Some serious security bug in INTEL CPUs?? Since Westmere possibly

[hikaruichijo]

Re: Some serious security bug in INTEL CPUs?? Since Westmere possibly

Thanks a lot, I'm sorry. My fault.

[Per Hansson]

Re: Some serious security bug in INTEL CPUs?? Since Westmere possibly

The fun continues with BranchScope Side-channel CPU Vulnerability:

https://cdn.badcaps-static.com/pdfs/...10593eb6a9.pdf

[bigbeark]

Re: Some serious security bug in INTEL CPUs?? Since Westmere possibly

The fun continues with BranchScope Side-channel CPU Vulnerability:

https://cdn.badcaps-static.com/pdfs/...10593eb6a9.pdf

Truly mind-boggling complexity - turned my brain to jello!

[stj]

Re: Some serious security bug in INTEL CPUs?? Since Westmere possibly

i learned to never trust Intel after the "floating point bug incident".

for those too young,
Intel found / was told of a bug in the floating point unit.
the company response to the customers was:

high precision floating point math will only effect scientists and similar people - so live with it!

"scientists" ??, yea and anybody else doing long floating point math, like AutoCad users and 3D gamers.
in other words - everybody!

[Curious.George]

Re: Some serious security bug in INTEL CPUs?? Since Westmere possibly

i learned to never trust Intel after the "floating point bug incident".

for those too young,
Intel found / was told of a bug in the floating point unit.
the company response to the customers was:

"scientists" ??, yea and anybody else doing long floating point math, like AutoCad users and 3D gamers.
in other words - everybody!

Intel has historically been reluctant to "admit" problems. And, liked to spin problems as "features". Or, as "obvious consequences of the implementation" (yeah, but the implementation details of the chip are ONLY KNOWN TO INTEL INSIDERS!)

By contrast, on my first NS16032 design, the folks from National showed up with a STACK of "errata" (bug list). Instead of being intimidated ("OhMiGosh! What sort of crappy processor IS this??"), it was refreshingly honest: they knew of the problems, didn't choose to hide them and, presumably, were addressing workarounds.

"OK, I can work with that..."

[stj]

Re: Some serious security bug in INTEL CPUs?? Since Westmere possibly

yea, i allways liked nat semi'
they sent me entire databooks with a single phonecall, all TI sent for years was an order-form with prices.
i dont pay for design-data, so i didnt use TI chips!

[Curious.George]

Re: Some serious security bug in INTEL CPUs?? Since Westmere possibly

yea, i allways liked nat semi'
they sent me entire databooks with a single phonecall, all TI sent for years was an order-form with prices.
i dont pay for design-data, so i didnt use TI chips!

When I lived in the midwest, I was reasonably friendly with the regional TI rep. So, "getting things" (databooks, samples -- even access to inside technical people) was relatively easy.

OTOH, once I moved away and had to deal with the "locals", things went downhill. I finally gave up on TI when they couldn't figure out how to get me off their "junk email" list (I simply closed the email account and cut them off, entirely!)

[retiredcaps]

Re: Some serious security bug in INTEL CPUs?? Since Westmere possibly

Looks like Wolfdale will never get any updates. Intel updated their guidance April 2. Sounds like a bunch of BS to me for the reasons.

[Per Hansson]

Re: Some serious security bug in INTEL CPUs?? Since Westmere possibly

Aye, bullshit it is
https://www.techspot.com/news/74014-...rocessors.html

While it's disappointing that Intel is leaving some legacy users out in the cold, it's not entirely surprising. As Tom's Hardware highlights, “the real reason Intel gave up on patching these systems seems to be that neither motherboard makers nor Microsoft may be willing to update systems sold a decade ago.”

[Behemot]

Re: Some serious security bug in INTEL CPUs?? Since Westmere possibly

Still could work on it on the SW level I guess. Though some ppl reported signifficant performance los on Core 2 Quads with the SW patches.

[mockingbird]

Re: Some serious security bug in INTEL CPUs?? Since Westmere possibly

Looks like Wolfdale will never get any updates. Intel updated their guidance April 2. Sounds like a bunch of BS to me for the reasons.

Just saw on Gamer Nexus, AMD is only releasing microcode updates from Bulldozer and up.

[stj]

Re: Some serious security bug in INTEL CPUs?? Since Westmere possibly

yes, but amd is not so badly effected.
nice try though!

[mockingbird]

Re: Some serious security bug in INTEL CPUs?? Since Westmere possibly

yes, but amd is not so badly effected.
nice try though!

If they're not affected, then why are they bothering to release a microcode update?

[mariushm]

Re: Some serious security bug in INTEL CPUs?? Since Westmere possibly

They're not as badly affected as Intel, but they're affected.
The CPUs are not vulnerable to all the issues Intel cpus are vulnerable to due to different design choices, and on those that affect them, the results are less severe.

See https://www.amd.com/en/corporate/sec...ragraph-290416

GPZ Variant 1 (Spectre) mitigation is provided through operating system updates that were made available previously by AMD ecosystem partners. GPZ Variant 3 (Meltdown) does not apply to AMD because of our processor design.

While we believe it is difficult to exploit Variant 2 on AMD processors, we actively worked with our customers and partners to deploy the above described combination of operating system patches and microcode updates for AMD processors to further mitigate the risk. A whitepaper detailing the AMD recommended mitigation for Windows is available, as well as links to ecosystem resources for the latest updates.


And for those other exploits reported by CTS Labs Research... I think there's already fixes for those even though all require administrative access (basically your system's fuxed already) : https://community.amd.com/community/...-labs-research

[mockingbird]

Re: Some serious security bug in INTEL CPUs?? Since Westmere possibly

They're not as badly affected as Intel, but they're affected.

Thanks for stating what is obvious to everyone... I'm still uisng my 45nm Intel system with XP, and I'll let you know if I get a rootkit while browsing the internet.

[Behemot]

Re: Some serious security bug in INTEL CPUs?? Since Westmere possibly

I wonder what the real-world risc is with XP or Vista? Both are unsupported now and won't be getting any update. They should run much faster than 7 with SW patches though. I've just recently run tests on a rig donated to ball tombola (to show how capable of such system still is) and the Q8400 had no troubles playing 4k teaser on a display with presentation on a projector. The whole thing (without the projector though) drew up to 150 W while doing that.

BTW - the screwed-up windblows update for Vista is something, I'll tell ya. Close to 3 days was still not enough for the thing to find updates. Think a customised image with at least 100 of the updates already in it will be necessary for further installations, that should take the time down to about 1-2 days.

[mockingbird]

Re: Some serious security bug in INTEL CPUs?? Since Westmere possibly

I wonder what the real-world risc is with XP or Vista? Both are unsupported now and won't be getting any update.

XP is still supported via PSOReady 2009 updates until April 2019.

Also:

1) You can still run pretty much run any new Intel CPU on XP. AHCI drivers are available even up to the latest generation (Don't quote me on that). So if you have a BIOS with the updated microcode, the risk is zero.
2) I'm pretty certain MS has already released the OS-level microcode update for XP.

I'm on XP64 though (not supported) with an undervolted, underclocked Wolfdale-DP, so I am wide open.

[Topcat]

Re: Some serious security bug in INTEL CPUs?? Since Westmere possibly

My poor harpertowns & westmeres....ohh well, unpatched and happy. Nothing has exploded.

[Behemot]

Re: Some serious security bug in INTEL CPUs?? Since Westmere possibly

XP is still supported via PSOReady 2009 updates until April 2019.

Also:

1) You can still run pretty much run any new Intel CPU on XP. AHCI drivers are available even up to the latest generation (Don't quote me on that). So if you have a BIOS with the updated microcode, the risk is zero.
2) I'm pretty certain MS has already released the OS-level microcode update for XP.

I'm on XP64 though (not supported) with an undervolted, underclocked Wolfdale-DP, so I am wide open.

M$ stated XP and Vista won't be getting the updates, nor will any of the systems based on that (PoS for example) because of kernel architecture and BS. Running XP or Vista on newer HW gives me NO benefit whatsoever, the focus is on old HW (Pentium D-C2Q) which 1st comes with XP/Vista CoAs, and 2nd may actually run (much) better with these systems.

But what I mean is, most of the most-used apps like web browsers got updates to mitigate it in the application itself. Besides that, it has been stated that the older the CPU architecture, the more complicated it actually is to gain anything with speculative execution as, well, they do not speculate so much, don't have all the instructions (we are already talking hundrends of new instructions between Wolfdale and say Ivy Bridge) and are generally slower in these things so it is quite challenging to bypass the system. Am I right here?

Cause what I also mean, I got tens of rigs with Vista (Business mostly) CoAs. If we ignore the broken update application crap, the systems will run happily most of current SW still, and actually faster than with 7. Yep. it is ironic, but because of 7 getting the patches, Vista will likely be faster from now on. So I can install original Vista (and activate via phone at the worst) with some (how great?) a risc, or pirated 7 without the risc. I do have chinese CoAs for 7 Pro too, but most of the folk wont be willing to pay extra 25 bucks for it.

[Hondaman]

Re: Some serious security bug in INTEL CPUs?? Since Westmere possibly

Close to 3 days was still not enough for the thing to find updates.

Last autumn I had the same problem with Win7. There are a few key patches that may bring your looooooong wait down to mere minutes. I googled it, downloaded a couple of updates, then started the Windows Update applet, and minutes later I had all the updates in a nice list, ready for my review.

Are there key patches that reduce the wait with Vista like they do with Seven?