selftests/bpf: validate .bss section bigger than 8MB is possible now

Add a simple big 16MB array and validate access to the very last byte of
it to make sure that kernel supports > KMALLOC_MAX_SIZE value_size for
BPF array maps (which are backing .bss in this case).

Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/r/20220715053146.1291891-5-andrii@kernel.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>

tools/testing/selftests/bpf/prog_tests/skeleton.c

@@ -122,6 +122,8 @@ void test_skeleton(void)
 
 	ASSERT_EQ(skel->bss->out_mostly_var, 123, "out_mostly_var");
 
+	ASSERT_EQ(bss->huge_arr[ARRAY_SIZE(bss->huge_arr) - 1], 123, "huge_arr");
+
 	elf_bytes = test_skeleton__elf_bytes(&elf_bytes_sz);
 	ASSERT_OK_PTR(elf_bytes, "elf_bytes");
 	ASSERT_GE(elf_bytes_sz, 0, "elf_bytes_sz");

tools/testing/selftests/bpf/progs/test_skeleton.c

@@ -51,6 +51,8 @@ int out_dynarr[4] SEC(".data.dyn") = { 1, 2, 3, 4 };
 int read_mostly_var __read_mostly;
 int out_mostly_var;
 
+char huge_arr[16 * 1024 * 1024];
+
 SEC("raw_tp/sys_enter")
 int handler(const void *ctx)
 {
@@ -71,6 +73,8 @@ int handler(const void *ctx)
 
 	out_mostly_var = read_mostly_var;
 
+	huge_arr[sizeof(huge_arr) - 1] = 123;
+
 	return 0;
 }