Add cuda::ceil_ilog10

[fbusato]

Description

The ceil_ilog10 can be used to covert integers to chars (e.g. to_chars).
In addition, we already have ilog2, ceil_ilog2, and ilog10, so would make sense to add ceil_log10 for completeness.

It tried a couple of optimizations but the simplest version should be also the most efficient.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: 0c72a2cc-fe32-4abf-b102-73575bf5c043

📥 Commits

Reviewing files that changed from the base of the PR and between 0838c9f and 3cb3009.

📒 Files selected for processing (3)

• docs/libcudacxx/extended_api/math.rst
• docs/libcudacxx/extended_api/math/ilog.rst
• libcudacxx/include/cuda/__cmath/ilog.h

✅ Files skipped from review due to trivial changes (1)

• docs/libcudacxx/extended_api/math.rst

🚧 Files skipped from review as they are similar to previous changes (2)

• docs/libcudacxx/extended_api/math/ilog.rst
• libcudacxx/include/cuda/__cmath/ilog.h

---

Note: CodeRabbit is enabled on this repository as a convenience for maintainers
and contributors. Use your best judgment when considering its review comments and
suggestions — a suggested change may be inadequate, unnecessary, or safe to ignore.
Contributors are not expected to address every comment. Human reviews are what
ultimately matter for merging.

Added cuda::ceil_ilog10 to libcu++ alongside the existing integer log helpers (ilog2, ceil_ilog2, ilog10).

What changed

• Implemented new public API cuda::ceil_ilog10 in libcudacxx/include/cuda/__cmath/ilog.h.
• Marked ilog2, ceil_ilog2, ilog10, and the new ceil_ilog10 as [[nodiscard]], and updated their parameter types to const _Tp.
• Adjusted internal compile-time helpers for powers-of-10 (moved to _CCCL_CONSTEVAL and qualified ::cuda::std::* types).
• Extended libcudacxx/test/libcudacxx/cuda/cmath/ilog.pass.cpp with new coverage for ceil_ilog10, including boundary/neighbor checks and max-value behavior.
• Updated documentation:
  • Added ceil_ilog10 to docs/libcudacxx/extended_api/math/ilog.rst (declaration, return-value wording, performance considerations, and examples).
  • Updated the math API compatibility matrix (docs/libcudacxx/extended_api/math.rst) to include ceil_ilog10.

Notes

• The function targets integer-to-string style use cases (e.g., to_chars) where rounding up the base-10 integer logarithm is useful.

Walkthrough

important: Adds cuda::ceil_ilog10 docs, header API, and tests, and updates related ilog2, ceil_ilog2, and ilog10 declarations and helper code.

Changes

ceil_ilog10 API updates

Layer / File(s)	Summary
Docs and compatibility entries docs/libcudacxx/extended_api/math.rst, docs/libcudacxx/extended_api/math/ilog.rst	ceil_ilog10 is added to the compatibility matrix, the ilog API listing, the return/performance text, and the example assertions.
Header updates libcudacxx/include/cuda/__cmath/ilog.h	Related ilog signatures and power-of-10 helpers are updated, one include is removed, and ceil_ilog10 is added.
Test coverage libcudacxx/test/libcudacxx/cuda/cmath/ilog.pass.cpp	ceil_ilog10 gets a new templated test helper and is wired into the existing dispatcher.

---

_{Comment @coderabbitai help to get the list of available commands.}

[coderabbitai]

Actionable comments posted: 2

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: 284ef720-5bac-4fd3-8ad9-8b2137beaa96

📥 Commits

Reviewing files that changed from the base of the PR and between 0b675e0 and 0838c9f.

📒 Files selected for processing (4)

• docs/libcudacxx/extended_api/math.rst
• docs/libcudacxx/extended_api/math/ilog.rst
• libcudacxx/include/cuda/__cmath/ilog.h
• libcudacxx/test/libcudacxx/cuda/cmath/ilog.pass.cpp

[github-actions]

🥳 CI Workflow Results

🟩 Finished in 1h 21m: Pass: 100%/120 | Total: 2d 15h | Max: 1h 19m | Hits: 82%/439967

See results here.

[davebayer]

Thank you!