Am 20.06.2016 um 07:40 schrieb John Crispin:
On 20/06/2016 07:39, Mathias Kresin wrote:
Am 20.06.2016 um 07:19 schrieb John Crispin:
On 20/06/2016 05:15, Zhao, Gang wrote:
Hi, Mathias Kresin
On Sun, Jun 19, 2016 at 10:30 PM, Mathias Kresin <d...@kresin.me> wrote:
So far the network in failsafe is setup only for one board. Use the
eth0 interface as lan interface for all boards for now.
If a board has its lan interface(s) on another eth, a special
handling based on the board name can be added.
Signed-off-by: Mathias Kresin <d...@kresin.me>
---
.../lantiq/base-files/lib/preinit/05_set_preinit_iface_lantiq |
10 +---------
1 file changed, 1 insertion(+), 9 deletions(-)
diff --git
a/target/linux/lantiq/base-files/lib/preinit/05_set_preinit_iface_lantiq
b/target/linux/lantiq/base-files/lib/preinit/05_set_preinit_iface_lantiq
index 3d7fabf..7ed0fab 100644
---
a/target/linux/lantiq/base-files/lib/preinit/05_set_preinit_iface_lantiq
+++
b/target/linux/lantiq/base-files/lib/preinit/05_set_preinit_iface_lantiq
@@ -3,15 +3,7 @@
. /lib/functions/lantiq.sh
set_preinit_iface() {
-
- board=$(lantiq_board_name)
-
- case "$board" in
- TDW8970)
- ifname=eth0
- ;;
- esac
-
+ ifname=eth0
What about to add a default branch in case statement?
*)
ifname=eth0
;;
i am wondering if we should add a switch default config here or simply
reset the switch with vlans turned off
Is it really necessary? I mean the failsafe kicks in long before the
network config and vlan setup is done.
I've tested this with a board using the xrx200 in-SoC switch and a board
using a RTL8306G switch. In both cases it wasn't necessary to turn off
the vlans since they had not yet been configured.
Mathias
this code also runs on danubes with rtl and ar8327 switches. some of
which have a default vlan setup provided by the bootloader
John
Indeed, for these boards a reset of the vlan config, depending on the
particular bootloader vlan config, might be necessary.
But I don't like the idea of adding a default switch config for boards
where it isn't strictly required. Especially in a failsafe script. The
more code we add, the more likely is a breakage or an unconsidered
side-effect of a change.
Since I don't have any of the affected danube boards, adding support for
this special case to the patch is out of scope for me. I don't like to
add untested code.
Mathias
_______________________________________________
Lede-dev mailing list
Lede-dev@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/lede-dev
