feat(mars): add WITH_MARS platform and htcc wrapper

[Ceng23333]

Summary

• Enable WITH_MARS / HPCC auto-detect in root CMakeLists.txt, set htcc via scripts/htcc_wrapper.sh, and link HPCC runtime/DNN/BLAS/OpenMP.
• Wire Mars into src/CMakeLists.txt source glob + DEVICE_LIST (no Mars op tree in this PR).
• Register Device::Type::kMars in src/device.h, "mars" in src/pybind11_utils.h, and Mars entries in scripts/generate_wrappers.py / examples/runtime_api.h.
• Stacked follow-ups: kernels (feat(mars): add native Mars operator specializations #942) and torch/smoke (fix(mars): enable torch bindings and CONTRIBUTING smoke #943). This PR intentionally excludes src/native/cuda/mars/ and torch host-compile plumbing.

Motivation

Mars/HPCC needs a dedicated InfiniOps backend (MetaX X203). Landing platform/toolchain first keeps the kernel PR reviewable in isolation and matches the MetaX split of CMake vs native/cuda/<vendor>/.

Stacked replacement for closed #939.

Type of Change

• feat — new feature / new operator / new platform
• fix — bug fix
• perf — performance improvement (no behavioral change)
• refactor — code restructuring without behavior change
• test — adding or fixing tests only
• docs — documentation only
• build / ci — build system or CI configuration
• chore — tooling, formatting, or other non-code changes
• Breaking change (requires a ! in the Conventional Commits prefix or a BREAKING CHANGE: footer)

Platforms Affected

• CPU (WITH_CPU)
• NVIDIA (WITH_NVIDIA)
• Iluvatar (WITH_ILUVATAR)
• MetaX (WITH_METAX)
• Cambricon (WITH_CAMBRICON)
• Moore (WITH_MOORE)
• Ascend (WITH_ASCEND)
• PyTorch C++ bindings (WITH_TORCH)
• Build system / CMake / CI
• Python bindings / user-facing API

Note: Mars / HPCC (WITH_MARS) is the affected GPU backend; it is not yet a checkbox in this template. This PR only enables the platform/toolchain (no Mars operator specializations).

Smoke Test Result

Host: MetaX X203 / HPCC. Branch feat/mars-platform-htcc.

# Library-only smoke configure/build (no Mars ops tree yet; shared native/cuda/*.cc may compile)
cmake -S . -B build-mars-smoke \
  -DWITH_CPU=ON -DWITH_MARS=ON \
  -DAUTO_DETECT_DEVICES=OFF -DAUTO_DETECT_BACKENDS=OFF \
  -DGENERATE_PYTHON_BINDINGS=OFF \
  -DINFINI_OPS_SMOKE_BUILD=ON \
  -DINFINI_RT_ROOT=<mars-rt-prefix> \
  -DCMAKE_BUILD_TYPE=Release
# Configuring / Generating done

cmake --build build-mars-smoke --target infiniops --parallel $(nproc)
# Built target infiniops

Full python -m pytest tests -m smoke -q is claimed on stack tip #943 (requires kernels + torch enablement).

Test Results on Supported Platforms

Platform	Affected	Build / Smoke Result	Full Result / Notes
NVIDIA	no	N/A - not affected
Iluvatar	no	N/A - not affected
MetaX	no	N/A - not affected
Cambricon	no	N/A - not affected
Moore	no	N/A - not affected
Ascend	no	N/A - not affected
Mars/HPCC (WITH_MARS)	yes	library smoke build (infiniops, INFINI_OPS_SMOKE_BUILD=ON)	No Mars Operator<…, kMars> yet — see #942. Pytest smoke on #943.

Full `pytest` output (optional)

N/A — pytest smoke is on stack tip #943.

Benchmark / Performance Impact

N/A

Notes for Reviewers

• Stack (merge order): this PR (feat(mars): add WITH_MARS platform and htcc wrapper #941) → feat(mars): add native Mars operator specializations #942 (kernels) → fix(mars): enable torch bindings and CONTRIBUTING smoke #943 (torch/smoke).
• After this merges, retarget feat(mars): add native Mars operator specializations #942 → master.
• htcc_wrapper.sh places .cu symlinks beside sources (not /tmp) so same-directory includes resolve.

[Ceng23333]

Closing as already landed.

This platform/htcc layer shipped as part of the stack tip merge:

• fix(mars): enable torch bindings and CONTRIBUTING smoke #943 (merged into master as 997ac66, including c60cca4 / c3feb67)

Do not merge this PR — it would be redundant with master.

Related: InfiniRT HPCC Mars dtype fix also landed — InfiniTensor/InfiniRT#46