| Age | Commit message (Collapse) | Author | Lines |
|---|
|
TypeTree support in autodiff
# TypeTrees for Autodiff
## What are TypeTrees?
Memory layout descriptors for Enzyme. Tell Enzyme exactly how types are structured in memory so it can compute derivatives efficiently.
## Structure
```rust
TypeTree(Vec<Type>)
Type {
offset: isize, // byte offset (-1 = everywhere)
size: usize, // size in bytes
kind: Kind, // Float, Integer, Pointer, etc.
child: TypeTree // nested structure
}
```
## Example: `fn compute(x: &f32, data: &[f32]) -> f32`
**Input 0: `x: &f32`**
```rust
TypeTree(vec![Type {
offset: -1, size: 8, kind: Pointer,
child: TypeTree(vec![Type {
offset: -1, size: 4, kind: Float,
child: TypeTree::new()
}])
}])
```
**Input 1: `data: &[f32]`**
```rust
TypeTree(vec![Type {
offset: -1, size: 8, kind: Pointer,
child: TypeTree(vec![Type {
offset: -1, size: 4, kind: Float, // -1 = all elements
child: TypeTree::new()
}])
}])
```
**Output: `f32`**
```rust
TypeTree(vec![Type {
offset: -1, size: 4, kind: Float,
child: TypeTree::new()
}])
```
## Why Needed?
- Enzyme can't deduce complex type layouts from LLVM IR
- Prevents slow memory pattern analysis
- Enables correct derivative computation for nested structures
- Tells Enzyme which bytes are differentiable vs metadata
## What Enzyme Does With This Information:
Without TypeTrees (current state):
```llvm
; Enzyme sees generic LLVM IR:
define float ``@distance(ptr*`` %p1, ptr* %p2) {
; Has to guess what these pointers point to
; Slow analysis of all memory operations
; May miss optimization opportunities
}
```
With TypeTrees (our implementation):
```llvm
define "enzyme_type"="{[]:Float@float}" float ``@distance(``
ptr "enzyme_type"="{[]:Pointer}" %p1,
ptr "enzyme_type"="{[]:Pointer}" %p2
) {
; Enzyme knows exact type layout
; Can generate efficient derivative code directly
}
```
# TypeTrees - Offset and -1 Explained
## Type Structure
```rust
Type {
offset: isize, // WHERE this type starts
size: usize, // HOW BIG this type is
kind: Kind, // WHAT KIND of data (Float, Int, Pointer)
child: TypeTree // WHAT'S INSIDE (for pointers/containers)
}
```
## Offset Values
### Regular Offset (0, 4, 8, etc.)
**Specific byte position within a structure**
```rust
struct Point {
x: f32, // offset 0, size 4
y: f32, // offset 4, size 4
id: i32, // offset 8, size 4
}
```
TypeTree for `&Point` (internal representation):
```rust
TypeTree(vec![
Type { offset: 0, size: 4, kind: Float }, // x at byte 0
Type { offset: 4, size: 4, kind: Float }, // y at byte 4
Type { offset: 8, size: 4, kind: Integer } // id at byte 8
])
```
Generates LLVM:
```llvm
"enzyme_type"="{[]:Float@float}"
```
### Offset -1 (Special: "Everywhere")
**Means "this pattern repeats for ALL elements"**
#### Example 1: Array `[f32; 100]`
```rust
TypeTree(vec![Type {
offset: -1, // ALL positions
size: 4, // each f32 is 4 bytes
kind: Float, // every element is float
}])
```
Instead of listing 100 separate Types with offsets `0,4,8,12...396`
#### Example 2: Slice `&[i32]`
```rust
// Pointer to slice data
TypeTree(vec![Type {
offset: -1, size: 8, kind: Pointer,
child: TypeTree(vec![Type {
offset: -1, // ALL slice elements
size: 4, // each i32 is 4 bytes
kind: Integer
}])
}])
```
#### Example 3: Mixed Structure
```rust
struct Container {
header: i64, // offset 0
data: [f32; 1000], // offset 8, but elements use -1
}
```
```rust
TypeTree(vec![
Type { offset: 0, size: 8, kind: Integer }, // header
Type { offset: 8, size: 4000, kind: Pointer,
child: TypeTree(vec![Type {
offset: -1, size: 4, kind: Float // ALL array elements
}])
}
])
```
|
|
|
|
These code paths are surprisingly hot in the `large-workspace` benchmark; it
would be handy to see some detailed timings and execution counts.
|
|
Signed-off-by: Karan Janthe <karanjanthe@gmail.com>
|
|
|
|
This makes it clearer that it is set by the build system rather than by
the rustc that compiles the current rustc. It also avoids bootstrap
needing to pass --check-cfg llvm_enzyme to rustc.
|
|
It is always false nowadays. ThinLTO summary writing is instead done by
llvm_optimize.
|
|
It is only used within cg_llvm.
|
|
|
|
A lot of places had special handling just in case they would get an
allocator module even though most of these places could never get one or
would have a trivial implementation for the allocator module. Moving all
handling of the allocator module to a single place simplifies things a
fair bit.
|
|
|
|
All other sanitizer symbols are handled in prepare_lto already.
|
|
Don't export them from cdylibs. There is no need to do so and it
complicates exported_non_generic_symbols. In addition the GCC backend
likely uses different symbols and may potentially not even need us to
explicitly tell it to export the symbols it needs.
|
|
Handle unwinding fatal errors in codegen workers
Also directly unwind on fatal errors at the point they are emitted inside the codegen backends.
Fixes the coordinator ICE of https://github.com/rust-lang/rust/issues/132240, https://github.com/rust-lang/rust/issues/135075 and https://github.com/rust-lang/rust/issues/145800.
|
|
This avoids the need for `#![allow(non_upper_case_globals)]`.
|
|
As opposed to passing it around through Result.
|
|
Remove `LlvmArchiveBuilder` and supporting code/bindings
Switching over to the newer Rust-based `ArArchiveBuilder` happened in rust-lang/rust#128936, a year ago.
Per the comment in `new_archive_builder`, that seems like enough time to justify removing the older, unused `LlvmArchiveBuilder` implementation and its associated bindings.
Fixes rust-lang/rust#128955.
|
|
cg_llvm: Small cleanups to `owned_target_machine`
This PR contains a few tiny cleanups to the `owned_target_machine` code.
Each individual commit should be fairly straightforward.
|
|
|
|
|
|
The `NonNull::as_mut` method returns a mut *reference*, rather than the mut
*pointer* that is intended here.
|
|
As noted in the `ffi` module docs, passing pointer/length byte strings from
Rust to C++ is easier if we declare them as `*const c_uchar` on the Rust side,
but `const char *` (possibly signed) on the C++ side. This is allowed because
both pointer types are ABI-compatible, regardless of char signedness.
|
|
Declaring these submodules directly in `lib.rs` was needlessly confusing.
|
|
Couple of minor cleanups
|
|
|
|
It used to be necessary on Apple platforms to ship with the App Store,
but XCode 15 has stopped embedding LLVM bitcode and the App Store no
longer accepts apps with bitcode embedded.
|
|
Rollup of 7 pull requests
Successful merges:
- rust-lang/rust#144072 (update `Atomic*::from_ptr` and `Atomic*::as_ptr` docs)
- rust-lang/rust#144151 (`tests/ui/issues/`: The Issues Strike Back [1/N])
- rust-lang/rust#144300 (Clippy fixes for miropt-test-tools)
- rust-lang/rust#144399 (Add a ratchet for moving all standard library tests to separate packages)
- rust-lang/rust#144472 (str: Mark unstable `round_char_boundary` feature functions as const)
- rust-lang/rust#144503 (Various refactors to the codegen coordinator code (part 3))
- rust-lang/rust#144530 (coverage: Infer `instances_used` from `pgo_func_name_var_map`)
r? `@ghost`
`@rustbot` modify labels: rollup
|
|
|
|
Nobody seems to actually use this, while still adding some extra
complexity to the already rather complex codegen coordinator code.
It is also not supported by any backend other than the LLVM backend.
|
|
|
|
Various refactors to the LTO handling code (part 2)
Continuing from https://github.com/rust-lang/rust/pull/143388 this removes a bit of dead code and moves the LTO symbol export calculation from individual backends to cg_ssa.
|
|
gpu offload host code generation
r? ghost
This will generate most of the host side code to use llvm's offload feature.
The first PR will only handle automatic mem-transfers to and from the device.
So if a user calls a kernel, we will copy inputs back and forth, but we won't do the actual kernel launch.
Before merging, we will use LLVM's Info infrastructure to verify that the memcopies match what openmp offloa generates in C++. `LIBOMPTARGET_INFO=-1 ./my_rust_binary` should print that a memcpy to and later from the device is happening.
A follow-up PR will generate the actual device-side kernel which will then do computations on the GPU.
A third PR will implement manual host2device and device2host functionality, but the goal is to minimize cases where a user has to overwrite our default handling due to performance issues.
I'm trying to get a full MVP out first, so this just recognizes GPU functions based on magic names. The final frontend will obviously move this over to use proper macros, like I'm already doing it for the autodiff work.
This work will also be compatible with std::autodiff, so one can differentiate GPU kernels.
Tracking:
- https://github.com/rust-lang/rust/issues/131513
|
|
|
|
|
|
And move exported_symbols_for_lto call from backends to cg_ssa.
|
|
|
|
The modules vec can already contain serialized modules and there is no
need to distinguish between cached and non-cached cgus at LTO time.
|
|
|
|
code generation
|
|
This is no longer implied by -wasm-enable-eh.
|
|
Various refactors to the LTO handling code
In particular reducing the sharing of code paths between fat and thin-LTO and making the fat LTO implementation more self-contained. This also moves some autodiff handling out of cg_ssa into cg_llvm given that Enzyme only works with LLVM anyway and an implementation for another backend may do things entirely differently. This will also make it a bit easier to split LTO handling out of the coordinator thread main loop into a separate loop, which should reduce the complexity of the coordinator thread.
|
|
|
|
fix: correct assertion to check for 'noinline' attribute presence before removal
|
|
|
|
|
|
|
|
Most uses of it either contain a fat or thin lto module. Only
WorkItem::LTO could contain both, but splitting that enum variant
doesn't complicate things much.
|
|
Signed-off-by: Karan Janthe <karanjanthe@gmail.com>
|
|
There is no safety contract and I don't think any of them can actually
cause UB in more ways than passing malicious source code to rustc can.
While LtoModuleCodegen::optimize says that the returned ModuleCodegen
points into the LTO module, the LTO module has already been dropped by
the time this function returns, so if the returned ModuleCodegen indeed
points into the LTO module, we would have seen crashes on every LTO
compilation, which we don't. As such the comment is outdated.
|
|
This patch also changes the Rust-side declaration to take `*const c_uchar`
instead of `*const c_char`, to avoid the need for `AsCCharPtr`.
|
