[merged mm-hotfixes-stable] tools-testing-selftests-fix-guard-region-test-tmpfs-assumption.patch removed from -mm tree

[Andrew Morton <akpm@linux-foundation.org>]

The quilt patch titled
     Subject: tools/testing/selftests: fix guard region test tmpfs assumption
has been removed from the -mm tree.  Its filename was
     tools-testing-selftests-fix-guard-region-test-tmpfs-assumption.patch

This patch was dropped because it was merged into the mm-hotfixes-stable branch
of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm

------------------------------------------------------
From: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Subject: tools/testing/selftests: fix guard region test tmpfs assumption
Date: Fri, 25 Apr 2025 17:24:36 +0100

The current implementation of the guard region tests assume that /tmp is
mounted as tmpfs, that is shmem.

This isn't always the case, and at least one instance of a spurious test
failure has been reported as a result.

This assumption is unsafe, rushed and silly - and easily remedied by
simply using memfd, so do so.

We also have to fixup the readonly_file test to explicitly only be
applicable to file-backed cases.

Link: https://lkml.kernel.org/r/20250425162436.564002-1-lorenzo.stoakes@oracle.com
Fixes: 272f37d3e99a ("tools/selftests: expand all guard region tests to file-backed")
Signed-off-by: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Reported-by: Ryan Roberts <ryan.roberts@arm.com>
Closes: https://lore.kernel.org/linux-mm/a2d2766b-0ab4-437b-951a-8595a7506fe9@arm.com/
Reviewed-by: Ryan Roberts <ryan.roberts@arm.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---

 tools/testing/selftests/mm/guard-regions.c |   16 ++++++++++------
 1 file changed, 10 insertions(+), 6 deletions(-)

--- a/tools/testing/selftests/mm/guard-regions.c~tools-testing-selftests-fix-guard-region-test-tmpfs-assumption
+++ a/tools/testing/selftests/mm/guard-regions.c
@@ -271,12 +271,16 @@ FIXTURE_SETUP(guard_regions)
 	self->page_size = (unsigned long)sysconf(_SC_PAGESIZE);
 	setup_sighandler();
 
-	if (variant->backing == ANON_BACKED)
+	switch (variant->backing) {
+	case ANON_BACKED:
 		return;
-
-	self->fd = open_file(
-		variant->backing == SHMEM_BACKED ? "/tmp/" : "",
-		self->path);
+	case LOCAL_FILE_BACKED:
+		self->fd = open_file("", self->path);
+		break;
+	case SHMEM_BACKED:
+		self->fd = memfd_create(self->path, 0);
+		break;
+	}
 
 	/* We truncate file to at least 100 pages, tests can modify as needed. */
 	ASSERT_EQ(ftruncate(self->fd, 100 * self->page_size), 0);
@@ -1696,7 +1700,7 @@ TEST_F(guard_regions, readonly_file)
 	char *ptr;
 	int i;
 
-	if (variant->backing == ANON_BACKED)
+	if (variant->backing != LOCAL_FILE_BACKED)
 		SKIP(return, "Read-only test specific to file-backed");
 
 	/* Map shared so we can populate with pattern, populate it, unmap. */
_

Patches currently in -mm which might be from lorenzo.stoakes@oracle.com are

maintainers-add-mm-gup-section.patch
mm-vma-fix-incorrectly-disallowed-anonymous-vma-merges.patch
tools-testing-add-procmap_query-helper-functions-in-mm-self-tests.patch
tools-testing-selftests-assert-that-anon-merge-cases-behave-as-expected.patch
mm-move-mmap-vma-locking-logic-into-specific-files.patch
mm-establish-mm-vma_execc-for-shared-exec-mm-vma-functionality.patch
mm-abstract-initial-stack-setup-to-mm-subsystem.patch
mm-move-dup_mmap-to-mm.patch
mm-perform-vma-allocation-freeing-duplication-in-mm.patch
mm-introduce-new-mmap_prepare-file-callback.patch
mm-secretmem-convert-to-mmap_prepare-hook.patch
mm-vma-remove-mmap-retry-merge.patch