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/amdxdna: Import AMD_PMF namespace Date: Tue, 03 Mar 2026 14:05:57 +1000 Message-ID: In-Reply-To: <20260301005028.367618-1-superm1@kernel.org> References: <20260301005028.367618-1-superm1@kernel.org> X-Mailer: Claude Code Patch Reviewer Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Overall Series Review Subject: accel/amdxdna: Import AMD_PMF namespace Author: "Mario Limonciello (AMD)" Patches: 2 Reviewed: 2026-03-03T14:05:57.418611 --- This is a single, trivially small patch that adds `MODULE_IMPORT_NS("AMD_PM= F")` to the amdxdna accelerator driver. The stated purpose is to allow the = driver to call `amd_pmf_get_npu_data()`, which is exported in the `AMD_PMF`= namespace from `drivers/platform/x86/amd/pmf/core.c`. The patch is **correct in form** =E2=80=94 adding a `MODULE_IMPORT_NS()` is= the right way to consume a symbol exported with `EXPORT_SYMBOL_NS_GPL()`. = However, there is a significant concern: **the current amdxdna driver on dr= m-next has no reference to `amd_pmf_get_npu_data()` or any `amd_pmf` API at= all**. This means the namespace import is being added ahead of the code th= at actually uses it, or depends on another patch (or series) that hasn't la= nded yet. This is fine if this patch is intended to be applied on top of another seri= es that adds the actual `amd_pmf_get_npu_data()` call, but the commit messa= ge should state this dependency explicitly (e.g., referencing the other ser= ies or commit). Without that context, the patch appears to add a namespace = import for a symbol that is never used, which would be dead code. --- Generated by Claude Code Patch Reviewer