Well, for one thing, modern CPUs aren&#x27;t microcoded much. Real microcoded CPUs went out with the 1980s.  The microcode in today&#x27;s fast CPUs is mostly initialization, and maybe emulation support for some instructions nobody uses any more, like decimal arithmetic or 16-bit mode. The microcode can maybe turn some stuff off and tweak some tuning parameters. Anyone know what Intel&#x27;s &quot;fixes&quot; really do?

Said current patches haven&#x27;t been merged. Just posted to the mailing list for review.Retpolines for Spectre, and PTI for Meltdown, are still in the kernel. Linus doesn&#x27;t have too much disagreement with them. It&#x27;s the IBRS &#x2F; IBPB code for Intel&#x27;s CPUs that Linus has a problem with.Also, not everything can be mitigated in microcode, for many reasons, not all of which I understand myself. Sometimes, only a fundamental hardware-redesign can fix the issues proper.

So are we going to unmerge this, and force Intel to fix this properly in microcode?

Would you rather have the kernel filled with false or botched security?

Retpoline and PTI patches are fine.IBRS and IBPB patches are what will be delayed and maybe never merged. Linus reckons that the Retpoline code will be quite sufficient.

So, it means more delays in upstream patches.

Being able to call out bullshit doesn&#x27;t mean having to curse it out. All Linus needs to do is say something like &quot;I don&#x27;t understand why this is here, and I&#x27;m not merging it until I do&quot;, rather than &quot;They do literally insane things. They do things that do not make sense&quot;. The latter is not a technical argument, and it doesn&#x27;t provide accuracy or clarity about next steps.&quot;So somebody isn&#x27;t telling the truth here. Somebody is pushing complete garbage for unclear reasons&quot; - maybe the patches are bad. Or maybe there is an undisclosed vulnerability (like <a href="https:&#x2F;&#x2F;skyfallattack.com&#x2F;" rel="nofollow">https:&#x2F;&#x2F;skyfallattack.com&#x2F;</a> ?) that needs this &quot;garbage&quot; to mitigate it, and no-one got Linus properly in the loop. If it&#x27;s the latter, all this shouting and cursing about &quot;They do things that do not make sense&quot; has likely attracted the attention of people interested in such things...

I suspect this isn&#x27;t even Linus at &quot;full throttle&quot;[1], as it were, since he&#x27;s still censoring his &quot;fuck&quot;s.[1] <a href="https:&#x2F;&#x2F;lkml.org&#x2F;lkml&#x2F;2012&#x2F;12&#x2F;23&#x2F;75" rel="nofollow">https:&#x2F;&#x2F;lkml.org&#x2F;lkml&#x2F;2012&#x2F;12&#x2F;23&#x2F;75</a>

You’re not the only one, he may be brash at times but god damn we need him and maybe we need his brash anti-bullshit meter too.

I must say that I&#x27;m really happy that Linux is taking a stance on this one. He doesn&#x27;t care what the legal consequences to Intel are, he is pushing for a proper technical solution damned be the consequences.

Thats a pretty superior attitude you&#x27;re taking regarding a disagreement about which you admit you understand nothing

&gt; See the difference between IBRS_ALL and RDCL_NO. One implies Intel will fix something. The other does not.I think that&#x27;s a pretty interesting claim on his part. I can&#x27;t weigh in too much on the evidence to support this claim, but it&#x27;s certainly more than just a rant it seems.

Crazy black magic which is insane and completely unnecessary, according to Linus.Seems like Intel just wants to half-arse a solution.

I read it like:Intel will make a fix but the fix will be disabled by default as that results in the best performance numbers, those that will be reported in the media.When you want to use the CPU in the cloud, you have to enable the fix and get much lower performance. But nobody in the media will care about that as they have the nice performance numbers and therefore Intel don&#x27;t care to implement a fix with high performance.

I interpreted it as: Intel isn&#x27;t going to fix Specter in HW (as opposed to Meltdown)Instead, they do some crazy crazy black magic in the patches.

It is patch from Amazon, not from hardware vendor.

I also have zero criticism of Linus for his style. However to be fair, the point being made by GP is I think that if the frequency of severe patch criticism by Linus is high enough (&quot;not newsworthy&quot;), then the information content of all his criticisms is reduced accordingly.

&gt; Linus having issues with a patch is, well, not particularly newsworthy.Why not? The lead engineer of the linux kernel project has specific issues with security patches from a major hardware vendor.I understand the constant distaste that some users here have with Linus&#x27; style, but I don&#x27;t understand the constant need to denigrate his positions by attempting to paint them as particularly diminutive.

Particularly interesting; the opt-in creates an incentive for OS vendors to comply, as being the only platform to go all-in would create an impression to the end user of poorer performance using your product.

But are we talking about fixing it on current chips or future chips? If the former I understand Linus’ position only if there is a better solution. But if it is a hardware problem it may not. If we are talking about future chips, I thought that the fix even on 7th and 8th generations had a low overhead. Why would it have a bigger overhead in future designs?

I don&#x27;t completely understand the technical issues, but reading between the lines I think the patches in question add super huge overhead. The overhead is so huge that nobody in their right mind would enable them by choice. However, the behaviour is governed by an opt-in configuration at boot time. What this means is that by default the patches will do nothing but if you opt in, you will get a huge performance hit.The end result is that Intel can say, &quot;We&#x27;ve fixed the security holes&quot;. Linus can apply the patches and say, &quot;I&#x27;ve fixed the security hole&quot;. However no distribution will enable it because they will suffer from a massive performance penalty if they do. The end result is that the security hole will remain, but the average person&#x2F;manager&#x2F;etc will not be aware of it.As far as I can tell Linus is calling that out as being complete BS. From what I can tell, his reasoning is that if there needs to be a performance hit, then Intel can eat it. They will have to get their act together and fix the problem in upcoming processors or else basically nobody will buy them. The implication is that Intel has no intention of fixing the problem in the next generation of processors and is hoping to sweep it all under the rug -- they have a fix. It&#x27;s not their fault if nobody enables it.The situation is a bit delicate because if Linus forces the Intel processors to be slow, then Intel will blame him -- their patches are only slow if you opt in to some completely-not-understandable-kernel-patch-option. He&#x27;s basically refusing to play ball at all -- which is probably the correct stance (assuming I understand what&#x27;s going on -- insert disclaimer about being a random person from the internet here).I don&#x27;t think Linus is necessarily blaming anyone in particular. That&#x27;s why he says the reasons behind the patches are unclear. I think he&#x27;s hoping that causing bad publicity will get Intel to abandon their tactic of being politically careful. Getting slagged of by Linus carries a certain stigma and Intel may consider it prudent to offer up information more quickly in order to stem the damage.

man, its been ages since I&#x27;ve read his rants... he can be such a fun keyboard warrior

Run through the Linus drama filter, the only thing I get from this is &quot;Linus doesn&#x27;t like the Meltdown&#x2F;Spectre patches&#x27;; Linus having issues with a patch is, well, not particularly newsworthy. Is there something more significant here than a bad first go at this?

It worries me that this isn&#x27;t higher both here and on Reddit. I read through the rest of the email thread this morning and wanted to say something but I&#x27;m worried that I don&#x27;t understand it well enough.

Agreed, but I don&#x27;t think it changes the bulk of what he&#x27;s talking about.

FWIW, at least a small part of that email is due to a confusion between Intel acronyms: <a href="http:&#x2F;&#x2F;lkml.iu.edu&#x2F;hypermail&#x2F;linux&#x2F;kernel&#x2F;1801.2&#x2F;04637.html" rel="nofollow">http:&#x2F;&#x2F;lkml.iu.edu&#x2F;hypermail&#x2F;linux&#x2F;kernel&#x2F;1801.2&#x2F;04637.html</a>

Management can seriously make or break a company in various ways. It sucks... :&#x2F;

Yea... and what about every other chip maker who also have the same exploits? Can someone explain whether AMD and Qualcomm licensed this tech or if they also independently made the same mistakes? Either way, we really haven&#x27;t seen much activity from them either.

The part of the lifecycle where the clueless have taken over:<a href="https:&#x2F;&#x2F;www.ribbonfarm.com&#x2F;2009&#x2F;10&#x2F;07&#x2F;the-gervais-principle-or-the-office-according-to-the-office" rel="nofollow">https:&#x2F;&#x2F;www.ribbonfarm.com&#x2F;2009&#x2F;10&#x2F;07&#x2F;the-gervais-principle-...</a>

Intel transformed from a technology company to a management company. That somewhat explains their highly unreasonable behavior.

&gt; Not only spot, but also invests the energy in calling it out and explaining it.Which is why FOSS is so important. Can you imagine a world where Linux was closed source? You’d never have any insight into security patches, or these discussions.

We should treasure it while it lasts, before PC culture silences him. And before downvoting, recall how many threads regarding Linus abusive behavior (sic) we&#x27;ve had.

&gt; Not only spot, but also invests the energy in calling it out and explaining it.It also requires the social capital necessary for the calls to be taken seriously.I don&#x27;t worry about Linus being silenced by prevailing culture; I worry that there won&#x27;t be another Linus after him.

&gt;Many business failures i&#x27;ve witnessed happened ultimately because nobody had the energy to cut down the bullshit forest that constantly regrows around a company.I&#x27;ve worked at a company that had 20 people and a company that&#x27;s had 5000 and some ones in between.Going from 2-3 people to 20 is pretty manageable. Everyone knows each other, work gets done, you can call people out on their bullshit, stuff is fixed, because everyone feels responsible and accountable.Then you go to 100 people and a few of the original 20 leave and the whole culture changes. 1000&#x27;s of people? Totally different ball game. Incompetence will slip through, no single person has a full grasp on everything, meetings and more meetings and more managers playing broken telephone all day and you got people in different cities and time zones and countries who&#x27;ve never seen each other face-to-face and it&#x27;s just a nightmare to manage that.It can be done, but that&#x27;s when you really need competent managers, and those are extremely hard to come by.I&#x27;m sure there are teams and divisions within Intel that are extremely competent, people at the top of their game who are looking at this debacle and just shaking their heads just as much as Linus.

&gt; Not only spot, but also invests the energy in calling it out and explaining it.&gt; Many business failures i&#x27;ve witnessed happened ultimately because nobody had the energy to cut down the bullshit forest that constantly regrows around a company.The sad thing is that people have unironically labeled Linus &quot;toxic&quot; for calling out crap very bluntly like this.

Not only spot, but also invests the energy in calling it out and explaining it.Many business failures i&#x27;ve witnessed happened ultimately because nobody had the energy to cut down the bullshit forest that constantly regrows around a company.

I&#x27;m not sure. Finns (and swedes) are considered introverted, with shoegazing issues. That kind of directness looks like a personality trait (acquired maybe) and not default.

&gt; From what I gather, some of this cultural. His country of origin is just generally blunt and direct about things.I&#x27;ve heard this blunt(&#x2F;direct&#x2F;honest) argument quite often (and the same about my country, The Netherlands), but I&#x27;ve also seen Americans in IT who can be very blunt. I attribute it not to country_of_origin, but to the [autism] spectrum.

I&#x27;m not sure you replied to the correct post since it didn&#x27;t really seem to be about Linus&#x27;s style.Anyway, this argument comes up again and again. Regardless of the style in his country and language or origin he is writing in English, in public, for an international audience. One expects a different sort of tone.

&gt; From what I gather, some of this is cultural. His country of origin is just generally blunt and direct about things.Maybe he just is what he is. The differences between individuals are greater than differences between cultures. I&#x27;m from Finland and I can assure you that Linus&#x27; style of communication is not a norm here.

As a fellow Finn, I agree that our country is probably much more blunt and direct than the impression I get how the US works. That&#x27;s also the way I like it, and I think you have gone way too far in the US in this politeness thing.However, Finland is culturally quite homogeneous, so I think the reason behind the bluntness is not what you suggest it may be.For a Finn, I think Linus is on the blunt side, but he&#x27;s not an extreme case. In fact, I&#x27;d argue he&#x27;s not an extreme case even among non-Finns; it&#x27;s just that most people only hear about his rants.Also, I find that Linus is somewhat quick to judge things, and often his reactions are based on a misunderstanding of the facts.

From what I gather, some of this is cultural. His country of origin is just generally blunt and direct about things.That directness is generally associated with multicultural environments where subtler context-dependent communication simply does not work. Some of the quotes I am seeing here from him are things I think are unnecessarily harsh, but most are not.Discussions about something he has said or written seem to inevitably comment on his speaking style and divide up between those who praise it and those who vilify it. I feel it is mostly a positive, but discussions could benefit from a little context and broader perspective of that detail.

Courage is in shorter supply than genius.

Linus, like Jobs, is a bully.He lives under the sword. The moment he stops being right, everyone will swiftly turn against them.What separates them is that they&#x27;re almost never wrong. Not when they pull out &quot;this is garbage&quot; card.This is a difficult ability to learn. You need to stay quiet most of the time, then be a complete asshole when it&#x27;s important.I really hate that aspect of our industry. But that&#x27;s just hating on humanity. Apes listen to apes that shout, but only when everyone else is listening too. Otherwise you&#x27;re just a dick with no influence.The key is to hate that Linus is a bully while also respecting him when he&#x27;s right.It&#x27;s super frustrating because I feel like any of us could have written that, and no one would care. Literally no one would listen. But Linus says it, so it bubbles to the top of HN and now we&#x27;re counting our blessings for how lucky we are.Try to remember this next time someone says something that sounds crazy. Your instinct should be to suppress your skepticism and think whether it has merit. It might piss you off, but is it true? That&#x27;s the important bit.And not just for design, but for everything.(I&#x27;m sure whoever came up with the current design isn&#x27;t too happy about Linus calling it garbage, but they&#x27;ll just have to deal with it.)

Actually that&#x27;s pretty scary. Are the rest of heavy weight kernel code maintainers too stupid for than then? If it weren&#x27;t for Linus&#x27; hard judgement? Would we then buying badly engineered products without ever questioning?

I think you are very harsh with &quot;they’ve shown they can’t be trusted with quality&quot;. Their product is incredibly complex, it&#x27;s amazing only so few mistakes like this happen.I totally agree with &quot;they can’t be trusted to be transparent or even take ownership of the problems they create&quot; though. When you play at Intel&#x27;s level and mess up on that order of magnitude, doing what they did make you loose all trust.

There&#x27;s no question that Intel&#x27;s response is entirely dictated by the legal ramifications of the coming lawsuits....but irrespective of that, it ALSO appears that there&#x27;s no leadership within Intel to actually fix the underlying issue.I don&#x27;t care if the PR is BS - that&#x27;s what I expect. But if the chips are not fixed the right way, I&#x27;m taking my business elsewhere, and I&#x27;m not alone.

I full agree with your sentiment.I also see Intel lawyers working overtime, closely interacting with marketing, to somehow ward of lawsuits coming from Google, Microsoft, Amazon, which must be seeking ways out of Intel contracts to turn towards AMD processors for their currently more attractive TCO characteristics.Could be the end of Intel as we know it.

You are far from alone, especially given the behaviors and attitudes of the most likely successor...

I fear for the future of post-Linus linux.

There&#x27;s lots of people who call out bullshit. But people usually don&#x27;t listen or respect them unless they have power - which Linus has. There are plenty of other professionals who have been complaining lately.

&gt; Intel &amp; friends have absolutely disgusted me lately [...] - they can’t be trusted to be transparent or even take ownership of the problems they create.Well, it&#x27;s the natural outcome of a system based entirely on competition. When the only incentive is &quot;beat everyone else, nothing else matters&quot; then this eventually happens.

I think we should count ourselves lucky we have someone so good at spotting bullsh*t like this - and isn’t afraid to call it out as they sees it.Intel &amp; friends have absolutely disgusted me lately, it could be all to easy to settle for less based on their standards, but let’s not let them set the standard - they’ve shown they can’t be trusted with quality and when they fail - they can’t be trusted to be transparent or even take ownership of the problems they create.

Maybe I wasn&#x27;t clear. Your explanation and the linked article is very informative, but I wanted to understand what&#x27;s the &quot;garbage&quot; Linus is talking about. As I said, I do understand retpolines from a high level.

Here&#x27;s an explanation of retpoline:<a href="https:&#x2F;&#x2F;support.google.com&#x2F;faqs&#x2F;answer&#x2F;7625886" rel="nofollow">https:&#x2F;&#x2F;support.google.com&#x2F;faqs&#x2F;answer&#x2F;7625886</a>“Retpoline” sequences are a software construct which allow indirect branches to be isolated from speculative execution. This may be applied to protect sensitive binaries (such as operating system or hypervisor implementations) from branch target injection attacks against their indirect branches.The name “retpoline” is a portmanteau of “return” and “trampoline.” It is a trampoline construct constructed using return operations which also figuratively ensures that any associated speculative execution will “bounce” endlessly.

(FYI you&#x27;re missing your [1] reference.)

&gt; Intel has no good hardware or microcode mitigation for spectre-like attacksI was under the impression that for at least one category of attack, there is no hardware mitigation possible because it&#x27;s a fundamental problem with the x86-64 design. Fixing it would require building a chip that uses some other architecture. Is that not the case?

Linus seems to have two complaints:1) Recent patch submissions imply that Intel has no good hardware or microcode mitigation for spectre-like attacks. There is a sub-complaint that Intel has a bad (i.e. kills performance) fix, but will not enable it by default because benchmarks matter.2) This series of patches in particular appears to be doing either something different, or more than what their descriptions implyThese patches do various things, presumably to manipulate the opaque internal state of the CPU, but only Intel knows for sure precisely what they do.

This doesn&#x27;t make sense though. If Intel indeed plans to leave chips vulnerable except if you set a flag, then how are these bullshit patches? This will be the only solution Intel is going to deliver right? I get that he doesn&#x27;t like it, but that doesn&#x27;t explain why he feels lied to. If Intel says &quot;we&#x27;re going to not turn the patch on by default&quot;, wherein does he suspect the lie?

Awesome explanation. That is indeed deserving of the word &quot;f*cked&quot;.My next CPU will be an AMD then.

The problem is that every fix that you could think for Spectre reduces the performance of the CPU.So not enabling this by default it&#x27;s a good choice, Spectre is very difficult to exploit: so if you do critical things you enable the fix, if you use the computer for gaming, video rendering, and things where you don&#x27;t care too much about security but you care about performance, you don&#x27;t enable it.

Wow.  That&#x27;s outright malicious behaviour from Intel there.

It&#x27;s not entirely clear to me, but in some of the followup emails it appears that Linus was mistaking the purpose of some patches (or flags at least) due to unobvious naming, but I&#x27;m unsure if that significantly alters his criticism.  He says it  still applies, but is much more muted in tone about it (and whether that&#x27;s from the original email possibly not being intended to by public or not, I don&#x27;t know).

Thanks for your informative explanation. However…&gt; &quot;boot an older os and it&#x27;s vulnerable since it doesn&#x27;t know to call this&quot;Presumably the hardware that fixes this is not even available on the market yet. How likely is it that somebody will go out of their way to install an obsolete OS version on their brand new hardware?

&gt; In other words, intel is using its presence and weight to try and push a shitty solution, but one that is better for them marketing wise. Linus is flabbergasted to be treated like an idiot or a obedient drone that should apply such obvious abusive patches.I see where the brashness comes from. Shady dealings on Intel&#x27;s part.

Linus also says that this shows Intel means never to fix Spectre2.  Of course, that would only be their current position -- they could change their minds later.  That strongly implies that the decision to disable by default is a marketing decision, but take this with salt -- it could also be a bad engineering decision.

Big simplification:Proper way to fix an hardware bug like this, is that newer cpu gets protected by default, and they answer they are when queried.So you can ask the CPU &quot;what&#x27;s your status on bug X&quot; and the cpu answers &quot;i&#x27;m good, you don&#x27;t need to do anything&quot; (newer fixed chips), or &quot;i know about it but was already built, and need microcode update&#x2F;special behavior to protect myself&quot; (current chips with microcode update), &quot;no answer &#x2F; I&#x27;m not good&quot; (old chips without update).So new stuff is protected, and you add more protection (and slowdowns, and special stuff) for older chips that don&#x27;t know how to deal with it.What Intel is trying to do here, is to go the other way: the chips, even the new ones, will stay vulnerable by default, and when queried they say &quot;I have a fix but I don&#x27;t use it, you can enable it by asking !&quot; and the kernel is supposed to enable it.It&#x27;s terrible for a lot of reasons, like &quot;boot an older os and it&#x27;s vulnerable since it doesn&#x27;t know to call this&quot;, &quot;additional code to enable this feature has to run for all of eternity for new chips now, instead of having to run for older chips and being phased out over time&quot;, etc ...The reason why Intel does that seems obvious: by default the chip does not lose speed since the fix is not enabled, and so instead of &quot;intel chips lose 30% speed over night because of a flaw&quot; it becomes &quot;intel adds a special security mode that protects you even more for critical applications, at the cost of some speed&quot;. 
Purely marketing speech and decision at the cost of proper engineering decisions, and they need and try to get OSes like Linux to play along. 
That&#x27;s what he means by &quot;[it] shows intel had no intention of fixing those flaws&quot;.Additionally there seems to be a second issue in that the quality and behavior of the patches they submitted are trying to hide this deceptively simple but technically terrible behavior by making it look&#x2F;sound obtuse and complicated.In other words, intel is using its presence and weight to try and push a shitty solution, but one that is better for them marketing wise. Linus is flabbergasted to be treated like an idiot or a obedient drone that should apply such obvious abusive patches.

There&#x27;s a follow up email that has much more details. Not ELI5 though.&gt; But since the peanut gallery is paying lots of attention it&#x27;s probably worth explaining it a little more for their benefit.<a href="http:&#x2F;&#x2F;lkml.iu.edu&#x2F;hypermail&#x2F;linux&#x2F;kernel&#x2F;1801.2&#x2F;05282.html" rel="nofollow">http:&#x2F;&#x2F;lkml.iu.edu&#x2F;hypermail&#x2F;linux&#x2F;kernel&#x2F;1801.2&#x2F;05282.html</a>

I&#x27;m going to be honest, I need an ELI5 here. I know what the meltdown&#x2F;spectre issues are, and kinda understand what retpoline is.

No worries at all, mailing lists are a PITA to navigate at the best of times.

I, too, missed the double-indenting at the start and thought David was replying to Linus. Once I got to the textbook swearing I double-checked and figured it out.

Thanks for this, the beginning where he quotes himself threw me off.

Thanks for pointing that out - I was under the impression he still worked for intel, before commenting I checked and all the top Google results suggested as much, I’ve corrected my comment.

This isn&#x27;t the case, he works for Amazon. Unless he also works for Intel but that seems unlikely.Edit: Just to make this clearer you can see his email in the Signed-off-by of the patch under discussion: <a href="https:&#x2F;&#x2F;lkml.org&#x2F;lkml&#x2F;2018&#x2F;1&#x2F;20&#x2F;163" rel="nofollow">https:&#x2F;&#x2F;lkml.org&#x2F;lkml&#x2F;2018&#x2F;1&#x2F;20&#x2F;163</a>

* Principal Engineer - Amazon Web Services - Employed Oct 2016 – Present - Employment Duration 1 yr 4 mo* Intel Corporation - Linux Kernel Engineer - Employed Jun 2008 – Oct 2016 - Employment Duration 8 yrs 5 mosSource: LinkedIn - <a href="https:&#x2F;&#x2F;www.linkedin.com&#x2F;in&#x2F;dwmw2&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.linkedin.com&#x2F;in&#x2F;dwmw2&#x2F;</a>Note: Also, this is not and please do not turn this into a personal witch-hunt or anything like that, I am just adding context &#x2F; clarification as to the mailing list thread.

In case it’s unclear to anyone, David Woodhouse (the person Linus is replying to &#x2F; previous poster) works for Amazon and previously Intel[1] aka @dwmw2 on Twitter, ironically his twitter profile is: “Kernel hacker. Known to occasionally promote an attitude of violence towards complete morons.”*[1] Correction, I thought (as does Google) that he still worked at Intel, but it was pointed out that he now works at Amazon UK.

I suspect it has more to do with the American culture. Every time I watch something American, be it a TV show or YouTube from a big name, things are constantly bleeped out or you notice they don&#x27;t mention swear words. It&#x27;s super annoying - either accept it and stop beeping, or stop swearing all the fucking time. It&#x27;s also not as if you have no idea what that beep meant.

They are not fooling anyone.They are fooling the content filters some corporate services use to block emails though.

Why do people censor words like that? They are not fooling anyone.

You&#x27;ll get a pass at being impolite if you are technically in the right...Here Intel fucked up so they should be really really polite instead.

If you want something from someone you ask it politely. If you produce BS, you get told you produce BS. I&#x27;m guessing it&#x27;s a cultural thing I&#x27;m also guessing you are an American. Americans coming here often tell us we are very direct and it takes getting used to. We often tell Americans we prefer they speak their mind and don&#x27;t pretend something is great or we are best friends when it isn&#x27;t the case. I&#x27;m from the Netherlands but I&#x27;m guessing Linus has his Finnish roots showing from time to time.Honestly, it&#x27;s nice in the US to have waiters&#x2F;waitresses that are nice, it feels insincere but I prefer it greatly to the sometimes outright moody waiters here. It&#x27;s a culturing thing as well.If Intel would be sincere, they would just admin they produced BS (Linus is the best person to call BS on such patches) and apologize to Linus for wasting his time and ask how he would like to see the patches.

I’d expect any such approach to take place off list

There&#x27;s more ways to be impolite than saying nasty words.Linus had been treating all this situation very fairly so far, but I guess Intel just had to repeatedly try pushing very bad fixes to a problem they created themselves.Imagine you go to a restaurant, and the waiter just repeatedly brings you the wrong order, repeatedly.

Why should they ask politely? It is not like the Linux mailing list would be polite environment.Nothing against impolite environment, people make case for it daily here and I don&#x27;t care. But it combined with the standard expectation that when someone else talks to them he should be always polite is getting funny in that context.

Ah, if they stopped now, imagine the PR nightmare.

Note that Intel is a platinum member of the Linux Foundation, which means they donate &gt;=500k$ per year.

Because there is no proper fix. The only fix comes with massive slowdown.

Why don&#x27;t they just work on a proper fix instead? I&#x27;m sure Linus would be glad to integrate a proper solution, but it&#x27;s clear that his opinion is that Intel is instead bringing &#x27;garbage&#x27; to the table.

Linus is also not actually qualified to solve this problem. He&#x27;s the guy we need to call bullshit, but as you said, Intel already knows what the problem - and the solution - is. Linus is simply pointing that out (which is certainly good).

Intel knows how to fix the problem, they don&#x27;t really need Linus. What they do need is to accept a slowdown for some of their processors, but they are absolutely terrified to look bad compared to AMD in the benchmarks.

Why doesn&#x27;t Intel just go to Linus and ask politely: &quot;Ok, we f-ed up, here is a million dollars (or whatever, 10x his normal salary or something) can you please work with us tightly for some time? Thank you.&quot;I would really respect Intel if they would.

But it would certainly be good to have some context on the quote (i.e. that it has to do with Linux, Intel and Meltdown)

Then you should have posted the comment that you did.Also, cherry picking a quote and putting it in submission title is against HN guidelines.

That&#x27;s a good blanket rule, but Linus&#x27; comments are uniquely important. He is a subject matter expert and context is important regarding who is doing the yelling. This isn&#x27;t some celebrity who is famous for something irrelevant.

I did that. We&#x27;ve observed over the years that celebrity names in titles tend to make for poorer discussion. HN is better when the focus is on content rather than personalities.<a href="https:&#x2F;&#x2F;hn.algolia.com&#x2F;?sort=byDate&amp;dateRange=all&amp;type=comment&amp;storyText=false&amp;prefix=false&amp;page=0&amp;query=by:dang%20content%20personalities" rel="nofollow">https:&#x2F;&#x2F;hn.algolia.com&#x2F;?sort=byDate&amp;dateRange=all&amp;type=comme...</a>Edit: ok, you guys have convinced me that this is a special case and we&#x27;ve put Linus back in the title above.

Whoever changed the title managed to make it less clear. With &quot;Linus&quot; in the title it was clear that it was Linux.

Alternatively, it could also be in the interest of many for these patches to have an inordinate negative affect on performance with Intel CPUs.I am reminded that Linus has experience in the CPU industry (Transmeta), so he is in a position to see both sides on this.

I could be very wrong but I read it as Intel trying to cover the fact that there are huge performance penalties with the patch enabled. Therefore Intel will continue to market chip performance sans patch while pushing down the responsibility of enabling them to OS vendors.

Why did you pick that specific example (power leakage)? Is there a proof of concept that does something similar?

The root question is what else is Intel trying to cover up with these garbage patches? Are they afraid of power leakage across gates allowing an attacker to gain a higher level of privilege in certain generations of silicon, and trying to cover it up with these patches (hence some of the seemingly crazy things they do)?

That bit sounds to me like Intel is trying to pull a &quot;Volkswagen&quot;: have it perform better in benchmarks than in real life (when hopefully secure execution will be enabled).

&quot;Intel: Where Security is Optional&quot;

He&#x27;s quite clear about his theory as to Intel&#x27;s motives:&gt; The whole IBRS_ALL feature to me very clearly says &quot;Intel is not
serious about this, we&#x27;ll have a ugly hack that will be so expensive
that we don&#x27;t want to enable it by default, because that would look
bad in benchmarks&quot;.&gt; So instead they try to push the garbage down to us. And they are doing
it entirely wrong, even from a technical standpoint.&quot;

Linus calls people stupid when they are usually being stupid, at least in his eyes. He doesn&#x27;t wantonly accuse people of stupidity for no reason. It&#x27;s just that when he perceives stupid activity... well, he generally goes off. It&#x27;s his main trigger.

That&#x27;s probably a better fit for Heinlein&#x27;s Razor:<a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Hanlon%27s_razor#Similar_quotations" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Hanlon%27s_razor#Similar_quota...</a>

Hanlon&#x27;s Razor is stupid... actually I misspoke. It&#x27;s just evil.

Hanlon&#x27;s razor should include an exception where PR and politics are involved. Discounting malicious people as just stupid is the reason so many &quot;stupid&quot; people are in power.

Wonder if perhaps Intel is under classified and gagged duress from the government? There has been plenty of evidence the government is not acting with citizen security foremost in its technical and telecommunications policies.

&gt;The most striking thing here is that Linus has apparently dismissed incompetence as a rational explanation. Yes, he is often brash, but usually he is accusing someone of sheer stupidity. He does not do that here. Linus alleges that we are being lied to - that we don’t know the full story, nor Intel’s motives.&quot;And that&#x27;s actually ignoring the much _worse_ issue, namely that the
whole hardware interface is literally mis-designed by morons.&quot;Maybe you missed this line? Some classic Linus right there...

The most striking thing here is that Linus has apparently dismissed incompetence as a rational explanation. Yes, he is often brash, but usually he is accusing someone of sheer stupidity. He does not do that here. Linus alleges that we are being lied to - that we don’t know the full story, nor Intel’s motives.Furthermore, we are left to wonder if Microsoft is also being fed “bullshit” patches, and if they may be less discerning than Linus regarding a proper solution.

Are there any people who care about what the “technology media” says?

Has a major technology media outlet taken a similar position as Bloomberg and The Economist?

Also if there are huge changes at the top of the internal food chain a huge corp, which is already slow by definition, will react even slower. This problem set is likely to extend more before it gets better.

Well, Torvalds has been heard from, along with Bloomberg and The Economist, as saying that Intel is in denial about this being a big, big problem.

The first reading I had the same trouble, but you should watch the &gt;&gt; (Linus&#x27;s previous post) and &gt; (David&#x27;s post) to figure out who said what. And this post is by Linus replying to David.

I don&#x27;t know, I&#x27;m 27 and a C++ programmer and found it difficult to figure out who&#x27;s talking at the time. I&#x27;m just more used to more modern forums where each &quot;block&quot; is wrapped in quote tags with signature on top.

People live in Slack and github today. The last time I had an actual conversation via email was probably 2 years ago.It&#x27;s rather refreshing to see the Linux kernel keeping the tradition of using mailing lists. Every company I&#x27;ve worked at has felt the need to gather a bunch of people in a room (or teleconference) and &quot;knock things out&quot; in real-time. Then, everyone leaves and develops a severe case of amnesia. Never mind that for technical debates you really need some time to think and reply in detail. Which Slack is not good at, but... surprise, email is. But alas, it&#x27;s &quot;too old.&quot;

The issue I have with most mailing list interfaces like this is how difficult it is to find context. This post is a perfect example, it&#x27;s pretty easy to find David&#x27;s email by clicking the &quot;In reply to:&quot; link but I&#x27;ve been searching for a good 10 minutes and I still can&#x27;t find Linus&#x27; original email. Going to &quot;Messages sorted by: [ thread ]&quot; doesn&#x27;t show it, going to &quot;Messages sorted by: [ date ]&quot; and looking for emails sent around 2018-01-21 11:34 doesn&#x27;t show it, and short of searching through every single one of Linus&#x27; emails I don&#x27;t really know how to find it.

It&#x27;s easier to read in a proper email agent, which usually color-code quotation levels, allows quotation collapse, and  similar things. But that means you&#x27;re an old fart reading their email in an actual, dedicated application instead of some webmail :)

Probably depends on how old you are. To me it was completely obvious and much, much clearer than most &#x27;modern&#x27; user interfaces. Every person replying prefixes each line with &gt; . Proper email clients do this automatically. This way you can tell from the number of &gt;&#x27;s how far back on the &#x27;stack&#x27; you have to go to know who wrote it. You start reading from the top, mentally associating an indent level (nr of &gt;&#x27;s) with a person&#x2F;email address. It&#x27;s easy to follow any mail with up to say 3 or 4 different people; after that it takes some thinking sometimes, in long emails.I&#x27;m actually surprised at the number of people here complaining about the legibility. This is absolutely basic knowledge; or maybe I&#x27;m just even older than I imagine myself to be. How can someone who can&#x27;t even figure this out write software? (I realize that not everybody on this site is a programmer nowadays, but still)

Well, &quot;modern&quot; technology does this too; Markdown quoting on many sites works exactly like that - the only difference is that mailing lists prefer not prettifying messages. This email would look like this if copied directly to Reddit (and GitHub):<a href="https:&#x2F;&#x2F;i.imgur.com&#x2F;eXODVHb.png" rel="nofollow">https:&#x2F;&#x2F;i.imgur.com&#x2F;eXODVHb.png</a>(which also reminds me of how Tumblr&#x27;s reblogs look: <a href="https:&#x2F;&#x2F;i.imgur.com&#x2F;e6UR8bn.png" rel="nofollow">https:&#x2F;&#x2F;i.imgur.com&#x2F;e6UR8bn.png</a>)

Even if one were confused as to who said what, the combination of &quot;shit&quot; and &quot;f*cking insane&quot; should have cleared it right up.

Linus Torvalds originaly said:&quot;All of this is pure garbage. Is Intel really planning on making this shit architectural? Has anybody talked to them and told them they are f*cking insane? Please, any Intel engineers here - talk to your managers.&quot;David Woodhouse replied with:If the alternative was a two-decade product recall and giving everyone free CPUs, I&#x27;m not sure it was entirely insane.&quot;The rest is written by Linus, with all further quoted segments being something David Woodhouse wrote.&gt; &quot;the user interface is horrendous&quot;The user interface is perfectly clear if you know how to read it.

&gt; Am I the only person who finds this extremely difficult to read?I found it clear in this case.I&#x27;m often confused by mailing lists and nobody in the HN thread mentions anything, but this time it was 100% clear to me on the first read and I see lots of complaints.

Am I the only person who finds this extremely difficult to read? The top says &quot;David wrote&quot; and then it looks like he&#x27;s replying to a message from Linus Torvalds. So any quoted text is by Linus, right? But then....the letter is signed by Linus? So who wrote it? I know this is some ancient technology but Jesus, the user interface is horrendous.

It is possible the patches are submitted on behalf of others.

&gt; Have you _looked_ at the patches you are talking about? You should have - several of them bear your name.This made me so nervous on his behalf

Meltdown is almost trivial to take advantage of, as these things go, and it was discovered independently by more than one researcher. That makes it somewhat more likely to have been exploited by a government level attacker.However, just being able to read all of kernelspace is more of a force-multiplier than being useful on its own. You still need to get your code executing in user-space to do anything, ant it&#x27;s value even then is at least partly tied up in its ability to easily defeat KASLR.

Why does HN not allow people to question things like this? Any attempt to bring some non hyperbolic view to the topic is continually voted into the ground.I think it is useful to know that these exploits just might not be as bad as they are being made out to be.Ice talked to a number of people that don&#x27;t seem to understand that there are limits on these: somebody cant just start reading data from random programs.

<a href="https:&#x2F;&#x2F;github.com&#x2F;dag-erling&#x2F;meltdown" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;dag-erling&#x2F;meltdown</a>One of many source codes online and they are actually pretty understandable by me. I am just a mediocre c programmer.

As far as we know, there has never been a single attack using either of specter or meltdown issues. No code had been found, nothing.It appears to be very difficult to take advantage of, and the initial idea of a JavaScript exploit seems to not be possible.Part of the reason for all of this might be that the attacker needs to run code on the target machine, and one the attacker can do that, there are far easier ways to read memory.

Are state actors or others trying to take advantage of meltdown needing fixes &amp; patches to insert their own version of fixes &amp; patches that end up benefiting them?

It&#x27;s upsetting that the media and some of the mainstream audience see Linus as a &quot;crank&quot; that only exists to complain and tear people down. Examples like this show how valuable his experience is. I would NEVER ask him to water himself down knowing it would likely make his technical analysis worse.

You and I both know there&#x27;s a (large) middle ground.

Like it or not, the common expectation is moving towards what you (sarcastically) wrote, and already much farther away than where Linus is at here.Linus gets away with it because he&#x27;s Linus. And surprisingly, I hear even he has toned down a bit.

Do you think Linus would get a maximally friendly reception if he said:&gt; You clearly mean well. Your heart is in the right place. Thank you so much for all your hard work. There&#x27;s this weird thing going on where it seems like a bunch of patches experiencing opportunity for improvement are being pushed. These patches also could maybe be better explained.&gt; I don&#x27;t understand. Can you help me understand, so we can fix our problems together? Again, thank you so much for your hard work!And, incidentally, does the above not just drip insincerity to you? It does to me, and I know how powerful the Dale Carnegie-esqe &quot;complement sandwich&quot; approach is.Linus is justifiably angry at someone screwing with code he is responsible for in ridiculous ways, without any real explanation of why a ridiculous approach was chosen. He&#x27;s maybe a smidge less than fully satisfied with either point. The person in question really should know better on all points.

Question is, is it constructive? Does ranting really yield better results than just stating the facts (both short term for fixing the problem and long term for working together)?It might work for Linus on the LKML, but it certainly doesn&#x27;t for everyone and in all circumstances.

As a Kiwi, living in Australia I totally agree and understand, during code review in ops if something crap - I’ll things like ‘hmm, this looks a bit crap, maybe refactor for improved readability?’ Or ‘this has code smell written all over it, we can’t accept this kind of quality into production’ - I could just as easily say ‘Ross, this is shit - just rewrite it’ and my not so hypothetical Ross in this situation would read that - probably laugh, agree and rewrite it the next day OR ask for clarification stating why he doesn’t think it’s shit.My point is, what matters is the product, the code, the result and the quality there of. The communication around it is entirely based on the relationship and expertise of those around you.

I think you&#x27;re hitting a cultural barrier.In my country, if someone brought me something terrible, in a professional environment, I would tell them it&#x27;s bullshit, and go try again. And that would be fine and dandy, and expected. (I&#x27;d expect the same treatment of myself. And it has happened.)I get the feeling, that in the US, you can&#x27;t criticise. You have to use some political language to say you like it, whilst meaning you don&#x27;t.Linus is Finnish. He says what he means. And it&#x27;s expected, and entirely acceptable.There are a handful of nations around the world, where saying exactly what you mean is far more valued than trying to preserve whatever emotion someone has invested into their work.

Really abusive language. There are ways to communicate his objections to the patches with extremely strong language WITHOUT resorting to abusive language.I wouldn&#x27;t tolerate anyone treating me like this. There are unseen consequences to a wild tongue.

Linus is a pure techie.No politics, no bullshitting around, your code sucks and he calls you out for it.The best kind of dictator to have.

&gt; I can see it very slowly starting to unravel and quality issues starting to sneak inI know what you mean here, and I agree with you 100%.That being said, I can&#x27;t help but chuckle because if you&#x27;ve ever looked at a vendor&#x27;s GPL source code for the kernel (Android phone, NAS&#x2F;router) you know it&#x27;s the SoCs vendor&#x27;s old SDK (lots of bodge) with the ODMs additional bodges for their product customizations.Yes, I am VERY glad that the mainline kernel has Linus and his nose for BS. But just go look at the vendor forks if you want to see some even worse shit than these patches!Allwinner&#x27;s GPL is basically garbage. You can find quality rants from tkaiser (Armbian) about this.MediaTek doesn&#x27;t even release theirs! Yes, this IS a GPL violation, and no they don&#x27;t seem to care.At least Amlogic seems to have woken up and realized an active kernel development community is an asset for them.

What about *BSD ? How are they managing their kernels and other important decisions?I know next to nothing about the BSD communities, my only interaction with the BSD is discussing the taste of a pizza Theo de Raadt and I shared about 20 years ago and young me had no idea who he was. I later heard he is at the helm of OpenBSD and is used to strongly defend his opinions but it is only hearsay.I have no idea how FreeBSD or OpenBSD communities are organized but I suspect they might not be relying on one specific individual for their major decisions?

It does make me worry about the day that Linus retires. I imagine the kernel has enough momentum to carry on for a while but without a leader like Linus I can see it very slowly starting to unravel and quality issues starting to sneak in, much like we&#x27;re seeing with post-Jobs Apple.Thankfully it doesn&#x27;t seem that time is coming any time soon.

Linus is angry, that always makes me happy.Lord knows where we would be today if we had leaders like this more frequently through history.

Are we reading the same thing? He states very clearly the specific technical issues and behaviour from Intel that he perceives to be &quot;FUCKING BULLSHIT&quot;.

&gt;&gt; But there&#x27;s a reason this style isn&#x27;t used more often in politics&#x2F;law&#x2F;philosophy or any other profession trying to change peoples&#x27; minds: it really doesn&#x27;t work.The United States has a president who literally just did that for an entire year.

I get that Linus is adored for &quot;rejecting political correctness&quot; or &quot;speaking truth to power&quot;, but this email just shows how little such rants can achieve: he makes almost no argument, relying on his authority instead.i&#x27;m actually not quite sure what he&#x27;s trying to achieve... There&#x27;s no chance that a patch will make it into the kernel against his objections, so he doesn&#x27;t really need to convince anybody.If he&#x27;s trying to make a case against Intel, he&#x27;s unlikely to succeed with the digital version of screaming expletives at people.I have absolutely no &quot;moral&quot; objections against the use of all the FUCKING BULLSHITS anybody wants. But there&#x27;s a reason this style isn&#x27;t used more often in politics&#x2F;law&#x2F;philosophy or any other profession trying to change peoples&#x27; minds: it really doesn&#x27;t work.If anything, these rants show that Linus is unable to effectively use language to make his arguments. It makes him appear small, weak, and angry.

I don&#x27;t think its possible for a CPU to read memory without going through the cache.
Moreover, the value of speculative execution drops really hard if you can&#x27;t use any cache, as it means you are restricted to operating only on registers.

Flushing the cache after every branch misprediction would probably make branch prediction a net loss in performance.

more or less... I&#x27;m surprised the CPU doesn&#x27;t already flush caches after speculative execution branches are abandoned

From what I understand of meltdown and spectre, they use a side channel attack to figure out CPU cache values after some speculative execution has been done. Can it not be solved by ensuring that the CPU either clears or purposefully does not cache memory that is read during speculative execution?

I feel like a lot of people judge Linus by some internal moral compass which just seems pointless to me. If anything is petulant and childish it&#x27;s trying to control others for not fitting within some arbitrary set of personal rules.Linus does a good job at maintaining kernel development and often makes sharp technical commentary. By those measures his communication style seems to have been successful and practically helpful so far.

As great as it is to sit in your chair and make snide remarks about him. The Linux Kernel seems to have a proven track record of success under that roleThat is to say, it is the largest collaborative software project in human historyNo amount of Linus being a &quot;petulant child&quot; can disprove the results of his style

That’s a pretty uncharitable analysis of the situation. The point of the attack are Intel’s future intentions as implied by the patches, and the degree which those intentions are at odds with Intel’s announced guidance.That seems like a very significant technical and security matter worth discussing, no matter the style of its delivery.

&gt; I wish people in this thread would stop turning the discussion into a straw-man argument about Linus’s language and focus their attention to the technical and security problems at hand.There wasn&#x27;t much of a technical and security discussion in what was posted. Linus spent his time writing about his feelings and how hurt they were that someone did something he didn&#x27;t like. Very little of substance or merit was communicated. If you consistently behave like a petulant child in a leadership role, expect people to gawk at it.

His language is not particularly new or interesting. His technical point is very interesting, as it exposes a simple question-test for evaluating Intel’s motivations as an outsider.

His language is why this particular post is getting such traction.  This is what he does.  You can&#x27;t get attention for your bluntness and not have your bluntness discussed.

Actually, if you pay close attention, nothing is said about X, where X is Linus&#x27; language. It&#x27;s a more general &quot;these sitautions exist, but let&#x27;s ignore them right now because there&#x27;s something else to address which is more concrete.&quot;You are correct though, in that the majority of the comment is spent at least dancing around that issue since little of substance was provided on the alternative that they suggest is more worthwhile.

I&#x27;m fascinated by the lack of self-awareness in your comment:
&quot;I wish people in this thread would stop talking about X and talk about Y instead. Anyway, here are my thoughts on X.&quot;-_-

I was really surprised at how polite Linus was being in this e-mail, for Linus.  He&#x27;s obviously being very cautious and contemplative, here.

You&#x27;re completely correct - I did indeed use the wrong terminology and I&#x27;ve done exactly that before with characterising the change in direction of an argument as a straw-man when as you rightly point out - is not correct. Thanks for politely pulling me up on this, I do appreciate it.

&gt; I wish people in this thread would stop turning the discussion into a straw-man argument about Linus’s language and focus their attention to the technical and security problems at hand.Nitpicking, since everyone can tell what you mean, but FWIW, the phrase you are looking for here is someting like &quot;ad hominem&quot; or &quot;tone argument&quot;, not &quot;straw man&quot;; that&#x27;s something completely different.

I wish people in this thread would stop turning the discussion into a straw-man argument about Linus’s language and focus their attention to the technical and security problems at hand.Some people are easily offended, some people easily offend, some people are direct, some people are indirect - get over it.What matters here is the quality of the code &#x2F; workarounds that are being presented, by whom and what alternatives they are.

In nature, we usually only consider two success metrics: hosts infected and hosts killed. To maximize hosts infected, keeping the host alive and transmitting is very successful, so the most successful examples are those we barely even take note of (common cold, intestinal worms, etc.). To maximize hosts killed, in the physical world you either need a long incubation period where the host stays alive and transmitting (AIDS) or agressive transmission methods that humans have been getting quite good at shutting down (with the exception of transmission by mosquito).In the world of digital viruses traditional metrics can apply too (Ransomware is measured by hosts &quot;killed&quot;, botnets by hosts infected), but those are no longer the only interesting metrics. A virus that spreads within a certain IP block and one month after inception bricks everything it can reach might be regarded as highly successful, despite killing itself in the process.

Nothing you said is really related to the topic... classic computer viruses are a thing of the past and there are so many other attack vectors.

It may interest you that a virus which kills its host won&#x27;t spread as well as viruses which co-exist. This is basic from epidemics transmission rate. The longer an infected host keeps spreading the infection on, the more individuals will be infected. If the host is killed it will no longer spread the infection and therefore the infection may infect fewer hosts and this reduction in transmission rate alone may prevent a major outbreak.<a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Mathematical_modelling_of_infectious_disease#The_SIR_model" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Mathematical_modelling_of_infe...</a>

Could this whole disaster and the shitty patches have something to do with the remote kill switch (Anti Theft) Intel have in their CPUs?If someone could enable and trigger that remotely, it would make for fun times.

But if he doesn&#x27;t think it&#x27;s a mistake... what could it be?

He certainly doesn&#x27;t think it is a mistake. And that would fall within the spectrum of possibilities, but I would doubt anyone considers that as a probable reason, since it would be stupid to push something like that right now. Like future-destroying stupid.

Doubtful.  Possibly a fix for an as-yet-undisclosed variant of the known attacks.  Possibly mere ineptitude.

So somebody isn&#x27;t telling the truth here. Somebody is pushing complete
garbage for unclear reasons. Sorry for having to point that out.As it is, the patches are COMPLETE AND UTTER GARBAGE.They do literally insane things. They do things that do not make
sense. That makes all your arguments questionable and suspicious. The
patches do things that are not sane.Is Linus suggesting these patches are so bad that there could possibly be a camouflaged backdoor?

This looks less like allowing hacks for states and more like covering their asses in benchmarks.

Sound like Intel is interested in allowing hacks, probably for government oversight.

People only post his rants. Most of his replies are pretty benign.

Sometimes you just get sick of people pushing off garbage like it&#x27;s a 3 star Michelin meal.As the primary kernel developer, the amount of crap he has to deal with gives him broad levity IMO.I&#x27;ve worked with engineers across many fields over my ~35 years of working. When they&#x27;re ripping you a new one, you pretty much always deserve it.Putting shit on a plate is being a jerk, because you&#x27;re not only wasting your time, you&#x27;re wasting his time as well.

Those other rants are neither entertaining, nor informative. The reasons why I like Linus&#x27; rants in particular are because they&#x27;re fun to read and I learn things.

Really? The internet is full of ranting commenters. They all have content of a sort. His don&#x27;t deserve any more attention just because he&#x27;s the subject of a hero-worship cult.Its actually possible to find thoughtful commentary on most any subject, without suffering jerks.

I always appreciate Linus&#x27; rants. In addition to being somewhat humorous, I usually learn a thing or two from them. It amazes me that people focus on how much of a jerk he&#x27;s being instead of actually looking at the contents of his emails and learning something.

It was actually flavor-aid :)<a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Drinking_the_Kool-Aid#Background" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Drinking_the_Kool-Aid#Backgrou...</a>

I want to create a new LKML account just to reply simply with&gt; s&#x2F;cool-aid&#x2F;kool-aid&#x2F;Sometimes I just want to see the world burn.

&gt;Not doing the right thing for meltdown
would be completely unacceptable.Like not patching it at all. Like selling new ARM Chromebooks with no plans to ever patch meltdown in the firmware or kernel. Because that&#x27;s what Google+Partners is doing with Rockchip Chromebooks AFAICT.

It seems the grandparent (authored by Linus, too) in the email thread is not available any more. Did Linus delete it?

Looks like Intel is trying to avoiding the best solution to not compromise benchmark scores.

Haha, two mails down, David Woodhouse doesn&#x27;t disappoint and sends a message that&#x27;m copying here:<a href="http:&#x2F;&#x2F;lkml.iu.edu&#x2F;hypermail&#x2F;linux&#x2F;kernel&#x2F;1801.2&#x2F;05282.html" rel="nofollow">http:&#x2F;&#x2F;lkml.iu.edu&#x2F;hypermail&#x2F;linux&#x2F;kernel&#x2F;1801.2&#x2F;05282.html</a><pre><code> I think we&#x27;ve covered the technical part of this now, not that you like
 it â not that any of us *like* it. But since the peanut gallery is
 paying lots of attention it&#x27;s probably worth explaining it a little
 more for their benefit.
 
 This is all about Spectre variant 2, where the CPU can be tricked into
 mispredicting the target of an indirect branch. And I&#x27;m specifically
 looking at what we can do on *current* hardware, where we&#x27;re limited to
 the hacks they can manage to add in the microcode.
 
 The new microcode from Intel and AMD adds three new features.
 
 One new feature (IBPB) is a complete barrier for branch prediction.
 After frobbing this, no branch targets learned earlier are going to be
 used. It&#x27;s kind of expensive (order of magnitude ~4000 cycles).
 
 The second (STIBP) protects a hyperthread sibling from following branch
 predictions which were learned on another sibling. You *might* want
 this when running unrelated processes in userspace, for example. Or
 different VM guests running on HT siblings.
 
 The third feature (IBRS) is more complicated. It&#x27;s designed to be
 set when you enter a more privileged execution mode (i.e. the kernel).
 It prevents branch targets learned in a less-privileged execution mode,
 BEFORE IT WAS MOST RECENTLY SET, from taking effect. But it&#x27;s not just
 a &#x27;set-and-forget&#x27; feature, it also has barrier-like semantics and
 needs to be set on *each* entry into the kernel (from userspace or a VM
 guest). It&#x27;s *also* expensive. And a vile hack, but for a while it was
 the only option we had.
 
 Even with IBRS, the CPU cannot tell the difference between different
 userspace processes, and between different VM guests. So in addition to
 IBRS to protect the kernel, we need the full IBPB barrier on context
 switch and vmexit. And maybe STIBP while they&#x27;re running.
 
 Then along came Paul with the cunning plan of &quot;oh, indirect branches
 can be exploited? Screw it, let&#x27;s not have any of *those* then&quot;, which
 is retpoline. And it&#x27;s a *lot* faster than frobbing IBRS on every entry
 into the kernel. It&#x27;s a massive performance win.
 
 So now we *mostly* don&#x27;t need IBRS. We build with retpoline, use IBPB
 on context switches&#x2F;vmexit (which is in the first part of this patch
 series before IBRS is added), and we&#x27;re safe. We even refactored the
 patch series to put retpoline first.
 
 But wait, why did I say &quot;mostly&quot;? Well, not everyone has a retpoline
 compiler yet... but OK, screw them; they need to update.
 
 Then there&#x27;s Skylake, and that generation of CPU cores. For complicated
 reasons they actually end up being vulnerable not just on indirect
 branches, but also on a &#x27;ret&#x27; in some circumstances (such as 16+ CALLs
 in a deep chain).
 
 The IBRS solution, ugly though it is, did address that. Retpoline
 doesn&#x27;t. There are patches being floated to detect and prevent deep
 stacks, and deal with some of the other special cases that bite on SKL,
 but those are icky too. And in fact IBRS performance isn&#x27;t anywhere
 near as bad on this generation of CPUs as it is on earlier CPUs
 *anyway*, which makes it not quite so insane to *contemplate* using it
 as Intel proposed.
 
 That&#x27;s why my initial idea, as implemented in this RFC patchset, was to
 stick with IBRS on Skylake, and use retpoline everywhere else. I&#x27;ll
 give you &quot;garbage patches&quot;, but they weren&#x27;t being &quot;just mindlessly
 sent around&quot;. If we&#x27;re going to drop IBRS support and accept the
 caveats, then let&#x27;s do it as a conscious decision having seen what it
 would look like, not just drop it quietly because poor Davey is too
 scared that Linus might shout at him again. :)
 
 I have seen *hand-wavy* analyses of the Skylake thing that mean I&#x27;m not
 actually lying awake at night fretting about it, but nothing concrete
 that really says it&#x27;s OK.
 
 If you view retpoline as a performance optimisation, which is how it
 first arrived, then it&#x27;s rather unconventional to say &quot;well, it only
 opens a *little* bit of a security hole but it does go nice and fast so
 let&#x27;s do it&quot;.
 
 But fine, I&#x27;m content with ditching the use of IBRS to protect the
 kernel, and I&#x27;m not even surprised. There&#x27;s a *reason* we put it last
 in the series, as both the most contentious and most dispensable part.
 I&#x27;d be *happier* with a coherent analysis showing Skylake is still OK,
 but hey-ho, screw Skylake.
 
 The early part of the series adds the new feature bits and detects when
 it can turn KPTI off on non-Meltdown-vulnerable Intel CPUs, and also
 supports the IBPB barrier that we need to make retpoline complete. That
 much I think we definitely *do* want. There have been a bunch of us
 working on this behind the scenes; one of us will probably post that
 bit in the next day or so.
 
 I think we also want to expose IBRS to VM guests, even if we don&#x27;t use
 it ourselves. Because Windows guests (and RHEL guests; yay!) do use it.
 
 If we can be done with the shouty part, I&#x27;d actually quite like to have
 a sensible discussion about when, if ever, we do IBPB on context switch
 (ptraceability and dumpable have both been suggested) and when, if
 ever, we set STIPB in userspace.</code></pre>

Are people reading the rest of the thread? Linus admitted mistaking an acronym and changed his tone from &quot;pushing garbage for unclear reasons&quot; to just &quot;garbage patches&quot;.

Sounds about right.As a rule of thumb if you see a’s with accents you’re loooking at a UTF-8 encoding problem.

Probably an encoding error. An m-dash is E2 80 94 in UTF-8. (An n-dash is E2 80 93.) I suspect that the original UTF-8 is misinterpreted as ISO 8859-1 or a similar encoding, where 0xE2 is “â”, and the other two are unassigned or control characters.

&gt; As a hack for existing CPUs, it&#x27;s just about tolerable â as long as it
&gt; can die entirely by the next generation.&gt; Certainly it&#x27;s a nasty hack, but hey â the world was on fire and in the
&gt; end we didn&#x27;t have to just turn the datacentres off and go back to goat
&gt; farming, so it&#x27;s not all bad.Off-topic, but what is with the &quot;â&quot; in these sentences? Is that just some weird encoding error, or does it actually have meaning?

Why would he do that? I completely agree with him on that, and I already told him I&#x27;ve been pushing back on it since I first heard about it a few weeks ago. Although there are technical reasons why we might need IBRS_ALL as a stop-gap before we can get to a proper solution, we bloody well ought to have line-of-sight to a proper solution in the same way that RDCL_NO says &quot;it&#x27;s OK, we fixed it&quot; for Meltdown.But that&#x27;s a separate topic. As I explicitly said, I limited that answer to the things we can do on current hardware.

That&#x27;s an interesting summary but I predict that Linus will destroy him for completely avoiding the hard question: why the fuck IBRS_ALL would not be on by default on future &quot;fixed&quot; chips, if on such CPU it would not be somehow crappy.
And the risk of it becoming architectural (with that absurd default) is insane too.

And here[1] is Woodhouse&#x27;s coherent latest reply as of a couple hours ago on the issue, which explains for all us in the peanut gallery (his words, which I think are spot on) exactly what this is about, why it&#x27;s included, how it actually affects the situation, why it was put forth at all, etc.It appears much less sinister than Linus was insinuating, but Linus has yet to reply.1: <a href="http:&#x2F;&#x2F;lkml.iu.edu&#x2F;hypermail&#x2F;linux&#x2F;kernel&#x2F;1801.2&#x2F;05282.html" rel="nofollow">http:&#x2F;&#x2F;lkml.iu.edu&#x2F;hypermail&#x2F;linux&#x2F;kernel&#x2F;1801.2&#x2F;05282.html</a>

I have a gut feeling that all of this madness is driven by legal and management terror. Doing the right thing likely means, to many people, admitting fault. Legal has likely banned anyone from doing anything that looks like an admission of guilt. Management and PR, etc. are all probably about saving face, too. The people who care about doing the right thing from a technology perspective likely don&#x27;t have any power.

Intel, please, hire Linus: If you can&#x27;t beat him, join him.

If someone were to call _you_ stupid, would you be more or less willing to help? How does calling other professionals stupid help the situation?

Yup, looks like it&#x27;s time for me to pull out the good ole Amiga 500 from the closet, and use it as my main PC again. Guru meditation errors anyone?

&quot;Boy that escalated quickly...&quot;
<a href="https:&#x2F;&#x2F;youtu.be&#x2F;ipsPgNEmAXI?t=301" rel="nofollow">https:&#x2F;&#x2F;youtu.be&#x2F;ipsPgNEmAXI?t=301</a>
Anchorman fight scene aftermath

<a href="http:&#x2F;&#x2F;lkml.iu.edu&#x2F;hypermail&#x2F;linux&#x2F;kernel&#x2F;1801.2&#x2F;05282.html" rel="nofollow">http:&#x2F;&#x2F;lkml.iu.edu&#x2F;hypermail&#x2F;linux&#x2F;kernel&#x2F;1801.2&#x2F;05282.html</a>I would recommend reading this response.Here are some great excerpts:&gt; 
That&#x27;s why my initial idea, as implemented in this RFC patchset, was to
stick with IBRS on Skylake, and use retpoline everywhere else. I&#x27;ll
give you &quot;garbage patches&quot;, but they weren&#x27;t being &quot;just mindlessly
sent around&quot;. If we&#x27;re going to drop IBRS support and accept the
caveats, then let&#x27;s do it as a conscious decision having seen what it
would look like, not just drop it quietly because poor Davey is too
scared that Linus might shout at him again. :)&gt; 
If we can be done with the shouty part, I&#x27;d actually quite like to have
a sensible discussion about when, if ever, we do IBPB on context switch
(ptraceability and dumpable have both been suggested) and when, if
ever, we set STIPB in userspace.Those seem relevant to the thread. That said - it&#x27;s really a good read for a technical overview of the issues and patches.

Is this public discussion an attempt to &quot;zero in&quot; on this problem? Does anyone outside of Intel &quot;agree&quot; with their current approach? This is a more reasonable situation for the Torvalds method than many others in which he uses it.

Complaining about downvotes is never cool.

&gt; I&#x27;m sure I&#x27;ll get downvoted though for calling Linus a baby, because somehow insulting Linus for his flaws is never ok but when Linus does it it&#x27;s so cool!I&#x27;m pretty sure people aren&#x27;t up in arms about Linus being a baby, but about your refusal to call out people on &quot;subpar performance&quot;. Replace &quot;subpar performance&quot; with whatever you feel is an acceptable name for executing wrongly on very bad ideas.And I feel like you might get downvoted because if there is one thing I HATE about companies, it&#x27;s that people responding like this always get rewarded.1) fuck up beyond belief (or just be assigned to a critical project and do nothing)2) talk about the problem with everyone, without, of course, without fixing anything3) inflict massive costs on everyone else because of the sudden urgent necessity of fixing your problem4) get insane rewards for thisAnd I for one call statements such as the one from the top post here &quot;optimism at it&#x27;s finest&quot;:&gt; if you call someone a moron in a thread, it&#x27;s likely to kill productivity and create a negative testosterone ladden atmosphere of people trying to avoid being wrong, and counter attacking others.Unfortunately, corporate America is FULL, and I mean overloaded to the brim with people who will never admit the slightest mistake, and explode at anyone implying they made a subpar decision.This attitude to &quot;avoid blame&quot; then proceeds to turn disaster into outright catastrophe. Many of these people who make some level of mistakes will see that not only can they get away with it, but aggressively attacking others for their mistakes comes with great rewards.I have personally worked with several well known people in my field, who have climbed up the ladder this way. It isn&#x27;t just that they didn&#x27;t contribute, but they made great mistakes, both on the technical front (taking disastrous architectural decisions and forcing them through, despite agreement that there were very wrong), and on the financial front (doing budget allocations based on what THEY KNEW was a lost cause, and refusing to admit any fault or change course).Once it does become clear that a disastrous mistake was made and propagated, and 90% of the budget was already spent (actually more like 150% in both cases), and none of the outcomes promised will be achieved, they then help organise a &quot;code red&quot;, a &quot;emergency sprint&quot;, an &quot;extra effort&quot;, ... to &quot;fix&quot; things. But things are never fixed. Everyone is worse off, and will be for years.The first few times this happens, these people are then rewarded. Promoted, in both cases. And of course, we are currently in the process of repeating one of these huge mistakes. Last time I so much as asked why this decision was made the guy very enthousiastically gave an answer, and I thought we both agreed it wasn&#x27;t a good answer.An hour later my manager asks to urgently see me.A day later my director urgently called my manager and me to see him.A week later we went to discuss &quot;safety&quot; with HR. (because of the 10 people thing, I&#x27;m sure, read on)And yet, everybody, and I do mean everybody, knows this is a bad idea. However, having 50 people work for 1.5 years on a bad idea ... well those 50 people call it &quot;a job&quot;, and nobody&#x27;s going to be risking their career on calling this guy out any further (even if I did get 10 of them to walk into that director&#x27;s office telling him that I did nothing wrong and that it&#x27;s a bad idea).But no worries ... no blame. We&#x27;re still executing on that idea.

This is obvious, and it&#x27;s extremely sad to me that people believe otherwise and vehemently argue that Linus is justified in his pety rants. It was always pathetic.The response is shockingly adult, given the culture Linus insists on trying to push:<a href="http:&#x2F;&#x2F;lkml.iu.edu&#x2F;hypermail&#x2F;linux&#x2F;kernel&#x2F;1801.2&#x2F;05282.html" rel="nofollow">http:&#x2F;&#x2F;lkml.iu.edu&#x2F;hypermail&#x2F;linux&#x2F;kernel&#x2F;1801.2&#x2F;05282.html</a>What an excellent writeup, despite the tantrum Linus threw just a few replies earlier.I&#x27;d highly recommend only reading the response and avoiding the Linus &#x27;rant&#x27;. The rant contains no valuable information, the response is excellent and far more informative.I&#x27;m sure I&#x27;ll get downvoted though for calling Linus a baby, because somehow insulting Linus for his flaws is never ok but when Linus does it it&#x27;s so cool!

A lot of people trying to defend being a jerk as necessary in these circumstances. I think Google culture proves the opposite: Googley code and peer reviews, blameless postmortems, and a host of other mechanisms that de-escalate and de-personalize things.The Meltdown work at Google probably didn&#x27;t begin with a rant about morons.It&#x27;s possible to construct a culture where people can air grievances and criticize others without inducing flame wars. It only works for Linus because people won&#x27;t go nuclear in the response, but for rank and file engineers, especially of equal stature, if you call someone a moron in a thread, it&#x27;s likely to kill productivity and create a negative testosterone ladden atmosphere of people trying to avoid being wrong, and counter attacking others.I&#x27;ve worked at a lot of places where engineer got into heated shouting matches. It&#x27;s is not a way to increase the probability of zeroing in on a problem, or reaching agreement faster.

Linus Torvalds: “Somebody is pushing complete garbage for unclear reasons.”