Refactors regmgr_call and privileges #663
Conversation
src/kernel/src/fs/ioctl.rs
Outdated
| pub fn try_from_raw_parts(cmd: u64, arg: *mut u8) -> Result<Self, SysErr> { | ||
| /// # Safety | ||
| /// `arg` has to be a pointer to the correct value | ||
| pub unsafe fn try_from_raw_parts(cmd: u64, arg: SysArg) -> Result<Self, SysErr> { |
There was a problem hiding this comment.
This is even dangerous than the pointer because it hide the fact that the returned value have a lifetime bind to some pointer.
I guess we should required every PRs that use unsafe to be reviewed by me (or someone else who can catch this kind of bug) so we don't have a dangerous bug get merged. Here is an example of what going to happen if we violate the Rust assumption due to unsafe: #48
There was a problem hiding this comment.
I tried something, take a look
There was a problem hiding this comment.
Better to use something like this: https://doc.rust-lang.org/std/ffi/struct.CStr.html#method.from_ptr
There was a problem hiding this comment.
It's that simple? :D
There was a problem hiding this comment.
Wait a minute. To be completely honest, sometimes I don't fully understand lifetimes. The CStr struct isn't lifetimed (which is why from_ptr has a lifetime parameter and returns a lifetimed reference), but our IoCmd and RegMgrCommand are. Doesn't that make any difference?
There was a problem hiding this comment.
The point is try_from_raw_parts need to have an explicit lifetime and document it how it bind to the pointer.
github-actions
bot
added
S-ready
|
We are closing this PR due to no any activities in the last 30 days. Feel free to re-open it if you would like to continue working on this. |
No description provided.