3 ms·

I sort of see the point of the article but I think talking about differences between pointers and integers obscures the real issue here: the problem as far as I

by simias 8y ago

I sort of see the point of the article but I think talking about differences between pointers and integers obscures the real issue here: the problem as far as I can tell is just that offsetting a pointer more than one byte after the end of the object it points to is UB. TFA argues that converting to integers first sidesteps the issue but I'm not convinced (although I'm not a C standard lawyer). Surely if you convert a pointer to an integer, add an offset that's greater than the size of the original object and then convert that to a pointer you trigger UB? How else could this be defined? You can't really expect the compiler to figure out what you're doing when you're casting "random" integers as pointers. Also the article mentions Rust in its introduction but in Rust anything surrounding pointers (including pointer arithmetics) is unsafe so I don't think there's any conflict whatsoever here.

cjcole 8y ago
Related discussion at internals.rust-lang.org: https://internals.rust-lang.org/t/pointers-are-complicated-or-whats-in-a-byte/8045 https://internals.rust-lang.org/t/pointers-are-complicated-o... "This summer, I will again work (amongst other things) on a “memory model” for Rust/MIR. However, before I can talk about the ideas I have for this year, I have to finally take the time and dispel the myth that “pointers are simple: they are just integers”. Both parts of this statement are false, at least in languages with unsafe features like Rust or C: Pointers are neither simple nor (just) integers. I also want to define a piece of the memory model that has to be fixed before we can even talk about some of the more complex parts: Just what is the data that is stored in memory? It is organized in bytes, the minimal addressable unit and the smallest piece that can be accessed (at least on most platforms), but what are the possible values of a byte? Again, it turns out “it’s just an 8-bit integer” does not actually work as the answer." Ralf Jung is a coauthor of the "What’s the difference between an integer and a pointer?" paper.
- simias 8y ago
  Okay I can agree with that but then there's no need to look for contrived examples, the C standard itself spells out a bunch of UB that arise from using pointers as "just integers". It's probably worth repeating, especially for people new to these languages, but it's not exactly breaking news. I guess I just expected something else when I started reading this article.
toast0 8y ago
The underlying problem is that memory access is slow, allocating registers is hard (so we let the compiler do it) and compilers don't always know what you're doing with memory. So, in order to generate faster programs, compilers are getting more and more agressive about reusing data that has already been copied into registers, hiding under the shadows of 'undefined behavior.' The standard doesn't indicate a defined way to mess with this data so I'll just keep using the copy I have. Writing through an untracked pointer probably convinced the compiler to dump its register caches. C used to be thought of as kind of a portable assembly language, but when the compiler is doing all these optimizations, there's a lot of room for clashing.

by simias 8y ago

cjcole 8y ago

Related discussion at internals.rust-lang.org: https://internals.rust-lang.org/t/pointers-are-complicated-or-whats-in-a-byte/8045 https://internals.rust-lang.org/t/pointers-are-complicated-o... "This summer, I will again work (amongst other things) on a “memory model” for Rust/MIR. However, before I can talk about the ideas I have for this year, I have to finally take the time and dispel the myth that “pointers are simple: they are just integers”. Both parts of this statement are false, at least in languages with unsafe features like Rust or C: Pointers are neither simple nor (just) integers. I also want to define a piece of the memory model that has to be fixed before we can even talk about some of the more complex parts: Just what is the data that is stored in memory? It is organized in bytes, the minimal addressable unit and the smallest piece that can be accessed (at least on most platforms), but what are the possible values of a byte? Again, it turns out “it’s just an 8-bit integer” does not actually work as the answer." Ralf Jung is a coauthor of the "What’s the difference between an integer and a pointer?" paper.

simias 8y ago
Okay I can agree with that but then there's no need to look for contrived examples, the C standard itself spells out a bunch of UB that arise from using pointers as "just integers". It's probably worth repeating, especially for people new to these languages, but it's not exactly breaking news. I guess I just expected something else when I started reading this article.

toast0 8y ago

The underlying problem is that memory access is slow, allocating registers is hard (so we let the compiler do it) and compilers don't always know what you're doing with memory. So, in order to generate faster programs, compilers are getting more and more agressive about reusing data that has already been copied into registers, hiding under the shadows of 'undefined behavior.' The standard doesn't indicate a defined way to mess with this data so I'll just keep using the copy I have. Writing through an untracked pointer probably convinced the compiler to dump its register caches. C used to be thought of as kind of a portable assembly language, but when the compiler is doing all these optimizations, there's a lot of room for clashing.