@@ -887,6 +887,15 @@ inline pte_t pte_set_vma_features(pte_t pte, struct vm_area_struct *vma)
return pte;
}
+inline bool pte_exclusive(pte_t pte, struct vm_area_struct *vma)
+{
+ if (vma->vm_flags & VM_SHSTK)
+ return pte_dirty_hw(pte);
+ else
+ return pte_dirty(pte);
+}
+
+#ifdef CONFIG_TRANSPARENT_HUGEPAGE
inline pmd_t pmd_set_vma_features(pmd_t pmd, struct vm_area_struct *vma)
{
if (vma->vm_flags & VM_SHSTK)
@@ -894,4 +903,13 @@ inline pmd_t pmd_set_vma_features(pmd_t pmd, struct vm_area_struct *vma)
else
return pmd;
}
+
+inline bool pmd_exclusive(pmd_t pmd, struct vm_area_struct *vma)
+{
+ if (vma->vm_flags & VM_SHSTK)
+ return pmd_dirty_hw(pmd);
+ else
+ return pmd_dirty(pmd);
+}
+#endif /* CONFIG_TRANSPARENT_HUGEPAGE */
#endif /* CONFIG_X86_INTEL_SHADOW_STACK_USER */
@@ -1202,18 +1202,30 @@ static inline pte_t pte_set_vma_features(pte_t pte, struct vm_area_struct *vma)
return pte;
}
+static inline bool pte_exclusive(pte_t pte, struct vm_area_struct *vma)
+{
+ return pte_dirty(pte);
+}
+
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
static inline pmd_t pmd_set_vma_features(pmd_t pmd, struct vm_area_struct *vma)
{
return pmd;
}
+
+static inline bool pmd_exclusive(pmd_t pmd, struct vm_area_struct *vma)
+{
+ return pmd_dirty(pmd);
+}
#endif
#else
bool arch_copy_pte_mapping(vm_flags_t vm_flags);
pte_t pte_set_vma_features(pte_t pte, struct vm_area_struct *vma);
+bool pte_exclusive(pte_t pte, struct vm_area_struct *vma);
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
pmd_t pmd_set_vma_features(pmd_t pmd, struct vm_area_struct *vma);
+bool pmd_exclusive(pmd_t pmd, struct vm_area_struct *vma);
#endif
#endif
#endif /* CONFIG_MMU */
@@ -164,10 +164,12 @@ static int follow_pfn_pte(struct vm_area_struct *vma, unsigned long address,
* FOLL_FORCE can write to even unwritable pte's, but only
* after we've gone through a COW cycle and they are dirty.
*/
-static inline bool can_follow_write_pte(pte_t pte, unsigned int flags)
+static inline bool can_follow_write(pte_t pte, unsigned int flags,
+ struct vm_area_struct *vma)
{
return pte_write(pte) ||
- ((flags & FOLL_FORCE) && (flags & FOLL_COW) && pte_dirty(pte));
+ ((flags & FOLL_FORCE) && (flags & FOLL_COW) &&
+ pte_exclusive(pte, vma));
}
static struct page *follow_page_pte(struct vm_area_struct *vma,
@@ -205,7 +207,7 @@ static struct page *follow_page_pte(struct vm_area_struct *vma,
}
if ((flags & FOLL_NUMA) && pte_protnone(pte))
goto no_page;
- if ((flags & FOLL_WRITE) && !can_follow_write_pte(pte, flags)) {
+ if ((flags & FOLL_WRITE) && !can_follow_write(pte, flags, vma)) {
pte_unmap_unlock(ptep, ptl);
return NULL;
}
@@ -1469,10 +1469,12 @@ vm_fault_t do_huge_pmd_wp_page(struct vm_fault *vmf, pmd_t orig_pmd)
* FOLL_FORCE can write to even unwritable pmd's, but only
* after we've gone through a COW cycle and they are dirty.
*/
-static inline bool can_follow_write_pmd(pmd_t pmd, unsigned int flags)
+static inline bool can_follow_write(pmd_t pmd, unsigned int flags,
+ struct vm_area_struct *vma)
{
return pmd_write(pmd) ||
- ((flags & FOLL_FORCE) && (flags & FOLL_COW) && pmd_dirty(pmd));
+ ((flags & FOLL_FORCE) && (flags & FOLL_COW) &&
+ pmd_exclusive(pmd, vma));
}
struct page *follow_trans_huge_pmd(struct vm_area_struct *vma,
@@ -1485,7 +1487,7 @@ struct page *follow_trans_huge_pmd(struct vm_area_struct *vma,
assert_spin_locked(pmd_lockptr(mm, pmd));
- if (flags & FOLL_WRITE && !can_follow_write_pmd(*pmd, flags))
+ if (flags & FOLL_WRITE && !can_follow_write(*pmd, flags, vma))
goto out;
/* Avoid dumping huge zero page */
Can_follow_write_pte() verifies that a read-only page is the task's own copy by ensuring the page has gone through faultin_page() and the PTE is Dirty. A Shadow Stack (SHSTK) PTE must be (read-only + _PAGE_DIRTY_HW). When a task does fork(), its SHSTK PTEs become (read-only + _PAGE_DIRTY_SW). This causes the next SHSTK access (i.e. CALL, RET, INCSSP) to trigger a fault; the page is then copied, and (read-only + _PAGE_DIRTY_HW) is restored. To update can_follow_write_pte() for SHSTK, introduce pte_exclusive(). It verifies a data PTE is Dirty and a SHSTK PTE has _PAGE_DIRTY_HW. Also rename can_follow_write_pte() to can_follow_write() to make its meaning clear; i.e. "Can we write to the page?", not "Is the PTE writable?" Also apply same changes to the huge memory case. Signed-off-by: Yu-cheng Yu <yu-cheng.yu@intel.com> --- arch/x86/mm/pgtable.c | 18 ++++++++++++++++++ include/asm-generic/pgtable.h | 12 ++++++++++++ mm/gup.c | 8 +++++--- mm/huge_memory.c | 8 +++++--- 4 files changed, 40 insertions(+), 6 deletions(-)