MDS: The Newest Speculative Execution Side-Channel Vulnerability

[Kimmono]

[QUOTE=dungeon;n1098876]

:~$ cat /sys/devices/system/cpu/vulnerabilities/mds
[B]Not affected[/B]

Is that a thing? Something you have installed?

[wpupkin]

well we have a new kernel now. does anybody know how to disable the mitigation?

It looks like they backported

mitigations=

option to longterm and stable kernel versions.

https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?h=v4.19.43&id=8cb932aca5d6728661a24eaecead9a34329903ff

[chithanh]

:~$ cat /sys/devices/system/cpu/vulnerabilities/mds
[B]Not affected[/B]

Is that a thing? Something you have installed?

If you upgrade to kernel 5.1.2 (or one of the stable series 5.0.16, 4.9.43, etc.) then you will get /sys/devices/system/cpu/vulnerabilities/mds
And "Not affected" are all CPUs that have NO_MDS flag set according to arch/x86/kernel/cpu/common.c

[flower]

It looks like they backported

mitigations=

option to longterm and stable kernel versions.

https://git.kernel.org/pub/scm/linux...ad9a34329903ff

sadly this does nothing atm. from your link:

Currently, these options are placeholders which don't actually do anything. They will be fleshed out in upcoming patches.

[RealNC]

If this goes on, Michael will need to change his benchmarking suite at some point to include the Commodore 64.

[wpupkin]

sadly this does nothing atm. from your link:
...

I picked a first patch in series, there are "upcoming patches": https://git.kernel.org/pub/scm/linux...og/?h=v4.19.43
But, tbh, not tested this yet.

[chithanh]

I don't understand how can a mitigation be implemented at all. The only option is to disable Hyperthreading.

No. Disabling HyperThreading is one mitigation which prevents data leaking from trusted to untrusted processes/VMs. It does however not prevent data leaking from kernel to userspace (cf. https://www.cyberus-technology.de/po...ion-techniques).

These CPU manufacturers are really shady. I mean, letting these kind of bugs slip through is utterly irresponsible and shamefull. How could they not see it? It looks like greed to the nth power to me.

It is actually much worse. Intel knew about this bug for a year (first report in June 2018), but said nothing and continued to sell vulnerable CPUs to unsuspecting customers.
During that time, several researchers independently rediscovered these vulnerabilities. Malicious actors could have, too.

Multiple teams and researchers independently discovered, studied, and reported MDS-class vulnerabilities and attacks to Intel. Unfortunately, the different timelines (see below), as well as the fine-grained vulnerability classification and embargo strategy enforced by Intel (which coordinated the disclosure process) made proper coordination across teams impossible. The year-long disclosure process (the longest to date) ultimately resulted in independent finders of even closely related MDS-class vulnerabilities to be completely unaware of one another until a few days before the May 14 disclosure date. We worked with Intel to make sure we'd give proper credit to everyone involved in the disclosure of these vulnerabilities, both in our papers and on this website. This website is our last-minute effort to coordinate the different teams, presenting the two independent papers (and counting?) on RIDL and Fallout.

[Cape]

Jesus Christ....

The FLOSS community must step up the hardware game badly...

We need RiscV fully functional ASAP.

[curaga]

POWER and AMD unaffected, and of course if you only run FOSS...

[Cattus_D]

Assuming that AMD CPUs really have fewer vulnerabilities than Intel processors, I wonder how the performance of older Intel CPUs compares with that of AMD CPUs of the same generation. I.e., are they still faster or have these mitigations affected them so much that their performance is now lower than, or on a par with, that of the AMD offerings that were available in the same time frame?