From mboxrd@z Thu Jan 1 00:00:00 1970 From: Claude Code Review Bot To: dri-devel-reviews@example.com Subject: Claude review: drm/msm/a6xx: Enable CLX feature on A840 Date: Sat, 16 May 2026 08:47:13 +1000 Message-ID: In-Reply-To: <20260516-a8xx-clx-support-v1-2-62508bf651ac@oss.qualcomm.com> References: <20260516-a8xx-clx-support-v1-0-62508bf651ac@oss.qualcomm.com> <20260516-a8xx-clx-support-v1-2-62508bf651ac@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 **Issue: `a840_limits_tbl` should be `static const`.** ```c struct a6xx_hfi_limits_tbl a840_limits_tbl[] =3D { ``` This is missing both `static` and `const`. All other comparable catalog arr= ays in this file (e.g., `a840_dyn_pwrup_reglist_regs[]`) are `static const`= . This table is only referenced through the `DECLARE_ADRENO_LIMITS_TABLE(a8= 40_limits)` macro and should be: ```c static const struct a6xx_hfi_limits_tbl a840_limits_tbl[] =3D { ``` Without `static`, the symbol is externally visible (unnecessary namespace p= ollution). Without `const`, the data goes into the writable data section in= stead of rodata. Also, as noted above, the `DECLARE_ADRENO_LIMITS_TABLE` ma= cro assigns `name ## _tbl` to a member in a `static const` struct, so havin= g the underlying array be non-const creates a type inconsistency. **Data tables look reasonable.** Four limits entries (IFF for GX, IFF for M= X, PCLX for GX, PCLX for MX) with the MX PCLX entry disabled (`.enable =3D = 0`) =E2=80=94 this is a sensible per-domain configuration. The CLX domain t= able has two domains (GX and MXG) with different current budgets (100 vs 50= ), which matches the asymmetric power characteristics. --- Generated by Claude Code Patch Reviewer