From mboxrd@z Thu Jan 1 00:00:00 1970 From: Claude Code Review Bot To: dri-devel-reviews@example.com Subject: Claude review: dma-buf: heaps: cma: Turn the heap into a module Date: Tue, 28 Apr 2026 14:45:01 +1000 Message-ID: In-Reply-To: <20260427-dma-buf-heaps-as-modules-v5-2-b6f5678feefc@kernel.org> References: <20260427-dma-buf-heaps-as-modules-v5-0-b6f5678feefc@kernel.org> <20260427-dma-buf-heaps-as-modules-v5-2-b6f5678feefc@kernel.org> X-Mailer: Claude Code Patch Reviewer Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 Patch Review **Change:**=20 - `drivers/dma-buf/heaps/Kconfig`: `DMABUF_HEAPS_CMA` changed from `bool` t= o `tristate` - `drivers/dma-buf/heaps/cma_heap.c`: Adds `MODULE_LICENSE("GPL")`, `MODULE= _IMPORT_NS("DMA_BUF")`, `MODULE_IMPORT_NS("DMA_BUF_HEAP")` The commit message correctly notes this heap can't be unloaded since there = is no `module_exit` function =E2=80=94 confirmed by grep showing no `module= _exit` in any heap source. This is a deliberate design choice; the kernel w= ill simply refuse to unload the module. The cover letter and v2 changelog m= ention this was documented. The two `MODULE_IMPORT_NS` lines match the namespaces used by `dma_heap_add= ()` (exported in `"DMA_BUF"`) and `mem_accounting` (exported in `"DMA_BUF_H= EAP"`). Even though `cma_heap.c` doesn't directly use `mem_accounting`, imp= orting `DMA_BUF_HEAP` is fine as the CMA heap uses other `DMA_BUF_HEAP`-nam= espaced symbols. No issues. --- Generated by Claude Code Patch Reviewer