This is not entirely accurate. You have unified memory access in CUDA, and it's been that way for years. The CUDA driver has system-level privileges.
A quick google turned up this NVIDIA blog post. You can dig into the details on CUDA Zone if you're really curious.
Unified memory is also supported in DirectX 12 if the underlying hardware supports it.
In both cases, the driver shuffles data transparently. There are already a lot of attacks that rely on manipulative accesses of memory/cache to ensure that data is being read from desirable locations, so it is conceivable that GPU code could expose kernel memory. After all, the driver that provides memory management would have access to it.
AMD is offers fuller preemptive multitasking. If anything, their hardware is more advanced than NVIDIA's for general compute. (NVIDIA has added some very nice hardware for deep learning though.)
AMD's bigger problem is that NVIDIA did a very good job of promoting CUDA and related proprietary tools. NVIDIA has a huge headstart in putting tools for programming their GPUs in front of developers. Honestly, they've always been better at software and industry partnership than AMD.
Intel isn't releasing the patches though. It's Microsoft, the Linux devs, the BSD devs, etc. who are releasing these patches.
personal desktop that is only used by 1 user who is also the owner of the hardware
You are using software made by other people. You either accept their judgement or roll your own.
Go ahead and fork pre-Spectre Linux if you really believe the Spectre mitigations are a bad idea. You might even find a few people who care about that marginal performance hit to help you---but I doubt it.
Divide the state up into sections for as equal a percentage of population as possible, grouped along county lines.
There is a lot of variation in populations between counties. Counties are not a good starting point.
There are mathematical measures for the compactness of a region. I would specify that each district must be as compact as possible while containing a nearly-identical number of voters. Hand over the census data to a computer or a handful of mathematicians, then wait for the results.
The individual right to equal representation is the primary concern. If a state attempts to suppress the political will of its people, its justification for state powers vanishes.
It's sad that the federal government needs to address this at all. There are both red and blue gerrymanders out there, and they are both morally wrong.
Regardless of the supposed intent, a legal precedent applies across the board. If gerrymandering is upheld as unconstitutional (I'm assuming there will be appeals), then both parties can rely on that precedent in court.
If you care about true representation of the people's will, then you very much want this decision to go to the Supreme Court and be upheld there. That way, it will be binding across the entire country.
Slashdot gerrymandered? Are you insane? It's basically reposting a news article.
The court case involved recent actions by the state of North Carolina. So yeah, Slashdot is talking specifically about North Carolina. Relax that persecution complex a bit, chief.
If you live in Maryland and believe you are affected by unjust gerrymandering, you can pursue a case against the state. In fact, this decision may have set the precedent that allows you to win.
What happens if you find that the concentration of registered citizens doesn't match the list of votes cast?
This is not relevant. A lot of registered voters don't actually go out and vote.
So what you are really arguing is Voter ID, that will verify a person to location mapping.
Most states already have this. You report to a designated polling station based on your home address. I've never lived in a state that worked differently.
If being at the right place on election day is difficult, you can submit an absentee ballot instead.
Without verifiable citizen locations, how could this automation be trusted?
Voter locations are determined based upon their legal residence. Basing it on anything more detailed is going to be both expensive and intrusive.
This is completely orthogonal to the question of gerrymandering anyway.
So what you are really arguing is Voter ID
Once your factual errors are corrected, you have no rational support for Voter ID.
This doesn't make sense. Stage 2 goes all the way to orbit (accepting SpaceX's assertion that Falcon worked entirely as planned). It only 'plunges into the sea' if they deliberately make a deorbit burn.
If you schedule a deorbit burn to occur after separation, but separation fails... where does the satellite end up?
SpaceX is probably waiting for the government or Northrop to admit that the satellite is ash because the docking adapter wouldn't let go.
Or maybe it released, but the satellite thrusters failed so it didn't move. When the Falcon deorbited, it would have either contained... or bumped... the unexpectedly close satellite.
Windows 10 fails to run quite a bit of legacy software developed for windows platform that functions just fine on windows 7 and earlier, this includes some very simple software using nothing but windows libraries.
People complained about this a lot going from XP to Vista/7. The culprit, in most cases, was a better security model for the operating system which broke old applications. I am perfectly fine with this kind of change.
Windows 10 fails to routinely respect user choice in things like allowed update install and reboot times.
This is largely resolved in v1709. You set your active hours, and you can override it when prompted if you're outside the active hours.
Or, for an unsanctioned fix, disable the Windows Update service until you want to install updates.
If you look on Nvidia forums theres a +100 page complaint thread about performance issues appearing over 2017 windows updates
There were all kinds of performance issues with XP, Vista, and 7 in the first couple years after release. And I expect 8 got a free pass because hardly anybody used it. This is not new, and FWIW, I think Vista still holds the crown as the most broken OS.
As a matter of fact I have hard time figuring any advantages windows 10 has over 7
There's lots of little things, which you would probably find if you tried the OS instead of complaining about it. Pinning windows to any/all desktops is the one thing I couldn't live without if I were considering going back to 7/8.
You are right. You can't expect Microsoft to fix their older operating systems that are under support.
They are fixing them.
There are patches for Windows 7, 8, and 10. Because the kernel architecture has evolved over time, the performance deltas are different when patching each OS. Microsoft has been streamlining and modularizing the Windows kernel for a decade, so this performance difference should be expected.
E.g., Microsoft kicked font rendering out of the kernel going from Windows 7 to 8---this means fewer transitions between user and kernel execution when rendering web pages or Office documents. Since the patch penalizes transitions between user and kernel code, this has a direct impact on how badly the patch will affect the system's performance.
Bear in mind, this is only one relevant change. Microsoft makes a lot of changes under the hood with each release of Windows, and I expect there are quite a few more examples. The overall trend of kicking non-essential code out of the kernel (good idea in its own right) is working to their benefit here.
It's pretty fucking obvious that MS are trying to take advantage of the situation to get more users to finally switch over to Spyware 10 by fearmongering.
Not at all.
Since Vista, Microsoft has been on a very deliberate path of reducing dependencies, streamlining interfaces, and stripping legacy cruft within Windows. They have been modularizing the Windows kernel and kicking more functionality into user space for over a decade. This isn't news. I'm sure they're glad that it finally paid off though.
You are conflating the architectural improvements with the intrusive telemetry. They are two entirely different issues. In fact, I wouldn't be surprised if the people who argued in favor of the current kernel design turned around and argued against the always-on telemetry. (Everyone wants telemetry during development; it only becomes stupid when you force users to transmit it.)
Windows 10 with ad hoc measures to suppress telemetry is a much more secure operating system than Windows 7 or 8. If you value security and require Windows, that is where you should land. This is doubly true for enterprise users, as they can acquire Windows licenses which allow them to completely disable telemetry.
Um, hate to break it to you, but that's getting fixed. Under his authority.
All existing payment systems are being merged under the Google Pay banner. Pichai inherited a fragmented patchwork of payment services, and now it's being rolled into a single entity.
He oversaw a lot of cool projects like Chrome, Chrome OS, Google Drive, Gmail, and Google Maps. He was not involved in any of the myriad payment systems.
You're criticizing a man for a problem he didn't create---and is working to fix. If I had to choose an inept party here, it would be the guy who is missing essential facts.
The Linux and Windows kernels are being rewritten in a rather complicated fashion, which includes a performance hit. These changes will have a bigger impact than a typical security patch. No one wants to do something like this unless it is truly necessary.
If all of the developers who have the details agree that something needs to be done, I'm willing to go along with it. When the guys who build something are worried about it falling over, you pay attention.
I would rather not see a POC until a patch is released, tested, and deployed. The implications of this bug are dire, and malware authors can turn a POC into real-world malware in under 48 hours---simple, historical fact.
Vendors have seen security patches reverse-engineered to produce malware within a week, so I'd be inclined to push this onto workstations and public-facing servers ASAP. Full details aren't available publicly yet, so maybe the danger is overblown. But it looks very bad right now, all things considered.
They're being honest, more or less. It's standard to describe what the exploit allows you to do directly.
Being able to read anything in kernel space will allow credential theft, true, but the exploit alone doesn't allow modification of data. Vulnerability reports typically describe exactly what is possible via the exploit and expect the reader to understand the implications---or to ask someone who does.
Anyone who rates vulnerabilities is going to put this into the highest risk category anyway, so it's not like it would slip under the radar. If your organization already prioritizes critical security patches, you're going to fix this as fast as possible.
Why are you posting an installer glitch from 16 years ago? It's not even the same distro mentioned in TFA. It certainly wouldn't have the GNOME release mentioned either.
I would call your post pointless FUD if that distro hadn't already been dead for years.
Trump has called out the intelligence community as being part of some "deep state" conspiracy. True or not, that clearly indicates he will view their work as suspect.
If anyone worked in that agency with a sincere desire to protect the American people and inform their national leaders of threats against the country, then that person's motivation is going to evaporate. When your leader has basically announced that he won't extend any consideration or trust to your organization, is there any value to your work? At the end of the day, what will all of your efforts accomplish?
This is not encouraging at all, but I can understand why they might feel this way.
Private keys for system-level crypto and user credentials are stored in kernel space. You want everyone on the system to be reading those? If you can read a private key or a Kerberos token, you can become that daemon/system/user.
This bug essentially destroys local security and severely compromises network security, subject to any limitations on where/when data can be read.
I'm not a microarchitecture guru who can dig through the details and figure out the limitations of potential attacks. Perhaps only a small portion of kernel memory can be exposed via this bug. I don't really know. The naive, simple scenario where all kernel memory is exposed, though---that is pretty damned bad. Infosec doomsday bad.
There are 12.5 million illegal immigrants in the US. In California, they are given a drivers license which is enough to be able to vote.
Calling bullshit right there. You have to be a citizen before you can register to vote. A driver's license doesn't get you on the voter rolls.
I know this may be confusing, but you can do two different things at one place. The DMV can both issue licenses and register voters---and they can have different rules for each thing. Amazing, right?
Anyway, if you think I'm as full of bullshit as you are, feel free to read it yourself:
Do you live in the US? There are far more polling stations than DMVs.
My nearest DMV is the next town over, and it shared between three cities. I pass 4-5 polling places on my way there, and that's the most direct route.
My polling place is about a mile away---I could easily walk over there to vote. And I didn't count it as being on the way to the DMV because it's down a side road that I wouldn't take.
If you go from (-10, 20, 20, 20, 35) to (-12, 25, 25, 25, 40), your highest and average values have both gone up---and your lowest has gone down. There are, in fact, infinitely many datasets where this can occur.
One could argue that the list isn't exhaustive, and swating could also fall under the felony-murder rule.
I'm not sure they're allowed to go fishing like that.
But didn't Andrew Finch have children? If swatting is known to be dangerous---and people have been shot before, just not killed---then it stands to reason that swatting endangers any children present.
And the law does include a provision for that:
(19) aggravated endangering a child, as defined in subsection (a)(1) of K.S.A. 21-3608a, and amendments thereto.
Android needs this feature badly. I assume there's one for the camera as well? (Not an iPhone user)
Both platforms need OS-level toggles for camera/mic, similar to the one for location. They can even dummy-proof it a little by allowing the native phone app to ignore it, and maybe even allowing other apps to read the setting so they can pop a warning if need be.
Slashdot Mirror
This is not entirely accurate. You have unified memory access in CUDA, and it's been that way for years. The CUDA driver has system-level privileges.
A quick google turned up this NVIDIA blog post. You can dig into the details on CUDA Zone if you're really curious.
Unified memory is also supported in DirectX 12 if the underlying hardware supports it.
In both cases, the driver shuffles data transparently. There are already a lot of attacks that rely on manipulative accesses of memory/cache to ensure that data is being read from desirable locations, so it is conceivable that GPU code could expose kernel memory. After all, the driver that provides memory management would have access to it.
AMD is offers fuller preemptive multitasking. If anything, their hardware is more advanced than NVIDIA's for general compute. (NVIDIA has added some very nice hardware for deep learning though.)
AMD's bigger problem is that NVIDIA did a very good job of promoting CUDA and related proprietary tools. NVIDIA has a huge headstart in putting tools for programming their GPUs in front of developers. Honestly, they've always been better at software and industry partnership than AMD.
Intel isn't releasing the patches though. It's Microsoft, the Linux devs, the BSD devs, etc. who are releasing these patches.
personal desktop that is only used by 1 user who is also the owner of the hardware
You are using software made by other people. You either accept their judgement or roll your own.
Go ahead and fork pre-Spectre Linux if you really believe the Spectre mitigations are a bad idea. You might even find a few people who care about that marginal performance hit to help you---but I doubt it.
Divide the state up into sections for as equal a percentage of population as possible, grouped along county lines.
There is a lot of variation in populations between counties. Counties are not a good starting point.
There are mathematical measures for the compactness of a region. I would specify that each district must be as compact as possible while containing a nearly-identical number of voters. Hand over the census data to a computer or a handful of mathematicians, then wait for the results.
The individual right to equal representation is the primary concern. If a state attempts to suppress the political will of its people, its justification for state powers vanishes.
It's sad that the federal government needs to address this at all. There are both red and blue gerrymanders out there, and they are both morally wrong.
Regardless of the supposed intent, a legal precedent applies across the board. If gerrymandering is upheld as unconstitutional (I'm assuming there will be appeals), then both parties can rely on that precedent in court.
If you care about true representation of the people's will, then you very much want this decision to go to the Supreme Court and be upheld there. That way, it will be binding across the entire country.
This could be a hugely important case.
Slashdot gerrymandered? Are you insane? It's basically reposting a news article.
The court case involved recent actions by the state of North Carolina. So yeah, Slashdot is talking specifically about North Carolina. Relax that persecution complex a bit, chief.
If you live in Maryland and believe you are affected by unjust gerrymandering, you can pursue a case against the state. In fact, this decision may have set the precedent that allows you to win.
What happens if you find that the concentration of registered citizens doesn't match the list of votes cast?
This is not relevant. A lot of registered voters don't actually go out and vote.
So what you are really arguing is Voter ID, that will verify a person to location mapping.
Most states already have this. You report to a designated polling station based on your home address. I've never lived in a state that worked differently.
If being at the right place on election day is difficult, you can submit an absentee ballot instead.
Without verifiable citizen locations, how could this automation be trusted?
Voter locations are determined based upon their legal residence. Basing it on anything more detailed is going to be both expensive and intrusive.
This is completely orthogonal to the question of gerrymandering anyway.
So what you are really arguing is Voter ID
Once your factual errors are corrected, you have no rational support for Voter ID.
This doesn't make sense. Stage 2 goes all the way to orbit (accepting SpaceX's assertion that Falcon worked entirely as planned). It only 'plunges into the sea' if they deliberately make a deorbit burn.
If you schedule a deorbit burn to occur after separation, but separation fails... where does the satellite end up?
SpaceX is probably waiting for the government or Northrop to admit that the satellite is ash because the docking adapter wouldn't let go.
Or maybe it released, but the satellite thrusters failed so it didn't move. When the Falcon deorbited, it would have either contained... or bumped... the unexpectedly close satellite.
Windows 10 fails to run quite a bit of legacy software developed for windows platform that functions just fine on windows 7 and earlier, this includes some very simple software using nothing but windows libraries.
People complained about this a lot going from XP to Vista/7. The culprit, in most cases, was a better security model for the operating system which broke old applications. I am perfectly fine with this kind of change.
Windows 10 fails to routinely respect user choice in things like allowed update install and reboot times.
This is largely resolved in v1709. You set your active hours, and you can override it when prompted if you're outside the active hours.
Or, for an unsanctioned fix, disable the Windows Update service until you want to install updates.
If you look on Nvidia forums theres a +100 page complaint thread about performance issues appearing over 2017 windows updates
There were all kinds of performance issues with XP, Vista, and 7 in the first couple years after release. And I expect 8 got a free pass because hardly anybody used it. This is not new, and FWIW, I think Vista still holds the crown as the most broken OS.
As a matter of fact I have hard time figuring any advantages windows 10 has over 7
There's lots of little things, which you would probably find if you tried the OS instead of complaining about it. Pinning windows to any/all desktops is the one thing I couldn't live without if I were considering going back to 7/8.
You are right. You can't expect Microsoft to fix their older operating systems that are under support.
They are fixing them.
There are patches for Windows 7, 8, and 10. Because the kernel architecture has evolved over time, the performance deltas are different when patching each OS. Microsoft has been streamlining and modularizing the Windows kernel for a decade, so this performance difference should be expected.
E.g., Microsoft kicked font rendering out of the kernel going from Windows 7 to 8---this means fewer transitions between user and kernel execution when rendering web pages or Office documents. Since the patch penalizes transitions between user and kernel code, this has a direct impact on how badly the patch will affect the system's performance.
Bear in mind, this is only one relevant change. Microsoft makes a lot of changes under the hood with each release of Windows, and I expect there are quite a few more examples. The overall trend of kicking non-essential code out of the kernel (good idea in its own right) is working to their benefit here.
Probably. I assume any such reductions are already included in the overall performance metric.
It's pretty fucking obvious that MS are trying to take advantage of the situation to get more users to finally switch over to Spyware 10 by fearmongering.
Not at all.
Since Vista, Microsoft has been on a very deliberate path of reducing dependencies, streamlining interfaces, and stripping legacy cruft within Windows. They have been modularizing the Windows kernel and kicking more functionality into user space for over a decade. This isn't news. I'm sure they're glad that it finally paid off though.
You are conflating the architectural improvements with the intrusive telemetry. They are two entirely different issues. In fact, I wouldn't be surprised if the people who argued in favor of the current kernel design turned around and argued against the always-on telemetry. (Everyone wants telemetry during development; it only becomes stupid when you force users to transmit it.)
Windows 10 with ad hoc measures to suppress telemetry is a much more secure operating system than Windows 7 or 8. If you value security and require Windows, that is where you should land. This is doubly true for enterprise users, as they can acquire Windows licenses which allow them to completely disable telemetry.
Um, hate to break it to you, but that's getting fixed. Under his authority.
All existing payment systems are being merged under the Google Pay banner. Pichai inherited a fragmented patchwork of payment services, and now it's being rolled into a single entity.
He oversaw a lot of cool projects like Chrome, Chrome OS, Google Drive, Gmail, and Google Maps. He was not involved in any of the myriad payment systems.
You're criticizing a man for a problem he didn't create---and is working to fix. If I had to choose an inept party here, it would be the guy who is missing essential facts.
The Linux and Windows kernels are being rewritten in a rather complicated fashion, which includes a performance hit. These changes will have a bigger impact than a typical security patch. No one wants to do something like this unless it is truly necessary.
If all of the developers who have the details agree that something needs to be done, I'm willing to go along with it. When the guys who build something are worried about it falling over, you pay attention.
I would rather not see a POC until a patch is released, tested, and deployed. The implications of this bug are dire, and malware authors can turn a POC into real-world malware in under 48 hours---simple, historical fact.
Vendors have seen security patches reverse-engineered to produce malware within a week, so I'd be inclined to push this onto workstations and public-facing servers ASAP. Full details aren't available publicly yet, so maybe the danger is overblown. But it looks very bad right now, all things considered.
They're being honest, more or less. It's standard to describe what the exploit allows you to do directly.
Being able to read anything in kernel space will allow credential theft, true, but the exploit alone doesn't allow modification of data. Vulnerability reports typically describe exactly what is possible via the exploit and expect the reader to understand the implications---or to ask someone who does.
Anyone who rates vulnerabilities is going to put this into the highest risk category anyway, so it's not like it would slip under the radar. If your organization already prioritizes critical security patches, you're going to fix this as fast as possible.
Why are you posting an installer glitch from 16 years ago? It's not even the same distro mentioned in TFA. It certainly wouldn't have the GNOME release mentioned either.
I would call your post pointless FUD if that distro hadn't already been dead for years.
Trump has called out the intelligence community as being part of some "deep state" conspiracy. True or not, that clearly indicates he will view their work as suspect.
If anyone worked in that agency with a sincere desire to protect the American people and inform their national leaders of threats against the country, then that person's motivation is going to evaporate. When your leader has basically announced that he won't extend any consideration or trust to your organization, is there any value to your work? At the end of the day, what will all of your efforts accomplish?
This is not encouraging at all, but I can understand why they might feel this way.
Private keys for system-level crypto and user credentials are stored in kernel space. You want everyone on the system to be reading those? If you can read a private key or a Kerberos token, you can become that daemon/system/user.
This bug essentially destroys local security and severely compromises network security, subject to any limitations on where/when data can be read.
I'm not a microarchitecture guru who can dig through the details and figure out the limitations of potential attacks. Perhaps only a small portion of kernel memory can be exposed via this bug. I don't really know. The naive, simple scenario where all kernel memory is exposed, though---that is pretty damned bad. Infosec doomsday bad.
There are 12.5 million illegal immigrants in the US. In California, they are given a drivers license which is enough to be able to vote.
Calling bullshit right there. You have to be a citizen before you can register to vote. A driver's license doesn't get you on the voter rolls.
I know this may be confusing, but you can do two different things at one place. The DMV can both issue licenses and register voters---and they can have different rules for each thing. Amazing, right?
Anyway, if you think I'm as full of bullshit as you are, feel free to read it yourself:
http://www.sos.ca.gov/election...
Do you live in the US? There are far more polling stations than DMVs.
My nearest DMV is the next town over, and it shared between three cities. I pass 4-5 polling places on my way there, and that's the most direct route.
My polling place is about a mile away---I could easily walk over there to vote. And I didn't count it as being on the way to the DMV because it's down a side road that I wouldn't take.
Someone has never taken a basic stats course. Or took it and failed horribly.
See my previous reply to you for a counterexample.
Not necessarily.
If you go from (-10, 20, 20, 20, 35) to (-12, 25, 25, 25, 40), your highest and average values have both gone up---and your lowest has gone down. There are, in fact, infinitely many datasets where this can occur.
Do you have any further oversimplifications?
One could argue that the list isn't exhaustive, and swating could also fall under the felony-murder rule.
I'm not sure they're allowed to go fishing like that.
But didn't Andrew Finch have children? If swatting is known to be dangerous---and people have been shot before, just not killed---then it stands to reason that swatting endangers any children present.
And the law does include a provision for that:
(19) aggravated endangering a child, as defined in subsection (a)(1) of K.S.A. 21-3608a, and amendments thereto.
The definition of aggravated child endangerment is at https://law.justia.com/codes/k...
So it may be possible to put him away for a long time. It would be nice to have a criminal attorney from Kansas weigh in on that.
Android needs this feature badly. I assume there's one for the camera as well? (Not an iPhone user)
Both platforms need OS-level toggles for camera/mic, similar to the one for location. They can even dummy-proof it a little by allowing the native phone app to ignore it, and maybe even allowing other apps to read the setting so they can pop a warning if need be.