Only add the new udev rule for x86 targets.

The fundamental issue is that there is not a hardware.definitions
defined for the non-x86 targets yet. so there is nothing in the config
to evaluate and iterate the pcidevices.

This is a tmp fix.

Signed-off-by: Brian McGillion <[email protected]>

modules/microvm/virtualization/microvm/netvm.nix

@@ -85,8 +85,11 @@ let
 
           services.openssh = config.ghaf.security.sshKeys.sshAuthorizedKeysCommand;
 
-          # WORKAROUND: Create a rule to temporary hardcode device name for Wi-Fi adapter
-          services.udev.extraRules = ''
+          # WORKAROUND: Create a rule to temporary hardcode device name for Wi-Fi adapter on x86
+          # TODO this is a dirty hack to guard against adding this to Nvidia/vm targets which
+          # dont have that definition structure yet defined. FIXME.
+          # TODO the hardware.definition should not even be exposed in targets that do not consume it
+          services.udev.extraRules = lib.mkIf (config.ghaf.hardware.definition.network.pciDevices != [ ]) ''
             SUBSYSTEM=="net", ACTION=="add", ATTRS{vendor}=="0x${(lib.head config.ghaf.hardware.definition.network.pciDevices).vendorId}", ATTRS{device}=="0x${(lib.head config.ghaf.hardware.definition.network.pciDevices).productId}", NAME="${(lib.head config.ghaf.hardware.definition.network.pciDevices).name}"
           '';