Quick Netgear R6700 question

Post new topic   Reply to topic    DD-WRT Forum Forum Index -> Broadcom SoC based Hardware
Goto page Previous  1, 2
Author Message
sdifox
DD-WRT Novice


Joined: 22 Jul 2017
Posts: 1

PostPosted: Sat Jul 22, 2017 19:25    Post subject: R6700V2 is MediaTek. Reply with quote
Ordered one on Prime Day and found myself with a R6700V2 instead of R6700.

https://wikidevi.com/wiki/MediaTek_MT7621

The ROM is horrid, takes like 30 seconds to save settings. It froze on me six times in one day with one active client and in AP mode.

Looks like Netgear is doing cost down the sneaky way. You only see R6700V2 on the side panel and the front removed the 1GHz reference so just reads Dual Core Processor.

Requested a R6700 as replacement, hopefully it is what they send me, otherwise it's refund.
Sponsor
MrLinux
DD-WRT Novice


Joined: 30 Jul 2017
Posts: 2

PostPosted: Sun Jul 30, 2017 7:59    Post subject: Reply with quote
@sdifox I'm in the same boat and tempted to ask for a replacement and mention to send a regular R6700 and not an R6700v2.

Please let us know if you were able to get the "R6700"

Thanks
broly
DD-WRT Novice


Joined: 03 Apr 2010
Posts: 30
Location: edmonton

PostPosted: Sat Aug 12, 2017 20:58    Post subject: Reply with quote
found out about the r6700v2/r6900/r6800 using the MT7621AT chipset a few weeks ago and finally ordered an R6700v2 today.

hopefully this is the platform i've been waiting for. have been working on the MT7621 platform for a bit, have managed to migrate all of crispin's work to the 4.9 build i had compiled earlier this year (maybe 4-5 months ago).

hopefully this r6700v2 uses a packet-ish bootloader that will make it easy, but we shall see.

if any of you can pass some information about this router's bootlog/boot sequence, it'd be appreciated.
MrLinux
DD-WRT Novice


Joined: 30 Jul 2017
Posts: 2

PostPosted: Sun Aug 13, 2017 21:27    Post subject: Reply with quote
I got another R6700v2 from Amazon. Both were returned.
broly
DD-WRT Novice


Joined: 03 Apr 2010
Posts: 30
Location: edmonton

PostPosted: Tue Aug 29, 2017 16:04    Post subject: Reply with quote
morning everyone.

in the past week i've made some progress on this router. i am excited but i remain pragmatic in light of the possibility that the mt7621 drivers cannot be disabled to enable a 'minimal boot' that allows me to bootstrap the devicetree source file (".dts") with the compiled kernel.

as it is, without the device tree, the kernel boots and successfully maps the memory/cpu but cannot go any further because the driver calls nodes defined in the devicetree binary (".dtb" a compiled dts file), which cause a panic.

here is a bunch of pastebins documenting the progress. i hope to have better news soon, but it may be tricky figuring out how to get the sercomm-ed u-boot to load the kernel with the binary, as i have to determine how exactly it's calling the kernel after initialisation.

the fact that it can load the kernel suggests i should be able to fix the problem, but the address the bootloader loads the kernel, and the load/entry points defined in the uImage.bin itself, are likely different. i will need to find the address called by the bootloader before calling the kernel, in hopes of calling kexec or something else that will let me bootstrap.

https://pastebin.com/K7yuBYfv


lastly: i am confident in stating that, if i can get this thing to run, it will probably be the best router on the market with a 4.9.44 glibc+bash+gudev (among other things) kernel. it has 4:3 antennas (better than my r7000 iirc, which is 3:3), using the solid MT7615EN radios, 256 (!!) mb ram, meaning it has a lot of untapped potential.

out of the box it uses a 2.6.36 kernel with several (admirable) "hacks" in the form of script files, so the poor performance experienced by others on stock fw is not a surprise. i do not think they released it hoping people use stock fw, as it is not suited for the hardware.

bless,
g
broly
DD-WRT Novice


Joined: 03 Apr 2010
Posts: 30
Location: edmonton

PostPosted: Tue Aug 29, 2017 19:15    Post subject: Reply with quote
edit, just an update. i have made progress and the boot has progressed much farther now i have figured out how to get the DTB in line.

i am hoping this will move faster now, but i want especially thank giuseppe lippolis for asking The Crispy One (John Crispin, aka blogic) about the DTB and how it is called. turned out this exchange was the nugget of truth i needed to get the DTB called.

https://pastebin.com/LAewauYh

hopefully it is smoother sailing from hereon out. no guarantees though. delighted i've come this far to be honest!
broly
DD-WRT Novice


Joined: 03 Apr 2010
Posts: 30
Location: edmonton

PostPosted: Thu Aug 31, 2017 18:08    Post subject: Reply with quote
'lo again.

below i present what-should-be a fully-working glibc 2.26 on linux 4.9.44 on an MT7621, with all cores showing up as 1004Kc, vga arbitration enabled, and a lot of other goodies (https://pastebin.com/T31eiR3f). the passage i've written below is also included in the pastebin, so please excuse me as i feel my overall experience must be shared.

i have fixed the "cannot find init" issue. turns out i had forgotten to put a symbolic ink for ld.so.1 in the lib dir. after that was done, i have now gotten a kernel panic that says "attempted to kill init" with an exit code that is not of much help.

i have asked Ralf about the stack dumping, but it seems this is a non-trivial request (http://elinux.org/images/0/07/Intricacies_of_a_MIPS_Stack_Backtrace_Implementation.pdf https://gcc.gnu.org/ml/gcc-help/2008-01/msg00067.html props to Daney for being on this so early. [he is active in #mipslinux, btw!]). i am pondering as to whether i really want to mess with the compilation flags to get the dump, as i tried inserting an -fno-omit-framepointer and it didn't seem to recompile all the source files.

in any case, it's clear to me that the issue here is glibc. i have found that people on ARM platforms have reported situations where identical setups in uClibc had no issue, but ran into this same exit code when attempting to switch to glibc (https://raspberrypi.stackexchange.com/questions/47002/kernel-panic-after-swtiching-from-uclibc-to-glibc), which is what i suspect is happening here.

given that the garbage platforms known as ARM have been already been lacking with their "FPU" support, and lack of proper "glibc" support would be another nail in their (already prepared) coffin, the exit code felt like something along the lines of the following:

"trying to use glibc on mips? spent 65 bucks on an MT7621 and now you want the world? shut it down ASAP!"
(with this facial expression )


i also noticed that approximately around the same time i had made my announcement that i received my r6700 (august 21/22 2017), that the losers at lede made mention of their broke-ass non-hardfloat ('eabi', not 'eabihf') 3.x ARM kernel having issues (http://lists.infradead.org/pipermail/lede-dev/2017-August/008607.html). this post was the first i've seen of any kind of success for ARM and glibc, and i suspect it was deliberate as "they" (openwrt/lede developers) have been aware of my desires for over a year (glibc hardfloat embedded platform). since this (alleged) build is not hardfloat, producing a 3.x glibc kernel for non-hardfloat ARM may be somewhat ubiquitous (though i'm not inclined to go that far).

all in all, i am very satisfied with the support provided by netgear, ralink, mediatek, and the boys on freenode. it seems that the privileged instructions ran by glibc were destined to get "locked out" as they not only provide legacy support, but additional access and power when compared to their bullshit variants that pretty strip down and modify glibc code.

selling people hardware under the guise of "open source" that doesn't include glibc compatibility is unacceptable. early on, glibc was too big to fit so it wasn't an issue. now this has changed and the RAM/ROM has reached a point where it's realistic expectation.
i suspect that, on the MIPS side, glibc compatibility is deliberately restricted since it affords great access to the hardware, and i don't think that was the goal of this whole "open source" wave.

contrast the "industry"'s deliberate attempts to impede MIPS' progress to the benefit of ARM, who has an army of garbage "techies" (coughcough Linaro coughcough) that have had over two years to develop a glibc-based 4.x kernel for ARM (with or out without hardfloat [latter scenario is me being generous, as it isn't comparable to mips w/ glibc since mips has FPU support]) and failed, only producing a 3.x softfloat kernel that has issues.

they've had all the resources (openwrt devs, linaro devs, pretty much all open source devs), money (rothschild was a main investor in ARM; he's the one trying to make out like a bandit) and time (two years since i expressed my interest at building a glibc kernel [initially for ARMv7, but that failed and i moved to MIPS earlier this year]).

i'll let you guys be the judge.
broly
DD-WRT Novice


Joined: 03 Apr 2010
Posts: 30
Location: edmonton

PostPosted: Thu Sep 07, 2017 1:40    Post subject: great news Reply with quote
ladies and gentlemen,

today i present to you a nearly-booting R6700v2/r6800V2/r6900fw with the RALINK SDK suite of drivers. that is, no net_mediatek_soc/ethernet driver, and a return to the esteemed RAETH driver.

i am delighted that after spending a large amount of time integrating the RALINK drivers into the fw, that i got the proportional results. that is, as you guys can see below, both radios boot up and are identified as MT7615 (both of which can do 5GHZ).

it seems that an initrd is required to complete the boot, which i am excited to figure out.

shout out to all the guys on freenode for your love and support!!

now without further ado....


https://pastebin.com/nqYDn4XD

this took quite a bit of work. a lot of googling. some lucky clicks on those hits. either way, i am delighted.

don't know if i will be able to share the fw due to copyright and stuff. i am not an expert in that area so i will have to do some checking there to make sure every contributor gets their long-awaited due (seems to be over ten years when fixing the code i had to add to make it boot).

cheers!
mjb32803
DD-WRT Novice


Joined: 11 Nov 2006
Posts: 41
Location: Orlando, FL USA

PostPosted: Fri Sep 15, 2017 1:17    Post subject: Re: great news Reply with quote
broly wrote:
ladies and gentlemen,

today i present to you a nearly-booting R6700v2/r6800V2/r6900fw with the RALINK SDK suite of drivers. that is, no net_mediatek_soc/ethernet driver, and a return to the esteemed RAETH driver.

i am delighted that after spending a large amount of time integrating the RALINK drivers into the fw, that i got the proportional results. that is, as you guys can see below, both radios boot up and are identified as MT7615 (both of which can do 5GHZ).

it seems that an initrd is required to complete the boot, which i am excited to figure out.

shout out to all the guys on freenode for your love and support!!

now without further ado....


https://pastebin.com/nqYDn4XD

this took quite a bit of work. a lot of googling. some lucky clicks on those hits. either way, i am delighted.

don't know if i will be able to share the fw due to copyright and stuff. i am not an expert in that area so i will have to do some checking there to make sure every contributor gets their long-awaited due (seems to be over ten years when fixing the code i had to add to make it boot).

cheers!



Appreciate all your hard work! Sincere Thanks
cuy
DD-WRT Novice


Joined: 24 Sep 2017
Posts: 1

PostPosted: Sun Sep 24, 2017 20:58    Post subject: Re: great news Reply with quote
broly wrote:
ladies and gentlemen,

today i present to you a nearly-booting R6700v2/r6800V2/r6900fw with the RALINK SDK suite of drivers. that is, no net_mediatek_soc/ethernet driver, and a return to the esteemed RAETH driver.

i am delighted that after spending a large amount of time integrating the RALINK drivers into the fw, that i got the proportional results. that is, as you guys can see below, both radios boot up and are identified as MT7615 (both of which can do 5GHZ).

it seems that an initrd is required to complete the boot, which i am excited to figure out.

shout out to all the guys on freenode for your love and support!!

now without further ado....


https://pastebin.com/nqYDn4XD

this took quite a bit of work. a lot of googling. some lucky clicks on those hits. either way, i am delighted.

don't know if i will be able to share the fw due to copyright and stuff. i am not an expert in that area so i will have to do some checking there to make sure every contributor gets their long-awaited due (seems to be over ten years when fixing the code i had to add to make it boot).

cheers!

hey I have an R6700v2 and I am willing to put some work to get this rolling, let me know if we can work together!
ncik0
DD-WRT Novice


Joined: 09 Oct 2017
Posts: 1

PostPosted: Mon Oct 09, 2017 7:44    Post subject: Re: great news Reply with quote
broly wrote:
ladies and gentlemen,

today i present to you a nearly-booting R6700v2/r6800V2/r6900fw with the RALINK SDK suite of drivers. that is, no net_mediatek_soc/ethernet driver, and a return to the esteemed RAETH driver.

i am delighted that after spending a large amount of time integrating the RALINK drivers into the fw, that i got the proportional results. that is, as you guys can see below, both radios boot up and are identified as MT7615 (both of which can do 5GHZ).

it seems that an initrd is required to complete the boot, which i am excited to figure out.

shout out to all the guys on freenode for your love and support!!

now without further ado....


https://pastebin.com/nqYDn4XD

this took quite a bit of work. a lot of googling. some lucky clicks on those hits. either way, i am delighted.

don't know if i will be able to share the fw due to copyright and stuff. i am not an expert in that area so i will have to do some checking there to make sure every contributor gets their long-awaited due (seems to be over ten years when fixing the code i had to add to make it boot).

cheers!


I have a 6700v2 with me as well. let me know if you want any sort of tests done. Thanks for all the work you are putting in!!!
phuz
DD-WRT User


Joined: 31 Jan 2011
Posts: 91

PostPosted: Wed Oct 18, 2017 12:36    Post subject: Reply with quote
Thanks for the work on this. Add me to the list of disappointed people who ordered an R6700 and received the R6700V2. Crying or Very sad
_________________
http://www.essentialintegrations.com
jpolhard
DD-WRT Novice


Joined: 13 Dec 2018
Posts: 3

PostPosted: Thu Dec 13, 2018 20:39    Post subject: Netgear r6700v2 Reply with quote
Where is the firmware then???? POst a link!!!
Goto page Previous  1, 2 Display posts from previous:    Page 2 of 2
Post new topic   Reply to topic    DD-WRT Forum Forum Index -> Broadcom SoC based Hardware All times are GMT

Navigation

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You can attach files in this forum
You can download files in this forum