Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1702473

[v4 09/15] sparc64: optimized struct page zeroing

Show all headers | View raw

Add an optimized mm_zero_struct_page(), so struct page's are zeroed without
calling memset(). We do eight regular stores, thus avoid cost of membar.

Signed-off-by: Pavel Tatashin <pasha.tatashin@oracle.com>
Reviewed-by: Steven Sistare <steven.sistare@oracle.com>
Reviewed-by: Daniel Jordan <daniel.m.jordan@oracle.com>
Reviewed-by: Bob Picco <bob.picco@oracle.com>
---
 arch/sparc/include/asm/pgtable_64.h | 18 ++++++++++++++++++
 1 file changed, 18 insertions(+)

diff --git a/arch/sparc/include/asm/pgtable_64.h b/arch/sparc/include/asm/pgtable_64.h
index 6fbd931f0570..23ad51ea5340 100644
--- a/arch/sparc/include/asm/pgtable_64.h
+++ b/arch/sparc/include/asm/pgtable_64.h
@@ -230,6 +230,24 @@ extern unsigned long _PAGE_ALL_SZ_BITS;
 extern struct page *mem_map_zero;
 #define ZERO_PAGE(vaddr)	(mem_map_zero)
 
+/* This macro must be updated when the size of struct page changes,
+ * so use static assert to enforce the assumed size.
+ */
+#define	mm_zero_struct_page(pp)					\
+	do {							\
+		unsigned long *_pp = (void *)(pp);		\
+								\
+		BUILD_BUG_ON(sizeof(struct page) != 64);	\
+		_pp[0] = 0;					\
+		_pp[1] = 0;					\
+		_pp[2] = 0;					\
+		_pp[3] = 0;					\
+		_pp[4] = 0;					\
+		_pp[5] = 0;					\
+		_pp[6] = 0;					\
+		_pp[7] = 0;					\
+	} while (0)
+
 /* PFNs are real physical page numbers.  However, mem_map only begins to record
  * per-page information starting at pfn_base.  This is to handle systems where
  * the first physical page in the machine is at some huge physical address,
-- 
2.13.3

Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread

Thread

[v4 00/15] complete deferred page initialization Pavel Tatashin <pasha.tatashin@oracle.com> - 2017-08-02 22:40 +0200
  [v4 14/15] mm: optimize early system hash allocations Pavel Tatashin <pasha.tatashin@oracle.com> - 2017-08-02 22:40 +0200
  [v4 03/15] sparc64/mm: setting fields in deferred pages Pavel Tatashin <pasha.tatashin@oracle.com> - 2017-08-02 22:50 +0200
  [v4 06/15] sparc64: simplify vmemmap_populate Pavel Tatashin <pasha.tatashin@oracle.com> - 2017-08-02 22:50 +0200
  [v4 01/15] x86/mm: reserve only exiting low pages Pavel Tatashin <pasha.tatashin@oracle.com> - 2017-08-02 22:50 +0200
  [v4 02/15] x86/mm: setting fields in deferred pages Pavel Tatashin <pasha.tatashin@oracle.com> - 2017-08-02 22:50 +0200
  [v4 15/15] mm: debug for raw alloctor Pavel Tatashin <pasha.tatashin@oracle.com> - 2017-08-02 22:50 +0200
  [v4 09/15] sparc64: optimized struct page zeroing Pavel Tatashin <pasha.tatashin@oracle.com> - 2017-08-02 22:50 +0200
  [v4 12/15] mm: explicitly zero pagetable memory Pavel Tatashin <pasha.tatashin@oracle.com> - 2017-08-02 22:50 +0200
  [v4 10/15] x86/kasan: explicitly zero kasan shadow memory Pavel Tatashin <pasha.tatashin@oracle.com> - 2017-08-02 22:50 +0200
  [v4 04/15] mm: discard memblock data later Pavel Tatashin <pasha.tatashin@oracle.com> - 2017-08-02 22:50 +0200
  [v4 08/15] mm: zero struct pages during initialization Pavel Tatashin <pasha.tatashin@oracle.com> - 2017-08-02 22:50 +0200
  [v4 11/15] arm64/kasan: explicitly zero kasan shadow memory Pavel Tatashin <pasha.tatashin@oracle.com> - 2017-08-02 22:50 +0200
  [v4 07/15] mm: defining memblock_virt_alloc_try_nid_raw Pavel Tatashin <pasha.tatashin@oracle.com> - 2017-08-02 22:50 +0200
  [v4 05/15] mm: don't accessed uninitialized struct pages Pavel Tatashin <pasha.tatashin@oracle.com> - 2017-08-02 22:50 +0200

csiph-web