From c887200da2b8c4bd093892f77888d61fa0ae6c11 Mon Sep 17 00:00:00 2001 From: Muchun Song Date: Mon, 20 Jun 2022 19:06:16 +0800 Subject: [PATCH] mm-memory_hotplug-make-hugetlb_optimize_vmemmap-compatible-with-memmap_on_memory-v5 walk vmemmap page tables to avoid false-positive Link: https://lkml.kernel.org/r/20220620110616.12056-3-songmuchun@bytedance.com Signed-off-by: Muchun Song Co-developed-by: Oscar Salvador Signed-off-by: Oscar Salvador Cc: David Hildenbrand Cc: Jonathan Corbet Cc: Mike Kravetz Cc: Paul E. McKenney Cc: Xiongchun Duan Signed-off-by: Andrew Morton --- mm/hugetlb_vmemmap.c | 69 ++++++++++++++++++++++++++++---------------- 1 file changed, 44 insertions(+), 25 deletions(-) diff --git a/mm/hugetlb_vmemmap.c b/mm/hugetlb_vmemmap.c index 982f27c24a95..6d9801bb3fec 100644 --- a/mm/hugetlb_vmemmap.c +++ b/mm/hugetlb_vmemmap.c @@ -10,6 +10,7 @@ */ #define pr_fmt(fmt) "HugeTLB: " fmt +#include #include "hugetlb_vmemmap.h" /* @@ -99,34 +100,52 @@ int hugetlb_vmemmap_alloc(struct hstate *h, struct page *head) static unsigned int vmemmap_optimizable_pages(struct hstate *h, struct page *head) { - struct mem_section *ms; - struct page *vmemmap_page; - unsigned long pfn = page_to_pfn(head); - if (READ_ONCE(vmemmap_optimize_mode) == VMEMMAP_OPTIMIZE_OFF) return 0; - ms = __pfn_to_section(pfn); - vmemmap_page = sparse_decode_mem_map(ms->section_mem_map, - pfn_to_section_nr(pfn)); - /* - * Only the vmemmap pages' vmemmap may be marked as VmemmapSelfHosted. - * - * Due to HugeTLB alignment requirements, and the vmemmap pages being - * at the start of the hotplugged memory region. Checking any vmemmap - * page's vmemmap is fine. - * - * [ hotplugged memory ] - * [ vmemmap ][ usable memory ] - * ^ | | | - * +---+ | | - * ^ | | - * +--------+ | - * ^ | - * +-----------------+ - */ - if (PageVmemmapSelfHosted(vmemmap_page)) - return 0; + if (IS_ENABLED(CONFIG_MEMORY_HOTPLUG)) { + pmd_t *pmdp, pmd; + struct page *vmemmap_page; + unsigned long vaddr = (unsigned long)head; + + /* + * Only the vmemmap page's vmemmap page can be self-hosted. + * Walking the page tables to find the backing page of the + * vmemmap page. + */ + pmdp = pmd_off_k(vaddr); + /* + * The READ_ONCE() is used to stabilize *pmdp in a register or + * on the stack so that it will stop changing under the code. + * The only concurrent operation where it can be changed is + * split_vmemmap_huge_pmd() (*pmdp will be stable after this + * operation). + */ + pmd = READ_ONCE(*pmdp); + if (pmd_leaf(pmd)) + vmemmap_page = pmd_page(pmd) + pte_index(vaddr); + else + vmemmap_page = pte_page(*pte_offset_kernel(pmdp, vaddr)); + /* + * Due to HugeTLB alignment requirements and the vmemmap pages + * being at the start of the hotplugged memory region in + * memory_hotplug.memmap_on_memory case. Checking any vmemmap + * page's vmemmap page if it is marked as VmemmapSelfHosted is + * sufficient. + * + * [ hotplugged memory ] + * [ section ][...][ section ] + * [ vmemmap ][ usable memory ] + * ^ | | | + * +---+ | | + * ^ | | + * +-------+ | + * ^ | + * +-------------------------------------------+ + */ + if (PageVmemmapSelfHosted(vmemmap_page)) + return 0; + } return hugetlb_optimize_vmemmap_pages(h); } -- 2.50.1