How fucked is Intel?
MEGA ULTRA FUCKED!!!!!!!!!!!!!!!!!!!!
https://lists.debian.org/debian-security-announce/2019/msg00090.html> Multiple researchers have discovered vulnerabilities in the way the> Intel processor designs have implemented speculative forwarding of data> filled into temporary microarchitectural structures (buffers). This> flaw could allow an attacker controlling an unprivileged process to> read sensitive information, including from the kernel and all other> processes running on the system or cross guest/host boundaries to read> host memory.
i would really like to see the face of a couple jerk companies asset managers which bought into the "xeon discounts" for their data centers only to have 50% of the performance taken away from themHAHAHAHAHAHA, I FUCKING TOLD YOU!
Another year, another 5% performance lost to some fatal design flaw bug.
>>70959152it's like 50% as hyper-threading must be disabled
How the fuck is the company still alive?
>>70958608Intel absolutely YEETED ON
>>70958608>buy zion>twice the price>twice the power consumption>half the performance>1/4 performance per watt>slash 30% cause spectre and meltdown>now shave half of what's left cause lol fuck yo threadsthose who bought xeon might as well replace them with raspberry pi clusters at this point
>>70959181>hyper-threading doubles performanceThe absolute state of this board
>>70958665debian still not update the fucking microcode
>>70958608are intel putting in those flaws on purpose to allow government agencies to spy on random people?or is it just incompetence?
>>70960399>hyper-threading doubles performancefor certain tasks. I would assume docker and similar server software will pay a penalty...
>>70960375>Buy used xeon>Same performance post spectre and meltdown patches as a Ryzen 2600 for 1/3rd the price>Hyperthreading increase performance anywhere from 0% to 100% depending upon application>Average is 20%>Now have 0.8 of a ryzen 2600 for 0.33 the priceStill a good deal
>>70958608At this point I'm using Raspberry Pis for mail, file, and web servers. They're slow but work alright for my uses. Just bought a new AMD ThinkPad A series. It's pretty nice. I also use PowerBooks with Debian. I'm torn between a Ryzen and a POWER9 workstation. I have $2500 allocated to either one. I ditched my Intel laptop and workstation last year because I'm not retarded and knew it wouldn't be over soon. >>70960442Probably both. If you want to be safe you should get a SISD (single instruction, single data) processor. They're slow but not unusably so. A Raspberry Pi 3 is fast enough for basic web browsing and office tasks if you install a lightweight Linux or BSD distro and window manager
>>70959181>30% performance loss for Spectre mitigation’s pushed out so far>another 40% performance loss for this vuln on top of thatIntel CPUs are running approximately 28% as fast as they were two years ago, this is assuming you are running all the patches and turned off hyperthreading.Literally more than two thirds performance penalty. Suddenly my Thinkpad X23 looks competitive again.
>>70960484and the power consumption over the lifespan of the build?
People mocked me when I bought my fx eight years ago, who’s laughing now.
Ok, so there is no clear explanation yet, is this an actual "you are seriously fucked, they can hack you remotely", or it's another exploit that requires physical and admin access to the machine to be performed, which at that point you have way bigger things to worry about than cpu vulnerabilities?
>>70960442they just want you to buy new chips, that's all
>>7096100512-30 watts difference?
Feels good being a Threadripper owner.
>>70961207Just delid + delap and everything will be fine goy
>>70961207it's another speculative computing exploit. If they can see your kernel space, they can get whatever is in there. The fix wipes the buffer whenever its done with the data, with a penalty of 3-9% based on model, according to techcrunch. Yep, that's on top of your Spectre/Meltdown penalty.I wish people actually knew how to read here.
>>70961318That doesn't answer the question you fucking retard
>>70961227The new chips has the same issues plus all the non-disclosed issues.
>>70958608>tfw still on kernel 4.19literally 30 years behindthx AMD
>>70961318forgot to include that hyper-threading buffer data is also up for grabs and is recommended to be disabled in coordination with the fix. Enjoy losing your extra threads.>>70961338of course, but you need a payload to do that for you and send it back. it's like you can't think for your self, damn.
>>70958608> i can screencap but can't post linksPOST THE LINK, YOU FAT USELESS CUNT.
>>70958630that's not how it works, iTODDLER.
>>70960562that's not how percentages work ya fuck
>>70961348then who'll buy their newer botnet with fresh vulnerabilities?
>>70961433They are cumulative dumbshit.
>>70960562cringe and bluepilled
>>70961351Just update your kernel??https://kernel.ubuntu.com/~kernel-ppa/mainline/v5.1.2/
>>70959181Up to 40% according to Applehttps://support.apple.com/en-us/HT210108
>>70961516Wow, “only” 40%, that’s a whole lot better than 50%. I guess Intel in the datacenter still makes a lot of sense, what with “only” a 40% slowdown.
>>70961516I really don't want to be the devil's advocate, but apple is moving to their own inhouse cpu's, I'd expect them to say that and give their customerbase the impression they must switch.
>>70960484>used xeon>same performance as a 2600>implyinghave fun with your dud, lmao
>>70961603Intel Pajeet detected, the performance gains from utilizing hyperthreading are well studied and “about 40%” is accurate.You should be ashamed of yourself.
>>70961679>Anyone who disagrees with me is a shill555-COMEON-NOW
>>70961726AMDrones anally raped by Intel for decades. 2019: nothing has changed.
Other threads have said this actually affects processors all the way to Cord Duo (2009) but Intel is gonna get away with ignoring the rest.
>>70958630Got a giggle out of me.
>>70961813> he doesn't know amd have to pay intel royaltiesAMDrones: anally raped for decades.
>>70961836Intel is using AMD64, Intel has to pay AMD royalties.
>>70958608>tfw the only intel cpu I own is in a thinkpad from 2006Feels good.By the way, does anybody know how this patch behaves on unaffected CPUs? I certainly hope I won't get a performance penalty on AMD, just because intel was greedy.
https://www.guru3d.com/news-story/new-ridl-vulnerability-hits-intel-advises-disabling-hyper-threading-below-8th9th-gen-cpus.html>To be able to exploit the vulnerability, no more is needed than hiding a few lines of malicious code on for example a website. A visitor that opens this site open, it will leak information.BTFOTFO
>>70961252Holy fuck, you are delusional.
>>70958630baste and hyperthreaded
>>70960399>50 percent is doubleThe state of your basic fucking math retard.
>>70961726Based Jim. Fucking Lugenpresse
>>70961726I'd suck Jim's tiny Scottish sausage.
>Intel will continue raking in money because of their pseudo-monopoly on chipsets
>>70962029It is when you decrease by it. It's a third when add it. And you(?) decreased by it, meaning going from the remaining 50% back to 100% (by enabling ht) would be doubling the performance. Mathlet.
Post your processor vulnshttps://www.ghacks.net/2019/05/15/mds-tool-find-out-if-you-are-vulnerable-to-microarchitectural-data-sampling-attacks-mds/
>>70961836amd doesn't pay intel royalties since at least 2003 retard
>>70962101this is pre-kernel updaterebooting now brb
>>70962196feeling much safer already
>>70962029you can't be this fucking stupid
nooooo bros delete this the goyim at the university were supposed to keep quiet
>>70958608bad news for the 5 people who can afford intel quality processors and a good OS, yet still choses to use linux.
So this is where the tech industry is at.Years spent building up a false sense of security and performance, with trash firmware on top of trash hardware, along with rising performance cost of running simple programs and websites.now the quick fixes implemented years ago are catching up to INCEL processors and things will be slower than theyve ever been.imagine if people actually started caring about optimization again because of this.I doubt it, though.
>>70962029oh no anon
>>70962873>bad news for the 5 people who can afford intel quality processors and a good OSfucking idiot, those vulnerabilities are on every OS.
>>70961884You should be able to just add a kernel parameter to disable the patch if it behaves negatively on unaffected CPUs.>>70958630Heh
>>70963008if you dont think 99% of this shit is just a bubble predicated on marking trends and funneling capital into 'tech startups' even though those tech startups are almost entirely predatory venture capitalists trying to make a quick buck
>>70961305Don't forget the chiller!
>>70961504Some people are Just retarded
>>70962101>>70962196>>70962257>>70962274It's like a list of STDs
Why can't Intel stop being pozzed
>>70963459WHAT IS THIS ? What is effecting storage speeds!?
>>70963480syscall overhead is my guess
better: How fucked is Linux?
>>70960423Debian doesn't upgrade the microcode all they do is upload Intel's patches to their repos.
>Apparently Intel attempted to play down the issue by trying to award the researchers with the 40,000 dollar tier reward and a separate 80,000 dollar reward as a "gift" (which the researchers kindly denied) instead of the maximum 100,000 reward for finding a critical vulnerability.>Intel was also planning to wait for at least another 6 months before bringing this to light if it wasn't for the researchers threatening to release the details in May.https://www.nrc.nl/nieuws/2019/05/14/hackers-mikken-op-het-intel-hart-a3960208
>>70959181>hyper-threading must be disabledSo my first gen i7 without hyperthreading will now perform like a modern i5 now? Nice.
>>70963719>NO GOYIM WAIT FOR THE NEXT GENERATION OF CHIPSAHAHAHHAA
>>70963719>400000 dollaridoos for finding a hole in the architecture equivalent of pic relatedBased. I want that job.
>>7096385840k is nothing, its probably split between an entire team.
>>70963008Wonder why the processors that Jim Keller has worked on don't suffer from these flaws? Does he know something?
>>70963931I'll take it.
>>70958608>WTF I AM PANICso just setup a rasp pi or alternate desktop as a network connection pass through/filter and then no one can access your exploited intel.
>>70963912AMD have a different kind of development philosophy which probably helps, then there's dumb luck.PS. Please stop the Keller worship, he's good but is/was just a part of a team.
>>70963912It seems AMD's engineers were more cautious and baked mitigations into the underlying hardware to avoid problems like this to begin with. Adding to that AMD didn't even support a full implementation of SMT until Ryzen. Bulldozer was a partial implementation and I don't think it was affected to the same extent.
I'm so sick of fucking restarting my computer for fixes because of your latest fuck up Jewtel! Get your shit together!
>>70963890The team who discovered ZombieLoad has 4 members. That's still 10 grand each, not bad.
>>70964470Fucking chump change. They can stream on twitch for like a week and get that return easily. Just by virtue of media attention.
>>70964609If Intel offered much more than that for every vulnerability found, they'd be taken to the cleaners.
>>70964626>Multi-Billion dollar company can’t afford to give at least a million in hush money per researcher per exploit.That would be no more than what 500 million at this point? It's jews being jews. Thats it.
WHAT DOES IT MEAN FOR MY OVH KVM VPS I am on kernel 3.16.0-6-amd64
>>70962274>>70962257>>70962196>>70962101does not work here. MDS Toolcrashing during start. Have x5470
>>70958608What does this mean for the i7 6950X?Oh fuck please no
>>70961338>calls others retards>can’t even answer the questionNot based.
>>70961726unless that money directly goes to AMD's bank account it literally doesn't matter
>>70964380Based bulldozer was cmt, which scales pretty damn good desu.
>>70964654Nothing overall. Keep installing security upgrades for the kernel, and upgrade to stretch before it runs out of support next year.
>>70958608Glad debian stable gets buster now so that it will have an older kernel even in 4 years. Nice. I guess I am gonna stay like this for the next 7 years. Fuck the haters, I enjoy my holes.
why is arch still on 5.0
>>70961836he doesn't know intel has to pay amd royaltiesnu/g/ can't even isa x64
>>70965136It gets backported you dumb incel
>>70958608Linux 5's out already? What the hell they just started 4. Give it a few years and they'll be caught up to Chrome's retarded version numbering.
Is a core 2 duo old enough not to be affected by this?
>>70964016That's not how speculative execution works you stupid nigger.
>>70964016Or just use a browser without scripts.
>>70965202They jumped to 5.0 because they reached some number of commits on Github if I remember right.
What does this mean to my 4.9.0-9-amd64 ?
how is this even possible anons?people sperg out screeching, flailing limbs, fiercely defending intel against anything, very concerning and damnable things, latching onto any piece of info that brings them a tiny bit of solace while cryying out that it isn't that bad, just to spite the 4 anons who would be same for the amd's side?how the fuck can you live that?what's the maximum amount of shit a fanboy can take from their fav tech company? (there's no maximum apparently, it's fucking infinite)
>>70958630You cheeky bastard
Lol, my cpu looks like a swiss-cheese.Thanks Intel!
>mfw currently on a core 2 quad from 2008>mfw thought of upgrading to 9900k just a month ago but decided to wait for ryzen 3000
So this shit only affects you if you have hyper threading right?
>>70965352AND no patch.
>>70965149>>x86_64 Core linux 5.1.2.arch1-1get better mirrors
>>70965378Were the fuck do i find the patch then?
>>70965403hussle.i'm not you're bitch and I have a ryzen.
>While programs normally only see their own data, a malicious program can exploit the fill buffers to get hold of secrets currently processed by other running programs. These secrets can be user-level secrets, such as browser history, website content, user keys, and passwords, or system-level secrets, such as disk encryption keys.https://zombieloadattack.com/I don't understand this reasoning. If a malicious program has already made it into my computer, somehow doesn't it already have access to all this, no matter if there is a security hole? What's preventing any program from reading from the filesystem or the ram?
>>70965423I am thinking the same way and I always loved the idea to be able to do this. I tried in the past with various hooks to read text from other programs but never got to do it and never understand the security risks. A program like tinytask combined with the functionality to read from clipboard always got me around.
>>70965423user privilegesif you run your malicious program as root, you're fucked. If you run it as normal user in a reasonably configured system, you're supposed to have a chance to keep your secrets.
>>70965466My whole GUI runs as root because I hate typing my password in.
>>70965423The biggest problem are virtual machines actually.
>>70961504my multihead setup doesn't work with newer kernels.
>>70965386wtf? i tried many different mirrors but only get 5.0.12whats your top mirror?
>>70965324is this a list of STDs?
>>70961073>fxEveryone is still laughing
openBSD and Theo are ALWAYS right. They disabled hyper threading months ago and everyone said it was autism. I’m going to have to make a separate thread on all the times they were right.
>>70965570Yeah, my CPU is a fucking slut.
>>70965477>hate typing my password indoes not require>GUI runs as rootfix your system
>>70965520https://ftp.halifax.rwth-aachen.de/archlinux/$repo/os/$archafterwards, use reflector to generate better mirrorlists whenever you miss something
>>70962101why am I so vulnerable even though I patched :c
>>70965666>fix your systemYeah I had to fix quite a few lines to be able to login into gnome as root but it works.
>>70965863I still don't understand why you think you need to run gnome as root to skip specific password checks.
>>70965826Because Intel is a retarded piece of shit of a company with infected architecture.
>>70966075I'm fucked aren't I
so what will VPS providers do?start offering a separate service for those with "sensitive data in memory" and charge a premium for it, everybody else is lumped onto their old unmitigated hardware?(I apologize for giving them ideas)
>>70960499>A Raspberry Pi 3 is fast enough for basic web browsing and office tasks if you install a lightweight Linux or BSD distro and window managerWhat about programming?
>>70966075Clean and beautiful.
>>70958608>does not affect AMD and ARM chipskek
>>70966131So much red...
>>70966211Feels good to be on Ryzen
>>70966131what app is that? i want to check myself ;_;
>>70966211Kikes on suicide watch
>>70966662Oh fuck, im fucked. Gigabyte won't release any new bios for my mobo...
>>70962101how the fuck do i downlkoad
>>70966178Cloud providers already offer premium plans to customers who process sensitive data, and run them on separate hardware. These vulnerabilities enable cross-VM attacks only if the attacker and victim are running on the same physical core.
>>70966688Microsoft will likely release a patch and intel will release a microcode update, I wouldn't worry too much.The issue isn't the vulnerability (although that is AN issue), it's the eventual performance hit once the patches are released.
>>70966705Try reading whats in the link.
>>70958608>more issuesGod, fuck intel. Bunch of jew retards. I'm not giving up on my thinkpad though
FX 8350 TRADES BLOWS WITH 3570K AHAHAHAHAHAHAHAHAAHAHAHAHAHAHAHAHAHAAHAHAHAHAHAHAHAHAHAAHhttps://www.youtube.com/watch?v=IEwwX7VEhDk
>>70962101THANK YOU, BASED INTEL!
>>70967194O O F
>>70961318are you diagnosed with autism?
When the fuck are the new ryzens coming out I am fucking done with Intel and Nvidia
>>70964626They spent $450 million on hiring black people without STEM degrees to stand around just 3-4 years ago, just because Obama told them to.
>>70965916Being root allows me to edit any file all over the system whenever I want by just using nautilus. Any terminal instantly opens as root without a password. Any program can write anywhere it wants. It's like in windows when having the UA deactivated. I never had problems with this setup in the last 10 years.
>>70967241It's official: only autists can read
i'm about to noibrs noibpb nopti nospectre_v2 nospectre_v1 l1tf=off nospec_store_bypass_disable no_stf_barrier mds=off mitigations=off on all these hoes I don't even care
>>70967371what is a person without deep knowledge of modern kernel architectures supposed to infer from the phrases "speculative computing exploit", "kernel space", "hyper-threading buffer data", the guy asked him a simple question and instead of answering directly he decided to sperg out
>>70967407You only need mitigations=off mds=offmitigations=off encompasses everything except mds
>>70962101>cleanest intelfag ITTSurely I'm getting false negatives right?
>>70958608"All users *must* do _____" beats the whole point of using Linux
>>70966178Realistically they will just switch to EPYC and Threadripper.
>>70967540and offer those to premium users?what will they do about their current hardware?throw it out?
>>70958630dare I say BASED
>>70967570Well yeah. If your hoster is worth a damn they regularly upgrade their hardware. This one will just be another gradual update with the old hardware either thrown out or rented at a discounted till they are also gone.I for one wouldn't take the risk or another 40% performance impact. At that point AMD is literally your only choice if you aren't clinically brain dead.
>>70966179Programming can be considered a basic office task. Most text editors (not IDEs) do less than the average word processor. All the standard GNU dev tools and compilers you're able to run on Debian normally will run just fine on ARM. If you want an ARM v8 64-bit environment to work in, try Devuan's Raspberry Pi 3 build. It's quite nice.
>>70961489Wouldn't it be 42% instead of 28% ?
Outside of the security fearmongering:Was Spectre, meltdown and this vulnerabilities really that bad in the wild?
>>70966705For the love of god i cant find a download link on that fucking page
>>70967841>Was Spectre, meltdown and this vulnerabilities really that bad in the wild?They were patched so that they couldn't be exploited, with a painful performance cost.
>>70958608wew im fine!https://en.wikipedia.org/wiki/List_of_Intel_Core_i3_microprocessors
>>70967987But ironically no one could exploit them because the vulnerability was too specific
They’re doing this shit on purpose. Intel is compromised. Only way forward is power9.
>>70968011>intelYou were never fine.
>>70968116I can actually see this as a way for Intel to bail out of CPU sector, or just becoming the Nokia of smartphones
>>70968095Bullshit, if it was then they would never have patched it given the steep performance penalty.
>>70968162>step performance penaltyHow much it was anyways?
>>70962101app crashes when i try to open it o well
>>70958630The fuck, I hated Satania in her series but love her now
>>70962029You are one stupid motherfucker.
>>70962101I gotchuu familia$ lscpu | grep "Model name"Model name: AMD Ryzen 7 2700X Eight-Core Processor
$ lscpu | grep "Model name"Model name: AMD Ryzen 7 2700X Eight-Core Processor
FalloutZombielandRIDL, microarchitectural load port data sampling (MLPDS)RIDL, microarchitectural data sampling uncacheable memory (MDSUM)CVE-2018-12126CVE-2018-12130 CVE-2018-12127 CVE-2019-11091https://blogs.oracle.com/security/intelmds
FalloutZombielandRIDL, microarchitectural load port data sampling (MLPDS)RIDL, microarchitectural data sampling uncacheable memory (MDSUM)
>>70958608>check tonkpad>P8400>Launch date: Q3, 2008safe.
>>70962101First and last time buying israeli
>>70968409I can't blame you, I almost bought a 5820k because it was mad value back then. Consider Threadripper if you need more lanes desu, I'm looking selling a 1700 for cheap and going all in on a 3950x if they don't skip TR this time around.
Hypothetically if I was a richfag with multiple PCs, I could just build with Intel and open up all these holes for dat speed but just do banking on my 2008 AMD laptop, yeah?
Adding poo to words isn't creative or amusing, you sound more mentally challenged than a literal child.
>>70968162according to most of the pictures in this thread, nobody patched anything.The anon you are refering is right though. There hasn't been any case known where this was used and oh boy, there are millions of unpatched windowx XP, 7, and linux boxes online.
>>70968573If you want to get all your passwords stolen, sure.
>>70969003Why not wait for Zen3/4 going tall? Having a bunch of DRAM so close is going to be a huge increase in performance.
>>70969042how will they dissipate all that heat?
>>70958630Fucking kek. Congrats on the (you)s.
>>70964747>crashing during startMy Avast kept trashing it so check your virus chest if you're a casual like me and use an AV, probably first time it bloody registered anything since I installed it too.
>>70958630>>70964953We're filled with comedians today, it seems
>>70969070Maybe the ram shit isn't affected by heat that much
>>70968613>there are millions of unpatched windowx XP, 7, and linux boxes online.LOL! Do you seriously think someone running unpatched windows XP, 7 and Linux boxes online would have the technical know-how to determine that they have been hacked ?And if they are unpatched there are likely easier exploits than Intel cpu bugs.
>>70969042So we're going to fully integrate the RAM into the CPU? That's really neat.
OpenBSD doesn't have this problem because they disabled Intel SMT since 6.4 I still remember some of you spergs laughing about that SECURE BY DEFAULT
>>70969042>>70969294So DRAM, NVRAM AND GPU INTO THE CPU?How? Heat management, hello?
>>70961318>tripfag has literal autismwoah......
>>70969070>>70969329I don't have the slightest idea, but Intel is definitely going the same way before going chiplets. Maybe looking at some patents will give you an idea, I hear most of this stuff is already resolved.
>>70965423filesystem, yes. ram, not really.
>>70967483>6600kHope not, that's the most recent Intel product I own and its nice to know its relatively unpozzed
>>70959181Hyperthreading is already turned off on scientific systems :)
>>70960442They’re unavoidable. These speculation-related bugs will continue to pop up and there is nothing we can do as so much processor performance depends on speculative execution now. >>70960499>>70961227https://arxiv.org/abs/1902.05178“Spectre is here to stay: An analysis of side-channels and speculative execution”
>>70965483>tfw 5 monitorsI'm using Linux-5.1.2 right now and it works just fine.
>>70969834>“Spectre is here to stay: An analysis of side-channels and speculative execution”:(
The Linux people are always fixing I tell problems in the kernel but does Microsoft do anything about it? I guess this affects any OS running on Intel, right?
>>70958630I love this meme and how it's used only in the most appropriate possible of postings and not spammed like /pol/, /v/, or /tv/ memes
>>70960499>A Raspberry Pi 3 is fast enough for basic web browsingif you use a text only browser maybe
I think Intel will stay pozzed forever, it's game over.
>>70958630Paste and delided.
>>70964953>>70958630These are both terrible.
>>70959152Still ahead of you pajeet.
aren't companies still buying intel cpus? apparently no one cares.
>>70970974No, only indians.
>>70962101wtf amdbros i thought i was safe???????????
>>70971031Slightly better than intel
>>70971065Why am I so vulnerable compared to some other amd people here? Is it my os? bios? unlucky bulldozer cpu? I don't know what most of those vulnerabilities are.
>>70958608Very, bring Linus back too.Let's make a new kernel actually cos why the fuck not.
>>70962101Good to know I've been making the right choice for sticking with AMD over the years. There are vulnerable Intel CPUs going all the way back to 2007 apparently
>>70971869I don't think that's Linux fault tho...
>>70960562>turned off hyperthreading.why in gods name would I ever do such a thing?I specifically paid extra to have hyperthreading.
>>70962101>ryzen is safe!lol
This a jewish trick, right? My Xeon station is behind a router and allows only signed ssh connections. Fuck your edge case 'hacker' scenarios.
>>70973213lol. 9700k ftw
>>70973373It's a risk if you run untrusted code. If you don't, then you shouldn't really care. But for VPS providers, cloud services, etc (IE the big boys) their entire business model is running untrusted code.
>>70973360The tool is designed for Intel processors only.
>>70973559>their entire business model is running untrusted codeAnyone really gives a fuck?Are they not insured?If one cretin decides to somehow hack a fucking VM in a node somewhere, nuke him and move on. Who in their right mind will play with hacking virtual machines running ubuntu somewhere? This is all bullshit. If Jews decide to hack some machine I assure you they will use something better than a "newly discovered" public exploit.
>>70973599Their customers care about not having their system's data breached, idiot. Insurance doesn't cover customers leaving and going to another provider because you won't install a patch that protects their data from Putin
>tfw bought a Xeon 2 weeks agoTHIS CAN'T BE HAPPENINGI'M IN CHARGE HERE
>mustthat's not very free as in freedom of you
l e g e n d a r y
So we all go to amd?
>>70969329go read amds patents. they are pretty interesting.
>>70958630based and redpilled
>>70973920Witch cpu>>70973951I will look it, thx
>>70958608RIP my i5 750Well it's not like I'm ever going to get a mitigation in illumos anyway, considering how long it took to get one for meltdown. The ONE GUY who writes kernel patches is gonna have a mental breakdown soon.
>>70974376Your CPU doesn't even have hyperthreading.
>>70974404I know, that saves me from the worst of the exploit, but disabling hyperthreading alone doesn't fix it entirely. It'll just take longer to get usable data.
>>70963717you mean kernel firmware? that's what fedora does to patch microcode
>>70960562you have to be so dumb to think this is true
>>70965296it means you will get 4.9.0-10-amd64 earlier!
>>70961516This is a worst case from just disabling hyperthreading. Go look up benchmarks with and without hyperthreading, in the average case the difference is small, it didn't magically become more important because the disabling is a result of security fears. Also, with security fixes its possible to have hyperthreading on, the paranoid advice is to outright disable hyperthreading until kernel fixes/microcode updates address this, which will reduce performance much less
>>70973560And the tool is so good that it can detect vulnerabilities on other processors too? amazing, good job intel!
>>70958608Hyperthreadless 2500K has nothing to worry about right?
>>70961489>>70960562that's not how it works dumbshit.reduced by 30% and then by 40% would be multiplying the initial by 0.7 (-30%) and 0.6 (-40%), giving you 0.42 the initial amount, 42%. Still intel is fucking dead
>>70974579Store/fill buffers, and load ports are still needed to be rewritten when core switches between processes (at least untrusted ones), so yeah. Update is still recommended.HT is just harder to mitigate due to required changes to OS scheduler (and I got no idea if it's implemented yet or not), but it's much harder to exploit too.
>>70962101>>70962274>>70966075>>70967483>>70968409>>70972974>>70973360If you are running windows you have bigger troubles to care about than some CPU vulnerability.
>>70967334You can do all of this by correctly setting up sudo and graphical sudo AND avoid giving root to random new applications. You say you had no problems, but in your setup, if you had some, you wouldn't notice.
>>70958608>2019>still using intelIt's like you brainlets were wanting this to happen
>tfw still using old i5 750 from 2009
>>70965826It's too late anon, hackers are raping your info as we speak.
> bugs : cpu_meltdown spectre_v1 spectre_v2 spec_store_bypass l1tf mdsINTEL!!!!
>>70958608One must wonder if they're "fixing" these backdoors because the Chinese found them.
>>70974898welcome to the swiss cheese security hole club
>>70974898Hey, I'm (>>70974376) not the only one.
>>70971241>I don't know what most of those vulnerabilities are.Probably why you are on /g/ right now.
>vulnerable since dozens months>never patched anything>still secureCan the AMD faggots stop spewing bullshit? Who cares about those cpu bugs, it's literally unexploitable.
Fuck yeah. So my non-hyperthreading i5-6300hq is a goddamn monster now.
>>70975180>run malware JS on browser on X.org>exploit can read memory of all other programs running on X.org session
>>70975311>X.orgI'm running Windows you dumb ass.
>>70973627IT'S NOT FAIR
>>70968573Let's say you do gaming. So you need steam. Your steam acct. is worth potentially thousands, and you HAVE to log in. Thus you need passwords, and security in general.
>>70968613And those exact unpatched boxes are part of the glorious collective of botnets nowadays (hey anon why is my computer so slow all the time ? me: shudder)
>>70975367If you somehow think that matters than good luck
>>70964626If Intel didn't pay much for vulnerabilities, they'd sell the information to someone else instead.