temac

17 ms·

151.
by temac 4y ago
If running in kernel mode under Linux, it is debatable whether most separate binary modules can and are effectively organized in a way they can be considered not a derived work of Linux. The only one I've personally seen managing to do
152.
by temac 4y ago
A lot of kernel code spawn threads. I was not aware that it can also spawn usermode processes running code embedded into the kernel. That should be advertised and developed more IMO. Tons of drivers should have nothing to do in kernelmode.
153.
by temac 4y ago
It can do anything theoretically. It doesn't do anything architecturally. For example, contrary to NT, syscalls can't callback to userspace.
154.
by temac 4y ago
meltdown / rowhammer are pretty fundamental design problems, spectre is even a pretty fundamental logic problem. It hasn't been comprehensively fixed in HW and probably will never be because it directly contradicts the need for
155.
by temac 4y ago
> then the defect is scheduled for being corrected in a new revision of the CPU Note that Intel very rarely issue new revisions of existing models nowadays. Probably creating new masks is too expensive. They simply include some fixes in
156.
by temac 4y ago
> Isn't my "cheap home machine" equivalent to your "most consumer and desktop computers"? Kind of, that's why I was I was saying it is not wrong. IMO though, there should be ECC on everything but the reall
157.
by temac 4y ago
The things is: they are not even rules, while some people think they are, but merely potential tradeoffs, interesting in some situations but that's it. And it is highly conterproductive when people start to cargo cult them, and oh my g
158.
by temac 4y ago
It depends on tons of factors. Yes there are things besides designing and coding, and they can be harder depending on people and projects, but some strong links to the technical parts are always needed, in the sense that a manager is not an
159.
by temac 4y ago
I don't completely understand your conversation. Maybe the error rate is inflated, but as for: > . I can't believe there's no ECC between ram and cpu, it's not credible for anything but a cheap home machine. it is abs
160.
by temac 4y ago
Or it can have catastrophic effects.
161.
by temac 4y ago
It is not reasonable at all for "non-Indian" English -- still that would perhaps be a reasonable explanation, just he should learn it is not reasonable.
162.
by temac 4y ago
We should allow neither.
163.
by temac 4y ago
Windows works (nearly) fine if you block all connections except those from the "Core Networking" group. Of course you won't have access to some services in that case, like Windows Update, etc. One particular weird thing I hav
164.
by temac 4y ago
You can probably configure the firewall to block.
165.
by temac 4y ago
It is disabled in all consumer Alder lake (and I don't remember if there will be Xeon of that gen with P-core only -- IIRC Intel stop the AVX512 validation late on those cores, but it was still before it was formally finished, so proba
166.
by temac 4y ago
VMWare Fusion on M1 was already way behind (to an absurd point: it has been > 1 year 1/2 after the M1 launch, and VMWare has not yet launched a release of Fusion for it, only betas, and they do not target Windows). Overwhelming most
167.
by temac 4y ago
x86 on M1 through Qemu is very slow though. Not really usable to run Windows 10 in good conditions, for example (good news is that this is rarely needed, Windows on Arm running x86 and x64 programs seems quite good)
168.
by temac 4y ago
Would it be to compete with one of their own product in such a major way and despite the integration MS did for Linux guests running on Hyper-V, yeah people would laugh. Not about them using Linux. About them having Hyper-V. Like we laughed
169.
by temac 4y ago
IIRC you can mix different editions in a project, and I think backward incompatibilities are also fewer (and detected at compilation)
170.
by temac 4y ago
I agree with you, and: > computers are incomparibly more efficient Yet to be clear the total world energy consumption for computers has increased, which is sad because we could certainly cope with 1/2 of the current total speed capa
171.
by temac 4y ago
> we need to increase the amount of watts we consume per chip. Not sure we need that, except in niches. At scale you often want at least some efficiency, which is certainly not max TDP per core (because the best efficiency point is wit
172.
by temac 4y ago
> I'd vote to undo these setbacks by moving to local data processing, where a large number of cores each have 1/N of the total memory, shared by M memory busses. Memory controllers would manage shuffling data to where it's
173.
by temac 4y ago
I'm not sure for TPM2 though.
174.
by temac 4y ago
I'm not sure win11 will accept to update if you do that. Or it is just when it detects an "unsupported" processor?
175.
by temac 4y ago
> So in the end, M1 Mini, MacBook Air and *MacBook Pro* differ only in the number of ports and the availability of a built-in display. Yes you were.
176.
by temac 4y ago
ARM ISA has probably little to do with the efficiency, although: * the memory model is slightly more relaxed * instruction decoding is obviously simpler, and while the impact of x86 decoding was not that major at a point, I suspect it has i
177.
by temac 5y ago
You seem to think AMD has no patent on Ryzen. Don't worry, they have. Hopefully not containing loads of diagrams from Intel :D
178.
by temac 5y ago
It's still blatant copyright infringement at least.
179.
by temac 5y ago
I'm using one as a coaster.
180.
by temac 5y ago
If I remember correctly, if an IOMMU is present it is enabled by default on modern Linux versions. Most drivers should be fine now.

More ›