linux-kernel - [PATCH] - GRU virtual -> physical translation

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>] [thread-next>] [day] [month] [year] [list]

Date:	Wed, 9 Jul 2008 14:14:39 -0500
From:	Jack Steiner <steiner@....com>
To:	akpm@...l.org, linux-kernel@...r.kernel.org
Cc:	linux-mm@...ck.org
Subject: [PATCH] - GRU virtual -> physical translation

Open code the equivalent to follow_page(). This eliminates the
requirement for an EXPORT of follow_page(). In addition, the code
is optimized for the specific case that is needed by the GRU and only
supports architectures supported by the GRU (ia64 & x86_64).

Signed-off-by: Jack Steiner <steiner@....com>

---

Andrew - do you want incremental patches or should I repost a V4 of
the original GRU driver.

After the GRU code is stabilized, I'll look into moving the open codes
PT lookup code into the kernel.


 drivers/misc/sgi-gru/grufault.c |   54 +++++++++++++++++++++++++++++++++-------
 1 file changed, 45 insertions(+), 9 deletions(-)

Index: linux/drivers/misc/sgi-gru/grufault.c
===================================================================
--- linux.orig/drivers/misc/sgi-gru/grufault.c	2008-07-09 13:54:16.786142769 -0500
+++ linux/drivers/misc/sgi-gru/grufault.c	2008-07-09 13:56:58.762219264 -0500
@@ -213,20 +213,56 @@ static int non_atomic_pte_lookup(struct 
 	return 0;
 }
 
+/*
+ *
+ * atomic_pte_lookup
+ *
+ * Convert a user virtual address to a physical address
+ * Only supports Intel large pages (2MB only) on x86_64.
+ *	ZZZ - hugepage support is incomplete
+ */
 static int atomic_pte_lookup(struct vm_area_struct *vma, unsigned long vaddr,
-			     int write, unsigned long *paddr, int *pageshift)
+	int write, unsigned long *paddr, int *pageshift)
 {
-	struct page *page;
+	pgd_t *pgdp;
+	pmd_t *pmdp;
+	pud_t *pudp;
+	pte_t pte;
+
+	WARN_ON(irqs_disabled());		/* ZZZ debug */
+
+	local_irq_disable();
+	pgdp = pgd_offset(vma->vm_mm, vaddr);
+	if (unlikely(pgd_none(*pgdp)))
+		goto err;
+
+	pudp = pud_offset(pgdp, vaddr);
+	if (unlikely(pud_none(*pudp)))
+		goto err;
+
+	pmdp = pmd_offset(pudp, vaddr);
+	if (unlikely(pmd_none(*pmdp)))
+		goto err;
+#ifdef CONFIG_X86_64
+	if (unlikely(pmd_large(*pmdp)))
+		pte = *(pte_t *) pmdp;
+	else
+#endif
+		pte = *pte_offset_kernel(pmdp, vaddr);
 
-	/* ZZZ Need to handle HUGE pages */
-	if (is_vm_hugetlb_page(vma))
-		return -EFAULT;
-	*pageshift = PAGE_SHIFT;
-	page = follow_page(vma, vaddr, (write ? FOLL_WRITE : 0));
-	if (!page)
+	local_irq_enable();
+
+	if (unlikely(!pte_present(pte) ||
+		     (write && (!pte_write(pte) || !pte_dirty(pte)))))
 		return 1;
-	*paddr = page_to_phys(page);
+
+	*paddr = pte_pfn(pte) << PAGE_SHIFT;
+	*pageshift = is_vm_hugetlb_page(vma) ? HPAGE_SHIFT : PAGE_SHIFT;
 	return 0;
+
+err:
+	local_irq_enable();
+	return 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/