Basically every indirect call (one instruction) turns into a seven-instruction sequence (retpoline) that will, due to preventing speculation, result in massive slowdowns.
Unlike the KPTI patches, which only affect things on each system call, this happens on every indirect call and probably bloats the code considerably too.
These are the patches that prompted Linus' angry response, and it's understandable: it's not an exaggeration to say these patches will be the biggest Linux kernel performance regression in history.
Merging has been halted until the Intel devs add configuration flags to allow it to be disabled selectively (for example, for AMD CPUs).
unless you're running a database or doing highly specialized stuff like live music production, you won't notice the hit
Sebastian White
>dont worry goy it doesnt matter
Levi Nguyen
NOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO THIS CAN'T BE HAPPENIIIIIIIIIING IT WAS NOT SUPPOSED TO BE LIKE THIS
Luke Campbell
>unless you're running a database or doing highly specialized stuff like live music production, you won't notice the hit you're an idiot. all modern software has a built in database, and makes a shit ton of system calls. how do you think it gets i/o?
also, thanks to webdevs who can't program, every website is going to get extremely slow. if you think your browser dosen't have a built in database, or doesnt make a shit ton of syscalls, you are mistaken.
Brayden Long
>spectre fix will only affect intel
Christian Reyes
None of this would happen if the kernel was just written in Rust.
Carter Bailey
>Merging has been halted until the Intel devs add configuration flags to allow it to be disabled selectively (for example, for AMD CPUs).
Those sneaky jews tried to make AMD pay for Intels fuckup?
Jordan Green
wrong again, friendo.
it doesn't matter what code is being run on the machine, it's a hardware side channel attack. also using this i can run code on your machine to do whatever i want, like turn it off, or block Sup Forums. it's an intel hardware backdoor lol
Ryan Cook
This actually has nothing to do with syscalls, have you even read the OP? Indirect calls are everywhere in regular code.
Brayden Reyes
Microsoft might as well patch Vista and XP which will kill the damned straggler's system performance.
Thomas Morales
This is an Intel-specific fix written by Intel engineers which uses Intel-specific architecture features, read the Andi Kleen emails.
Jeremiah Long
My emulation already took a hit with the meltdown fix. I'm just gonna take my chances with this. No more slowdown.
retpoline is a compile-time option with a patch set being pushed by Intel that does not allow a way to disable it dynamically for CPUs that aren't vulnerable.
Level that playing field, goy!
Bentley Edwards
I'm never buying a Hebrew processor again
John Sanders
yes wasn't it clear by their sleazy inclusion of others but in general way without specifics just to make them look not that bad because you see it's not just them but in reality it is mostly them
Jayden Lee
>block Sup Forums Imagine waves of hackers going about to millions of machines and blocking Sup Forums across the globe.
Andrew Walker
>already using an i5 3320M in my day to day laptop >see these fucking threads everywhere I haven't updated my OS for like 3 weeks now. What happens if I update Debian? Am I going to get bent over and fucked by this patch? AMD shills give me an honest answer, does this impact performance on AMD mobile chips? I'm due for a laptop upgrade and I'm thinking about dropping like $800 on an A-series ThinkPad. Please respond.
Dylan Long
So... is this going to give insight into reverse engineering certain consoles architecture for emulation and piracy?
Easton Green
the fuck??
Nathan Hall
AMD isn't affected performance-wise by any of this so far. Maybe in the future an AMD-specific Spectre vulnerability is discovered and a performance-degrading patch is needed, but to date that hasn't happened and I bet Intel getting even more crippling fixes is more likely.
Gabriel Jackson
Sup Forums is here
Sebastian Rivera
How do I install homebrew on my 3ds?
Easton Flores
>tfw no T/X/W thinkpads with AMD >tfw no cheap servers with AMD fucking jews with their almost monopoly position
Ayden Martin
Joke's on you, Intel doesn't have heavy R&D in Tel Aviv.
Lincoln Gutierrez
Thinkpads are botnet.
Alexander Phillips
...
Ryan Green
>list of affect intel CPUs
Intel® Core™ i3 processor (45nm and 32nm) Intel® Core™ i5 processor (45nm and 32nm) Intel® Core™ i7 processor (45nm and 32nm) Intel® Core™ M processor family (45nm and 32nm) 2nd generation Intel® Core™ processors 3rd generation Intel® Core™ processors 4th generation Intel® Core™ processors 5th generation Intel® Core™ processors 6th generation Intel® Core™ processors 7th generation Intel® Core™ processors 8th generation Intel® Core™ processors Intel® Core™ X-series Processor Family for Intel® X99 platforms Intel® Core™ X-series Processor Family for Intel® X299 platforms Intel® Xeon® processor 3400 series Intel® Xeon® processor 3600 series Intel® Xeon® processor 5500 series Intel® Xeon® processor 5600 series Intel® Xeon® processor 6500 series Intel® Xeon® processor 7500 series Intel® Xeon® Processor E3 Family Intel® Xeon® Processor E3 v2 Family Intel® Xeon® Processor E3 v3 Family Intel® Xeon® Processor E3 v4 Family Intel® Xeon® Processor E3 v5 Family Intel® Xeon® Processor E3 v6 Family Intel® Xeon® Processor E5 Family Intel® Xeon® Processor E5 v2 Family Intel® Xeon® Processor E5 v3 Family Intel® Xeon® Processor E5 v4 Family Intel® Xeon® Processor E7 Family Intel® Xeon® Processor E7 v2 Family Intel® Xeon® Processor E7 v3 Family Intel® Xeon® Processor E7 v4 Family Intel® Xeon® Processor Scalable Family Intel® Xeon Phi™ Processor 3200, 5200, 7200 Series Intel® Atom™ Processor C Series Intel® Atom™ Processor E Series Intel® Atom™ Processor A Series Intel® Atom™ Processor x3 Series Intel® Atom™ Processor Z Series Intel® Celeron® Processor J Series Intel® Celeron® Processor N Series Intel® Pentium® Processor J Series Intel® Pentium® Processor N Series
Camden Brown
>reverse engineering certain consoles architecture If you're thinking of recent consoles, they're all on AMD except Nintendo, so I don't think so. Meltdown affects Intel, Spectre would've affected AMD but AMD came right out and said no effect to 2 out of 3 of the variants. It's only the first spectre variant where if you're running a custom BIOS instead of default on certain FX machines that you're fucked unless it got patched.
Logan Butler
I haven't seen this outside of opensuse, but there's an amd microcode patch that disables branch prediction on Zen to harden against spectre, so uh, rip Ryzen performance.
Shit is so bad that even the unrelenting shintel shills have toned down a bit.
Josiah Kelly
>C2D not affected Comfy af using my Q6600
Andrew Clark
Wow, that's almost all the letters of the alphabet!
Nolan Smith
>Basically every indirect call (one instruction) turns into a seven-instruction sequence (retpoline)
do not assume i know what I'm talking about, but this wouldn't need to be that bad (going for the "term" sequence), or are you explicitly saying that for every instruction there will be a 6 extra cycles demand?
Lincoln Sanders
>no C2Q on the list never imagined I'd pull out my q9550 from the wreckage
Aaron Nguyen
Is this a seperate bug or the same bug? What's Spectre 1 and Spectre 2?
Elijah Smith
>Intel® Celeron® Processor N Series Processors that are already crawling going to be hit with a huge performance penalty. Oy vey!
Jose Wood
How much performance does branch prediction really add? What if Intel were to completely remove it instead of patching it in software?
Ayden Gray
I'm disheartened by the number of comments here who are taking the stance that Intel has idiot designers or that management doesn't care about security. This attack is very clever and unexpected, nobody could have predicted this. Intel still the dominant and most trusted industry player and will remain so in the future.
Kevin Williams
>those digits >this shill
Josiah Harris
Right there mate: 2nd generation Intel® Core™ processors
Austin Lopez
YFW
Eli Miller
>This attack > This attack it's a fucking design CHOICE
Parker Lopez
Did you test your system pre-after patch? Mine had some degradation on win10 after patch, using Q9550. That's Core ix-2xxx series
Jaxon Taylor
you've got it on point
someone give him an oscar
Jace Jenkins
Fucking kill yourself you idiot shill
Thomas Carter
>tfw Intel Pentium G series
Brandon Ramirez
The fuck is this webm? Any context?
Nicholas Long
>nobody couldve predicted this
i can think of someone
Jonathan Parker
See the 5th row. >2nd generation Intel Core I.e Core2 duo.
Christian Edwards
Forgot image
Luis Russell
a """girl" with a very very small penis trying to get a neovagina
Isaac Bennett
Witnessed, mein neger!
Jose Flores
That kid is a tranny. Google transexual dilatation.
Lincoln Clark
Actually this has been know since 2007 but swept under the rug.
Camden Adams
proof please?
Ryder Parker
Thank dog I bought a cheap g4560 in february and didn't spend in a more expensive CPU. Thanks ryzen hype I guess.
so did i fuck by upgrading my t430 to a quad core few days ago?
Evan Roberts
Meant to reply to
Justin Wood
>Allowing privileged reads in an unprivileged context >Not completely idiotic
Thomas Campbell
Emulation is CPU heavy. CPU usage gone up after patch.
Luis Ross
I'm not sure what this would mean for performance, but Intel's modern pipeline is about ~20 steps long, so I'd assume that it means any conditionals are gonna cost that many extra cycles, if the OoOE can't fill in the gaps in the pipeline.
Someone with ryzen should go benchmark the opensuse with the microcode, if they want to test how much a lack of branch prediction will affect performance.
Logan Harris
So was this vulnerability the reason why Intel has IPC numbers and will the fix lower IPC numbers
Ian Perry
bbs.archlinux.org/viewtopic.php?id=233233 Only PassMark from all of the tested use cases took a drastic hit performance wise. The tests have been performed by Phoronix. Can anyone provide additional benchmarks?
Robert Butler
...
Easton Turner
No, you're going to need the extra power now.
Logan Harris
He is wrong. They know that there was something they took 10 year to find it. You know. Because it was hard to find? Probably hard to put into practice too. Better stick to scamming old folk with fake paypal emails. less work.
Eli Myers
What CPU, emulator and game?
Joshua Richardson
That's for KPTI (Meltdown fix), this is different.
Eli Ross
No The worst is yet to come. There are more vulnerabilities screencap this post. Intel is still not telling how bad it really is.
Jeremiah Anderson
It goes all the way done to Pentium Pro from 1995, everyone on Intel is fucked.
Adam Morgan
Are you sure Core 2 Duo isn't affected? I've got Thinkpads that use them.
John Brown
If you don't use your donger much before The Surgery™, it can cause them to not have enough material to work with (since HRT shrinks the dingus).
Angel Roberts
>I'm disheartened Of course you are, you fucking corporate shill, since it means fewer will be tricked to use your corrupt company's shitty hardware in the future.
Justin Wilson
so, it hasn't even started yet?
Adam Flores
The fact that guy sold his shares tells me he expects more then just a dip. He expects a hard fall and he's bailing out. Otherwise, he'd ride it out.
Owen Moore
How badly and what processor is in it? I've disconnected my Core 2 Duo machines until there's more information out.
Jace Kelly
Amerimutt at it again. lol
Lincoln Fisher
>Your system either does not have the appropriate patch, or it may not support the information class required.
Sigh, I wish Alex would stop shilling and add a check for a system running on AMD.
Elijah Clark
Of fucking course
Ethan Hall
Exactly this. This one is very serious.
Caleb Butler
wrong img
Aaron Morgan
This looks extremely painful
Asher James
Underrated.
Eli Reed
For you
Daniel Brooks
One can only hope
Aiden Reyes
debian haven't patched yet so nothing
Brody Gutierrez
It's not on the list.
Grayson Gonzalez
Is Core 2 Duo on the list under a slightly less obvious name or not, lads? And is that list complete?
David Myers
The 3 variants includes both meltdown and specter-based attacks. I say "*-based" attacks because the specter attack vector still exists on every single CPU with speculative execution.
Alexander Wilson
It was predicted within the context of side channel attacks, that's why multiple researchers discovered Meltdown independently. It just took a while for the actual side channel techniques to catch up and create a practical exploit.
It's a little more subtle than that because the privileged reads don't affect the architectural state but information is leaked via side-channel. Of course it's still idiotic because side-channel attacks have been predicted for ages.
Nolan Cox
>At the time of writing, Google believes that "every Intel processor which implements out-of-order execution is potentially affected, which is effectively every processor since 1995 (except Intel Itanium and Intel Atom before 2013)" is affected by Meltdown.
