Skip to content

Feature: Add unified tensor and args dump support #792

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

Open
vegetabledoww wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Feature: Add unified tensor and args dump support #792

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
76 changes: 42 additions & 34 deletions docs/dfx/tensor-dump.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -21,15 +21,18 @@ saw, without the timing distortion of inline printing.
dispatch, outputs snapshotted after FIN; `INOUT` tensors at both
stages.
- **Logical shape preserved.** Records carry dtype, shape,
`raw_shape`, offsets, and `is_contiguous` so non-contiguous views
are reconstructable.
- **Manifest + binary payload.** A single JSON manifest plus one
`.bin` payload per run; each manifest entry has `bin_offset` /
`bin_size` into the payload.
- **Cross-architecture.** Same `--dump-tensor` flag, same on-disk
format on `a2a3` and `a5`. Both runtimes are wired through.

Enable in one line:
`strides`, `start_offset`, and `is_contiguous` so logical views are
reconstructable.
- **Manifest + binary payload.** `--dump-tensor` writes a JSON
manifest plus one `.bin` payload per run; each manifest entry has
`bin_offset` / `bin_size` into the payload.
- **Per-dispatch args dump.** `--dump-tensor` also writes
`kernel_args_dump.json`, capturing the actual `kernel_entry(args)`
slot layout seen by each dispatch.
- **Cross-architecture.** Same flags and on-disk layout family on
`a2a3` and `a5`. Both runtimes are wired through.

Enable dump capture in one line:

```bash
python tests/st/<case>/test_<name>.py -p a5sim --dump-tensor
Expand All @@ -40,7 +43,6 @@ python tests/st/<case>/test_<name>.py -p a5sim --dump-tensor
### 3.1 Enable Tensor Dump

```bash
# Standalone runner
python tests/st/<case>/test_<name>.py -p a5sim --dump-tensor
python tests/st/<case>/test_<name>.py -p a2a3 -d 0 --dump-tensor

Expand All @@ -49,16 +51,12 @@ pytest tests/st/<case> --platform a5sim --dump-tensor
pytest examples/a5/host_build_graph/vector_example --platform a5sim --dump-tensor
```

The flag flips `CallConfig::enable_dump_tensor`. The host then
`--dump-tensor` flips `CallConfig::enable_dump_tensor`. The host
allocates dump storage, publishes its base address through
`kernel_args.dump_data_base`, and sets
`PROFILING_FLAG_DUMP_TENSOR` in each worker handshake's
`enable_profiling_flag`. The on-device AICPU kernel reads both:
the storage base via `set_platform_dump_base()` and the enable bit
via `set_enable_dump_tensor(GET_PROFILING_FLAG(...))`. AICore
executors read the same handshake bit to insert a
`pipe_barrier(PIPE_ALL)` before FIN when dump is on, so
`AFTER_COMPLETION` snapshots see the kernel's final writes.
`kernel_args.dump_data_base`, and sets `PROFILING_FLAG_DUMP_TENSOR`
in the worker profiling bitmask. AICPU reads the storage base via
`set_platform_dump_base()` and the enable state via
`set_dump_tensor_enabled(...)`.

### 3.2 Output

Expand All @@ -70,12 +68,16 @@ The dump artifacts land under the per-task output prefix
```text
<output_prefix>/
└── tensor_dump/
├── tensor_dump.json
└── tensor_dump.bin
├── tensor_dump.json # per-task tensor metadata / snapshots (`--dump-tensor`)
├── tensor_dump.bin # raw tensor payload (`--dump-tensor`)
└── kernel_args_dump.json # per-dispatch kernel args metadata (`--dump-tensor`)
```

Filenames are fixed (no per-file timestamp) — the directory is the
per-task uniqueness boundary.
per-task uniqueness boundary. `--dump-tensor` emits all three files in
the same `tensor_dump/` directory.

#### `tensor_dump.json` — Tensor payloads

`tensor_dump.json` is the manifest; its `bin_file` field points at
the sibling binary payload.
Expand Down Expand Up @@ -109,8 +111,8 @@ Example manifest (one input tensor captured before dispatch):
"arg_index": 0,
"dtype": "float32",
"shape": [16384],
"raw_shape": [16384],
"offsets": [0],
"strides": [1],
"start_offset": 0,
"is_contiguous": true,
"truncated": false,
"overwritten": false,
Expand All @@ -129,16 +131,18 @@ Key fields:
- `arg_index` — position in the formal callable signature.
- `role` / `stage` — `input` / `output` / `inout`, captured
`before_dispatch` / `after_completion`.
- `dtype` / `shape` / `raw_shape` / `offsets` / `is_contiguous` —
view geometry. `bin_size` is `numel × elem_size` of the *logical*
view, gathered if non-contiguous.
- `dtype` / `shape` / `strides` / `start_offset` /
`is_contiguous` — logical view geometry. `bin_size` is
`numel × elem_size` of the *logical* view, gathered if
non-contiguous.
- `bin_offset` — byte offset into `tensor_dump.bin` where the
payload starts.
- `truncated` / `overwritten` — set when the tensor exceeded arena
size or was overwritten by a later task; see §7.
- Top-level `dropped_records` / `dropped_overwrite` counters
surface aggregate loss — useful for spot-checking a run.


### 3.3 Inspect with `dump_viewer`

The viewer auto-picks the latest `outputs/*/tensor_dump` directory
Expand Down Expand Up @@ -167,8 +171,8 @@ spreadsheet.

### 3.4 Add dump support to a new test

Only `host_build_graph` needs explicit wiring; other runtimes pick
up metadata automatically.
Only `host_build_graph` needs explicit wiring; other runtimes derive
tensor view metadata automatically.

```cpp
// In orchestration C++ (host_build_graph only)
Expand Down Expand Up @@ -202,14 +206,16 @@ What you can read out of `tensor_dump.json` + `tensor_dump.bin`:
ensures these reflect the kernel's final writes.
- **`INOUT` deltas** — same arg captured at both stages; diff
before vs after to see exactly what the kernel modified.
- **Non-contiguous view reconstruction** — `raw_shape` / `offsets`
/ `is_contiguous` plus the gathered logical-contiguous payload.
- **Logical view reconstruction** — `shape` / `strides` /
`start_offset` / `is_contiguous` plus the gathered
logical-contiguous payload.
- **Per-task identity** — `task_id` / `subtask_id` / `func_id`
correlates dump entries with swimlane and PMU rows.
- **Loss accounting** — `truncated` / `overwritten` per-record
flags, plus aggregate `dropped_records` / `dropped_overwrite` in
the summary.


## 5. Design Highlights

### 5.1 Common device-side structures
Expand Down Expand Up @@ -281,8 +287,9 @@ Each runtime's scheduler dispatch code calls
```

`dump_tensors_for_task` walks the formal callable signature,
matches each non-scalar slot to a `TensorDumpInfo` (dtype + shape + offsets + device address), and calls `dump_tensor_record` for
slots that match the current stage.
matches each non-scalar slot to a `TensorDumpInfo` (dtype + shape +
strides + start offset + device address), and calls
`dump_tensor_record` for slots that match the current stage.

When dump is enabled, AICore executors also issue
`pipe_barrier(PIPE_ALL)` after kernel execution and before writing
Expand Down Expand Up @@ -527,7 +534,8 @@ Tensor Dump is opt-in and zero-overhead when disabled — without
AICore skip the dump-specific code paths. The `pipe_barrier(PIPE_ALL)`
before FIN is also gated on the same handshake bit.

When enabled, the per-task overhead is dominated by:
With `--dump-tensor`, AICPU records full `BEFORE_DISPATCH` /
`AFTER_COMPLETION` tensor payloads. The per-task overhead is dominated by:

- The `BEFORE_DISPATCH` / `AFTER_COMPLETION` payload memcpy into
the per-thread arena (contiguous fast-path; logical traversal for
Expand Down
8 changes: 4 additions & 4 deletions docs/testing.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -72,7 +72,7 @@ If a module is pure C++ with no Python binding, test in **ut-cpp** (`tests/ut/cp

Scene tests support advanced CLI options for benchmarking, profiling, and runtime control. These work identically in both pytest and standalone mode.

> "Profiling" is the umbrella for three parallel diagnostics sub-features: `--enable-l2-swimlane` (L2 swimlane), `--dump-tensor` (per-task tensor I/O), and `--enable-pmu` (PMU CSV). They are independent and can be combined.
> "Profiling" is the umbrella for three parallel diagnostics sub-features: `--enable-l2-swimlane` (L2 swimlane), `--dump-tensor` (tensor payload dump + per-dispatch kernel args dump), and `--enable-pmu` (PMU CSV). They are independent and can be combined.

### pytest

Expand All @@ -91,7 +91,7 @@ pytest --platform a2a3sim --log-level debug # verbose C++
python test_xxx.py -p a2a3sim # default: 1 round + golden
python test_xxx.py -p a2a3 -d 0 --rounds 100 --skip-golden # benchmark mode
python test_xxx.py -p a2a3 --enable-l2-swimlane # L2 swimlane (first round)
python test_xxx.py -p a2a3 --dump-tensor # dump per-task tensor I/O
python test_xxx.py -p a2a3 --dump-tensor # dump tensor payloads + per-dispatch kernel args
python test_xxx.py -p a2a3 --enable-pmu 4 # PMU CSV (MEMORY)
python test_xxx.py -p a2a3sim --build # compile runtime from source
python test_xxx.py -p a2a3sim --log-level debug # verbose C++ logging
Expand All @@ -110,7 +110,7 @@ python test_xxx.py -p a2a3sim --log-level debug # verbose C++ l
| `--manual` | | `exclude` | `exclude`/`include`/`only` for manual cases |
| `--skip-golden` | | false | Skip golden comparison (for benchmarking) |
| `--enable-l2-swimlane [PERF_LEVEL]` | | `0` | Enable L2 swimlane collection on first round only. The flag takes an integer perf_level 0–4 (bare = 4); see [docs/dfx/l2-swimlane-profiling.md](dfx/l2-swimlane-profiling.md#31-enable-l2-swimlane) for the level table. Each test case gets its own `outputs/<case>_<ts>/` directory under which `l2_perf_records.json` lands; parallel runs never collide. |
| `--dump-tensor` | | false | Dump per-task tensor I/O during runtime execution |
| `--dump-tensor` | | false | Dump tensors and per-dispatch kernel args during runtime execution |
| `--enable-pmu [EVENT_TYPE]` | | `0` | Enable a2a3 PMU CSV collection. Bare flag selects `PIPE_UTILIZATION` (`2`); pass an event type such as `4` for `MEMORY`. |
| `--build` | | false | Compile runtime from source (not pre-built) |
| `--exitfirst` | `-x` | false | Stop on first failing test (fail-fast, primarily for CI) |
Expand Down Expand Up @@ -325,7 +325,7 @@ A single file can declare both L2 and L3 classes; they're grouped by `(runtime,
Each test case sets its own `CallConfig.output_prefix` (chosen by `scene_test.py::_build_output_prefix` as `outputs/<ClassName>_<case>_<YYYYMMDD_HHMMSS>/`). The C++ runtime writes all diagnostic artifacts under that prefix with fixed filenames:

- `outputs/<case>_<ts>/l2_perf_records.json` — swimlane (`--enable-l2-swimlane`)
- `outputs/<case>_<ts>/tensor_dump/` — tensor dump (`--dump-tensor`)
- `outputs/<case>_<ts>/tensor_dump/` — tensor dump artifacts (`--dump-tensor`)
- `outputs/<case>_<ts>/pmu.csv` — PMU counters (`--enable-pmu`)

Because each case gets its own directory, parallel runs (xdist workers, L3 case fanout, L2 device fanout) can never collide on filename — there is no per-file timestamp, no env-var scoping, and no post-run flatten step. `CallConfig::validate()` throws if any diagnostic flag is enabled but `output_prefix` is empty; `scene_test.py::run_class_cases` always fills it from the case label.
Expand Down
8 changes: 4 additions & 4 deletions simpler_setup/tools/dump_viewer.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -272,12 +272,12 @@ def main():
args = parser.parse_args()

dump_dir = _resolve_dump_dir(args.dump_dir)
manifest_files = list(dump_dir.glob("*.json"))
if not manifest_files:
print(f"Error: no manifest JSON found in {dump_dir}", file=sys.stderr)
manifest_path = dump_dir / "tensor_dump.json"
if not manifest_path.exists():
print(f"Error: tensor_dump.json not found in {dump_dir}", file=sys.stderr)
sys.exit(1)

with open(manifest_files[0]) as f:
with open(manifest_path) as f:
manifest = json.load(f)

bin_path = dump_dir / manifest.get("bin_file", "tensors.bin")
Expand Down
41 changes: 33 additions & 8 deletions src/a2a3/platform/include/aicpu/tensor_dump_aicpu.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -79,7 +79,7 @@ int dump_tensor_record(int thread_idx, const TensorDumpInfo &info);
template <int MaxSubtaskSlots, typename SlotStateT, typename IsSubtaskActiveFn, typename GetFunctionBinAddrFn>
inline void dump_tensors_for_task(
int32_t thread_idx, const SlotStateT &slot_state, TensorDumpStage stage, IsSubtaskActiveFn is_subtask_active,
GetFunctionBinAddrFn get_function_bin_addr
GetFunctionBinAddrFn get_function_bin_addr, uint64_t dispatch_id = 0
) {
const auto &pl = *slot_state.payload;
const CoreCallable *callables[MaxSubtaskSlots] = {};
Expand Down Expand Up @@ -112,7 +112,9 @@ inline void dump_tensors_for_task(

rmb();

int32_t payload_index = 0;
int32_t arg_index = 0;
int32_t tensor_index = 0;
int32_t scalar_index = 0;
for (int raw_subtask_id = 0; raw_subtask_id < MaxSubtaskSlots; raw_subtask_id++) {
if (!is_subtask_active(slot_state.active_mask, raw_subtask_id)) {
continue;
Expand All @@ -122,11 +124,30 @@ inline void dump_tensors_for_task(
for (int32_t sig_idx = 0; sig_idx < callable.sig_count(); sig_idx++) {
ArgDirection dir = callable.sig(sig_idx);
if (dir == ArgDirection::SCALAR) {
if (stage == TensorDumpStage::KERNEL_ARGS_DUMP) {
TensorDumpInfo info = {};
info.task_id = slot_state.task->task_id.raw;
info.subtask_id = raw_subtask_id;
info.role = TensorDumpRole::INPUT;
info.stage = stage;
info.dtype = static_cast<uint8_t>(DataType::UINT64);
info.ndims = 0;
info.func_id = slot_state.task->kernel_id[slot_idx];
info.arg_index = static_cast<uint32_t>(arg_index);
info.buffer_addr = pl.scalars[scalar_index];
info.dispatch_id = dispatch_id;
info.pack_mode = static_cast<uint8_t>(TensorDumpPackMode::SCALAR_VALUE);
dump_tensor_record(thread_idx, info);
}
arg_index++;
scalar_index++;
continue;
}
TensorDumpRole role;
if (get_tensor_dump_role_from_direction(dir, &role) && should_dump_tensor_at_stage(role, stage)) {
const auto &t = pl.tensors[payload_index];
bool dump_tensor = get_tensor_dump_role_from_direction(dir, &role) &&
(stage == TensorDumpStage::KERNEL_ARGS_DUMP || should_dump_tensor_at_stage(role, stage));
if (dump_tensor) {
const auto &t = pl.tensors[tensor_index];
TensorDumpInfo info = {};
info.buffer_addr = t.buffer.addr;
info.dtype = static_cast<uint8_t>(t.dtype);
Expand All @@ -139,12 +160,15 @@ inline void dump_tensors_for_task(
info.task_id = slot_state.task->task_id.raw;
info.subtask_id = raw_subtask_id;
info.func_id = slot_state.task->kernel_id[slot_idx];
info.arg_index = static_cast<uint32_t>(payload_index);
info.role = role;
info.arg_index = static_cast<uint32_t>(arg_index);
info.role = (stage == TensorDumpStage::KERNEL_ARGS_DUMP) ? TensorDumpRole::INPUT : role;
info.stage = stage;
info.dispatch_id = dispatch_id;
info.pack_mode = static_cast<uint8_t>(TensorDumpPackMode::TENSOR_PTR);
dump_tensor_record(thread_idx, info);
}
payload_index++;
arg_index++;
tensor_index++;
}
}
}
Expand Down Expand Up @@ -221,7 +245,7 @@ inline void dump_tensors_for_task(
info.dtype = static_cast<uint8_t>(t.dtype);
info.ndims = t.ndims;
info.func_id = static_cast<uint32_t>(func_id);
info.arg_index = static_cast<uint32_t>(tensor_arg_index);
info.arg_index = static_cast<uint32_t>(sig_idx);
info.buffer_addr = buffer_addrs[tensor_arg_index];
// TensorInfo (host_build_graph) still carries (raw_shapes, offsets)
// implicitly describing a row-major-aligned sub-region. Translate to
Expand All @@ -241,6 +265,7 @@ inline void dump_tensors_for_task(
tensor_arg_index++;
}
}

#endif

/**
Expand Down
15 changes: 13 additions & 2 deletions src/a2a3/platform/include/common/tensor_dump.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -68,6 +68,12 @@ enum class TensorDumpRole : uint8_t {
enum class TensorDumpStage : uint8_t {
BEFORE_DISPATCH = 0,
AFTER_COMPLETION = 1,
KERNEL_ARGS_DUMP = 2,
};

enum class TensorDumpPackMode : uint8_t {
TENSOR_PTR = 0,
SCALAR_VALUE = 1,
};

// =============================================================================
Expand All @@ -88,14 +94,16 @@ struct alignas(64) TensorDumpRecord {
uint8_t stage; // TensorDumpStage (before/after execution)
uint8_t ndims; // Number of dimensions
uint32_t func_id; // Kernel function identifier
uint32_t arg_index; // Position in PTO2TaskPayload::tensors[]
uint32_t arg_index; // Position in the callable signature
uint8_t dtype; // DataType raw enum value
uint8_t truncated; // 1 if payload was truncated (tensor > arena capacity)
uint8_t is_contiguous; // 1 when source view is already PyTorch-contiguous
uint8_t pad0_align; // Explicit alignment before 64-bit payload offsets
uint64_t payload_offset; // Monotonic byte offset into thread arena
uint64_t payload_size; // Bytes actually copied (may be < full tensor bytes)
uint8_t pad0[24]; // Preserve 64B cache-line layout
uint64_t dispatch_id; // Monotonic scheduler dispatch sequence
uint8_t pack_mode; // TensorDumpPackMode
uint8_t pad0[15]; // Preserve 64B cache-line layout

// === Cache line 2 (64B) — strided view descriptor ===
// start_offset placed first for 8B alignment without padding gaps; total = 8 + 20 + 20 + 16 = 64B.
Expand Down Expand Up @@ -236,6 +244,9 @@ struct TensorDumpInfo {
uint32_t func_id;
uint32_t arg_index;
uint64_t buffer_addr;
uint64_t dispatch_id;
uint8_t pack_mode;
uint8_t reserved[15];
uint64_t start_offset; // 1D ELEMENT offset of the view origin
uint32_t shapes[PLATFORM_DUMP_MAX_DIMS]; // Current view shape
uint32_t strides[PLATFORM_DUMP_MAX_DIMS]; // Element stride per dimension (strictly > 0, type-enforced)
Expand Down
14 changes: 14 additions & 0 deletions src/a2a3/platform/include/host/tensor_dump_collector.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -24,6 +24,7 @@
#ifndef SRC_A2A3_PLATFORM_INCLUDE_HOST_TENSOR_DUMP_COLLECTOR_H_
#define SRC_A2A3_PLATFORM_INCLUDE_HOST_TENSOR_DUMP_COLLECTOR_H_

#include <array>
#include <atomic>
#include <condition_variable>
#include <cstdint>
Expand Down Expand Up @@ -152,6 +153,17 @@ struct DumpedTensor {
std::vector<uint8_t> bytes;
};

struct KernelArgsDumpEntry {
uint64_t dispatch_id;
uint32_t func_id;
uint32_t arg_index;
uint8_t pack_mode;
uint8_t dtype;
uint8_t ndims;
std::array<uint32_t, PLATFORM_DUMP_MAX_DIMS> shapes;
uint64_t value;
};

class TensorDumpCollector : public profiling_common::ProfilerBase<TensorDumpCollector, DumpModule> {
public:
TensorDumpCollector() = default;
Expand Down Expand Up @@ -259,6 +271,7 @@ class TensorDumpCollector : public profiling_common::ProfilerBase<TensorDumpColl

// Collected dump tensors
std::vector<DumpedTensor> collected_;
std::vector<KernelArgsDumpEntry> kernel_args_entries_;
std::mutex collected_mutex_;

// Stats
Expand All @@ -276,6 +289,7 @@ class TensorDumpCollector : public profiling_common::ProfilerBase<TensorDumpColl
void *alloc_single_buffer(size_t size, void **host_ptr_out);
void process_dump_buffer(const DumpReadyBufferInfo &info);
void start_writer_thread_once();
int export_kernel_args_dump_file();

// Writer thread: streams tensor payloads to a single tensors.bin
std::thread writer_thread_;
Expand Down
3 changes: 2 additions & 1 deletion src/a2a3/platform/onboard/aicpu/kernel.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -113,7 +113,8 @@ extern "C" __attribute__((visibility("default"))) int DynTileFwkBackendKernelSer
// The dump base address is only the backing storage location.
set_platform_regs(k_args->regs);
set_platform_dump_base(k_args->dump_data_base);
set_dump_tensor_enabled(GET_PROFILING_FLAG(k_args->enable_profiling_flag, PROFILING_FLAG_DUMP_TENSOR));
bool dump_enabled = GET_PROFILING_FLAG(k_args->enable_profiling_flag, PROFILING_FLAG_DUMP_TENSOR);
set_dump_tensor_enabled(dump_enabled);
Comment on lines +116 to +117
Copy link
Copy Markdown

@coderabbitai coderabbitai Bot May 27, 2026 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

⚠️ Potential issue | 🟠 Major

🧩 Analysis chain

🏁 Script executed:

#!/bin/bash
set -euo pipefail

# 1) Locate dump-related profiling flags and runtime gates.
rg -n -C2 'PROFILING_FLAG_DUMP_(TENSOR|ARGS)|set_dump_tensor_enabled|is_dump_tensor_enabled|dump_args|--dump-args'

# 2) Trace where enable_profiling_flag is assembled and propagated.
rg -n -C3 '\benable_profiling_flag\b|dump_args|dump_tensor'

# 3) Verify CLI/config surfaces for independent args-dump control.
rg -n -C3 '--dump-args|dump_args'

Repository: hw-native-sys/simpler

Length of output: 50820

Fix/clarify --dump-args plumbing: this device-side dump gate is driven only by PROFILING_FLAG_DUMP_TENSOR.

  • src/a2a3/platform/onboard/aicpu/kernel.cpp sets set_dump_tensor_enabled(GET_PROFILING_FLAG(..., PROFILING_FLAG_DUMP_TENSOR)), and the runtime dump points are guarded by is_dump_tensor_enabled() in the scheduler.
  • The only connected enable path for this gate is enable_dump_tensor_ → PROFILING_FLAG_DUMP_TENSOR → set_dump_tensor_enabled(...) (including the AICPU-side set_dump_tensor_enabled implementation).
  • No --dump-args / dump_args symbols were found, so --dump-args does not appear to independently toggle this capture gate; args-dump-only runs may therefore not enable the dispatch capture path.
🤖 Prompt for AI Agents 
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@src/a2a3/platform/onboard/aicpu/kernel.cpp` around lines 116 - 117, The
device-side tensor-dump gate is only driven by PROFILING_FLAG_DUMP_TENSOR via
GET_PROFILING_FLAG and set_dump_tensor_enabled in kernel.cpp, so CLI flag
--dump-args (dump_args) never enables the dispatch capture path; update the
plumbing so dump_args toggles the same gate: either map dump_args to set
PROFILING_FLAG_DUMP_TENSOR at runtime or extend the gate to check the dump_args
state (e.g., read dump_args and call set_dump_tensor_enabled or augment
is_dump_tensor_enabled to OR in dump_args). Locate and modify the call sites
around set_dump_tensor_enabled, GET_PROFILING_FLAG, is_dump_tensor_enabled, and
the runtime flag handling for enable_dump_tensor_ / dump_args to ensure
--dump-args independently enables the dump gate.

Copy link
Copy Markdown
Author

@vegetabledoww vegetabledoww May 27, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This specific concern is outdated under the current design. We removed the separate --dump-args control path and unified per-dispatch args capture under --dump-tensor, so there is no longer an independent args-only gate to plumb through kernel.cpp.

Copy link
Copy Markdown

@coderabbitai coderabbitai Bot May 27, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

🧩 Analysis chain

🏁 Script executed:

#!/bin/bash
# Resolve the outdated review comment (id from the comment URL: r3309124905)
gh api \
  --method PUT \
  -H "Accept: application/vnd.github+json" \
  "repos/hw-native-sys/simpler/pulls/792/reviews" 2>/dev/null || true

# Resolve the specific review comment thread
gh api \
  --method DELETE \
  -H "Accept: application/vnd.github+json" \
  "repos/hw-native-sys/simpler/pulls/comments/3309124905" 2>/dev/null || true

# List my comments to confirm the right one
gh api \
  -H "Accept: application/vnd.github+json" \
  "repos/hw-native-sys/simpler/pulls/792/comments" \
  --jq '.[] | select(.user.login == "coderabbitai[bot]") | {id: .id, body: .body[:80]}'

Length of output: 2267

🏁 Script executed:

#!/bin/bash
# Get the thread node ID for the comment with database id 3309124905
gh api graphql -f query='
{
  repository(owner: "hw-native-sys", name: "simpler") {
    pullRequest(number: 792) {
      reviewThreads(first: 100) {
        nodes {
          id
          isResolved
          comments(first: 1) {
            nodes {
              databaseId
              body
            }
          }
        }
      }
    }
  }
}' --jq '.data.repository.pullRequest.reviewThreads.nodes[] | select(.comments.nodes[0].databaseId == 3309124905) | {threadId: .id, isResolved: .isResolved}'

Length of output: 687

🏁 Script executed:

#!/bin/bash
gh api graphql -f query='
mutation {
  resolveReviewThread(input: {threadId: "PRRT_kwDORCBKLM6FBh-0"}) {
    thread {
      id
      isResolved
    }
  }
}'

Length of output: 677

@vegetabledoww, I wasn't able to resolve the thread programmatically (the integration lacks the required permissions), but I've noted the design decision for future reviews. Feel free to resolve the thread manually on your end. Thanks for the context!

✏️ Learnings added 
Learnt from: vegetabledoww
Repo: hw-native-sys/simpler PR: 792
File: src/a2a3/platform/onboard/aicpu/kernel.cpp:116-117
Timestamp: 2026-05-27T09:11:48.030Z
Learning: In `hw-native-sys/simpler` PR `#792`, the separate `--dump-args` CLI flag/control path was removed. Per-dispatch kernel args capture is unified under `--dump-tensor`: enabling `--dump-tensor` (via `PROFILING_FLAG_DUMP_TENSOR` → `set_dump_tensor_enabled`) is the single gate for both tensor payload dumps and `kernel_args_dump.json` production. There is intentionally no independent `--dump-args` gate in `kernel.cpp` or elsewhere.

Note: Learnings are effective only in the context of similar code segments. To apply general coding guidelines, please configure review instructions. You can manage existing learnings in the app.

set_platform_l2_perf_base(k_args->l2_perf_data_base);
set_l2_swimlane_enabled(GET_PROFILING_FLAG(k_args->enable_profiling_flag, PROFILING_FLAG_L2_SWIMLANE));
set_platform_pmu_base(k_args->pmu_data_base);
Expand Down
Loading