[V5,2/2] selftests: vm: Add test for Soft-Dirty PTE bit

From: Gabriel Krisman Bertazi <krisman@collabora.com>

This introduces three tests:
1) Sanity check soft dirty basic semantics: allocate area, clean, dirty,
check if the SD bit is flipped.
2) Check VMA reuse: validate the VM_SOFTDIRTY usage
3) Check soft-dirty on huge pages

This was motivated by Will Deacon's fix commit 912efa17e512 ("mm: proc:
Invalidate TLB after clearing soft-dirty page state"). I was tracking the
same issue that he fixed, and this test would have caught it.

Cc: Will Deacon <will@kernel.org>
Signed-off-by: Gabriel Krisman Bertazi <krisman@collabora.com>
Signed-off-by: Muhammad Usama Anjum <usama.anjum@collabora.com>
---
Andrew has dropped the V3 of this patch.

Changes in V5:
- Correct indentation of macros
- Tested by reverting 912efa17e512 and one sub test failed
- Updated the writing values to the memory and added a comment

Changes in V4:
Cosmetic changes
Removed global variables
Replaced ksft_print_msg with ksft_exit_fail_msg to exit the program at
once
Some other minor changes
Correct the authorship of the patch

Tests of soft dirty bit in this patch and in madv_populate.c are
non-overlapping. madv_populate.c has only one soft-dirty bit test in the
context of different advise (MADV_POPULATE_READ and
MADV_POPULATE_WRITE). This new test adds more tests.

Tab width of 8 has been used to align the macros. This alignment may look
odd in shell or email. But it looks alright in editors.

Test output:
TAP version 13
1..5
ok 1 Test test_simple
ok 2 Test test_vma_reuse reused memory location
ok 3 Test test_vma_reuse dirty bit of previous page
ok 4 Test test_hugepage huge page allocation
ok 5 Test test_hugepage huge page dirty bit
 # Totals: pass:5 fail:0 xfail:0 xpass:0 skip:0 error:0

Or

TAP version 13
1..5
ok 1 Test test_simple
ok 2 Test test_vma_reuse reused memory location
ok 3 Test test_vma_reuse dirty bit of previous page
ok 4 # SKIP Test test_hugepage huge page allocation
ok 5 # SKIP Test test_hugepage huge page dirty bit
 # Totals: pass:3 fail:0 xfail:0 xpass:0 skip:2 error:0

Changes in V3:
Move test to selftests/vm
Use kselftest macros
Minor updates to make code more maintainable
Add configurations in config file

V2 of this patch:
https://lore.kernel.org/lkml/20210603151518.2437813-1-krisman@collabora.com/
---
 tools/testing/selftests/vm/.gitignore   |   1 +
 tools/testing/selftests/vm/Makefile     |   2 +
 tools/testing/selftests/vm/config       |   2 +
 tools/testing/selftests/vm/soft-dirty.c | 146 ++++++++++++++++++++++++
 4 files changed, 151 insertions(+)
 create mode 100644 tools/testing/selftests/vm/soft-dirty.c

Muhammad Usama Anjum <usama.anjum@collabora.com> writes:

> new file mode 100644
> index 0000000000000..3153ebac6909b
> --- /dev/null
> +++ b/tools/testing/selftests/vm/soft-dirty.c
> @@ -0,0 +1,146 @@
> +// SPDX-License-Identifier: GPL-2.0
> +#include <stdio.h>
> +#include <string.h>
> +#include <stdbool.h>
> +#include <fcntl.h>
> +#include <stdint.h>
> +#include <malloc.h>
> +#include <sys/mman.h>
> +#include "../kselftest.h"
> +#include "vm_util.h"
> +
> +#define PAGEMAP_FILE_PATH "/proc/self/pagemap"
> +#define TEST_ITERATIONS 10000
> +
> +static void test_simple(int pagemap_fd, int pagesize)
> +{
> +	int i;
> +	char *map;
> +
> +	map = aligned_alloc(pagesize, pagesize);
> +	if (!map)
> +		ksft_exit_fail_msg("mmap failed\n");
> +
> +	clear_softdirty();
> +
> +	for (i = 0 ; i < TEST_ITERATIONS; i++) {
> +		if (pagemap_is_softdirty(pagemap_fd, map) == 1) {
> +			ksft_print_msg("dirty bit was 1, but should be 0 (i=%d)\n", i);
> +			break;
> +		}
> +
> +		clear_softdirty();
> +		// Write something to the page to get the dirty bit enabled on the page
> +		map[0] = i % 255;

you don't need this mod at all but at least it should be 256 :).  I think
Either 'map[0] = !map[0]' or keeping the original 'map[0]++' is fine.

> +
> +		if (pagemap_is_softdirty(pagemap_fd, map) == 0) {
> +			ksft_print_msg("dirty bit was 0, but should be 1 (i=%d)\n", i);
> +			break;
> +		}
> +
> +		clear_softdirty();
> +	}
> +	free(map);
> +
> +	ksft_test_result(i == TEST_ITERATIONS, "Test %s\n", __func__);
> +}
> +
> +static void test_vma_reuse(int pagemap_fd, int pagesize)
> +{
> +	char *map, *map2;
> +
> +	map = mmap(NULL, pagesize, (PROT_READ | PROT_WRITE), (MAP_PRIVATE | MAP_ANON), -1, 0);
> +	if (map == MAP_FAILED)
> +		ksft_exit_fail_msg("mmap failed");
> +
> +	clear_softdirty();
> +
> +	/* Write to the page before unmapping and map the same size region again to check
> +	 * if same memory region is gotten next time and if dirty bit is preserved across
> +	 * this type of allocations.
> +	 */

This reads weird.  It should *not* be preserved across different
mappings.  Also, we are not testing if the same region is reused, we are
depending on it to test the sd bit.

/* Ensures the soft-dirty bit is reset accross different mappings on the
same address.  */

> +	map[0]++;

This is inconsistent with the other two tests.

> +
> +	munmap(map, pagesize);
> +
> +	map2 = mmap(NULL, pagesize, (PROT_READ | PROT_WRITE), (MAP_PRIVATE | MAP_ANON), -1, 0);
> +	if (map2 == MAP_FAILED)
> +		ksft_exit_fail_msg("mmap failed");
> +
> +	ksft_test_result(map == map2, "Test %s reused memory location\n", __func__);

if map != map2, the test itself is broken, meaning we should skip it, not
fail, i guess.

On 3/19/22 12:23 AM, Gabriel Krisman Bertazi wrote:
>> +static void test_vma_reuse(int pagemap_fd, int pagesize)
>> +{
>> +	char *map, *map2;
>> +
>> +	map = mmap(NULL, pagesize, (PROT_READ | PROT_WRITE), (MAP_PRIVATE | MAP_ANON), -1, 0);
>> +	if (map == MAP_FAILED)
>> +		ksft_exit_fail_msg("mmap failed");
>> +
>> +	clear_softdirty();
>> +
>> +	/* Write to the page before unmapping and map the same size region again to check
>> +	 * if same memory region is gotten next time and if dirty bit is preserved across
>> +	 * this type of allocations.
>> +	 */
> This reads weird.  It should *not* be preserved across different
> mappings.  Also, we are not testing if the same region is reused, we are
> depending on it to test the sd bit.
> 
> /* Ensures the soft-dirty bit is reset accross different mappings on the
> same address.  */
Soft dirty bit is always set for new regions (reused or not). I'll
correct the comment and improve this sub test.