Re: [PATCH RFC 3/3] KVM: selftests: exercise guest_memfd folio migration

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] The test `test_migrate_folio` hardcodes NUMA nodes 0 and 1, leading to false-positive test failures on systems with sparse NUMA topologies or restricted cgroups.
--

commit 857b251fdce0e2a9ddc74b27904d9c410d3010ca
Author: Shivank Garg <shivankg@amd.com>

KVM: selftests: exercise guest_memfd folio migration

This commit adds a migration test to guest_memfd_test for the
MMAP | INIT_SHARED configuration on multi-NUMA systems. It migrates folios
between node 0 and node 1 using move_pages(2) and verifies data integrity.

> diff --git a/tools/testing/selftests/kvm/guest_memfd_test.c b/tools/testing/selftests/kvm/guest_memfd_test.c
> index 832ef4dfb99fa..04931d3add46c 100644
> --- a/tools/testing/selftests/kvm/guest_memfd_test.c
> +++ b/tools/testing/selftests/kvm/guest_memfd_test.c
[ ... ]
> @@ -76,6 +76,82 @@ static void test_mmap_supported(int fd, size_t total_size)
>  	kvm_munmap(mem, total_size);
>  }
>  
> +/*
> + * Each page is filled with a distinct byte (its index). Check every byte that
> + * data is intact after migration.
> + */
> +static void verify_page(const char *page, int page_idx, size_t size,
> +			const char *when)
> +{
[ ... ]
> +static void test_migrate_folio(int fd, size_t total_size)
> +{
> +	const unsigned long nodemask_0 = 1; /* nid: 0 */
> +	unsigned long maxnode = BITS_PER_TYPE(nodemask_0);
> +	int page_count = total_size / page_size;
> +	void **addr;
> +	int *status, *nodes;
> +	char *mem;
> +	int i;
> +
> +	if (!is_multi_numa_node_system())
> +		return;
[ ... ]
> +	kvm_move_pages(0, page_count, addr, NULL, status, 0);
> +	for (i = 0; i < page_count; i++)
> +		TEST_ASSERT(status[i] == 0, "Page %d should be on node 0", i);
> +
> +	/* Migrate node 0 -> 1, then check both the location and the data. */
> +	for (i = 0; i < page_count; i++)
> +		nodes[i] = 1;
> +	kvm_move_pages(0, page_count, addr, nodes, status, MPOL_MF_MOVE);

[Severity: Medium]
Can this hardcoded target node cause false-positive test failures on
systems with sparse NUMA topologies or restricted cpusets?

is_multi_numa_node_system() verifies that get_max_numa_node() >= 1, but
it does not guarantee that nodes 0 and 1 specifically are online, have
memory, or are accessible within the current cpuset (for example, a system
might only have nodes 0 and 8).

If node 1 is unavailable, kvm_move_pages() or kvm_mbind() will fail with
-ENODEV or -EINVAL, triggering a test assertion failure.

Should this test dynamically discover available NUMA nodes instead of
hardcoding nodes 0 and 1?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260611-shivank-gmem-migrate-v1-0-2d266bfc6f95@amd.com?part=3