bpf: Clear subreg_def for global function return values

[ Upstream commit 45159b2 ]

test_global_func4 fails on s390 as reported by Yauheni in [1].

The immediate problem is that the zext code includes the instruction,
whose result needs to be zero-extended, into the zero-extension
patchlet, and if this instruction happens to be a branch, then its
delta is not adjusted. As a result, the verifier rejects the program
later.

However, according to [2], as far as the verifier's algorithm is
concerned and as specified by the insn_no_def() function, branching
insns do not define anything. This includes call insns, even though
one might argue that they define %r0.

This means that the real problem is that zero extension kicks in at
all. This happens because clear_caller_saved_regs() sets BPF_REG_0's
subreg_def after global function calls. This can be fixed in many
ways; this patch mimics what helper function call handling already
does.

  [1] https://lore.kernel.org/bpf/20200903140542.156624-1-yauheni.kaliuta@redhat.com/
  [2] https://lore.kernel.org/bpf/CAADnVQ+2RPKcftZw8d+B1UwB35cpBhpF5u3OocNh90D9pETPwg@mail.gmail.com/

Fixes: 51c39bb ("bpf: Introduce function-by-function verification")
Reported-by: Yauheni Kaliuta <yauheni.kaliuta@redhat.com>
Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20210212040408.90109-1-iii@linux.ibm.com
Signed-off-by: Sasha Levin <sashal@kernel.org>

kernel/bpf/verifier.c

@@ -4786,8 +4786,9 @@ static int check_func_call(struct bpf_verifier_env *env, struct bpf_insn *insn,
 					subprog);
 			clear_caller_saved_regs(env, caller->regs);
 
-			/* All global functions return SCALAR_VALUE */
+			/* All global functions return a 64-bit SCALAR_VALUE */
 			mark_reg_unknown(env, caller->regs, BPF_REG_0);
+			caller->regs[BPF_REG_0].subreg_def = DEF_NOT_SUBREG;
 
 			/* continue with next insn after call */
 			return 0;