[PATCH AUTOSEL 6.1 05/42] bpf, x64: Fix a jit convergence issue

[PATCH AUTOSEL 6.1 05/42] bpf, x64: Fix a jit convergence issue

[Sasha Levin]

From: Yonghong Song <yonghong.song@linux.dev>

[ Upstream commit c8831bdbfbab672c006a18006d36932a494b2fd6 ]

Daniel Hodges reported a jit error when playing with a sched-ext program.
The error message is:
  unexpected jmp_cond padding: -4 bytes

But further investigation shows the error is actual due to failed
convergence. The following are some analysis:

  ...
  pass4, final_proglen=4391:
    ...
    20e:    48 85 ff                test   rdi,rdi
    211:    74 7d                   je     0x290
    213:    48 8b 77 00             mov    rsi,QWORD PTR [rdi+0x0]
    ...
    289:    48 85 ff                test   rdi,rdi
    28c:    74 17                   je     0x2a5
    28e:    e9 7f ff ff ff          jmp    0x212
    293:    bf 03 00 00 00          mov    edi,0x3

Note that insn at 0x211 is 2-byte cond jump insn for offset 0x7d (-125)
and insn at 0x28e is 5-byte jmp insn with offset -129.

  pass5, final_proglen=4392:
    ...
    20e:    48 85 ff                test   rdi,rdi
    211:    0f 84 80 00 00 00       je     0x297
    217:    48 8b 77 00             mov    rsi,QWORD PTR [rdi+0x0]
    ...
    28d:    48 85 ff                test   rdi,rdi
    290:    74 1a                   je     0x2ac
    292:    eb 84                   jmp    0x218
    294:    bf 03 00 00 00          mov    edi,0x3

Note that insn at 0x211 is 6-byte cond jump insn now since its offset
becomes 0x80 based on previous round (0x293 - 0x213 = 0x80). At the same
time, insn at 0x292 is a 2-byte insn since its offset is -124.

pass6 will repeat the same code as in pass4. pass7 will repeat the same
code as in pass5, and so on. This will prevent eventual convergence.

Passes 1-14 are with padding = 0. At pass15, padding is 1 and related
insn looks like:

    211:    0f 84 80 00 00 00       je     0x297
    217:    48 8b 77 00             mov    rsi,QWORD PTR [rdi+0x0]
    ...
    24d:    48 85 d2                test   rdx,rdx

The similar code in pass14:
    211:    74 7d                   je     0x290
    213:    48 8b 77 00             mov    rsi,QWORD PTR [rdi+0x0]
    ...
    249:    48 85 d2                test   rdx,rdx
    24c:    74 21                   je     0x26f
    24e:    48 01 f7                add    rdi,rsi
    ...

Before generating the following insn,
  250:    74 21                   je     0x273
"padding = 1" enables some checking to ensure nops is either 0 or 4
where
  #define INSN_SZ_DIFF (((addrs[i] - addrs[i - 1]) - (prog - temp)))
  nops = INSN_SZ_DIFF - 2

In this specific case,
  addrs[i] = 0x24e // from pass14
  addrs[i-1] = 0x24d // from pass15
  prog - temp = 3 // from 'test rdx,rdx' in pass15
so
  nops = -4
and this triggers the failure.

To fix the issue, we need to break cycles of je <-> jmp. For example,
in the above case, we have
  211:    74 7d                   je     0x290
the offset is 0x7d. If 2-byte je insn is generated only if
the offset is less than 0x7d (<= 0x7c), the cycle can be
break and we can achieve the convergence.

I did some study on other cases like je <-> je, jmp <-> je and
jmp <-> jmp which may cause cycles. Those cases are not from actual
reproducible cases since it is pretty hard to construct a test case
for them. the results show that the offset <= 0x7b (0x7b = 123) should
be enough to cover all cases. This patch added a new helper to generate 8-bit
cond/uncond jmp insns only if the offset range is [-128, 123].

Reported-by: Daniel Hodges <hodgesd@meta.com>
Signed-off-by: Yonghong Song <yonghong.song@linux.dev>
Link: https://lore.kernel.org/r/20240904221251.37109-1-yonghong.song@linux.dev
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 arch/x86/net/bpf_jit_comp.c | 54 +++++++++++++++++++++++++++++++++++--
 1 file changed, 52 insertions(+), 2 deletions(-)

diff --git a/arch/x86/net/bpf_jit_comp.c b/arch/x86/net/bpf_jit_comp.c
index 7913440c0fd46..92db785a0a8eb 100644
--- a/arch/x86/net/bpf_jit_comp.c
+++ b/arch/x86/net/bpf_jit_comp.c
@@ -57,6 +57,56 @@ static bool is_imm8(int value)
 	return value <= 127 && value >= -128;
 }
 
+/*
+ * Let us limit the positive offset to be <= 123.
+ * This is to ensure eventual jit convergence For the following patterns:
+ * ...
+ * pass4, final_proglen=4391:
+ *   ...
+ *   20e:    48 85 ff                test   rdi,rdi
+ *   211:    74 7d                   je     0x290
+ *   213:    48 8b 77 00             mov    rsi,QWORD PTR [rdi+0x0]
+ *   ...
+ *   289:    48 85 ff                test   rdi,rdi
+ *   28c:    74 17                   je     0x2a5
+ *   28e:    e9 7f ff ff ff          jmp    0x212
+ *   293:    bf 03 00 00 00          mov    edi,0x3
+ * Note that insn at 0x211 is 2-byte cond jump insn for offset 0x7d (-125)
+ * and insn at 0x28e is 5-byte jmp insn with offset -129.
+ *
+ * pass5, final_proglen=4392:
+ *   ...
+ *   20e:    48 85 ff                test   rdi,rdi
+ *   211:    0f 84 80 00 00 00       je     0x297
+ *   217:    48 8b 77 00             mov    rsi,QWORD PTR [rdi+0x0]
+ *   ...
+ *   28d:    48 85 ff                test   rdi,rdi
+ *   290:    74 1a                   je     0x2ac
+ *   292:    eb 84                   jmp    0x218
+ *   294:    bf 03 00 00 00          mov    edi,0x3
+ * Note that insn at 0x211 is 6-byte cond jump insn now since its offset
+ * becomes 0x80 based on previous round (0x293 - 0x213 = 0x80).
+ * At the same time, insn at 0x292 is a 2-byte insn since its offset is
+ * -124.
+ *
+ * pass6 will repeat the same code as in pass4 and this will prevent
+ * eventual convergence.
+ *
+ * To fix this issue, we need to break je (2->6 bytes) <-> jmp (5->2 bytes)
+ * cycle in the above. In the above example je offset <= 0x7c should work.
+ *
+ * For other cases, je <-> je needs offset <= 0x7b to avoid no convergence
+ * issue. For jmp <-> je and jmp <-> jmp cases, jmp offset <= 0x7c should
+ * avoid no convergence issue.
+ *
+ * Overall, let us limit the positive offset for 8bit cond/uncond jmp insn
+ * to maximum 123 (0x7b). This way, the jit pass can eventually converge.
+ */
+static bool is_imm8_jmp_offset(int value)
+{
+	return value <= 123 && value >= -128;
+}
+
 static bool is_simm32(s64 value)
 {
 	return value == (s64)(s32)value;
@@ -1589,7 +1639,7 @@ st:			if (is_imm8(insn->off))
 				return -EFAULT;
 			}
 			jmp_offset = addrs[i + insn->off] - addrs[i];
-			if (is_imm8(jmp_offset)) {
+			if (is_imm8_jmp_offset(jmp_offset)) {
 				if (jmp_padding) {
 					/* To keep the jmp_offset valid, the extra bytes are
 					 * padded before the jump insn, so we subtract the
@@ -1671,7 +1721,7 @@ st:			if (is_imm8(insn->off))
 				break;
 			}
 emit_jmp:
-			if (is_imm8(jmp_offset)) {
+			if (is_imm8_jmp_offset(jmp_offset)) {
 				if (jmp_padding) {
 					/* To avoid breaking jmp_offset, the extra bytes
 					 * are padded before the actual jmp insn, so
-- 
2.43.0

[PATCH AUTOSEL 6.1 33/42] netfilter: nf_reject: Fix build warning when CONFIG_BRIDGE_NETFILTER=n

[Sasha Levin]

From: Simon Horman <horms@kernel.org>

[ Upstream commit fc56878ca1c288e49b5cbb43860a5938e3463654 ]

If CONFIG_BRIDGE_NETFILTER is not enabled, which is the case for x86_64
defconfig, then building nf_reject_ipv4.c and nf_reject_ipv6.c with W=1
using gcc-14 results in the following warnings, which are treated as
errors:

net/ipv4/netfilter/nf_reject_ipv4.c: In function 'nf_send_reset':
net/ipv4/netfilter/nf_reject_ipv4.c:243:23: error: variable 'niph' set but not used [-Werror=unused-but-set-variable]
  243 |         struct iphdr *niph;
      |                       ^~~~
cc1: all warnings being treated as errors
net/ipv6/netfilter/nf_reject_ipv6.c: In function 'nf_send_reset6':
net/ipv6/netfilter/nf_reject_ipv6.c:286:25: error: variable 'ip6h' set but not used [-Werror=unused-but-set-variable]
  286 |         struct ipv6hdr *ip6h;
      |                         ^~~~
cc1: all warnings being treated as errors

Address this by reducing the scope of these local variables to where
they are used, which is code only compiled when CONFIG_BRIDGE_NETFILTER
enabled.

Compile tested and run through netfilter selftests.

Reported-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Closes: https://lore.kernel.org/netfilter-devel/20240906145513.567781-1-andriy.shevchenko@linux.intel.com/
Signed-off-by: Simon Horman <horms@kernel.org>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 net/ipv4/netfilter/nf_reject_ipv4.c | 10 ++++------
 net/ipv6/netfilter/nf_reject_ipv6.c |  5 ++---
 2 files changed, 6 insertions(+), 9 deletions(-)

diff --git a/net/ipv4/netfilter/nf_reject_ipv4.c b/net/ipv4/netfilter/nf_reject_ipv4.c
index fc761915c5f6f..675b5bbed638e 100644
--- a/net/ipv4/netfilter/nf_reject_ipv4.c
+++ b/net/ipv4/netfilter/nf_reject_ipv4.c
@@ -239,9 +239,8 @@ static int nf_reject_fill_skb_dst(struct sk_buff *skb_in)
 void nf_send_reset(struct net *net, struct sock *sk, struct sk_buff *oldskb,
 		   int hook)
 {
-	struct sk_buff *nskb;
-	struct iphdr *niph;
 	const struct tcphdr *oth;
+	struct sk_buff *nskb;
 	struct tcphdr _oth;
 
 	oth = nf_reject_ip_tcphdr_get(oldskb, &_oth, hook);
@@ -266,14 +265,12 @@ void nf_send_reset(struct net *net, struct sock *sk, struct sk_buff *oldskb,
 	nskb->mark = IP4_REPLY_MARK(net, oldskb->mark);
 
 	skb_reserve(nskb, LL_MAX_HEADER);
-	niph = nf_reject_iphdr_put(nskb, oldskb, IPPROTO_TCP,
-				   ip4_dst_hoplimit(skb_dst(nskb)));
+	nf_reject_iphdr_put(nskb, oldskb, IPPROTO_TCP,
+			    ip4_dst_hoplimit(skb_dst(nskb)));
 	nf_reject_ip_tcphdr_put(nskb, oldskb, oth);
 	if (ip_route_me_harder(net, sk, nskb, RTN_UNSPEC))
 		goto free_nskb;
 
-	niph = ip_hdr(nskb);
-
 	/* "Never happens" */
 	if (nskb->len > dst_mtu(skb_dst(nskb)))
 		goto free_nskb;
@@ -290,6 +287,7 @@ void nf_send_reset(struct net *net, struct sock *sk, struct sk_buff *oldskb,
 	 */
 	if (nf_bridge_info_exists(oldskb)) {
 		struct ethhdr *oeth = eth_hdr(oldskb);
+		struct iphdr *niph = ip_hdr(nskb);
 		struct net_device *br_indev;
 
 		br_indev = nf_bridge_get_physindev(oldskb, net);
diff --git a/net/ipv6/netfilter/nf_reject_ipv6.c b/net/ipv6/netfilter/nf_reject_ipv6.c
index 71d692728230e..c8f5196d752e6 100644
--- a/net/ipv6/netfilter/nf_reject_ipv6.c
+++ b/net/ipv6/netfilter/nf_reject_ipv6.c
@@ -283,7 +283,6 @@ void nf_send_reset6(struct net *net, struct sock *sk, struct sk_buff *oldskb,
 	const struct tcphdr *otcph;
 	unsigned int otcplen, hh_len;
 	const struct ipv6hdr *oip6h = ipv6_hdr(oldskb);
-	struct ipv6hdr *ip6h;
 	struct dst_entry *dst = NULL;
 	struct flowi6 fl6;
 
@@ -339,8 +338,7 @@ void nf_send_reset6(struct net *net, struct sock *sk, struct sk_buff *oldskb,
 	nskb->mark = fl6.flowi6_mark;
 
 	skb_reserve(nskb, hh_len + dst->header_len);
-	ip6h = nf_reject_ip6hdr_put(nskb, oldskb, IPPROTO_TCP,
-				    ip6_dst_hoplimit(dst));
+	nf_reject_ip6hdr_put(nskb, oldskb, IPPROTO_TCP, ip6_dst_hoplimit(dst));
 	nf_reject_ip6_tcphdr_put(nskb, oldskb, otcph, otcplen);
 
 	nf_ct_attach(nskb, oldskb);
@@ -355,6 +353,7 @@ void nf_send_reset6(struct net *net, struct sock *sk, struct sk_buff *oldskb,
 	 */
 	if (nf_bridge_info_exists(oldskb)) {
 		struct ethhdr *oeth = eth_hdr(oldskb);
+		struct ipv6hdr *ip6h = ipv6_hdr(nskb);
 		struct net_device *br_indev;
 
 		br_indev = nf_bridge_get_physindev(oldskb, net);
-- 
2.43.0