linux-kernel - [PATCH 2/2] mm: pefer fault_around_pages to fault_around

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

Message-Id: <4995bad07fe9baa51c786fa0d81819dddfb57654.1679089214.git.lstoakes@gmail.com>
Date:   Fri, 17 Mar 2023 21:58:26 +0000
From:   Lorenzo Stoakes <lstoakes@...il.com>
To:     linux-mm@...ck.org, linux-kernel@...r.kernel.org,
        Andrew Morton <akpm@...ux-foundation.org>
Cc:     David Hildenbrand <david@...hat.com>,
        Matthew Wilcox <willy@...radead.org>,
        Vlastimil Babka <vbabka@...e.cz>,
        Lorenzo Stoakes <lstoakes@...il.com>
Subject: [PATCH 2/2] mm: pefer fault_around_pages to fault_around_bytes

All use of this value is now at page granularity, so specify the variable
as such too. This simplifies the logic.

We maintain the debugfs entry to ensure that there are no user-visible
changes.

Signed-off-by: Lorenzo Stoakes <lstoakes@...il.com>
---
 mm/memory.c | 26 +++++++++++++++-----------
 1 file changed, 15 insertions(+), 11 deletions(-)

diff --git a/mm/memory.c b/mm/memory.c
index 3d85aa7106b0..ae01f541ad30 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -4393,13 +4393,13 @@ vm_fault_t finish_fault(struct vm_fault *vmf)
 	return ret;
 }
 
-static unsigned long fault_around_bytes __read_mostly =
-	rounddown_pow_of_two(65536);
+static unsigned long fault_around_pages __read_mostly =
+	65536 >> PAGE_SHIFT;
 
 #ifdef CONFIG_DEBUG_FS
 static int fault_around_bytes_get(void *data, u64 *val)
 {
-	*val = fault_around_bytes;
+	*val = fault_around_pages << PAGE_SHIFT;
 	return 0;
 }
 
@@ -4411,10 +4411,13 @@ static int fault_around_bytes_set(void *data, u64 val)
 {
 	if (val / PAGE_SIZE > PTRS_PER_PTE)
 		return -EINVAL;
-	if (val > PAGE_SIZE)
-		fault_around_bytes = rounddown_pow_of_two(val);
-	else
-		fault_around_bytes = PAGE_SIZE; /* rounddown_pow_of_two(0) is undefined */
+
+	/*
+	 * The minimum value is 1 page, however this results in no fault-around
+	 * at all. See should_fault_around().
+	 */
+	fault_around_pages = max(rounddown_pow_of_two(val) >> PAGE_SHIFT, 1UL);
+
 	return 0;
 }
 DEFINE_DEBUGFS_ATTRIBUTE(fault_around_bytes_fops,
@@ -4440,18 +4443,18 @@ late_initcall(fault_around_debugfs);
  * This function doesn't cross VMA or page table boundaries, in order to call
  * map_pages() and acquire a PTE lock only once.
  *
- * fault_around_bytes defines how many bytes we'll try to map.
+ * fault_around_pages defines how many pages we'll try to map.
  * do_fault_around() expects it to be set to a power of two less than or equal
  * to PTRS_PER_PTE.
  *
  * The virtual address of the area that we map is naturally aligned to
- * fault_around_bytes rounded down to the machine page size
+ * fault_around_pages * PAGE_SIZE rounded down to the machine page size
  * (and therefore to page order).  This way it's easier to guarantee
  * that we don't cross page table boundaries.
  */
 static vm_fault_t do_fault_around(struct vm_fault *vmf)
 {
-	pgoff_t nr_pages = READ_ONCE(fault_around_bytes) >> PAGE_SHIFT;
+	pgoff_t nr_pages = READ_ONCE(fault_around_pages);
 	pgoff_t pte_off = pte_index(vmf->address);
 	/* The page offset of vmf->address within the VMA. */
 	pgoff_t vma_off = vmf->pgoff - vmf->vma->vm_pgoff;
@@ -4486,7 +4489,8 @@ static inline bool should_fault_around(struct vm_fault *vmf)
 	if (uffd_disable_fault_around(vmf->vma))
 		return false;
 
-	return fault_around_bytes >> PAGE_SHIFT > 1;
+	/* A single page implies no faulting 'around' at all. */
+	return fault_around_pages > 1;
 }
 
 static vm_fault_t do_read_fault(struct vm_fault *vmf)
-- 
2.39.2