Unable to change wifi network because raspi-config does not do a rescan

[ylexus]

Tried to change the wifi network on my Raspberry Pi, but raspi-config kept immediately failing with

Error: No network with SSID 'XXX' found.

nmcli device wifi list was returning only one network, the one it's currently connected to. nmcli device wifi connect was failing with same issue as raspi-config. But after I ran nmcli device wifi rescan, nmcli device wifi list started hosing all networks around and nmcli device wifi connect succeeded.

I think raspi-config should execute rescan before reporting the error.

raspi-confog Version: 20250210

~ $ cat /etc/os-release
PRETTY_NAME="Debian GNU/Linux 12 (bookworm)"
NAME="Debian GNU/Linux"
VERSION_ID="12"
VERSION="12 (bookworm)"
VERSION_CODENAME=bookworm
ID=debian
HOME_URL="https://www.debian.org/"
SUPPORT_URL="https://www.debian.org/support"
BUG_REPORT_URL="https://bugs.debian.org/"

~ $ uname -a
Linux xxxxx 6.6.74+rpt-rpi-2712 #1 SMP PREEMPT Debian 1:6.6.74-1+rpt1 (2025-01-27) aarch64 GNU/Linux

[XECDesign]

I'm seeing rescan exit immediately, but I'm sure it actually takes a while for the scan results to come in. You're saying connect also exits immediately if the SSID is not already in the list.

I'm not entirely sure about how NM works under the hood, but wouldn't that means that if we ran one command after the other, it would still fail in your scenario?

It seems like we'd need list SSIDs check if it's in the list, if not, rescan, wait until it shows up or time out.

We should remove all the legacy wpa_cli sections anyway, so maybe it's something we can take a look at when we re-work this functionality for Trixie.

[ylexus]

Yes rescan does exit pretty quickly, but then it's documented to "initialise" the rescan, which presumably happens in the background. In my experiments this "fixed" connect and list as they were both now able to see other networks. So your suggested approach is sensible, need to initialise rescan, then wait for a while for the network to appear in the list, and only then execute connect. Alternatively, instead of retrying list just keep retrying connect until it times out or succeeds.