cfallin

24 ms·

211.
by cfallin 11y ago
Not sure about the posted paper, but another recent work that comes to mind is from David Brumley's group at CMU: http://security.ece.cmu.edu/aeg/ "Automatic Exploit Generation" which does basically this
212.
by cfallin 11y ago
That's just about the only possibility, I think -- both forms should generate the same sequence of uops.
213.
by cfallin 11y ago
One example where I've found SFINAE useful is in using "type traits" to tag different classes, and then writing containers to handle the classes differently based on their tags (see [1]). For example, in arena-allocation supp
214.
by cfallin 11y ago
If it's the first time the CPU has seen this code, then the predictor has nothing to go off of, but it will learn "taken" pretty quickly when it sees the `while` loop continue to iterate. However, branch predictors also learn
215.
by cfallin 11y ago
Cool, I have some more reading to do -- thanks for that link! My understanding of TRIPS is that compiler complexity is an issue -- it depends on being able to chunk the program into hyperblocks with fairly high control locality (i.e., remai
216.
by cfallin 11y ago
Interesting! I saw Lava briefly and understood it as basically a DSL to describe logic; additional abstractions to build pipelines on top of that would be very cool. Looking forward to seeing what comes out of your work! And yup, there&#x27
217.
by cfallin 11y ago
Hey cool, someone pointed me here -- surprised to find my toy side-project on HN! I'm not actively hacking on autopiper at the moment (busy with other things) but I still believe the semantics/ideas are sound and I'd love to
218.
by cfallin 11y ago
> but their instruction window does not seem to scale. The issue with scaling instruction windows really has more to do with (i) the scheduling logic and (ii) the physical register file than it does with ISA semantics. In particular, the
219.
by cfallin 11y ago
Yup, all those things are useful, but they don't solve everything. I've worked on projects where a single change to a header file would lead to a 20-minute rebuild adventure and where individual .cc files would take 10 minutes to
220.
by cfallin 11y ago
Well, even if money does matter, the tradeoff is often more complex than just the one-time hardware cost. If you are a software engineer whose time is expensive, and a slower workstation processor means you wait an extra 5 seconds for a com
221.
by cfallin 11y ago
I'm nowhere near an expert in this, but I've heard that electromigration in the really thin wires can cause wearout over time. I imagine that this is probably near-negligible in the micrometer-sized interconnect layers in the 286
222.
by cfallin 11y ago
I'd be interested in seeing a citation for that "next button" work, too. In general HCI is a serious (and useful) field. And while I agree that there's high variability in quality of work, both between and within discipl
223.
by cfallin 11y ago
I think most of the opportunity for "invisible" performance improvement today is below the ISA -- new branch predictors, new prefetchers, smarter cache hierarchy eviction/replacement heuristics. These are problems that are fu
224.
by cfallin 11y ago
Possibly, though it got better with warmer weather, so my first guess was grease viscosity. shrug I hear you on road salt though. I had to be pretty religious about oiling my chain to keep it happy and unrusted.
225.
by cfallin 11y ago
Did you have to do anything special when maintaining your bike? E.g. last time in did winter biking (in Pittsburgh, so not nearly as crazy as you, maybe 10 F / -12 C) the ratchet mechanism in the freehub would often skip and I'd f
226.
by cfallin 11y ago
Yup, in some domains it definitely still makes sense. GPUs work well for highly-data-parallel applications (they're essentially vector machines, modulo branch divergence) and VLIW-style DSPs work because the code is numerical and easy
227.
by cfallin 11y ago
FWIW, conversations originated from Pidgin do show up in chat history in the Hangouts interface. Or at least that's been my experience. I agree that sync in the other direction (to Pidgin), and support for offline messages, would be
228.
by cfallin 11y ago
In addition to what others have said (mainly, it would be specific to a particular version of the CPU, which is probably the biggest thing -- imagine if no x86 binary older than ~1 year could run on your chip), one more point: The programme
229.
by cfallin 11y ago
Oh definitely, I've done my fair share of Digi-Key orders too. The lack of a good actual-local-physical-store option for that one missing resistor on a Saturday is a little unfortunate though. I do wonder what the overall trend is. I m
230.
by cfallin 11y ago
> it's in a position to be cutting edge for the maker scene I remember in the early 2000's there were a lot more discrete electronic components. Last time I went you could still find the soldering irons and oscilloscopes and re
231.
by cfallin 11y ago
You joke a bit but this is one of the biggest things I miss now (in a small apartment in the middle of a city). Growing up in suburbia with a well-stocked garage for tinkering was a wonderful experience. Yeah, hackerspaces; not quite the sa
232.
by cfallin 11y ago
Interesting, I thought that was a PNW (OR/WA/BC) thing (I've heard it described as the "Northwest vowel shift")? I say "bag" and "lag" rhyming with "vague" or "bagel" too, and
233.
by cfallin 11y ago
> Interpreters don't necessarily do GC Of course, but no one claimed that interpreters necessarily do GC either, so this statement is not very useful. You keep arguing against claims that aren't there. The only claim ever mad
234.
by cfallin 11y ago
> GCing has nothing to do with a language being functional Yep, I didn't say it did! You mentioned Haskell and OCaml as counterexamples. I referred to them as "the compiled functional languages". This isn't a claim th
235.
by cfallin 11y ago
I never said precise GC (or even GC in general -- see Boehm GC which I cited) is restricted to JIT! Only that this is perhaps the most common case, because type info still exists at runtime. As I mentioned, the compiler can emit the neces
236.
by cfallin 11y ago
Yep, you're right, it is. And I totally forgot about those compiled functional languages. Thanks! I mentioned JITs and interpreters because those are the obvious cases where type information still exists at runtime, but the compiler ca
237.
by cfallin 11y ago
If you want an imprecise (conservative) garbage collector, the GC and the compiler are almost disjoint. The GC only needs to know where the heap and stack are, and be able to capture register values at GC time. In other words, it's a r
238.
by cfallin 11y ago
> Skylake (2016) and Sandy Bridge (2010) have roughly comparable single core desktop benchmarks. That's not quite true -- Intel aims for a big IPC (instructions/clock) improvement for each "tock" generation (Nehalem -
239.
by cfallin 11y ago
I did something like this when I was doing lots of data analysis for a research project: [1] The problem space I was addressing was more restricted than a fully generic spreadsheet model, but I found everything you say about reproducibility
240.
by cfallin 11y ago
Doesn't this cause issues with cached bogus A records once the user pays and is granted access? I suppose you could return really short TTLs, but there would still be a delay of at least a few seconds. (I'm not doubting they do th

More ›