Yes but linux kernel need a direct ram mapping. And kernel mapping is 1G.
That's why highmem exist http://linux-mm.org/HighMemory .

However, many people insist on using more than 1GB of physical memory on such a 32 bit system. This makes it necessary for the Linux kernel to jump through some interesting hoops... Basically the system uses the following tactics: * Memory above the physical address of 896MB are temporarily mapped into kernel virtual memory whenever the kernel needs to access that memory. * Data which the kernel frequently needs to access is allocated in the lower 896MB of memory (ZONE_NORMAL) and can be immediately accessed by the kernel (see Temporary mapping). * Data which the kernel only needs to access occasionally, including page cache, process memory and page tables, are preferentially allocated from ZONE_HIGHMEM. * The system can have additional physical memory zones to deal with devices that can only perform DMA to a limited amount of physical memory, ZONE_DMA and ZONE_DMA32. * Allocations and pageout pressure on the various memory zones need to be balanced (see Memory Balancing).

Cortex A15 uses 40-bit physical addressing is really useless : it is x86 PAE

But 64 bits is really interesting for the kernel. With Linux you can only map 1G of map without using segmentation (highmem).

Using 64 bits kernel with 32 bit programs solve the issue. And that's what is done on sparc, ppc, mips...

Also some algorithm really need 64 bits arithmetic to be faster.

No they should have construct IP adress like phone number : an extend it by adding prefix/suffix.

It should have simplified router job, because they could have done the routing only on some part of the number

Yes PAE is ugly and slow. It's been long time MIPS switch to 64 bits. I wonder why arm don't do it. Smartphone tends to get bigger and bigger memory...

It's mathematically impossible to make a device completely safe from someone who has complete physical control over it. You can encrypt this and that all you like, but it's literally only a matter of time before someone applies enough computing power and breaks said rights-management. Boot loaders can be heavily obfuscated against reverse engineering, but since the device has to actually boot and work at some point, there's a key to the proverbial lock in that haystack somewhere. I hope I'm making sense, coffee hasn't kicked in yet. You are wrong. First encryption is not used, it is digital signature (rsa). Second to be safe the first signature check should be done on the CPU ROM using internal RAM (no MITM). In this case if the signature algorithm is mathematically safe, you're device is broken if : - if somebody publish private key - if there bug in the ROM There can be others bugs/holes in upper layer of software (bootloader, kernel, ...), they can be patched by an update.

could be great to propagate rootkited OS ;) Of course authentication is needed (but the code for doing that could be big and complex).

Did you notice that the author of the blog entry is a developer of theorarm. His point of vue is not necessary the same as google...

And people seems to also forget that nouveau is ATM Linux only. Bsd users will have to use vesa or the binary blob...

http://x264dev.multimedia.cx/?p=292 list some missing feature a html5 :
- buffering control
- signalling back to server
- Quality of implementations

So why nobody have wrote a codec for theora on this dsp ?
AFAIK theora spec is quite old

PS : a dsp is not magic, you have limitation and can't make all codec work on it

Not really some architectures like arm doesn't have this bug (because page 0 is used for something else).
Also one of the most use of mapping page0 is for running ... dosemulator on x86 via vm86 [1].

So they seems related.

Especially since on a 64 bit distro pretty much everything, with very few exceptions is 64 bit.
You should look at ppc and sparc distro : pretty much everything is ... 32 bit. That because 32bit is more efficiant than 64 bits (less code/data size). But on x86 that's different because that's on the same arch at all (not the same register/feature, ...).

You will be glad to learn that last arm cpu (cortex A9) do OoOE. Also newer arm got mixed 16/32 bits instruction with thumb2. And yes newer arm CPU eat lot's of more power that older generation (armv5), but there are more powerfull.

But lot's of media oriented ARM platform already got h264 (and other) hardware accelerator...
It will be difficult to beat them with pure software.

And how will you recover the data ? The keyboard can't call home on its own.