intel-iommu: PAE memory corruption fix

PAGE_MASK is 0xFFFFF000 on i386 -- even with PAE.

So it's not sufficient to ensure that you use phys_addr_t or uint64_t
everywhere you handle physical addresses -- you also have to avoid using
the construct 'addr & PAGE_MASK', because that will strip the high 32
bits of the address.

This patch avoids that problem by using PHYSICAL_PAGE_MASK instead of
PAGE_MASK where appropriate. It leaves '& PAGE_MASK' in a few instances
that don't matter -- where it's being used on the virtual bus addresses
we're dishing out, which are 32-bit anyway.

Since PHYSICAL_PAGE_MASK is not present on other architectures, we have
to define it (to PAGE_MASK) if it's not already defined.

Maybe it would be better just to fix PAGE_MASK for i386/PAE?

Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

diff --git a/drivers/pci/intel-iommu.c b/drivers/pci/intel-iommu.c
index 001b328adf804e1e75d9c561df9e704b424b5468..a563fbe559d0505b90d1f736beb3a8e8bf197717 100644 (file)
--- a/drivers/pci/intel-iommu.c
+++ b/drivers/pci/intel-iommu.c
@@ -59,6 +59,10 @@
 #define DMA_32BIT_PFN          IOVA_PFN(DMA_BIT_MASK(32))
 #define DMA_64BIT_PFN          IOVA_PFN(DMA_BIT_MASK(64))
 
+#ifndef PHYSICAL_PAGE_MASK
+#define PHYSICAL_PAGE_MASK PAGE_MASK
+#endif
+
 /* global iommu list, set NULL for ignored DMAR units */
 static struct intel_iommu **g_iommus;
 
@@ -1216,7 +1220,7 @@ static void dmar_init_reserved_ranges(void)
                        if (!r->flags || !(r->flags & IORESOURCE_MEM))
                                continue;
                        addr = r->start;
-                       addr &= PAGE_MASK;
+                       addr &= PHYSICAL_PAGE_MASK;
                        size = r->end - addr;
                        size = PAGE_ALIGN(size);
                        iova = reserve_iova(&reserved_iova_list, IOVA_PFN(addr),
@@ -2173,7 +2177,8 @@ static dma_addr_t __intel_map_single(struct device *hwdev, phys_addr_t paddr,
         * is not a big problem
         */
        ret = domain_page_mapping(domain, start_paddr,
-               ((u64)paddr) & PAGE_MASK, size, prot);
+                                 ((u64)paddr) & PHYSICAL_PAGE_MASK,
+                                 size, prot);
        if (ret)
                goto error;
 
@@ -2463,8 +2468,8 @@ static int intel_map_sg(struct device *hwdev, struct scatterlist *sglist, int ne
                addr = page_to_phys(sg_page(sg)) + sg->offset;
                size = aligned_size((u64)addr, sg->length);
                ret = domain_page_mapping(domain, start_addr + offset,
-                       ((u64)addr) & PAGE_MASK,
-                       size, prot);
+                                         ((u64)addr) & PHYSICAL_PAGE_MASK,
+                                         size, prot);
                if (ret) {
                        /*  clear the page */
                        dma_pte_clear_range(domain, start_addr,