x86/kprobes: Fix crash when probing CS CALL instructionsgrafted

When using eBPF to probe CS CALL instructions within a function,
a crash can be triggered.

The eBPF tool probes offset 257 of the __hrtimer_run_queues()
function:

<__hrtimer_run_queues+249>:  nopl   0x0(%rax,%rax,1)
<__hrtimer_run_queues+254>:  mov    %r14,%rdi
<__hrtimer_run_queues+257>:  cs call <__x86_indirect_thunk_r12>
<__hrtimer_run_queues+263>:  mov    %eax,%r12d
<__hrtimer_run_queues+266>:  xchg   %ax,%ax
<__hrtimer_run_queues+268>:  mov    %r13,%rdi

Which triggers this crash:

  BUG: unable to handle page fault for address: 00000000000f41c9
  #PF: supervisor write access in kernel mode
  #PF: error_code(0x0002) - not-present page
  PGD 0 P4D 0
  Oops: 0002 [#1] SMP NOPTI
  CPU: 1 PID: 0 Comm: swapper/1 Kdump: loaded Tainted: P
  RIP: 0010:__hrtimer_run_queues+0x106/0x230

Note that __hrtimer_run_queues+0x106 is __hrtimer_run_queues+262, which is
at the 6th byte of the above CS CALL instruction. Since the CS CALL
instruction occupies 6 bytes, the exception occurred in the middle of that
call instruction.

The root cause is that when using eBPF tools to probe in the middle of a
function, a kprobe with INT3 is used as the underlying implementation.

During single-step emulation of the original CALL instruction,
int3_emulate_call() assumes that the probed CALL instruction is 5 bytes
long. However, the actual CS-prefixed CALL instruction occupies 6 bytes,
so it constructs an incorrect exception return address. When the CPU
returns from the kprobe handler, the next instruction to be executed is at
the address of the last byte of that CS CALL instruction. Coincidentally,
starting from that address, the CPU fetches and decodes a completely
different instruction, which ultimately triggers a kernel crash.

Fix the issue by using the actual instruction length obtained from
the instruction decoder when constructing the exception return
address, rather than relying on the hardcoded CALL_INSN_SIZE macro.

[ mingo: Refined the changelog ]

Fixes: 6256e668b7af ("x86/kprobes: Use int3 instead of debug trap for single-step")
Suggested-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Signed-off-by: Jinke Han <jinkehan@didiglobal.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Acked-by: Yafang Shao <laoar.shao@gmail.com>
Acked-by: Borislav Petkov <bp@alien8.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://patch.msgid.link/20260908073742.GA10517@didi-ThinkCentre-M920t-N000

1 files changed, 435 insertions, 0 deletions

diff --git a/net/ipv6/esp6_offload.c b/net/ipv6/esp6_offload.c
new file mode 100644
index 000000000..22895521a
--- /dev/null
+++ b/net/ipv6/esp6_offload.c
@@ -0,0 +1,435 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * IPV6 GSO/GRO offload support
+ * Linux INET implementation
+ *
+ * Copyright (C) 2016 secunet Security Networks AG
+ * Author: Steffen Klassert <steffen.klassert@secunet.com>
+ *
+ * ESP GRO support
+ */
+
+#include <linux/skbuff.h>
+#include <linux/init.h>
+#include <net/protocol.h>
+#include <crypto/aead.h>
+#include <crypto/authenc.h>
+#include <linux/err.h>
+#include <linux/module.h>
+#include <net/gro.h>
+#include <net/gso.h>
+#include <net/ip.h>
+#include <net/xfrm.h>
+#include <net/esp.h>
+#include <linux/scatterlist.h>
+#include <linux/kernel.h>
+#include <linux/slab.h>
+#include <linux/spinlock.h>
+#include <net/ip6_route.h>
+#include <net/ipv6.h>
+#include <linux/icmpv6.h>
+
+static __u16 esp6_nexthdr_esp_offset(struct ipv6hdr *ipv6_hdr, int nhlen)
+{
+	int off = sizeof(struct ipv6hdr);
+	struct ipv6_opt_hdr *exthdr;
+
+	/* ESP or ESPINUDP */
+	if (likely(ipv6_hdr->nexthdr == NEXTHDR_ESP ||
+		   ipv6_hdr->nexthdr == NEXTHDR_UDP))
+		return offsetof(struct ipv6hdr, nexthdr);
+
+	while (off < nhlen) {
+		exthdr = (void *)ipv6_hdr + off;
+		if (exthdr->nexthdr == NEXTHDR_ESP)
+			return off;
+
+		off += ipv6_optlen(exthdr);
+	}
+
+	return 0;
+}
+
+static struct sk_buff *esp6_gro_receive(struct list_head *head,
+					struct sk_buff *skb)
+{
+	int offset = skb_gro_offset(skb);
+	struct xfrm_offload *xo;
+	struct xfrm_state *x;
+	int encap_type = 0;
+	__be32 seq;
+	__be32 spi;
+	int nhoff;
+
+	if (NAPI_GRO_CB(skb)->proto == IPPROTO_UDP)
+		encap_type = UDP_ENCAP_ESPINUDP;
+
+	if (!pskb_pull(skb, offset))
+		return NULL;
+
+	if (xfrm_parse_spi(skb, IPPROTO_ESP, &spi, &seq) != 0)
+		goto out;
+
+	xo = xfrm_offload(skb);
+	if (!xo || !(xo->flags & CRYPTO_DONE)) {
+		struct sec_path *sp = secpath_set(skb);
+
+		if (!sp)
+			goto out;
+
+		if (sp->len == XFRM_MAX_DEPTH)
+			goto out_reset;
+
+		x = xfrm_input_state_lookup(dev_net(skb->dev), skb->mark,
+					    (xfrm_address_t *)&ipv6_hdr(skb)->daddr,
+					    spi, IPPROTO_ESP, AF_INET6);
+
+		if (unlikely(x && x->dir && x->dir != XFRM_SA_DIR_IN)) {
+			/* non-offload path will record the error and audit log */
+			xfrm_state_put(x);
+			x = NULL;
+		}
+
+		if (!x)
+			goto out_reset;
+
+		skb->mark = xfrm_smark_get(skb->mark, x);
+
+		sp->xvec[sp->len++] = x;
+		sp->olen++;
+
+		xo = xfrm_offload(skb);
+		if (!xo)
+			goto out_reset;
+	}
+
+	xo->flags |= XFRM_GRO;
+
+	nhoff = esp6_nexthdr_esp_offset(ipv6_hdr(skb), offset);
+	if (!nhoff)
+		goto out;
+
+	IP6CB(skb)->nhoff = nhoff;
+	XFRM_TUNNEL_SKB_CB(skb)->tunnel.ip6 = NULL;
+	XFRM_SPI_SKB_CB(skb)->family = AF_INET6;
+	XFRM_SPI_SKB_CB(skb)->daddroff = offsetof(struct ipv6hdr, daddr);
+	XFRM_SPI_SKB_CB(skb)->seq = seq;
+
+	/* We don't need to handle errors from xfrm_input, it does all
+	 * the error handling and frees the resources on error. */
+	xfrm_input(skb, IPPROTO_ESP, spi, encap_type);
+
+	return ERR_PTR(-EINPROGRESS);
+out_reset:
+	secpath_reset(skb);
+out:
+	skb_push(skb, offset);
+	NAPI_GRO_CB(skb)->same_flow = 0;
+	NAPI_GRO_CB(skb)->flush = 1;
+
+	return NULL;
+}
+
+static void esp6_gso_encap(struct xfrm_state *x, struct sk_buff *skb)
+{
+	struct ip_esp_hdr *esph;
+	struct ipv6hdr *iph = ipv6_hdr(skb);
+	struct xfrm_offload *xo = xfrm_offload(skb);
+	u8 proto = iph->nexthdr;
+
+	skb_push(skb, -skb_network_offset(skb));
+
+	if (x->outer_mode.encap == XFRM_MODE_TRANSPORT) {
+		__be16 frag;
+
+		ipv6_skip_exthdr(skb, sizeof(struct ipv6hdr), &proto, &frag);
+	}
+
+	esph = ip_esp_hdr(skb);
+	*skb_mac_header(skb) = IPPROTO_ESP;
+
+	esph->spi = x->id.spi;
+	esph->seq_no = htonl(XFRM_SKB_CB(skb)->seq.output.low);
+
+	xo->proto = proto;
+}
+
+static struct sk_buff *xfrm6_tunnel_gso_segment(struct xfrm_state *x,
+						struct sk_buff *skb,
+						netdev_features_t features)
+{
+	struct xfrm_offload *xo = xfrm_offload(skb);
+	const struct xfrm_mode *inner_mode = xfrm_ip2inner_mode(x, xo->proto);
+	__be16 type = inner_mode->family == AF_INET ? htons(ETH_P_IP)
+						    : htons(ETH_P_IPV6);
+
+	return skb_eth_gso_segment(skb, features, type);
+}
+
+static struct sk_buff *xfrm6_transport_gso_segment(struct xfrm_state *x,
+						   struct sk_buff *skb,
+						   netdev_features_t features)
+{
+	const struct net_offload *ops;
+	struct sk_buff *segs = ERR_PTR(-EINVAL);
+	struct xfrm_offload *xo = xfrm_offload(skb);
+
+	skb->transport_header += x->props.header_len;
+	ops = rcu_dereference(inet6_offloads[xo->proto]);
+	if (likely(ops && ops->callbacks.gso_segment))
+		segs = ops->callbacks.gso_segment(skb, features);
+
+	return segs;
+}
+
+static struct sk_buff *xfrm6_beet_gso_segment(struct xfrm_state *x,
+					      struct sk_buff *skb,
+					      netdev_features_t features)
+{
+	struct xfrm_offload *xo = xfrm_offload(skb);
+	struct sk_buff *segs = ERR_PTR(-EINVAL);
+	const struct net_offload *ops;
+	u8 proto = xo->proto;
+
+	skb->transport_header += x->props.header_len;
+
+	if (x->sel.family != AF_INET6) {
+		skb->transport_header -=
+			(sizeof(struct ipv6hdr) - sizeof(struct iphdr));
+
+		if (proto == IPPROTO_BEETPH) {
+			struct ip_beet_phdr *ph =
+				(struct ip_beet_phdr *)skb->data;
+
+			skb->transport_header += ph->hdrlen * 8;
+			proto = ph->nexthdr;
+		} else {
+			skb->transport_header -= IPV4_BEET_PHMAXLEN;
+		}
+
+		if (proto == IPPROTO_TCP)
+			skb_shinfo(skb)->gso_type |= SKB_GSO_TCPV6;
+	} else {
+		__be16 frag;
+
+		skb->transport_header +=
+			ipv6_skip_exthdr(skb, 0, &proto, &frag);
+	}
+
+	if (proto == IPPROTO_IPIP)
+		skb_shinfo(skb)->gso_type |= SKB_GSO_IPXIP6;
+
+	__skb_pull(skb, skb_transport_offset(skb));
+	ops = rcu_dereference(inet6_offloads[proto]);
+	if (likely(ops && ops->callbacks.gso_segment))
+		segs = ops->callbacks.gso_segment(skb, features);
+
+	return segs;
+}
+
+static struct sk_buff *xfrm6_outer_mode_gso_segment(struct xfrm_state *x,
+						    struct sk_buff *skb,
+						    netdev_features_t features)
+{
+	switch (x->outer_mode.encap) {
+	case XFRM_MODE_TUNNEL:
+		return xfrm6_tunnel_gso_segment(x, skb, features);
+	case XFRM_MODE_TRANSPORT:
+		return xfrm6_transport_gso_segment(x, skb, features);
+	case XFRM_MODE_BEET:
+		return xfrm6_beet_gso_segment(x, skb, features);
+	}
+
+	return ERR_PTR(-EOPNOTSUPP);
+}
+
+static struct sk_buff *esp6_gso_segment(struct sk_buff *skb,
+				        netdev_features_t features)
+{
+	struct xfrm_state *x;
+	struct ip_esp_hdr *esph;
+	struct crypto_aead *aead;
+	netdev_features_t esp_features = features;
+	struct xfrm_offload *xo = xfrm_offload(skb);
+	struct sec_path *sp;
+
+	if (!xo)
+		return ERR_PTR(-EINVAL);
+
+	if (!(skb_shinfo(skb)->gso_type & SKB_GSO_ESP))
+		return ERR_PTR(-EINVAL);
+
+	sp = skb_sec_path(skb);
+	x = sp->xvec[sp->len - 1];
+	aead = x->data;
+	esph = ip_esp_hdr(skb);
+
+	if (esph->spi != x->id.spi)
+		return ERR_PTR(-EINVAL);
+
+	if (!pskb_may_pull(skb, sizeof(*esph) + crypto_aead_ivsize(aead)))
+		return ERR_PTR(-EINVAL);
+
+	__skb_pull(skb, sizeof(*esph) + crypto_aead_ivsize(aead));
+
+	skb->encap_hdr_csum = 1;
+
+	if (!(features & NETIF_F_HW_ESP) || x->xso.dev != skb->dev)
+		esp_features = features & ~(NETIF_F_SG | NETIF_F_CSUM_MASK |
+					    NETIF_F_SCTP_CRC);
+	else if (!(features & NETIF_F_HW_ESP_TX_CSUM))
+		esp_features = features & ~(NETIF_F_CSUM_MASK |
+					    NETIF_F_SCTP_CRC);
+
+	xo->flags |= XFRM_GSO_SEGMENT;
+
+	return xfrm6_outer_mode_gso_segment(x, skb, esp_features);
+}
+
+static int esp6_input_tail(struct xfrm_state *x, struct sk_buff *skb)
+{
+	struct crypto_aead *aead = x->data;
+	struct xfrm_offload *xo = xfrm_offload(skb);
+
+	if (!pskb_may_pull(skb, sizeof(struct ip_esp_hdr) + crypto_aead_ivsize(aead)))
+		return -EINVAL;
+
+	if (!(xo->flags & CRYPTO_DONE))
+		skb->ip_summed = CHECKSUM_NONE;
+
+	return esp6_input_done2(skb, 0);
+}
+
+static int esp6_xmit(struct xfrm_state *x, struct sk_buff *skb,  netdev_features_t features)
+{
+	int len;
+	int err;
+	int alen;
+	int blksize;
+	struct xfrm_offload *xo;
+	struct crypto_aead *aead;
+	struct esp_info esp;
+	bool hw_offload = true;
+	__u32 seq;
+
+	esp.inplace = true;
+
+	xo = xfrm_offload(skb);
+
+	if (!xo)
+		return -EINVAL;
+
+	if (!(features & NETIF_F_HW_ESP) || x->xso.dev != skb->dev) {
+		xo->flags |= CRYPTO_FALLBACK;
+		hw_offload = false;
+	}
+
+	esp.proto = xo->proto;
+
+	/* skb is pure payload to encrypt */
+
+	aead = x->data;
+	alen = crypto_aead_authsize(aead);
+
+	esp.tfclen = 0;
+	/* XXX: Add support for tfc padding here. */
+
+	blksize = ALIGN(crypto_aead_blocksize(aead), 4);
+	esp.clen = ALIGN(skb->len + 2 + esp.tfclen, blksize);
+	esp.plen = esp.clen - skb->len - esp.tfclen;
+	esp.tailen = esp.tfclen + esp.plen + alen;
+
+	if (!hw_offload || !skb_is_gso(skb)) {
+		esp.nfrags = esp6_output_head(x, skb, &esp);
+		if (esp.nfrags < 0)
+			return esp.nfrags;
+	}
+
+	seq = xo->seq.low;
+
+	esp.esph = ip_esp_hdr(skb);
+	esp.esph->spi = x->id.spi;
+
+	skb_push(skb, -skb_network_offset(skb));
+
+	if (xo->flags & XFRM_GSO_SEGMENT) {
+		esp.esph->seq_no = htonl(seq);
+
+		if (!skb_is_gso(skb))
+			xo->seq.low++;
+		else
+			xo->seq.low += skb_shinfo(skb)->gso_segs;
+	}
+
+	if (xo->seq.low < seq)
+		xo->seq.hi++;
+
+	esp.seqno = cpu_to_be64(xo->seq.low + ((u64)xo->seq.hi << 32));
+
+	len = skb->len - sizeof(struct ipv6hdr);
+	if (len > IPV6_MAXPLEN)
+		len = 0;
+
+	ipv6_hdr(skb)->payload_len = htons(len);
+
+	if (hw_offload) {
+		if (!skb_ext_add(skb, SKB_EXT_SEC_PATH))
+			return -ENOMEM;
+
+		xo = xfrm_offload(skb);
+		if (!xo)
+			return -EINVAL;
+
+		xo->flags |= XFRM_XMIT;
+		return 0;
+	}
+
+	err = esp6_output_tail(x, skb, &esp);
+	if (err)
+		return err;
+
+	secpath_reset(skb);
+
+	if (skb_needs_linearize(skb, skb->dev->features) &&
+	    __skb_linearize(skb))
+		return -ENOMEM;
+	return 0;
+}
+
+static const struct net_offload esp6_offload = {
+	.callbacks = {
+		.gro_receive = esp6_gro_receive,
+		.gso_segment = esp6_gso_segment,
+	},
+};
+
+static const struct xfrm_type_offload esp6_type_offload = {
+	.owner		= THIS_MODULE,
+	.proto	     	= IPPROTO_ESP,
+	.input_tail	= esp6_input_tail,
+	.xmit		= esp6_xmit,
+	.encap		= esp6_gso_encap,
+};
+
+static int __init esp6_offload_init(void)
+{
+	if (xfrm_register_type_offload(&esp6_type_offload, AF_INET6) < 0) {
+		pr_info("%s: can't add xfrm type offload\n", __func__);
+		return -EAGAIN;
+	}
+
+	return inet6_add_offload(&esp6_offload, IPPROTO_ESP);
+}
+
+static void __exit esp6_offload_exit(void)
+{
+	xfrm_unregister_type_offload(&esp6_type_offload, AF_INET6);
+	inet6_del_offload(&esp6_offload, IPPROTO_ESP);
+}
+
+module_init(esp6_offload_init);
+module_exit(esp6_offload_exit);
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Steffen Klassert <steffen.klassert@secunet.com>");
+MODULE_ALIAS_XFRM_OFFLOAD_TYPE(AF_INET6, XFRM_PROTO_ESP);
+MODULE_DESCRIPTION("IPV6 GSO/GRO offload support");