В Mon, 24 Oct 2011 16:10:14 +0200
Daniel Golle <dgo...@allnet.de> пишет:
> Not particularly beautiful but doing the trick...
>
> Signed-off-by: Daniel Golle <dgo...@allnet.de>
>
> diff --git a/utils/usb-modeswitch/files/modeswitch.hotplug
> b/utils/usb-modeswitch/files/modeswitch.hotplug
> index 1aecb1f..7f9ce94 100644
> --- a/utils/usb-modeswitch/files/modeswitch.hotplug
> +++ b/utils/usb-modeswitch/files/modeswitch.hotplug
> @@ -120,7 +120,20 @@ if [ "$ACTION" = add ]; then
> # If a candidate is remaining, start usb-modeswitch
> [ -n "$configs" ] && {
> log "$DEVICENAME: Selecting ${configs%% *} for mode
> switching"
> - $modeswitch -c "${configs%% *}"
> + # ugly workaround, but working for all hw we got for
> testing
> + switching_done=0
> + switching_tries=0
> + local usb_dir="/sys/$DEVPATH"
> + [ -f "$usb_dir/idVendor" ] || usb_dir="${usb_dir%/*}"
> + while [ $switching_done -lt 1 -a $switching_tries -le 6
> ]; do
> + $modeswitch -I -D -n -s 30 -c "${configs%% *}"
> + if [ $(sanitize "$usb_dir/idProduct") -eq $uPid
> ]; then
> + log "switching seemingly failed"
> + else
> + switching_done=1
> + fi
> + switching_tries=$(( $switching_tries + 1 ))
> + done
> }
> }
> fi
Can you please send the log of usb_modeswitch and dmesg when the
problem occurs?
Probably it can be fixed easier and much more reliable in the
usb_modeswitch code. Also other platforms will benefit from it.
--
Alexander
_______________________________________________
openwrt-devel mailing list
openwrt-devel@lists.openwrt.org
https://lists.openwrt.org/mailman/listinfo/openwrt-devel
