Added two new changes based on observations from Ethan. Like the rest of
the fixes in this series, these do not trigger problems today given the
limited feature set supported. Re-tested the series with emulated devices,
VFIO passthrough usage with amd-iommu is not possible since merge of commit
31753d5a336f ("hw/i386/amd_iommu: Fix device setup failure when PT is on.")
regardless of guest kernel iommu mode.Changes since v3: - Made an additional change in PATCH 1 with correct encoding for AMDVI_MAX_GVA_ADDR, adding Ethan as co-author. Dropped Vasant R-b. - Added patch by Ethan fixing truncation bug. Thank you, Alejandro v3: https://lore.kernel.org/all/20250529193023.3590780-1-alejandro.j.jime...@oracle.com/ Alejandro Jimenez (7): amd_iommu: Fix Miscellaneous Information Register 0 encoding amd_iommu: Fix Device ID decoding for INVALIDATE_IOTLB_PAGES command amd_iommu: Update bitmasks representing DTE reserved fields amd_iommu: Fix masks for various IOMMU MMIO Registers amd_iommu: Fix mask to retrieve Interrupt Table Root Pointer from DTE amd_iommu: Fix the calculation for Device Table size amd_iommu: Remove duplicated definitions Ethan Milon (1): amd_iommu: Fix truncation of oldval in amdvi_writeq hw/i386/amd_iommu.c | 17 +++++++------ hw/i386/amd_iommu.h | 59 ++++++++++++++++++++++----------------------- 2 files changed, 38 insertions(+), 38 deletions(-) base-commit: a6f02277595136832c9e9bcaf447ab574f7b1128 -- 2.43.5
