fweimer

10 ms·

151.
by fweimer 9mo ago
I think GCC calls the IPA (inter-procedural analysis) clones. https://gcc.gnu.org/onlinedocs/gccint/IPA-passes.html https://gcc.gnu.org/onlinedocs/gccint/Regular-IPA-passes.htm... https
152.
by fweimer 9mo ago
There should be plenty of existing programming models that can be reused because HPC used single-image multi-hop NUMA systems a lot before the Beowulf clusters took over. Even today, I think very large enterprise systems (where a single ker
153.
by fweimer 9mo ago
I think it's an interesting model. Somehow, the maintenance needs to be funded, and that is an ongoing effort. Charging for security updates is not ideal, but I'm not sure what the alternative would be.
154.
by fweimer 9mo ago
What's wrong with Maven and building fat JARs? It's not dynamic linking, despite excellent support for very late binding in historic Java versions. (Newer versions require specific launcher configurations to use certain platform f
155.
by fweimer 9mo ago
Bundling the JRE in the bundle typically results in something that is not redistributable with the default OpenJDK license: The Java ecosystem is heavily tilted towards the Apache license, but Hotspot is licensed under the GPL v2 only (no c
156.
by fweimer 9mo ago
> I am also deeply concerned about the “speculative” data center market. The “build it and they will come”strategy is a trap. If you are a hyperscaler, you will own your own data centers. Is this actually true? I thought that hyperscaler
157.
by fweimer 9mo ago
> If an LLM is typing that code - and it can maintain a test suite that shows everything works correctly - maybe we don't need that abstraction after all. I'm worried that there is a tendency in LLM-generated code to avoid even
158.
by fweimer 9mo ago
Note that those 13.1 TFLOPs are FP64, which isn't supported natively on the MacBook GPU. On the other hand, local/per-node memory bandwidth is significantly higher on the MacBook. (Apparently, SGI Altix only had 8.5 to 12.8 GB&#x2
159.
by fweimer 9mo ago
Before clusters with fast interconnects were a thing, there were quite a few systems that had more than a thousand hardware threads: https://linuxdevices.org/worlds-largest-single-kernel-linux-... We're slowly getting
160.
by fweimer 9mo ago
But its license is not open source. It even disallows reverse engineering.
161.
by fweimer 9mo ago
At least they are clear about it: It's an editor for open-source AI code. It's not intended to be used for maintaining internal software projects.
162.
by fweimer 9mo ago
Unfortunately, this board seems to be using the CIX CPU that has power management issues: > 15W at idle, which is fairly high
163.
by fweimer 9mo ago
The AGPL does not prevent offering the software as a service. It's got a reputation as the GPL variant for an open-core business model, but it really isn't that. Most companies trying to sell open-source software probably lose mor
164.
by fweimer 9mo ago
These days, people solve similar problems by wrapping their data in an OCI container image and distribute it through one of the container registries that do not have a practically meaningful pull rate limit. Not really a joke, unfortunately
165.
by fweimer 9mo ago
I understood what you meant. I just disagree about priorities. Conceptually, every array access (absent dependent types) can produce a null value because the index might be out of bounds. Languages that eliminate null values in other areas
166.
by fweimer 9mo ago
I'm worried that if memory tagging becomes mandatory, it sucks the air out of the room for solutions that might have a more long-lasting impact. Keep in mind that memory tagging is just heuristics beyond very specific bug scenarios (li
167.
by fweimer 9mo ago
It's possible in practice (at least more so than with Go), but it's highly unusual. Back when free Java became a thing, I used it at first to obtain a memory-safe TLS implementation. It worked out well, I think, but there is a str
168.
by fweimer 9mo ago
I dislike Java's abstraction-through-indirection approach, which is related to the non-representable invalid states you mention. But I think it's more of a matter of taste. Somewhat controversially, I think Java is actually doing
169.
by fweimer 9mo ago
People have tried, and so far, achieving safety through trusted compilers and (fairly complicated) run-time support has been much more efficient. A small team could probably design a RISC-V CPU with extensions for hardware-assisted bounds c
170.
by fweimer 9mo ago
I find it strange that this web site completely ignores the Java ecosystem, which offers memory-safe implementations for most of the protocols and services listed.
171.
by fweimer 9mo ago
These approaches can only detect linear overflows deterministically. Use-after-frees (temporal safety violations) are only detected with some probability. It's mostly a debugging tool. And MTE requires special firmware, which is usuall
172.
by fweimer 9mo ago
At a certain point, it's a trade-off. A systems language will offer facilities that can be used to break encapsulation and abstractions, and access memory as a sequences of bytes. (Anything capable of file I/O on stock Linux can w
173.
by fweimer 9mo ago
The standard does not assign meaning to this sequence of execution, so an implementation can detect this and abort. This is not just hypothetical: existing implementations with pointer capabilities (Fil-C, CHERI targets, possibly even compi
174.
by fweimer 9mo ago
I'm not involved in Go development, only watching from the sidelines. I think it's very likely due to the project dynamics that after the first (published) exploit against real software, the compiler will be changed so that low-le
175.
by fweimer 10mo ago
Did it involve bitfields? GCC is notoriously bad at optimizing them. There are some target-specific optimizations, but pretty much nothing in the middle-end.
176.
by fweimer 10mo ago
Weren't legal protections for semiconductor masks rather lax in the 70s, at least in the United States? You might need certain patent licenses for the manufacturing process, but the chip itself was largely unprotected.
177.
by fweimer 10mo ago
But Unifi should be able to implement this with zero extra hardware, just with VPN-style clients on phones and laptops? I'm just surprised this needs an extra device. It would make sense if the device provided its own connectivity (wit
178.
by fweimer 10mo ago
It seems to me that interest in AArch64 for on-promise general-purpose compute workloads has largely waned. Are Dell/HPE/Lenovo currently selling AArch64 servers? Maybe there is a rack-mounted Nvidia DGX variant, but that's m
179.
by fweimer 10mo ago
I don't think it works that way in practice. Some distributions like Debian or Fedora will make newer features (such as AVX/VEX) mandatory only after the patents expire, if ever. So a new entrant could implement the original x86-6
180.
by fweimer 10mo ago
I think there have been many improvements since 2005 that are not dependent at all on the process node.

More ›