Max HDD Size on USB / SATA Port with BPI-M3

so, here the direct Info from Sinovoip: we use GL830 for USB to sata.

I would appreciate more direct support, so i must pull this Information from their Nose, which isn’t funny for me. If I had a wish free, I would wish for me that they read more in their forum and give more detailed info about the questions to their Hardware.

ty Sinovoip!

we use GL830 for USB to sata

BPI-M3 sata schematic diagram and GL830 datasheet download

https://drive.google.com/folderview?id=0B4PAo2nW2KfnfjNVdkxhVktnTUJEdXhlZXZrN2YtVW1LSXJlNW1MZUpDQllzVVFTYmFjUFE&usp=sharing

Since your post was directed to me… I knew that since I asked the very same frustrating questions some time ago back when the official forum wasn’t this one but another (see the links above!). And did they answer your real question whether you can use the full capacity of a HDD greater than 2TB? Will they update their so called ‘specs’? The answer is no because they simply don’t give a sh*t about correct informations or customer needs.

To find answers to questions many users have you have to crawl through different forums where the company’s CEO posts some links to google drive late at night. Whether these answers are correct or not… at least the person who gave them doesn’t care. The SinoVoip people call their hardware ‘open source hardware’ but they have no clue what ‘open source’ means. That’s more than just releasing parts of schematics if a user asks. It’s about providing any hardware information available in a reasonable way.

Like Olimex does for example: They release their hardware as Open Source Hardware and therefore you find anything you need on Github: https://github.com/OLIMEX/OLINUXINO

What does SinoVoip put on Github? Countless and useless cloned sources that do not contain the latest fixes even for their own hardware. They seem to be not able to understand the meaning of ‘open source’ :frowning:

Nope they did not answer the question.

Just assume there is a 2TB limit.

Why do you focus so much on “open source hardware”. Its hardware. They have OS for it, though with bugs. What did you want to do with the BPI?

I’m a hardware Developer, but if getting Input to this Device is so arduous. I will not buy it. If this begins so heavy, I won’t :frowning: I am not willing to pull all the Input out of their nose.

BUT: And I will tell all my Friends how easy I’m getting Information to Develop. I think they won’t buy, too, because there is no centralised Support that all the needed Inrformations know and give to the Developer.

It’s a shame, not telling directly the needed Informations, that I could dev things working with the M3.

I’m sorry and very disappointed.

I think they should work more on supporting their Customers, and will learn how to spontaneously give the needed Input to the Community. If they will not learn to Support, they will not sell a huge amount of these BaPi’s such the Suppliers from the Cubiboard CC-A80. They haven’t sold a huge amount of them, and it’s not sold anymore.

I thought at Sinovoip the World would be better than this.

Sorry for all this…

Developing is hard work man. That is why not many people do serious developing.

Thanks for your input and wish you good luck with other devices. There are lots of fishes in the ocean.

SinoVoip claims to produce “open source hardware”. That is simply a lie (but as usual, they don’t care at all – “f*ck the customers!”). If they would produce Open Source Hardware they would release documentation early and complete. And therefore not a single of this moronic threads would happen since the informations would all be both available and correct. That’s one huge advantage of the ‘open source’ attempt.

But what to expect from a company (or lets better say its CEO) that doesn’t even realise that hardware without appropriate software support will have a severely limited functionality (if you buy a product based on the crappy pseudo specs SinoVoip provides just to realise that your specific use case isn’t adressed at all due to no available drivers or bald lies like ‘the M3 has SATA support’ then you can use the piece of hardware as a paperweight and that’s it – and as usual: The manufacturer won’t care because he doesn’t even get the idea behind providing correct informations to customers).

If you keep in mind how the situation with the Lamobo R1 or the BPi M2 was (SinoVoip holding back necessary hardware informations/descriptions and actively preventing the linux-suxi community from jumping in to help) then it’s clear what will happen with the M3 again. But this time it will get even worse because the SoC used on the M3 has only basic community support (and the manufacturer doesn’t have the skills to do kernel development, in this area they rehash only software they took from Allwinner and the linux-sunxi community, obfuscate what they’ve – not – changed and provide millions of lines of code in countless/useless github repositories to let people believe an active software development would happen)

But maybe the whole ‘information policy’ is due to the target audience of the M3: Dumb/clueless people believing many CPU cores are great. The M3 with its new Allwinner A83T SoC is behind the old M1 in so many areas that it’s unbelievable anyone wants a board based on that.

A83T is better than the A20 for its android 5.1 support. Not everything is about Linux you know.

But you know that Android uses the linux kernel, yes? And that’s exactly the point or lets better say the difference between closed and open source development.

Do you find Allwinner here: https://android.googlesource.com/kernel/ – no. The reason is simple: Allwinner’s kernel sources are proprietary stuff. They don’t support their older hardware internally so you are forced to buy a new one. But you must be really mad to rely in 2015 on an outdated 3.4.37 kernel without any community support. But hey, maybe people are happy to get an Android 5.1 version running on the M3 and nothing interesting on an SBC (like GPIO stuff) really works like it’s the case with the M2 today.

Yea that is my point. People may not need to use he “functions” of a “open source board”. They can just treat it like a android board or something.

O yea tkaiser, A quick question, if you can find a kernel there, how can one compile an android image for that device? I want to make an android image for x86_64 chip. There already exists android images with devices, but how do i get one without buying it.

i want you can login our github. all image source code is open source on it, you can build yourself image , i know ,we have to do more work . but just need some time ,and get your support.

With included ‘expiration date’ like it happens now with older Allwinner SoCs where the vendors doesn’t gave a sh*t about mainlining support for their device but put together an SDK with proprietary stuff and drivers not maintainable by the community.

Do you get Android 5.x support for the A20? Nope. For the A31s? Maybe, but not from Allwinner since that SoC is even discontinued but supported partially by mainline kernel. Will you get Android 6.x support for the A83T? Nope.

This all applies only to Allwinner’s kernels. If they would’ve chosen to support the community and release their stuff in a way it could be included into mainline kernel (that’s the stuff ‘open source’ is about, not just taking only from the community but becoming a team player and giving back) then the ‘expiration date’ is gone since now the hardware is fully supported by mainline kernel.

That’s the point: Once the software for a piece of hardware is sent upstream and is included into the official kernel sources it’s easy to maintain support for it even with newer kernels (since it’s already in). When Android 6.x will be released and will require eg. kernel 3.18 for all new features than it would be an easy one to use the official Android SDK and the official mainline kernel to get a working Android on an older SoC. That’s the difference, the open source attempt makes. Even if you just want to use Android on the device and don’t care about its ‘openness’ at all this makes the difference regarding ‘expiration date’ and ‘vendor lock’.

Please remember: SinoVoip chose the A31s SoC for the M2 and sold it as being compatible with the M1 thereby tricking many many now really unsatisfied customers into buying this piece of hardware with wrong expectations. The OS images SinoVoip provided were based on the Allwinner SDK they had: Using kernel 3.0.0 (!!!) and u-boot-2011-something (both really crappy). All OS images based on this combination still suck more or less. The only good thing about the A31s is that it’s old enough that the linux-sunxi community jumped in and started mainline support from scratch (yes, it happens this way: You can not use code from Allwinner’s outdated 3.x kernels but you start writing drivers and device tree stuff from scratch to get the hardware supported in mainline kernel)

What’s the situation with Allwinner these days and with the A83T SinoVoip chose for the M3? Allwinner is already on kernel 3.10 (for their most recent SoCs like the A64 they did the work. Mostly due to the requirement of Android 5.x needing kernel 3.10 for all features). What do we get with the A83T instead? SinoVoip ships with Allwinner’s aging 3.4.37 kernel. The same kernel Allwinner used when they started with the A20 back then (think about what that means regarding potential support for Android 5.x on A20).

SinoVoip ignores everything the community did in the meantime (we’re on 3.4.110 now!) so you will be assured that the ‘SinoVoip experience’ with the M2 will happen again: Worst software support ever. And what does this mean regarding expiration date? Allwinner already stopped supporting their A83T SoC since they didn’t provide an SDK based at least on kernel 3.10. You’re already cut off from every Allwinner development right now and the community’s efforts to get A83T support in mainline kernel are in very early stages and its questionable whether the A83T will ever get full support (since Allwinner annoys developers by releasing every few months a new SoC with crappy outdated GPUs and a few cores more and crappy SDK).

To sum it up: The situation with the M3 will be even worse as with the M2 today regardless whether you’re trying to use Linux or Android on it. And that’s due to software issues and ‘open’ vs. ‘closed’ (something especially the manufacturer of the device is not able to understand)

Completely bullshit. You still have no idea what ‘open source development’ means. I know that you maintain a bunch of useless Github ressources. Have a look at this issue:

You release hardware and don’t provide the informations necessary to be supported by software. You’re holding back stuff again and again. Since you have no clue how this works. This affects the use of your devices regardless whether they should run Android or Linux. Since without correct device tree descriptions in more recent kernels or correct fex contents for older Allwinner kernels hardware simply doesn’t work.

Remember the GBit LAN issues with the M2? Or non working Wi-Fi and zero useable GPIO pins?! It was either wrong contents of the fex files or wrong device tree descriptions of YOUR OWN hardware. And we as the community asked you again and again for the fex file you used but you refused to provide. We asked to help you. But you still don’t get it! You said ‘just need some time’ – but that’s bullsh*t. If you would’ve released the fex file you use earlier we as the community would’ve helped you to fix the errors inside months earlier as it happened back then.

And since you still rely on this moronic attempt to ‘finish things internally and release them when ready’ the very same crap will happen again with the M3. Remember how much efforts it took to get correct hardware descriptions in your OS images since you prevented the community from helping you due to your release policy? Surely you don’t.

Your words above are the confirmation that you’ve learned nothing. The M3 will start with crappy OS images, many things won’t work as expected and you will prevent any attempts to fix that.