gnufx

37 ms·

451.
by gnufx 5y ago
> carefully-thought out Unicode policies Like https://www.unicode.org/reports/tr39/tr39-24.html ? The phrase ‘It's not difficult’ is probably rarely applicable in the context of multilingual text — notice
452.
by gnufx 5y ago
I don't know what date "Visual Studio 6" implies, but I'd been using Irix for a few years before GNU Emacs 20 and coding conversion; I can't speak for anything vi-le. Anyway, it's not uncommon to find mixed li
453.
by gnufx 5y ago
See also at least https://github.com/LLNL/RAJA and https://www.khronos.org/sycl/ The great thing about portability layers...
454.
by gnufx 5y ago
Darcs (and Pijul?) can support more patch types than textual diffs, but I doubt much use has ever been made of that. I don't know about the more general case, but it supports the extra type now for identifier replacement, at least as
455.
by gnufx 5y ago
I don't know what exactly is required, but the basic business is gcc -fprofile-generate ... ./a.out ... gcc -fprofile-use ... I don't know the current state of https://gcc.gnu.org/wiki/AutoFDO
456.
by gnufx 5y ago
Depending on what you want, OpenNebula (opennebula.io) has been making a thing of Firecracker.
457.
by gnufx 5y ago
I suspect there's a big variation, particularly in whether the university has placed the keys to the kingdom in the hands of a "partner" of, shall we say, dubious competence and understanding of academia, after getting rid of
458.
by gnufx 5y ago
And we've noticed compromises of multiple "Enterprise VPNs" reported recently.
459.
by gnufx 5y ago
There's a free software self-hostable server called "Headscale" somewhere on Github (not that I've used it). I saw a comment somewhere that Tailscale approved of it. I wonder whether a smaller company of experts(?) whos
460.
by gnufx 5y ago
Every time this sort of thing happens, it's the Home Secretary et al's opportunity to propose yet more measures attacking civil liberties, despite the absence (usually, always?) of evidence what they propose would even have helped
461.
Maqao (Modular Assembly Quality Analyzer and Optimizer)(maqao.org)
2 points by gnufx 5y ago|1 comments
462.
by gnufx 5y ago
I referred to this apropos optimization and I can't see the previous reference I thought there was. x86 only, unfortunately, and the profiler is non-free.
463.
by gnufx 5y ago
Yes, and that's sometimes a good idea. GCC has a pragma and function attribute (unfortunately not for Fortran, at least).
464.
by gnufx 5y ago
End-user experience with "enterprise" VPNs (GlobalProtect currently) leaves me much more trusting of WireGuard and OpenSSH, despite the Enterprise thinking SSH is too insecure to use without the VPN. Early on, I looked at the Glob
465.
by gnufx 5y ago
I see something rather different with that example, using GCC 10 on SKX with that code and program argument, bound to a core, +/- ~2%: -O2 -g: 1.07s -O3 -g: 1.17s -O3 -g -fprofile-generate/use: 0.84s (maqao cqa from https:/&
466.
by gnufx 5y ago
Regarding unrolling, -O3 contains -funroll-and-jam but not -funroll-loops. You may want one or the other, maybe both, depending on circumstances. I don't see much benefit from the available pragmas on HPC-type code unless for OpenMP,
467.
by gnufx 5y ago
Python, specifically, is supposed to get most benefit from -fno-semantic-interposition
468.
by gnufx 5y ago
Yes to -fprofile-use. That brought gfortran to essentially the same geometric mean as ifort with reasonable "good" flags for me, running a benchmark set which I think is supposed to show off proprietary compilers. ifort got litt
469.
by gnufx 5y ago
See https://gcc.gnu.org/gcc-12/changes.html It's still not necessarily a guaranteed win, I think.
470.
by gnufx 5y ago
https://sshuttle.readthedocs.io/ seems worth mentioning for convenient VPN-like tunnelling.
471.
Democracy Risk Report Released(digitalfutures.manchester.ac.uk)
1 points by gnufx 5y ago|0 comments
472.
by gnufx 5y ago
You are very wise. Visualization and analysis in particular, but also facilities, from what I've seen. I couldn't find anything in the Python world to measure a Python application calling native code (like linear algebra), but a
473.
by gnufx 5y ago
Free performance tools (tracing, profiling, visualization) from the HPC world, but more widely applicable, some with decades of R&D behind them, include: BSC's https://tools.bsc.es/ TAU https://www.cs.uo
474.
by gnufx 5y ago
I think in a lot of cases, it's partly a question of where the material was (on paper, not necessarily in libraries, possibly at obscure conferences). That's the case for a various things I'd like to refer to, and the system
475.
by gnufx 5y ago
I find the freshest view is typically in the foundational papers (not necessarily in computer science, but generally). Always go back and look.
476.
by gnufx 5y ago
If you build against, say, RHEL5, you presumably acquire vulnerabilities in relevant libraries, give up hardware support, and still can't guarantee it will run correctly. That's at least because Linux interfaces aren't stabl
477.
by gnufx 5y ago
Sorry, I was talking generally. I try to avoid being personal in posts.
478.
by gnufx 5y ago
I was thinking of the usual complaint about Intel not supporting AMD hardware that is common in HPC. We don't know what that example was actually measuring, except apparently not the same thing for BLIS and MKL. On the basis of only t
479.
by gnufx 5y ago
There are examples of the sort of reason I want dynamic linking under https://news.ycombinator.com/item?id=28734875 and you also really want it for profiling/tracing, and possibly debugging, e.g. in HPC with the MPI PM
480.
by gnufx 5y ago
Drepper's scheme obviously works tolerably well with glibc, in particular, even if he didn't know what he was talking about. If you don't have symbol versioning, and you want to statically link something with parts that depen

More ›