[PATCH bpf-next v2 3/3] bpf/selftests: add bpf_get_task_stack retval bounds test_prog

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Add a libbpf test prog which feeds bpf_get_task_stack's return value
into seq_write after confirming it's positive. No attempt to bound the
value from above is made.

Load will fail if verifier does not refine retval range based on buf sz
input to bpf_get_task_stack.

Signed-off-by: Dave Marchevsky <davemarchevsky@xxxxxx>
Acked-by: Song Liu <songliubraving@xxxxxx>
---
 .../selftests/bpf/prog_tests/bpf_iter.c       |  1 +
 .../selftests/bpf/progs/bpf_iter_task_stack.c | 27 +++++++++++++++++++
 2 files changed, 28 insertions(+)

diff --git a/tools/testing/selftests/bpf/prog_tests/bpf_iter.c b/tools/testing/selftests/bpf/prog_tests/bpf_iter.c
index 74c45d557a2b..2d3590cfb5e1 100644
--- a/tools/testing/selftests/bpf/prog_tests/bpf_iter.c
+++ b/tools/testing/selftests/bpf/prog_tests/bpf_iter.c
@@ -147,6 +147,7 @@ static void test_task_stack(void)
 		return;
 
 	do_dummy_read(skel->progs.dump_task_stack);
+	do_dummy_read(skel->progs.get_task_user_stacks);
 
 	bpf_iter_task_stack__destroy(skel);
 }
diff --git a/tools/testing/selftests/bpf/progs/bpf_iter_task_stack.c b/tools/testing/selftests/bpf/progs/bpf_iter_task_stack.c
index 50e59a2e142e..43c36f5f7649 100644
--- a/tools/testing/selftests/bpf/progs/bpf_iter_task_stack.c
+++ b/tools/testing/selftests/bpf/progs/bpf_iter_task_stack.c
@@ -35,3 +35,30 @@ int dump_task_stack(struct bpf_iter__task *ctx)
 
 	return 0;
 }
+
+SEC("iter/task")
+int get_task_user_stacks(struct bpf_iter__task *ctx)
+{
+	struct seq_file *seq = ctx->meta->seq;
+	struct task_struct *task = ctx->task;
+	uint64_t buf_sz = 0;
+	int64_t res;
+
+	if (task == (void *)0)
+		return 0;
+
+	res = bpf_get_task_stack(task, entries,
+			MAX_STACK_TRACE_DEPTH * SIZE_OF_ULONG, BPF_F_USER_STACK);
+	if (res <= 0)
+		return 0;
+
+	buf_sz += res;
+
+	/* If the verifier doesn't refine bpf_get_task_stack res, and instead
+	 * assumes res is entirely unknown, this program will fail to load as
+	 * the verifier will believe that max buf_sz value allows reading
+	 * past the end of entries in bpf_seq_write call
+	 */
+	bpf_seq_write(seq, &entries, buf_sz);
+	return 0;
+}
-- 
2.30.2





[Index of Archives]     [Linux Samsung SoC]     [Linux Rockchip SoC]     [Linux Actions SoC]     [Linux for Synopsys ARC Processors]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]


  Powered by Linux