Skip to content

fix(zram): the cuda feature says why it cannot build, instead of a cascade #3267

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.

Sign up for GitHub

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

Closed
noahgift wants to merge 1 commit into from
Closed

fix(zram): the cuda feature says why it cannot build, instead of a cascade #3267

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 5 additions & 0 deletions crates/aprender-zram-core/Cargo.toml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -44,6 +44,11 @@ std = []
nightly = []
# Enable CUDA GPU acceleration
cuda = ["dep:cudarc", "dep:rayon", "dep:aprender-gpu"]
# Private. Carries the REAL gate on the GPU batch path in src/gpu/batch/mod.rs,
# so `--features cuda` reports the one named refusal in that file instead of a
# cascade of unresolved imports. Set it only once the three gaps that refusal
# names are closed. See #3179.
__cuda-linked = []

[lints.rust]
unsafe_code = "allow"
Expand Down
59 changes: 44 additions & 15 deletions crates/aprender-zram-core/src/gpu/batch/mod.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,6 +5,35 @@
//!
//! For 4KB pages with ~2:1 ratio, batch 1000+ pages to amortize PCIe overhead.

// #3179 — the GPU batch path cannot run, and has not compiled since APR-MONO.
// Three gaps, each verified by reading the tree rather than inferred:
//
// 1. `trueno_gpu::driver::PinnedBuffer` and `GpuBuffer::copy_{from,to}_pinned_async`
// do not exist and never did in this monorepo (`git log -S PinnedBuffer --
// crates/aprender-gpu/` is empty). Only synchronous pageable copies exist.
// 2. This file launches the entry point `lz4_compress_warp_shuffle` with 32
// threads and no shared memory. No kernel of that name exists; the one that
// does, `lz4_compress_warp`, wants 128 threads and shared memory.
// 3. `Lz4WarpDecompressKernel::build_ptx` is a STUB whose whole body is
// `ctx.label("L_exit")`. There is no GPU LZ4 decompressor in this workspace.
//
// So this is not one API away from working: it is wired to a kernel ABI removed
// at consolidation, and its decompressor does not exist. Making it merely
// COMPILE would hand back a feature that is green in the matrix and dead at
// runtime — the theater this repo refuses. The refusal names all three gaps and
// the private `__cuda-linked` feature carries the real gate, so nothing here is
// deleted and the diagnostic is one line instead of a cascade.
#[cfg(all(feature = "cuda", not(feature = "__cuda-linked")))]
compile_error!(
"aprender-zram-core's `cuda` feature cannot build: the GPU batch path needs \
(1) trueno_gpu::driver::PinnedBuffer + GpuBuffer::copy_{from,to}_pinned_async, \
which do not exist; (2) a kernel named `lz4_compress_warp_shuffle`, which does \
not exist (aprender-gpu ships `lz4_compress_warp`, 128 threads + shared memory); \
and (3) a real LZ4 decompress kernel — Lz4WarpDecompressKernel::build_ptx is a \
stub that exits immediately. See #3179. The non-cuda build of this crate is \
unaffected."
);

use crate::error::{Error, Result};
use crate::page::CompressedPage;
use crate::{Algorithm, PAGE_SIZE};
Expand Down Expand Up @@ -43,7 +72,7 @@ impl Default for GpuBatchConfig {
#[allow(dead_code)] // Fields used when CUDA kernels are fully implemented
pub struct GpuBatchCompressor {
config: GpuBatchConfig,
#[cfg(feature = "cuda")]
#[cfg(feature = "__cuda-linked")]
context: Option<GpuContext>,
/// Cached SIMD compressor for parallel CPU path (Arc for thread-safety)
simd_compressor: std::sync::Arc<Box<dyn crate::PageCompressor>>,
Expand All @@ -64,7 +93,7 @@ impl std::fmt::Debug for GpuBatchCompressor {
}
}

#[cfg(feature = "cuda")]
#[cfg(feature = "__cuda-linked")]
struct GpuContext {
/// CUDA context handle (trueno-gpu driver).
context: trueno_gpu::driver::CudaContext,
Expand Down Expand Up @@ -92,7 +121,7 @@ struct GpuContext {
max_batch_size: usize,
}

#[cfg(feature = "cuda")]
#[cfg(feature = "__cuda-linked")]
#[allow(clippy::missing_fields_in_debug)]
impl std::fmt::Debug for GpuContext {
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
Expand Down Expand Up @@ -208,7 +237,7 @@ impl GpuBatchCompressor {
.build()?,
);

#[cfg(feature = "cuda")]
#[cfg(feature = "__cuda-linked")]
{
// Initialize CUDA context with pre-allocated buffers for batch_size
let context = Self::init_cuda(config.device_index, config.batch_size)?;
Expand All @@ -223,7 +252,7 @@ impl GpuBatchCompressor {
})
}

#[cfg(not(feature = "cuda"))]
#[cfg(not(feature = "__cuda-linked"))]
{
Ok(Self {
config,
Expand All @@ -236,7 +265,7 @@ impl GpuBatchCompressor {
}
}

#[cfg(feature = "cuda")]
#[cfg(feature = "__cuda-linked")]
fn init_cuda(device_index: u32, batch_size: usize) -> Result<GpuContext> {
use trueno_gpu::driver::{CudaContext, CudaModule, CudaStream, GpuBuffer, PinnedBuffer};
use trueno_gpu::kernels::lz4::{Lz4DecompressKernel, Lz4WarpShuffleKernel};
Expand Down Expand Up @@ -416,20 +445,20 @@ impl GpuBatchCompressor {
});
}

#[cfg(feature = "cuda")]
#[cfg(feature = "__cuda-linked")]
{
self.compress_batch_cuda(pages)
}

#[cfg(not(feature = "cuda"))]
#[cfg(not(feature = "__cuda-linked"))]
{
Err(Error::GpuNotAvailable(
"CUDA feature not enabled".to_string(),
))
}
}

#[cfg(feature = "cuda")]
#[cfg(feature = "__cuda-linked")]
fn compress_batch_cuda(&mut self, pages: &[[u8; PAGE_SIZE]]) -> Result<BatchResult> {
use std::time::Instant;

Expand Down Expand Up @@ -472,7 +501,7 @@ impl GpuBatchCompressor {
})
}

#[cfg(feature = "cuda")]
#[cfg(feature = "__cuda-linked")]
fn transfer_to_device(
&self,
pages: &[[u8; PAGE_SIZE]],
Expand All @@ -498,7 +527,7 @@ impl GpuBatchCompressor {
Ok(device_buffer)
}

#[cfg(feature = "cuda")]
#[cfg(feature = "__cuda-linked")]
fn execute_compression_kernel(&self, pages: &[[u8; PAGE_SIZE]]) -> Result<Vec<Vec<u8>>> {
use std::ffi::c_void;
use std::time::Instant;
Expand Down Expand Up @@ -718,7 +747,7 @@ impl GpuBatchCompressor {
}

/// CPU fallback for compression when GPU is not available
#[cfg(feature = "cuda")]
#[cfg(feature = "__cuda-linked")]
fn execute_compression_kernel_cpu(&self, pages: &[[u8; PAGE_SIZE]]) -> Result<Vec<Vec<u8>>> {
use rayon::prelude::*;
use trueno_gpu::kernels::lz4::lz4_compress_block;
Expand All @@ -737,7 +766,7 @@ impl GpuBatchCompressor {
results
}

#[cfg(feature = "cuda")]
#[cfg(feature = "__cuda-linked")]
fn transfer_from_device(&self, data: Vec<Vec<u8>>) -> Result<Vec<CompressedPage>> {
// In the current hybrid approach, data is already on host.
// When nvCOMP is integrated, this will do D2H transfer.
Expand Down Expand Up @@ -786,7 +815,7 @@ impl GpuBatchCompressor {
/// The kernel achieves 32 GB/s internally.
///
/// G.119 COMPLIANT: Eliminates 350ms allocation overhead by reusing buffers.
#[cfg(feature = "cuda")]
#[cfg(feature = "__cuda-linked")]
pub fn decompress_batch_gpu(
&mut self,
compressed: &[Vec<u8>],
Expand Down Expand Up @@ -1036,7 +1065,7 @@ impl GpuBatchCompressor {
///
/// Zero pages are extremely common in memory (>30% typically) and compress
/// to just a few bytes with LZ4's RLE-style encoding.
#[cfg(feature = "cuda")]
#[cfg(feature = "__cuda-linked")]
#[allow(dead_code)]
fn encode_lz4_zero_page() -> Vec<u8> {
use trueno_gpu::kernels::lz4::lz4_compress_block;
Expand Down
14 changes: 7 additions & 7 deletions crates/aprender-zram-core/src/gpu/hybrid.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -172,7 +172,7 @@ impl HybridScheduler {
/// # Errors
///
/// Returns error if decompression fails.
#[cfg(feature = "cuda")]
#[cfg(feature = "__cuda-linked")]
pub fn decompress_batch_gpu(
&mut self,
compressed: &[Vec<u8>],
Expand Down Expand Up @@ -273,7 +273,7 @@ impl HybridScheduler {
/// # Errors
///
/// Returns error if decompression fails.
#[cfg(feature = "cuda")]
#[cfg(feature = "__cuda-linked")]
pub fn decompress_parallel(
&mut self,
compressed: &[Vec<u8>],
Expand Down Expand Up @@ -338,7 +338,7 @@ impl HybridScheduler {
/// # Errors
///
/// Returns error if decompression fails or buffer is too small.
#[cfg(feature = "cuda")]
#[cfg(feature = "__cuda-linked")]
pub fn decompress_parallel_into(
&mut self,
compressed: &[Vec<u8>],
Expand Down Expand Up @@ -500,7 +500,7 @@ mod tests {
}

#[test]
#[cfg(feature = "cuda")]
#[cfg(feature = "__cuda-linked")]
#[ignore] // SIGSEGV during large batch allocation - needs investigation
fn test_hybrid_scheduler_creation() {
let config = HybridConfig::default();
Expand All @@ -513,7 +513,7 @@ mod tests {
}

#[test]
#[cfg(feature = "cuda")]
#[cfg(feature = "__cuda-linked")]
#[ignore] // Depends on test_hybrid_scheduler_creation
fn test_hybrid_scheduler_compress() {
if !crate::gpu::gpu_available() {
Expand All @@ -540,7 +540,7 @@ mod tests {
}

#[test]
#[cfg(feature = "cuda")]
#[cfg(feature = "__cuda-linked")]
#[ignore] // Depends on test_hybrid_scheduler_creation
fn test_hybrid_scheduler_decompress_gpu() {
if !crate::gpu::gpu_available() {
Expand Down Expand Up @@ -585,7 +585,7 @@ mod tests {
}

#[test]
#[cfg(feature = "cuda")]
#[cfg(feature = "__cuda-linked")]
fn test_hybrid_scheduler_g119_estimate() {
let stats = HybridStats {
pages_decompressed_gpu: 1_000_000, // 1M pages = 4GB
Expand Down
Loading