The A10, what is it?
David R. Lane (77) 766 posts |
There is an Allwinner A10 SoC and there is an Apple A10 and maybe other chips with the same name. Are they related? Whether or not they are related, my Son, who has an Apple pad with one in, asks does, or could, RISC OS run on it? Is it a realistic target for RISC OS? |
Chris Mahoney (1684) 2165 posts |
As far as I’m aware, they’re completely different products with the same name (although still ARM). In any case, you’d need cooperation from Apple to get RISC OS on there as by default it’ll only run code that Apple has “signed”. Good luck with that… |
Rick Murray (539) 13840 posts |
Not only that, but there’s a lot more to running RISC OS than “is it an ARM?”. Yes, both devices ought to be perfectly capable of running RISC OS if they support an ARMv7-like 32 bit world (they should, I’m not aware of any AArch64 only chips yet). But where is the USB interface (in memory)? What does it look like (register bits etc)? How do you talk to the display? What is the protocol for initialising the MMU? What timing does the RAM expect? It’s there an RTC? How do you talk to it? IIC? How does storage appear? Is it a faked IDE or SD or what? Can one format the entire device, or is some of it reserved for a recovery bootloader? How do you access a serial port or JTAG for initial development? If the user tattoos the display, this will probably generate an interrupt. What, where, what data is provided to register the position of a tap? If I tap-drag, does the interface continually interrupt or should I poll until there’s no more tap? I don’t think RISC OS supports automatic screen rotation, but if it did, how do I tell if the screen is this way or that way? What’s the battery state? Does the OS deal with battery charging or is that autonomous? It would be reasonable to force a shut down if there’s only 5% battery life remaining. How do I determine this? Pretty much every ARM SoC is capable of running RISC OS, it’s just all the other stuff is much more important and usually either simply not documented or is buried under a pile of NDAs… Things are easier for Linux. It’s a pretty generic kernel that talks to a number of pre-built system modules for the platform specific bits, and most of these modules are not open source (such as the plethora of cheap and cheerful Allwinner A10 devices). |
James Wheeler (3283) 344 posts |
You’d need to hack the bootloader first. Second, Apple’s A-series chips don’t provide public docs, afaik. I remember Android rom cooks complaining about Samsung’s CPUs (Exynos?) being undocumented and causing nothing but headaches in bringing custom roms to Samsung phones, so I cant imagine that porting RO to the iPad is a simple task. In theory, yes. In practice, very unlikely at the moment. Edit: Fix typos. This is what I get for replying on a iPad. Thanks fatigue and AC. |
Rick Murray (539) 13840 posts |
There, fixed that for you. Remember, it’s Apple we’re talking about. Raising the dead would be more successful than getting any technical information whatsoever out of Apple. There are zero unsanctioned firmwares for iPads. No Android, no Linux, nothing. |
James Wheeler (3283) 344 posts |
Thanks. You’ve always got my back
I never would expect Apple to accommodate. Our only hope is a hacker trying to crack the Apple bootloader, and reverse engineering an A-series chip. |
Ian Clark (1610) 2 posts |
The Apple A11 is A64 only. Not sure of any others. Shame really as the Apple ARM stuff would make amazing desktop CPUS. |
Timothy Baldwin (184) 242 posts |
Cavium ThunderX and ThunderX2 are also A64 only. Scaleway and Packet have them available to rent by the hour. If there’s a way to intercept SWI calls and writable and executable memory is available one could port RISC OS to iOS. |