Quantcast
[ 3 / biz / cgl / ck / diy / fa / g / ic / jp / lit / sci / tg / vr / vt ] [ index / top / reports / report a bug ] [ 4plebs / archived.moe / rbt ]

Due to resource constraints, /g/ and /tg/ will no longer be archived or available. Other archivers continue to archive these boards.Become a Patron!

/g/ - Technology


View post   

[ Toggle deleted replies ]
File: 296 KB, 695x707, Intel Bug.png [View same] [iqdb] [saucenao] [google] [report]
64109858 No.64109858 [Reply] [Original] [archived.moe] [rbt]

INTEL IS FINISHED

https://www.theregister.co.uk/2018/01/02/intel_cpu_design_flaw/

>> No.64109883

>>64109858
AMD guy myself, but all I see is intel gaining 30% more performance in some cases in future

>> No.64109889

>Another exploit that requires 90 other layers of incredibly specific access that makes the exploit moot in the first place
Ah yes, if the attacker has physical access to my box with root access and I'm running this particular build of the OS with this specific patch missing and this specific build of Adobe Acrobat installed, it's all over!

>> No.64109892

>>64109858
>if true
>olly place reporting is a tech opinion website from the uk
I have a hard time beleiving this. When you google it the first result is reddit using 4chan as a source. Hoax?

>> No.64109905

>>64109883

Wat?

>> No.64109954

>>64109889
>incredibly specific access
Yes. Such as running javascript in your browser.
Or deploying a docker container to the same physical machine as your container is running on.

>> No.64109989

>>64109892
"that intel bug is definitely sweating a bunch of people i know who work in ops

basically there is an issue with intel CPUs and VRAM that leads to protected pages of memory becoming accessible in userspace. it's both a HUGE security issue and something that is so fundamental to OSes that requires a massive rework of the VRAM systems on all the OSes. the patch that will need to go in is apparently so significant that it'll cause a 5-30% performance drop on Linux."

>> No.64110135

>>64109989
Yeah i read that. What i am curious of is its legitimacy. Some unknown tech site, 4chan, and reddit doesnt inspire confidence. /g/ is overrun with amd shills so i trust this place the least

>> No.64110146

>>64109889
Apart from it can be exploited by javascript and it's serious enough they're giving everyone a 5-30% performance penalty solely in order to address this issue.

>> No.64110213

>>64110135
>theregister
>unknown tech site

>> No.64110223

>>64109892
>I have a hard time beleiving this.
Of course you do, little Intel fanboi.

>> No.64110251

>>64109858
its over, amd won

>> No.64110263

>>64109308
>>64109308
>>64109308

>> No.64110319

>>64109858
IT'S NOT A BUG GOYIM

IT'S A FEATURE

>> No.64110388

>>64110223
Screenshotted to laugh at you later when its confirmed a hoax

>> No.64110402

>>64110213
Is that where bongs get their tech news?

>> No.64110533

>>64109858
Will this effect my CPU?

I've got Intel Core i3-4005U and its x64 based.

>> No.64110553

>>64110533
Yes, all Core i3 models are affected.

>> No.64110583

>>64110533
Every intel CPU made in the past 10 years is what I read...I think

>> No.64110623

>>64110402
Nigga you retarded

>> No.64110659

>>64109883
>>64109889
>>64109892
>>64110135
>>64110388
Better sell some more stock, Brian.

>> No.64110689

>>64110623
Okay keep pretending then

>> No.64110767

>>64110553
Fuck. I give up with Intel, I'm just going to sell this laptop and buy something else. Can anybody recommend a good cheap laptop that I could buy which wouldn't be affected by this? Preferably available in the UK

>> No.64110783

>>64110767
Lenovo Thinkpad A470.

>> No.64110796

>>64110783

Just don't install Ubuntu 17.10 on it.

>> No.64110816

>>64110796
You shouldn't install Ubuntu in the first place.

>> No.64110890

>>64110783
First page of Google brings up a shit load of replacement screens and one laptop for over 800 USD, hardly cheap lol. I only want it for YouTube and Facebook so doesn't need to be anything special but just not fucked by this bug

>> No.64110901

>>64110816
u r right, u should use gentoo

>> No.64110917
File: 350 KB, 672x794, 1502055422354.png [View same] [iqdb] [saucenao] [google] [report]
64110917

NOOOOOOO! IT ISN'T TRUE! THIS IS AN AMD PLOT!

>> No.64110927

>>64110890
Get a core 2 duo macbook, later models. Those are just fine for YT/FB and shouldn't be affected.

>> No.64110951

This is fake af.
If im wrong show some proofs

>> No.64111007

>>64110927
>shouldn't
Not very reassuring lol

I don't understand why a list of affected and unaffected CPUs hasn't been compiled yet, the register says winblows is going to release a fix next week so I might try it and see how badly processes get slowed down, although I rarely seem to receive any updates on 8.1

>> No.64111014

>>64110890
In that case buy a mid-2009 Macbook. Can find one on Amazon for like $105 max, and they're not affected by this bug. Can upgrade to El Capitan, too.

>> No.64111031

>>64110927
>core 2 duo
most recent comments on the various forums seem to indicate that everything from Core and up is affected, so all processors from 2006 and onwards.

>> No.64111037

>>64111007
Intel has embargoed flaw details. What we know comes almost entirely from people actually installing the patch and testing, or from trying to decipher the patch source code.

>>64110951
https://lkml.org/lkml/2018/1/2/703

>> No.64111055

>>64111031
The "Intel Core" series does not include the original Core Solo/Duo and Core 2 Solo/Duo/Quad as a marketing umbrella. I've not seen anything that *confirms* that pre-Nehalem is affected, although it's been speculated that processors as old as the Pentium II might be.

>> No.64111057

>>64111031
/g/ was right about 90s thinkpads all along....

>> No.64111072

>>64111037
Appreciate it :3

>> No.64111108
File: 95 KB, 1024x576, stallman.jpg [View same] [iqdb] [saucenao] [google] [report]
64111108

>>64111057

>> No.64111110

>>64111057
Fuck it man, if Pentium Pros are vulnerable to this too then the ONLY solution will be going back to PowerPC Macbooks.

>> No.64111129

>>64111110
or you can just buy AMD like a normal person

>> No.64111143

>>64111110
>>64111111
So fucking close to making a prophecy. Damnit

>> No.64111186

>>64110767
>I give up with Intel

So the fuck what?
It's not like you were on their team or someshit, you happen to own a computer with a CPU they made.

Quite acting like anything you do is relevant enough that someone would want to hack you anyways.

>> No.64111203

>>64111129
Well, yeah. Here's to hoping Lenovo drops an A1 Carbon with a touch screen.

>> No.64111214

>>64111186
Suck my gooch you Intel shillbot

>> No.64111278
File: 97 KB, 800x480, idiot-mud-girl.jpg [View same] [iqdb] [saucenao] [google] [report]
64111278

>>64111214
You should go back to your Ipad like the retarded faggot you are. You're so dumb you don't need to worry about anything tech related.

>> No.64111291

>>64109905
>Wat
How long before intel fixes the bug and releases a patch?

>> No.64111321

>>64111055

I've seen some people speculate this affects all processors back to Pentium Pro. It's not just "Core" processors, its everything with ME

>> No.64111331

>>64111186
why are you so insecure

>> No.64111341

Intel is full of diversity and engineering has led by a pajeet for years. Need I say more?

>> No.64111346

>>64111278
Should I install genpoo daddi?

>> No.64111357

>>64111331
He's an Intel processor

>> No.64111370
File: 33 KB, 500x375, you-are-a-fggt.jpg [View same] [iqdb] [saucenao] [google] [report]
64111370

>>64111331
Why are you so fucking dumb? Which one of your ancestors fucked their sister to result in a dumb fuck like you?

>> No.64111377

>>64109858
>google Ryan Shrout
>known amd shill

going all out

>> No.64111403
File: 76 KB, 640x480, 1513846718898.jpg [View same] [iqdb] [saucenao] [google] [report]
64111403

>>64111357

>> No.64111406

>>64111291
i dont know a shit about computers but it seems that linux and windows are patching, if what i read is correct

>> No.64111423

>>64111291
The patch IS what causes the 30% drop in performance. Not having the patch leaves your system wide open to total compromization.

>> No.64111454

>>64111321
It's Virtmem that matters here, not ME, although ME is itself cancerous.

>>64111357
AYYYY

>>64111406
The patch reduces performance by X%. The flaw (not a bug, it's an actual design flaw) is in the hardware of the processors and cannot be ultimately fixed by software patches.

>> No.64111483

why are intels cpus so fucking insecure all of the sudden

>> No.64111492

>>64111454
Thanks for the clarification!!

>> No.64111506
File: 162 KB, 633x900, 1514920622752.png [View same] [iqdb] [saucenao] [google] [report]
64111506

NOOOOOOOOOOOOOOOOOO

>> No.64111536

>>64111483
Intel has been doing mostly the same shit over the past 9 years so even rather old research into Intel's security still applies to newer CPUs, It's just now starting to catch up

>> No.64111546

>>64109858
does this affect my shitty dial core from 2014?

>> No.64111564

>>64111483
They always were and anyone with a brain knew it. Intel has been exercising abysmal business practices since the Pentium 3. What makes you think their vicious cost cutting, relentless gas lighting and anti-competitive bribery/coercion was actually sitting on top of a well designed product?

>> No.64111566
File: 106 KB, 300x449, 1514791552815.jpg [View same] [iqdb] [saucenao] [google] [report]
64111566

So, now what?
As an average home user faggot, do I need to buy a new AMD processor and replace my i7? Or download the patch and take a performance hit?
Or or just do nothing and carry on as usual because this is mostly a concern for large businesses and I know I can trust the handful of pages I visit on a daily basis to not implement a script which can fuck me over?

>> No.64111571
File: 232 KB, 800x449, intel ceo yells at clouds.png [View same] [iqdb] [saucenao] [google] [report]
64111571

>>64111377

>> No.64111574

>>64111546
it is a 1 core now performance wise my friend, welcome back to 2003

>> No.64111588

>>64111546
>2014
yep. Everything back to Nehalem and possibly farther.

>> No.64111596

>>64111566
You're better off patching and keeping up to date.

>> No.64111598

>>64111546
Read the thread this could effect all intel CPUs all the way back to the 90s

>> No.64111600

>>64111566
You're going to get compromised. Just take the hit or buy a Ryzen.

>> No.64111603

>>64111546
Lets put it in a simpler way to stop people from constantly asking this, if you managed to get to /g/ on an intel based system, yes its highly likely it will affect you.

>> No.64111617

>>64111566
Once the details of the vulnerability are released the potential attackers will learn about it. And if it's really Level 3 to Level 0 vulnerability it will mean that you can get fucked up by running Javascript on your browser

>> No.64111619

>>64111598

and it'll be completely patched by tuesday. it's an OS issue

>> No.64111624

>>64111566

windows will have an update that will give you ~5% less performance

>> No.64111636

>>64111619
Its not an OS issue, the OS patch is a crude workaround, hence the potentially huge performance impact.

>> No.64111638

>>64111624
5%-65% less performance. Fixed.

>> No.64111649
File: 75 KB, 419x480, 1485408164323.jpg [View same] [iqdb] [saucenao] [google] [report]
64111649

>>64111357

>> No.64111670

>>64111638

if you have an x86 processor, it'll be 5%

>> No.64111671

>>64111574
>>64111588
>>64111598
>>64111603
>tfw almost fell for the intel meme entirely
cheaped out with a dual core with a possible future upgrade in mind.
still havent upgraded. at least now i can ditch it and build new system.

>> No.64111700

Good thing I shitpost with an 80186

>> No.64111701

>>64111670
Source: Your ass.

>> No.64111713

Thomas Gleixner (25):
x86/cpufeatures: Add X86_BUG_CPU_INSECURE
x86/mm/pti: Add infrastructure for page table isolation
x86/mm/pti: Force entry through trampoline when PTI active
x86/entry: Align entry text section to PMD boundary
x86/mm/pti: Share entry text PMD
x86/cpu_entry_area: Add debugstore entries to cpu_entry_area
x86/mm/dump_pagetables: Check user space page table for WX pages
x86/mm/dump_pagetables: Allow dumping current pagetables
x86/ldt: Make the LDT mapping RO
perf/x86/intel: Plug memory leak in intel_pmu_init()
x86/apic: Switch all APICs to Fixed delivery mode
gpio: brcmstb: Make really use of the new lockdep class
genirq/msi: Handle reactivation only on success
genirq: Introduce IRQD_CAN_RESERVE flag
x86/vector: Use IRQD_CAN_RESERVE flag
genirq/irqdomain: Rename early argument of irq_domain_activate_irq()
genirq/msi, x86/vector: Prevent reservation mode for non maskable MSI
timers: Reinitialize per cpu bases on hotplug
nohz: Prevent a timer interrupt storm in tick_nohz_stop_sched_tick()
timers: Invoke timer_start_debug() where it makes sense
timerqueue: Document return values of timerqueue_add/del()
x86/smpboot: Remove stale TLB flush invocations
x86/mm: Remove preempt_disable/enable() from __native_flush_tlb()
x86/ldt: Plug memory leak in error path
x86/ldt: Make LDT pgtable free conditional

>> No.64111737

>>64111624
I sure hope the update can be downloaded and installed as standalone because I haven't updated Windows in years.

>> No.64111745

>>64109889
>>64109889
You can bet nsa/cia are aware of it and have been using malware that exploits this particular bug.
In fact intel probably told them about it (if it wasnt there intentionally in the first place)
Amd probably leaked to public/ linux devs.


Kek so we will be seeing 30% performance hit. Does it mean intel is now equal to Amd?

>> No.64111747
File: 39 KB, 511x509, ww2.jpg [View same] [iqdb] [saucenao] [google] [report]
64111747

>>64109858
Ok, so microshit will release fix on that next week, and this 30% performance impact is a bullshit. It will mostly affect io operations, such as disk access... but still not -30%.
Good news to gaymers - you will not see any difference in gaymes. So you might want to fuck off already to >>>/v/.
Also saging and hidding this shit thread. AyyMD drones - please get back to sleep.

>> No.64111754

>>64111737
Can't you manually update your drivers?

>> No.64111755

>>64111737
You're already compromised

>> No.64111769
File: 436 KB, 600x580, 1514151062965.gif [View same] [iqdb] [saucenao] [google] [report]
64111769

>not using IBM Power Solutions or Oracle Sparc Solutions
>2018
>using Intel or AMD

>> No.64111772

>>64111713
What did I just read

>> No.64111781

>>64110767
Here tard, this one is safe
https://www.youtube.com/watch?v=ThItRZqFKX4

>> No.64111787

>>64111747
It won't affect games? Are GPUs not controlled through syscalls?

>> No.64111793

>>64111747
>Good news to gaymers - you will not see any difference in gaymes
Whut. dGPU drivers are pure syscalls.

>> No.64111801
File: 421 KB, 497x593, [Erai-raws] New Game!! - 10 [720p][HardShit][484F2738].mkv_snapshot_09.03_[2017.09.12_15.16.13].png [View same] [iqdb] [saucenao] [google] [report]
64111801

>>64111747
>>64111747
>you will not see any difference in gaymes
What a baka.

>> No.64111802

>>64111617
It won't be. It'll have the same risk as a local escalation vulnerability, meaning you need local user rights first to escalate this (so not something a browser could use, but software the user runs could).

>> No.64111807

>>64111747
>AyyMD drones - please get back to sleep.
okay I'll be over here sleeping while you rush out a patch for a bug in every CPU intel built for a decade

>> No.64111809

>>64111745
>Does it mean intel is now equal to Amd?
Intel is now *significantly* below AMD.

>> No.64111812
File: 2.36 MB, 640x640, ExhaustedChillyArabianoryx.webm [View same] [iqdb] [saucenao] [google] [report]
64111812

Main rig: i7-4790k
Weekend rig: i3-2120

How fucked am I? Or will I be fine because they are too old and not effected?

>> No.64111825

This is fucking HUGE

>> No.64111827

>>64111812
100% fucked on both. No joke.

>> No.64111828

>>64111801
I want to fuck Nenecchi.

>> No.64111837

>>64111812
100% fucked. Every processor with 'Core' in the name is affected by this.

>> No.64111845

>>64111787
Unless its amd vs nvidia, latest cod benchmarks, and speeds/overclocking/how many gb of ram I am not interested gtfo from g aka technology board.

>> No.64111849

>>64109858
does this only effect systems with INTEGRATED graphics?

>> No.64111850

>>64111812
They're both affected. We don't know how much it'll hurt your vidya gay men yet

>> No.64111853

>>64111807
>implying AyyMD doesn't produce erratas
brainlet detected. You clearly have no idea how complicated microprocessors are.

>> No.64111856

Is 2018 the year of great things?
>AMD Ryzen refresh
>This
whats next?

>> No.64111866

are you guys retarded? why is anyone "fucked" with a intel cpu? do you realize that this literally means the already best cpu's can potentially be even better with the fix?

>> No.64111867
File: 63 KB, 638x472, 1263280289831.jpg [View same] [iqdb] [saucenao] [google] [report]
64111867

> patch causes a 30% hit performance.

fuck that,
I rather stay like this, I only have porn on my PC anyway.

>> No.64111870

>>64111849
Every fucking Intel processor of the past ten fucking years.

>> No.64111872

>>64111849
It affects all computers that have an intel CPU from ??? to 2017.

>> No.64111891

>>64111872
2006 Core architecture

>> No.64111895

I have 7980XE, lol.

>> No.64111901

>>64111891
Source.

>> No.64111906

>>64111755
Quoth the Slav: "Oh well."
>>64111754
Yeah, Device Manager -> Processors right?

>> No.64111910

>>64111747
this reminds me, can't anyone with an intel processor and linux apply the patch and test how it affects gpu performance? at least this will settle those endless questions.

>> No.64111912

>>64111895
>being the literal definition of cuckold

>> No.64111918

>>64110583
>>64110553
>>64110533
All of the Intel iCore (i3/5/7/9/Xeons) lineup since 2010 (Westmere) is affected.

>> No.64111920

>>64111866
>better
>flushing the TLA cache on every syscall
The absolute state of Intel fanboys.

>> No.64111922
File: 101 KB, 740x416, Intel-AMD-Naples-Reply-2-1080.1870784274-740x416.png [View same] [iqdb] [saucenao] [google] [report]
64111922

Does anyone still have the scosystem quote?

>> No.64111923

>>64111866
No. It's a hardware flaw making the CPUs insecure.
OSes (Linux, windows) are patching it and the crude patch will be a 5-60% performance hit.

>> No.64111929

>>64111866
How is a 30% drop in performance a 'better'?
Because that's what the patch does. Reduces performance by as much as 30%.

>> No.64111930

I'm not very good with actual computer hardware, is there a quick rundown for brainlet.

>> No.64111943

whoops our hardware backdoors are exposed

funny how we are learning about all these backdoors after Trump is elected

>> No.64111947
File: 2.85 MB, 640x480, Final Destination.webm [View same] [iqdb] [saucenao] [google] [report]
64111947

>>64111837
>>64111827
>>64111850
Well god damn fucking shit. I already fell for the GTX970 meme and now I get fucked by Intel too.

I am so sick of this shit.

>> No.64111949

>>64111930
Intel is L I T E R A L L Y on suicide watch.
It's not too late to buy a mommy Su portrait!

>> No.64111957

>>64111930
Yes you either install a patch that drops performance between 5% to 60% depending on the workload or get buttfucked by javascript browser exploits

>> No.64111976

>>64111930
Basically you now have to choose between opening your computer wide for anyone on the internet to take full control of it, or a 30% performance drop.
But ONLY if you have an Intel CPU.

>> No.64111979

>>64111957
Well, fuck. Class Action Suit /when/

>> No.64111989
File: 50 KB, 645x729, 1484445291877.png [View same] [iqdb] [saucenao] [google] [report]
64111989

>>64111930

here it is from another brainlet

>inlel has had a hardware bug on their processors for the the past few millenia
>someone fucking finds out
>it's a huge security issue
>a patch is about to hit that's gonna reduce performance of all inlel based systems newer then 2010

>> No.64111991

>>64111947
>Well god damn fucking shit. I already fell for the GTX970 meme and now I get fucked by Intel too.
>I am so sick of this shit.
you did not listen, /g/ always said buy AMD

>> No.64112003

>>64111989
if Intel knew about this beforehand and hid it, can't people in U.S. sue the fucking shit out of them

>> No.64112004

>>64111979
It ain't gonna happen. You can't prove malignance; they fucked up big time, sure, but it wasn't on purpose.

>> No.64112006
File: 659 KB, 800x450, lewd_pasta.png [View same] [iqdb] [saucenao] [google] [report]
64112006

>>64111910
initial benchmarks related to file operations
https://www.phoronix.com/scan.php?page=article&item=linux-415-x86pti&num=2

>> No.64112009
File: 740 KB, 1102x1129, 1503233095843.jpg [View same] [iqdb] [saucenao] [google] [report]
64112009

>>64111357

>> No.64112018

>At one point, Forcefully Unmap Complete Kernel With Interrupt Trampolines, aka FUCKWIT

https://www.theregister.co.uk/2018/01/02/intel_cpu_design_flaw/

is it 1st of april already

>> No.64112024

>>64111976
I have three Intel CPUs, lol.

>> No.64112025

>>64112003
Intel *did* know about this beforehand...for roughly a month. They didn't know about the exploits (based on currently available information) when actually designing the CPUs.

>> No.64112026

>>64111891
>since 2006

>>64111918
>since 2010

Make up your minds, people are throwing around random numbers. I hope the embargo ends soon so we can all hear from Intel themselves about just how much lube shall be required

>> No.64112027

So is isolating the kernel page table(s) the only solution? Because that's a HUGE performance impact when you're syscall/IO-bound. Consider the difference between:
- syscall to self/standard entry point, do stuff, if sensitive data: clear cache entries, return
- trampoline with kernel trap, swap page table(s), flush TLB, do stuff, jump back, swap page table(s), flush TLB, return

>> No.64112030

>>64111930
Intel fucked up.
OS devs have to fix it in the short term with a performance crippling patch.
Intel will see no repercussions for this despite destroying years worth of optimizations in a single fuck up.
Intel will buy silence.
AMD will not step up and take over market share.
The status quo will continue because everyone is a fucking retard and the CPU "market" is dead.

>> No.64112039

>>64109858

>30% performance hit
>tfw I bought 20 shares of AMD early Dec

Here we go bros!!!! Blow and hoes for the whole board.

>> No.64112041
File: 33 KB, 620x670, 1487425630258.jpg [View same] [iqdb] [saucenao] [google] [report]
64112041

>>64111357

>> No.64112043

>>64112018
Engineers/Hackers always hid jokes and what not in code

>> No.64112044

>>64109883
>>64109905
>>64109858
>not intel can show a 30 percent gain when they fix it hardware side, and all their marketing wank makes people forget about the bug.

>> No.64112046

>>64112027
>So is isolating the kernel page table(s) the only solution?
Yes.
The hardware is straight-up --flawed--.

>> No.64112058
File: 466 KB, 500x375, giphy.gif [View same] [iqdb] [saucenao] [google] [report]
64112058

>>64110767
>I'm just going to sell this laptop and buy something else.
Ain't no nigga gonna buy that shit after THIS week!

>> No.64112061

I spent about 3400 euros to Intel CPUs in the past six month. Guess I'm on a suicide watch.

Wev lads.

>> No.64112062

>More recent Intel chips have features – specifically, PCID – to reduce the performance hit.
Hmmm

>> No.64112077

>>64112044
You aren't reading it right

the PATCH TO FIX IT causes a 30% performance loss

>> No.64112078
File: 45 KB, 640x480, 1401426740999.jpg [View same] [iqdb] [saucenao] [google] [report]
64112078

> tfw you just dropped 400 bucks on an i7 about 3 days ago and now hear about this.

fucking Intel

>> No.64112085

>>64112006
>https://www.phoronix.com/scan.php?page=article&item=linux-415-x86pti&num=2
This is just disk access WTF!

M.2 NVME is halved in speed, FS-Mark 54% performance drop on i7-8700k

>> No.64112087

>>64112046
Elaborate for someone who doesn't do OS development. Assuming the hardware is flawed, why is this the only solution?

>> No.64112090

>>64112026
>I hope the embargo ends soon
Day after tomorrow, iirc.

>> No.64112094

This series is a major overhaul of the KAISER patches:

1) Entry code

Mostly the same, except for a handful of fixlets and delta
improvements folded into the corresponding patches

New: Map TSS read only into the user space visible mapping

This is 64bit only, as 32bit needs the TSS mapped RW

AMD confirmed that there is no issue with that. It would be nice to
get confirmation from Intel as well.

2) Namespace

Several people including Linus requested to change the KAISER name.

We came up with a list of technically correct acronyms:

User Address Space Separation, prefix uass_

Forcefully Unmap Complete Kernel With Interrupt Trampolines, prefix fuckwit_

but we are politically correct people so we settled for

Kernel Page Table Isolation, prefix kpti_

Linus, your call :)

3) The actual isolation patches

- Replaced the magic kaiser_add/remove_mapping() code by mapping everything
which needs to be shared with user space into the fixmap

- PMD aligned the shared fixmap so the PTE page can be shared between
user and kernel space page tables

- Integrated all fixes and Peters rewrite of the PCID/TLB flush code.

- Restructured the patch set in a way that it is simpler to review

- Got rid of the strange wording of shadow page tables, because they are
not shadowish at all. KASAN, virt etc. use shadows, but these tables
are actively in use and integral part of the functionality

- Moved the debugfs files into a new directory so they don't clutter the
debugfs root directory.

LIMITATIONS:

- allmod/yes config builds fail right now because the fixmap grows
too large and breaks the EFI assumptions. This is still investigated.

A possible solution is just to use one of the address space holes
and grab a separate pgdir to map the cpu entry area. Not hard to do
and it wont change much of the principle of these patches.

>> No.64112099

>>64112044
Intel can't even design a 10nm die given three years and more money than anyone else in the market. What makes you think they can fix this hardware flaw in a reasonable timeframe when they literally don't have a release roadmap for Cannon Lake anymore?

>> No.64112108
File: 1.44 MB, 720x404, chicken neger.webm [View same] [iqdb] [saucenao] [google] [report]
64112108

>>64111991
I guess my old 1090T+GTX480 system I gave to my mom so she can browse youtube will stay unaffected.

>> No.64112111

>>64112078
I dropped 2000 euros to 7980XE last week lol

>> No.64112112

>>64112094
TODOs:

- This needs a thorough review again. Sorry.

- Please verify that all fixlets have been integrated. The mail threads
are horribly scattered so I might have missed something.

- Rewrite documentation. I dropped the documentation patch as it not
longer applies.

- Handle native vsyscalls. Right now the patch set supports only
emulation, but it should be possible to support native as well.
Nothing urgent, I'd rather prefer to kill them completely.

- Populate a branch with minimal prerequisite patches to apply.

https://lkml.org/lkml/2017/12/4/709

>> No.64112115

>>64111357
Hahahahaha

>> No.64112117

>>64112058
>thinks Facebook and eBay users care or understand security

>> No.64112119

>>64112099
Never underestimate a Jew.

>> No.64112120

>>64112087
Well, the *correct* solution would be removing speculative execution from Intel CPUs. But that's impossible, because the CPUs' hardware is designed to do that on a hardware level. The only software-based solution that fixes the exploit is page table isolation, which causes the decrease in performance.

>> No.64112130

>>64112099
>Intel can't even design a 10nm die given three years and more money than anyone else in the market.
Honestly you can't blame them for this. They're hitting quantum tunneling. It's not exactly an issue that anyone has solved.

>> No.64112135

So I'm hearing about Windows and Linux patches, but what about Apple?

>> No.64112136

>>64112078
ebay

>>64112111
proof with a timestamp, i need proof you are that monumentally retarded

>> No.64112139

>>64112078
>>64112111
Return it

>> No.64112140

>>64112087
To put in potato terms, the processor is wired to only work efficiently in one way.

>> No.64112145

>>64112046
Christ almighty, you'd expect the idiots at Intel NOT TO SPECULATIVELY EXECUTE MEMORY READ/WRITES ON INSECURE ADDRESSES. No surprise AMD got it right, it's such an obvious hole, it along with hazards are the hardware equivalent of buffer overflows.

Also https://www.fool.com/investing/2017/12/19/intels-ceo-just-sold-a-lot-of-stock.aspx

>> No.64112163

>>64112135
Apple must be patching it too, though with their macOS root bug maybe not lel

>> No.64112164

>>64112130
>They're hitting quantum tunneling.
That's not the problem.
Retards went with Co for M0&M1.
And some other nasty for yields changes (like fucking SAQP for metal layers, hoy!).

>> No.64112172

>>64112135
Allegedly this bug affects hypervisors and VMs. Something that applel users don't need to worry about. Afaik the *BSDs were told of this exploit and should be fixing it too.

>> No.64112178
File: 11 KB, 600x334, Screenshot-2018-1-2 Initial Benchmarks Of The Performance Impact Resulting From Linux's x86 Security Changes - Phoronix.png [View same] [iqdb] [saucenao] [google] [report]
64112178

>> No.64112183

>>64112094
>uass
Perfect.

>> No.64112190
File: 127 KB, 600x600, 1514911512213.gif [View same] [iqdb] [saucenao] [google] [report]
64112190

Phonorix did a test on disk performance

No performance hit on slower SATA samsung 850

But on NVME 950 samsung speed more than halved!

>> No.64112194
File: 1.59 MB, 996x1168, 1508041400161.png [View same] [iqdb] [saucenao] [google] [report]
64112194

>the warranty on this intel card JUST ran out

I'm not kidding. I literally bought my card on new year's day, just barely more than a year ago. I cannot tell you how pissed I am right now.

>> No.64112199

>>64112178
Good lord.

What the hell

Is this real life

>> No.64112208
File: 58 KB, 512x351, 1508033982678.jpg [View same] [iqdb] [saucenao] [google] [report]
64112208

>>64111357

>> No.64112209

>>64112111
why

>> No.64112214

>>64112172

yes macos is affected...you'd know this if you read the article

>Similar operating systems, such as Apple's 64-bit macOS, will also need to be updated – the flaw is in the Intel x86 hardware, and it appears a microcode update can't address it. It has to be fixed in software at the OS level, or buy a new processor without the design blunder.

>> No.64112219
File: 1.50 MB, 1280x720, laughing patricians.webm [View same] [iqdb] [saucenao] [google] [report]
64112219

>>64112178
How

>> No.64112225

>>64112209
I wanted to compute fast and hard lol

>> No.64112230

It is understood the bug is present in modern Intel processors produced in the past decade. It allows normal user programs – from database applications to JavaScript in web browsers – to discern to some extent the contents of protected kernel memory.

The fix is to separate the kernel's memory completely from user processes using what's called Kernel Page Table Isolation, or KPTI. At one point, Forcefully Unmap Complete Kernel With Interrupt Trampolines, aka FUCKWIT, was mulled by the Linux kernel team, giving you an idea of how annoying this has been for the developers.

Whenever a running program needs to do anything useful – such as write to a file or open a network connection – it has to temporarily hand control of the processor to the kernel to carry out the job. To make the transition from user mode to kernel mode and back to user mode as fast and efficient as possible, the kernel is present in all processes' virtual memory address spaces, although it is invisible to these programs. When the kernel is needed, the program makes a system call, the processor switches to kernel mode and enters the kernel. When it is done, the CPU is told to switch back to user mode, and reenter the process. While in user mode, the kernel's code and data remains out of sight but present in the process's page tables.

Think of the kernel as God sitting on a cloud, looking down on Earth. It's there, and no normal being can see it, yet they can pray to it.

These KPTI patches move the kernel into a completely separate address space, so it's not just invisible to a running process, it's not even there at all. Really, this shouldn't be needed, but clearly there is a flaw in Intel's silicon that allows kernel access protections to be bypassed in some way.

>> No.64112240

>>64112225
threadripper

>> No.64112241

>>64112240
Was too slow

>> No.64112248

>>64112230

The downside to this separation is that it is relatively expensive, time wise, to keep switching between two separate address spaces for every system call and for every interrupt from the hardware. These context switches do not happen instantly, and they force the processor to dump cached data and reload information from memory. This increases the kernel's overhead, and slows down the computer.

Your Intel-powered machine will run slower as a result.

>> No.64112252

>>64111802
>a browser is not software
>i am an idiot

>> No.64112277

>>64112190
Please link
have they done tests on 980 NVME I have that one (2x2TB)

>> No.64112278

>>64112241
like 10% slower and 100% cheaper, and now threadripper is gonna be even better

>> No.64112284

How the fuck does one even update a CPU

>> No.64112296

>>64112278
kek I deserve that, I'm terrified if the 7980XE & 980 NVME SSDs will tank hard

>> No.64112300
File: 74 KB, 320x454, 1513313055657.jpg [View same] [iqdb] [saucenao] [google] [report]
64112300

>>64112094
>uass

>> No.64112309

>>64112284
In this case you can't but in certain cases it's possible to update them at boot up through microcode patches.

>> No.64112315

>>64112277
https://www.phoronix.com/scan.php?page=article&item=linux-415-x86pti&num=2

>> No.64112316

Ryzen 2 in two months, right?

>> No.64112318

>>64112284
microcode, but wont help in this case.

>> No.64112320

>>64112284
Microcode update.

>> No.64112322

>>64112277
first graph
https://www.phoronix.com/scan.php?page=article&item=linux-415-x86pti&num=2

>> No.64112323

>>64112078
return it, and do it now before the news drops so they don't deny the return

>> No.64112326

>>64112296
i feel sorry for you

>> No.64112328

>>64112284
It's a hardware design flaw.
The patch is to kernels as a work around

>> No.64112353

>>64112322
wev lad

>> No.64112358

Holeeeey-fuck, this is the APOCALYPSE for DCG.
"Nobody got ever fired for buying Intel", huh.

>> No.64112360

>>64112328
>>64112320
>>64112318
>>64112309
So how will we be receiving the kernal update? Is the only way through windows update? Or can you manually do it? I thought you could do that via device manager.

>> No.64112362

This is even worse than the F00F bug, because at least then there were few Pentiums compared to today's Core series
Can't even recall them, would probably cost them in the hundred millions
cf. https://groups.google.com/forum/m/#!topic/comp.sys.intel/dmGyQZT6xGU

>> No.64112369

>>64112026
I got my number (2010) from that GCF Security tweet that was shown earlier in another thread. As far as they could tell it's only the i#-#### / ring bus stuff affected.

>> No.64112380

i'm installing phoronix-test-suite here on ubuntu.

will run several phoronix tests (CPU/multicore/disk access) before and after the patch to compare the results.

>> No.64112384

>>64112358
>DCG
Whats t hat

>> No.64112397

>>64111747
Does this mean it will kill any SSD random read/write performance?

>> No.64112399

>>64112384
Data Center Group.
Might as well rename the DOAG.

>> No.64112403

>>64112380
What's your rig, high end or?

>> No.64112405
File: 441 KB, 500x600, [Erai-raws] New Game!! - 10 [720p][HardShit][484F2738].mkv_snapshot_14.28_[2017.09.12_15.23.31].png [View same] [iqdb] [saucenao] [google] [report]
64112405

>>64112178
Holy shit.

>> No.64112408

>>64112380
anyone knows when the linux kernel patch will be available for end users?

>> No.64112422
File: 123 KB, 638x599, 1490566832793.jpg [View same] [iqdb] [saucenao] [google] [report]
64112422

>>64109858
INTEL IS BANKRUPT AND FINISHED
HOW WILL INTELLETS EVER RECOVER?

>> No.64112425

>>64112397
Yes, read the thread.
There's only one platform for hot storage now, and it's caleld EPYC.
Unfortunately, AMD is about to become a monopoly.

>> No.64112426

>>64112403
no it's an old q8300 (core 2 quad) from 2008, i guess it will be seriously impacted.

>> No.64112431
File: 1.24 MB, 640x640, lil sis 2.webm [View same] [iqdb] [saucenao] [google] [report]
64112431

So which FX should I get for the AM3+ board I still have?

>> No.64112439

Linux should have called it
>Total Inability To Supply Usable Processors - aka titsup_

>> No.64112442
File: 5 KB, 435x269, 1513600836225.png [View same] [iqdb] [saucenao] [google] [report]
64112442

who here /waitingforzen+/?

>> No.64112444
File: 37 KB, 269x270, !!!!.jpg [View same] [iqdb] [saucenao] [google] [report]
64112444

>>64112145
Jumping ship while the price is still up, EH BRIAN.

>> No.64112447

>>64112442
More like /waiting for A485/.

>> No.64112448

>>64112431
Sauce

>> No.64112451

>>64112360
On any supported OS the patch will be applied directly to the OS kernel. You'll get it through windows update on windows. On *nix you'll get a kernel update in your package manager.

>> No.64112453

>>64112431
>So which FX should I get for the AM3+ board I still have?
FX-8370e, it is the same as the fx-8370 but specially selected for low power leakage and low voltage. at base clock it is 90 watts

The latest ones can reach 5Ghz on all cores with only 1.42V

>> No.64112454

>>64112078
It's the GTX 970 and EVO 840 from January 2015 all over again.

>> No.64112458

>>64112425
eh, maybe Nvidia will buy Intel and then we'll have an AMD/ATI vs Intel/Nvidia battle.

>> No.64112459

>>64112300
Wow that bitch is fat.

>> No.64112471

>>64112458
NV has no shares or cash to buy Intel RN.
And, well, I won't trush Jensen&folks to do CPUs after Denver.

>> No.64112472

>>64112360
The same way you get any update for windows. Microsoft will force you to upgrade and then reboot your computer at your least convenience.

>> No.64112473

>>64112431

homegirl spends a lot of time in the gym, good for her

>> No.64112476

>>64112451
Crap, so I have to turn on windows update?

>> No.64112481

>>64112397
Anyone got buttmarks of SSDs?

>> No.64112489

>>64112453
You can just undervolt your normal FX chips

>> No.64112497

>>64111812
*unzips PSU bag

>> No.64112500
File: 262 KB, 1064x1330, 1495532425380.jpg [View same] [iqdb] [saucenao] [google] [report]
64112500

>>64112425
>There's only one platform for hot storage now, and it's caleld EPYC.
>Unfortunately, AMD is about to become a monopoly.
AMD is too incompetent to see a chance and use it.

>>64112473
>That
>gym

>> No.64112501

>>64112316
No, Ryzen+ in two, Ryzen2 (7nm) is in the 4th quarter.

>> No.64112506
File: 45 KB, 1024x582, 1496502779982.jpg [View same] [iqdb] [saucenao] [google] [report]
64112506

>>64109858
>mfw 3570k + R9 270x
My CPU is the bottleneck now.

>> No.64112511

>>64112500
>Lisa's AMD
>incompetent
What the fuck are you smoking.
Their 1P EPYC hits right into Intel's heart.

>> No.64112515

>>64112476
You can wait for it to land and then update manually.

>> No.64112528

>>64109858
This is only a rumour, fucking faggots

>> No.64112529

>>64112511
Too many times has AMD had a chance and not used it. I won't believe it.

>> No.64112539

>>64112529
What the fuck are you talking about?
They are already using it with their fucking 1P platforms.
Nothing ever comes close.

>> No.64112552

>>64112284
Would it be possible to open a laptop and physically remove and replace the CPU without bricking the whole thing? Complete technology brainlet here

>> No.64112559

>>64112552
Only if your laptop uses mPGA socket (i.e. CPU is not soldered to the mobo).

>> No.64112564

>>64112481
Well imagine your SSD, but functioning 50% slower.

>> No.64112565

>>64112564
I got nvme tho

>> No.64112578

>>64112565
That's the point.
NVMe SSDs are slower.

>> No.64112582

>>64112564
So still slightly faster than my HDD, got it.

>> No.64112583

Anyone wanna make some predictions on how the resulting malware is going to stack up, in terms of numbers infected, compared to, say, WannaCry?

>> No.64112590

>M.2 NVME is halved in speed with current fix.
How on earth will datacenters and servers deal with this lol?

>> No.64112591

>>64109858
How do I apply the patch in Arch? Do I just get the latest linux-mainline?

>> No.64112600

>>64112590
>that's literally the one I'm using
Well, fuck.

>> No.64112606

>>64112590
Literally by buying 1P EPYC-based storage platforms.

>> No.64112613

>>64112606
>Updating hardware
That will be the fucking day

>> No.64112615

>>64112448
Elizabeth Zaks

>> No.64112624

>>64112613
They don't have any other fucking choice for hot storage.
EPYC, or death.

>> No.64112631

>>64112583
as ti affects almost every intel computer i guess malwares will be rampant unless they're really hard to implement. any unpatched intel core computer would be a potential target..

>> No.64112650

>>64110319
Planned obsolescence. They can't possibly increase performance any more, so they neuter the software, and then slowly build up to the same performance using a new architecture. Rinse and repeat as the stupid masses forget about it each time and just hand over their money.

>> No.64112662

>>64112506
>tfw your 3570k and 16GB of RAM are starting to get inadequate
>tfw you get visited by the magical fairy of 5-30 percent performance loss

>> No.64112677

Will there be any lawsuit?

>> No.64112694

>>64112677
there should be public executions

>> No.64112703

>>64112677
I hope so, I'll be wanting the money to hop from my 6 year old i5 to Ryzen after this.

>> No.64112711

>>64112662
AMD drivers aren't helping me either.

>> No.64112720

>>64109883
The patch is what's going to CAUSE the performance drop, you goddamn retard. The bug itself has no performance impact.

>> No.64112727
File: 694 KB, 320x248, hehehehehe.webm [View same] [iqdb] [saucenao] [google] [report]
64112727

>>64112703
I bet it will only benefit people that are still covered by their waranty

>> No.64112732
File: 2.81 MB, 480x270, 1514662904059.gif [View same] [iqdb] [saucenao] [google] [report]
64112732

>>64112694
I was never one for capital punishment

but

Now would be the time

>> No.64112735

>>64112677
yes a class action lawsuit where every party gets $3 and intel doesn't feel a fucking thing because it barely eats into their profits.

>> No.64112742

>>64112711
>using intel CPU and AMD GPU

>> No.64112745

>>64112631
Just install the patch once released in a few days. I'd rather lose some processing speed instead of getting infected with some ruski botnet client and involuntarily becoming part of a pizza sharing network

>> No.64112747

>>64112735
$3 for every chip would be billions

>> No.64112761

Will any of this actually affect the average person, apparently this shit has been around for a decade without anyone fucking noticing. Is there something I'm missing? The last fucking thing I want is to have to buy a new fucking CPU, I don't have the kind of money to do that right now.

>> No.64112763

8400 owner here, should i kill myself?

>> No.64112772

>>64112735
>implying datacenters wont be SEEETHING

>> No.64112778

>>64112515
And I do that through device manager, right?

>> No.64112781

>>64112763
>should i kill myself?
Yeah, you could've gotten a fucking 1600 earlier (and probably cheaper).

>> No.64112789
File: 95 KB, 530x762, ssd.jpg [View same] [iqdb] [saucenao] [google] [report]
64112789

Arnt you guys overexaggerating?

>> No.64112794

>>64112761
See the phoronix.com benchmarks my dude
and soon the exploit will be OUT and your computer unsecure against Russians and Norks

>> No.64112799
File: 726 KB, 932x1080, ln0YPOk.png [View same] [iqdb] [saucenao] [google] [report]
64112799

>>64112528
>Performance harming updates pushed to current linux kernel and backported
>Linus allowing performance hit into the kernel

Yeah totally a rumor.

>> No.64112805

>>64112778
Just windows updates, it will list all available updates and you just need to select the latest security related update which apparently ought to be released on the 4th

>> No.64112807

>>64112799
This.
They're backporting to stable versions too. No way this isn't huge.

>> No.64112810

What does this mean for the pajeet that jumped ship to intel? You know, the one that sabotaged the radeon division. Just desserts? Karma?

>> No.64112812
File: 17 KB, 439x196, 14568784869.jpg [View same] [iqdb] [saucenao] [google] [report]
64112812

>>64112761
>Will any of this actually affect the average person
Yes

>>64112761
>The last fucking thing I want is to have to buy a new fucking CPU
Don't worry, you won't have to.
For example, if you have a 2013 cpu, it will continue working just fine and speedy like a 2007 cpu.

>> No.64112815
File: 2.58 MB, 1280x720, cliffhanger 2.webm [View same] [iqdb] [saucenao] [google] [report]
64112815

>>64112735
I have 6 CPUs affected by this.
That would be $18 just for me.

They would lose billions.

>> No.64112822
File: 90 KB, 250x250, 1439500672520.gif [View same] [iqdb] [saucenao] [google] [report]
64112822

So what does it means for someone who is swimming in script blockers, adblockers, not visiting shady sites and not retarded enough to download potentially harmless shit on his PC?

>> No.64112824

Has anyone read the comments on the kernel patches? Every time you enter the kernel there's multiple LDT &c. checks, and they must be performed inside critical sections...

>> No.64112827

>>64112794
I don't see my CPU on here
>that 8700
my fucking god my 4770k is going to get fucking slaughtered

>> No.64112829
File: 270 KB, 700x700, 1500687146346.png [View same] [iqdb] [saucenao] [google] [report]
64112829

BROS
AMD IS LAUGHING AT US AGAIN
IT'S NOT FAIR

>> No.64112830

>>64112747
I said per party, not per chip. They'll make individual arrangements with large customers to satisfy them, but everyone else else who owns an intel CPU will get fuck all and have to deal with the performance decrease.

>> No.64112831

>>64112807
>They're backporting to stable versions too.
Like with every normal security update my man

>> No.64112842

>>64112810
Pajeet didn't sabotage anything, or anyone.
He even shat out a bunch of interesting patents.

>> No.64112845
File: 1010 KB, 2000x1333, Lisa-Su-AMD-MIT-00_0.jpg [View same] [iqdb] [saucenao] [google] [report]
64112845

what is mommy thinking right now?

>> No.64112848

>>64112812
I have a 4770k am I fucked?

>> No.64112858
File: 104 KB, 682x600, 1502997829133.jpg [View same] [iqdb] [saucenao] [google] [report]
64112858

>get bullied by friends when i bought my ryzen

thanks for validating my purchase inlel

>> No.64112867

>>64112848
depends on what you're doing but so far the numbers are between 5 to 50% performance loss and yes, also for your CPU.

>> No.64112873

>>64112827

they dont test every single cpu out there, do you have any idea how much work it is to test a single cpu? it's already confirmed that it literally affects every intel cpu up to 10 years ago

>> No.64112882
File: 226 KB, 560x577, 1514939745955[1].jpg [View same] [iqdb] [saucenao] [google] [report]
64112882

today is a good day to be on team red

>> No.64112891

Will it affect games at 720p? If not then fuck it lol, who cares.

>> No.64112897

>>64112631
Sure. But given how wide open this will potentially leave people, I'm actually pretty fucking scared, not as an Intel user, but as a computer user. Look how many retards in just these threads are saying they won't accept the patches. Now think about all of the non-informed, non-updating, clueless, people there are, and how many of them are running Intel machines. There's gotta be hundreds of millions of them. Far outnumbering the total of dipshits running XP, who got infected by WannaCry. This really has the potential to create world panic.

>> No.64112903

>>64112867
I use it on videogames, to emulate and go on the internet (obviously)
How the fuck am I supposed to know exactly how hard I'm being hit

>> No.64112904
File: 1.02 MB, 2025x9065, 1483857980040.png [View same] [iqdb] [saucenao] [google] [report]
64112904

My God.

F/S

>> No.64112907

>>64112891
>slowdowns of up to 30%
>CPU
>resolution dependent

>> No.64112909
File: 42 KB, 578x495, V576S9SY.jpg [View same] [iqdb] [saucenao] [google] [report]
64112909

>>64112858
>went for intel this time around
>woke up to getting spammed by smug anime girls from amd friends
Fucks sake

>> No.64112912

>>64112891
>look it barely impacts crysis at 600x800 lmao amdeets

>> No.64112913

>>64112891
It will, since GPU driver interaction with your game is stream of smelly syscalls.

>> No.64112914

Is the vulnerability confirmed to only affect i series processors? Or are other processor families like Atom affected?

Am curious to how it will affect GPD UMPCs such as Win and Pocket.

>> No.64112935
File: 280 KB, 752x548, 1412582918709.png [View same] [iqdb] [saucenao] [google] [report]
64112935

>>64112891
>Will it affect games at 720p? If not then fuck it lol, who cares.
Count yourself lucky if you can game at 240p when the patch hits my friend.

>> No.64112937

>>64112897
We should just make malicious hacking a death penalty tier crime, if we just killed these hacking faggots instead of arresting them there would be far less of them.

>> No.64112938

>>64112914
Everything since at least Westmere

>> No.64112943 [DELETED] 

>>64109858
So is the Pentium with MMX affected? I'm hoping it doesn't have speculative execution when I grab one from goybay.

>> No.64112948

>>64112873
Fuck, even the early i5's?

>> No.64112954

So, what AMD cpu and what Mobo should I get?

>> No.64112960

EVERY INTEL CPU SINCE AT LEAST 2010 IS

F U C K E D.

8700K IS FUCKING FRIED IN LATEST LINUX KERNEL ONLY MANAGING 50% SPEEDS WITH NVME

REST IN PEACE

>> No.64112967
File: 91 KB, 631x537, 537368637.jpg [View same] [iqdb] [saucenao] [google] [report]
64112967

>>64112903
>How the fuck am I supposed to know exactly how hard I'm being hit
You will most certainly notice it after patching.

>> No.64112985

>>64112960
>Intel is literally about to be fucking banished from the datacenter
What the fuck
The
Fuck

>> No.64112992

>>64112891
>GPUs use syscalls heavily
Yeah gaymers are fucked

>> No.64112995

>>64112954
I wait for Ryzen+ in few months and always buy ASRock middle tier boards.

>> No.64113000
File: 234 KB, 800x612, 1514936077077.png [View same] [iqdb] [saucenao] [google] [report]
64113000

>>64109858
NOOOOOOO! That isn't true!

>> No.64113003

>>64112615

Many thanks, anon!

>> No.64113009

>>64112967
I want to know exactly how hard I'm going to be fucked, because otherwise I have to go and buy a fucking new CPU and I don't have the money to do that

>> No.64113011

If anyone cares (companies I guess), paravirtualization is losing kernel support and hypervisors are not supported anymore, because they both expect page tables to be left more or less alone, or at least relatively untouched

>> No.64113014

>>64112891
This impacts syscalls, GPUs need a fuckton.

>>64112858
>>64112909
>every new CPU in my group of friends is Ryzen because I'm that good of a shill

Revving up the smug anime girls for the 4th.

>> No.64113017

>>64112992
Only people who play at 1440p and 240 FPS. I don't mind a minor downgrade, but I won't be forgiving Intel for this.

>> No.64113023
File: 155 KB, 301x429, iIS4fnj.png [View same] [iqdb] [saucenao] [google] [report]
64113023

>>64112985
2018 starting off with a bang

>> No.64113028

Guys i'm retarded and I don't get it. Is it a bad thing or a good thing? If i get this right, so Intel CPUs will slow down for about 30%?
Sorry for my stupidity

>> No.64113048

>>64113028
bad, very fucking bad

>> No.64113057

>>64113028
From what I understand, if you're playing a game that runs at 30% CPU usage, it'll likely go up to 50% if predictions are right. People say it's more, others say it's less. We actually don't know all the details.

>> No.64113074
File: 1.97 MB, 1920x1038, 1514936573394.png [View same] [iqdb] [saucenao] [google] [report]
64113074

>>64113028
>java ad loads
>lol the code has ring 0 privilege now lmao

>> No.64113079

>>64111423
KEK

>> No.64113087

>>64109954
Would be nice if browsers could just absorb the performance hit, since they’re the primary reason this exploit is so dangerous. I don’t think the entire system needs a slowdown because of this.

>> No.64113109

>>64113057
>From what I understand, if you're playing a game that runs at 30% CPU usage, it'll likely go up to 50% if predictions are right. People say it's more, others say it's less. We actually don't know all the details.


Looking at the disk IO benchmark a Samsung NVME SSD maxes a i7-8700K 100% only reaching 50% of what that SSD is capable of, how the fuck do you think games will not stutter when textures and other shit is being loaded

>> No.64113112

>>64113087
Any VM is a danger too.
So the entire cloud infrastructure of Google, Amazon and so on is fucked.

>> No.64113119

>>64113028
If rumors about this are correct this might be the worst CPU bug ever potentially. Based on the rumors it has the ability to infect your OS from anything capable of executing on the CPU, something as simple as javascript embedded in a webpage could infect your OS. This can all be done silently as well like for cryptolockers. On the server side it can allow virtual machine guests to infect one another through a host system

>> No.64113149

>>64113119
>Based on the rumors it has the ability to infect your OS
Sorry. A malware that makes use of this flaw could infect your OS*

>> No.64113199

>>64111377
>ryan shrout
>amd shill

I won't even call you a retard, because not even a severe mental deficit could come with something like that. That sentence is a result of your fanboyism, denial and apologetic desperation.

>> No.64113205

Gonna go Ryzen+ this year then lads.

>> No.64113221

>>64113109
Why would they? Unless the game is constantly maxing out CPU.

>> No.64113224

>>64113087
No, think about this. You install ApplicationX and their update server gets hacked and serves up a malicious update. You have no idea and run the application and get COMPLETELY owned. This "bug" in design is basically game over for you if you get compromised.

>> No.64113238

people should be in FUCKING PANIC MODE
but the good goys don't know shit, and will never even care about this.

>> No.64113256
File: 62 KB, 1064x438, Screenshot from 2018-01-03 12-04-02.png [View same] [iqdb] [saucenao] [google] [report]
64113256

Ah, going full AMD was not a bad idea after all.

>> No.64113263

I'm not even mad. Time for the anticompetitive trick-playing monopoly to die.

>> No.64113282

>>64113221
Syscalls are the worst affected. It's not a hit across the board but biggest when making many syscalls because the kernel now has to trampoline to a different memory space for every single syscall.
GPUs use a tonne of syscalls.

>> No.64113289

>>64113238
It's going to reach a point where you'll need a license to operate a computer. In a way, I guess the HAM's have the right idea.

>> No.64113295

I can already hear the smug Scott in my head.

>> No.64113307

I don't get it, if this has been an issue for over a decade, why now is it a problem? It obviously being a huge issue aside, what makes it so much worse now?

>> No.64113331

>>64113256
>7199MB
Ishmael, I..

>> No.64113335

>globally turn off java in my browser with a very rare exceptions
what now?

>> No.64113338

>>64113307
No one knew about it until like 3 months ago

>> No.64113352

Wow i was just about to switch from Ryzen 1700 to I7-8700k (don't ask me why). Should i wait for their next gen of CPUs or they'll fix it? And can it actually be fixed without buying a new gen of processors?

>> No.64113360

>>64113307
Recently discovered by Linux Kernel devs

>> No.64113361

>>64113331
What message are you trying to convey here?

>> No.64113365

>>64113307
>I don't get it, if this has been an issue for over a decade, why now is it a problem?
It wasn't known about before.

>> No.64113367

>>64113282
Again, as long as the system isn't being pushed to its constant limit, how should that be a problem? The way I understand it, it's forcing the system to perform more for less performance. That definitely sucks, don't get me wrong, but it's not gonna destroy a game that barely takes 20% of your CPU.

>> No.64113370

>>64113307
Like for any serious flaw, the more wildly known it becomes, thenmore chance for it to be exploited. That's why Intel still puts an embargo on the details. But of course it could have been exploited for a very long time without people noticing.

>> No.64113385
File: 50 KB, 512x512, 1514850045803.jpg [View same] [iqdb] [saucenao] [google] [report]
64113385

>>64112742
Am I still screwed?

>> No.64113387

>>64113352
>can it actually be fixed without buying a new gen of processors?
Hardware design flaw, so no.

A lot of gaymers run at around 60% CPU or more. The effect on I/O could be over 50%.

>> No.64113402
File: 2.86 MB, 450x253, ab0.gif [View same] [iqdb] [saucenao] [google] [report]
64113402

>tfw there will be a fix soon
>tfw intel forces M$ hand and it applies to all of x86 isa

>> No.64113433

>>64113402
AMD committed to the linux patch so that it wouldn't affect their chips.

>> No.64113477
File: 129 KB, 1083x1200, DSkyq4HUQAAdaE2.jpg [View same] [iqdb] [saucenao] [google] [report]
64113477

uhm...
https://twitter.com/aionescu/status/948344914990964736

>> No.64113491
File: 57 KB, 1920x1080, maxresdefault.jpg [View same] [iqdb] [saucenao] [google] [report]
64113491

>>64112662
>mfw hit the cpu lottery and I've had my 3570K oc'd to 4.6 GHz for years with no overheating or stability issues

>> No.64113496
File: 195 KB, 413x350, 1514780442079.png [View same] [iqdb] [saucenao] [google] [report]
64113496

>>64113477

SHEEEEEEEEEEIIIIIIIIITTTTTTTTTTT

>> No.64113555

>>64113477
Isn't this a blatant case of insider trading?

>> No.64113569

>>64113555
>Inb4 it was a planned sale of stock ;^)

>> No.64113570

>>64113477
woah, he actually did it early december or late nov, iirc.

this shit can must've been brewing for a lot longer than we know, and krzanich saw that the compartmentalization wouldn't hold for much longer, since it must've been brought up to them by a major player like google or amazon...

>> No.64113574

>>64112822
Depends on your OS and whether you're blocking *all* JS or just some JS that another person has considered 'malicious.'

>> No.64113583

>>64113555
it's not insider trading if jews are behind it you shabbos goy

>> No.64113609

>>64112914
EVERY intel processor since Nehalem, probably older too.

>> No.64113612

>>64113555
Yeah, annnnnd? At most this cocksucker will get a month in Martha Stewart's old cell.

>> No.64113621
File: 294 KB, 850x861, 1514086327805.jpg [View same] [iqdb] [saucenao] [google] [report]
64113621

>>64113477
>sold off everything bu the minimum 250k shares required to be held by the ceo

>> No.64113653

>>64113555
yeah, and he should be executed in public for that.

>> No.64113674

here is the content of the linux kernel fix if anyone is interested

https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=5aa90a84589282b87666f92b6c3c917c8080a9bf

>> No.64113679

>>64113352
>Should i wait for their next gen of CPUs or they'll fix it?
It's unlikely that even their next gen of CPUs will fix this flaw, because it's a design flaw intrinsic to how Intel CPUs function. Buy an AMD CPU.

>> No.64113693

>>64113555
lol, what are shareholders and potential investors supposed to think when the CEO sells out like that? he should be fired instantly

>> No.64113696

>>64113367
Because the actual performance of that 20% will decrease. The actual usage % of your processor doesn't matter much when what that % can do goes down.

>> No.64113731

guise, is this the worst hardware level bug that became known?

>> No.64113734

>>64113367
It's an os fix. Games communicate to the cpu through the os using direct x. Latency starts becoming a problem since the os has to make sure it's not using the bug to access memory before it can pass it along to the cpu and then back to the gpu. That 30% each way starts adding up.

Do you actually think the gpu can simply just talk to the cpu on it's own? It would be extremely easy for things to install themselves and hijack your pc if it weren't for the os sitting in between. The patch effectively changes how the os handles things, in a more inefficient manner.

>> No.64113741

>>64113731
yeah it certainly is.

>> No.64113747

>>64113731
Honestly, yes.

>> No.64113753

>>64113731
Yeah, since it affects literally DECADES of CPUs sold.

>> No.64113764

>>64113731
Pretty much, it wouldn't be so bad if intel didn't have a massive market share in CPUs and the go to for networking

>> No.64113772

Im conflicted. On one hand I have Intel CPU and suffer from this but on the other I love seeing Intel JUST'd.

>> No.64113774

>>64113731
It will impact all intel based computers/servers/mainframes/super computers with a 5-50% performance hit. so yeah, it's a major fuck up.

plus it will leave hundreds of millions of unpatched machines wide open.

>> No.64113787

>>64113731
Yes

>> No.64113791

>>64113764
Oh wait, this shit primarely hits I/O, right?
The storage and networking folk are gonna be really suicidal this week.

>> No.64113800

Wow I bet AMD fucking wishes this happened at their bulldozer launch instead, shit this would have made it actually look good.

>> No.64113804

>>64113772
same here really
I already made some plans for the future, but I'll wait some days before pulling the trigger
I wanted to upgrade anyways

>> No.64113808

>>64113772
You won't be conflicted if you invest in AMD now and ride sell at peak price hikes when the patch rolls out and Intel has to beg people not to sell their stocks

>> No.64113815

>>64113731
Yeah, pretty much. Things are gonna be interesting for quite a while.

>> No.64113818

>>64113791
it's everyone using processing power, anyone in fact.

like the world global processing power will instantly go backwards 30%.

it will cost billions and billioins of dollars just in electrical power consumption to maintain a constant processing power.

intel is so much fucked

>> No.64113823

>>64113772
Me too.

>> No.64113841

>>64113818
>intel is so much fucked
AMD might be, too. How the fuck are they ever going to keep up with the demand?

>> No.64113844

>>64110796
Why?

>> No.64113856

>>64113841
GloFo is raming up Malta fab capacity by another 20%.

>> No.64113874

>>64113844
Always go with LTS, son.

>> No.64113886

from what I understood it stems from the memory prefetch mechanism, so even they're next gen products must have it from legacy. They'll be doing some seriously hardcore overtime in intel's engineering team to meet 2018 release dates....

>> No.64113894

>>64113856
I'm still skeptical.

>> No.64113900

>>64113886
Cannon Lake's release date was supposed to be January 2016. They're already a year overdue and with still no release in sight.
Intel is finished.

>> No.64113909

>tfw been waiting to sell an i7 2700 for months
>it's now worth dick

just

>> No.64113917

>>64113874
But why not on an A470 specifically?

>> No.64113919

can you like "overclock" mainframes , servers, supercomputers etc...?

of not, many companies will need to scale up their infrastructures to catch up with the % loss due to the patch.

intel may sell even more CPUs AFTER this fuck up, i hope there will be lawsuits and huge penalties for them...

>> No.64113927

>>64113886
The 2018 line up is the least of their worries when they could be facing corporate class action lawsuits and fucking bankruptcy

>> No.64113939
File: 92 KB, 1280x720, 2111767321001_5257424931001_5257417085001-vs.jpg [View same] [iqdb] [saucenao] [google] [report]
64113939

2018 will be epyc

>> No.64113948

>>64112003
>if Intel knew about this beforehand and hid it
Companies usually don't announce huge security flaws until they have a fix ready so no one can develop an exploit in the meantime. They've been provately communicating with Amazon, Google and Microsoft for months.

>> No.64113986

Is this the happening /g/? Is this truly the happening that will kill intel? That will kill x86 as we know it?

>> No.64114004

>>64113986
>That will kill x86 as we know it?
Ehm, AMD still cooks it the right way™.

>> No.64114023

>>64113927
damn, I actually never thought of that possibility.
but it's a lot more complex legal debate than contra-revenue for instance (which they still didn't pay fyi), intel could very well ride a 20 year case and another couple of decades stalling the settlements.

But if their revenue comes down, from bad press and whatnot, yeah, krzanich did the right thing it seems.

>> No.64114024
File: 74 KB, 852x480, meanwhile_at_intel.jpg [View same] [iqdb] [saucenao] [google] [report]
64114024

>>64113927
This. It's going to be years before they can pull their shit together enough to be competitive again.

>> No.64114084

>>64113948
well, that's actually the modus operandi for these things. You don't advertise 0 day (hardware at that) exploits, it's common sense. keeping a lid on it isn't in anyway foul play for this circumstance.

dumping your stocks ahead of it though, now that's a juicy case for the SEC.

>> No.64114139

Can this mitigation theoretically be possible to be bypassed?

>> No.64114179

>>64113986
>Is this truly the happening that will kill intel?
Probably.
>That will kill x86 as we know it?
Absolutely not. This bug has nothing to do with x86, only Intel processors.

>> No.64114183

people responsible for this should be shot

>> No.64114231

>>64112711
>nvidia literal windows xp drivers
>good
dumb meme

>> No.64114237

>>64114139
It shouldn't be possible I think. I think this patch basically emulates what the CPU is supposed to be doing already. Intel's CPUs apparently don't do it correctly so it has to be done through software instead which makes it less efficient.

>> No.64114251

>>64114139
Nope.

>> No.64114261

Is this the inverse y2k, a problem stretching to all hardware (of intel) in existence?

>> No.64114289

man, I wanted to see the world burn, but I didn't think it would be like this.

>> No.64114302

>>64114261
No.
The 80286 is not affected.

>> No.64114359

>>64114289
>The world ends not in a nuclear mad scenario but instead Intel fucking literally everyone including themselves by fixing a bug they made
Just as god intended

>> No.64114370

>>64113734
My 1070 already makes a ton of syscalls

>> No.64114395

>>64112937
>there would be far less of them.
not in north korea pal I can tell you that much

>> No.64114580

>>64113674
don't forget https://lkml.org/lkml/2017/12/27/2
- /* Assume for now that ALL x86 CPUs are insecure */
- setup_force_cpu_bug(X86_BUG_CPU_INSECURE);
+ if (c->x86_vendor != X86_VENDOR_AMD)
+ setup_force_cpu_bug(X86_BUG_CPU_INSECURE);

>> No.64114661

Is this something I should actually be worried about?
My poor 2600k shits the bed when rendering and losing 30% of my CPU would kill it.

>> No.64114689

>>64114661
just don't patch it anon, and you'll be perfectly fine.

>> No.64114703
File: 373 KB, 490x318, sand.png [View same] [iqdb] [saucenao] [google] [report]
64114703

>>64109858
Does this affect me if I'm running a sandy bridge i5?
Does this affect me if I'm running macOS on non-mac hardware? It only mentioned Windows and LInux, wtf...

>> No.64114711

>>64109889
more like when police power trips and confiscate your pc they can back the fucking door into your system with or without your encryptions. ezpz

>> No.64114746

>>64114703
>Does this affect me if I'm running a sandy bridge i5?
Yes, very much so.
>>64114703
>Does this affect me if I'm running macOS on non-mac hardware?
Yes, even more so.

>> No.64114758

>t-t-twe..nty percent slowdown at the most guise.
>up to 4x slower in some cases

>> No.64114825

>>64114703
It's a hardware issue that the OS needs to fix

>>
Name (leave empty)
Comment (leave empty)
Name
E-mail
Subject
Comment
Password [?]Password used for file deletion.
Captcha
Action