one moment! wrt1900ac != wrt1200ac! the port config is not the same!
on wrt1200ac port6 is the cpu port.

what do you want? trunk ports too?

so long

ps: it mabye "bricked" because u have to update your network config too? for example eth0 would be eth0.1 and so onf?!

(Last edited by trustno1foxm on 24 Oct 2015, 16:35)

I've tried the lasted trunk and driver for half a day, but just returned to OpenWRT CC final release with the 10.3.0.3 driver. 10.3.0.10 is seriously broken with anything that even remotely smells like a broadcom chipset. Performance of the remaining devices was about 20% less then under 10.3.0.3 as well.

Anyway, at least it was a nice learning experience how to properly compile your own image with all the needed packages. Once a fix is released, i'll give it another try.

It's working with one of my devices which uses a Broadcom BCM5352 chip rev 0

@muronghan
guys read wiki first ..
wrt1900ac => lan = 0 1 2 3 5 and wan = 4 6(almost missed switch layout inside wiki because of improper naming)
wrt1200ac => lan = 0 1 2 3 6 and wan = 4 5
both 5 and 6 are cpu ports; 0-4 are the plugs on the back of the router
also i think sw.port0=lan4 and sw.port3=lan1 (reference required; sow a picture somewhere)

so on wrt1200ac you tag sw.port5 and on wrt1900ac you tag sw.port6 for second wan

    config switch
        option name 'switch0'
        option reset '1'
        option enable_vlan '1'

    config switch_vlan
        option device 'switch0'
        option vlan '1'
        option vid '1'
        option ports '0 1 2 6'

    config switch_vlan
        option device 'switch0'
        option vlan '2'
        option vid '2'
        option ports '4 5t'

   config switch_vlan
        option device 'switch0'
        option vlan '3'
        option vid '3'
        option ports '3 5t'

ps: so many edits on wrt1900ac wiki page
ps2: wont those wiki linked picture be gone if that service goes?? howto upload pictures to openwrt wiki?

(Last edited by makarel on 25 Oct 2015, 00:09)

Fails on my Roku 2 XS and Apple TV. Both have Broadcom chips. The 10.3.0.3 driver works without issue.

Quote from Kong on DD-WRT site:
"Most likely these are phones with broadcom wireless chips. The new driver drops packages in case the client side is a broadcom radio. Thus you hardly get a connection and performance is really bad.

The patch, that TireMeat posted on github is improving things for broadcom radios, but it makes it worse for other radios.

amsdu and ampdu is still severly broken. I don't expect this to be solved shortly. Obviously only one marvel developer works on this partly."

--bill

@makarel, thx. It did work (wan can not access at very beginning until I bridged wan with eth0.2). I configured eth0.3 as wan2 and mwan3 worked perfectly. Thanks a lot!

@muronghan

can you confirm that pls

Yeah, just as the diagram says, also the same order as my WNDRMAC2, port3=LAN 1 on the panel which is the closest to WAN port, port2=LAN2, port1=LAN3, and port0=LAN4.

WRT1200AC:
==========
WAN                  1   2   3   4 (ports)
 |                   |   |   |   |
 4                   3   2   1   0 (mapping)
----                 ---------------
 |                          |
 5 (internal)               6 (internal)

sorry for my awful code, but thats the wrt1200ac layout.

any ideas when there will be a new build for the wifi issues available? The wlan problems are really awful

@muronghan and @trustno1foxm
ty for info
i updated wiki switch section
http://wiki.openwrt.org/toh/linksys/wrt1900ac
http://wiki.openwrt.org/toh/linksys/wrt1200ac

I don't mean to be a bug but I posted a question in the General Discussion a week ago and haven't gotten any response.
You can find it here: https://forum.openwrt.org/viewtopic.php … 59#p296359

Basically I'm looking to find out what it would take to get an open-source router supported by OpenWrt.
My project is here: https://www.crowdsupply.com/krtkl/snick … gigglebits
and here: https://github.com/krtkl/schematics

Thanks.

Hi Makarel, are you the maintainer of the wiki page? Some small remarks:

1. The info on the V2 is wrong.
I have one and is has 512MB of ram, and a 1.6Ghz processor (allthough the box specifies something else)
2. Maybe mention that the new wrt1900ACS verion has not been tested yet? (I did saw Kaloz add some shelby code a couple of weeks back though)

Some proof for above statements on cpu and memory:

Memory:

:~# cat /proc/meminfo                              
MemTotal:         515584 kB                                    
MemFree:          453444 kB                                    
MemAvailable:     472912 kB

All clock speeds:

:/sys/kernel/debug/clk# cat clk_summary               
clock              enable_cnt  prepare_cnt        rate   accuracy   phase                                       
------------------------------------------------------------------------------                                  
oscillator                  0            0    25000000          0 0                                              
mainpll                     1            1  2000000000          0 0                                                
nand                        1            1   666666666          0 0                                              
cpuclk                      1            1  1600000000          0 0                                               
ddrclk                      0            0   800000000          0 0                                                 
l2clk                       1            1   800000000          0 0                                              
tclk                       11           11   200000000          0 0                                                 
sata1                       0            0   200000000          0 0                                                 
xor1                        1            1   200000000          0 0                                                 
tdm                         0            0   200000000          0 0                                                 
crypto0                     0            0   200000000          0 0                                                 
xor0                        1            1   200000000          0 0                                                 
crypto1                     0            0   200000000          0 0                                                 
usb2                        1            1   200000000          0 0                                                 
sdio                        0            0   200000000          0 0                                                 
crypto1z                    0            0   200000000          0 0                                                 
sata0                       1            1   200000000          0 0                                                 crypto0z                    0            0   200000000          0 0                                                
bm                          0            0   200000000          0 0                                                 
usb3d                       0            0   200000000          0 0                                                 usb3h1                      1            1   200000000          0 0                                                 usb3h0                      0            0   200000000          0 0                                                 
pex0                        1            1   200000000          0 0                                                 
pex3                        0            0   200000000          0 0                                                 
pex2                        0            0   200000000          0 0                                                 
pex1                        1            1   200000000          0 0                                                 
ge0                         2            2   200000000          0 0                                                 
ge1                         0            0   200000000          0 0                                                 
ge2                         1            1   200000000          0 0                                                 
audio                       0            0   200000000          0 0                    

(Last edited by JohnnySL on 25 Oct 2015, 20:32)

Correct the OpenWRT wiki on the WRT1900ACv2 is indeed wrong or to be more precise Linksys's specs on the box of the WRT1900ACv2 are wrong.

The WRT1900ACS is in fact a rebadged WRT1900ACv2 they use identical hardware and thus have the same specs.

You can read all about it here:
https://wikidevi.com/wiki/Talk:Linksys_WRT1900AC_v2

nope no maintainer just casual modifier of wikis..
as i understand wiki can be edited by anybody just follow the general formatting of the page

regarding 1.6Ghz processor version.. i understand its an 1900ACS model..
seems Linksys is pulling all sort of snakes out of the hat.. but from specs.. v2!=ACS .. read wiki especially cputype switch type and flash type and size

handle snakes with care.. especially poisonous ones

(Last edited by makarel on 25 Oct 2015, 21:31)

hi, i would like to know, if there is an working openwrt for the ac1900, two months ago i testes and failed, or maybe is there a step by step instruction how i may compile one with the working drivers (i am not expirienced in making sth with custom "linux")

D0nar, the 15.05 chaos calmer version of openwrt that you can just download from the website is currently one of the most workable releases. It still has wifi issues, but when daily rebooted it is for me at least more or less workable.

No need to start compiling your own version.

RC3 still remains the most stable and is recommended if you don't want to report bug fixes on CC Final.  I found CC Final to crash a lot, so I reverted back to RC3, and haven't had 1 crash in over a month. Of course, depending on your environment, stability could be different.

(Last edited by davidc502 on 25 Oct 2015, 22:37)

What wiki? Wikidevi shows identical switches for the ACS and the V2.
There is a big difference beteween the 1200AC and the V2, maybe that's what you mean?

hehe your in such a hurry to test things..
none's holding you from loading the fw and posting proper results?? from the link above seems ddwrt is working go see if owrt is ruling, share wifi performance thoughts as that is the brake factor of previews linksys snakes which it didnt fixed, yet it throws on another frog would you be the kind princess to kiss it
i would recommend though you read wiki and get a proper image.. 2 penny edit^^

edit: seems wrt1900acs is in fact wrt1200ac cpu 1.6.. are the 1200ac radios better regarding stability.. what a lame coverup and a branding mess^^

(Last edited by makarel on 26 Oct 2015, 01:23)

FYI; I know of a user that loaded a WRT1900AC V2 with the Linksys WRT1900ACS firmware without issue.

yeap.. well from wiki specs wrt1200ac(base 1.2Ghz 512Mram)=wrt1900ac-v2(1.3Ghz 256Mram)=wrt1900acs(1.6Ghz 512Mram) all Marvell Armada 385

Anecdotal report:

I have NEVER had a crash on my WRT1900AC V1 running gargoyle 1.9.X (based on openwrt 47165).

I have on my network:
4 iPhones
4 iPads
1 iPod touch
1 iMac
2 MacBooks

More fruity devices than you can poke a stick at.

please don't jump in the middle of another thread for something completely unrelated.

If you have a device that runs linux, then support in openwrt is pretty trivial. you need to make a definition that will compile things for your system and submit it to be included.

If you ren't the one developing the router, then you need to have someone who is comfortable with compiling the kernel to be your maintainer.

I took a glance at recent 40 pages and did not find the similar situation as mine took place on my WRT1200AC from yesterday (I got the router the day before yesterday and flashed OpenWrt CC final) as below. Forgive me if such issue was mentioned or discussed before.

It is about the 2.4GHz wireless connection. I configured a guest wifi on 2.4GHz (but not  for 5G), country code US or CN, security WPA2-AES, channel 6 or 11, width 20 or 40MHz (as I tried a few times).

The issue was, the devices connected with 5G radio would remain the connection, while those with 2.4GHz's would lose the connection (maybe standby?) after a while, and the 2.4G LED on the front panel of the router was gone. Then, when 2.4GHz devices (iPhone4, iPhone 5S with iOS8, Kindle and Huawei phone) came back, they could not be connected. I re-input password onto those devices, but it did not word. And restarting the 2.4GHz radio did not help, until a rebooting fixed it.

Will appreciate some advice on it, or such thing might come from the immature Marvell wireless driver and is supposed to be fixed someday in the near future?

==========================
Maybe a cron job of rebooting my router in the middle night everyday may fix it temperately but I am not fond of it.

(Last edited by muronghan on 26 Oct 2015, 07:25)