On 5/24/23 3:02 PM, Tian, Kevin wrote:
From: Liu, Yi Lyi.l.liu@intel.com Sent: Thursday, May 11, 2023 10:51 PM
From: Lu Baolubaolu.lu@linux.intel.com
The nested domain fields are exclusive to those that used for a DMA remapping domain. Use union to avoid memory waste.
Signed-off-by: Lu Baolubaolu.lu@linux.intel.com Signed-off-by: Yi Liuyi.l.liu@intel.com
drivers/iommu/intel/iommu.h | 35 +++++++++++++++++++++++++++++------ 1 file changed, 29 insertions(+), 6 deletions(-)
diff --git a/drivers/iommu/intel/iommu.h b/drivers/iommu/intel/iommu.h index 1c5e1d88862b..e818520f4068 100644 --- a/drivers/iommu/intel/iommu.h +++ b/drivers/iommu/intel/iommu.h @@ -596,15 +596,38 @@ struct dmar_domain { spinlock_t lock; /* Protect device tracking lists */ struct list_head devices; /* all devices' list */
- struct dma_pte *pgd; /* virtual address */
- int gaw; /* max guest address width */
- /* adjusted guest address width, 0 is level 2 30-bit */
- int agaw; int iommu_superpage;/* Level of superpages supported: 0 == 4KiB (no superpages), 1 ==
2MiB, 2 == 1GiB, 3 == 512GiB, 4 == 1TiB */
- u64 max_addr; /* maximum mapped address */
- union {
/* DMA remapping domain */struct {/* virtual address */struct dma_pte *pgd;/* max guest address width */int gaw;/** adjusted guest address width:* 0: level 2 30-bit* 1: level 3 39-bit* 2: level 4 48-bit* 3: level 5 57-bit*/int agaw;/* maximum mapped address */u64 max_addr;};what about 'nid'?
"nid" represents which NUMA node should we allocate pages from for this domain. It's updated every time when a domain is attached/detached to/from a device or pasid.
Generally speaking, "nid" is common for all types of domain. But in this case, only a DMA remapping domain has a need to allocate pages. I intend to keep it as it for now. There's more cleanup rooms if we limit it only for DMA remapping domain.
Best regards, baolu