Hello
I have been having a strange issue with wpa_supplicant and scanning for ssids.
I have a linux box running mainline linux version 5.15.23, compiled myself from a kernel.org tarball, no patches applied, but with a custom config including needed modules for the wifi card (iwlwifi and iwlmvm). It has an intel ax210 wifi card installed and has firmware for said wifi card (/lib/firmware/iwlwifi-ty-a0-gf-a0-66.ucode and /lib/firmware/iwlwifi-ty-a0-gf-a0.pnvm). The wifi card has both antennae connected. wpa_supplicant is version 2.10, also compiled from unpatched official source tarball.
At first glance, everything looks like it is working correctly, I can scan, get scan results, and connect to and communicate with an AP. I can check dmesg and confirm that the correct firmware is loaded successfully.
However, the problem is that the scan results look rather short, even though I know I am in a very noisy environment. It can detect an APs that are a good distance away, so its not a signal strength issue, but it only ever finds about 65 BSSes with each scan, even though Windows machines with other wifi modules can detect hundreds of ssids in the same environment (which excludes duplicate and hidden ssids, which wpa_supplicant reports as seperate BSSes). I have searched online for others complaining of the same issue, but I seem to be the only one.
I have wpa_supplicant running with pretty much a blank config. only country=US, autoscan=0, and ignore_old_scan_res=1.
bss_max_count is set to its default value of 200, which is a limit that I expected to need to increase, but 65 is nowhere near 200... I tried raising bss_max_count to 1024, but the number of results returned didnt change at all.
The results I get do look unusual sometimes, on some scans I only get 5ghz results, even though I know every AP has both 2.4 and 5 radios, other scans I get a mix of both. The list seems to usually be the APs with the highest signal, but not always.
I am afraid I am hitting some limit somewhere that everybody knows about except me. Does anybody know of something I am missing? Is it possible that wpa_supplicant has some other undocumented limit to the number of APs it can scan? Does the ax210 have an internal limit to the number of APs it can scan? Does anybody know of a way to improve wpa_supplicant's scan results? Would newer firmware for the ax210 help (I have a file called iwlwifi-ty-a0-gf-a0-68.ucode, but I think I need a newer .pnvm file in order to load that microcode)?
Sorry for the long email and thanks for any advice you guys have.
Philip Rushik