Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The Art of 64-bit Assembly

nostarch.com

Post reply on HN
Print view View original

41–50 of 122 posts

Re: The Art of 64-bit Assembly

#41

Post by ndesaulniers »

sousousou wrote: post #8
amelius wrote: post #6

Is anyone still writing assembly in the age of LLMs? Asking seriously because most assembly is just doing one very simple thing very fast and because it's so simple conceptually, an LLM can easily code it without errors.

Yep. In my experience, LLMs easily go in circles with even simple assembly, creating fixes that result in 2x slower code and then fixing those with even slower code. They are pretty great as a reference or finding needle in the haystack bugs though!

I was able to use Claude to translate a few thousand lines of assembler from one syntax+toolchain to another. Process was definitely iterative; had to keep adding a few rules along the lines of "don't do this...here's the equivalent pattern." But in the end it did a good job and saved me a ton of time. Allowed me to move a ton of unit tests off a hand rolled broken+bad assembler to a modern production toolchain.

Re: The Art of 64-bit Assembly

#42

Post by adamddev1 »

rajeevk wrote: post #17

>> You can ask an AI to explain how vtables work in x86. It will give you something that sounds right. What it won’t give you is what Windows actually expects the vtable to look like, why method dispatch behaves the way it does at the instruction level, or what breaks when you deviate from convention. This volume of The Art of 64-Bit Assembly closes the gap between a plausible explanation and genuine understanding. O…

One book probably doesn't do it. I know the LLMs have been trained on one particular book that explains some crucial things about the grammar of a low-resource language. The LLMs will sort of use the information, terms, and concepts in the book to explain things but they fail to understand and continue to confidently mangle the language and plagerize the content of that book.

Re: The Art of 64-bit Assembly

#43

Post by zero-sharp »

slashdave wrote: post #33
Keyframe wrote: post #14

Earlier quoted context omitted.

an LLM can easily code it without errors can't tell if sarcasm or not. Just in case it's not.. in that case why don't we all vibe everything in assembly? No need for abstractions anymore since that's a human concept. LLM can do it without errors, as you say, and we'll reap the benefits of speed!

Except for some odd corner cases, it is very hard for a good engineer to beat compilers these days (they are very good now). Not to mention, tuning depends on the platform, and often the exact model of chip.

there were two lead developers from FFMPEG on Lex Fridman's youtube channel who talked about performance gain from assembly. They claimed the opposite

https://www.youtube.com/watch?v=IUo0UwZOaRw

Re: The Art of 64-bit Assembly

#45

Post by sitzkrieg »

amelius wrote: post #6

Is anyone still writing assembly in the age of LLMs? Asking seriously because most assembly is just doing one very simple thing very fast and because it's so simple conceptually, an LLM can easily code it without errors.

yes, i write c and asm professionally for small hard real-time mcus. llms give me wrong code nonstop in what little i have tried in my field. not a force multiplier

Re: The Art of 64-bit Assembly

#46

Post by astrange »

rajeevk wrote: post #17

>> You can ask an AI to explain how vtables work in x86. It will give you something that sounds right. What it won’t give you is what Windows actually expects the vtable to look like, why method dispatch behaves the way it does at the instruction level, or what breaks when you deviate from convention. This volume of The Art of 64-Bit Assembly closes the gap between a plausible explanation and genuine understanding. O…

Given the tricolon in that text, I think the content of the book may have come out of an LLM.

Re: The Art of 64-bit Assembly

#47

Post by fasterik »

rajeevk wrote: post #17

>> You can ask an AI to explain how vtables work in x86. It will give you something that sounds right. What it won’t give you is what Windows actually expects the vtable to look like, why method dispatch behaves the way it does at the instruction level, or what breaks when you deviate from convention. This volume of The Art of 64-Bit Assembly closes the gap between a plausible explanation and genuine understanding. O…

I'm skeptical of that statement to begin with. I would be surprised if models weren't trained on plenty of x86 and Windows internals information. If nothing else, they could be prompted to download the relevant material and then write well-commented source code to explain the implementation details.

This reads more like marketing copy targeting anti-AI sentiment than anything.

Re: The Art of 64-bit Assembly

#48

Post by jcranmer »

Retr0id wrote: post #2

> what Windows actually expects the vtable to look like I'm not a Windows-knower, are vtable layouts part of the user kernel ABI?

All of the major kernel ABIs are using C function interfaces as their stable ABI (except Linux, which uses an assembly syscall instruction as the stable interface, although you still rely on a lot of the ancillary C ABI for things like struct layout or stack layout).

For C++ ABIs, there are really only 2.5 major ABIs: the MSVC ABI, used by MSVC and clang wanting to be compatible with MSVC, and the Itanium ABI, used for everything else. There are some slight variants on the Itanium ABI which makes the ".5": ARM uses a different layout for exception handling tables, and there are some flags you can set to use a more compressed vtable layout (32-bit offsets instead of 64-bit pointers). (There are other older ABIs, but either companies stopped making a C++ compiler or they switched to Itanium.)

Windows COM APIs (which isn't part of the kernel, they're still userspace libraries) rely on an IDL which is meant to be directly compatible with the C++ vtable. That said, they also use a restricted subset of C++ that all of the ABIs are going to agree on for vtable layout--if you don't overload any functions, and you don't have any virtual inheritance, there's pretty much only one possible sane vtable layout, and everyone does that.

Re: The Art of 64-bit Assembly

#49

Post by alain94040 »

mdp2021 wrote: post #10
amelius wrote: post #6

Is anyone still writing assembly in the age of LLMs? Asking seriously because most assembly is just doing one very simple thing very fast and because it's so simple conceptually, an LLM can easily code it without errors.

> Is anyone still writing assembly in the age of LLMs ... Yes. Its "simplicity" is exactly why we pick and assemble piece by hand - we imagine the leanest way. LLMs are (not just in Assembly, but especially) very precious as a natural language manual. > an LLM can easily code it without errors One day it will probably also be able to have sex, and yet we think we will not pass on the experience - unless, like some ki…

> Yes. Its "simplicity" is exactly why we pick and assemble piece by hand - we imagine the leanest way

Professionally, assembly is the last thing I would let an LLM generate for me. That's because if you need to write something in assembly, it's because it's critical: operating system context switch, interrupt handler, that kind of thing. You don't want to be 99% correct, you need that code to be 100% good.

I can imagine some other professional scenarios that I'm less familiar with, where you want to write a tight numeric loop of NEON64 assembly and let the LLM do it, because it's frankly a pain to do by hand, and you can test it to some degree of confidence.

Re: The Art of 64-bit Assembly

#50

Post by slashdave »

zero-sharp wrote: post #43
slashdave wrote: post #33

Earlier quoted context omitted.

Except for some odd corner cases, it is very hard for a good engineer to beat compilers these days (they are very good now). Not to mention, tuning depends on the platform, and often the exact model of chip.

there were two lead developers from FFMPEG on Lex Fridman's youtube channel who talked about performance gain from assembly. They claimed the opposite https://www.youtube.com/watch?v=IUo0UwZOaRw

Yeah, a corner case (streaming) and good developers working very hard.

The original SSE (SIMD) extensions were actually written with this use case in mind. When they first came out, the compilers were awful. They've since caught up.

Post reply on HN