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


Groups > linux.kernel > #1636496

[v3 4/9] mm: do not zero vmemmap_buf

From Pavel Tatashin <pasha.tatashin@oracle.com>
Newsgroups linux.kernel
Subject [v3 4/9] mm: do not zero vmemmap_buf
Date 2017-05-05 19:10 +0200
Message-ID <tDOTU-4Dx-13@gated-at.bofh.it> (permalink)
References <tDOTT-4Dx-3@gated-at.bofh.it>
Organization linux.* mail to news gateway

Show all headers | View raw


alloc_block_buf() can either use external allocator by calling
vmemmap_alloc_block() or when available use pre-allocated vmemmap_buf
to do allocation. In either case, alloc_block_buf() knows when to zero
memory based on the "zero" argument.  This is why it is not needed to
zero vmemmap_buf beforehand. Let clients of alloc_block_buf() to
decide whether that is needed.

Signed-off-by: Pavel Tatashin <pasha.tatashin@oracle.com>
---
 mm/sparse-vmemmap.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/mm/sparse-vmemmap.c b/mm/sparse-vmemmap.c
index 5d255b0..1e9508b 100644
--- a/mm/sparse-vmemmap.c
+++ b/mm/sparse-vmemmap.c
@@ -305,7 +305,7 @@ void __init sparse_mem_maps_populate_node(struct page **map_map,
 
 	size = ALIGN(size, PMD_SIZE);
 	vmemmap_buf_start = __earlyonly_bootmem_alloc(nodeid, size
-			* map_count, PMD_SIZE, __pa(MAX_DMA_ADDRESS), true);
+			* map_count, PMD_SIZE, __pa(MAX_DMA_ADDRESS), false);
 
 	if (vmemmap_buf_start) {
 		vmemmap_buf = vmemmap_buf_start;
-- 
1.7.1

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


Thread

[v3 4/9] mm: do not zero vmemmap_buf Pavel Tatashin <pasha.tatashin@oracle.com> - 2017-05-05 19:10 +0200

csiph-web