Re: [PATCH v14 05/10] mm: introduce memfd_secret system call to create "secret" memory areas

On Thu, Dec 03, 2020 at 08:29:44AM +0200, Mike Rapoport wrote:

+static vm_fault_t secretmem_fault(struct vm_fault *vmf)
+{
+	struct address_space *mapping = vmf->vma->vm_file->f_mapping;
+	struct inode *inode = file_inode(vmf->vma->vm_file);
+	pgoff_t offset = vmf->pgoff;
+	vm_fault_t ret = 0;
+	unsigned long addr;
+	struct page *page;
+	int err;
+
+	if (((loff_t)vmf->pgoff << PAGE_SHIFT) >= i_size_read(inode))
+		return vmf_error(-EINVAL);
+
+	page = find_get_page(mapping, offset);
+	if (!page) {
+
+		page = secretmem_alloc_page(vmf->gfp_mask);
+		if (!page)
+			return vmf_error(-ENOMEM);

Just use VM_FAULT_OOM directly.

+		err = add_to_page_cache(page, mapping, offset, vmf->gfp_mask);
+		if (unlikely(err))
+			goto err_put_page;

What if the error is EEXIST because somebody else raced with you to add
a new page to the page cache?

+		err = set_direct_map_invalid_noflush(page, 1);
+		if (err)
+			goto err_del_page_cache;

Does this work correctly if somebody else has a reference to the page
in the meantime?

+		addr = (unsigned long)page_address(page);
+		flush_tlb_kernel_range(addr, addr + PAGE_SIZE);
+
+		__SetPageUptodate(page);

Once you've added it to the cache, somebody else can come along and try
to lock it.  They will set PageWaiter.  Now you call __SetPageUptodate
and wipe out their PageWaiter bit.  So you won't wake them up when you
unlock.

You can call __SetPageUptodate before adding it to the page cache,
but once it's visible to another thread, you can't do that.

+		ret = VM_FAULT_LOCKED;
+	}
+
+	vmf->page = page;

You're supposed to return the page locked, so use find_lock_page() instead
of find_get_page().

+	return ret;
+
+err_del_page_cache:
+	delete_from_page_cache(page);
+err_put_page:
+	put_page(page);
+	return vmf_error(err);
+}