linux-kernel - [PATCH] tile: fix some issues in hugepage support

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Message-ID: <201308072040.r77KeaiE016083@farm-0012.internal.tilera.com>
Date:	Wed, 7 Aug 2013 11:00:45 -0400
From:	Chris Metcalf <cmetcalf@...era.com>
To:	<linux-kernel@...r.kernel.org>
Subject: [PATCH] tile: fix some issues in hugepage support

First, in huge_pte_offset(), we were erroneously checking
pgd_present(), which is always true, rather than pud_present(),
which is the thing that tells us if there is a top-level (L0) PTE.
Fixing this means we properly look up huge page entries only when
the Present bit is actually set in the PTE.

Second, use the standard pte_alloc_map() instead of the hand-rolled
pte_alloc_hugetlb() routine that basically was written to avoid
worrying about CONFIG_HIGHPTE.  However, we no longer plan to support
HIGHPTE, so a separate routine was just unnecessary code duplication.

Signed-off-by: Chris Metcalf <cmetcalf@...era.com>
---
 arch/tile/mm/hugetlbpage.c | 38 +++-----------------------------------
 1 file changed, 3 insertions(+), 35 deletions(-)

diff --git a/arch/tile/mm/hugetlbpage.c b/arch/tile/mm/hugetlbpage.c
index 650ccff..e514899 100644
--- a/arch/tile/mm/hugetlbpage.c
+++ b/arch/tile/mm/hugetlbpage.c
@@ -49,38 +49,6 @@ int huge_shift[HUGE_SHIFT_ENTRIES] = {
 #endif
 };
 
-/*
- * This routine is a hybrid of pte_alloc_map() and pte_alloc_kernel().
- * It assumes that L2 PTEs are never in HIGHMEM (we don't support that).
- * It locks the user pagetable, and bumps up the mm->nr_ptes field,
- * but otherwise allocate the page table using the kernel versions.
- */
-static pte_t *pte_alloc_hugetlb(struct mm_struct *mm, pmd_t *pmd,
-				unsigned long address)
-{
-	pte_t *new;
-
-	if (pmd_none(*pmd)) {
-		new = pte_alloc_one_kernel(mm, address);
-		if (!new)
-			return NULL;
-
-		smp_wmb(); /* See comment in __pte_alloc */
-
-		spin_lock(&mm->page_table_lock);
-		if (likely(pmd_none(*pmd))) {  /* Has another populated it ? */
-			mm->nr_ptes++;
-			pmd_populate_kernel(mm, pmd, new);
-			new = NULL;
-		} else
-			VM_BUG_ON(pmd_trans_splitting(*pmd));
-		spin_unlock(&mm->page_table_lock);
-		if (new)
-			pte_free_kernel(mm, new);
-	}
-
-	return pte_offset_kernel(pmd, address);
-}
 #endif
 
 pte_t *huge_pte_alloc(struct mm_struct *mm,
@@ -109,7 +77,7 @@ pte_t *huge_pte_alloc(struct mm_struct *mm,
 		else {
 			if (sz != PAGE_SIZE << huge_shift[HUGE_SHIFT_PAGE])
 				panic("Unexpected page size %#lx\n", sz);
-			return pte_alloc_hugetlb(mm, pmd, addr);
+			return pte_alloc_map(mm, NULL, pmd, addr);
 		}
 	}
 #else
@@ -144,14 +112,14 @@ pte_t *huge_pte_offset(struct mm_struct *mm, unsigned long addr)
 
 	/* Get the top-level page table entry. */
 	pgd = (pgd_t *)get_pte((pte_t *)mm->pgd, pgd_index(addr), 0);
-	if (!pgd_present(*pgd))
-		return NULL;
 
 	/* We don't have four levels. */
 	pud = pud_offset(pgd, addr);
 #ifndef __PAGETABLE_PUD_FOLDED
 # error support fourth page table level
 #endif
+	if (!pud_present(*pud))
+		return NULL;
 
 	/* Check for an L0 huge PTE, if we have three levels. */
 #ifndef __PAGETABLE_PMD_FOLDED
-- 
1.8.3.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/