selftests/vm: add KSM unmerge tests

Patch series "mm/ksm: break_ksm() cleanups and fixes", v2.

This series cleans up and fixes break_ksm(). In summary, we no longer use
fake write faults to break COW but instead FAULT_FLAG_UNSHARE. Further,
we move away from using follow_page() --- that we can hopefully remove
completely at one point --- and use new walk_page_range_vma() instead.

Fortunately, we can get rid of VM_FAULT_WRITE and FOLL_MIGRATION in common
code now.

Extend the existing ksm tests by an unmerge benchmark, and a some new
unmerge tests.

Also, add a selftest to measure MADV_UNMERGEABLE performance. In my setup
(AMD Ryzen 9 3900X), running the KSM selftest to test unmerge performance
on 2 GiB (taskset 0x8 ./ksm_tests -D -s 2048), this results in a
performance degradation of ~6% -- 7% (old: ~5250 MiB/s, new: ~4900 MiB/s).
I don't think we particularly care for now, but it's good to be aware of
the implication.


This patch (of 9):

Let's add three unmerge tests (MADV_UNMERGEABLE unmerging all pages in the
range).

test_unmerge(): basic unmerge tests
test_unmerge_discarded(): have some pte_none() entries in the range
test_unmerge_uffd_wp(): protect the merged pages using uffd-wp

ksm_tests.c currently contains a mixture of benchmarks and tests, whereby
each test is carried out by executing the ksm_tests binary with specific
parameters. Let's add new ksm_functional_tests.c that performs multiple,
smaller functional tests all at once.

Link: https://lkml.kernel.org/r/[e-mail]
Link: https://lkml.kernel.org/r/[e-mail]
Signed-off-by: David Hildenbrand <[e-mail]>
Cc: Andrea Arcangeli <[e-mail]>
Cc: Hugh Dickins <[e-mail]>
Cc: Jason Gunthorpe <[e-mail]>
Cc: John Hubbard <[e-mail]>
Cc: Matthew Wilcox (Oracle) <[e-mail]>
Cc: Peter Xu <[e-mail]>
Cc: Shuah Khan <[e-mail]>
Cc: Vlastimil Babka <[e-mail]>
Signed-off-by: Andrew Morton <[e-mail]>

tools/testing/selftests/vm/Makefile

···
 TEST_GEN_PROGS += soft-dirty
 TEST_GEN_PROGS += split_huge_page_test
 TEST_GEN_FILES += ksm_tests
+TEST_GEN_PROGS += ksm_functional_tests
 
 ifeq ($(MACHINE),x86_64)
 CAN_BUILD_I386 := $(shell ./../x86/check_cc.sh "$(CC)" ../x86/trivial_32bit_program.c -m32)
···
 
 $(OUTPUT)/cow: vm_util.c
 $(OUTPUT)/khugepaged: vm_util.c
+$(OUTPUT)/ksm_functional_tests: vm_util.c
 $(OUTPUT)/madv_populate: vm_util.c
 $(OUTPUT)/soft-dirty: vm_util.c
 $(OUTPUT)/split_huge_page_test: vm_util.c

tools/testing/selftests/vm/ksm_functional_tests.c

···
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * KSM functional tests
+ *
+ * Copyright 2022, Red Hat, Inc.
+ *
+ * Author(s): David Hildenbrand <[e-mail]>
+ */
+#define _GNU_SOURCE
+#include <stdlib.h>
+#include <string.h>
+#include <stdbool.h>
+#include <stdint.h>
+#include <unistd.h>
+#include <errno.h>
+#include <fcntl.h>
+#include <sys/mman.h>
+#include <sys/syscall.h>
+#include <sys/ioctl.h>
+#include <linux/userfaultfd.h>
+
+#include "../kselftest.h"
+#include "vm_util.h"
+
+#define KiB 1024u
+#define MiB (1024 * KiB)
+
+static int ksm_fd;
+static int ksm_full_scans_fd;
+static int pagemap_fd;
+static size_t pagesize;
+
+static bool range_maps_duplicates(char *addr, unsigned long size)
+{
+	unsigned long offs_a, offs_b, pfn_a, pfn_b;
+
+	/*
+	 * There is no easy way to check if there are KSM pages mapped into
+	 * this range. We only check that the range does not map the same PFN
+	 * twice by comaring each pair of mapped pages.
+	 */
+	for (offs_a = 0; offs_a < size; offs_a += pagesize) {
+		pfn_a = pagemap_get_pfn(pagemap_fd, addr + offs_a);
+		/* Page not present or PFN not exposed by the kernel. */
+		if (pfn_a == -1ull || !pfn_a)
+			continue;
+
+		for (offs_b = offs_a + pagesize; offs_b < size;
+		     offs_b += pagesize) {
+			pfn_b = pagemap_get_pfn(pagemap_fd, addr + offs_b);
+			if (pfn_b == -1ull || !pfn_b)
+				continue;
+			if (pfn_a == pfn_b)
+				return true;
+		}
+	}
+	return false;
+}
+
+static long ksm_get_full_scans(void)
+{
+	char buf[10];
+	ssize_t ret;
+
+	ret = pread(ksm_full_scans_fd, buf, sizeof(buf) - 1, 0);
+	if (ret <= 0)
+		return -errno;
+	buf[ret] = 0;
+
+	return strtol(buf, NULL, 10);
+}
+
+static int ksm_merge(void)
+{
+	long start_scans, end_scans;
+
+	/* Wait for two full scans such that any possible merging happened. */
+	start_scans = ksm_get_full_scans();
+	if (start_scans < 0)
+		return start_scans;
+	if (write(ksm_fd, "1", 1) != 1)
+		return -errno;
+	do {
+		end_scans = ksm_get_full_scans();
+		if (end_scans < 0)
+			return end_scans;
+	} while (end_scans < start_scans + 2);
+
+	return 0;
+}
+
+static char *mmap_and_merge_range(char val, unsigned long size)
+{
+	char *map;
+
+	map = mmap(NULL, size, PROT_READ|PROT_WRITE,
+		   MAP_PRIVATE|MAP_ANON, -1, 0);
+	if (map == MAP_FAILED) {
+		ksft_test_result_fail("mmap() failed\n");
+		return MAP_FAILED;
+	}
+
+	/* Don't use THP. Ignore if THP are not around on a kernel. */
+	if (madvise(map, size, MADV_NOHUGEPAGE) && errno != EINVAL) {
+		ksft_test_result_fail("MADV_NOHUGEPAGE failed\n");
+		goto unmap;
+	}
+
+	/* Make sure each page contains the same values to merge them. */
+	memset(map, val, size);
+	if (madvise(map, size, MADV_MERGEABLE)) {
+		ksft_test_result_fail("MADV_MERGEABLE failed\n");
+		goto unmap;
+	}
+
+	/* Run KSM to trigger merging and wait. */
+	if (ksm_merge()) {
+		ksft_test_result_fail("Running KSM failed\n");
+		goto unmap;
+	}
+	return map;
+unmap:
+	munmap(map, size);
+	return MAP_FAILED;
+}
+
+static void test_unmerge(void)
+{
+	const unsigned int size = 2 * MiB;
+	char *map;
+
+	ksft_print_msg("[RUN] %s\n", __func__);
+
+	map = mmap_and_merge_range(0xcf, size);
+	if (map == MAP_FAILED)
+		return;
+
+	if (madvise(map, size, MADV_UNMERGEABLE)) {
+		ksft_test_result_fail("MADV_UNMERGEABLE failed\n");
+		goto unmap;
+	}
+
+	ksft_test_result(!range_maps_duplicates(map, size),
+			 "Pages were unmerged\n");
+unmap:
+	munmap(map, size);
+}
+
+static void test_unmerge_discarded(void)
+{
+	const unsigned int size = 2 * MiB;
+	char *map;
+
+	ksft_print_msg("[RUN] %s\n", __func__);
+
+	map = mmap_and_merge_range(0xcf, size);
+	if (map == MAP_FAILED)
+		return;
+
+	/* Discard half of all mapped pages so we have pte_none() entries. */
+	if (madvise(map, size / 2, MADV_DONTNEED)) {
+		ksft_test_result_fail("MADV_DONTNEED failed\n");
+		goto unmap;
+	}
+
+	if (madvise(map, size, MADV_UNMERGEABLE)) {
+		ksft_test_result_fail("MADV_UNMERGEABLE failed\n");
+		goto unmap;
+	}
+
+	ksft_test_result(!range_maps_duplicates(map, size),
+			 "Pages were unmerged\n");
+unmap:
+	munmap(map, size);
+}
+
+#ifdef __NR_userfaultfd
+static void test_unmerge_uffd_wp(void)
+{
+	struct uffdio_writeprotect uffd_writeprotect;
+	struct uffdio_register uffdio_register;
+	const unsigned int size = 2 * MiB;
+	struct uffdio_api uffdio_api;
+	char *map;
+	int uffd;
+
+	ksft_print_msg("[RUN] %s\n", __func__);
+
+	map = mmap_and_merge_range(0xcf, size);
+	if (map == MAP_FAILED)
+		return;
+
+	/* See if UFFD is around. */
+	uffd = syscall(__NR_userfaultfd, O_CLOEXEC | O_NONBLOCK);
+	if (uffd < 0) {
+		ksft_test_result_skip("__NR_userfaultfd failed\n");
+		goto unmap;
+	}
+
+	/* See if UFFD-WP is around. */
+	uffdio_api.api = UFFD_API;
+	uffdio_api.features = UFFD_FEATURE_PAGEFAULT_FLAG_WP;
+	if (ioctl(uffd, UFFDIO_API, &uffdio_api) < 0) {
+		ksft_test_result_fail("UFFDIO_API failed\n");
+		goto close_uffd;
+	}
+	if (!(uffdio_api.features & UFFD_FEATURE_PAGEFAULT_FLAG_WP)) {
+		ksft_test_result_skip("UFFD_FEATURE_PAGEFAULT_FLAG_WP not available\n");
+		goto close_uffd;
+	}
+
+	/* Register UFFD-WP, no need for an actual handler. */
+	uffdio_register.range.start = (unsigned long) map;
+	uffdio_register.range.len = size;
+	uffdio_register.mode = UFFDIO_REGISTER_MODE_WP;
+	if (ioctl(uffd, UFFDIO_REGISTER, &uffdio_register) < 0) {
+		ksft_test_result_fail("UFFDIO_REGISTER_MODE_WP failed\n");
+		goto close_uffd;
+	}
+
+	/* Write-protect the range using UFFD-WP. */
+	uffd_writeprotect.range.start = (unsigned long) map;
+	uffd_writeprotect.range.len = size;
+	uffd_writeprotect.mode = UFFDIO_WRITEPROTECT_MODE_WP;
+	if (ioctl(uffd, UFFDIO_WRITEPROTECT, &uffd_writeprotect)) {
+		ksft_test_result_fail("UFFDIO_WRITEPROTECT failed\n");
+		goto close_uffd;
+	}
+
+	if (madvise(map, size, MADV_UNMERGEABLE)) {
+		ksft_test_result_fail("MADV_UNMERGEABLE failed\n");
+		goto close_uffd;
+	}
+
+	ksft_test_result(!range_maps_duplicates(map, size),
+			 "Pages were unmerged\n");
+close_uffd:
+	close(uffd);
+unmap:
+	munmap(map, size);
+}
+#endif
+
+int main(int argc, char **argv)
+{
+	unsigned int tests = 2;
+	int err;
+
+#ifdef __NR_userfaultfd
+	tests++;
+#endif
+
+	ksft_print_header();
+	ksft_set_plan(tests);
+
+	pagesize = getpagesize();
+
+	ksm_fd = open("/sys/kernel/mm/ksm/run", O_RDWR);
+	if (ksm_fd < 0)
+		ksft_exit_skip("open(\"/sys/kernel/mm/ksm/run\") failed\n");
+	ksm_full_scans_fd = open("/sys/kernel/mm/ksm/full_scans", O_RDONLY);
+	if (ksm_full_scans_fd < 0)
+		ksft_exit_skip("open(\"/sys/kernel/mm/ksm/full_scans\") failed\n");
+	pagemap_fd = open("/proc/self/pagemap", O_RDONLY);
+	if (pagemap_fd < 0)
+		ksft_exit_skip("open(\"/proc/self/pagemap\") failed\n");
+
+	test_unmerge();
+	test_unmerge_discarded();
+#ifdef __NR_userfaultfd
+	test_unmerge_uffd_wp();
+#endif
+
+	err = ksft_get_fail_cnt();
+	if (err)
+		ksft_exit_fail_msg("%d out of %d tests failed\n",
+				   err, ksft_test_num());
+	return ksft_exit_pass();
+}

tools/testing/selftests/vm/run_vmtests.sh

···
 
 CATEGORY="ksm" run_test ./ksm_functional_tests
 
+run_test ./ksm_functional_tests
+
 # protection_keys tests
 if [ -x ./protection_keys_32 ]
 then

tools/testing/selftests/vm/vm_util.c

···
 	return entry & 0xc000000000000000ull;
 }
 
+unsigned long pagemap_get_pfn(int fd, char *start)
+{
+	uint64_t entry = pagemap_get_entry(fd, start);
+
+	/* If present (63th bit), PFN is at bit 0 -- 54. */
+	if (entry & 0x8000000000000000ull)
+		return entry & 0x007fffffffffffffull;
+	return -1ull;
+}
+
 void clear_softdirty(void)
 {
 	int ret;

tools/testing/selftests/vm/vm_util.h

···
 bool pagemap_is_softdirty(int fd, char *start);
 bool pagemap_is_swapped(int fd, char *start);
 bool pagemap_is_populated(int fd, char *start);
+unsigned long pagemap_get_pfn(int fd, char *start);
 void clear_softdirty(void);
 bool check_for_pattern(FILE *fp, const char *pattern, char *buf, size_t len);
 uint64_t read_pmd_pagesize(void);