ramips: add support for Netgear EX3700 by jclehner · Pull Request #937 · lede-project/source

jclehner · 2017-03-10T07:37:21Z

Specifications:

SoC: MT7620A
RAM: 64 MB DDR
Flash: 8MB NOR SPI flash
WiFi: MT7612E (5Ghz) and builtin MT7620A (2.4GHz)
LAN: 1x100M

This patch would not have been possible without the help and input of
Paul Oranje (@poranje).

This patch makes specifying NETGEAR_REGION optional, in which case mkchkimage will default to region 1 (WW). Signed-off-by: Joseph C. Lehner <joseph.c.lehner@gmail.com>

mkresin

Hey Joseph,

excellent work!

Would you please add Pauls Signed-off-by to the ocmmit message as well. Please add a short description how the initial install of LEDE can/should be done to the commit message.

mkresin · 2017-03-10T07:51:44Z

target/linux/ramips/base-files/etc/board.d/01_leds

Drop this LED here, it is already handled by diag.sh.

I would suggests to use the green:power LED in diag.sh. Everything reddish indicates an error to me. You can drop the default-state = "on"; from the green power LED in dts afterwards.

In case the amber LED is is switched on by the bootloader, add a default-state = "off"; to the amber power led in the DTS (http://lxr.free-electrons.com/source/Documentation/devicetree/bindings/leds/leds-gpio.txt).

The blinking amber led on bootup is used by the stock firmware on this device, so we wanted to replicate this in LEDE. I did a similar thing in the EX2700 port.

The blinking amber led on bootup is used by the stock firmware on this device, so we wanted to replicate this in LEDE. I did a similar thing in the EX2700 port.

We had similar discussion about LEDs in one of recent PR: #865 (comment).

If this device contains two LEDs which can be used as status LED, I also prefer to select the green one over the red/amber/orange.

mkresin · 2017-03-10T07:52:39Z

target/linux/ramips/base-files/etc/diag.sh

Please move the whole block a few lines up so that the file is in alphabetical order again.

mkresin · 2017-03-10T07:54:38Z

target/linux/ramips/dts/EX3700.dts

Would you please swap the include lines so that they are in alphabetical order.

mkresin · 2017-03-10T07:56:09Z

target/linux/ramips/dts/EX3700.dts

Drop this blank line.

f00b4r0 · 2017-03-10T12:31:10Z

target/linux/ramips/image/mt7620.mk

 endef
 TARGET_DEVICES += wn3000rpv3

+define Device/ex3700


Shouldn't that be after ex2700 (alphabetically-wise)?

nope. this file is a mess anyway. I will sort it after the patch is merged.

f00b4r0 · 2017-03-10T17:40:07Z

target/linux/ramips/dts/EX3700.dts

+
+&wmac {
+	ralink,mtd-eeprom = <&factory 0x0>;
+};


what about mac-address? Is it correctly set? On the WN3000RPv3 the wireless MAC was stored at a different offset in the factory partition...

Several tests have shown that the MAC address is correct (same as when using the stock image, on two different devices).

Ok, thanks. I'm curious: if you change either of the MACs in u-boot, is that change correctly reflected in LEDE?
I notice this device's MTD appears to be structured quite differently from the ex2700/wn3000, which surprises me. Maybe there's room for improvement on these two?

@f00b4r0 not sure what you are trying to find here. They have tested it on two devices and it was fine on both. It should be enough confirmation that the way it is done is the correct way.

There is no need to let them do more work than necessary. Especially since @jclehner has a good track record in submitting patches.

@mkresin I noticed on the wn3000 that if I didn't specify the correct MAC address location, the wlan would seemingly get the right MAC address, but that address wasn't updated when changed in u-boot (contrary to stock firmware). Hence my concern. Forgive me if I look like I'm splitting hair here, I'm just trying to ensure that we get most accurate support.

@mkresin ok. It's quite possible the ex2700/wn3000 are more different from the ex3700 than I expect, which would explain the discrepancy in DTS. As for the ability of changing the device MAC address, I'm told the primary use is to resolve collisions, though I never experienced one myself :)

@f00b4r0 The EX2700 and WN3000* are both manufactured by Delta Networks, while the EX3700's ODM is Foxconn, so this explains the difference in MTD layouts. The stock firmware does store MAC addresses in the config and nvram partitions, but so far, these matched the EEPROM MACs on all tested devices.

@mkresin I've updated the patch, so the diagnostic LED thingy is the only thing holding back this PR at this point, correct?

@f00b4r0 The EX2700 and WN3000* are both manufactured by Delta Networks, while the EX3700's ODM is Foxconn, so this explains the difference in MTD layouts. The stock firmware does store MAC addresses in the config and nvram partitions, but so far, these matched the EEPROM MACs on all tested devices.

Thanks for the clarification! Re MAC: I suppose this means that the WLAN MAC cannot be changed with the current DTS config then?

@mkresin I've updated the patch, so the diagnostic LED thingy is the only thing holding back this PR at this point, correct?

Correct

Latest patch fixes that too.

Specifications: * SoC: MT7620A * RAM: 64 MB DDR * Flash: 8MB NOR SPI flash * WiFi: MT7612E (5Ghz) and builtin MT7620A (2.4GHz) * LAN: 1x100M The -factory images can be flashed from the device's web interface or via nmrpflash. Co-authored-by: Paul Oranje <por@xs4all.nl> Signed-off-by: Paul Oranje <por@xs4all.nl> Signed-off-by: Joseph C. Lehner <joseph.c.lehner@gmail.com>

mkresin · 2017-03-12T08:29:14Z

merged, thanks.

poranje · 2017-03-13T11:28:45Z

tested snapshot images of monday 2017-03-13:
sysupgrade works
flash of factory image works
failsafe (with green blinking power LED) works.

build: make NETGEAR_REGION optional in netgear-chk

4a1bf68

This patch makes specifying NETGEAR_REGION optional, in which case mkchkimage will default to region 1 (WW). Signed-off-by: Joseph C. Lehner <joseph.c.lehner@gmail.com>

mkresin requested changes Mar 10, 2017

View reviewed changes

mkresin added the ramips label Mar 10, 2017

jclehner force-pushed the ramips-ex3700-final branch 2 times, most recently from 7932cbc to ab4ce9d Compare March 10, 2017 08:38

f00b4r0 reviewed Mar 10, 2017

View reviewed changes

jclehner force-pushed the ramips-ex3700-final branch from ab4ce9d to 4755f82 Compare March 11, 2017 16:49

mkresin self-assigned this Mar 11, 2017

mkresin closed this Mar 12, 2017

Uh oh!

Conversation

jclehner commented Mar 10, 2017

Uh oh!

mkresin left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

jclehner Mar 10, 2017 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

f00b4r0 Mar 11, 2017 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

mkresin commented Mar 12, 2017

Uh oh!

poranje commented Mar 13, 2017

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

Comments

jclehner Mar 10, 2017 •

edited

Loading

f00b4r0 Mar 11, 2017 •

edited

Loading