Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit f03152bb authored by Joerg Roedel's avatar Joerg Roedel
Browse files

x86/amd-iommu: Make amd_iommu_iova_to_phys aware of multiple page sizes 



This patch extends the amd_iommu_iova_to_phys() function to
handle different page sizes correctly. It doesn't use
fetch_pte() anymore because we don't know (or care about)
the page_size used for mapping the given iova.

Signed-off-by: default avatarJoerg Roedel <joerg.roedel@amd.com>
parent 24cd7723

arch/x86/kernel/amd_iommu.c

+9 −4
Original line number Diff line number Diff line
@@ -2556,17 +2556,22 @@ static phys_addr_t amd_iommu_iova_to_phys(struct iommu_domain *dom, 
					  unsigned long iova)

{

	struct protection_domain *domain = dom->priv;

	unsigned long offset = iova & ~PAGE_MASK;

	unsigned long offset_mask;

	phys_addr_t paddr;

	u64 *pte;

	u64 *pte, __pte;



	pte = fetch_pte(domain, iova);



	if (!pte || !IOMMU_PTE_PRESENT(*pte))

		return 0;



	paddr  = *pte & IOMMU_PAGE_MASK;

	paddr |= offset;

	if (PM_PTE_LEVEL(*pte) == 0)

		offset_mask = PAGE_SIZE - 1;

	else

		offset_mask = PTE_PAGE_SIZE(*pte) - 1;



	__pte = *pte & PM_ADDR_MASK;

	paddr = (__pte & ~offset_mask) | (iova & offset_mask);



	return paddr;

}