alex/gba
1
0
Fork 0
mirror of https://github.com/italicsjenga/gba.git synced 2024-12-23 19:01:30 +11:00
Code Issues Packages Projects Releases Wiki Activity

Update Volatile description to match new definition

Browse source
This commit is contained in:
Lokathor 2018-12-20 17:20:59 -07:00
parent a1b85fa98f
commit 2f3becee59
2 changed files with 281 additions and 283 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
book/src/00-introduction/02-goals_and_style.md
View file

@ -21,7 +21,7 @@ Overall the book is sorted for easy review once you're trying to program
something, and the GBA has a few interconnected concepts, so some parts of the
book end up having to refer you to portions that you haven't read yet. The
chapters and sections are sorted so that _minimal_ future references are
required, but it's unavoidable.
required, but it's unavoidable that it'll happen sometimes.
The actual "tutorial order" of the book is the
[Examples](../05-examples/00-index.md) chapter. Each section of that chapter

560
book/src/01-quirks/03-volatile_destination.md
View file

@ -1,321 +1,293 @@
# Volatile Destination
TODO: replace all this one "the rant" is finalized
There's a reasonable chance that you've never heard of `volatile` before, so
what's that? Well, it's a term that can be used in more than one context, but
basically it means "get your grubby mitts off my stuff you over-eager compiler".
TODO: update this when we can make more stuff `const`
## Volatile Memory
The first, and most common, form of volatile thing is volatile memory. Volatile
memory can change without your program changing it, usually because it's not a
location in RAM, but instead some special location that represents an actual
hardware device, or part of a hardware device perhaps. The compiler doesn't know
what's going on in this situation, but when the program is actually run and the
CPU gets an instruction to read or write from that location, instead of just
accessing some place in RAM like with normal memory, it accesses whatever bit of
hardware and does _something_. The details of that something depend on the
hardware, but what's important is that we need to actually, definitely execute
that read or write instruction.
This is not how normal memory works. Normally when the compiler
sees us write values into variables and read values from variables, it's free to
optimize those expressions and eliminate some of the reads and writes if it can,
and generally try to save us time. Maybe it even knows some stuff about the data
dependencies in our expressions and so it does some of the reads or writes out
of order from what the source says, because the compiler knows that it won't
actually make a difference to the operation of the program. A good and helpful
friend, that compiler.
Volatile memory works almost the opposite way. With volatile memory we
need the compiler to _definitely_ emit an instruction to do a read or write and
they need to happen _exactly_ in the order that we say to do it. Each volatile
read or write might have any sort of side effect that the compiler
doesn't know about, and it shouldn't try to be clever about the optimization. Just do what we
say, please.
In Rust, we don't mark volatile things as being a separate type of thing,
instead we use normal raw pointers and then call the
[read_volatile](https://doc.rust-lang.org/core/ptr/fn.read_volatile.html) and
[write_volatile](https://doc.rust-lang.org/core/ptr/fn.write_volatile.html)
functions (also available as methods, if you like), which then delegate to the
LLVM
[volatile_load](https://doc.rust-lang.org/core/intrinsics/fn.volatile_load.html)
and
[volatile_store](https://doc.rust-lang.org/core/intrinsics/fn.volatile_store.html)
intrinsics. In C and C++ you can tag a pointer as being volatile and then any
normal read and write with it becomes the volatile version, but in Rust we have
to remember to use the correct alternate function instead.
I'm told by the experts that this makes for a cleaner and saner design from a
_language design_ perspective, but it really kinda screws us when doing low
level code. References, both mutable and shared, aren't volatile, so they
compile into normal reads and writes. This means we can't do anything we'd
normally do in Rust that utilizes references of any kind. Volatile blocks of
memory can't use normal `.iter()` or `.iter_mut()` based iteration (which give
`&T` or `&mut T`), and they also can't use normal `Index` and `IndexMut` sugar
like `a + x[i]` or `x[i] = 7`.
Unlike with normal raw pointers, this pain point never goes away. There's no way
to abstract over the difference with Rust as it exists now, you'd need to
actually adjust the core language by adding an additional pointer type (`*vol
T`) and possibly a reference type to go with it (`&vol T`) to get the right
semantics. And then you'd need an `IndexVol` trait, and you'd need
`.iter_vol()`, and so on for every other little thing. It would be a lot of
work, and the Rust developers just aren't interested in doing all that for such
a limited portion of their user population. We'll just have to deal with not
having any syntax sugar.
### VolatilePtr
No syntax sugar doesn't mean we can't at least make things a little easier for
ourselves. Enter the `VolatilePtr<T>` type, which is a newtype over a `*mut T`.
One of those "manual" newtypes I mentioned where we can't use our nice macro.
The compiler is an eager friend, so when it sees a read or a write that won't
have an effect, it eliminates that read or write. For example, if we write
```rust
#[derive(Debug, Clone, Copy, Hash, PartialEq, Eq, PartialOrd, Ord)]
let mut x = 5;
x = 7;
```
The compiler won't actually ever put 5 into `x`. It'll skip straight to putting
7 in `x`, because we never read from `x` when it's 5, so that's a safe change to
make. Normally, values are stored in RAM, which has no side effects when you
read and write from it. RAM is purely for keeping notes about values you'll need
later on.
However, what if we had a bit of hardware where we wanted to do a write and that
did something _other than_ keeping the value for us to look at later? As you saw
in the `hello_magic` example, we have to use a `write_volatile` operation.
Volatile means "just do it anyway". The compiler thinks that it's pointless, but
we know better, so we can force it to really do exactly what we say by using
`write_volatile` instead of `write`.
This is kinda error prone though, right? Because it's just a raw pointer, so we
might forget to use `write_volatile` at some point.
Instead, we want a type that's always going to use volatile reads and writes.
Also, we want a pointer type that lets our reads and writes to be as safe as
possible once we've unsafely constructed the initial value.
### Constructing The VolAddress Type
First, we want a type that stores a location within the address space. This can
be a pointer, or a `usize`, and we'll use a `usize` because that's easier to
work with in a `const` context (and we want to have `const` when we can get it).
We'll also have our type use `NonZeroUsize` instead of just `usize` so that
`Option<VolAddress<T>>` stays as a single machine word. This helps quite a bit
when we want to iterate over the addresses of a block of memory (such as
locations within the palette memory). Hardware is never at the null address
anyway. Also, if we had _just_ an address number then we wouldn't be able to
track what type the address is for. We need some
[PhantomData](https://doc.rust-lang.org/core/marker/struct.PhantomData.html),
and specifically we need the phantom data to be for `*mut T`:
* If we used `*const T` that'd have the wrong
[variance](https://doc.rust-lang.org/nomicon/subtyping.html).
* If we used `&mut T` then that's fusing in the ideas of _lifetime_ and
_exclusive access_ to our type. That's potentially important, but that's also
an abstraction we'll build _on top of_ this `VolAddress` type if we need it.
One abstraction layer at a time, so we start with just a phantom pointer. This gives us a type that looks like this:
```rust
#[derive(Debug)]
#[repr(transparent)]
pub struct VolatilePtr<T>(pub *mut T);
```
Obviously we want to be able to read and write:
```rust
impl<T> VolatilePtr<T> {
/// Performs a `read_volatile`.
pub unsafe fn read(self) -> T {
self.0.read_volatile()
}
/// Performs a `write_volatile`.
pub unsafe fn write(self, data: T) {
self.0.write_volatile(data);
}
```
And we want a way to jump around when we do have volatile memory that's in
blocks. This is where we can get ourselves into some trouble if we're not
careful. We have to decide between
[offset](https://doc.rust-lang.org/std/primitive.pointer.html#method.offset) and
[wrapping_offset](https://doc.rust-lang.org/std/primitive.pointer.html#method.wrapping_offset).
The difference is that `offset` optimizes better, but also it can be Undefined
Behavior if the result is not "in bounds or one byte past the end of the same
allocated object". I asked [ubsan](https://github.com/ubsan) (who is the expert
that you should always listen to on matters like this) what that means exactly
when memory mapped hardware is involved (since we never allocated anything), and
the answer was that you _can_ use an `offset` in statically memory mapped
situations like this as long as you don't use it to jump to the address of
something that Rust itself allocated at some point. Cool, we all like being able
to use the one that optimizes better. Unfortunately, the downside to using
`offset` instead of `wrapping_offset` is that with `offset`, it's Undefined
Behavior _simply to calculate the out of bounds result_ (with `wrapping_offset`
it's not Undefined Behavior until you _use_ the out of bounds result). We'll
have to be quite careful when we're using `offset`.
```rust
/// Performs a normal `offset`.
pub unsafe fn offset(self, count: isize) -> Self {
VolatilePtr(self.0.offset(count))
}
```
Now, one thing of note is that doing the `offset` isn't `const`. The math for it
is something that's possible to do in a `const` way of course, but Rust
basically doesn't allow you to fiddle raw pointers much during `const` right
now. Maybe in the future that will improve.
If we did want to have a `const` function for finding the correct address within
a volatile block of memory we'd have to do all the math using `usize` values,
and then cast that value into being a pointer once we were done. It'd look
something like this:
```rust
const fn address_index<T>(address: usize, index: usize) -> usize {
address + (index * std::mem::size_of::<T>())
pub struct VolAddress<T> {
address: NonZeroUsize,
marker: PhantomData<*mut T>,
}
```
But, back to methods for `VolatilePtr`, well we sometimes want to be able to
cast a `VolatilePtr` between pointer types. Since we won't be able to do that
with `as`, we'll have to write a method for it:
Now, because of how `derive` is specified, it derives traits _if the generic
parameter_ supports those traits. Since our type is like a pointer, the traits
it supports are distinct from whatever traits the target type supports. So we'll
provide those implementations manually.
```rust
/// Performs a cast into some new pointer type.
pub fn cast<Z>(self) -> VolatilePtr<Z> {
VolatilePtr(self.0 as *mut Z)
impl<T> Clone for VolAddress<T> {
fn clone(&self) -> Self {
*self
}
}
impl<T> Copy for VolAddress<T> {}
impl<T> PartialEq for VolAddress<T> {
fn eq(&self, other: &Self) -> bool {
self.address == other.address
}
}
impl<T> Eq for VolAddress<T> {}
impl<T> PartialOrd for VolAddress<T> {
fn partial_cmp(&self, other: &Self) -> Option<Ordering> {
Some(self.address.cmp(&other.address))
}
}
impl<T> Ord for VolAddress<T> {
fn cmp(&self, other: &Self) -> Ordering {
self.address.cmp(&other.address)
}
}
```
### Volatile Iterating
Boilerplate junk, not interesting. There's a reason that you derive those traits
99% of the time in Rust.
How about that `Iterator` stuff I said we'd be missing? We can actually make
_an_ Iterator available, it's just not the normal "iterate by shared reference
or unique reference" Iterator. Instead, it's more like a "throw out a series of
`VolatilePtr` values" style Iterator. Other than that small difference it's
totally normal, and we'll be able to use map and skip and take and all those
neat methods.
### Constructing A VolAddress Value
So how do we make this thing we need? First we check out the [Implementing
Iterator](https://doc.rust-lang.org/core/iter/index.html#implementing-iterator)
section in the core documentation. It says we need a struct for holding the
iterator state. Right-o, probably something like this:
Okay so here's the next core concept: If we unsafely _construct_ a
`VolAddress<T>`, then we can safely _use_ the value once it's been properly
created.
```rust
#[derive(Debug, Clone, Hash, PartialEq, Eq)]
pub struct VolatilePtrIter<T> {
vol_ptr: VolatilePtr<T>,
// you'll need these features enabled and a recent nightly
#![feature(const_int_wrapping)]
#![feature(min_const_unsafe_fn)]
impl<T> VolAddress<T> {
pub const unsafe fn new_unchecked(address: usize) -> Self {
VolAddress {
address: NonZeroUsize::new_unchecked(address),
marker: PhantomData,
}
}
pub const unsafe fn cast<Z>(self) -> VolAddress<Z> {
VolAddress {
address: self.address,
marker: PhantomData,
}
}
pub unsafe fn offset(self, offset: isize) -> Self {
VolAddress {
address: NonZeroUsize::new_unchecked(self.address.get().wrapping_add(offset as usize * core::mem::size_of::<T>())),
marker: PhantomData,
}
}
}
```
So what are the unsafety rules here?
* Non-null, obviously.
* Must be aligned for `T`
* Must always produce valid bit patterns for `T`
* Must not be part of the address space that Rust's stack or allocator will ever
uses.
So, again using the `hello_magic` example, we had
```rust
(0x400_0000 as *mut u16).write_volatile(0x0403);
```
And instead we could declare
```rust
const MAGIC_LOCATION: VolAddress<u16> = unsafe { VolAddress::new_unchecked(0x400_0000) };
```
### Using A VolAddress Value
Now that we've named the magic location, we want to write to it.
```rust
impl<T> VolAddress<T> {
pub fn read(self) -> T
where
T: Copy,
{
unsafe { (self.address.get() as *mut T).read_volatile() }
}
pub unsafe fn read_non_copy(self) -> T {
(self.address.get() as *mut T).read_volatile()
}
pub fn write(self, val: T) {
unsafe { (self.address.get() as *mut T).write_volatile(val) }
}
}
```
So if the type is `Copy` we can `read` it as much as we want. If, somehow, the
type isn't `Copy`, then it might be `Drop`, and that means if we read out a
value over and over we could cause the `drop` method to trigger UB. Since the
end user might really know what they're doing, we provide an unsafe backup
`read_non_copy`.
On the other hand, we can `write` to the location as much as we want. Even if
the type isn't `Copy`, _not running `Drop` is safe_, so a `write` is always
safe.
Now we can write to our magical location.
```rust
MAGIC_LOCATION.write(0x0403);
```
### VolAddress Iteration
We've already seen that sometimes we want to have a base address of some sort
and then offset from that location to another. What if we wanted to iterate over
_all the locations_. That's not particularly hard.
```rust
impl<T> VolAddress<T> {
pub const unsafe fn iter_slots(self, slots: usize) -> VolAddressIter<T> {
VolAddressIter { vol_address: self, slots }
}
}
#[derive(Debug)]
pub struct VolAddressIter<T> {
vol_address: VolAddress<T>,
slots: usize,
}
```
impl<T> Clone for VolAddressIter<T> {
fn clone(&self) -> Self {
VolAddressIter {
vol_address: self.vol_address,
slots: self.slots,
}
}
}
impl<T> PartialEq for VolAddressIter<T> {
fn eq(&self, other: &Self) -> bool {
self.vol_address == other.vol_address && self.slots == other.slots
}
}
impl<T> Eq for VolAddressIter<T> {}
impl<T> Iterator for VolAddressIter<T> {
type Item = VolAddress<T>;
And then we just implement
[core::iter::Iterator](https://doc.rust-lang.org/core/iter/trait.Iterator.html)
on that struct. Wow, that's quite the trait though! Don't worry, we only need to
implement two small things and then the rest of it comes free as a bunch of
default methods.
So, the code that we _want_ to write looks like this:
```rust
impl<T> Iterator for VolatilePtrIter<T> {
type Item = VolatilePtr<T>;
fn next(&mut self) -> Option<VolatilePtr<T>> {
fn next(&mut self) -> Option<Self::Item> {
if self.slots > 0 {
let out = Some(self.vol_ptr);
let out = self.vol_address;
unsafe {
self.slots -= 1;
self.vol_ptr = unsafe { self.vol_ptr.offset(1) };
out
self.vol_address = self.vol_address.offset(1);
}
Some(out)
} else {
None
}
}
}
impl<T> FusedIterator for VolAddressIter<T> {}
```
Except we _can't_ write that code. What? The problem is that we used
`derive(Clone, Copy` on `VolatilePtr`. Because of a quirk in how `derive` works,
this means `VolatilePtr<T>` will only be `Copy` if the `T` is `Copy`, _even
though the pointer itself is always `Copy` regardless of what it points to_.
Ugh, terrible. We've got three basic ways to handle this:
### VolAddressBlock
* Make the `Iterator` implementation be for `<T:Clone>`, and then hope that we
always have types that are `Clone`.
* Hand implement every trait we want `VolatilePtr` (and `VolatilePtrIter`) to
have so that we can override the fact that `derive` is basically broken in
this case.
* Make `VolatilePtr` store a `usize` value instead of a pointer, and then cast
it to `*mut T` when we actually need to read and write. This would require us
to also store a `PhantomData<T>` so that the type of the address is tracked
properly, which would make it a lot more verbose to construct a `VolatilePtr`
value.
None of those options are particularly appealing. I guess we'll do the first one
because it's the least amount of up front trouble, and I don't _think_ we'll
need to be iterating non-Clone values. All we do to pick that option is add the
bound to the very start of the `impl` block, where we introduce the `T`:
Obviously, having a base address and a length exist separately is error prone.
There's a good reason for slices to keep their pointer and their length
together. We want something like that, which we'll call a "block" because
"array" and "slice" are already things in Rust.
```rust
impl<T: Clone> Iterator for VolatilePtrIter<T> {
type Item = VolatilePtr<T>;
fn next(&mut self) -> Option<VolatilePtr<T>> {
if self.slots > 0 {
let out = Some(self.vol_ptr.clone());
self.slots -= 1;
self.vol_ptr = unsafe { self.vol_ptr.clone().offset(1) };
out
} else {
None
}
}
}
```
What's going on here? Okay so our iterator has a number of slots that it'll go
over, and then when it's out of slots it starts producing `None` forever. That's
actually pretty simple. We're also masking some unsafety too. In this case,
we'll rely on the person who made the `VolatilePtrIter` to have selected the
correct number of slots. This gives us a new method for `VolatilePtr`:
```rust
pub unsafe fn iter_slots(self, slots: usize) -> VolatilePtrIter<T> {
VolatilePtrIter {
vol_ptr: self,
slots,
}
}
```
With this design, making the `VolatilePtrIter` at the start is `unsafe` (we have
to trust the caller that the right number of slots exists), and then using it
after that is totally safe (if the right number of slots was given we'll never
screw up our end of it).
### VolatilePtr Formatting
Also, just as a little bonus that we probably won't use, we could enable our new
pointer type to be formatted as a pointer value.
```rust
impl<T> core::fmt::Pointer for VolatilePtr<T> {
/// Formats exactly like the inner `*mut T`.
fn fmt(&self, f: &mut core::fmt::Formatter) -> core::fmt::Result {
write!(f, "{:p}", self.0)
}
}
```
Neat!
### VolatilePtr Complete
That was a lot of small code blocks, let's look at it all put together:
```rust
#[derive(Debug, Clone, Copy, PartialEq, Eq, PartialOrd, Ord, Hash)]
#[repr(transparent)]
pub struct VolatilePtr<T>(pub *mut T);
impl<T> VolatilePtr<T> {
pub unsafe fn read(self) -> T {
self.0.read_volatile()
}
pub unsafe fn write(self, data: T) {
self.0.write_volatile(data);
}
pub unsafe fn offset(self, count: isize) -> Self {
VolatilePtr(self.0.offset(count))
}
pub fn cast<Z>(self) -> VolatilePtr<Z> {
VolatilePtr(self.0 as *mut Z)
}
pub unsafe fn iter_slots(self, slots: usize) -> VolatilePtrIter<T> {
VolatilePtrIter {
vol_ptr: self,
slots,
}
}
}
impl<T> core::fmt::Pointer for VolatilePtr<T> {
fn fmt(&self, f: &mut core::fmt::Formatter) -> core::fmt::Result {
write!(f, "{:p}", self.0)
}
}
#[derive(Debug, Clone, Hash, PartialEq, Eq)]
pub struct VolatilePtrIter<T> {
vol_ptr: VolatilePtr<T>,
#[derive(Debug)]
pub struct VolAddressBlock<T> {
vol_address: VolAddress<T>,
slots: usize,
}
impl<T: Clone> Iterator for VolatilePtrIter<T> {
type Item = VolatilePtr<T>;
fn next(&mut self) -> Option<VolatilePtr<T>> {
if self.slots > 0 {
let out = Some(self.vol_ptr.clone());
self.slots -= 1;
self.vol_ptr = unsafe { self.vol_ptr.clone().offset(1) };
out
impl<T> Clone for VolAddressBlock<T> {
fn clone(&self) -> Self {
VolAddressBlock {
vol_address: self.vol_address,
slots: self.slots,
}
}
}
impl<T> PartialEq for VolAddressBlock<T> {
fn eq(&self, other: &Self) -> bool {
self.vol_address == other.vol_address && self.slots == other.slots
}
}
impl<T> Eq for VolAddressBlock<T> {}
impl<T> VolAddressBlock<T> {
pub const unsafe fn new_unchecked(vol_address: VolAddress<T>, slots: usize) -> Self {
VolAddressBlock { vol_address, slots }
}
pub const fn iter(self) -> VolAddressIter<T> {
VolAddressIter {
vol_address: self.vol_address,
slots: self.slots,
}
}
pub unsafe fn index_unchecked(self, slot: usize) -> VolAddress<T> {
self.vol_address.offset(slot as isize)
}
pub fn index(self, slot: usize) -> VolAddress<T> {
if slot < self.slots {
unsafe { self.vol_address.offset(slot as isize) }
} else {
panic!("Index Requested: {} >= Bound: {}", slot, self.slots)
}
}
pub fn get(self, slot: usize) -> Option<VolAddress<T>> {
if slot < self.slots {
unsafe { Some(self.vol_address.offset(slot as isize)) }
} else {
None
}
@ -323,6 +295,26 @@ impl<T: Clone> Iterator for VolatilePtrIter<T> {
}
```
Now we can have something like:
```rust
const OTHER_MAGIC: VolAddressBlock<u16> = unsafe {
VolAddressBlock::new_unchecked(
VolAddress::new_unchecked(0x600_0000),
240 * 160
)
};
OTHER_MAGIC.index(120 + 80 * 240).write_volatile(0x001F);
OTHER_MAGIC.index(136 + 80 * 240).write_volatile(0x03E0);
OTHER_MAGIC.index(120 + 96 * 240).write_volatile(0x7C00);
```
### Docs?
If you wanna see these types and methods with a full docs write up you should
check the GBA crate's source.
## Volatile ASM
In addition to some memory locations being volatile, it's also possible for
@ -343,3 +335,9 @@ safely (otherwise the GBA won't ever actually wake back up from the low power
state), but the `asm!` you use once you're ready is just a single instruction
with no return value. The compiler can't tell what's going on, so you just have
to say "do it anyway".
Note that if you use a linker script to include any ASM with your Rust program
(eg: the `crt0.s` file that we setup in the "Development Setup" section), all of
that ASM is "volatile" for these purposes. Volatile isn't actually a _hardware_
concept, it's just an LLVM concept, and the linker script runs after LLVM has
done its work.