OpenWrt Forum Archive

Topic: Update on Linksys WRT1900AC support

The content of this topic has been archived between 16 Sep 2014 and 7 May 2018. Unfortunately there are posts – most likely complete pages – missing.

Since the unit was very new and just released when i bought it , I believed the promises of the manufacturer . So i did not return it to the shop and waited .
My right to return this unit has now long been due , but no promise has been fulfilled . I am scammed and tricked by Belkin/Linksys . They have sold me something that was not to be found anywhere in the box when I opened it , and still cant be found anywhere .

Its time for Belkin to step up and create a full refund policy . How do we go about this ?

(Last edited by CrimInalA on 31 Jul 2014, 08:26)

greymattr wrote:

Some users may be disappointed to learn that the release will contain a pre-built library.  While this may not be ideal for some users, it is good progress and will allow for developers to recompile the driver as updates are made to the kernel.  Going forward I am sure efforts will be made to incorporate a wireless driver where developers have access to 100% of the source.

  Personally, I would like to thank you for your patience, and understanding through this process.  As soon as we are able to release the driver, I will post another update.

Quite frankly, this is completely ridiculous. Belkin has already released the source code of this driver in a GPL tarball. Interestingly enough, the driver there has GPL license headers apparently added by Marvell.

Having seen the driver, I can say that even with full source code available it would be hard to get it accepted into OpenWrt for a number of reasons:
- Crappy non-standard ioctl interface instead of standard Linux wireless APIs.
- No support for upstream hostapd, requires an old version hacked up with a bunch of crappy patches.
- Bad code quality.

Of course not even having full source code means this thing is even less useful than it would have been with complete source code access.

It seems to me that the only sane way to get it working is to add support for the chipset to mwl8k or mwifiex-pcie, but who's going to do the work on that? Probably not Belkin...

From people with contacts at Marvell I heard that this situation is unlikely to change in the future. It seems to me that Belkin got scammed by Marvell and is now trying to push the costs of that onto their users.

I would recommend that anybody who bought this device believing Belkin's promises of OpenWrt support should either ask for a refund or sell it to somebody who doesn't care about OpenWrt support.

blogic wrote:

The person to send the complains to is called Mike Chen

The more people that send him their reports of frustration the better.

Mike will be the person in charge of resolving the mess as he created the mess.

Wrong. People should push Marvell, as they are the ones behind this mess. Asked around, they've changed their mind and now one of the reply I've received states:

"I was told that the 8864 firmware file will not be available for public in near future."

You can express your feelings at http://www.marvell.com/company/contact.jsp and http://www.marvell.com/support/contact/cust_support.do

People should not push Marvell . In fact people should not be pushing anything if no false promises were made by Belkin/Linksys . It is not openwrt that made false promises , nor Marvell . Its Belkin/Linksys .

We all bought a Linksys product , its written all over the box in big letters . They are responsible for the false advertsing and the mess they created .

(Last edited by CrimInalA on 31 Jul 2014, 14:52)

jalyst wrote:
dpdurst wrote:

Actually I'm pretty excited about the news, can't wait to see what comes of it...


And tell us, exactly what "news" would that be?
The news that there's no final word about the closed binaries for the umpteenth time?


It sounded like the drivers where getting somewhere, its such a fluid environment changing. Just because your pissed that is on you not me. I was quite excited to hear we may have had something that is going work. Apparently that is not the case. Anything else?

Why does it matter that a driver has binary components? My understanding is most linux wireless drivers has binary components, notable exceptions are the older atheros drivers. Also isn't it true that wireless drivers are supposed to submitted to the kernel maintainers and not to openwrt developers to ensure regulatory support? My understanding is Marvell, as the device manufacturer, is not allowed to provide a driver that is completely FOSS unless it was certified by the FCC with a FOSS driver. I have experience with SDR development not with linux wireless driver development. I only done a cursory review of policies around linux FOSS wireless driver development. Does anyone know the specifics of this process? I'm starting to wonder if waiting for a completely FOSS driver from Marvell to be submitted to openwrt developers a reasonable expectation.

ducphuc wrote:

Why does it matter that a driver has binary components?

It matters because this particular manufacturer has released backdoored firmware in the past.  It matters because this isn't a binary blob, this is a very non compliant driver with a binary LIBRARY counterpart.

ducphuc wrote:

Why does it matter that a driver has binary components? My understanding is most linux wireless drivers has binary components, notable exceptions are the older atheros drivers. Also isn't it true that wireless drivers are supposed to submitted to the kernel maintainers and not to openwrt developers to ensure regulatory support? My understanding is Marvell, as the device manufacturer, is not allowed to provide a driver that is completely FOSS unless it was certified by the FCC with a FOSS driver. I have experience with SDR development not with linux wireless driver development. I only done a cursory review of policies around linux FOSS wireless driver development. Does anyone know the specifics of this process? I'm starting to wonder if waiting for a completely FOSS driver from Marvell to be submitted to openwrt developers a reasonable expectation.

The reason it matters is because the driver has code issues that can be solved and improved upon by other developers. Also the Open Source developers are saying they need the source code to fix bugs in the current OpenWRT solution. DD-WRT is saying the same thing.

WRT54G was a huge success with Broadcom radios in the Open Source community because Broadcom released the driver source code.

I wish the same for the WRT1900AC.

blogic wrote:

The person to send the complains to is called Mike Chen

The more people that send him their reports of frustration the better.

Kaloz wrote:

People should push Marvell, as they are the ones behind this mess.

You can express your feelings at http://www.marvell.com/company/contact.jsp and http://www.marvell.com/support/contact/cust_support.do

I guess our last resort is to make noise.
1. Use your insider contacts. This dude, anclar, posted on the Linksys forum an email to Marvell PR Director Ms Sue Kim.
2. Contact Belkin support.
3. Contact Marvell as Kaloz says.

Ask for the sources, for a FOSS driver. Ask for open source support. Ask for a refund. And let us know here how it goes.

nyt wrote:
ducphuc wrote:

Why does it matter that a driver has binary components?

It matters because this particular manufacturer has released backdoored firmware in the past.  It matters because this isn't a binary blob, this is a very non compliant driver with a binary LIBRARY counterpart.

OK, I see. The binary components are more than the firmware. Thanks. Still, shouldn't it be submitted to the kernel developers? I can see why Marvell legal would be reluctant to release source to openwrt developers first.

Kaloz wrote:
blogic wrote:

The person to send the complains to is called Mike Chen

The more people that send him their reports of frustration the better.

Mike will be the person in charge of resolving the mess as he created the mess.

Wrong. People should push Marvell, as they are the ones behind this mess. Asked around, they've changed their mind and now one of the reply I've received states:

"I was told that the 8864 firmware file will not be available for public in near future."

You can express your feelings at http://www.marvell.com/company/contact.jsp and http://www.marvell.com/support/contact/cust_support.do

I agree, it's a Marvell mess.

gufus wrote:
Kaloz wrote:
blogic wrote:

The person to send the complains to is called Mike Chen

The more people that send him their reports of frustration the better.

Mike will be the person in charge of resolving the mess as he created the mess.

Wrong. People should push Marvell, as they are the ones behind this mess. Asked around, they've changed their mind and now one of the reply I've received states:

"I was told that the 8864 firmware file will not be available for public in near future."

You can express your feelings at http://www.marvell.com/company/contact.jsp and http://www.marvell.com/support/contact/cust_support.do

I agree, it's a Marvell mess.

Marvell didn't make these promises. Marvell didn't release hardware claiming open source / openwrt compatibility when it wasn't true.

nyt wrote:
gufus wrote:
Kaloz wrote:

Wrong. People should push Marvell, as they are the ones behind this mess. Asked around, they've changed their mind and now one of the reply I've received states:

"I was told that the 8864 firmware file will not be available for public in near future."

You can express your feelings at http://www.marvell.com/company/contact.jsp and http://www.marvell.com/support/contact/cust_support.do

I agree, it's a Marvell mess.

Marvell didn't make these promises. Marvell didn't release hardware claiming open source / openwrt compatibility when it wasn't true.

So,

Whats the bottom line here?

nyt wrote:

Marvell didn't make these promises. Marvell didn't release hardware claiming open source / openwrt compatibility when it wasn't true.

How do you know? From what it seems above Marvell did make these promises and Belkin/Linksys were passing them on.

drawz wrote:

Does Broadcom have open source drivers for that would support the R8000 that none of us know about? How about Quantenna in the case of the AC-87U? Nope. So it doesn't sound much better than the WRT1900AC in that respect.

WTH is wrong with this forum sw, sick of unsub'ing & resub'ing just so I can get email notifications again about new posts. R8000 will prolly be supported, going by Broadcom's track-record there's a good chance*, Marvell was always a gamble, RE AC-87U I had a temporary memory lapse, actually I recalled at the time of posting but CBF'd amending. For background on what will happen with it see here, approach is "similar" to 1900AC, only BSlayer/Kong can say whether it's been better to work with, ask them: http://www.dd-wrt.com/phpBB2/viewtopic. … 960#907960

dpdurst wrote:

It sounded like the drivers where getting somewhere, its such a fluid environment changing. Just because your pissed that is on you not me. I was quite excited to hear we may have had something that is going work. Apparently that is not the case. Anything else?

Try reading the last 6 pages (at least) & comprehending the posts within, that should prove invaluable to you...

CrimInalA wrote:

People should not push Marvell . In fact people should not be pushing anything if no false promises were made by Belkin/Linksys . It is not openwrt that made false promises , nor Marvell . Its Belkin/Linksys .

We kinda think the 2 OWRT devs who posted before you have a better idea... >.>
Regardless, as others have said it can't hurt to "make noise" to both parties...

nyt wrote:

It matters because this particular manufacturer has released backdoored firmware in the past.  It matters because this isn't a binary blob, this is a very non compliant driver with a binary LIBRARY counterpart.

+1


*Yes it's had it's speed-bumps, but it's been far less rocky than Marvell

(Last edited by jalyst on 2 Aug 2014, 05:59)

So what to expect?! Is this "fight" lost yet?!

The original linksys firmware is still not working as promised (e.g. Guest network) and there's no fix at all.
And neither Belkin nor Marvell seems very interested (besides greymatter) to push openwrt firmware development.

So this is over?!

The story is never over. Companies leverage open standards and open software to reduce costs and increase market size while simultaneously leveraging differences in their product to increase increase market share and profit.

We are at an interesting juncture where most of router services are handled by open source software. The  new profit generating portions of the routers are associated with the radios, particularly how they do beamforming. Everyone will be trying to keep this part of their routers secret.

dfarning wrote:

We are at an interesting juncture where most of router services are handled by open source software. The  new profit generating portions of the routers are associated with the radios, particularly how they do beamforming. Everyone will be trying to keep this part of their routers secret.

If it is in fact an issue with the way it does beam forming, they could potentially release an older implementation or a simple form of beam forming that they aren't using and doesn't really offer a benefit and then the open source community could modify this themselves vs having it be a proprietary algorithm. This might mean better/worse beamforming for the open source community but at least they would only have themselves to blame vs the manufacturers. This might be a worthwhile compromise if this is what is causing the delays/legal/other issues potentially?

All I know is I'm likely returning my WRT1900AC as there are faster routers already out and as the open support isn't there as advertised not worth me keeping. May get another one down the road if this ever changes as aside from this issue it seems to be very well built compared to many others I've had in the past.

dfarning wrote:

how they do beamforming. Everyone will be trying to keep this part of their routers secret.

How they do beamforming is defined by the 802.11ac specifications. What exactly would they gain from keeping something secret that's already public and everyone knows about?

javawolfpack wrote:

All I know is I'm likely returning my WRT1900AC as there are faster routers already out and as the open support isn't there as advertised not worth me keeping. May get another one down the road if this ever changes as aside from this issue it seems to be very well built compared to many others I've had in the past.

My WRT1900AC went back for a refund today. I've hung onto it for a while in the hope that Linksys would deliver, but the most announcement finally forced me to conclude that this is unlikely. I know there are people inside and outside who are working to resolve this, but I've reached the point where I'm no longer prepared to wait. It's a shame - it was a nice piece of hardware.

So, Linksys has lost a sale - and a customer. I will never believe their marketing again, and I will avoid their products in future - not because I believe they're bad, but because I believe in voting with my feet, and I'm not prepared send further business in the direction of a company that misrepresents its products. This probably wasn't malice on their part, but it was careless, unprofessional, and stupid to sell this router on the basis of OpenWRT support that didn't exist, doubly so when they obviously didn't even have legal authority to release the necessary code.

Huge thanks to greymattr for al his work trying to get this fixed. It's a great shame it couldn't have ended better.

sauron11 wrote:

My WRT1900AC went back for a refund today. I've hung onto it for a while in the hope that Linksys would deliver, but the most announcement finally forced me to conclude that this is unlikely. I know there are people inside and outside who are working to resolve this, but I've reached the point where I'm no longer prepared to wait. It's a shame - it was a nice piece of hardware.

So, Linksys has lost a sale - and a customer. I will never believe their marketing again, and I will avoid their products in future - not because I believe they're bad, but because I believe in voting with my feet, and I'm not prepared send further business in the direction of a company that misrepresents its products. This probably wasn't malice on their part, but it was careless, unprofessional, and stupid to sell this router on the basis of OpenWRT support that didn't exist, doubly so when they obviously didn't even have legal authority to release the necessary code.

Huge thanks to greymattr for al his work trying to get this fixed. It's a great shame it couldn't have ended better.

Very well said . Sums up my feelings very well , altough I can only dream of a refund . But if I could , i would .
I vote with my wallet too . So -1 customer again .

(Last edited by CrimInalA on 3 Aug 2014, 14:58)

I'm returning my WRT1900AC too sad

I'm as well returning my WRT1900AC - Belkin/Linksys just lost another customer sad

It should be named...

I'm returning my WRT1900AC forum

not

Update on Linksys WRT1900AC support forum

Good grief!

(Last edited by gufus on 3 Aug 2014, 22:26)

Meh, serves them right for believing marketing hype. Learning the hard way that you should never believe any marketing claims from a manufacturer without third-party verification first.

I would never buy a product based on the manufacturer claiming OSF support, after routers that have shipped with Openwrt or DD-Wrt out of the box in the past have almost always had ancient, out of date, or hacked/proprietary versions of them that are years behind the main trunk.