revelation

14 ms·

91.
by revelation 9y ago
They are doing the same as the Intel microcode update presumably, offering kernel a bit to enable/disable indirect branch speculation. Usually it will be disabled on entry into kernel and reenabled on exit, so it helps the kernel, not
92.
by revelation 9y ago
I think you're missing that the misdirection happens in the kernel. On an indirect branch (i.e. "jmp <register>") the CPU picks a value for <register> from a small cache when speculatively executing, since the rea
93.
by revelation 9y ago
I don't see how requiring grant holders to spend money on the Elsevier "open access" fees is a win for the council? You can always pay more, the point of negotiation is to get more for less. Mandating that researchers pay t
94.
by revelation 9y ago
They "edit" them in the sense that they pay an Indian outsourcing company to fuck up your formatting, then send you a list of typos that aren't.
95.
by revelation 9y ago
If a critical mass decamps, they still don't have a contract with Elsevier. I don't see how granting free access unilaterally improves their negotiation position other than goodwill in manipulating the executive and legislative go
96.
by revelation 9y ago
I would think it is their huge failure of a drone.
97.
by revelation 9y ago
It feels like you should be able to make a version of branch prediction with speculative execution that behaves the same as the non-speculative-fo-real execution. As in accessing memory you can't actually access doesn't cause a lo
98.
by revelation 9y ago
There is an ARM processor in your SSD
99.
by revelation 9y ago
There is still branch prediction for normal calls or jumps, which are the majority and should be in performance conscious code. It's just that some language features such as virtual functions in C++ often require indirect invocation wh
100.
by revelation 9y ago
retpoline is just a convoluted way of doing an indirect jump/call designed to make branch prediction entirely useless. It's a novel concept because doing this is completely opposite to making a program run faster. Here is an examp
101.
by revelation 9y ago
Attributes is just another word for "global state". When we speak of elegance in computing, it often means functional programming languages where eradicating global state is the primary motivation. So really fork is the opposite o
102.
by revelation 9y ago
Of course. But it needs spelling out, because branch prediction is essential to good performance and why todays CPUs can be much faster despite not reaching higher clocks. It's much more than just the cost of a few more instructions.
103.
by revelation 9y ago
Here is an example with the very common programming patterns that result in indirect jumps/calls: https://godbolt.org/g/LPycpY Every single call rax and jump rax that you see on the right will be replaced with
104.
by revelation 9y ago
I missed that, actually. But you are right, it's not just the crucial bits of inline assembly they are patching here, they expect your kernel to be built with their special version of GCC (yet unreleased) that mutilates every single f
105.
by revelation 9y ago
This isn't the fix with the 5% quoted performance loss (that one clears all of TLB on every syscall), I don't think this is as costly. But it's certainly not free and it's certainly not clean. I'm disappointed Intel
106.
by revelation 9y ago
Love it, an LFENCE on every indirect jump in core 32/64bit entry assembler code, not to mention the cost of the trampoline itself. Linus is right to say that this hack should be reserved for configured as broken CPUs..
107.
by revelation 9y ago
It's counter intuitive at first, but the key to understand how this works is that while you can use an MMU to assign chunks of physical memory to a process, you can of course also just use the MMU to assign the memory mapped IO registe
108.
by revelation 9y ago
Looks like I'm apologizing. Sorry, I was wrong.
109.
by revelation 9y ago
There aren't really any special Android ARM CPUs, maybe they are confident it doesn't really work on Android because it's very difficult to get the timing precision and low-level assembly sequences in Java/ART compiled c
110.
by revelation 9y ago
Applications like this where the syscall overhead (and latency) starts to be a significant factor in processing time and latency have moved to userland drivers anyway: DPDK for 10-100 Gbps networking: https://dpdk.org/ SPDK
111.
by revelation 9y ago
Ideally a mutex is just a cmpxchg. It gets more expensive when it is contended. See the Drepper paper on futexes: http://www.akkadia.org/drepper/futex.pdf
112.
by revelation 9y ago
Linux Format is owned by Future plc, a publicly traded company with more than 200 magazines in their portfolio. Linux Pro is ultimately owned by Marquard Media Group AG, a large swiss media conglomerate. So a lot of those have parents with
113.
by revelation 9y ago
I checked the stock of Intel during the FDIV bug (1994/1995) where they had to go as far as recalling the affected processors at a cost of $500M in January 1995 and there was basically zero effect. By the end of 1995 the stock had actu
114.
by revelation 9y ago
Which is why commercial AV routinely throws up entirely bogus reports. Not a good example. Comparing statistics (accuracy, reaction time, chat messages, what have you) server-side to a not cheating player population seems like a very promis
115.
by revelation 9y ago
Thats because Gentoo decided to switch on a new compile flag, then didn't bother to test that the kernel still boots: https://lkml.org/lkml/2017/12/29/449
116.
by revelation 9y ago
Sure, with the appropriate baking time. But I don't see a cloud company taking an intermediate version of this patchset, backporting it themselves and then sending it out to all their customers in a hurry.
117.
by revelation 9y ago
But no one is hiding the bug this "fixes". If this bug + something else can be a hypervisor escape, well that's too bad for the AWS of this world, but I don't see how this patchset would leak the "something else&quo
118.
by revelation 9y ago
Well Intel for one manufactures the insecure CPUs.. This will be merged for 4.16, when there is no 4.15 release yet. No idea what your cloud computing companies run but it's not 4.15-dirty, and backporting this monster is a great recip
119.
by revelation 9y ago
No ones hiding anything, this patchset was developed in the open for many many months. The hysteria and intrigue in this random tumblr blog is completely superfluous. It's a hardware bug anyway. Here is a good hint to when something is
120.
by revelation 9y ago
No, this makes no sense at all. This is a skit from Dwight of The Office. The guy is standing fully exposed in the open, blinded by light while a bunch of cops have taken cover behind their cars and are aiming their guns at him. What kind o

More ›