Re: [PATCH v2] selftests/mm: pagemap_scan ioctl: add PFN ZERO test cases

[David Hildenbrand <david@redhat.com>]

On 02.07.25 16:20, Muhammad Usama Anjum wrote:
> Add test cases to test the correctness of PFN ZERO flag of pagemap_scan
> ioctl. Test with normal pages backed memory and huge pages backed
> memory.
> 
> Cc: David Hildenbrand <david@redhat.com>
> Signed-off-by: Muhammad Usama Anjum <usama.anjum@collabora.com>
> ---
> The bug has been fixed [1].
> 
> [1] https://lore.kernel.org/all/20250617143532.2375383-1-david@redhat.com
> Changes since v1:
> - Skip if madvise() fails
> - Skip test if use_zero_page isn't set to 1
> - Keep on using memalign()+free() to allocate huge pages
> ---
>   tools/testing/selftests/mm/pagemap_ioctl.c | 86 +++++++++++++++++++++-
>   1 file changed, 85 insertions(+), 1 deletion(-)
> 
> diff --git a/tools/testing/selftests/mm/pagemap_ioctl.c b/tools/testing/selftests/mm/pagemap_ioctl.c
> index 57b4bba2b45f3..976ab357f4651 100644
> --- a/tools/testing/selftests/mm/pagemap_ioctl.c
> +++ b/tools/testing/selftests/mm/pagemap_ioctl.c
> @@ -1,4 +1,5 @@
>   // SPDX-License-Identifier: GPL-2.0
> +
>   #define _GNU_SOURCE
>   #include <stdio.h>
>   #include <fcntl.h>
> @@ -1480,6 +1481,86 @@ static void transact_test(int page_size)
>   			      extra_thread_faults);
>   }
>   
> +bool is_use_zero_page_set(void)
> +{
> +	ssize_t bytes_read;
> +	char buffer[2];
> +	int fd;
> +
> +	fd = open("/sys/kernel/mm/transparent_hugepage/use_zero_page", O_RDONLY);
> +	if (fd < 0)
> +		return 0;
> +
> +	bytes_read = read(fd, buffer, sizeof(buffer) - 1);
> +	if (bytes_read <= 0) {
> +		close(fd);
> +		return 0;
> +	}
> +
> +	close(fd);
> +	if (atoi(buffer) != 1)
> +		return 0;
> +
> +	return 1;
> +}

You should probably factor out detect_huge_zeropage() from cow.c into 
vm_utils.c, and let it return the result.

> +
> +void zeropfn_tests(void)
> +{
> +	unsigned long long mem_size;
> +	struct page_region vec;
> +	int i, ret;
> +	char *mem;
> +
> +	/* Test with normal memory */
> +	mem_size = 10 * page_size;
> +	mem = mmap(NULL, mem_size, PROT_READ, MAP_PRIVATE | MAP_ANON, -1, 0);
> +	if (mem == MAP_FAILED)
> +		ksft_exit_fail_msg("error nomem\n");
> +
> +	/* Touch each page to ensure it's mapped */
> +	for (i = 0; i < mem_size; i += page_size)
> +		(void)((volatile char *)mem)[i];
> +
> +	ret = pagemap_ioctl(mem, mem_size, &vec, 1, 0,
> +			    (mem_size / page_size), PAGE_IS_PFNZERO, 0, 0, PAGE_IS_PFNZERO);
> +	if (ret < 0)
> +		ksft_exit_fail_msg("error %d %d %s\n", ret, errno, strerror(errno));
> +
> +	ksft_test_result(ret == 1 && LEN(vec) == (mem_size / page_size),
> +			 "%s all pages must have PFNZERO set\n", __func__);
> +
> +	munmap(mem, mem_size);
> +
> +	/* Test with huge page if user_zero_page is set to 1 */
> +	if (!is_use_zero_page_set()) {
> +		ksft_test_result_skip("%s use_zero_page not supported or set to 1\n", __func__);
> +		return;
> +	}
> +
> +	mem_size = 10 * hpage_size;
> +	mem = memalign(hpage_size, mem_size);
> +	if (!mem)
> +		ksft_exit_fail_msg("error nomem\n");

Didn't we discuss using mmap() instead?

See run_with_huge_zeropage() in cow.c on how to do the alignemnt 
yourself easily.


-- 
Cheers,

David / dhildenb