Device Info
- Device: Tecno Pova Neo 2 (LG6n), MediaTek MT6768
- Android version: 12 (build SP1A.210812.016 / 230510V1293)
- Root: Magisk (Kitsune fork), ReZygisk, bootloader unlocked,
secure: no
- Interfaces available: both
wlan0 and wlan1 exist under /sys/class/net/ (confirmed via ls)
Problem
After running virtualap start, the script completes with no errors and reports the AP as fully active:
Auto upstream: table wlan0 (wlan0)
Creating virtual AP interface 'ap0' on wlan0...
Configuring gateway address 192.168.42.1...
Installing routes and policy rules...
Network configuration completed.
Generating service configs (band 5GHz, channel 36)...
Starting hostapd (5GHz ch 36, 80MHz)...
hostapd running (PID 7609, 80MHz)
Starting dnsmasq...
dnsmasq running (PID 7624)
VirtualAP is ACTIVE!
SSID : PAK-Link 5G (5GHz, channel 36, 80MHz)
Gateway : 192.168.42.1 (static - never changes)
Upstream: wlan0 (table wlan0)
ip link show ap0 also confirms the interface is fully up:
48: ap0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
link/ether fa:c7:45:9c:8a:4d brd ff:ff:ff:ff:ff:ff
However, the SSID never appears in a WiFi scan on any test device — tried:
- Windows laptop (multiple driver resets, airplane mode toggle, no result)
- A second Android phone
- Both report zero visible network with this SSID, at any band/channel
What I've tried (all failed to fix visibility)
- Band changes: 2.4GHz (ch 6, 20MHz) and 5GHz (ch 36 and ch 157, 80MHz) — no difference
- Regulatory domain:
iw reg get showed country 00: DFS-UNSET with PASSIVE-SCAN flagged on all 5GHz ranges. Manually patched the script's hardcoded country_code=US to country_code=AE (my actual region) in write_hostapd_conf(). Confirmed the generated hostapd.conf picks up country_code=AE correctly on next start — no change in visibility.
- Stock/native Android hotspot (not VirtualAP) was also tested as a baseline — also invisible to all test devices, at any band. This suggests the issue may not be VirtualAP-specific, but I'm filing here since VirtualAP is the tool I'm trying to get working, and the identical failure mode may still be relevant to how
ap0 is being created here (e.g., wrong physical interface being targeted).
- Confirmed
wlan1 exists as a separate interface alongside wlan0 (ls /sys/class/net/) — not yet tried forcing VirtualAP to use it via -w wlan1, wanted to raise this first in case there's a known reason it wouldn't apply.
Question / request
- Is there a known MT6768 same-channel/same-radio concurrency limitation where
hostapd believes it's transmitting (PID alive, interface UP) but the firmware silently drops AP-mode beacon frames when bound to wlan0 (the STA-mode interface)?
- Would forcing the AP interface onto
wlan1 (if it's a distinct hardware-mapped vif on this chipset) be expected to help, or is that not how MTK's vif allocation works?
- Any diagnostic logs (hostapd.log, dnsmasq.log, dmesg/logcat around AP start) that would help narrow this down further? Happy to provide.
Let me know what additional info would help debug this.
Device Info
secure: nowlan0andwlan1exist under/sys/class/net/(confirmed vials)Problem
After running
virtualap start, the script completes with no errors and reports the AP as fully active:ip link show ap0also confirms the interface is fully up:However, the SSID never appears in a WiFi scan on any test device — tried:
What I've tried (all failed to fix visibility)
iw reg getshowedcountry 00: DFS-UNSETwithPASSIVE-SCANflagged on all 5GHz ranges. Manually patched the script's hardcodedcountry_code=UStocountry_code=AE(my actual region) inwrite_hostapd_conf(). Confirmed the generatedhostapd.confpicks upcountry_code=AEcorrectly on next start — no change in visibility.ap0is being created here (e.g., wrong physical interface being targeted).wlan1exists as a separate interface alongsidewlan0(ls /sys/class/net/) — not yet tried forcing VirtualAP to use it via-w wlan1, wanted to raise this first in case there's a known reason it wouldn't apply.Question / request
hostapdbelieves it's transmitting (PID alive, interface UP) but the firmware silently drops AP-mode beacon frames when bound towlan0(the STA-mode interface)?wlan1(if it's a distinct hardware-mapped vif on this chipset) be expected to help, or is that not how MTK's vif allocation works?Let me know what additional info would help debug this.