[Info-vax] Intel junk...Kernel-memory-leaking Intel processor design flaw forces Linux, Windows redesign
Johnny Billquist
bqt at softjar.se
Sat Jan 6 23:23:23 EST 2018
On 2018-01-07 01:28, Jan-Erik Soderholm wrote:
> Den 2018-01-06 kl. 23:34, skrev Tim Streater:
>> In article <p2rh0g$876$1 at Iltempo.Update.UU.SE>, Johnny Billquist
>> <bqt at softjar.se> wrote:
>>
>>> On 2018-01-06 19:23, Tim Streater wrote:
>>>> In article <p2qtjk$ul0$1 at Iltempo.Update.UU.SE>, Johnny Billquist
>>>> <bqt at softjar.se> wrote:
>>>>
>>>>> And then they figured out a clever way of mining the contents of
>>>>> the cache.
>>>>>
>>>>> One could argue that the cache should be invalidated in such a
>>>>> scenario, but that is not happening either.
>>>>
>>>> Never mind invalidating it. WTF is going on if a non-priv process has
>>>> the right to do anything at all to the cache? Non-priv processes
>>>> shouldn't even be aware that there *is* a cache, never mind having the
>>>> right to execute instructions *about* the cache.
>>>
>>> Normally, that is true. But clever people can do a lot around this.
>>> When I was doing my CS major, we had a course on advanced computer
>>> architectures, in where we learned how to write a very simple program
>>> that told us everything about cache size, associativeness, line size,
>>> TLB size, TLB associativeness, and so on...
>>> All you need to do is understand how the computer is affected by
>>> these things, and then write programs that detect the effects.
>>>
>>> In short, you write small loops that exercise the cache in different
>>> ways, and time the whole thing. You don't even need any high
>>> precision timers for it. All user level, and all very simple.
>>
>> Does that involve instructions that operate on the cache. Such as
>> "clear cache"?
>>
>
> You just read some unrelated (to the actual tests) data, so that the
> test that you are running are 100% non-chached. If that is what your
> tests are about.
Right.
> And anyway, you can never "clear" any memory, being it the cache or any
> other memory. Each byte will always have a value between x'00' and x'FF'.
True for normal memory, but not for caches. Caches actually have a
separate bit saying if the content is valid or not.
Johnny
--
Johnny Billquist || "I'm on a bus
|| on a psychedelic trip
email: bqt at softjar.se || Reading murder books
pdp is alive! || tryin' to stay hip" - B. Idol
More information about the Info-vax
mailing list