Intel Skylake/Kaby Lake processors: Broken Hyper-Threading?

richierich1212

Platinum Member
Jul 5, 2002
2,662
1
126
#1
Just saw this warning posted across other Tech sites. It says the issue isn't just limited to Linux:

[WARNING] Intel Skylake/Kaby Lake processors: broken hyper-threading
https://lists.debian.org/debian-devel/2017/06/msg00308.html


"This warning advisory is relevant for users of systems with the Intel
processors code-named "Skylake" and "Kaby Lake". These are: the 6th and
7th generation Intel Core processors (desktop, embedded, mobile and
HEDT), their related server processors (such as Xeon v5 and Xeon v6), as
well as select Intel Pentium processor models.

TL;DR: unfixed Skylake and Kaby Lake processors could, in some
situations, dangerously misbehave when hyper-threading is enabled.
Disable hyper-threading immediately in BIOS/UEFI to work around the
problem. Read this advisory for instructions about an Intel-provided
fix.

SO, WHAT IS THIS ALL ABOUT?
---------------------------

This advisory is about a processor/microcode defect recently identified
on Intel Skylake and Intel Kaby Lake processors with hyper-threading
enabled. This defect can, when triggered, cause unpredictable system
behavior: it could cause spurious errors, such as application and system
misbehavior, data corruption, and data loss.


It was brought to the attention of the Debian project that this defect
is known to directly affect some Debian stable users (refer to the end
of this advisory for details), thus this advisory.

Please note that the defect can potentially affect any operating system
(it is not restricted to Debian, and it is not restricted to Linux-based
systems). It can be either avoided (by disabling hyper-threading), or
fixed (by updating the processor microcode).

Due to the difficult detection of potentially affected software, and the
unpredictable nature of the defect, all users of the affected Intel
processors are strongly urged to take action as recommended by this advisory."



So is this something that people with Skylake/Kabylake CPUs should really worry about? Seems like it isn't that huge of a deal if it took this long to discover.
 

WhoBeDaPlaya

Diamond Member
Sep 15, 2000
7,367
2
126
jwtioh.bluesonic.net
#2
Some functional verification guys at Intel are going to get an a*s-whippin' come Monday...
 
Last edited:

Ajay

Diamond Member
Jan 8, 2001
5,240
240
136
#3
There is already a microcode fix. Hopefully mobo vendors will get this out quickly before ppl panic.
 

JoeRambo

Senior member
Jun 13, 2013
689
73
136
#4
There is already a microcode fix. Hopefully mobo vendors will get this out quickly before ppl panic.
Is there a need to panic at all tho? The conditions described are pretty much "antipattern" of programming - mixing usage of low 8 bit part of register access with full register access ( that is for example AH + EAH or RAX ). Those carry hefty performance penalties and are avoided in compiler generated and hand written asm code. So that relegates this bug to once in a blue moon category.

And these days Intel/AMD microcode on Windows is distributed by Windows Update, there is very high chance we all have required microcode fix already.
 

Ajay

Diamond Member
Jan 8, 2001
5,240
240
136
#5
No, nobody with a technical background needs to panic, but how long before this shows up on Semiaccurate and then a million clickbait sites?
 

JoeRambo

Senior member
Jun 13, 2013
689
73
136
#6
There has been similar and even more nasty errata in the past. Heck Intel even disabled TSX on Haswell stuff. This is as rare as hen teeth as it takes effort to generate such instructions, guys in original report "massaged" gcc to do so.

But yeah, "social" click-bait media is a factor, just ask Mrs Clinton.
 

Edrick

Golden Member
Feb 18, 2010
1,883
2
106
#7
Turn HT off anyways. Too much extra heat and voltage for minimal gain.
 

ZGR

Golden Member
Oct 26, 2012
1,827
42
126
#9
This appears to be a patchable fix at least. So no big deal right? :D
 

USER8000

Golden Member
Jun 23, 2012
1,493
3
136
#13
It just shows you that Intel after over a decade of using HT still can have some issues,so people jumping on AMD having some issues in games scaling at launch with their SMT implementation were being a bit unfair!! :p
 
Aug 10, 2009
11,951
0
106
#16
It just shows you that Intel after over a decade of using HT still can have some issues,so people jumping on AMD having some issues in games scaling at launch with their SMT implementation were being a bit unfair!! :p
Well, this is an Intel thread. So I'll bite my tongue. Hopefully you aren't too surprised by the IDF (Intel defense force) though. :)
 

NTMBK

Diamond Member
Nov 14, 2011
8,321
315
126
#17
It's taken this long since the release of Skylake for the bug to be discovered, so it's pretty obviously not a massive deal. Hopefully we'll get fixed microcode soon.
 

Nothingness

Golden Member
Jul 3, 2013
1,905
35
106
#19
No, nobody with a technical background needs to panic, but how long before this shows up on Semiaccurate and then a million clickbait sites?
I have a technical background, and I consider this kind of bug as a big deal. I know Intel does a very good job at validation so this likely shows they went quite far in optimizing some cases and they failed at validating it. Chips have become incredibly complex and that is an issue in general :)
 

Torn Mind

Diamond Member
Nov 25, 2012
3,246
0
91
#21
More like Debian is once again 5 years behind the times and is not a worthy desktop OS for brand new hardware.
 

WhoBeDaPlaya

Diamond Member
Sep 15, 2000
7,367
2
126
jwtioh.bluesonic.net
#23
Quoting from a comment on /. (no idea as to the veracity - don't have time to check up on it @ work)
>>>
Apparently, the fix works only for some models of Skylake (models 78 and 94, stepping 3).
On any other Skylakes and all Kaby Lakes there's no way other than disabling hyperthreading entirely.
A fix might or might not be released in the future, Intel doesn't say a word about the issue.
<<<
 

R0H1T

Platinum Member
Jan 12, 2013
2,559
8
106
#24
Quoting from a comment on /. (no idea as to the veracity - don't have time to check up on it @ work)
>>>
Apparently, the fix works only for some models of Skylake (models 78 and 94, stepping 3).
On any other Skylakes and all Kaby Lakes there's no way other than disabling hyperthreading entirely.
A fix might or might not be released in the future, Intel doesn't say a word about the issue.
<<<
Who's to say that this bug wasn't baked in at the behest of you know who :eek:
 


ASK THE COMMUNITY

TRENDING THREADS