The PTX emitted can sometimes be of the form:
```
{ code
}
```
Later on if the parser sees that the current and last line were closing
and opening braces, it drops both lines. This prevents any single-line
instructions from showing up. This PR changes the logic to make opcodes
on the same line as opening braces valid, and also adds unit tests for
matching.
## What
Adds [Revive's Resolc](https://github.com/paritytech/revive) compiler
for compiling Solidity and Yul (Solidity IR) to RISC-V and PolkaVM
assembly.
### Main Additions
- [x] Implement new `ResolcCompiler`
- [x] Implement Yul language definition and config for Monaco
- [x] Add Resolc as a compiler for the Solidity and Yul languages
- The `ResolcCompiler` handles both kinds of language input
- [x] Implement initial `PolkaVMAsmParser` (no source mappings)
- [x] Enable viewing LLVM IR in a supplementary view
- [x] Implement a new LLVM IR backend option for toggling between
optimized and unoptimized ll
- Affects non-resolc files ([see
commit](606bab9a59))
- Disabled by default
- (Enable by setting `this.compiler.supportsIrViewOptToggleOption =
true` in a compiler's constructor)
- The compiler's `getIrOutputFilename()` will receive the LLVM IR
backend options
### CE Infra
Accompanying CE Infra PR:
https://github.com/compiler-explorer/infra/pull/1855
## Overall Usage
### Steps
(See screenshots)
* Choose between two input languages:
* Solidity
* Yul (Solidity IR)
* Choose a Resolc compiler
* View assembly:
* PolkaVM assembly (if enabling "Compile to binary")
* RISC-V (64 bits) assembly
* View intermediate results:
* Optimized LLVM IR (if enabling "Show Optimized" in the LLVM IR view)
* Unoptimized LLVM IR
### Notes
Source mappings currently only exist between:
- Yul and RISC-V
- Yul and LLVM-IR
## Screenshots
<img width="1502" height="903" alt="CE Yul RISC-V LLVM IR"
src="https://github.com/user-attachments/assets/7503b9b5-0f2c-4ddf-9405-669e4bdcd02d"
/>
<img width="1502" height="903" alt="CE Solidity PolkaVM"
src="https://github.com/user-attachments/assets/eeb51c99-3eaa-4dda-b13c-ac7783e66cb8"
/>
---------
Co-authored-by: Matt Godbolt <matt@godbolt.org>
After directive filtering, there will be dangling `{}` spanning two
lines. This PR attempts to remove them.
For example,
```
.section .debug_abbrev
{
.b8 1
}
.section .debug_info
{
.b32 70
.b32 .debug_abbrev
}
```
The above assembly used to become the following text after filtering
directive
```
{
}
{
}
```
Recogonize lines with predicate (e.g., `@%p1`, `@!%p1`) as instructions:
- `@%p1 ld.global.v4.b32 { %r9, %r10, %r11, %r12 }, [ %rd3 + 0 ];`
- `@%p2 bra $L__BB0_28;`
- `@!%p2 bra $L__BB0_28;`
| Before | After |
| --- | --- |
| <img width="580" height="320" alt="image"
src="https://github.com/user-attachments/assets/c3913b09-ce04-44c5-a4c0-fa69805cf1d7"
/> | <img width="566" height="323" alt="image"
src="https://github.com/user-attachments/assets/08308b3d-df00-4cf1-8321-203c374c27c3"
/> |
From
https://docs.nvidia.com/cuda/parallel-thread-execution/#instruction-statements:
> Instructions have an optional guard predicate which controls
conditional execution. The guard predicate follows the optional label
and precedes the opcode, and is written as @p, where p is a predicate
register. The guard predicate may be optionally negated, written as @!p.
Implementation-wise, added `(@!?%\w+\s+)?` to existing default
`hasOpcodeRe` (`/^\s*(%[$.A-Z_a-z][\w$.]*\s*=\s*)?[A-Za-z]/`)
Close#5530. Infra:
https://github.com/compiler-explorer/infra/pull/1711. Previous work by
@siboehm at #5531
## Summary
This pull request introduces support for the
[Triton](https://github.com/triton-lang/triton) language, a Python-based
DSL for writing highly efficient GPU kernels.
- [x] **New Language Support**: I've added comprehensive support for the
Triton programming language, allowing users to compile and inspect
Triton kernels within Compiler Explorer. (c.f.,
`lib/compilers/triton.ts`)
- [x] **Python Wrapper for Compilation**: A new Python wrapper script
(`triton_wrapper.py`) has been introduced to manage Triton compilation,
patching its behavior to dump compiled kernels and intermediate
representations without requiring actual execution, and consolidating
the output for Compiler Explorer.
- [x] **Device Assembly View**: Enables viewing of generated device
assembly code (e.g., PTX, AMDGCN) and various intermediate
representations (MLIR, LLVM IR) produced by the Triton compiler.
- [x] **MLIR Parsing**: New parsers (`asm-parser-mlir.ts` and
`mlir-pass-dump-parser.ts`) have been added to correctly interpret and
display MLIR assembly and optimization pass dumps, including source
location information.
- [x] **Multi-Version & Multi-Backend Support**: Painstakingly includes
all 8 versions (from 2.2.0 to 3.3.1) of Triton that supports Python
3.12. Supports both CUDA and HIP backend for Triton 3.
## Screenshots
Source and assembly:
<img width="1354" height="789" alt="image"
src="https://github.com/user-attachments/assets/c29650ff-2073-40e0-a9e6-ff8377094b5e"
/>
Device view for MLIR and LLVM IR:
<img width="1402" height="670" alt="image"
src="https://github.com/user-attachments/assets/43dd5c68-ca78-41b1-9865-e97ffe3ef73c"
/>
Opt pipeline viewer:
<img width="1408" height="668" alt="image"
src="https://github.com/user-attachments/assets/429eef8c-aaac-4781-aafa-39ef0ffc7241"
/>
Diff of TTIR in Triton 3.3.1 vs 2.3.0:
<img width="1580" height="726" alt="image"
src="https://github.com/user-attachments/assets/a928c893-dd9a-4c3a-a048-14046e56a14c"
/>
CUDA & HIP:
<img width="1596" height="800" alt="image"
src="https://github.com/user-attachments/assets/c18800c3-cfad-4e5e-96de-ba92c9f236ea"
/>
## Implementation Details (and Notes for Reviewers)
- For Device Assembly View, I Implemented `MlirAsmParser` for parsing
MLIR assembly. Technically MLIR is not an assembly language, but there
is no better choice to make the source line map work w/ device view.
- I Implemented `MlirPassDumpParser` for processing MLIR optimization
pass dumps. I tried to subclass `LlvmPassDumpParser`, but they turn out
to be too different to worth doing it.
- `LlvmPassDumpParser` made some assumptions that do not hold true for
MLIR passed. Some effort is put to make sure that the passes are
properly diff-ed, since some passes can run multiple times and also
sometimes pass can be nested (i.e., some number of `before`s followed by
some number of `after`s)
- A lot of effort is put into `patch_triton` to make sure that the we
only compile the kernel without actually running it, and that needs to
work across all the versions we support.
## Steps to Run Locally
1. Clone https://github.com/ShawnZhong/compiler-explorer-infra.git
2. Install Triton to `/opt/compiler-explorer/triton`:
```sh
$ cd compiler-explorer-infra
$ ./bin/ce_install install triton
$ ls /opt/compiler-explorer/triton
# v2.2.0 v2.3.0 v2.3.1 v3.0.0 v3.1.0 v3.2.0 v3.3.0 v3.3.1
```
3. Clone https://github.com/ShawnZhong/compiler-explorer.git and
checkout branch `triton`
4. Run Compiler Explorer
```sh
make EXTRA_ARGS='--language triton' dev
```
5. Enjoy
---------
Co-authored-by: Matt Godbolt <matt@godbolt.org>
- latest biome, and fix its configuration
- fixes "static" content to be globally configured too (instead of
per-line)
- fixes issues:
- imports fixed up
- `Date.now()` vs `+new Date()`
- some unused things `_` prefixed
After discussion with the team, turned off the unused parameter warning.
Ensured labels are not mistaken for comments by skipping comment
filtering when the line matches a label definition
Added a regression test verifying that labels starting with “@” remain
visible even when comment filtering is enabled
Fixes#7889
Tested with
npm run lint
npm run ts-check
npm run test-min
npm run check-frontend-imports
make check
