Jérémy Bobbio wrote: > Wild guess after having a look at the log: is the firmware only > requested when trying to brought the interface up? > > If you still have access to the necessary hardware, can you look how far > I am from what is really happening? Hardware is not mine and already in production unfortunatly. But looking at the log again, I think you may be right, the driver only complains about microcode once upped by netcfg: Jul 21 02:40:45 kernel: [ 29.419310] iwl4965: Detected Intel Wireless WiFi Link 4965AGN Jul 21 02:40:45 kernel: [ 29.469768] iwl4965: Tunable channels: 11 802.11bg, 13 802.11a channels Jul 21 02:40:47 main-menu[1466]: INFO: Menu item 'netcfg' selected Jul 21 02:40:47 kernel: [ 30.851885] iwl4965: iwlwifi-4965-1.ucode firmware file req failed: Reason -2 Jul 21 02:40:47 kernel: [ 30.851892] iwl4965: Could not read microcode: -2 FWIW, this doesn't affect other wifi drivers that use firmware. At least, the one I can easily test, ipw2200, requests firmware immediatly on load. > If something close happens, we probably need to find a way to "up" then > "down" every interfaces in ethdetect, just to be sure that we have the > necessary firmware. There's another way.. netcfg could run check-missing-firmware after it ups the interface (either after dhcp fails, or after static configuration). Or netcfg could bring interfaces up with a fake IP before trying dhcp at all, to avoid dhcp timeouts. But that may be overkill. -- see shy jo
Attachment:
signature.asc
Description: Digital signature