Commit Graph

6 Commits (fbd4214bb0727268aa535af5b600d38939ec8dbc)

Author SHA1 Message Date
Chuanhong Guo a43cbfe2e3 ramips: remove default switch setup in 02_network
ramips images now relies on explicit switch setup for proper failsafe
functionality. Remove default cases where it relies on vlan setup in
dts and add switch setup for devices affected.

Signed-off-by: Chuanhong Guo <gch981213@gmail.com>
4 years ago
Adrian Schmutzler 03e742dcd8 Revert "ramips: explicitly disable built-in switch for lan-only devices"
This reverts commit a1693bf626.

The rt288x and rt3883 devices in question don't have switches.
Only keep the merged case for rt305x.

Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
4 years ago
Sungbo Eo a1693bf626 ramips: explicitly disable built-in switch for lan-only devices
Commit 8f6334eb94 ("ramips: explicitly disable built-in switch when needed")
did not fix rt288x and rt3883 devices. This patch deals with them.

While at it, consolidate duplicate cases in interface setup.

Signed-off-by: Sungbo Eo <mans0n@gorani.run>
4 years ago
Adrian Schmutzler a858d6d0e2 ramips: rt288x: use flash location for wan_mac in 02_network
This uses the flash locations instead of eth0 MAC address to
calculate MAC address increments for WAN.

The change will make the MAC address setup of a particular device
more obvious and removes the dependency of 02_network on the eth0
initialization.

This removes the wan_mac setup for ralink,v11st-fe as this device
does not set up a MAC address for ethernet in the first place.

Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
5 years ago
Adrian Schmutzler d2b9333ea9 ramips: remove default case for MAC address assignment
So far, MAC address assignment in ramips has contained a default
case, which defined wan_mac = eth0 + 1 for _every_ device not
having an explicit case there.

This is not desirable, as many device supporters will just not
care or know about this definition, so another MAC address will be
introduced by accident. In some cases the wan_mac is assigned
although it is not needed, in other cases even addresses not
dedicated to the device will be used (e.g. wan_mac actually is
eth0 - 1, but during support nobody cared, so eth0 + 1 is used now,
which might actually belong to another device ...).

Thus, in this PR the former default case is converted to an
explicit case. This one comprises all devices not being accounted
for by other cases, reduced by those not having wan at all.
The big number of entries for this node might be another indication
that many of them wouldn't actually be there if there hadn't been
default wan_mac setup.

In exchange, the current "do nothing" case can be removed, as it
will be the new default case.

The devices being put in the newly created explicit case were
determined as follows:

1. Create a list of all devices based on the DTS files.

2. Remove all devices already having an explicit entry setting
   their address.

3. Remove all devices that only have lan set up in the first part
   of 02_network:

mt7620:
   - alfa-network,tube-e4g
   - asus,rp-n53
   - buffalo,wmr-300
   - comfast,cf-wr800n
   - edimax,ew-7476rpc
   - edimax,ew-7478ac
   - elecom,wrh-300cr
   - hnet,c108
   - kimax,u25awf-h1
   - kimax,u35wf
   - kingston,mlw221
   - kingston,mlwg2
   - microduino,microwrt
   - netgear,ex2700
   - netgear,ex3700
   - netgear,wn3000rp-v3
   - planex,cs-qr10
   - planex,mzk-ex300np
   - planex,mzk-ex750np
   - ravpower,wd03
   - sercomm,na930
   - yukai,bocco
   - zbtlink,zbt-cpe102
   - zte,q7

mt7621:
   - gnubee,gb-pc1
   - gnubee,gb-pc2
   - linksys,re6500
   - mikrotik,rbm11g
   - netgear,ex6150
   - thunder,timecloud
   - tplink,re350-v1
   - tplink,re650-v1

mt76x8:
   - alfa-network,awusfree1
   - d-team,pbr-d1
   - glinet,vixmini
   - vocore,vocore2-lite
   - tama,w06
   - tplink,tl-mr3020-v3
   - tplink,tl-wa801nd-v5
   - tplink,tl-wr802n-v4
   - tplink,tl-wr902ac-v3
   - vocore,vocore2
   - widora,neo-16m
   - widora,neo-32m

rt288x:
   - buffalo,wli-tx4-ag300n
   - dlink,dap-1522-a1

rt305x:
   - allnet,all0256n-4m
   - allnet,all0256n-8m
   - allnet,all5002
   - allnet,all5003
   - alphanetworks,asl26555-16m
   - alphanetworks,asl26555-8m
   - asus,wl-330n
   - aximcom,mr-102n
   - dlink,dcs-930
   - easyacc,wizard-8800
   - hame,mpr-a2
   - hootoo,ht-tm02
   - huawei,d105
   - intenso,memory2move
   - planex,mzk-dp150n
   - rt305x dlink,dcs-930l-b1
   - sparklan,wcr-150gn
   - tenda,3g150b
   - tenda,3g300m
   - tenda,w150m
   - trendnet,tew-638apb-v2
   - unbranded,a5-v11
   - vocore,vocore-16m
   - vocore,vocore-8m
   - wansview,ncs601w
   - zorlik,zl5900v2

rt3883:
   - loewe,wmdr-143n
   - omnima,hpm

4. Put the remaining devices in the new case.

Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
5 years ago
Adrian Schmutzler 19724e28c8 ramips: split base-files into subtargets
While most of the target's contents are split into subtargets, the
base-files are maintained for the target as a whole.

However, OpenWrt already implements a mechanism that will use (and
even prefer) files in the subtargets' directories. This can be
exploited to make several scripts subtarget-specific and thus save
some space.

In certain cases, keeping files in parent (=target) base-files was
more convenient, and thus no splitting was performed for those.

Note that this will increase overall code lines, but reduce code
per subtarget.

base-files ipk size reduction:
master (mt7621)   60958 B
split (mt7620)    46358 B (- 14.3 kiB)
split (mt7621)    48759 B (- 11.9 kiB)
split (mt76x8)    44948 B (- 15.6 kiB)
split (rt288x)    43508 B (- 17.0 kiB)
split (rt305x)    45616 B (- 15.0 kiB)
split (rt3883)    44176 B (- 16.4 kiB)

Run-tested on:
GL.iNet GL-MT300N-V2 (mt76x8)
D-Link DWR-116 (mt7620)

Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
5 years ago