bpf: add meta pointer for direct access

This work enables generic transfer of metadata from XDP into skb. The
basic idea is that we can make use of the fact that the resulting skb
must be linear and already comes with a larger headroom for supporting
bpf_xdp_adjust_head(), which mangles xdp->data. Here, we base our work
on a similar principle and introduce a small helper bpf_xdp_adjust_meta()
for adjusting a new pointer called xdp->data_meta. Thus, the packet has
a flexible and programmable room for meta data, followed by the actual
packet data. struct xdp_buff is therefore laid out that we first point
to data_hard_start, then data_meta directly prepended to data followed
by data_end marking the end of packet. bpf_xdp_adjust_head() takes into
account whether we have meta data already prepended and if so, memmove()s
this along with the given offset provided there's enough room.

xdp->data_meta is optional and programs are not required to use it. The
rationale is that when we process the packet in XDP (e.g. as DoS filter),
we can push further meta data along with it for the XDP_PASS case, and
give the guarantee that a clsact ingress BPF program on the same device
can pick this up for further post-processing. Since we work with skb
there, we can also set skb->mark, skb->priority or other skb meta data
out of BPF, thus having this scratch space generic and programmable
allows for more flexibility than defining a direct 1:1 transfer of
potentially new XDP members into skb (it's also more efficient as we
don't need to initialize/handle each of such new members). The facility
also works together with GRO aggregation. The scratch space at the head
of the packet can be multiple of 4 byte up to 32 byte large. Drivers not
yet supporting xdp->data_meta can simply be set up with xdp->data_meta
as xdp->data + 1 as bpf_xdp_adjust_meta() will detect this and bail out,
such that the subsequent match against xdp->data for later access is
guaranteed to fail.

The verifier treats xdp->data_meta/xdp->data the same way as we treat
xdp->data/xdp->data_end pointer comparisons. The requirement for doing
the compare against xdp->data is that it hasn't been modified from it's
original address we got from ctx access. It may have a range marking
already from prior successful xdp->data/xdp->data_end pointer comparisons
though.

Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

1 files changed, 86 insertions, 28 deletions

diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index b914fbe1383e..f849eca36052 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -177,6 +177,12 @@ static __printf(1, 2) void verbose(const char *fmt, ...)
 	va_end(args);
 }
 
+static bool type_is_pkt_pointer(enum bpf_reg_type type)
+{
+	return type == PTR_TO_PACKET ||
+	       type == PTR_TO_PACKET_META;
+}
+
 /* string representation of 'enum bpf_reg_type' */
 static const char * const reg_type_str[] = {
 	[NOT_INIT]		= "?",
@@ -187,6 +193,7 @@ static const char * const reg_type_str[] = {
 	[PTR_TO_MAP_VALUE_OR_NULL] = "map_value_or_null",
 	[PTR_TO_STACK]		= "fp",
 	[PTR_TO_PACKET]		= "pkt",
+	[PTR_TO_PACKET_META]	= "pkt_meta",
 	[PTR_TO_PACKET_END]	= "pkt_end",
 };
 
@@ -226,7 +233,7 @@ static void print_verifier_state(struct bpf_verifier_state *state)
 			verbose("(id=%d", reg->id);
 			if (t != SCALAR_VALUE)
 				verbose(",off=%d", reg->off);
-			if (t == PTR_TO_PACKET)
+			if (type_is_pkt_pointer(t))
 				verbose(",r=%d", reg->range);
 			else if (t == CONST_PTR_TO_MAP ||
 				 t == PTR_TO_MAP_VALUE ||
@@ -519,6 +526,31 @@ static void mark_reg_known_zero(struct bpf_reg_state *regs, u32 regno)
 	__mark_reg_known_zero(regs + regno);
 }
 
+static bool reg_is_pkt_pointer(const struct bpf_reg_state *reg)
+{
+	return type_is_pkt_pointer(reg->type);
+}
+
+static bool reg_is_pkt_pointer_any(const struct bpf_reg_state *reg)
+{
+	return reg_is_pkt_pointer(reg) ||
+	       reg->type == PTR_TO_PACKET_END;
+}
+
+/* Unmodified PTR_TO_PACKET[_META,_END] register from ctx access. */
+static bool reg_is_init_pkt_pointer(const struct bpf_reg_state *reg,
+				    enum bpf_reg_type which)
+{
+	/* The register can already have a range from prior markings.
+	 * This is fine as long as it hasn't been advanced from its
+	 * origin.
+	 */
+	return reg->type == which &&
+	       reg->id == 0 &&
+	       reg->off == 0 &&
+	       tnum_equals_const(reg->var_off, 0);
+}
+
 /* Attempts to improve min/max values based on var_off information */
 static void __update_reg_bounds(struct bpf_reg_state *reg)
 {
@@ -702,6 +734,7 @@ static bool is_spillable_regtype(enum bpf_reg_type type)
 	case PTR_TO_STACK:
 	case PTR_TO_CTX:
 	case PTR_TO_PACKET:
+	case PTR_TO_PACKET_META:
 	case PTR_TO_PACKET_END:
 	case CONST_PTR_TO_MAP:
 		return true;
@@ -1047,7 +1080,10 @@ static int check_ptr_alignment(struct bpf_verifier_env *env,
 
 	switch (reg->type) {
 	case PTR_TO_PACKET:
-		/* special case, because of NET_IP_ALIGN */
+	case PTR_TO_PACKET_META:
+		/* Special case, because of NET_IP_ALIGN. Given metadata sits
+		 * right in front, treat it the very same way.
+		 */
 		return check_pkt_ptr_alignment(reg, off, size, strict);
 	case PTR_TO_MAP_VALUE:
 		pointer_desc = "value ";
@@ -1124,8 +1160,8 @@ static int check_mem_access(struct bpf_verifier_env *env, int insn_idx, u32 regn
 		err = check_ctx_access(env, insn_idx, off, size, t, &reg_type);
 		if (!err && t == BPF_READ && value_regno >= 0) {
 			/* ctx access returns either a scalar, or a
-			 * PTR_TO_PACKET[_END].  In the latter case, we know
-			 * the offset is zero.
+			 * PTR_TO_PACKET[_META,_END]. In the latter
+			 * case, we know the offset is zero.
 			 */
 			if (reg_type == SCALAR_VALUE)
 				mark_reg_unknown(state->regs, value_regno);
@@ -1170,7 +1206,7 @@ static int check_mem_access(struct bpf_verifier_env *env, int insn_idx, u32 regn
 		} else {
 			err = check_stack_read(state, off, size, value_regno);
 		}
-	} else if (reg->type == PTR_TO_PACKET) {
+	} else if (reg_is_pkt_pointer(reg)) {
 		if (t == BPF_WRITE && !may_access_direct_pkt_data(env, NULL, t)) {
 			verbose("cannot write into packet\n");
 			return -EACCES;
@@ -1310,6 +1346,7 @@ static int check_helper_mem_access(struct bpf_verifier_env *env, int regno,
 
 	switch (reg->type) {
 	case PTR_TO_PACKET:
+	case PTR_TO_PACKET_META:
 		return check_packet_access(env, regno, reg->off, access_size);
 	case PTR_TO_MAP_VALUE:
 		return check_map_access(env, regno, reg->off, access_size);
@@ -1342,7 +1379,7 @@ static int check_func_arg(struct bpf_verifier_env *env, u32 regno,
 		return 0;
 	}
 
-	if (type == PTR_TO_PACKET &&
+	if (type_is_pkt_pointer(type) &&
 	    !may_access_direct_pkt_data(env, meta, BPF_READ)) {
 		verbose("helper access to the packet is not allowed\n");
 		return -EACCES;
@@ -1351,7 +1388,8 @@ static int check_func_arg(struct bpf_verifier_env *env, u32 regno,
 	if (arg_type == ARG_PTR_TO_MAP_KEY ||
 	    arg_type == ARG_PTR_TO_MAP_VALUE) {
 		expected_type = PTR_TO_STACK;
-		if (type != PTR_TO_PACKET && type != expected_type)
+		if (!type_is_pkt_pointer(type) &&
+		    type != expected_type)
 			goto err_type;
 	} else if (arg_type == ARG_CONST_SIZE ||
 		   arg_type == ARG_CONST_SIZE_OR_ZERO) {
@@ -1375,7 +1413,8 @@ static int check_func_arg(struct bpf_verifier_env *env, u32 regno,
 		 */
 		if (register_is_null(*reg))
 			/* final test in check_stack_boundary() */;
-		else if (type != PTR_TO_PACKET && type != PTR_TO_MAP_VALUE &&
+		else if (!type_is_pkt_pointer(type) &&
+			 type != PTR_TO_MAP_VALUE &&
 			 type != expected_type)
 			goto err_type;
 		meta->raw_mode = arg_type == ARG_PTR_TO_UNINIT_MEM;
@@ -1401,7 +1440,7 @@ static int check_func_arg(struct bpf_verifier_env *env, u32 regno,
 			verbose("invalid map_ptr to access map->key\n");
 			return -EACCES;
 		}
-		if (type == PTR_TO_PACKET)
+		if (type_is_pkt_pointer(type))
 			err = check_packet_access(env, regno, reg->off,
 						  meta->map_ptr->key_size);
 		else
@@ -1417,7 +1456,7 @@ static int check_func_arg(struct bpf_verifier_env *env, u32 regno,
 			verbose("invalid map_ptr to access map->value\n");
 			return -EACCES;
 		}
-		if (type == PTR_TO_PACKET)
+		if (type_is_pkt_pointer(type))
 			err = check_packet_access(env, regno, reg->off,
 						  meta->map_ptr->value_size);
 		else
@@ -1590,8 +1629,8 @@ static int check_raw_mode(const struct bpf_func_proto *fn)
 	return count > 1 ? -EINVAL : 0;
 }
 
-/* Packet data might have moved, any old PTR_TO_PACKET[_END] are now invalid,
- * so turn them into unknown SCALAR_VALUE.
+/* Packet data might have moved, any old PTR_TO_PACKET[_META,_END]
+ * are now invalid, so turn them into unknown SCALAR_VALUE.
  */
 static void clear_all_pkt_pointers(struct bpf_verifier_env *env)
 {
@@ -1600,18 +1639,15 @@ static void clear_all_pkt_pointers(struct bpf_verifier_env *env)
 	int i;
 
 	for (i = 0; i < MAX_BPF_REG; i++)
-		if (regs[i].type == PTR_TO_PACKET ||
-		    regs[i].type == PTR_TO_PACKET_END)
+		if (reg_is_pkt_pointer_any(&regs[i]))
 			mark_reg_unknown(regs, i);
 
 	for (i = 0; i < MAX_BPF_STACK; i += BPF_REG_SIZE) {
 		if (state->stack_slot_type[i] != STACK_SPILL)
 			continue;
 		reg = &state->spilled_regs[i / BPF_REG_SIZE];
-		if (reg->type != PTR_TO_PACKET &&
-		    reg->type != PTR_TO_PACKET_END)
-			continue;
-		__mark_reg_unknown(reg);
+		if (reg_is_pkt_pointer_any(reg))
+			__mark_reg_unknown(reg);
 	}
 }
 
@@ -1871,7 +1907,7 @@ static int adjust_ptr_min_max_vals(struct bpf_verifier_env *env,
 		}
 		dst_reg->var_off = tnum_add(ptr_reg->var_off, off_reg->var_off);
 		dst_reg->off = ptr_reg->off;
-		if (ptr_reg->type == PTR_TO_PACKET) {
+		if (reg_is_pkt_pointer(ptr_reg)) {
 			dst_reg->id = ++env->id_gen;
 			/* something was added to pkt_ptr, set range to zero */
 			dst_reg->range = 0;
@@ -1931,7 +1967,7 @@ static int adjust_ptr_min_max_vals(struct bpf_verifier_env *env,
 		}
 		dst_reg->var_off = tnum_sub(ptr_reg->var_off, off_reg->var_off);
 		dst_reg->off = ptr_reg->off;
-		if (ptr_reg->type == PTR_TO_PACKET) {
+		if (reg_is_pkt_pointer(ptr_reg)) {
 			dst_reg->id = ++env->id_gen;
 			/* something was added to pkt_ptr, set range to zero */
 			if (smin_val < 0)
@@ -2421,7 +2457,8 @@ static int check_alu_op(struct bpf_verifier_env *env, struct bpf_insn *insn)
 }
 
 static void find_good_pkt_pointers(struct bpf_verifier_state *state,
-				   struct bpf_reg_state *dst_reg)
+				   struct bpf_reg_state *dst_reg,
+				   enum bpf_reg_type type)
 {
 	struct bpf_reg_state *regs = state->regs, *reg;
 	int i;
@@ -2483,7 +2520,7 @@ static void find_good_pkt_pointers(struct bpf_verifier_state *state,
 	 * dst_reg->off is known < MAX_PACKET_OFF, therefore it fits in a u16.
 	 */
 	for (i = 0; i < MAX_BPF_REG; i++)
-		if (regs[i].type == PTR_TO_PACKET && regs[i].id == dst_reg->id)
+		if (regs[i].type == type && regs[i].id == dst_reg->id)
 			/* keep the maximum range already checked */
 			regs[i].range = max_t(u16, regs[i].range, dst_reg->off);
 
@@ -2491,7 +2528,7 @@ static void find_good_pkt_pointers(struct bpf_verifier_state *state,
 		if (state->stack_slot_type[i] != STACK_SPILL)
 			continue;
 		reg = &state->spilled_regs[i / BPF_REG_SIZE];
-		if (reg->type == PTR_TO_PACKET && reg->id == dst_reg->id)
+		if (reg->type == type && reg->id == dst_reg->id)
 			reg->range = max_t(u16, reg->range, dst_reg->off);
 	}
 }
@@ -2856,19 +2893,39 @@ static int check_cond_jmp_op(struct bpf_verifier_env *env,
 	} else if (BPF_SRC(insn->code) == BPF_X && opcode == BPF_JGT &&
 		   dst_reg->type == PTR_TO_PACKET &&
 		   regs[insn->src_reg].type == PTR_TO_PACKET_END) {
-		find_good_pkt_pointers(this_branch, dst_reg);
+		find_good_pkt_pointers(this_branch, dst_reg, PTR_TO_PACKET);
 	} else if (BPF_SRC(insn->code) == BPF_X && opcode == BPF_JLT &&
 		   dst_reg->type == PTR_TO_PACKET &&
 		   regs[insn->src_reg].type == PTR_TO_PACKET_END) {
-		find_good_pkt_pointers(other_branch, dst_reg);
+		find_good_pkt_pointers(other_branch, dst_reg, PTR_TO_PACKET);
 	} else if (BPF_SRC(insn->code) == BPF_X && opcode == BPF_JGE &&
 		   dst_reg->type == PTR_TO_PACKET_END &&
 		   regs[insn->src_reg].type == PTR_TO_PACKET) {
-		find_good_pkt_pointers(other_branch, &regs[insn->src_reg]);
+		find_good_pkt_pointers(other_branch, &regs[insn->src_reg],
+				       PTR_TO_PACKET);
 	} else if (BPF_SRC(insn->code) == BPF_X && opcode == BPF_JLE &&
 		   dst_reg->type == PTR_TO_PACKET_END &&
 		   regs[insn->src_reg].type == PTR_TO_PACKET) {
-		find_good_pkt_pointers(this_branch, &regs[insn->src_reg]);
+		find_good_pkt_pointers(this_branch, &regs[insn->src_reg],
+				       PTR_TO_PACKET);
+	} else if (BPF_SRC(insn->code) == BPF_X && opcode == BPF_JGT &&
+		   dst_reg->type == PTR_TO_PACKET_META &&
+		   reg_is_init_pkt_pointer(&regs[insn->src_reg], PTR_TO_PACKET)) {
+		find_good_pkt_pointers(this_branch, dst_reg, PTR_TO_PACKET_META);
+	} else if (BPF_SRC(insn->code) == BPF_X && opcode == BPF_JLT &&
+		   dst_reg->type == PTR_TO_PACKET_META &&
+		   reg_is_init_pkt_pointer(&regs[insn->src_reg], PTR_TO_PACKET)) {
+		find_good_pkt_pointers(other_branch, dst_reg, PTR_TO_PACKET_META);
+	} else if (BPF_SRC(insn->code) == BPF_X && opcode == BPF_JGE &&
+		   reg_is_init_pkt_pointer(dst_reg, PTR_TO_PACKET) &&
+		   regs[insn->src_reg].type == PTR_TO_PACKET_META) {
+		find_good_pkt_pointers(other_branch, &regs[insn->src_reg],
+				       PTR_TO_PACKET_META);
+	} else if (BPF_SRC(insn->code) == BPF_X && opcode == BPF_JLE &&
+		   reg_is_init_pkt_pointer(dst_reg, PTR_TO_PACKET) &&
+		   regs[insn->src_reg].type == PTR_TO_PACKET_META) {
+		find_good_pkt_pointers(this_branch, &regs[insn->src_reg],
+				       PTR_TO_PACKET_META);
 	} else if (is_pointer_value(env, insn->dst_reg)) {
 		verbose("R%d pointer comparison prohibited\n", insn->dst_reg);
 		return -EACCES;
@@ -3298,8 +3355,9 @@ static bool regsafe(struct bpf_reg_state *rold, struct bpf_reg_state *rcur,
 			return false;
 		/* Check our ids match any regs they're supposed to */
 		return check_ids(rold->id, rcur->id, idmap);
+	case PTR_TO_PACKET_META:
 	case PTR_TO_PACKET:
-		if (rcur->type != PTR_TO_PACKET)
+		if (rcur->type != rold->type)
 			return false;
 		/* We must have at least as much range as the old ptr
 		 * did, so that any accesses which were safe before are