You are not logged in.
Pages: 1
To begin with my internet had been originally around 500-600 mbps —now I get a measly 13 mbps. I use a wireless connection using Network Manager and wpa as the back-end. The specific chip is RTL8922AE. I am unsure as to what could of caused this change as I have went to try multiple solutions. I had tried disabling power-save mode via /etc/modprobe.d/rtw.conf as well as configurations to network manager. I am using the 5 GHZ channel as well. To further note, I run a dual boot system with Windows and it has no issues with WiFi speeds, and this dual boot has been active for half a year without issue. Further on, I have downloaded RTW89 drives from another similar post to see if that would help fix the issue: https://github.com/morrownr/rtw89/. I also tried using linux-lts to see if it had been related, but it does not seem to be. Afterwards I saw no improvement. If possible could anyone help as I am entirely lost after trying numerous fixes to no avail. Also, I apologize as I am unsure what logs to post along with this as nothing seems unusual.
Last edited by computer_spider (2026-04-09 23:15:29)
Offline
I run a dual boot system with Windows
3rd link below. Mandatory.
Disable it (it's NOT the BIOS setting!) and reboot windows and linux twice for voodo reasons.
I had tried disabling power-save mode via /etc/modprobe.d/rtw.conf
rtw89_pci.disable_aspm_l1=y rtw89_pci.disable_aspm_l1ss=y rtw89_core.disable_ps_mode=yhttps://wiki.archlinux.org/title/Kernel_parameters
But also
lsmod | grep rtwto make sure this isn't handled by rtw88
Ultimately please post your complete system journal for the boot:
sudo journalctl -b | nc termbin.com 9999Behavior sounds like you're on 802.11a instead of 802.11n (or ac/x)
Offline
Thank you again for helping!
3rd link below. Mandatory. Disable it (it's NOT the BIOS setting!) and reboot windows and linux twice for voodo reasons.
Unfortunately, this was something I had already tried. I checked once again, but fast boot was still disabled. I also have hibernation turned off as that caused other issues at some point.
But also
lsmod | grep rtwto make sure this isn't handled by rtw88
It does not seem to be the case. Here is the output as confirmation:
rtw89_8922ae_git 12288 0
rtw89_8922a_git 77824 1 rtw89_8922ae_git
rtw89_pci_git 126976 1 rtw89_8922ae_git
rtw89_core_git 1298432 2 rtw89_pci_git,rtw89_8922a_git
mac80211 1728512 2 rtw89_pci_git,rtw89_core_git
cfg80211 1470464 3 rtw89_core_git,rtw89_8922a_git,mac80211
rfkill 45056 8 rtw89_core_git,bluetooth,cfg80211Ultimately please post your complete system journal for the boot:
sudo journalctl -b | nc termbin.com 9999
As per requested; however, I was unable to use termbin. I doubt it relates to this issue but nevertheless I mention it. I even tested on my laptop and seem to be unable to upload to termbin. I hope pastebin is acceptable.
Behavior sounds like you're on 802.11a instead of 802.11n (or ac/x)
I am unsure how to test this or if this is related, but seemingly this issue is not isolated to my computer. I thought this issue was bizarre due to no updates being made on my computer when this issue began recently, last few days. It seems others are experiencing the same issue, that is Xfinity (Comcast). Unfortunately, I do not have ease of access to manipulate Xfinity settings. To further add, my laptop, that is also Linux, has slow WiFi as well now, despite being able to reach higher download speeds on other networks. I am unsure if this issue can be resolved via these forums. I apologize if this is now the inappropriate place to post this issue.
The aforementioned thread complaining about Xfinity wifi issues:
Offline
Please disable systemd-networkd
But
Apr 09 11:08:35 silk kernel: wlan0: required MCSes not supported, disabling HTWelcome to the late 90ies there on 802.11a
You're the second one this week and the exact band of "freq=5220 MHz" … Canada or Trumpistan?
Apr 09 11:08:26 silk kernel: cfg80211: failed to load regulatory.db
…
Apr 09 11:08:32 silk wpa_supplicant[1043]: wlan0: CTRL-EVENT-REGDOM-CHANGE init=BEACON_HINT type=UNKNOWN
Apr 09 11:08:32 silk wpa_supplicant[1043]: wlan0: CTRL-EVENT-REGDOM-BEACON-HINT before freq=5200 max_tx_power=2000 no_ir=1
Apr 09 11:08:32 silk wpa_supplicant[1043]: wlan0: CTRL-EVENT-REGDOM-BEACON-HINT after freq=5200 max_tx_power=2000https://wiki.archlinux.org/title/Networ … ory_domain
Completely unrelated sidebar
Apr 09 11:08:37 silk sddm-helper[1052]: Starting Wayland user session: "/usr/share/sddm/scripts/wayland-session" "uwsm start -e -D Hyprland hyprland.desktop"uwsm is heavily discouraged by upstream hyprland
Offline
Thank you, it seems the issue has been resolved. I am hitting ~700 mbps again. Which is faster than the windows counter part.
As suggested I disabled systemd-network, removed uwsm, and downloaded wireless-regdb and uncommented the US line for my case, or Trumpistan as I just learned from this post. ![]()
Offline
I'm also getting the same issue right now, not sure when it started but must be recent.
[ 8.714138] wlan0: required MCSes not supported, disabling HT
[ 8.719361] wlan0: authenticate with 72:13:01:87:79:82 (local address=60:f2:62:f8:50:02)
[ 8.720445] wlan0: send auth to 72:13:01:87:79:82 (try 1/3)
[ 8.755643] wlan0: authenticated
[ 8.756790] wlan0: associate with 72:13:01:87:79:82 (try 1/3)
[ 8.769045] wlan0: RX AssocResp from 72:13:01:87:79:82 (capab=0x1011 status=0 aid=19)
[ 8.773648] wlan0: associatedAlso on 5220MHz in Canada. Something's up with Canadian regulatory it seems like
$ iw dev wlan0 info
Interface wlan0
ifindex 3
wdev 0x1
addr 60:f2:62:f8:50:02
ssid
type managed
wiphy 0
channel 44 (5220 MHz), width: 20 MHz (no HT), center1: 5220 MHz
txpower 22.00 dBm
multicast TXQ:
qsz-byt qsz-pkt flows drops marks overlmt hashcol tx-bytes tx-packets
0 0 0 0 0 0 0 0 0I downloaded wireless-regdb and enabed CA, but it doesn't seem to fix it. My wifi card shows up as
00:14.3 Network controller: Intel Corporation Comet Lake PCH-LP CNVi WiFithough so something might be different?
Offline
Looks like many people who's using Xfinity/Rogers is having the same issue
https://linuxcommunity.io/t/intel-ax210 … os/6708/30
https://www.reddit.com/r/Comcast_Xfinit … computers/
https://www.reddit.com/r/Comcast_Xfinit … _breaking/
https://www.reddit.com/r/Rogers/comment … t_54_mbps/
Offline
x-ref, https://bbs.archlinux.org/viewtopic.php … 4#p2293764
Same issue w/ some older live distro?
Offline
Hi Computer_Spider & Rio6, I have Videotron Helix modem here in Quebec with same issue. I got myself a dedicated router which completely solved the issue, went from 20mbps to 1100mbps.
Last edited by Grafdude (2026-04-10 22:30:34)
Offline
Same issue w/ some older live distro?
Good idea. I tried with a live Artix usb circa 2023 and the bandwidth shows 80MHz there. Do you know which packages would be relevant here if I want to try roll back?
Last edited by Rio6 (2026-04-10 22:27:16)
Offline
Most likely the kernel - there're a bunch of older https://aur.archlinux.org/packages?O=0&K=linux-lts ones in the AUR
You can try one from the ALA, but there's no guarantee it will still boot if you're going back in time too much, so install an LTS kernel next to the main one, don't downgrade your only kernel to figure that you can no longer boot the system.
Offline
Ok, I don't have much time to rebuild kernels right now. But if it is indeed the kernel I can probably git bisect it later
Offline
OK I found this repo https://github.com/WoodyWoodster/mac80211-mcs-patch that removes the MCS check from mac80211 module, applying the patch get me back to 80MHz channels.
$ iw dev wlan0 info
Interface wlan0
ifindex 3
wdev 0x1
addr 60:f2:62:f8:50:02
ssid
type managed
wiphy 0
channel 157 (5785 MHz), width: 80 MHz, center1: 5775 MHz
txpower 22.00 dBm
multicast TXQ:
qsz-byt qsz-pkt flows drops marks overlmt hashcol tx-bytes tx-packets
0 0 0 0 0 0 0 0 0I also found the patch where the check was introduced: https://lore.kernel.org/all/20250205110 … @changeid/. Interestingly there exists this "strict flag" that disables the MCS check.
/*
* Many APs are incorrectly advertising an all-zero value here,
* which really means MCS 0-7 are required for 1-8 streams, but
* they don't really mean it that way.
* Some other APs are incorrectly advertising 3 spatial streams
* with MCS 0-7 are required, but don't really mean it that way
* and we'll connect only with HT, rather than even HE.
* As a result, unfortunately the VHT basic MCS/NSS set cannot
* be used at all, so check it only in strict mode.
*/
if (!ieee80211_hw_check(&sdata->local->hw, STRICT))
return true;But that code path doesn't seem like it's being taken, even though according to debugfs the strict flag isn't set.
For now the patch on mac80211 works, but I do want to find a proper way to fix this.
Offline
Message is from https://elixir.bootlin.com/linux/v6.19. … me.c#L1168 after failing https://elixir.bootlin.com/linux/v6.19. … lme.c#L347 which isn't gated by the strict setting - there's also no comment about quirky APs ![]()
Either the implementation is wrong or (rather) some buggy APs are in the wild (mostly the wilderness of canada
)
Offline
OK I found this repo https://github.com/WoodyWoodster/mac80211-mcs-patch that removes the MCS check from mac80211 module, applying the patch get me back to 80MHz channels.
For now the patch on mac80211 works, but I do want to find a proper way to fix this.
I have also applied this patch and got the data link speed up to around 1Gbps on an 80Mhz wide channel, but it does not fully solve the problem for me at all. I get higher wifi transfer speeds, but they are still only up from 15Mbps to around 50Mbps (when they used to be around 300Mbps).
Have your wifi transfer speeds improved to what they were before this fiasco?
(Regarding the fiasco, it seems like a lot of people are having this problem with the Technicolor modems from Comcast. Apparently it is something in new firmware that is taken up by various ISPs that use them including Rogers, which is where I experienced the problem. I am eager to hear any more information about this that anyone has. Probably though, I am going to buy another modem and see if it can be used in any of Bridge mode (preferred, but will TV still work), AP mode (but the locked down Technicolor is pretty awful and it would be nice to use a nice router), or double Nat mode (with nothing but TV handled but Technicolor, and everything else on the other router which is in the DMZ))
Offline
Have your wifi transfer speeds improved to what they were before this fiasco?
For me yes, I am getting 150-200Mbps after the patch while without the patch it's capped at 20Mbps.
Offline
@BensonBear make sure to not have colliding network services enabled, can you test the WIFI in the LAN using eg. iperf to rule out WAN-related restrictions?
Offline
@BensonBear make sure to not have colliding network services enabled, can you test the WIFI in the LAN using eg. iperf to rule out WAN-related restrictions?
Thanks for your reply.
Yes, we are talking about wifi speed, so I have tried to test that as directly as possible.
One of the tests is with iperf3 from a single wired connection to a single wifi connection, with no other devices on the network. Then I have tested also with iperf3 between two wifi connections.
It's all very much slower than it used to be, even after applying the patch.
Offline
Do you get decent throughput if you eg. use your phone as hotspot (or any other router than the concerned one)?
What does the interface/link status look like?
https://wiki.archlinux.org/title/Networ … _interface
50Mbps, not 50MB/s (but 6¼)?
50Mbps would still be in the 802.11a realm, no HT/VHT
What kind of NIC is this (realtek, intel, …)?
You probably should open a new thread about your situation (or report your first post to be split off by a mod)
Offline
Do you get decent throughput if you eg. use your phone as hotspot (or any other router than the concerned one)?
I thought I mentioned that but apparently not. Yes, if I take my notebooks (any of three) out into our building lobby the internet speed test from there measures generally over 200Mbps. (So the Wi-Fi transfer rate would also be high)
What does the interface/link status look like?
Various for different cards:
Before applying the patch that ignored the (supposedly unreasonable) BCS requirements, the link was 802.11g at best with 54Mps tx rate, and no indication of any HT/VHT/HE being used, and a 20Mhz channel only.
After the patch, I got a much higher tx rate with the Realtek card, and indication the HE (802.11ax) was being used, and a 80Mh width, but the actual Wi-Fi speeds were much lower. Going out to the lobby, connect to building's network, then much faster.
In the case of the Intel cards, I did not apply the patch because it was not indicated (the journal indicates that mac80211 driver has denied any HT or higher only in the Realtek case) Still, a very low tx rate, low width (20Mhz), no indication of HT/VHT/HE being used. Out in the lobby on the other network, indication of VHT (802.11ac, the limit of this card), high tx rate, 80Mhz width, and fast Wi-Fi speeds.
50Mbps, not 50MB/s (but 6¼)?
Yes, 50Mbps.
50Mbps would still be in the 802.11a realm, no HT/VHT
Correct (or 802.11g)
What kind of NIC is this (realtek, intel, …)?
Two older Intel 7265 Wi-Fi cards, and a newer Realtek RTL8852BE Wi-Fi card.
You probably should open a new thread about your situation (or report your first post to be split off by a mod)
Well, it is not an arch problem. I just posted here because this specific issue was being discussed so the specific people here were already up on the issue. I have posted about it in many other forums including Rogers on reddit , where the XB7 (i.e. Technicolor ) involved is being deployed. So instead of new thread will just bid you adieu if no one else in this one has any more insights.
ETA: Oops! Above I said that the mac80211 patch was "not indicated" for older NICs. I don't know where I got that idea, and I thought I had checked in the journal, but in fact even with the 7265s HT and above was being disabled by that module. So I will have to patch the other versions of mac80211 as well when I get a chance and hopefully that will improve those systems. (Still probably won't reach where they were before on the Rogers XB7)
Last edited by BensonBear (2026-04-13 12:49:06)
Offline
Pages: 1