tools/testing/selftests/mm/Makefile | 1 + .../selftests/mm/folio_split_race_test.c | 293 ++++++++++++++++++ tools/testing/selftests/mm/run_vmtests.sh | 2 + 3 files changed, 296 insertions(+) create mode 100644 tools/testing/selftests/mm/folio_split_race_test.c
The added folio_split_race_test is a modified C port of the race condition
test from [1]. The test creates shmem huge pages, where the main thread
punches holes in the shmem to cause folio_split() in the kernel and
a set of 16 threads reads the shmem to cause filemap_get_entry() in the
kernel. filemap_get_entry() reads the folio and xarray split by
folio_split() locklessly. The original test[2] is written in rust and uses
memfd (shmem backed). This C port uses shmem directly and use a single
process.
Note: the initial rust to C conversion is done by Cursor.
Link: https://lore.kernel.org/all/CAKNNEtw5_kZomhkugedKMPOG-sxs5Q5OLumWJdiWXv+C9Yct0w@mail.gmail.com/ [1]
Link: https://github.com/dfinity/thp-madv-remove-test [2]
Signed-off-by: Bas van Dijk <bas@dfinity.org>
Signed-off-by: Adam Bratschi-Kaye <adam.bratschikaye@dfinity.org>
Signed-off-by: Zi Yan <ziy@nvidia.com>
---
From V3:
1. fixed for loop stepping issue
2. used PRIu64 instead of %zu for uint64_t.
From V2:
1. simplied the program by removing fork.
From V1:
1. added prctl(PR_SET_PDEATHSIG, SIGTERM) to avoid child looping
forever.
2. removed page_idx % PUNCH_INTERVAL >= 0, since it is a nop. Added a
comment.
3. added a child process status check to prevent parent looping forever
and record that as a failure.
4. used ksft_exit_skip() instead of ksft_finished() when the program is
not running as root.
5. restored THP settings properly when the program exits abnormally.
tools/testing/selftests/mm/Makefile | 1 +
.../selftests/mm/folio_split_race_test.c | 293 ++++++++++++++++++
tools/testing/selftests/mm/run_vmtests.sh | 2 +
3 files changed, 296 insertions(+)
create mode 100644 tools/testing/selftests/mm/folio_split_race_test.c
diff --git a/tools/testing/selftests/mm/Makefile b/tools/testing/selftests/mm/Makefile
index 7a5de4e9bf520..cd24596cdd27e 100644
--- a/tools/testing/selftests/mm/Makefile
+++ b/tools/testing/selftests/mm/Makefile
@@ -105,6 +105,7 @@ TEST_GEN_FILES += droppable
TEST_GEN_FILES += guard-regions
TEST_GEN_FILES += merge
TEST_GEN_FILES += rmap
+TEST_GEN_FILES += folio_split_race_test
ifneq ($(ARCH),arm64)
TEST_GEN_FILES += soft-dirty
diff --git a/tools/testing/selftests/mm/folio_split_race_test.c b/tools/testing/selftests/mm/folio_split_race_test.c
new file mode 100644
index 0000000000000..c264cc625a7cb
--- /dev/null
+++ b/tools/testing/selftests/mm/folio_split_race_test.c
@@ -0,0 +1,293 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * The test creates shmem PMD huge pages, fills all pages with known patterns,
+ * then continuously verifies non-punched pages with 16 threads. Meanwhile, the
+ * main thread punches holes via MADV_REMOVE on the shmem.
+ *
+ * It tests the race condition between folio_split() and filemap_get_entry(),
+ * where the hole punches on shmem lead to folio_split() and reading the shmem
+ * lead to filemap_get_entry().
+ */
+
+#define _GNU_SOURCE
+#include <errno.h>
+#include <inttypes.h>
+#include <linux/mman.h>
+#include <pthread.h>
+#include <stdatomic.h>
+#include <stdbool.h>
+#include <stdint.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <sys/mman.h>
+#include <signal.h>
+#include <unistd.h>
+#include "vm_util.h"
+#include "kselftest.h"
+#include "thp_settings.h"
+
+uint64_t page_size;
+uint64_t pmd_pagesize;
+#define NR_PMD_PAGE 5
+#define FILE_SIZE (pmd_pagesize * NR_PMD_PAGE)
+#define TOTAL_PAGES (FILE_SIZE / page_size)
+
+/* Every N-th to N+M-th pages are punched; not aligned with huge page boundaries. */
+#define PUNCH_INTERVAL 50 /* N */
+#define PUNCH_SIZE_FACTOR 3 /* M */
+
+#define NUM_READER_THREADS 16
+#define FILL_BYTE 0xAF
+#define NUM_ITERATIONS 100
+
+/* Shared control block: control reading threads and record stats */
+struct shared_ctl {
+ atomic_uint_fast32_t stop;
+ atomic_size_t reader_failures;
+ atomic_size_t reader_verified;
+};
+
+static void fill_page(unsigned char *base, size_t page_idx)
+{
+ unsigned char *page_ptr = base + page_idx * page_size;
+ uint64_t idx = (uint64_t)page_idx;
+
+ memset(page_ptr, FILL_BYTE, page_size);
+ memcpy(page_ptr, &idx, sizeof(idx));
+}
+
+/* Returns true if valid, false if corrupted. */
+static bool check_page(unsigned char *base, size_t page_idx)
+{
+ unsigned char *page_ptr = base + page_idx * page_size;
+ uint64_t expected_idx = (uint64_t)page_idx;
+ uint64_t got_idx;
+
+ memcpy(&got_idx, page_ptr, 8);
+
+ if (got_idx != expected_idx) {
+ size_t off;
+ int all_zero = 1;
+
+ for (off = 0; off < page_size; off++) {
+ if (page_ptr[off] != 0) {
+ all_zero = 0;
+ break;
+ }
+ }
+ if (all_zero) {
+ ksft_print_msg(
+ "CORRUPTED: page %zu (huge page %zu) is ALL ZEROS\n",
+ page_idx,
+ (page_idx * page_size) / pmd_pagesize);
+ } else {
+ ksft_print_msg(
+ "CORRUPTED: page %zu (huge page %zu): expected idx %zu, got %lu\n",
+ page_idx, (page_idx * page_size) / pmd_pagesize,
+ page_idx, (unsigned long)got_idx);
+ }
+ return false;
+ }
+ return true;
+}
+
+struct reader_arg {
+ unsigned char *base;
+ struct shared_ctl *ctl;
+ int tid;
+ atomic_size_t *failures;
+ atomic_size_t *verified;
+};
+
+static void *reader_thread(void *arg)
+{
+ struct reader_arg *ra = (struct reader_arg *)arg;
+ unsigned char *base = ra->base;
+ struct shared_ctl *ctl = ra->ctl;
+ int tid = ra->tid;
+ atomic_size_t *failures = ra->failures;
+ atomic_size_t *verified = ra->verified;
+ size_t page_idx;
+
+ while (atomic_load_explicit(&ctl->stop, memory_order_acquire) == 0) {
+ for (page_idx = (size_t)tid; page_idx < TOTAL_PAGES;
+ page_idx += NUM_READER_THREADS) {
+ /*
+ * page_idx % PUNCH_INTERVAL is in [0, PUNCH_INTERVAL),
+ * skip [0, PUNCH_SIZE_FACTOR)
+ */
+ if (page_idx % PUNCH_INTERVAL < PUNCH_SIZE_FACTOR)
+ continue;
+ if (check_page(base, page_idx))
+ atomic_fetch_add_explicit(verified, 1,
+ memory_order_relaxed);
+ else
+ atomic_fetch_add_explicit(failures, 1,
+ memory_order_relaxed);
+ }
+ if (atomic_load_explicit(failures, memory_order_relaxed) > 0)
+ break;
+ }
+
+ return NULL;
+}
+
+static void create_readers(pthread_t *threads, struct reader_arg *args,
+ unsigned char *base, struct shared_ctl *ctl)
+{
+ int i;
+
+ for (i = 0; i < NUM_READER_THREADS; i++) {
+ args[i].base = base;
+ args[i].ctl = ctl;
+ args[i].tid = i;
+ args[i].failures = &ctl->reader_failures;
+ args[i].verified = &ctl->reader_verified;
+ if (pthread_create(&threads[i], NULL, reader_thread,
+ &args[i]) != 0)
+ ksft_exit_fail_msg("pthread_create failed\n");
+ }
+}
+
+/* Run a single iteration. Returns total number of corrupted pages. */
+static size_t run_iteration(void)
+{
+ size_t reader_failures, reader_verified;
+ struct reader_arg args[NUM_READER_THREADS];
+ pthread_t threads[NUM_READER_THREADS];
+ unsigned char *mmap_base;
+ struct shared_ctl ctl;
+ size_t i;
+
+ memset(&ctl, 0, sizeof(struct shared_ctl));
+
+ mmap_base = mmap(NULL, FILE_SIZE, PROT_READ | PROT_WRITE,
+ MAP_SHARED | MAP_ANONYMOUS, -1, 0);
+
+ if (mmap_base == MAP_FAILED)
+ ksft_exit_fail_msg("mmap failed: %d\n", errno);
+
+ if (madvise(mmap_base, FILE_SIZE, MADV_HUGEPAGE) != 0)
+ ksft_exit_fail_msg("madvise(MADV_HUGEPAGE) failed: %d\n",
+ errno);
+
+ for (i = 0; i < TOTAL_PAGES; i++)
+ fill_page(mmap_base, i);
+
+ if (!check_huge_shmem(mmap_base, NR_PMD_PAGE, pmd_pagesize))
+ ksft_exit_fail_msg("No shmem THP is allocated\n");
+
+ create_readers(threads, args, mmap_base, &ctl);
+
+ for (i = 0; i < TOTAL_PAGES; i++) {
+ if (i % PUNCH_INTERVAL != 0)
+ continue;
+ if (madvise(mmap_base + i * page_size,
+ PUNCH_SIZE_FACTOR * page_size, MADV_REMOVE) != 0) {
+ ksft_exit_fail_msg(
+ "madvise(MADV_REMOVE) failed on page %zu: %d\n",
+ i, errno);
+ }
+
+ i += PUNCH_SIZE_FACTOR - 1;
+ }
+
+ atomic_store_explicit(&ctl.stop, 1, memory_order_release);
+
+ for (i = 0; i < NUM_READER_THREADS; i++)
+ pthread_join(threads[i], NULL);
+
+ reader_failures = atomic_load_explicit(&ctl.reader_failures,
+ memory_order_acquire);
+ reader_verified = atomic_load_explicit(&ctl.reader_verified,
+ memory_order_acquire);
+ if (reader_failures)
+ ksft_print_msg("Child: %zu pages verified, %zu failures\n",
+ reader_verified, reader_failures);
+
+ munmap(mmap_base, FILE_SIZE);
+
+ return reader_failures;
+}
+
+static void thp_cleanup_handler(int signum)
+{
+ thp_restore_settings();
+ /*
+ * Restore default handler and re-raise the signal to exit.
+ * This is to ensure the test process exits with the correct
+ * status code corresponding to the signal.
+ */
+ signal(signum, SIG_DFL);
+ raise(signum);
+}
+
+static void thp_settings_cleanup(void)
+{
+ thp_restore_settings();
+}
+
+int main(void)
+{
+ struct thp_settings current_settings;
+ bool failed = false;
+ size_t failures;
+ size_t iter;
+
+ ksft_print_header();
+
+ if (!thp_is_enabled())
+ ksft_exit_skip("Transparent Hugepages not available\n");
+
+ if (geteuid() != 0)
+ ksft_exit_skip("Please run the test as root\n");
+
+ thp_save_settings();
+ /* make sure thp settings are restored */
+ if (atexit(thp_settings_cleanup) != 0)
+ ksft_exit_fail_msg("atexit failed\n");
+
+ signal(SIGINT, thp_cleanup_handler);
+ signal(SIGTERM, thp_cleanup_handler);
+
+ thp_read_settings(¤t_settings);
+ current_settings.shmem_enabled = SHMEM_ADVISE;
+ thp_write_settings(¤t_settings);
+
+ ksft_set_plan(1);
+
+ page_size = getpagesize();
+ pmd_pagesize = read_pmd_pagesize();
+
+ ksft_print_msg("folio split race test\n");
+ ksft_print_msg("===================================================\n");
+ ksft_print_msg("Shmem size: %" PRIu64 " MiB\n", FILE_SIZE / 1024 / 1024);
+ ksft_print_msg("Total pages: %" PRIu64 "\n", TOTAL_PAGES);
+ ksft_print_msg("Child readers: %d\n", NUM_READER_THREADS);
+ ksft_print_msg("Punching every %dth to %dth page\n", PUNCH_INTERVAL,
+ PUNCH_INTERVAL + PUNCH_SIZE_FACTOR);
+ ksft_print_msg("Iterations: %d\n", NUM_ITERATIONS);
+
+ for (iter = 1; iter <= NUM_ITERATIONS; iter++) {
+ failures = run_iteration();
+ if (failures > 0) {
+ failed = true;
+ ksft_print_msg(
+ "FAILED on iteration %zu: %zu pages corrupted by MADV_REMOVE!\n",
+ iter, failures);
+ break;
+ }
+ }
+
+ if (failed) {
+ ksft_test_result_fail("Test failed\n");
+ ksft_exit_fail();
+ } else {
+ ksft_test_result_pass("All %d iterations passed\n",
+ NUM_ITERATIONS);
+ ksft_exit_pass();
+ }
+
+ return 0;
+}
diff --git a/tools/testing/selftests/mm/run_vmtests.sh b/tools/testing/selftests/mm/run_vmtests.sh
index 606558cc3b098..530980fdf3227 100755
--- a/tools/testing/selftests/mm/run_vmtests.sh
+++ b/tools/testing/selftests/mm/run_vmtests.sh
@@ -553,6 +553,8 @@ if [ -n "${MOUNTED_XFS}" ]; then
rm -f ${XFS_IMG}
fi
+CATEGORY="thp" run_test ./folio_split_race_test
+
CATEGORY="migration" run_test ./migration
CATEGORY="mkdirty" run_test ./mkdirty
--
2.51.0On 3/20/26 15:22, Zi Yan wrote:
> The added folio_split_race_test is a modified C port of the race condition
> test from [1]. The test creates shmem huge pages, where the main thread
> punches holes in the shmem to cause folio_split() in the kernel and
> a set of 16 threads reads the shmem to cause filemap_get_entry() in the
> kernel. filemap_get_entry() reads the folio and xarray split by
> folio_split() locklessly. The original test[2] is written in rust and uses
> memfd (shmem backed). This C port uses shmem directly and use a single
> process.
>
> Note: the initial rust to C conversion is done by Cursor.
>
> Link: https://lore.kernel.org/all/CAKNNEtw5_kZomhkugedKMPOG-sxs5Q5OLumWJdiWXv+C9Yct0w@mail.gmail.com/ [1]
> Link: https://github.com/dfinity/thp-madv-remove-test [2]
> Signed-off-by: Bas van Dijk <bas@dfinity.org>
> Signed-off-by: Adam Bratschi-Kaye <adam.bratschikaye@dfinity.org>
You are likely missing two Co-developed-by.
See Documentation/process/submitting-patches.rst on how to handle such
SOBs.
> Signed-off-by: Zi Yan <ziy@nvidia.com>
> ---
> From V3:
> 1. fixed for loop stepping issue
> 2. used PRIu64 instead of %zu for uint64_t.
>
> From V2:
> 1. simplied the program by removing fork.
>
> From V1:
> 1. added prctl(PR_SET_PDEATHSIG, SIGTERM) to avoid child looping
> forever.
> 2. removed page_idx % PUNCH_INTERVAL >= 0, since it is a nop. Added a
> comment.
> 3. added a child process status check to prevent parent looping forever
> and record that as a failure.
> 4. used ksft_exit_skip() instead of ksft_finished() when the program is
> not running as root.
> 5. restored THP settings properly when the program exits abnormally.
> tools/testing/selftests/mm/Makefile | 1 +
> .../selftests/mm/folio_split_race_test.c | 293 ++++++++++++++++++
> tools/testing/selftests/mm/run_vmtests.sh | 2 +
> 3 files changed, 296 insertions(+)
> create mode 100644 tools/testing/selftests/mm/folio_split_race_test.c
>
> diff --git a/tools/testing/selftests/mm/Makefile b/tools/testing/selftests/mm/Makefile
> index 7a5de4e9bf520..cd24596cdd27e 100644
> --- a/tools/testing/selftests/mm/Makefile
> +++ b/tools/testing/selftests/mm/Makefile
> @@ -105,6 +105,7 @@ TEST_GEN_FILES += droppable
> TEST_GEN_FILES += guard-regions
> TEST_GEN_FILES += merge
> TEST_GEN_FILES += rmap
> +TEST_GEN_FILES += folio_split_race_test
>
> ifneq ($(ARCH),arm64)
> TEST_GEN_FILES += soft-dirty
> diff --git a/tools/testing/selftests/mm/folio_split_race_test.c b/tools/testing/selftests/mm/folio_split_race_test.c
> new file mode 100644
> index 0000000000000..c264cc625a7cb
> --- /dev/null
> +++ b/tools/testing/selftests/mm/folio_split_race_test.c
> @@ -0,0 +1,293 @@
> +// SPDX-License-Identifier: GPL-2.0
> +/*
> + * The test creates shmem PMD huge pages, fills all pages with known patterns,
> + * then continuously verifies non-punched pages with 16 threads. Meanwhile, the
> + * main thread punches holes via MADV_REMOVE on the shmem.
> + *
> + * It tests the race condition between folio_split() and filemap_get_entry(),
> + * where the hole punches on shmem lead to folio_split() and reading the shmem
> + * lead to filemap_get_entry().
> + */
> +
> +#define _GNU_SOURCE
> +#include <errno.h>
> +#include <inttypes.h>
> +#include <linux/mman.h>
> +#include <pthread.h>
> +#include <stdatomic.h>
> +#include <stdbool.h>
> +#include <stdint.h>
> +#include <stdio.h>
> +#include <stdlib.h>
> +#include <string.h>
> +#include <sys/mman.h>
> +#include <signal.h>
> +#include <unistd.h>
> +#include "vm_util.h"
> +#include "kselftest.h"
> +#include "thp_settings.h"
[...] some comment son the main() part :)
> +int main(void)
> +{
> + struct thp_settings current_settings;
> + bool failed = false;
> + size_t failures;
> + size_t iter;
Why are iterations a "size_t" ? Similarly for "failures". Just use int / unsigned long?
> +
> + ksft_print_header();
> +
> + if (!thp_is_enabled())
> + ksft_exit_skip("Transparent Hugepages not available\n");
Is checking thp_is_enabled() the right thing to do when you perform your own setup below either way?
I think you should just use thp_available(). Then, configure THP accordingly below?
> +
> + if (geteuid() != 0)
> + ksft_exit_skip("Please run the test as root\n");
> +
> + thp_save_settings();
> + /* make sure thp settings are restored */
> + if (atexit(thp_settings_cleanup) != 0)
> + ksft_exit_fail_msg("atexit failed\n");
> +
> + signal(SIGINT, thp_cleanup_handler);
> + signal(SIGTERM, thp_cleanup_handler);
> +
> + thp_read_settings(¤t_settings);
> + current_settings.shmem_enabled = SHMEM_ADVISE;
> + thp_write_settings(¤t_settings);
> +
> + ksft_set_plan(1);
> +
> + page_size = getpagesize();
> + pmd_pagesize = read_pmd_pagesize();
I wonder whether we should check for 0 here and skip the test (older kernels?).
> +
> + ksft_print_msg("folio split race test\n");
> + ksft_print_msg("===================================================\n");
> + ksft_print_msg("Shmem size: %" PRIu64 " MiB\n", FILE_SIZE / 1024 / 1024);
> + ksft_print_msg("Total pages: %" PRIu64 "\n", TOTAL_PAGES);
> + ksft_print_msg("Child readers: %d\n", NUM_READER_THREADS);
> + ksft_print_msg("Punching every %dth to %dth page\n", PUNCH_INTERVAL,
> + PUNCH_INTERVAL + PUNCH_SIZE_FACTOR);
> + ksft_print_msg("Iterations: %d\n", NUM_ITERATIONS);
I don't think printing static test information is that helpful.
Do we need all that at all?
> +
> + for (iter = 1; iter <= NUM_ITERATIONS; iter++) {
Why not start at 0? You know, to confuse less people :)
for (iter = 0; iter < NUM_ITERATIONS; iter++) {
> + failures = run_iteration();
"corrupted_pages" ?
> + if (failures > 0) {
> + failed = true;
Do you really need that variable?
> + ksft_print_msg(
> + "FAILED on iteration %zu: %zu pages corrupted by MADV_REMOVE!\n",
> + iter, failures);
Can that simply be printed below?
Like
if (iter < NUM_ITERATIONS) {
ksft_test_result_fail("Test failed on iterations %zu: %zu pages ...\n",
iter + 1, corrupted_pages);
} else {
ksft_test_result_pass ...
}
> + break;
> + }
> + }
> +
> + if (failed) {
> + ksft_test_result_fail("Test failed\n");
> + ksft_exit_fail();
> + } else {
> + ksft_test_result_pass("All %d iterations passed\n",
> + NUM_ITERATIONS);
> + ksft_exit_pass();
> + }
> +
> + return 0;
--
Cheers,
David
On 23 Mar 2026, at 8:48, David Hildenbrand (Arm) wrote:
> On 3/20/26 15:22, Zi Yan wrote:
>> The added folio_split_race_test is a modified C port of the race condition
>> test from [1]. The test creates shmem huge pages, where the main thread
>> punches holes in the shmem to cause folio_split() in the kernel and
>> a set of 16 threads reads the shmem to cause filemap_get_entry() in the
>> kernel. filemap_get_entry() reads the folio and xarray split by
>> folio_split() locklessly. The original test[2] is written in rust and uses
>> memfd (shmem backed). This C port uses shmem directly and use a single
>> process.
>>
>> Note: the initial rust to C conversion is done by Cursor.
>>
>> Link: https://lore.kernel.org/all/CAKNNEtw5_kZomhkugedKMPOG-sxs5Q5OLumWJdiWXv+C9Yct0w@mail.gmail.com/ [1]
>> Link: https://github.com/dfinity/thp-madv-remove-test [2]
>> Signed-off-by: Bas van Dijk <bas@dfinity.org>
>> Signed-off-by: Adam Bratschi-Kaye <adam.bratschikaye@dfinity.org>
>
> You are likely missing two Co-developed-by.
>
> See Documentation/process/submitting-patches.rst on how to handle such
> SOBs.
Will add them.
>
>> Signed-off-by: Zi Yan <ziy@nvidia.com>
>> ---
>> From V3:
>> 1. fixed for loop stepping issue
>> 2. used PRIu64 instead of %zu for uint64_t.
>>
>> From V2:
>> 1. simplied the program by removing fork.
>>
>> From V1:
>> 1. added prctl(PR_SET_PDEATHSIG, SIGTERM) to avoid child looping
>> forever.
>> 2. removed page_idx % PUNCH_INTERVAL >= 0, since it is a nop. Added a
>> comment.
>> 3. added a child process status check to prevent parent looping forever
>> and record that as a failure.
>> 4. used ksft_exit_skip() instead of ksft_finished() when the program is
>> not running as root.
>> 5. restored THP settings properly when the program exits abnormally.
>> tools/testing/selftests/mm/Makefile | 1 +
>> .../selftests/mm/folio_split_race_test.c | 293 ++++++++++++++++++
>> tools/testing/selftests/mm/run_vmtests.sh | 2 +
>> 3 files changed, 296 insertions(+)
>> create mode 100644 tools/testing/selftests/mm/folio_split_race_test.c
>>
>> diff --git a/tools/testing/selftests/mm/Makefile b/tools/testing/selftests/mm/Makefile
>> index 7a5de4e9bf520..cd24596cdd27e 100644
>> --- a/tools/testing/selftests/mm/Makefile
>> +++ b/tools/testing/selftests/mm/Makefile
>> @@ -105,6 +105,7 @@ TEST_GEN_FILES += droppable
>> TEST_GEN_FILES += guard-regions
>> TEST_GEN_FILES += merge
>> TEST_GEN_FILES += rmap
>> +TEST_GEN_FILES += folio_split_race_test
>>
>> ifneq ($(ARCH),arm64)
>> TEST_GEN_FILES += soft-dirty
>> diff --git a/tools/testing/selftests/mm/folio_split_race_test.c b/tools/testing/selftests/mm/folio_split_race_test.c
>> new file mode 100644
>> index 0000000000000..c264cc625a7cb
>> --- /dev/null
>> +++ b/tools/testing/selftests/mm/folio_split_race_test.c
>> @@ -0,0 +1,293 @@
>> +// SPDX-License-Identifier: GPL-2.0
>> +/*
>> + * The test creates shmem PMD huge pages, fills all pages with known patterns,
>> + * then continuously verifies non-punched pages with 16 threads. Meanwhile, the
>> + * main thread punches holes via MADV_REMOVE on the shmem.
>> + *
>> + * It tests the race condition between folio_split() and filemap_get_entry(),
>> + * where the hole punches on shmem lead to folio_split() and reading the shmem
>> + * lead to filemap_get_entry().
>> + */
>> +
>> +#define _GNU_SOURCE
>> +#include <errno.h>
>> +#include <inttypes.h>
>> +#include <linux/mman.h>
>> +#include <pthread.h>
>> +#include <stdatomic.h>
>> +#include <stdbool.h>
>> +#include <stdint.h>
>> +#include <stdio.h>
>> +#include <stdlib.h>
>> +#include <string.h>
>> +#include <sys/mman.h>
>> +#include <signal.h>
>> +#include <unistd.h>
>> +#include "vm_util.h"
>> +#include "kselftest.h"
>> +#include "thp_settings.h"
>
>
> [...] some comment son the main() part :)
>
>> +int main(void)
>> +{
>> + struct thp_settings current_settings;
>> + bool failed = false;
>> + size_t failures;
>> + size_t iter;
>
> Why are iterations a "size_t" ? Similarly for "failures". Just use int / unsigned long?
Will change them.
>
>> +
>> + ksft_print_header();
>> +
>> + if (!thp_is_enabled())
>> + ksft_exit_skip("Transparent Hugepages not available\n");
>
> Is checking thp_is_enabled() the right thing to do when you perform your own setup below either way?
>
> I think you should just use thp_available(). Then, configure THP accordingly below?
OK, will use thp_available().
>
>> +
>> + if (geteuid() != 0)
>> + ksft_exit_skip("Please run the test as root\n");
>> +
>> + thp_save_settings();
>> + /* make sure thp settings are restored */
>> + if (atexit(thp_settings_cleanup) != 0)
>> + ksft_exit_fail_msg("atexit failed\n");
>> +
>> + signal(SIGINT, thp_cleanup_handler);
>> + signal(SIGTERM, thp_cleanup_handler);
>> +
>> + thp_read_settings(¤t_settings);
>> + current_settings.shmem_enabled = SHMEM_ADVISE;
>> + thp_write_settings(¤t_settings);
>> +
>> + ksft_set_plan(1);
>> +
>> + page_size = getpagesize();
>> + pmd_pagesize = read_pmd_pagesize();
>
> I wonder whether we should check for 0 here and skip the test (older kernels?).
OK, will take care of that.
>
>> +
>> + ksft_print_msg("folio split race test\n");
>> + ksft_print_msg("===================================================\n");
>> + ksft_print_msg("Shmem size: %" PRIu64 " MiB\n", FILE_SIZE / 1024 / 1024);
>> + ksft_print_msg("Total pages: %" PRIu64 "\n", TOTAL_PAGES);
>> + ksft_print_msg("Child readers: %d\n", NUM_READER_THREADS);
>> + ksft_print_msg("Punching every %dth to %dth page\n", PUNCH_INTERVAL,
>> + PUNCH_INTERVAL + PUNCH_SIZE_FACTOR);
>> + ksft_print_msg("Iterations: %d\n", NUM_ITERATIONS);
>
>
> I don't think printing static test information is that helpful.
> Do we need all that at all?
To provide some information on what this test is doing? I am OK with
removing them, but one will need to check the source code to get an idea.
>
>
>> +
>> + for (iter = 1; iter <= NUM_ITERATIONS; iter++) {
>
> Why not start at 0? You know, to confuse less people :)
>
> for (iter = 0; iter < NUM_ITERATIONS; iter++) {
Will change it.
>
>> + failures = run_iteration();
>
> "corrupted_pages" ?
>
>> + if (failures > 0) {
>> + failed = true;
>
> Do you really need that variable?
>
>> + ksft_print_msg(
>> + "FAILED on iteration %zu: %zu pages corrupted by MADV_REMOVE!\n",
>> + iter, failures);
>
> Can that simply be printed below?
>
> Like
>
> if (iter < NUM_ITERATIONS) {
> ksft_test_result_fail("Test failed on iterations %zu: %zu pages ...\n",
> iter + 1, corrupted_pages);
> } else {
> ksft_test_result_pass ...
> }
>
Sure, will simplify it. Thanks for the feedback.
>> + break;
>> + }
>> + }
>> +
>> + if (failed) {
>> + ksft_test_result_fail("Test failed\n");
>> + ksft_exit_fail();
>> + } else {
>> + ksft_test_result_pass("All %d iterations passed\n",
>> + NUM_ITERATIONS);
>> + ksft_exit_pass();
>> + }
>> +
>> + return 0;
> --
> Cheers,
>
> David
Best Regards,
Yan, Zi
On 20 Mar 2026, at 10:22, Zi Yan wrote:
> The added folio_split_race_test is a modified C port of the race condition
> test from [1]. The test creates shmem huge pages, where the main thread
> punches holes in the shmem to cause folio_split() in the kernel and
> a set of 16 threads reads the shmem to cause filemap_get_entry() in the
> kernel. filemap_get_entry() reads the folio and xarray split by
> folio_split() locklessly. The original test[2] is written in rust and uses
> memfd (shmem backed). This C port uses shmem directly and use a single
> process.
>
> Note: the initial rust to C conversion is done by Cursor.
>
> Link: https://lore.kernel.org/all/CAKNNEtw5_kZomhkugedKMPOG-sxs5Q5OLumWJdiWXv+C9Yct0w@mail.gmail.com/ [1]
> Link: https://github.com/dfinity/thp-madv-remove-test [2]
> Signed-off-by: Bas van Dijk <bas@dfinity.org>
> Signed-off-by: Adam Bratschi-Kaye <adam.bratschikaye@dfinity.org>
> Signed-off-by: Zi Yan <ziy@nvidia.com>
> ---
> From V3:
> 1. fixed for loop stepping issue
> 2. used PRIu64 instead of %zu for uint64_t.
>
> From V2:
> 1. simplied the program by removing fork.
>
> From V1:
> 1. added prctl(PR_SET_PDEATHSIG, SIGTERM) to avoid child looping
> forever.
> 2. removed page_idx % PUNCH_INTERVAL >= 0, since it is a nop. Added a
> comment.
> 3. added a child process status check to prevent parent looping forever
> and record that as a failure.
> 4. used ksft_exit_skip() instead of ksft_finished() when the program is
> not running as root.
> 5. restored THP settings properly when the program exits abnormally.
> tools/testing/selftests/mm/Makefile | 1 +
> .../selftests/mm/folio_split_race_test.c | 293 ++++++++++++++++++
> tools/testing/selftests/mm/run_vmtests.sh | 2 +
> 3 files changed, 296 insertions(+)
> create mode 100644 tools/testing/selftests/mm/folio_split_race_test.c
>
Hi Andrew,
The fixup below addressed the new issues (first and third) raised by sashiko[1].
The second issue is that the test only verifies first 8 bytes. Because
the test is intended to verify the race condition causing a wrong page index
is used.
The fourth issue is addressed in Q3 from V3[2]
[1] https://sashiko.dev/#/patchset/20260320142219.375118-1-ziy%40nvidia.com
[2] https://lore.kernel.org/all/8B720FB8-DE4D-487A-9AEF-AC204E9F5755@nvidia.com/
From a66945de00f33c163cf814ac7c2d9620a725bfed Mon Sep 17 00:00:00 2001
From: Zi Yan <ziy@nvidia.com>
Date: Sun, 22 Mar 2026 19:53:54 -0400
Subject: [PATCH] selftests/mm: fix sashiko complains on folio_split_race_test
1. used PRIu64 for uint64_t
2. added pthread_barrier_t to ensure main thread starts to punch holes when
all reader threads are spawned.
Signed-off-by: Zi Yan <ziy@nvidia.com>
---
.../selftests/mm/folio_split_race_test.c | 17 ++++++++++++++---
1 file changed, 14 insertions(+), 3 deletions(-)
diff --git a/tools/testing/selftests/mm/folio_split_race_test.c b/tools/testing/selftests/mm/folio_split_race_test.c
index c264cc625a7cb..ab6868e7e2efe 100644
--- a/tools/testing/selftests/mm/folio_split_race_test.c
+++ b/tools/testing/selftests/mm/folio_split_race_test.c
@@ -46,6 +46,7 @@ struct shared_ctl {
atomic_uint_fast32_t stop;
atomic_size_t reader_failures;
atomic_size_t reader_verified;
+ pthread_barrier_t barrier;
};
static void fill_page(unsigned char *base, size_t page_idx)
@@ -78,14 +79,14 @@ static bool check_page(unsigned char *base, size_t page_idx)
}
if (all_zero) {
ksft_print_msg(
- "CORRUPTED: page %zu (huge page %zu) is ALL ZEROS\n",
+ "CORRUPTED: page %zu (huge page %" PRIu64 ") is ALL ZEROS\n",
page_idx,
(page_idx * page_size) / pmd_pagesize);
} else {
ksft_print_msg(
- "CORRUPTED: page %zu (huge page %zu): expected idx %zu, got %lu\n",
+ "CORRUPTED: page %zu (huge page %" PRIu64 "): expected idx %zu, got %" PRIu64 "\n",
page_idx, (page_idx * page_size) / pmd_pagesize,
- page_idx, (unsigned long)got_idx);
+ page_idx, got_idx);
}
return false;
}
@@ -110,6 +111,8 @@ static void *reader_thread(void *arg)
atomic_size_t *verified = ra->verified;
size_t page_idx;
+ pthread_barrier_wait(&ctl->barrier);
+
while (atomic_load_explicit(&ctl->stop, memory_order_acquire) == 0) {
for (page_idx = (size_t)tid; page_idx < TOTAL_PAGES;
page_idx += NUM_READER_THREADS) {
@@ -178,8 +181,14 @@ static size_t run_iteration(void)
if (!check_huge_shmem(mmap_base, NR_PMD_PAGE, pmd_pagesize))
ksft_exit_fail_msg("No shmem THP is allocated\n");
+ if (pthread_barrier_init(&ctl.barrier, NULL, NUM_READER_THREADS + 1) != 0)
+ ksft_exit_fail_msg("pthread_barrier_init failed\n");
+
create_readers(threads, args, mmap_base, &ctl);
+ /* Wait for all reader threads to be ready before punching holes. */
+ pthread_barrier_wait(&ctl.barrier);
+
for (i = 0; i < TOTAL_PAGES; i++) {
if (i % PUNCH_INTERVAL != 0)
continue;
@@ -198,6 +207,8 @@ static size_t run_iteration(void)
for (i = 0; i < NUM_READER_THREADS; i++)
pthread_join(threads[i], NULL);
+ pthread_barrier_destroy(&ctl.barrier);
+
reader_failures = atomic_load_explicit(&ctl.reader_failures,
memory_order_acquire);
reader_verified = atomic_load_explicit(&ctl.reader_verified,
--
2.53.0
--
Best Regards,
Yan, Zi
On 20 Mar 2026, at 10:22, Zi Yan wrote: > The added folio_split_race_test is a modified C port of the race condition > test from [1]. The test creates shmem huge pages, where the main thread > punches holes in the shmem to cause folio_split() in the kernel and > a set of 16 threads reads the shmem to cause filemap_get_entry() in the > kernel. filemap_get_entry() reads the folio and xarray split by > folio_split() locklessly. The original test[2] is written in rust and uses > memfd (shmem backed). This C port uses shmem directly and use a single > process. > > Note: the initial rust to C conversion is done by Cursor. > > Link: https://lore.kernel.org/all/CAKNNEtw5_kZomhkugedKMPOG-sxs5Q5OLumWJdiWXv+C9Yct0w@mail.gmail.com/ [1] > Link: https://github.com/dfinity/thp-madv-remove-test [2] > Signed-off-by: Bas van Dijk <bas@dfinity.org> > Signed-off-by: Adam Bratschi-Kaye <adam.bratschikaye@dfinity.org> > Signed-off-by: Zi Yan <ziy@nvidia.com> > --- > From V3: > 1. fixed for loop stepping issue > 2. used PRIu64 instead of %zu for uint64_t. > > From V2: > 1. simplied the program by removing fork. > > From V1: > 1. added prctl(PR_SET_PDEATHSIG, SIGTERM) to avoid child looping > forever. > 2. removed page_idx % PUNCH_INTERVAL >= 0, since it is a nop. Added a > comment. > 3. added a child process status check to prevent parent looping forever > and record that as a failure. > 4. used ksft_exit_skip() instead of ksft_finished() when the program is > not running as root. > 5. restored THP settings properly when the program exits abnormally. > tools/testing/selftests/mm/Makefile | 1 + > .../selftests/mm/folio_split_race_test.c | 293 ++++++++++++++++++ > tools/testing/selftests/mm/run_vmtests.sh | 2 + > 3 files changed, 296 insertions(+) > create mode 100644 tools/testing/selftests/mm/folio_split_race_test.c > Hi Andrew, The fixup below adds folio_split_race_test binary to .gitignore. Thank Lorenzo for pointing this out. From bd164a7090c0e8b6e3013502c64c161214e38714 Mon Sep 17 00:00:00 2001 From: Zi Yan <ziy@nvidia.com> Date: Fri, 20 Mar 2026 13:56:04 -0400 Subject: [PATCH] git: add generated binary into gitignore Signed-off-by: Zi Yan <ziy@nvidia.com> --- tools/testing/selftests/mm/.gitignore | 1 + 1 file changed, 1 insertion(+) diff --git a/tools/testing/selftests/mm/.gitignore b/tools/testing/selftests/mm/.gitignore index 83ad9454dd9d1..b0c30c5ee9e30 100644 --- a/tools/testing/selftests/mm/.gitignore +++ b/tools/testing/selftests/mm/.gitignore @@ -61,3 +61,4 @@ guard-regions merge prctl_thp_disable rmap +folio_split_race_test -- 2.51.0 Best Regards, Yan, Zi
© 2016 - 2026 Red Hat, Inc.