On 25.03.21 12:31, Wang Wensheng wrote:
sparse_buffer_init() and sparse_buffer_fini() should appear in pair, or a WARN issue would be through the next time sparse_buffer_init() runs.Add the missing sparse_buffer_fini() in error branch. Fixes: 85c77f791390 ("mm/sparse: add new sparse_init_nid() and sparse_init()") Signed-off-by: Wang Wensheng <wangwenshe...@huawei.com> --- mm/sparse.c | 1 + 1 file changed, 1 insertion(+) diff --git a/mm/sparse.c b/mm/sparse.c index 7bd23f9..33406ea 100644 --- a/mm/sparse.c +++ b/mm/sparse.c @@ -547,6 +547,7 @@ static void __init sparse_init_nid(int nid, unsigned long pnum_begin, pr_err("%s: node[%d] memory map backing failed. Some memory will not be available.", __func__, nid); pnum_begin = pnum; + sparse_buffer_fini(); goto failed; } check_usemap_section_nr(nid, usage);
Did you actually manage to reproduce this? I would have thought that __populate_section_memmap() barely fails during boot ...
Anyhow, looks like the right think to do to me Reviewed-by: David Hildenbrand <da...@redhat.com> -- Thanks, David / dhildenb
