[rocm-libraries] ROCm/rocm-libraries#6302 (commit 8d419e8)

CK: Remove 41 commented-out dead code blocks (~200 lines)
 (#6302)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Depends on #6300

## Summary

Remove 41 commented-out code blocks across 33 files in Composable
Kernel, totaling ~200 lines.

Identified using an automated dead code scanning skill (`ck-dead-code`)
with a calibrated two-stage pipeline:
1. **Pre-filter**: Keyword-based scan found 1,338 `//`-commented blocks.
Calibrated heuristics (trained on 50-sample expert classification)
reduced to 89 high-confidence candidates — 93% noise reduction.
2. **Expert triage**: LLM expert classified each block in context as
CODE_REMOVE, CODE_KEEP, or NOT_CODE.

| Classification | Count |
|---------------|-------|
| Removed (this PR) | 41 |
| Kept (debug helpers, alt configs, reference impls) | 32 |
| Not code (false positives) | 16 |

Removed blocks include: superseded implementations, old test data,
abandoned stubs, unreachable code, and buggy dead code.
This commit is contained in:
Aviral Goel
2026-04-10 15:18:02 +00:00
committed by assistant-librarian[bot]
parent 4d0bbe5d17
commit e0dfe58d66
82 changed files with 22 additions and 2883 deletions

View File

@@ -120,10 +120,6 @@ struct BlockNormReduceSync
constexpr index_t idim_p_lane = NDimP - 1;
// const auto ps_idx = make_array<index_t>(get_warp_id(), get_lane_id());
// const auto rs_idx =
// mean_tensor.get_tile_distribution().calculate_rs_index_from_ps_index(ps_idx);
constexpr index_t thread_buf_size = MeanDistributedTensor_::get_thread_buffer_size();
static_assert(thread_buf_size == VarDistributedTensor_::get_thread_buffer_size());
@@ -360,17 +356,6 @@ struct BlockNormReduceCrossWarpSync
template <typename BlockShape>
CK_TILE_DEVICE constexpr index_t block_tile_welford_calculate_max_count(int row_size)
{
#if 0
using S = BlockShape;
index_t LastloopN = row_size % S::Block_N == 0 ? S::Block_N : row_size % S::Block_N;
constexpr index_t NThread = S::WarpPerBlock_N * S::ThreadPerWarp_N;
index_t iNLane = get_thread_id() % NThread;
index_t iN0 = LastloopN / (S::Vector_N * S::ThreadPerWarp_N);
index_t iN1 = (LastloopN % (S::Vector_N * S::ThreadPerWarp_N)) / S::Vector_N;
index_t N2 = (LastloopN % (S::Vector_N * S::ThreadPerWarp_N)) % S::Vector_N;
index_t iN3 = iNLane < iN1 ? S::Vector_N : iNLane == iN1 ? N2 : 0;
return iN0 * S::Vector_N + iN3;
#endif
using S_ = BlockShape;
constexpr index_t ThreadsPerBlock_N = S_::WarpPerBlock_N * S_::ThreadPerWarp_N;