Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

TTL should be regarded as octal number (8-bit unsigned integer) #73158

Open
safwanrahman opened this issue Jun 9, 2020 · 7 comments
Open

TTL should be regarded as octal number (8-bit unsigned integer) #73158

safwanrahman opened this issue Jun 9, 2020 · 7 comments
Labels
A-io Area: `std::io`, `std::fs`, `std::net` and `std::path` C-bug Category: This is a bug. T-libs-api Relevant to the library API team, which will review and decide on the PR/issue.

Comments

@safwanrahman
Copy link

safwanrahman commented Jun 9, 2020

I tried this code:

use std::net::UdpSocket;
fn main(){
    let mut socket = UdpSocket::bind("0.0.0.0:1234").unwrap();
    socket.set_ttl(300);
}

I expected to see this happen: I expected that it would raise compilation error because the ttl is octal and maximum value is 255. But I could set 300 as ttl.

Instead, this happened: It does not raise any error and I could send packet with ttl more than 255. It was causing trouble as my software was setting ttl greater than 255 and was providing wrong data.
Upon looking more into the source code and documentation, it seems like set_ttl method is taking 32-bit unsigned integer type u32 as ttl, where is should take 8-bit unsigned integer type u8 as ttl.

Meta

rustc --version --verbose:

rustc 1.43.1 (8d69840ab 2020-05-04)
binary: rustc
commit-hash: 8d69840ab92ea7f4d323420088dd8c9775f180cd
commit-date: 2020-05-04
host: x86_64-apple-darwin
release: 1.43.1
LLVM version: 9.0
Backtrace

<backtrace>

@safwanrahman safwanrahman added the C-bug Category: This is a bug. label Jun 9, 2020
@safwanrahman safwanrahman changed the title TTL should be regarded as octal number TTL should be regarded as octal number (8-bit unsigned integer) Jun 9, 2020
@tesuji
Copy link
Contributor

tesuji commented Jun 9, 2020

Actually in this case, set_ttl returns an error by converting the C API error to std::io::Result:

warning: unused `std::result::Result` that must be used
 --> src/main.rs:4:5
  |
4 |     socket.set_ttl(300);
  |     ^^^^^^^^^^^^^^^^^^^^
  |
  = note: `#[warn(unused_must_use)]` on by default
  = note: this `Result` may be an `Err` variant, which should be handled

Code: https://play.rust-lang.org/?version=nightly&mode=debug&edition=2018&gist=bfa7dfb16d218160b8758458cd68da21
Output:

[src/main.rs:4] socket.set_ttl(300) = Err(
    Os {
        code: 22,
        kind: InvalidInput,
        message: "Invalid argument",
    },
)

@safwanrahman
Copy link
Author

@lzutao interesting! I did not call unwrap so it did not panic. But I think the parameter should be u8 instead of u32. What do you think?

@tesuji
Copy link
Contributor

tesuji commented Jun 9, 2020

I read the Unix code a bit.

pub fn set_ttl(&self, ttl: u32) -> io::Result<()> {
setsockopt(&self.inner, c::IPPROTO_IP, c::IP_TTL, ttl as c_int)
}

pub fn setsockopt<T>(sock: &Socket, opt: c_int, val: c_int, payload: T) -> io::Result<()> {
unsafe {
let payload = &payload as *const T as *const c_void;
cvt(c::setsockopt(
*sock.as_inner(),
opt,
val,
payload,
mem::size_of::<T>() as c::socklen_t,
))?;
Ok(())
}
}

Maybe we could implement the check for ttl range in Rust side? But what are the advantages
by doing that?

@tesuji
Copy link
Contributor

tesuji commented Jun 9, 2020

Also we might not be able to change the set_ttl parameter type to u8. That would be a breaking change.

@wesleywiser
Copy link
Member

We can't change the parameter type because this is a stable API. Also, because you don't use the Result, you will get a warning:

warning: unused `std::result::Result` that must be used
 --> src/main.rs:4:5
  |
4 |     socket.set_ttl(300);
  |     ^^^^^^^^^^^^^^^^^^^^
  |
  = note: `#[warn(unused_must_use)]` on by default
  = note: this `Result` may be an `Err` variant, which should be handled

@safwanrahman
Copy link
Author

safwanrahman commented Jun 9, 2020

@wesleywiser I understand it is a stable API. but taking u32 as ttl does not make much sense as we can avoide the issue in compilation time rathar than panicing in runtime.

@nagisa
Copy link
Member

nagisa commented Jun 12, 2020

One way to change the input type is to deprecate this function and introduce a new one with a correct type.

@Enselic Enselic added T-libs-api Relevant to the library API team, which will review and decide on the PR/issue. A-io Area: `std::io`, `std::fs`, `std::net` and `std::path` and removed needs-triage-legacy labels Aug 19, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
A-io Area: `std::io`, `std::fs`, `std::net` and `std::path` C-bug Category: This is a bug. T-libs-api Relevant to the library API team, which will review and decide on the PR/issue.
Projects
None yet
Development

No branches or pull requests

6 participants