The fix is to not read from memory into the CPU cache during the speculative execution when that block of data is not there already. Changing this in the CPUs core would solve both Spectre and Meltdown, at a reasonable cost (would not defeat much current optimizations).
This is the correct answer. Proceed to Go and collect your $200.
You may want set aside more silicon for caching and less for handling the speculation.
You may also want more cores rather than greater complexity. This would not have been a good choice ten years ago, but now people are learning how to use the extra cores, it will probably sell well.
Alternatively, you could set a flag to say whether your application cares about the risk (if the entire machine is dedicated to a single offline task, you probably don't).
Actually, implementing the logic gates for it can very well be worth a patent. Since there doesn't seem to be any prior art for actually implementing it with logic gates in a quick patent DB search, that definitely makes it a justifiable patent.
In most countries, patents require the invention not to be "obvious to anyone sufficiently skilled in the art". Only in America is "not blatantly obvious to a blithering idiot with some $$$" the criterion.
In the olden days (1970's) we did the computation in standard and double precision - if the answers were different, then you probably needed to change the method. (The underlying problem is "underflow" - you do not have any significant digits).
The exact same thing was said about Linux in 1991-1992, that it would never compete against "real" operating systems like Solaris, ULTRIX, and others.
No sane person ever said any such thing. The whole of Unix was written by Ken and Dennis in their lunch breaks, not just the kernel. Minix was not such a great deal either.
An OS was smaller then, because
(a) there were no GUIs
(b) most of the "utilities" we expect now were not considered part of the system (and that often included the compiler and linker)
(c) there was less bloat (Multics excepted).
As is said elsewhere - these days, a processor is a lot more than the CPU.
However, most of the rest is well known - how to cache - or externally defined (Memory and peripheral interfaces).
In reality, the system architecture is defined in software (VHDL). Most generation of test software should be highly automated. So, yes... the problems of an open source processor are mostly political.
There may be no money in open source hardware NOW - but the argument is that so many people have now been shafted by closed source hardware that some might be interested in an alternative.
No one is talking about "free as in beer".
Someone who wants to break into the CPU market (can you say "Chinese") might want a sales pitch that can overcome their current image problems ("it might keep phoning home") - and open source is probably the ONLY tool that can crack that particular nut.
And "It keeps phoning the mother ship" (or hackers.ru) is becoming a blunderbuss aimed deep into the heart of closed source. Meltdown and Spectre are "me, too" arguments in reality. You don't need bugs for closed source kit to be a security risk - you cannot tell if it is a security risk even in the absence of bugs. As any reliable and honest crook/casino owner will tell you: "if you can't tell if you are being cheated: you are being cheated".
A lot of large customers buying expensive kit is a very valuable market. "Not being American" is potentially a massive potential selling point to the 90% of the world's population that is not American, but without open access to the design, very few will buy into a product from any of the potential alternative suppliers.
However, I suspect what is really needed is not just open source, but also with a credible second source (eg from two different BRIC countries).
Any project involving a government committing to spending money is doomed, and any IT project involving the British Government, doubly or even quadruply so.
Never mind Babbage, and Harrison's clocks, look what killed the Transputer - Mrs Thatcher promised £50M the same week United Technologies scrapped their microprocessor project saying "in the world of computers, $50M is peanuts". The funding was indeed too little, too late, and the project was sold to the French - who killed it by accident.
It's not going to happen "open source" unless some very wealthy individual or organization decides to do so for altruistic reasons.
That is certainly the Windows buyer PHB perspective.
In reality, large corporations (you have probably heard of IBM) put money into open source because it is a way to share the cost of creating and supporting infrastructure which their actual product depends on.
We have to hope/pray that Larry Ellison has a "Road to Damascus" event, and realises that a truly open Sparc system (not just CPU) might lead to Oracle being in a far stronger position than it already is, and seriously weaken his competition. It might also mean that the nerds who currently detest him might see him as a saviour. (I am not holding my breath).
Currently, Oracle is in the business of preventing access to drivers and microcode for machines over 5 years old and out of support, thus depriving them of second hand value - not exactly helpful to their own business - one can only assume that that Larry is possessed by the ghost of Ebenezer Scrooge and hell bent on wasting the best publicity event that he could dream of.
Only in America. The rest of the world does not have Trump and his followers, and people go outside and see for themselves.
Some of us have lived longer than 50 years, and been to many parts of the world, where we have families, and we can confirm the changes they describe from our own experience. The climate science is in agreement with this experience as far as we can tell.
The deniers world exists only in America, where most people appear to live in basements, experiencing virtual reality.
Except they're starving because of assholes like Robert Mugabe, not because of climate anything (or weather, for that matter). Productive farmland became a desolate waste because people stopped farming, not because crops wouldn't grow. African starvation is the result of African land management policies and "charitable" donations of millions of tons of food from the developed world over decades that drove local farmers out of business, not CO2.
I am not defending Mugabe (I will, however, point out that he was deposed, peacefully), or that he destroyed the farming productivity in Zimbabwe, BUT Zimbabwe is not big in the scale of Africa. Its goats that have destroyed good farmland in Africa, and to a lesser extent, continue to do so.
Much of Africa is not starving, partly as a result of proactive policies (eg widespread fish farming in West Africa means fish is cheaper and better than in Europe, massive reforestation program crossing Africa below the Sahara). Good news is not very profitable - people want doom and gloom - and by God, Trump is going to give them that!
There is a massive problem with politics in Africa - though whether its worse than America is somewhat debatable - and much of the problem was due to large amounts of outside influence destroying traditional political structures. (eg colonisation, oil and arms industries).
All food is local? Not here (UK) we import about 75% of our food, and for the same reason Trump was elected, are busy biting the had that feeds us (Brexit).
My globules are feeing pretty warm, despite the freezing weather.
Go back to Africa. See if they welcome you with open arms... or if they machete you to death because the 40% African in you is from the wrong tribe.
In most of Africa (ie those parts where there is no actual declared war), the murder rate is significantly lower than in the USA (much the same as in Europe).
otherwise you might learn that the people in north Africa have been selling slaves (including whites) since B.C.
Or, if you were to look outside of/., you would find that in a lot of North Africa and the Arab world, nothing has changed in this regard. Saudis selling African slaves (and killing or castrating them) has been much in the news over the past week.
Retard detected. Apache 2.0 was an enormous rewrite.
And mostly works exactly the same.
A closed source rewrite would have cost an arm and a leg to upgrade, had twice as many bugs, and rewriting your configuration files would have taken a year to get stable because of piss-poor documentation,
In 1987, most companies shared software for their hardware within a user group - and it was not much use to anyone not using the hardware.
The Pascal compiler available on DECUS was definitely ported to other hardware, as was a bunch of the other stuff there - and I am sure some other programs began on Data General and moved to DEC.
In fact, if you got software from your hardware supplier, you normally could get the source if you wanted, and probably wanted to if you had people capable of fixing the bugs - which were many.
Closed Source started with Bill Gates - who wrote a famous letter about how programmers (meaning himself) should get paid. Before the PC, most people would not have bought hardware if they could not get the source. (Of course, with computers costing slightly more than a large house, "most people" would not have bought a computer). With cheaper computers came PT Barnum.
They knew damn well it was shady. It was widely discussed at the time. And then they patented it so AMD could not do it!
Intel are up to their neck in it.
This is the correct answer. Proceed to Go and collect your $200.
You may want set aside more silicon for caching and less for handling the speculation.
You may also want more cores rather than greater complexity. This would not have been a good choice ten years ago, but now people are learning how to use the extra cores, it will probably sell well.
Alternatively, you could set a flag to say whether your application cares about the risk (if the entire machine is dedicated to a single offline task, you probably don't).
In most countries, patents require the invention not to be "obvious to anyone sufficiently skilled in the art". Only in America is "not blatantly obvious to a blithering idiot with some $$$" the criterion.
No special hardware or patents needed.
No sane person ever said any such thing. The whole of Unix was written by Ken and Dennis in their lunch breaks, not just the kernel. Minix was not such a great deal either.
An OS was smaller then, because
(a) there were no GUIs
(b) most of the "utilities" we expect now were not considered part of the system (and that often included the compiler and linker) (c) there was less bloat (Multics excepted).
As is said elsewhere - these days, a processor is a lot more than the CPU. However, most of the rest is well known - how to cache - or externally defined (Memory and peripheral interfaces).
In reality, the system architecture is defined in software (VHDL). Most generation of test software should be highly automated. So, yes ... the problems of an open source processor are mostly political.
There may be no money in open source hardware NOW - but the argument is that so many people have now been shafted by closed source hardware that some might be interested in an alternative.
No one is talking about "free as in beer".
Someone who wants to break into the CPU market (can you say "Chinese") might want a sales pitch that can overcome their current image problems ("it might keep phoning home") - and open source is probably the ONLY tool that can crack that particular nut. And "It keeps phoning the mother ship" (or hackers.ru) is becoming a blunderbuss aimed deep into the heart of closed source. Meltdown and Spectre are "me, too" arguments in reality. You don't need bugs for closed source kit to be a security risk - you cannot tell if it is a security risk even in the absence of bugs. As any reliable and honest crook/casino owner will tell you: "if you can't tell if you are being cheated: you are being cheated".
A lot of large customers buying expensive kit is a very valuable market. "Not being American" is potentially a massive potential selling point to the 90% of the world's population that is not American, but without open access to the design, very few will buy into a product from any of the potential alternative suppliers.
However, I suspect what is really needed is not just open source, but also with a credible second source (eg from two different BRIC countries).
No, No, a thousand times NO!!!!!!
Any project involving a government committing to spending money is doomed, and any IT project involving the British Government, doubly or even quadruply so.
Never mind Babbage, and Harrison's clocks, look what killed the Transputer - Mrs Thatcher promised £50M the same week United Technologies scrapped their microprocessor project saying "in the world of computers, $50M is peanuts". The funding was indeed too little, too late, and the project was sold to the French - who killed it by accident.
That is certainly the Windows buyer PHB perspective.
In reality, large corporations (you have probably heard of IBM) put money into open source because it is a way to share the cost of creating and supporting infrastructure which their actual product depends on.
We have to hope/pray that Larry Ellison has a "Road to Damascus" event, and realises that a truly open Sparc system (not just CPU) might lead to Oracle being in a far stronger position than it already is, and seriously weaken his competition. It might also mean that the nerds who currently detest him might see him as a saviour. (I am not holding my breath).
Currently, Oracle is in the business of preventing access to drivers and microcode for machines over 5 years old and out of support, thus depriving them of second hand value - not exactly helpful to their own business - one can only assume that that Larry is possessed by the ghost of Ebenezer Scrooge and hell bent on wasting the best publicity event that he could dream of.
You don't think that content might be responsible for that?
If you go Bluetooth:
OTOH, with a cable
However, being deaf, you can ignore the whole POS and get on with your life in peace and quiet. Stop wingeing.
Think of the headphones, dammit!
Only pirates require thinner phones!
You seem to mistake journalists for scientists.
Only in America. The rest of the world does not have Trump and his followers, and people go outside and see for themselves.
Some of us have lived longer than 50 years, and been to many parts of the world, where we have families, and we can confirm the changes they describe from our own experience. The climate science is in agreement with this experience as far as we can tell.
The deniers world exists only in America, where most people appear to live in basements, experiencing virtual reality.
I am not defending Mugabe (I will, however, point out that he was deposed, peacefully), or that he destroyed the farming productivity in Zimbabwe, BUT Zimbabwe is not big in the scale of Africa. Its goats that have destroyed good farmland in Africa, and to a lesser extent, continue to do so.
Much of Africa is not starving, partly as a result of proactive policies (eg widespread fish farming in West Africa means fish is cheaper and better than in Europe, massive reforestation program crossing Africa below the Sahara). Good news is not very profitable - people want doom and gloom - and by God, Trump is going to give them that!
There is a massive problem with politics in Africa - though whether its worse than America is somewhat debatable - and much of the problem was due to large amounts of outside influence destroying traditional political structures. (eg colonisation, oil and arms industries).
All food is local? Not here (UK) we import about 75% of our food, and for the same reason Trump was elected, are busy biting the had that feeds us (Brexit).
My globules are feeing pretty warm, despite the freezing weather.
The Darwin Awards should be on prime time TV with scantily clad women and rappers.
That should help control the population!
I drive a Ford - I'd be astonished if it did Mach 0.5!
In most of Africa (ie those parts where there is no actual declared war), the murder rate is significantly lower than in the USA (much the same as in Europe).
Or, if you were to look outside of /., you would find that in a lot of North Africa and the Arab world, nothing has changed in this regard. Saudis selling African slaves (and killing or castrating them) has been much in the news over the past week.
Why bother with self-driving cars? I am waiting for self-driving Pizzas!
I am totally inspired by your post. I am going right out back to build a canoe out of dollar bills!
Marie Antoinette - is that you?
17th century coats (think Pirates of the Caribbean) would be good too! Men's grey business suits are just boring.
I understand that putting your head in the Lion's mouth gets a lot of applause - but not because it is a sensible thing to do!
And mostly works exactly the same.
A closed source rewrite would have cost an arm and a leg to upgrade, had twice as many bugs, and rewriting your configuration files would have taken a year to get stable because of piss-poor documentation,
Disclaimer: I am using Nginx.
The Pascal compiler available on DECUS was definitely ported to other hardware, as was a bunch of the other stuff there - and I am sure some other programs began on Data General and moved to DEC.
In fact, if you got software from your hardware supplier, you normally could get the source if you wanted, and probably wanted to if you had people capable of fixing the bugs - which were many.
Closed Source started with Bill Gates - who wrote a famous letter about how programmers (meaning himself) should get paid. Before the PC, most people would not have bought hardware if they could not get the source. (Of course, with computers costing slightly more than a large house, "most people" would not have bought a computer). With cheaper computers came PT Barnum.