uprobe: use FOLL_SPLIT_PMD instead of FOLL_SPLIT (5a52c9df) · Commits · 戴 / test

Use the newly added FOLL_SPLIT_PMD in uprobe. This preserves the huge page when the uprobe is enabled. When the uprobe is disabled, newer instances of the same application could still benefit from huge page. For the next step, we will enable khugepaged to regroup the pmd, so that existing instances of the application could also benefit from huge page after the uprobe is disabled. Link: http://lkml.kernel.org/r/20190815164525.1848545-5-songliubraving@fb.com Signed-off-by:

Song Liu <songliubraving@fb.com> Acked-by:

Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Reviewed-by:

Srikar Dronamraju <srikar@linux.vnet.ibm.com> Reviewed-by:

Oleg Nesterov <oleg@redhat.com> Signed-off-by:

Andrew Morton <akpm@linux-foundation.org> Signed-off-by:

Linus Torvalds <torvalds@linux-foundation.org>

kernel/events/uprobes.c

+2 −4

Original line number	Diff line number	Diff line
		@@ -155,7 +155,7 @@ static int __replace_page(struct vm_area_struct *vma, unsigned long addr,
		{
		struct mm_struct *mm = vma->vm_mm;
		struct page_vma_mapped_walk pvmw = {
		.page = old_page,
		.page = compound_head(old_page),
		.vma = vma,
		.address = addr,
		};
		@@ -166,8 +166,6 @@ static int __replace_page(struct vm_area_struct *vma, unsigned long addr,
		mmu_notifier_range_init(&range, MMU_NOTIFY_CLEAR, 0, vma, mm, addr,
		addr + PAGE_SIZE);

		VM_BUG_ON_PAGE(PageTransHuge(old_page), old_page);

		if (new_page) {
		err = mem_cgroup_try_charge(new_page, vma->vm_mm, GFP_KERNEL,
		&memcg, false);
		@@ -481,7 +479,7 @@ int uprobe_write_opcode(struct arch_uprobe auprobe, struct mm_struct mm,
		retry:
		/* Read the page with vaddr into memory */
		ret = get_user_pages_remote(NULL, mm, vaddr, 1,
		FOLL_FORCE \| FOLL_SPLIT, &old_page, &vma, NULL);
		FOLL_FORCE \| FOLL_SPLIT_PMD, &old_page, &vma, NULL);
		if (ret <= 0)
		return ret;

Admin message