From mboxrd@z Thu Jan 1 00:00:00 1970 From: Claude Code Review Bot To: dri-devel-reviews@example.com Subject: Claude review: accel/qda: Add initial FastRPC attach and release support Date: Tue, 24 Feb 2026 18:20:13 +1000 Message-ID: In-Reply-To: <20260224-qda-firstpost-v1-13-fe46a9c1a046@oss.qualcomm.com> References: <20260224-qda-firstpost-v1-0-fe46a9c1a046@oss.qualcomm.com> <20260224-qda-firstpost-v1-13-fe46a9c1a046@oss.qualcomm.com> X-Mailer: Claude Code Patch Reviewer Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Patch Review This is the largest patch (548+ lines). Adds `fastrpc_invoke_context`, argu= ment marshalling, and INIT_ATTACH/INIT_RELEASE. - **Should be split**: Dmitry suggested "the FastRPC backing code should be= a separate commit, INIT_ATTACH another, separate commit." - **kref in fastrpc_invoke_context**: Dmitry asked "why does it embed kref?= " if contexts are tracked in an xarray. The lifecycle management needs clea= rer justification. - **Too many allocations per RPC call**: Dmitry noted "Generally it feels l= ike there are too many allocations and frees for a single RPC call. Can all= these buffers be embedded into the context instead?" This is a performance= concern for a hot path. - **Unclear data definitions**: Multiple struct members lack documentation = about what address space they refer to =E2=80=94 is `ptr` a CPU virtual add= ress, a DSP address, or an IOMMU address? - **"olaps" / "wist" / "direct buffer"**: Abbreviations and concepts are us= ed without definition. --- Generated by Claude Code Patch Reviewer