This releases includes some pretty nice improvements to the usage of the crate.
If you want to know how the View types I talk about in the release post are built, take a look at my post from back when I contributed them:
This releases includes some pretty nice improvements to the usage of the crate.
If you want to know how the View types I talk about in the release post are built, take a look at my post from back when I contributed them:
Alloca for buffers is generally a bad idea:
The only reasonable usage I have seen is in the Swift ABI. See this blog post on the topic by a rust compiler developer: https://faultlore.com/blah/swift-abi/ (and even there it is only for some cases that it can be used).
First of all: alloca is fun.
Second of all: sub is my favourite allocator
Third of all: doesn’t llvm literally have an alloca primitive for stack allocation? Like it should handle it, no?
Fourth of all: second point is literally a skill issue idk, especially if your compiler is already proving bounds anyway.
Fifth of all: oh damn, that swift thing looks cool!
Sixth of all: don’t take any of this too seriously
Seventh of all: but like actually tho, the nontrivial alloca case not being easily optimizable does not mean that we can’t optimize the bounded case, no?
Eight of all: ig rust doesn’t really do that kind of low-level access.
Ninth of all: this reminds me of those dynamic stack array warcrime gcc extensions, like shit’s craazy, you just size a stack array with an int from stdin.
Tenth level: I suppose the main benefit of this over a stack allocator allocator interface is compile time bounds checks.
Eleventh stage: and maybe it doesn’t quite fit the allocator interface?
12th verse: I’m sane, I promise
Hmm…
As to LLVM and alloca, it doesn’t optimise or even work well in practise. Some basic cases work, others are less well tested. There are lots of “should” that “doesn’t” in practice in LLVM.
I have not looked at alloca in LLVM myself but from what I have heard from those who are experts on this, it is quite brittle.
https://docs.rs/bumpalo/latest/bumpalo/ (and bump allocators in general).
In general proving bounds for stack growth is very difficult. With recursion it is undecidable. This follows directly from Rice’s Theorem. (This is my favourite theorem, it is nice to know that something is impossible rather than a skill issue.)
(Of course you could have a static analyser that instead of yes/no returns yes/no/don’t know, and then you assign don’t know to be either of the other classes depending on if you care more about false positives or false negatives. This is how the rust borrow checker works: forbid if it can’t prove it is safe, but there will be safe code that it doesn’t allow.)