From: Edward Cree <ecree@solarflare.com>
Date: Mon, 7 Aug 2017 15:29:11 +0100
Subject: selftests/bpf: don't try to access past MAX_PACKET_OFF in
 test_verifier
Patch-mainline: v4.14-rc1
Git-commit: 1f9ab38f8a155913c9a587a673e61eedb75c9bc8
References: bsc#1083647

A number of selftests fell foul of the changed MAX_PACKET_OFF handling.
For instance, "direct packet access: test2" was potentially reading four
 bytes from pkt + 0xffff, which could take it past the verifier's limit,
 causing the program to be rejected (checks against pkt_end didn't give
 us any reg->range).
Increase the shifts by one so that R2 is now mask 0x7fff instead of
 mask 0xffff.

NOTE from Gary Lin:
  This commit is only applied partially to one test case:
  "direct packet access: test2"

Signed-off-by: Edward Cree <ecree@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Acked-by: Gary Lin <glin@suse.com>
---
 tools/testing/selftests/bpf/test_verifier.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

--- a/tools/testing/selftests/bpf/test_verifier.c
+++ b/tools/testing/selftests/bpf/test_verifier.c
@@ -2331,8 +2331,8 @@ static struct bpf_test tests[] = {
 				    offsetof(struct __sk_buff, data)),
 			BPF_ALU64_REG(BPF_ADD, BPF_REG_3, BPF_REG_4),
 			BPF_MOV64_REG(BPF_REG_2, BPF_REG_1),
-			BPF_ALU64_IMM(BPF_LSH, BPF_REG_2, 48),
-			BPF_ALU64_IMM(BPF_RSH, BPF_REG_2, 48),
+			BPF_ALU64_IMM(BPF_LSH, BPF_REG_2, 49),
+			BPF_ALU64_IMM(BPF_RSH, BPF_REG_2, 49),
 			BPF_ALU64_REG(BPF_ADD, BPF_REG_3, BPF_REG_2),
 			BPF_MOV64_REG(BPF_REG_2, BPF_REG_3),
 			BPF_ALU64_IMM(BPF_ADD, BPF_REG_2, 8),