FW16 - Weak Wifi on Fedora?

Framework Laptop 16 Linux
1

Is there any reason I’d be getting a weaker wifi signal on Fedora than with Windows on the same hardware? It doesn’t appear to be affecting performance. Is it just a difference in opinions between the two operating systems as to what constitutes a strong signal? I can see my neighbors’ wifi seemingly equally poorly (names redacted to protect the innocent), but I can’t get 3 bars of my own on Fedora even if I sit 5 feet from the AP.

Screenshots taken back to back in the same exact location, one room over from the wifi AP:

wifis
wifis1203×544 300 KB

  • Dual booting Fedora 39 and Windows 10
  • Installed Windows 11 driver pack, then Windows 10 AMD chipset drivers
  • Fedora drivers presumably in kernel
  • output from modinfo mt7921e:
filename:       /lib/modules/6.7.7-200.fc39.x86_64/kernel/drivers/net/wireless/mediatek/mt76/mt7921/mt7921e.ko.xz
license:        Dual BSD/GPL
author:         Lorenzo Bianconi <lorenzo@kernel.org>
author:         Sean Wang <sean.wang@mediatek.com>
firmware:       mediatek/WIFI_MT7922_patch_mcu_1_1_hdr.bin
firmware:       mediatek/WIFI_RAM_CODE_MT7922_1.bin
firmware:       mediatek/WIFI_MT7961_patch_mcu_1_2_hdr.bin
firmware:       mediatek/WIFI_RAM_CODE_MT7961_1.bin
rhelversion:    9.99
alias:          pci:v000014C3d00000616sv*sd*bc*sc*i*
alias:          pci:v000014C3d00000608sv*sd*bc*sc*i*
alias:          pci:v00000B48d00007922sv*sd*bc*sc*i*
alias:          pci:v000014C3d00007922sv*sd*bc*sc*i*
alias:          pci:v000014C3d00007961sv*sd*bc*sc*i*
depends:        mt76-connac-lib,mt7921-common,mt792x-lib,mt76
retpoline:      Y
intree:         Y
name:           mt7921e
vermagic:       6.7.7-200.fc39.x86_64 SMP preempt mod_unload 
sig_id:         PKCS#7
signer:         Fedora kernel signing key
sig_key:        some hex
sig_hashalgo:   sha256
signature:      a lot of hex
parm:           disable_aspm:disable PCI ASPM support (bool)
2

The WiFi display “bars” unfortunately aren’t an accurate or comparable metric for signal strength. Can you compare the actual values as reported in dB? That would help.

3

Windows reports a dB of around -66 (which seems low given the full bars), but I can’t figure out for the life of me how to get the dB values on Fedora. In attempting, I did find “nmcli dev wifi” which showed something interesting:

Screenshot_20240309_065809
Screenshot_20240309_0658091121×81 33.8 KB

It looks like it’s picking the faster, but weaker 5GHz band. But logging into the AP, it doesn’t show my MAC in the connection list, though my router sees my FW. In fact, I doesn’t look like I even have 5GHz enabled! What the heck is this network?!

Update was available for the AP, and after applying, here is my connection list:

Screenshot_20240309_071904
Screenshot_20240309_0719041112×86 30.7 KB

The F2:9F device is gone now, presumably the 5GHz radio, though when I enable the 5GHz band F2:9F shows up along with a different MAC (identical except it begins with F0) which carries the actuall 5GHz SSID, something default based on the MAC.

Screenshot_20240309_073709
Screenshot_20240309_0737091153×137 55.8 KB

I think both I, Fedora, and my AP are confused as to what’s going on, but it seems to be better now that I’ve updated the AP and made sure 5GHz is disabled. Maybe someone more familiar with how wifi works can make sense of this, but I’m guessing it’s a combination of user error and the wifi AP needing a reboot.

Side note: does anyone know how to get dB signal strength from Fedora?

4

iw dev will list your wireless devices, in my case the interface is called wlp4s0.

Then to list the link information, which includes the RSSI, run iw dev wlp4s0 link making sure the device name is correct.

Example output from my machine right now:

Connected to <REDACTED> (on wlp4s0)
	SSID: <REDACTED>
	freq: 5200
	RX: 23131826 bytes (20757 packets)
	TX: 1011587 bytes (6147 packets)
	signal: -66 dBm
	rx bitrate: 866.7 MBit/s VHT-MCS 9 80MHz short GI VHT-NSS 2
	tx bitrate: 780.0 MBit/s VHT-MCS 9 80MHz VHT-NSS 2

	bss flags:	short-slot-time
	dtim period:	1
	beacon int:	100
1 Like
5

There it is, thanks! Interestingly, even though I’m definitely on the same (only) wifi band and I just read -61dB in Fedora, I still only get 1 bar compared to Windows which was showing full strength at -66dB, which as I understand it is actually a weaker signal. It appears that the two OS’s do really have differing opinions as to what is a strong signal.

1 Like
6

Other than the difference in UI, are you noticing a major difference in performance?

7

No, I’m fairly convinced it really is just a difference in UI. An odd difference, odd on both sides with Windows being overoptimistic and Fedora saying 5 feet away is still not good enough, but whatever, it still works just as good. Been doing a lot of setup on both OS’s, and all the downloading seemed pretty normal and smooth.

8

Not sure if this applies to Fedora as well (considering it’s supposed to be an officially supported distro), but for Arch, there is a comment about needing to set the regulatory domain, otherwise you’ll be limited. It’s possible that the Windows Drivers are somehow setting that, but it’s not getting set in Linux.

I’m not certain that this would be the “issue” but it seems like it could be a possibility considering it only seems to be the case with the 5GHz networks and the regulatory domain impacts the 5GHz functionality more than 2.4GHz.

Try running iw reg get and see if it specifies a country or if it reports 00.

1 Like
9

I am indeed in the US, under the jurisdiction of the FCC:

global
country US: DFS-FCC