[atharva-potdar]

Rationale for this change

Currently, the slicing constructors for Buffer (src/arrow/buffer.h) and MutableBuffer (src/arrow/buffer.cc) unconditionally add offset to the parent buffer's pointer/address. When slicing a null-backed buffer where parent->data_ is nullptr or parent->mutable_address() is 0 with a non-zero offset, this performs undefined pointer arithmetic on a null pointer (producing a corrupt non-null address 0x0 + offset) and triggers clang-analyzer-core.NullPointerArithm in clang-tidy:

src/arrow/buffer.cc:167:75: warning: Addition of a null pointer and a nonzero integer value (from variable 'offset') results in undefined behavior [clang-analyzer-core.NullPointerArithm]

What changes are included in this PR?

• Guard Buffer's slice constructor in src/arrow/buffer.h against null parent->data_ before adding offset.
• Guard MutableBuffer's slice constructor in src/arrow/buffer.cc using parent->mutable_address() (preserving non-CPU device compatibility from ARROW-2447) and performing integer addition before casting when non-zero.
• Add TestBuffer.SliceNullBuffer in src/arrow/buffer_test.cc.

Are these changes tested?

Yes, added TestBuffer.SliceNullBuffer in arrow-buffer-test (which fails on unpatched main where sliced->data() evaluates to 0x8 instead of nullptr, and passes with this fix) and verified that clang-tidy no longer emits clang-analyzer-core.NullPointerArithm.

Are there any user-facing changes?

No.

Was AI used for this PR?

In accordance to the AI generation guidelines, please disclose below whether and how AI was used in this PR.

PR code written by:

• Human
• AI

PR description written by:

• Human
• AI

I asked Gemini 4 Argon to write a PR description after I wrote the test, the code, and pasted the clang-tidy and google test logs to it.

Reviewed before submission by:

• Human
• AI
• Not reviewed

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has insufficient remaining quota. Try requesting a review at a lower effort level.

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has insufficient remaining quota. Try requesting a review at a lower effort level.