Make GraphicsResource inherit from Buffer

[rparolin]

Summary

• Refactors GraphicsResource to inherit from Buffer instead of wrapping a separate Buffer object, eliminating the _MappedBufferContext intermediary class
• map() now populates the GraphicsResource itself with device pointer/size (since it IS-A Buffer) and returns self
• Makes stream a required parameter for map() and unmap() (removes implicit default stream usage)
• Adds stream parameter to from_gl_buffer() for a convenient register-and-map-in-one-step pattern
• Renames handle property to resource_handle to avoid conflicting with Buffer.handle (which now exposes the mapped device pointer)

Motivation

Review feedback requested that GraphicsResource be a Buffer directly rather than producing a separate Buffer via map(). This simplifies the API surface: a mapped GraphicsResource can be passed directly anywhere a Buffer is accepted (e.g., StridedMemoryView.from_buffer()), without needing a wrapper context manager.

Key changes

• GraphicsResource(Buffer) — inherits from Buffer; handle and size are valid while mapped
• Removed _MappedBufferContext — no longer needed since GraphicsResource itself is the buffer
• __enter__/__exit__ — moved onto GraphicsResource directly; __exit__ auto-unmaps using the stream from map()
• stream required — map(stream=) and unmap(stream=) no longer default to stream 0
• from_gl_buffer(stream=) — optional stream param to register + map in one call, enabling with GraphicsResource.from_gl_buffer(vbo, stream=s) as buf: pattern
• close(stream=None) — accepts stream kwarg for compatibility with Buffer.close()
• resource_handle — renamed from handle to avoid shadowing Buffer.handle

🤖 Generated with Claude Code

[copy-pr-bot]

Auto-sync is disabled for draft pull requests in this repository. Workflows must be run manually.

Contributors can view more details about this message here.

[copy-pr-bot]

Auto-sync is disabled for ready for review pull requests in this repository. Workflows must be run manually.

Contributors can view more details about this message here.

[rparolin]

/ok to test

[github-actions]

Doc Preview CI
🚀 View preview at https://nvidia.github.io/cuda-python/pr-preview/pr-1701/
https://nvidia.github.io/cuda-python/pr-preview/pr-1701/cuda-core/
https://nvidia.github.io/cuda-python/pr-preview/pr-1701/cuda-bindings/
https://nvidia.github.io/cuda-python/pr-preview/pr-1701/cuda-pathfinder/
Preview will be ready when the GitHub Pages deployment is complete.