From mboxrd@z Thu Jan 1 00:00:00 1970 From: Claude Code Review Bot To: dri-devel-reviews@example.com Subject: Claude review: iommu: remove deprecated iova_to_phys from domain_ops and io_pgtable_ops Date: Thu, 04 Jun 2026 15:00:36 +1000 Message-ID: In-Reply-To: <20260531093637.3893199-10-guanghuifeng@linux.alibaba.com> References: <20260531093637.3893199-1-guanghuifeng@linux.alibaba.com> <20260531093637.3893199-10-guanghuifeng@linux.alibaba.com> X-Mailer: Claude Code Patch Reviewer Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7bit MIME-Version: 1.0 Patch Review Clean removal. The patch correctly: - Removes the wrapper functions from io-pgtable backends - Removes the fallback in `iommu_iova_to_phys_length()` - Removes the ops fields from both structs - Updates documentation comments One concern: this patch assumes ALL drivers have been migrated. If any out-of-tree or staging driver still sets `.iova_to_phys` without `.iova_to_phys_length`, it will silently compile (the field is just gone) but `iommu_iova_to_phys()` will return 0 for all translations. This is inherent to the approach and acceptable for in-tree code. --- **Summary of issues to address:** 1. **Bug (patch 7)**: The `npages` calculation in `batch_from_domain` and `raw_pages_from_domain` incorrectly adds `page_offset` to the PTE-remaining calculation. 2. **Missing coverage**: Selftests don't verify the new `mapped_length` output. 3. **Semantic mismatch (patch 2)**: DART `pgsize_bitmap` used as a size works by accident. 4. **Bundled bug fixes**: The generic_pt error path fix (patch 3) and mtk_iommu_v1 page offset fix (patch 6) should be separate patches. 5. **Functionality gap (patch 4)**: arm-smmu hardware translation path doesn't report `mapped_length`. 6. **Series structure (v1)**: Too many drivers per patch; v2 fixes this. --- Generated by Claude Code Patch Reviewer