fork: lock VMAs of the parent process when forking

When forking a child process, parent write-protects an anonymous page
and COW-shares it with the child being forked using copy_present_pte().
Parent's TLB is flushed right before we drop the parent's mmap_lock in
dup_mmap(). If we get a write-fault before that TLB flush in the parent,
and we end up replacing that anonymous page in the parent process in
do_wp_page() (because, COW-shared with the child), this might lead to
some stale writable TLB entries targeting the wrong (old) page.
Similar issue happened in the past with userfaultfd (see flush_tlb_page()
call inside do_wp_page()).
Lock VMAs of the parent process when forking a child, which prevents
concurrent page faults during fork operation and avoids this issue.
This fix can potentially regress some fork-heavy workloads. Kernel build
time did not show noticeable regression on a 56-core machine while a
stress test mapping 10000 VMAs and forking 5000 times in a tight loop
shows ~5% regression. If such fork time regression is unacceptable,
disabling CONFIG_PER_VMA_LOCK should restore its performance. Further
optimizations are possible if this regression proves to be problematic.

Suggested-by: David Hildenbrand <[email protected]>
Reported-by: Jiri Slaby <[email protected]>
Closes: https://lore.kernel.org/all/[email protected]/
Reported-by: Holger Hoffstätte <[email protected]>
Closes: https://lore.kernel.org/all/[email protected]/
Reported-by: Jacob Young <[email protected]>
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=217624
Fixes: 0bff0aa ("x86/mm: try VMA lock-based page fault handling first")
Cc: [email protected]
Signed-off-by: Suren Baghdasaryan <[email protected]>
Acked-by: David Hildenbrand <[email protected]>

Author: surenbaghdasaryan

kernel/fork.c

@@ -662,6 +662,12 @@ static __latent_entropy int dup_mmap(struct mm_struct *mm,
 		retval = -EINTR;
 		goto fail_uprobe_end;
 	}
+#ifdef CONFIG_PER_VMA_LOCK
+	/* Disallow any page faults before calling flush_cache_dup_mm */
+	for_each_vma(old_vmi, mpnt)
+		vma_start_write(mpnt);
+	vma_iter_init(&old_vmi, oldmm, 0);
+#endif
 	flush_cache_dup_mm(oldmm);
 	uprobe_dup_mmap(oldmm, mm);
 	/*