ehhhhhhhhhhhhhhhh.. I mean sure, you could be technically correct here- a single temperature cannot drive an engine. However, a gradient of temperature absolutely can drive an engine. Waste heat is a temperature gradient that is too small to do work.
So your heat engine sitting in the hottest part of the universe merely needs a way to channel that heat to a colder part of the universe, which should be hard if it's sitting in the hottest part.
Erm, yes?
Space is a perfect insulator. There's no method of heat exchange other than black body radiation, the least efficient method.
Cooling stuff in space is hard.
Not with fission plants, man.
You couldn't have radiators large enough to keep the people alive.
Aneutronic fusion is the critical nuclear technology for space travel.
I wish you had researched a bit before posting this. A lot of people modded incorrect information as insightful.
There are working rowhammer privilege escalations.
The main one I remember involves mmap()ing a whole asston of anonymous pages, essentially to the point where you fill an appreciable amount of RAM with page table entries. At that point you can use some inference to target those PTEs with good probability of success. As soon as you can modify one PTE to give your process RW access to an arbitrary page in RAM (hopefully full of PTE entries as well) you now have unfettered access to the entire address space of the machine.
You on the other hand, got knocked down and started whining "life isn't fair."
That's a rather boneheaded interpretation of what I wrote, don't you think?
I'd say it's more accurate that to say that I dusted myself off, made it to the top 6% of income earners and then reflected upon the journey and said, "wow. life really isn't fair."
Luck is absolutely garbage mumbo-jumbo.
Followed with...
There is chance, and chance can be modified by your own actions.
I think it's clear you're not entirely familiar with the words chance and luck.
Also, anything can be modified by your actions, but the hubris displayed here regarding your apparent belief that you can overcome all odds, and that anyone who cannot obviously just didn't have what it takes is mind-blowing. You may be a skilled person, but you're certainly not intelligent.
Not sure from where you drew that twisted ass logic to come to that conclusion. I'm pretty sure I said luck was the luck part.
Seriously, how the hell did you make it past grade school?
Was it luck that got you your skills?
Nope. It was luck that allowed me to use them.
Luck that made you decide that you didn't like living dirt poor?
Nope. It was luck that allowed me to use my skills, and pull myself out of poverty.
It was the desire not to be in the same place your parents were because you saw how much suffering it caused them.
Pretty sure that isn't a desire unique to the successful. I think the fact that you say that says a lot more about you than you realize.
I think it's the tortured logic your brain has developed to cope with the cognitive dissonance you feel inside over the knowledge that those of us in the white collar 10%+ percentiles aren't as special as luck has allowed us to position ourselves.
Luck is about the furthest thing from reality in the world in being an actual factor.
An interesting claim... One you haven't actually made any argument to support.
I have, however, already give you a counter-argument to that assertion- so keep trying. You can have all the skill in the world, but it's luck that gets you to where you can use it. Or do you really think all those kids in the third world are just lazy brown colored animals?
The arrogance in believing that luck though screams that you don't think things through.
Another claim you haven't actually backed up with any logic to speak of.
Though I fundamentally pity you, because I'm an asshole.
I find that assholes often pity their betters. I think it's another coping mechanism.
It sure is, and I have monetarily supported their experiment, and will continue to do so.
I wasn't aware the Aspire line now had officially supported linux models- that's awesome. What I'd really like to see though are some ultra-light Linux celeron laptops with silly battery life. I'm currently using a ChomeBook that's been converted to running Kubuntu natively, and I've absolutely fallen in love with it. All it needs is a real keyboard. This Bay Trail Celeron may be pretty gutless, but it gets the job done for my work and home use, and whether I'm watching movies on it, browsing, or working- it goes forever. I haven't pulled out the beefy i7 laptop in almost 2 months.
I don't disagree with your assertion that there's value in having mindshare... and I share your dislike of locked down systems.
I wouldn't say open source is embracing locked down systems, but the success of those systems is also an advertisement for open source.
I think Linux benefits from the WSL. It increases mindshare, and increases use of the GNU/Linux ecosystem, particularly for developers who are using Windows. It could be their preparatory course for making the switch.
ChromeOS is an actively supported platform. They make improvements to the kernel (as well as things that are useless to mainline- but who cares about that) and they offer the knowledge and fixes gleaned from experience bye having a large installed attack surface.
So really, I don't know. I get what you're saying. But we have benefited from the exposure.
At least in the case of ChromeOS, they have a good reason for locking down the system (It's essentially supposed to be a cloud thin-client with absolute security) *and* every single model provides a developer mode with unfettered access to the OS.
Android... That's a bit sadder of a situation, and I do agree it sucks.
I call someone out for calling a legal behavior that he doesn't like illegal, making no argument whatsoever on the intelligence or ethics of the said action... and you counter by giving your argument as to why it shouldn't be done as well? You're an imbecile.
Ya, I think I was trying to be conciliatory or something. When you get right down to it, there's nothing less esoteric about the incantations you have to pump into Windows' pathetic excuse for a terminal when something on it doesn't work right, and most of the time, you can actually fix something that isn't working right with your popular Linux DE, while in Windows you're forced to suffer a million 'might work' measures because nobody actually really understands wtf the Windows Update agent actually does.
Linux on the desktop is fucking wonderful. It's just not very user-friendly, and so doesn't appeal to a large audience.
I can't imagine having my primary desktop being anything else anymore, after years of using Windows and MacOS.
I of course acknowledge that that is an opinion, and inherently worthless- much like your post.
Well, it *is* Linux, in that it uses the Linux kernel.
Its userspace is also very similar to a busybox userspace (actually uses toolbox, which is their own version of busybox) which is pretty similar to GNU...
I have no problem calling ChromeOS Linux. I'm also completely sure that you're correct that only a tiny fraction of new non-server PC sales come with a flavor of Linux that isn't ChromeOS.
It's too early to tell. It's possible AMD is affected as well, as this doesn't really sound like a bug by any sort.
It appears to be a new vector for well-known crypto key-derivation side channel timing attacks.
The CPU isn't really leaking anything from the sounds of it, they're just using the fact that the TLB is shared between threads as a way to statistically derive the key schedule of a well known and profiled encryption library.
I'm not familiar with AMDs architecture, so I can't speak as to whether or not their TLB is shared, but I'd guess it was.
The real problem here is a shitty encryption library. Encryption routines that are aware of the existence of side-channel attacks have constant-timing mechanisms built in to prevent exactly this kind of thing, from any vector, because there are probably a near infinite amount of vectors we have yet to even have thought up. It's the nature of having shared resources and multiple processes. Timing is going to change based on what someone else is doing.
After this hits, it will take about 45 minutes for someone to have a Linux distro up and running on it.
I wish, but nope.
There are Qualcomm 835 laptops on the market already. I own one. Still no Linux.
It's a little harder than you think building a BSP for the kernel when you have literally no idea what's going on in the boot process, hardware or software wise.
More than 1 instruction per cycle isn't a sign of RISC or CISC, it's a sign of a design which is superscalar.
Yep... And RISC itself was an architecture designed to make it so that your program required more single-cycle instructions to operate, as opposed to less multi-cycle instructions in a CISC machine. As you perfectly parroted what I said- a distinction largely irrelevant with superscalar processors.
All modern processors are internally RISC, with CISC glued onto the front of them.
Yep. Pretty sure that was implied in what I said?
Yes, RISC operations.
You betcha. But the ISA that is exposed to the programmer is the determining factor of whether or not an *ISA* is CISC or RISC, not the microarchitecture.
CISC machines have been RISC at the core since forever- that's literally a design requirement. The microarchitecture is inherently RISC.
Yes, RISC operations.
Again, all processor microarchitectures are and generally speaking have always been RISC. The point is that RISC is meaningless in an ISA context anymore because even RISC processors are CISC. It's a requirement for superscalar operation. RISC was always a poor design. Requiring more instructions to flow across the entire code path in order to do the same amount of work that can be done efficiently in tight hardware was silly.
All the stuff that doesn't decompose into RISC is implemented as a coprocessor
Not quite, but you're on the right track. ARM handles SIMD in-line as separate paths in the pipeline after the instruction decoders. It's not kicked to a coprocessor. This started around ARMv6 architecture.
One would *hope* that AMD's HT design is safe, as they had ample time to learn from the first time Intel made this fuck-up, long before AMD had a HT core, back in 2005 with the P4.
Regardless, Random Slashdotter is correct.
I'm not Theo, but I do have published kernel-level exploits for Linux/Android in the NVD, so I'm not speaking completely out of my ass. A lot of these recent CPU architecture attacks are pretty weaksauce. They certainly are real fuckups in the concept of process isolation, but the paranoid people like to act like inferring random memory a few bits at a time is going to yield you useful information.
I've done a lot of work breaking into kernels, and getting access to their memory was never the hard part. Navigating it was. And that was at full line speed.
Different, but also very much the same.
It was designed to run on an engine that had asm.js optimizations in place, and was itself merely a binary (and descriptive assembly) format for applications that were designed to run on an asm.js optimized VM.
The initial versions of WebAssembly were in fact demoed using asm.js shims, and currently, WebAssembly appliations can be converted to asm.js on the fly for browsers that don't support the binary format. WebAssembly introduces even more restrictions to the operational set than asm.js did so that the optimizations required to make it fast could be standardized and implemented on more platforms easier.
Even FTM is misleading:
"The technology is a compact binary language that a browser will convert into machine code and run it directly on the CPU."
That is in fact not misleading at all. It's so true I'm almost shocked at its accuracy.
WebAssembly is in fact a compact binary language that a browser (at least most) will convert into machine code on the fly and run directly on the CPU via their wasm-optimized JavaScript JIT compilers.
Sadly, I don't know. You can't run linux yet on an HP Envy x2, as stupid as that sounds.
I am curious, there isn't much information available other than on platforms like the RaPi which have bad IO congestion holding them back.
Agreed... And I suspect that's the real cause of the perceived slowness. Not I/O specifically, but just chunks of the hardware stack in general that just don't have the throughput full OS stacks designed to run on x86 machines are accustomed to.
Now, I've never thrown a full Linux DE on my RPi 3, but I have put it on an Exynos big.LITTLE SBC I got from hardkernel, and while it performs decently enough... It's the same story. It feels more like my Celeron ChromeBook than even my little i3 Asus laptop (which is running Linux).
It could also be that people feel the lack of individual core performance between ARMs and AMD/Intel x86 more than is pushed by those comparing multithreaded benchmark scores that show close parity between architectures.
This all said- I'd still shell out money again for an ARM laptop that could run Linux, and I'm not terribly unhappy about the one I have purchased.. But if we're being honest, it's still a $1000 ARM laptop that is best compared against $300 Intel laptops.
I'm also one of those people, and never once had a desire to see others struggle.
You start your reply out with a lie. You know it, I know it, we all know it. It's hard to admit, so I get it.
I did, however, witness a lot of stupidity and bad choices.
How wonderful it must be to have been to have not made bad choices, or to not have had good choices still fuck you over.
Choices for short term pleasure instead of long term gain.
Ah, yes. This is why poverty exists. Idiots and their short-term pleasure, working their 60 hours a week for piddly sums while you (I presume) and I sit here with our 6 figure salaries.
Having witnessed the choices being made, I now find it difficult to give a fuck about the predictable results.
Here we go- you equate lack of success across the board with one of the reasons you've seen for it. The only thing predictable here is how the system let an asshole like you succeed where you should have fallen flat on your face for being a generalizing toolshed.
Go to school. Do your homework. Stay off drugs. Keep your job, even when it isn't fun, at least until you have another.
Speaking of vast majorities... That's precisely what the vast majority does.
There is luck, but that is just opportunity meeting preparation, and it is the ones who always refused to prepare by cultivating skill, ability and competence that are whining about missing out on it.
And it's also the people who prepared their entire lives, but that luck just never showed up.
I worked fast food for the first half a decade of my life before luck put me in a position to show that I can excel.
I worked with better people than myself. Better people than you, judging from my first impression of you.
So, how about you getting over yourself, asshole. Nobody owes you anything.
That statement shows you literally lacked the reading comprehension required to even understand my post.
You're a perfect example of someone who had the luck, and not the merit, I think.
Slashdot Mirror
s/should/shouldn\'t/;
See, temperature itself cannot drive an engine.
ehhhhhhhhhhhhhhhh.. I mean sure, you could be technically correct here- a single temperature cannot drive an engine. However, a gradient of temperature absolutely can drive an engine. Waste heat is a temperature gradient that is too small to do work.
So your heat engine sitting in the hottest part of the universe merely needs a way to channel that heat to a colder part of the universe, which should be hard if it's sitting in the hottest part.
Other than that, your post is correct.
Erm, yes?
Space is a perfect insulator. There's no method of heat exchange other than black body radiation, the least efficient method.
Cooling stuff in space is hard.
Not with fission plants, man.
You couldn't have radiators large enough to keep the people alive.
Aneutronic fusion is the critical nuclear technology for space travel.
I wish you had researched a bit before posting this. A lot of people modded incorrect information as insightful.
There are working rowhammer privilege escalations.
The main one I remember involves mmap()ing a whole asston of anonymous pages, essentially to the point where you fill an appreciable amount of RAM with page table entries. At that point you can use some inference to target those PTEs with good probability of success. As soon as you can modify one PTE to give your process RW access to an arbitrary page in RAM (hopefully full of PTE entries as well) you now have unfettered access to the entire address space of the machine.
You on the other hand, got knocked down and started whining "life isn't fair."
That's a rather boneheaded interpretation of what I wrote, don't you think?
I'd say it's more accurate that to say that I dusted myself off, made it to the top 6% of income earners and then reflected upon the journey and said, "wow. life really isn't fair."
Luck is absolutely garbage mumbo-jumbo.
Followed with...
There is chance, and chance can be modified by your own actions.
I think it's clear you're not entirely familiar with the words chance and luck.
Also, anything can be modified by your actions, but the hubris displayed here regarding your apparent belief that you can overcome all odds, and that anyone who cannot obviously just didn't have what it takes is mind-blowing. You may be a skilled person, but you're certainly not intelligent.
So luck was the hardworking part?
Not sure from where you drew that twisted ass logic to come to that conclusion. I'm pretty sure I said luck was the luck part.
Seriously, how the hell did you make it past grade school?
Was it luck that got you your skills?
Nope. It was luck that allowed me to use them.
Luck that made you decide that you didn't like living dirt poor?
Nope. It was luck that allowed me to use my skills, and pull myself out of poverty.
It was the desire not to be in the same place your parents were because you saw how much suffering it caused them.
Pretty sure that isn't a desire unique to the successful. I think the fact that you say that says a lot more about you than you realize.
I think it's the tortured logic your brain has developed to cope with the cognitive dissonance you feel inside over the knowledge that those of us in the white collar 10%+ percentiles aren't as special as luck has allowed us to position ourselves.
Luck is about the furthest thing from reality in the world in being an actual factor.
An interesting claim... One you haven't actually made any argument to support.
I have, however, already give you a counter-argument to that assertion- so keep trying. You can have all the skill in the world, but it's luck that gets you to where you can use it. Or do you really think all those kids in the third world are just lazy brown colored animals?
The arrogance in believing that luck though screams that you don't think things through.
Another claim you haven't actually backed up with any logic to speak of.
Though I fundamentally pity you, because I'm an asshole.
I find that assholes often pity their betters. I think it's another coping mechanism.
and rapidly growing
It sure is, and I have monetarily supported their experiment, and will continue to do so.
I wasn't aware the Aspire line now had officially supported linux models- that's awesome. What I'd really like to see though are some ultra-light Linux celeron laptops with silly battery life. I'm currently using a ChomeBook that's been converted to running Kubuntu natively, and I've absolutely fallen in love with it. All it needs is a real keyboard. This Bay Trail Celeron may be pretty gutless, but it gets the job done for my work and home use, and whether I'm watching movies on it, browsing, or working- it goes forever. I haven't pulled out the beefy i7 laptop in almost 2 months.
I don't disagree with your assertion that there's value in having mindshare... and I share your dislike of locked down systems.
I wouldn't say open source is embracing locked down systems, but the success of those systems is also an advertisement for open source.
I think Linux benefits from the WSL. It increases mindshare, and increases use of the GNU/Linux ecosystem, particularly for developers who are using Windows. It could be their preparatory course for making the switch.
ChromeOS is an actively supported platform. They make improvements to the kernel (as well as things that are useless to mainline- but who cares about that) and they offer the knowledge and fixes gleaned from experience bye having a large installed attack surface.
So really, I don't know. I get what you're saying. But we have benefited from the exposure.
At least in the case of ChromeOS, they have a good reason for locking down the system (It's essentially supposed to be a cloud thin-client with absolute security) *and* every single model provides a developer mode with unfettered access to the OS.
Android... That's a bit sadder of a situation, and I do agree it sucks.
I call someone out for calling a legal behavior that he doesn't like illegal, making no argument whatsoever on the intelligence or ethics of the said action... and you counter by giving your argument as to why it shouldn't be done as well? You're an imbecile.
Ya, I think I was trying to be conciliatory or something. When you get right down to it, there's nothing less esoteric about the incantations you have to pump into Windows' pathetic excuse for a terminal when something on it doesn't work right, and most of the time, you can actually fix something that isn't working right with your popular Linux DE, while in Windows you're forced to suffer a million 'might work' measures because nobody actually really understands wtf the Windows Update agent actually does.
Finally time to go HURD, anyone?
I just installed Debian/HURD on a VM 2 evenings ago to see how it was coming along...
Let me answer that question for you: No.
Linux on the desktop is fucking wonderful. It's just not very user-friendly, and so doesn't appeal to a large audience.
I can't imagine having my primary desktop being anything else anymore, after years of using Windows and MacOS.
I of course acknowledge that that is an opinion, and inherently worthless- much like your post.
Well, it *is* Linux, in that it uses the Linux kernel.
Its userspace is also very similar to a busybox userspace (actually uses toolbox, which is their own version of busybox) which is pretty similar to GNU...
I have no problem calling ChromeOS Linux. I'm also completely sure that you're correct that only a tiny fraction of new non-server PC sales come with a flavor of Linux that isn't ChromeOS.
Not trolling-
Are the existing Fuse/KDE/Gnome VFS plugins not what you'd like to see?
It's too early to tell. It's possible AMD is affected as well, as this doesn't really sound like a bug by any sort.
It appears to be a new vector for well-known crypto key-derivation side channel timing attacks.
The CPU isn't really leaking anything from the sounds of it, they're just using the fact that the TLB is shared between threads as a way to statistically derive the key schedule of a well known and profiled encryption library.
I'm not familiar with AMDs architecture, so I can't speak as to whether or not their TLB is shared, but I'd guess it was.
The real problem here is a shitty encryption library. Encryption routines that are aware of the existence of side-channel attacks have constant-timing mechanisms built in to prevent exactly this kind of thing, from any vector, because there are probably a near infinite amount of vectors we have yet to even have thought up. It's the nature of having shared resources and multiple processes. Timing is going to change based on what someone else is doing.
After this hits, it will take about 45 minutes for someone to have a Linux distro up and running on it.
I wish, but nope.
There are Qualcomm 835 laptops on the market already. I own one. Still no Linux.
It's a little harder than you think building a BSP for the kernel when you have literally no idea what's going on in the boot process, hardware or software wise.
More than 1 instruction per cycle isn't a sign of RISC or CISC, it's a sign of a design which is superscalar.
Yep... And RISC itself was an architecture designed to make it so that your program required more single-cycle instructions to operate, as opposed to less multi-cycle instructions in a CISC machine. As you perfectly parroted what I said- a distinction largely irrelevant with superscalar processors.
All modern processors are internally RISC, with CISC glued onto the front of them.
Yep. Pretty sure that was implied in what I said?
Yes, RISC operations.
You betcha. But the ISA that is exposed to the programmer is the determining factor of whether or not an *ISA* is CISC or RISC, not the microarchitecture.
CISC machines have been RISC at the core since forever- that's literally a design requirement. The microarchitecture is inherently RISC.
Yes, RISC operations.
Again, all processor microarchitectures are and generally speaking have always been RISC. The point is that RISC is meaningless in an ISA context anymore because even RISC processors are CISC. It's a requirement for superscalar operation. RISC was always a poor design. Requiring more instructions to flow across the entire code path in order to do the same amount of work that can be done efficiently in tight hardware was silly.
All the stuff that doesn't decompose into RISC is implemented as a coprocessor
Not quite, but you're on the right track. ARM handles SIMD in-line as separate paths in the pipeline after the instruction decoders. It's not kicked to a coprocessor. This started around ARMv6 architecture.
One would *hope* that AMD's HT design is safe, as they had ample time to learn from the first time Intel made this fuck-up, long before AMD had a HT core, back in 2005 with the P4.
Regardless, Random Slashdotter is correct.
I'm not Theo, but I do have published kernel-level exploits for Linux/Android in the NVD, so I'm not speaking completely out of my ass. A lot of these recent CPU architecture attacks are pretty weaksauce. They certainly are real fuckups in the concept of process isolation, but the paranoid people like to act like inferring random memory a few bits at a time is going to yield you useful information.
I've done a lot of work breaking into kernels, and getting access to their memory was never the hard part. Navigating it was. And that was at full line speed.
VM separation protects you from Murphy, not Machiavelli.
Brilliant.
Different, but also very much the same.
It was designed to run on an engine that had asm.js optimizations in place, and was itself merely a binary (and descriptive assembly) format for applications that were designed to run on an asm.js optimized VM.
The initial versions of WebAssembly were in fact demoed using asm.js shims, and currently, WebAssembly appliations can be converted to asm.js on the fly for browsers that don't support the binary format. WebAssembly introduces even more restrictions to the operational set than asm.js did so that the optimizations required to make it fast could be standardized and implemented on more platforms easier.
Even FTM is misleading: "The technology is a compact binary language that a browser will convert into machine code and run it directly on the CPU."
That is in fact not misleading at all. It's so true I'm almost shocked at its accuracy.
WebAssembly is in fact a compact binary language that a browser (at least most) will convert into machine code on the fly and run directly on the CPU via their wasm-optimized JavaScript JIT compilers.
Is that true even with a light Linux distro?
Sadly, I don't know. You can't run linux yet on an HP Envy x2, as stupid as that sounds.
I am curious, there isn't much information available other than on platforms like the RaPi which have bad IO congestion holding them back.
Agreed... And I suspect that's the real cause of the perceived slowness. Not I/O specifically, but just chunks of the hardware stack in general that just don't have the throughput full OS stacks designed to run on x86 machines are accustomed to.
Now, I've never thrown a full Linux DE on my RPi 3, but I have put it on an Exynos big.LITTLE SBC I got from hardkernel, and while it performs decently enough... It's the same story. It feels more like my Celeron ChromeBook than even my little i3 Asus laptop (which is running Linux).
It could also be that people feel the lack of individual core performance between ARMs and AMD/Intel x86 more than is pushed by those comparing multithreaded benchmark scores that show close parity between architectures.
This all said- I'd still shell out money again for an ARM laptop that could run Linux, and I'm not terribly unhappy about the one I have purchased.. But if we're being honest, it's still a $1000 ARM laptop that is best compared against $300 Intel laptops.
I'm also one of those people, and never once had a desire to see others struggle.
You start your reply out with a lie. You know it, I know it, we all know it. It's hard to admit, so I get it.
I did, however, witness a lot of stupidity and bad choices.
How wonderful it must be to have been to have not made bad choices, or to not have had good choices still fuck you over.
Choices for short term pleasure instead of long term gain.
Ah, yes. This is why poverty exists. Idiots and their short-term pleasure, working their 60 hours a week for piddly sums while you (I presume) and I sit here with our 6 figure salaries.
Having witnessed the choices being made, I now find it difficult to give a fuck about the predictable results.
Here we go- you equate lack of success across the board with one of the reasons you've seen for it. The only thing predictable here is how the system let an asshole like you succeed where you should have fallen flat on your face for being a generalizing toolshed.
Go to school. Do your homework. Stay off drugs. Keep your job, even when it isn't fun, at least until you have another.
Speaking of vast majorities... That's precisely what the vast majority does.
There is luck, but that is just opportunity meeting preparation, and it is the ones who always refused to prepare by cultivating skill, ability and competence that are whining about missing out on it.
And it's also the people who prepared their entire lives, but that luck just never showed up.
I worked fast food for the first half a decade of my life before luck put me in a position to show that I can excel.
I worked with better people than myself. Better people than you, judging from my first impression of you.
So, how about you getting over yourself, asshole. Nobody owes you anything.
That statement shows you literally lacked the reading comprehension required to even understand my post.
You're a perfect example of someone who had the luck, and not the merit, I think.