OpenWrt Forum Archive

Topic: GPL source code VIOLATION + Comfast-CF-E316NV2-0

The content of this topic has been archived on 18 Apr 2018. There are no obvious gaps in this topic, but there may still be some posts missing at the end.

I recently bought a Comfast-CF-E316NV2-0


http://www.aliexpress.com/item/300Mbps- … 9.html?s=p

However although it does, as advertised, come with openwrt already pre-installed on it the firmware is a very poor implementation of openwrt with no SSH etc etc

I have recently been trying the persuade COMFAST to provide the GPL'ed source code which they must do under the terms of the GPL.

However COMFAST have refused to do so.  I wrote to them attempting to explain the benefits of providing the GPL'ed source code however they have replied thus :-

Dear John ,

I thank you for your advise and email .

But you might not know that we have our own R&D team more than 25 software&hardware engineers .

If we want to change or add some firmware , we would do it ourseft and won't looking for support from outside anyone else. Will you able to understand .

In a word , thank you for your attention , I'm afraid we have no chance to going cooperation at this field .

You don't need to write email to me or anyone of our company .

Thanks .


B.R.
Anna Ling

Shenzhen Four Seas Global Link Network Technology Co., Ltd.
[Foreign Trade Dept.]
--------------------------------------------------------------------------------------------
Address: Room 607-610, Block B, TAOJINDI Electronic Business Incubation Base, Tenglong Road, Longhua District, Shenzhen, China.
Postal code: 518000
Tel: 0755-83790659 ext 887
Fax: 0755-83790059-808
Mobile: 0086-13924672008
Skype ID: comfast01
QQ:  896326054
Email: sales35@szcomfast.com
Website: www.szcomfast.com

(Last edited by triggerpoint on 11 Aug 2015, 13:20)

With that kind of attitude, there is a strong possibility they don't design their own hardware, but rather copy from other manufacturers who own the copyrights. Perhaps, it is time to boycott this company. For instance, on our OpenWRT Makefile for the packages, include the following statement after the OpenWRT copyright statements:

##########
In view of the response from Ana Ling of Shenzhen Four Seas Global Link Network Technology Co., Ltd. as shown in this post (https://forum.openwrt.org/viewtopic.php?id=58973) that clearly indicates a violation of CopyRight licenses from GPL and/or OpenWRT, we hereby declare to boycott Shenzhen Four Seas Global Link Network Technology Co.

We reserve our rights to follow up this violation with a law suite.
##########

This fancy company is a joke.. Do not buy from them they pretend to be the new Apple onto the market and they save cost on everything especially the hardware... ( probably the only similar point they have with Apple.. )

I sent them a similar series of emails that were completely ignored.

Just an update to this.  I recently bought some samples from this company and was aware of this forum thread, I asked them about running openwrt on their devices and the possibility of getting a copy of the source code.  They sent me a patch to add all their products to the openwrt buildroot, plus QCA proprietary drivers.

I'm planning on importing a large amount of their products to NZ and I made them aware of my intentions right from the start, plus my wife is from China and was also talking to them in Chinese at one stage, so this probably made a huge difference.

The patch they sent me is here:
http://www.gareth.net.nz/openwrt/Comfas … 160311.rar
The patch is against trunk, commit 44ae85c49f629ec18a584b48ee2f7f1cfb825b2a
March 25th 2015

There are some issues with their code in the mach-comfast.c file, causing a kernel panic which i'm fixing and will post later.

(Last edited by chessman on 19 Apr 2016, 08:22)

chessman wrote:

Just an update to this.  I recently bought some samples from this company and was aware of this forum thread, I asked them about running openwrt on their devices and the possibility of getting a copy of the source code.  They sent me a patch to add all their products to the openwrt buildroot, plus QCA proprietary drivers.

The patch they sent me is here:
http://www.gareth.net.nz/openwrt/Comfas … 160311.rar
The patch is against trunk, commit 44ae85c49f629ec18a584b48ee2f7f1cfb825b2a
March 25th 2015

Nice and thank you.

There are some issues with their code in the mach-comfast.c file, causing a kernel panic which i'm fixing and will post later.

I am looking forward to the fix.

I've been doing most testing on the WR650AC.  Its a dual band gigabit router.  Getting a strange problem now with the internal switch on this device, possibly someone here might be able to shed some light on it.

Chipset is QCA9558/QCA9880 and switch is a 7 port QCA8337N
Port0 (CPU) goes to eth0 and port6 goes to eth1.
Default switch config provided by comfast is as follows:

config switch_vlan
        option device 'switch0'
        option vlan '1'
        option ports '0 2 3 4 5'

config switch_vlan
        option device 'switch0'
        option vlan '2'
        option ports '1 6'

However with the ath9k kmod driver installed, Port0 (CPU) stops working and hence LAN ports do not work.  The interfaces br-lan and eth0 still show using ifconfig.  Port6 still works and so does eth1 (wan).

As a work around I enabled tagging on port6 back to the CPU for lan/wan and left port0 off.  If I disable the ath9k driver in make menuconfig, port0 (CPU) works no problem, but then there is no wifi on 2.4g. I also discarded the proprietary QCA drivers, never imported them to the buildroot so they are not the cause.

Any ideas?

Edit:
Got this fixed now, had to do a mac swap on the 8337 switch, port0 and port 6.

Added to machine file:
.mac06_exchange_en = true,

Documented here:
https://lists.openwrt.org/pipermail/ope … 32798.html

The kernel panic I described earlier was due to the comfast provided patch trying to disable JTAG using:
ath79_gpio_function_setup(AR934X_GPIO_FUNC_JTAG_DISABLE, 0);

and there is also a Pericom PT7A7514 on the pcb which is used by the watchdog

So it looks like their patch may have been against a slightly different board of the same model.

(Last edited by chessman on 21 Apr 2016, 12:31)

The discussion might have continued from here.