[Info-vax] Intel junk...Kernel-memory-leaking Intel processor design flaw forces Linux, Windows redesign

Wolf K wolfmac at sympatico.ca
Fri Jan 5 10:51:31 EST 2018


On 2018-01-05 09:22, Jan-Erik Soderholm wrote:

[snip remarks about probable performance hits]]
> As I understand, as in Linux, the kernel memeory is mapped into each user
> process memory space (for performance reasons). The speculative fetch done
> by the hardware can read kernel memory directly. And when the protection
> schemas detects this, the data is already in the internal CPU cache.
> 
> The solution seems to be to separate kernel and user memoery into separate
> virtual memory areas. So a re-mapp of the memory mapping is needed each
> time the process needs to read kernel memory, and that adds a perf cost.
> 
> And yes, it looks like different "levels" in the hardware are bit
> out of sync...

So, in order to reduce the performance hit, would it make sense to 
redesign the CPU with a larger on-board cache to store both kernel and 
user memory? Or, what am I missing in the protected memory concept?

Anyhow, I think most users will see no performance hit. I mean, how many 
people are rendering CGI on their laptops? Etc.

I'm more worried about server farms used by big data, banks, ISPs etc. 
These already show performance hits tied to time-of-day, as user access 
(ie demand) varies. Even a few % slowdown in overall throughput will be 
noticeable at peak demand times.

-- 
Wolf K
kirkwood40.blogspot.com
"The next conference for the time travel design team will be held two 
weeks ago."



More information about the Info-vax mailing list