Nice try, but it runs in ring 0, so it can jump into the kernel anywhere it wants.
Worse than that after boot, the BIOS runs in System Management Mode, which is delberatly designed to be non-interceptable by the OS.
The reason I ask is that the Linux model for 32 bit is to have a 4G/4G address space, where the user and kernel address space are completely disjoint
A 4/4 split is completely impossible. x86 doesn't switch CR3 or TR automatically. At an absolute minimum, the kernel needs all its entry points (exception, interrupt and syscall/sysenter targets), and hardware structures (active pagetables, GDT, LDT and TSS) mapped into all virtual address spaces.
When will Linux auto-detect the 'static' components of the hardware and do an install with a kernel compiled and optimized for that hardware - along with a vanilla kernel for emergency backup? (Something like Gentoo does)
Never - it is not in a general distros best interest. It is hard enough keeping on top of the kernel bugs with a signle configuration in use, let alone customising it to everyones hardware. Anyway, when it comes to drivers, the optimisations you can do are "dont load drivers you dont need", which is what already happens
The problem is, the fix I had to employ was to physically replace the co-opted DNS servers with more advanced equipment because the system software that was on them had no throttling capabilities nor was is capable of recognizing and rejecting suspicious queries.
Protecting against DDoS reflection attacks is very easy, but it requires all 1st-tier ISPs to perform egress IP validation, so packets coming from the end users trying to get onto the internet are checked that the IP address is correct. Filtering anywhere else is impossible because of transit routes, so by the time the second AS gets to inspect the packet, it could legitimatly be from anywhere.
The problem is that this costs money to implement and isn't in the interest of 1st-tier ISPs, so is unlikely to ever get done.
Beta is not bad code. It is bad design.
It is unknown code and bad design, which is arguably worse.
AMD64 has a prefix byte before first op code byte, so in 64bit mode no instruction is smaller then 2bytes
Nope - stack operations are just a single byte, even in 64bit mode.
I use a MacBook Pro for my main machine, but also have a Ubuntu desktop. I get irritated about switching between command-oriented hotkeys and ctrl-oriented hotkeys (cmd-a on OSX = ctrl-a on Linux/windows). I've looked over a lot of forums and have found that Gnome doesn't seem capable of changing hotkeys, while xfce and fluxbox can. The ideal solution would be a way to change system keys in X, or at the system level — that way I can keep compiz. Does anyone have any ideas or know a trick to change system hot keys?