devices.
I was wondering, does anyone know the exact build that caused the iOS stability issue? I was thinking of trying to investigate the issue myself, and see if maybe I could find where the issue occurred. _________________ WRT3200ACM
I don't know what exact change but the very next build, 44085, had IOS and MACOS issues (maybe other devices) where wifi remained connected and strong but all traffic stopped flowing. Stopping/starting wifi on the device immediately but temporarily fixed the problem.
Last week I installed r46854 on my 3200ACM after thoroughly reseting it, and it has been much better than any release since 44048. I've had a handful of IOS wifi hangs and my first macos wifi hang today. See my notes in the r46854 thread.
Last week I installed r46854 on my 3200ACM after thoroughly reseting it, and it has been much better than any release since 44048. I've had a handful of IOS wifi hangs and my first macos wifi hang today. See my notes in the r46854 thread.
When I saw your success I immediately tried it myself. Reset nvram multiple times, reloaded multiple times (since it switches partitions each time), tried the reset button multiple times. I've seen no difference at all in the IOS hangups...
Sorry to hear that. The wifi on my 3200ACM isn't used as heavily or by the variety of devices that my 1900ACS was. I keep the 3200ACM around to serve wifi to a few rooms upstairs, including my office, and as a gateway for wireguard. The 1900ACS that only worked on 44048 has been replaced by by a Netgear r7800 running r46854 - it's the main wifi AP in the house and has 15+ devices connected at any one time. Runs without a problem. Someday I'll reload the 1900ACS and see if I can get it to work.
The only things I can think of as to why the 3200ACM is working fairly well with 46854 are 1) resetting the crap out of it and manually typing everything in, or 2) some magic combo of wifi settings (see my screen shots in the 46854 thread), or 3) the 3200ACM is not being hit as hard or by the right combo of devices to knock it out.
Regarding #3, I was able to get my 1900ACS to hang with a single IOS or Macos client attached so it's not like it took a bunch of weird devices to knock it out. Also, the 3200ACM is still in the same spot in my house as when I tried several releases on it last Fall and wifi hung up constantly, so its wifi landscape hasn't changed that much but the 3200ACM is much more stable. I'm stumped.
Posted: Tue Jun 15, 2021 8:23 Post subject: Re: What was the breaking build for iOS/Apple?
whynotwrt wrote:
So I'm still using build 44048 since I need to support iOS/Apple devices. After research, I've seen that some build after 44048 caused instability in iOS devices.
I was wondering, does anyone know the exact build that caused the iOS stability issue? I was thinking of trying to investigate the issue myself, and see if maybe I could find where the issue occurred.
The problem started when Apple decided to update their firmware and ever since all the Apple problems started. I can't pinpoint the exact revision but it's one of the latest ones
Not sure about that. Others have reported the same problem with non-apple devices. This issue started with r44085 last August - well before IOS 14 and MacOS 11. Falling back to 44048 fixes the problem regardless of the IOS/MacOS version your device is running.
Any version newer than 44048 on our wrt1200acv2 will have problems with Windows 10 clients after a few hours. It stays connected, but the throughput falls to nearly nothing. Restarting the connection will temporarily fix the problem, but it always comes back. _________________ Netgear R9000
Netgear R7800 (spare)
Linksys WRT1200ac v2 (spare)
Posted: Sat Jun 19, 2021 0:03 Post subject: Re: What was the breaking build for iOS/Apple?
jifffy wrote:
whynotwrt wrote:
So I'm still using build 44048 since I need to support iOS/Apple devices. After research, I've seen that some build after 44048 caused instability in iOS devices.
I was wondering, does anyone know the exact build that caused the iOS stability issue? I was thinking of trying to investigate the issue myself, and see if maybe I could find where the issue occurred.
The problem started when Apple decided to update their firmware and ever since all the Apple problems started. I can't pinpoint the exact revision but it's one of the latest ones
I want to say again that I don't think it's apple. I use an LG G7 ThinQ Android device that I connect to my router via wifi from bed to RealVNC into my PC in the room. I am currently still on 44048 and haven't tried any other firmware later than this because I had, for the first time ever, problems connecting to my PC with my android device, so I went back to 44048. I would be willing to try a newer firmware if it's been shown to be as stable as 44048, but until then, nope! Nothing personal!
Posted: Sat Jun 19, 2021 2:58 Post subject: Re: What was the breaking build for iOS/Apple?
PainbowDDWRT wrote:
jifffy wrote:
whynotwrt wrote:
So I'm still using build 44048 since I need to support iOS/Apple devices. After research, I've seen that some build after 44048 caused instability in iOS devices.
I was wondering, does anyone know the exact build that caused the iOS stability issue? I was thinking of trying to investigate the issue myself, and see if maybe I could find where the issue occurred.
The problem started when Apple decided to update their firmware and ever since all the Apple problems started. I can't pinpoint the exact revision but it's one of the latest ones
I want to say again that I don't think it's apple. I use an LG G7 ThinQ Android device that I connect to my router via wifi from bed to RealVNC into my PC in the room. I am currently still on 44048 and haven't tried any other firmware later than this because I had, for the first time ever, problems connecting to my PC with my android device, so I went back to 44048. I would be willing to try a newer firmware if it's been shown to be as stable as 44048, but until then, nope! Nothing personal!
Apple changed "something" and Linksys as manufacturer of your device ceased support long ago.
Until they urge someone over at NXP to solve this, no fast solution to await (the Marvell WiFi hardware branch was sold long time ago).
Over the time unsolved problems, resulting in this "Wifi is somehow garbage" situation piled up to the current state, for some users with Marvell based hardware.
I don't understand why people keep saying that it has nothing to do with dd-wrt and it's just the mwlwifi driver. This doesn't make any sense to me. The most recent update to https://github.com/kaloz/mwlwifi is a single bin that is 1.5 yr old but most files are several years old. The last good release 44048, is not even a year old. Is it the case that the driver was updated after 44048, using the firmware that was months old at that point? Seems unlikely, but even if that is the case, why can't we just get a build with the older firmware? It seems more likely that a bug was introduced into dd-wrt after 44048.
Joined: 08 May 2018 Posts: 14126 Location: Texas, USA
Posted: Fri Aug 13, 2021 13:47 Post subject:
Consider following browser recommendations for accessing webUI, disabling pre-fetching, and even adjusting the
auto-refresh interval in the webUI. NOTE: Check setting "tcp_tw_recycle" (in "Administration->Sysctl") and if set
to 1, set to 0; "save" and reboot as it appears to solve issues with Apple products. Read the first four links.