1 |
On 02/01/18 22:58, Adam Carter wrote: |
2 |
> AMD coder's patch to disable the new code (to avoid the performance hit) |
3 |
> where he states the issue doesnt exist on AMD processors; |
4 |
> https://lkml.org/lkml/2017/12/27/2 |
5 |
|
6 |
Read LWN, specifically the links to the people who covered the bug. |
7 |
|
8 |
It's a flaw in speculative forward processing, where the security does |
9 |
not travel with the speculative processing. So user code can trigger a |
10 |
page fault that references kernel code, causing that page to be |
11 |
retrieved. OOPPSSSS. AMD keeps security context with the code, causing |
12 |
an attempt to exploit the bug to fail with "invalid security context". |
13 |
|
14 |
And as I understand it the code can be disabled with either a compile |
15 |
time option or command line switch to the kernel. The relevant code is |
16 |
called KAISER, which forces kernel and user address space into different |
17 |
contexts, and causes a nasty context-switching overhead on both Intel |
18 |
and AMD cpus. |
19 |
|
20 |
Cheers, |
21 |
Wol |