wyldfire

30 ms·

571.
by wyldfire 3y ago
> 2x is a lot better than 50x though. the 50x metric was using something like this AST analysis mechanism (and not a system metric like resident set size)?
572.
by wyldfire 3y ago
> In it he describes how a parsed clang AST regularly consumes 50x more memory than the original source code! I guess that seems like a lot but the context I'm missing is - how good could it get? I mean, if you want to preserve sou
573.
by wyldfire 3y ago
> phase ordering precisely because they don't impact correctness. lol let's say they're less likely to impact correctness than an arbitrary new optimization.
574.
by wyldfire 3y ago
> understanding. We evaluate on a large suite of test programs. Our approach achieves a 3.0% improvement in reducing instruction counts over the compiler, 3% code size reduction is really good. The challenge will be having codegen like
575.
by wyldfire 3y ago
Sure, I get that part -- but then how do people use the predictions?
576.
by wyldfire 3y ago
> Japans only hope was finishing the us Navy early on at Pearl Harbor and midway but once the US ramped up production the US never lost their military lead From Admiral Yamamoto's diary: > I fear all we have done is to awaken a s
577.
by wyldfire 3y ago
> Before I decide to build or modify a tool myself, I thought I'd ask here: if you do decide to, start with ggml/whisper
578.
by wyldfire 3y ago
The firmware blobs execute on the peripherals that happen to be software programmable - not on the CPU (or "apps" CPU). And therefore they're not bound to a particular kernel. IMO the blobs aren't very objectionable w
579.
by wyldfire 3y ago
I'm pretty sure I'm not the intended audience but I also have no idea what this is used for. Surveying? Real estate tycoons? Oil & gas exploration?
580.
by wyldfire 3y ago
I love the way LaTeX output looks and so I've dabbled. I've been very happy with the results using templates written by others. But I occasionally need something slightly customized and I feel a bit hopeless. I've just neve
581.
by wyldfire 3y ago
Why we ever left the PDP I'll never know.
582.
by wyldfire 3y ago
Distro? I dunno probably not without at least some tweaking at least to the DT. Stock android? Yes, perhaps. Some functionality from devices is provided via closed firmware blobs that you'll need too. I'm thinking of the modem o
583.
by wyldfire 3y ago
"we will not" would suggest that any drivers that are present are reverse engineered which is not the case. "Linux 6.2 Expands Support For More Qualcomm Snapdragon SoCs, Apple M1 Pro/Ultra/Max" https:/&#x
584.
by wyldfire 3y ago
> Oh, and we will not upstream our drivers as usual so your devices will be landfill after 4 years as usual This is not genuine argument. Check the sources at kernel.org and you will see tons of Qualcomm drivers upstreamed by Qualcomm.
585.
by wyldfire 3y ago
I'm pretty sure you're mistaken and those were only ARM's claims in their suit filed last year. The trial was set to begin on the 23rd. Not sure what the date is now. But you can see depositions being taken in August [1]. [1
586.
by wyldfire 3y ago
... and is now (again) an architectural licensee.
587.
by wyldfire 3y ago
The "Support WASIX" [1] link shows up there, that's what led to this. [1] https://github.com/ziglang/zig/issues/17115
588.
by wyldfire 3y ago
> Instead it seems like each one of these diagnosticians felt they'd found the "real problem" and pushed ahead with treatment as such. More likely that they heard hoofbeats and were content to use the effectiveness of thei
589.
by wyldfire 3y ago
Interesting blog post by Fangrui Song "ELF interposition and -Bsymbolic" [1] - talks about the real world cost of this feature that's imposed by ELF. On one hand it enables incredible extensibility, OTOH codegen is obligated
590.
by wyldfire 3y ago
It almost seems like it gives credence to self-inflicted problems by implementing changes like that. The creativity of people to break things by misusing them is unbound. Fix your bugs, fix non-bugs that seem like "foreseeable misuse&q
591.
by wyldfire 3y ago
That's not what that says. It refers to unchecked drug use as not worse than firearms, not whether or not firearms could be worse than they are now.
592.
by wyldfire 3y ago
> Forgive me the snark, but I'll believe it when I see it. The last years didn't inspire me with confidence. Indeed, it will be hard to change the minds of all the folks at Qualcomm who have operated here for years under the st
593.
by wyldfire 3y ago
Qualcomm employee here - I agree that we have fallen short of expectations but I don't think it's fair to say "none". I know several folks whose job is kernel development and upstreaming is part of that. It would be gre
594.
by wyldfire 3y ago
The reports say 8cx gen 4 is due out in 2024 and IIRC that's the new Nuvia-designed CPU which should ideally be more competitive than the reference designs from ARM.
595.
by wyldfire 3y ago
Probably couldn't go any worse than it has with firearms.
596.
by wyldfire 3y ago
Wow this sounds like stagefright was way back when. So for those keeping score, is Android now ahead of iOS in this aspect of security?
597.
by wyldfire 3y ago
But ... he didn't need a lawyer. He needed to threaten litigation but you don't need an attorney to do that for you.
598.
by wyldfire 3y ago
Yes, -ish. Qualcomm uses(d) reference designs from ARM from ~2016 to ~2024. Snapdragon 8cx Gen 4, the first Nuvia-designed CPU, is due early next year [1]. [1] https://www.gsmarena.com/qualcomms_new_oryonbased_chipsets_w.
599.
by wyldfire 3y ago
It would be great if C++, Rust, Zig or similar languages could have language features to support invariant assertions. The library solutions are not good enough IMO.
600.
by wyldfire 3y ago
Great job! The world needs more folks to take initiative like this.

More ›