Merge git://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf

Daniel Borkmann says:

====================
pull-request: bpf 2021-07-29

The following pull-request contains BPF updates for your *net* tree.

We've added 9 non-merge commits during the last 14 day(s) which contain
a total of 20 files changed, 446 insertions(+), 138 deletions(-).

The main changes are:

1) Fix UBSAN out-of-bounds splat for showing XDP link fdinfo, from Lorenz Bauer.

2) Fix insufficient Spectre v4 mitigation in BPF runtime, from Daniel Borkmann,
   Piotr Krysiuk and Benedict Schlueter.

3) Batch of fixes for BPF sockmap found under stress testing, from John Fastabend.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

arch/arm/net/bpf_jit_32.c

@@ -1602,6 +1602,9 @@ static int build_insn(const struct bpf_insn *insn, struct jit_ctx *ctx)
 		rn = arm_bpf_get_reg32(src_lo, tmp2[1], ctx);
 		emit_ldx_r(dst, rn, off, ctx, BPF_SIZE(code));
 		break;
+	/* speculation barrier */
+	case BPF_ST | BPF_NOSPEC:
+		break;
 	/* ST: *(size *)(dst + off) = imm */
 	case BPF_ST | BPF_MEM | BPF_W:
 	case BPF_ST | BPF_MEM | BPF_H:

arch/arm64/net/bpf_jit_comp.c

@@ -823,6 +823,19 @@ static int build_insn(const struct bpf_insn *insn, struct jit_ctx *ctx,
 			return ret;
 		break;
 
+	/* speculation barrier */
+	case BPF_ST | BPF_NOSPEC:
+		/*
+		 * Nothing required here.
+		 *
+		 * In case of arm64, we rely on the firmware mitigation of
+		 * Speculative Store Bypass as controlled via the ssbd kernel
+		 * parameter. Whenever the mitigation is enabled, it works
+		 * for all of the kernel code with no need to provide any
+		 * additional instructions.
+		 */
+		break;
+
 	/* ST: *(size *)(dst + off) = imm */
 	case BPF_ST | BPF_MEM | BPF_W:
 	case BPF_ST | BPF_MEM | BPF_H:

arch/mips/net/ebpf_jit.c

@@ -1355,6 +1355,9 @@ static int build_one_insn(const struct bpf_insn *insn, struct jit_ctx *ctx,
 		}
 		break;
 
+	case BPF_ST | BPF_NOSPEC: /* speculation barrier */
+		break;
+
 	case BPF_ST | BPF_B | BPF_MEM:
 	case BPF_ST | BPF_H | BPF_MEM:
 	case BPF_ST | BPF_W | BPF_MEM:

arch/powerpc/net/bpf_jit_comp32.c

@@ -737,6 +737,12 @@ int bpf_jit_build_body(struct bpf_prog *fp, u32 *image, struct codegen_context *
 			}
 			break;
 
+		/*
+		 * BPF_ST NOSPEC (speculation barrier)
+		 */
+		case BPF_ST | BPF_NOSPEC:
+			break;
+
 		/*
 		 * BPF_ST(X)
 		 */

arch/powerpc/net/bpf_jit_comp64.c

@@ -627,6 +627,12 @@ int bpf_jit_build_body(struct bpf_prog *fp, u32 *image, struct codegen_context *
 			}
 			break;
 
+		/*
+		 * BPF_ST NOSPEC (speculation barrier)
+		 */
+		case BPF_ST | BPF_NOSPEC:
+			break;
+
 		/*
 		 * BPF_ST(X)
 		 */

arch/riscv/net/bpf_jit_comp32.c

@@ -1251,6 +1251,10 @@ int bpf_jit_emit_insn(const struct bpf_insn *insn, struct rv_jit_context *ctx,
 			return -1;
 		break;
 
+	/* speculation barrier */
+	case BPF_ST | BPF_NOSPEC:
+		break;
+
 	case BPF_ST | BPF_MEM | BPF_B:
 	case BPF_ST | BPF_MEM | BPF_H:
 	case BPF_ST | BPF_MEM | BPF_W:

arch/riscv/net/bpf_jit_comp64.c

@@ -939,6 +939,10 @@ int bpf_jit_emit_insn(const struct bpf_insn *insn, struct rv_jit_context *ctx,
 		emit_ld(rd, 0, RV_REG_T1, ctx);
 		break;
 
+	/* speculation barrier */
+	case BPF_ST | BPF_NOSPEC:
+		break;
+
 	/* ST: *(size *)(dst + off) = imm */
 	case BPF_ST | BPF_MEM | BPF_B:
 		emit_imm(RV_REG_T1, imm, ctx);

arch/s390/net/bpf_jit_comp.c

@@ -1153,6 +1153,11 @@ static noinline int bpf_jit_insn(struct bpf_jit *jit, struct bpf_prog *fp,
 			break;
 		}
 		break;
+	/*
+	 * BPF_NOSPEC (speculation barrier)
+	 */
+	case BPF_ST | BPF_NOSPEC:
+		break;
 	/*
 	 * BPF_ST(X)
 	 */

arch/sparc/net/bpf_jit_comp_64.c

@@ -1287,6 +1287,9 @@ static int build_insn(const struct bpf_insn *insn, struct jit_ctx *ctx)
 			return 1;
 		break;
 	}
+	/* speculation barrier */
+	case BPF_ST | BPF_NOSPEC:
+		break;
 	/* ST: *(size *)(dst + off) = imm */
 	case BPF_ST | BPF_MEM | BPF_W:
 	case BPF_ST | BPF_MEM | BPF_H:

arch/x86/net/bpf_jit_comp.c

@@ -1219,6 +1219,13 @@ static int do_jit(struct bpf_prog *bpf_prog, int *addrs, u8 *image,
 			}
 			break;
 
+			/* speculation barrier */
+		case BPF_ST | BPF_NOSPEC:
+			if (boot_cpu_has(X86_FEATURE_XMM2))
+				/* Emit 'lfence' */
+				EMIT3(0x0F, 0xAE, 0xE8);
+			break;
+
 			/* ST: *(u8*)(dst_reg + off) = imm */
 		case BPF_ST | BPF_MEM | BPF_B:
 			if (is_ereg(dst_reg))

arch/x86/net/bpf_jit_comp32.c

@@ -1886,6 +1886,12 @@ static int do_jit(struct bpf_prog *bpf_prog, int *addrs, u8 *image,
 			i++;
 			break;
 		}
+		/* speculation barrier */
+		case BPF_ST | BPF_NOSPEC:
+			if (boot_cpu_has(X86_FEATURE_XMM2))
+				/* Emit 'lfence' */
+				EMIT3(0x0F, 0xAE, 0xE8);
+			break;
 		/* ST: *(u8*)(dst_reg + off) = imm */
 		case BPF_ST | BPF_MEM | BPF_H:
 		case BPF_ST | BPF_MEM | BPF_B:

include/linux/bpf_types.h

@@ -134,4 +134,5 @@ BPF_LINK_TYPE(BPF_LINK_TYPE_CGROUP, cgroup)
 BPF_LINK_TYPE(BPF_LINK_TYPE_ITER, iter)
 #ifdef CONFIG_NET
 BPF_LINK_TYPE(BPF_LINK_TYPE_NETNS, netns)
+BPF_LINK_TYPE(BPF_LINK_TYPE_XDP, xdp)
 #endif

include/linux/bpf_verifier.h

@@ -340,8 +340,8 @@ struct bpf_insn_aux_data {
 	};
 	u64 map_key_state; /* constant (32 bit) key tracking for maps */
 	int ctx_field_size; /* the ctx field size for load insn, maybe 0 */
-	int sanitize_stack_off; /* stack slot to be cleared */
 	u32 seen; /* this insn was processed by the verifier at env->pass_cnt */
+	bool sanitize_stack_spill; /* subject to Spectre v4 sanitation */
 	bool zext_dst; /* this insn zero extends dst reg */
 	u8 alu_state; /* used in combination with alu_limit */
 
@@ -414,6 +414,7 @@ struct bpf_verifier_env {
 	u32 used_map_cnt;		/* number of used maps */
 	u32 used_btf_cnt;		/* number of used BTF objects */
 	u32 id_gen;			/* used to generate unique reg IDs */
+	bool explore_alu_limits;
 	bool allow_ptr_leaks;
 	bool allow_uninit_stack;
 	bool allow_ptr_to_map_access;

include/linux/filter.h

@@ -73,6 +73,11 @@ struct ctl_table_header;
 /* unused opcode to mark call to interpreter with arguments */
 #define BPF_CALL_ARGS	0xe0
 
+/* unused opcode to mark speculation barrier for mitigating
+ * Speculative Store Bypass
+ */
+#define BPF_NOSPEC	0xc0
+
 /* As per nm, we expose JITed images as text (code) section for
  * kallsyms. That way, tools like perf can find it to match
  * addresses.
@@ -390,6 +395,16 @@ static inline bool insn_is_zext(const struct bpf_insn *insn)
 		.off   = 0,					\
 		.imm   = 0 })
 
+/* Speculation barrier */
+
+#define BPF_ST_NOSPEC()						\
+	((struct bpf_insn) {					\
+		.code  = BPF_ST | BPF_NOSPEC,			\
+		.dst_reg = 0,					\
+		.src_reg = 0,					\
+		.off   = 0,					\
+		.imm   = 0 })
+
 /* Internal classic blocks for direct assignment */
 
 #define __BPF_STMT(CODE, K)					\

include/linux/skmsg.h

@@ -285,11 +285,45 @@ static inline struct sk_psock *sk_psock(const struct sock *sk)
 	return rcu_dereference_sk_user_data(sk);
 }
 
+static inline void sk_psock_set_state(struct sk_psock *psock,
+				      enum sk_psock_state_bits bit)
+{
+	set_bit(bit, &psock->state);
+}
+
+static inline void sk_psock_clear_state(struct sk_psock *psock,
+					enum sk_psock_state_bits bit)
+{
+	clear_bit(bit, &psock->state);
+}
+
+static inline bool sk_psock_test_state(const struct sk_psock *psock,
+				       enum sk_psock_state_bits bit)
+{
+	return test_bit(bit, &psock->state);
+}
+
+static inline void sock_drop(struct sock *sk, struct sk_buff *skb)
+{
+	sk_drops_add(sk, skb);
+	kfree_skb(skb);
+}
+
+static inline void drop_sk_msg(struct sk_psock *psock, struct sk_msg *msg)
+{
+	if (msg->skb)
+		sock_drop(psock->sk, msg->skb);
+	kfree(msg);
+}
+
 static inline void sk_psock_queue_msg(struct sk_psock *psock,
 				      struct sk_msg *msg)
 {
 	spin_lock_bh(&psock->ingress_lock);
-	list_add_tail(&msg->list, &psock->ingress_msg);
+	if (sk_psock_test_state(psock, SK_PSOCK_TX_ENABLED))
+		list_add_tail(&msg->list, &psock->ingress_msg);
+	else
+		drop_sk_msg(psock, msg);
 	spin_unlock_bh(&psock->ingress_lock);
 }
 
@@ -406,24 +440,6 @@ static inline void sk_psock_restore_proto(struct sock *sk,
 		psock->psock_update_sk_prot(sk, psock, true);
 }
 
-static inline void sk_psock_set_state(struct sk_psock *psock,
-				      enum sk_psock_state_bits bit)
-{
-	set_bit(bit, &psock->state);
-}
-
-static inline void sk_psock_clear_state(struct sk_psock *psock,
-					enum sk_psock_state_bits bit)
-{
-	clear_bit(bit, &psock->state);
-}
-
-static inline bool sk_psock_test_state(const struct sk_psock *psock,
-				       enum sk_psock_state_bits bit)
-{
-	return test_bit(bit, &psock->state);
-}
-
 static inline struct sk_psock *sk_psock_get(struct sock *sk)
 {
 	struct sk_psock *psock;

kernel/bpf/core.c

@@ -32,6 +32,8 @@
 #include <linux/perf_event.h>
 #include <linux/extable.h>
 #include <linux/log2.h>
+
+#include <asm/barrier.h>
 #include <asm/unaligned.h>
 
 /* Registers */
@@ -1377,6 +1379,7 @@ static u64 ___bpf_prog_run(u64 *regs, const struct bpf_insn *insn)
 		/* Non-UAPI available opcodes. */
 		[BPF_JMP | BPF_CALL_ARGS] = &&JMP_CALL_ARGS,
 		[BPF_JMP | BPF_TAIL_CALL] = &&JMP_TAIL_CALL,
+		[BPF_ST  | BPF_NOSPEC] = &&ST_NOSPEC,
 		[BPF_LDX | BPF_PROBE_MEM | BPF_B] = &&LDX_PROBE_MEM_B,
 		[BPF_LDX | BPF_PROBE_MEM | BPF_H] = &&LDX_PROBE_MEM_H,
 		[BPF_LDX | BPF_PROBE_MEM | BPF_W] = &&LDX_PROBE_MEM_W,
@@ -1621,7 +1624,21 @@ static u64 ___bpf_prog_run(u64 *regs, const struct bpf_insn *insn)
 	COND_JMP(s, JSGE, >=)
 	COND_JMP(s, JSLE, <=)
 #undef COND_JMP
-	/* STX and ST and LDX*/
+	/* ST, STX and LDX*/
+	ST_NOSPEC:
+		/* Speculation barrier for mitigating Speculative Store Bypass.
+		 * In case of arm64, we rely on the firmware mitigation as
+		 * controlled via the ssbd kernel parameter. Whenever the
+		 * mitigation is enabled, it works for all of the kernel code
+		 * with no need to provide any additional instructions here.
+		 * In case of x86, we use 'lfence' insn for mitigation. We
+		 * reuse preexisting logic from Spectre v1 mitigation that
+		 * happens to produce the required code on x86 for v4 as well.
+		 */
+#ifdef CONFIG_X86
+		barrier_nospec();
+#endif
+		CONT;
 #define LDST(SIZEOP, SIZE)						\
 	STX_MEM_##SIZEOP:						\
 		*(SIZE *)(unsigned long) (DST + insn->off) = SRC;	\

kernel/bpf/disasm.c

@@ -206,15 +206,17 @@ void print_bpf_insn(const struct bpf_insn_cbs *cbs,
 			verbose(cbs->private_data, "BUG_%02x\n", insn->code);
 		}
 	} else if (class == BPF_ST) {
-		if (BPF_MODE(insn->code) != BPF_MEM) {
+		if (BPF_MODE(insn->code) == BPF_MEM) {
+			verbose(cbs->private_data, "(%02x) *(%s *)(r%d %+d) = %d\n",
+				insn->code,
+				bpf_ldst_string[BPF_SIZE(insn->code) >> 3],
+				insn->dst_reg,
+				insn->off, insn->imm);
+		} else if (BPF_MODE(insn->code) == 0xc0 /* BPF_NOSPEC, no UAPI */) {
+			verbose(cbs->private_data, "(%02x) nospec\n", insn->code);
+		} else {
 			verbose(cbs->private_data, "BUG_st_%02x\n", insn->code);
-			return;
 		}
-		verbose(cbs->private_data, "(%02x) *(%s *)(r%d %+d) = %d\n",
-			insn->code,
-			bpf_ldst_string[BPF_SIZE(insn->code) >> 3],
-			insn->dst_reg,
-			insn->off, insn->imm);
 	} else if (class == BPF_LDX) {
 		if (BPF_MODE(insn->code) != BPF_MEM) {
 			verbose(cbs->private_data, "BUG_ldx_%02x\n", insn->code);

kernel/bpf/verifier.c

@@ -2610,6 +2610,19 @@ static int check_stack_write_fixed_off(struct bpf_verifier_env *env,
 	cur = env->cur_state->frame[env->cur_state->curframe];
 	if (value_regno >= 0)
 		reg = &cur->regs[value_regno];
+	if (!env->bypass_spec_v4) {
+		bool sanitize = reg && is_spillable_regtype(reg->type);
+
+		for (i = 0; i < size; i++) {
+			if (state->stack[spi].slot_type[i] == STACK_INVALID) {
+				sanitize = true;
+				break;
+			}
+		}
+
+		if (sanitize)
+			env->insn_aux_data[insn_idx].sanitize_stack_spill = true;
+	}
 
 	if (reg && size == BPF_REG_SIZE && register_is_bounded(reg) &&
 	    !register_is_null(reg) && env->bpf_capable) {
@@ -2632,47 +2645,10 @@ static int check_stack_write_fixed_off(struct bpf_verifier_env *env,
 			verbose(env, "invalid size of register spill\n");
 			return -EACCES;
 		}
-
 		if (state != cur && reg->type == PTR_TO_STACK) {
 			verbose(env, "cannot spill pointers to stack into stack frame of the caller\n");
 			return -EINVAL;
 		}
-
-		if (!env->bypass_spec_v4) {
-			bool sanitize = false;
-
-			if (state->stack[spi].slot_type[0] == STACK_SPILL &&
-			    register_is_const(&state->stack[spi].spilled_ptr))
-				sanitize = true;
-			for (i = 0; i < BPF_REG_SIZE; i++)
-				if (state->stack[spi].slot_type[i] == STACK_MISC) {
-					sanitize = true;
-					break;
-				}
-			if (sanitize) {
-				int *poff = &env->insn_aux_data[insn_idx].sanitize_stack_off;
-				int soff = (-spi - 1) * BPF_REG_SIZE;
-
-				/* detected reuse of integer stack slot with a pointer
-				 * which means either llvm is reusing stack slot or
-				 * an attacker is trying to exploit CVE-2018-3639
-				 * (speculative store bypass)
-				 * Have to sanitize that slot with preemptive
-				 * store of zero.
-				 */
-				if (*poff && *poff != soff) {
-					/* disallow programs where single insn stores
-					 * into two different stack slots, since verifier
-					 * cannot sanitize them
-					 */
-					verbose(env,
-						"insn %d cannot access two stack slots fp%d and fp%d",
-						insn_idx, *poff, soff);
-					return -EINVAL;
-				}
-				*poff = soff;
-			}
-		}
 		save_register_state(state, spi, reg);
 	} else {
 		u8 type = STACK_MISC;
@@ -6561,6 +6537,12 @@ static int sanitize_ptr_alu(struct bpf_verifier_env *env,
 		alu_state |= off_is_imm ? BPF_ALU_IMMEDIATE : 0;
 		alu_state |= ptr_is_dst_reg ?
 			     BPF_ALU_SANITIZE_SRC : BPF_ALU_SANITIZE_DST;
+
+		/* Limit pruning on unknown scalars to enable deep search for
+		 * potential masking differences from other program paths.
+		 */
+		if (!off_is_imm)
+			env->explore_alu_limits = true;
 	}
 
 	err = update_alu_sanitation_state(aux, alu_state, alu_limit);
@@ -9936,8 +9918,8 @@ static void clean_live_states(struct bpf_verifier_env *env, int insn,
 }
 
 /* Returns true if (rold safe implies rcur safe) */
-static bool regsafe(struct bpf_reg_state *rold, struct bpf_reg_state *rcur,
-		    struct bpf_id_pair *idmap)
+static bool regsafe(struct bpf_verifier_env *env, struct bpf_reg_state *rold,
+		    struct bpf_reg_state *rcur, struct bpf_id_pair *idmap)
 {
 	bool equal;
 
@@ -9963,6 +9945,8 @@ static bool regsafe(struct bpf_reg_state *rold, struct bpf_reg_state *rcur,
 		return false;
 	switch (rold->type) {
 	case SCALAR_VALUE:
+		if (env->explore_alu_limits)
+			return false;
 		if (rcur->type == SCALAR_VALUE) {
 			if (!rold->precise && !rcur->precise)
 				return true;
@@ -10053,9 +10037,8 @@ static bool regsafe(struct bpf_reg_state *rold, struct bpf_reg_state *rcur,
 	return false;
 }
 
-static bool stacksafe(struct bpf_func_state *old,
-		      struct bpf_func_state *cur,
-		      struct bpf_id_pair *idmap)
+static bool stacksafe(struct bpf_verifier_env *env, struct bpf_func_state *old,
+		      struct bpf_func_state *cur, struct bpf_id_pair *idmap)
 {
 	int i, spi;
 
@@ -10100,9 +10083,8 @@ static bool stacksafe(struct bpf_func_state *old,
 			continue;
 		if (old->stack[spi].slot_type[0] != STACK_SPILL)
 			continue;
-		if (!regsafe(&old->stack[spi].spilled_ptr,
-			     &cur->stack[spi].spilled_ptr,
-			     idmap))
+		if (!regsafe(env, &old->stack[spi].spilled_ptr,
+			     &cur->stack[spi].spilled_ptr, idmap))
 			/* when explored and current stack slot are both storing
 			 * spilled registers, check that stored pointers types
 			 * are the same as well.
@@ -10159,10 +10141,11 @@ static bool func_states_equal(struct bpf_verifier_env *env, struct bpf_func_stat
 
 	memset(env->idmap_scratch, 0, sizeof(env->idmap_scratch));
 	for (i = 0; i < MAX_BPF_REG; i++)
-		if (!regsafe(&old->regs[i], &cur->regs[i], env->idmap_scratch))
+		if (!regsafe(env, &old->regs[i], &cur->regs[i],
+			     env->idmap_scratch))
 			return false;
 
-	if (!stacksafe(old, cur, env->idmap_scratch))
+	if (!stacksafe(env, old, cur, env->idmap_scratch))
 		return false;
 
 	if (!refsafe(old, cur))
@@ -11906,35 +11889,33 @@ static int convert_ctx_accesses(struct bpf_verifier_env *env)
 
 	for (i = 0; i < insn_cnt; i++, insn++) {
 		bpf_convert_ctx_access_t convert_ctx_access;
+		bool ctx_access;
 
 		if (insn->code == (BPF_LDX | BPF_MEM | BPF_B) ||
 		    insn->code == (BPF_LDX | BPF_MEM | BPF_H) ||
 		    insn->code == (BPF_LDX | BPF_MEM | BPF_W) ||
-		    insn->code == (BPF_LDX | BPF_MEM | BPF_DW))
+		    insn->code == (BPF_LDX | BPF_MEM | BPF_DW)) {
 			type = BPF_READ;
-		else if (insn->code == (BPF_STX | BPF_MEM | BPF_B) ||
-			 insn->code == (BPF_STX | BPF_MEM | BPF_H) ||
-			 insn->code == (BPF_STX | BPF_MEM | BPF_W) ||
-			 insn->code == (BPF_STX | BPF_MEM | BPF_DW))
+			ctx_access = true;
+		} else if (insn->code == (BPF_STX | BPF_MEM | BPF_B) ||
+			   insn->code == (BPF_STX | BPF_MEM | BPF_H) ||
+			   insn->code == (BPF_STX | BPF_MEM | BPF_W) ||
+			   insn->code == (BPF_STX | BPF_MEM | BPF_DW) ||
+			   insn->code == (BPF_ST | BPF_MEM | BPF_B) ||
+			   insn->code == (BPF_ST | BPF_MEM | BPF_H) ||
+			   insn->code == (BPF_ST | BPF_MEM | BPF_W) ||
+			   insn->code == (BPF_ST | BPF_MEM | BPF_DW)) {
 			type = BPF_WRITE;
-		else
+			ctx_access = BPF_CLASS(insn->code) == BPF_STX;
+		} else {
 			continue;
+		}
 
 		if (type == BPF_WRITE &&
-		    env->insn_aux_data[i + delta].sanitize_stack_off) {
+		    env->insn_aux_data[i + delta].sanitize_stack_spill) {
 			struct bpf_insn patch[] = {
-				/* Sanitize suspicious stack slot with zero.
-				 * There are no memory dependencies for this store,
-				 * since it's only using frame pointer and immediate
-				 * constant of zero
-				 */
-				BPF_ST_MEM(BPF_DW, BPF_REG_FP,
-					   env->insn_aux_data[i + delta].sanitize_stack_off,
-					   0),
-				/* the original STX instruction will immediately
-				 * overwrite the same stack slot with appropriate value
-				 */
 				*insn,
+				BPF_ST_NOSPEC(),
 			};
 
 			cnt = ARRAY_SIZE(patch);
@@ -11948,6 +11929,9 @@ static int convert_ctx_accesses(struct bpf_verifier_env *env)
 			continue;
 		}
 
+		if (!ctx_access)
+			continue;
+
 		switch (env->insn_aux_data[i + delta].ptr_type) {
 		case PTR_TO_CTX:
 			if (!ops->convert_ctx_access)
@@ -12752,37 +12736,6 @@ static void free_states(struct bpf_verifier_env *env)
 	}
 }
 
-/* The verifier is using insn_aux_data[] to store temporary data during
- * verification and to store information for passes that run after the
- * verification like dead code sanitization. do_check_common() for subprogram N
- * may analyze many other subprograms. sanitize_insn_aux_data() clears all
- * temporary data after do_check_common() finds that subprogram N cannot be
- * verified independently. pass_cnt counts the number of times
- * do_check_common() was run and insn->aux->seen tells the pass number
- * insn_aux_data was touched. These variables are compared to clear temporary
- * data from failed pass. For testing and experiments do_check_common() can be
- * run multiple times even when prior attempt to verify is unsuccessful.
- *
- * Note that special handling is needed on !env->bypass_spec_v1 if this is
- * ever called outside of error path with subsequent program rejection.
- */
-static void sanitize_insn_aux_data(struct bpf_verifier_env *env)
-{
-	struct bpf_insn *insn = env->prog->insnsi;
-	struct bpf_insn_aux_data *aux;
-	int i, class;
-
-	for (i = 0; i < env->prog->len; i++) {
-		class = BPF_CLASS(insn[i].code);
-		if (class != BPF_LDX && class != BPF_STX)
-			continue;
-		aux = &env->insn_aux_data[i];
-		if (aux->seen != env->pass_cnt)
-			continue;
-		memset(aux, 0, offsetof(typeof(*aux), orig_idx));
-	}
-}
-
 static int do_check_common(struct bpf_verifier_env *env, int subprog)
 {
 	bool pop_log = !(env->log.level & BPF_LOG_LEVEL2);
@@ -12859,9 +12812,6 @@ static int do_check_common(struct bpf_verifier_env *env, int subprog)
 	if (!ret && pop_log)
 		bpf_vlog_reset(&env->log, 0);
 	free_states(env);
-	if (ret)
-		/* clean aux data in case subprog was rejected */
-		sanitize_insn_aux_data(env);
 	return ret;
 }
 

net/core/skmsg.c

@@ -584,29 +584,42 @@ static int sk_psock_handle_skb(struct sk_psock *psock, struct sk_buff *skb,
 	return sk_psock_skb_ingress(psock, skb);
 }
 
-static void sock_drop(struct sock *sk, struct sk_buff *skb)
+static void sk_psock_skb_state(struct sk_psock *psock,
+			       struct sk_psock_work_state *state,
+			       struct sk_buff *skb,
+			       int len, int off)
 {
-	sk_drops_add(sk, skb);
-	kfree_skb(skb);
+	spin_lock_bh(&psock->ingress_lock);
+	if (sk_psock_test_state(psock, SK_PSOCK_TX_ENABLED)) {
+		state->skb = skb;
+		state->len = len;
+		state->off = off;
+	} else {
+		sock_drop(psock->sk, skb);
+	}
+	spin_unlock_bh(&psock->ingress_lock);
 }
 
 static void sk_psock_backlog(struct work_struct *work)
 {
 	struct sk_psock *psock = container_of(work, struct sk_psock, work);
 	struct sk_psock_work_state *state = &psock->work_state;
-	struct sk_buff *skb;
+	struct sk_buff *skb = NULL;
 	bool ingress;
 	u32 len, off;
 	int ret;
 
 	mutex_lock(&psock->work_mutex);
-	if (state->skb) {
+	if (unlikely(state->skb)) {
+		spin_lock_bh(&psock->ingress_lock);
 		skb = state->skb;
 		len = state->len;
 		off = state->off;
 		state->skb = NULL;
-		goto start;
+		spin_unlock_bh(&psock->ingress_lock);
 	}
+	if (skb)
+		goto start;
 
 	while ((skb = skb_dequeue(&psock->ingress_skb))) {
 		len = skb->len;
@@ -621,9 +634,8 @@ static void sk_psock_backlog(struct work_struct *work)
 							  len, ingress);
 			if (ret <= 0) {
 				if (ret == -EAGAIN) {
-					state->skb = skb;
-					state->len = len;
-					state->off = off;
+					sk_psock_skb_state(psock, state, skb,
+							   len, off);
 					goto end;
 				}
 				/* Hard errors break pipe and stop xmit. */
@@ -722,6 +734,11 @@ static void __sk_psock_zap_ingress(struct sk_psock *psock)
 		skb_bpf_redirect_clear(skb);
 		sock_drop(psock->sk, skb);
 	}
+	kfree_skb(psock->work_state.skb);
+	/* We null the skb here to ensure that calls to sk_psock_backlog
+	 * do not pick up the free'd skb.
+	 */
+	psock->work_state.skb = NULL;
 	__sk_psock_purge_ingress_msg(psock);
 }
 
@@ -773,8 +790,6 @@ static void sk_psock_destroy(struct work_struct *work)
 
 void sk_psock_drop(struct sock *sk, struct sk_psock *psock)
 {
-	sk_psock_stop(psock, false);
-
 	write_lock_bh(&sk->sk_callback_lock);
 	sk_psock_restore_proto(sk, psock);
 	rcu_assign_sk_user_data(sk, NULL);
@@ -784,6 +799,8 @@ void sk_psock_drop(struct sock *sk, struct sk_psock *psock)
 		sk_psock_stop_verdict(sk, psock);
 	write_unlock_bh(&sk->sk_callback_lock);
 
+	sk_psock_stop(psock, false);
+
 	INIT_RCU_WORK(&psock->rwork, sk_psock_destroy);
 	queue_rcu_work(system_wq, &psock->rwork);
 }

tools/testing/selftests/bpf/verifier/value_ptr_arith.c

+{
+	"map access: known scalar += value_ptr unknown vs const",
+	.insns = {
+	BPF_LDX_MEM(BPF_W, BPF_REG_0, BPF_REG_1,
+		    offsetof(struct __sk_buff, len)),
+	BPF_ST_MEM(BPF_DW, BPF_REG_10, -8, 0),
+	BPF_MOV64_REG(BPF_REG_2, BPF_REG_10),
+	BPF_ALU64_IMM(BPF_ADD, BPF_REG_2, -8),
+	BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 1, 3),
+	BPF_LD_MAP_FD(BPF_REG_1, 0),
+	BPF_JMP_IMM(BPF_JNE, BPF_REG_0, 1, 2),
+	BPF_LD_MAP_FD(BPF_REG_1, 0),
+	BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0, BPF_FUNC_map_lookup_elem),
+	BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 0, 9),
+	BPF_LDX_MEM(BPF_B, BPF_REG_4, BPF_REG_0, 0),
+	BPF_JMP_IMM(BPF_JEQ, BPF_REG_4, 1, 4),
+	BPF_MOV64_IMM(BPF_REG_1, 6),
+	BPF_ALU64_IMM(BPF_NEG, BPF_REG_1, 0),
+	BPF_ALU64_IMM(BPF_AND, BPF_REG_1, 0x7),
+	BPF_JMP_IMM(BPF_JA, 0, 0, 1),
+	BPF_MOV64_IMM(BPF_REG_1, 3),
+	BPF_ALU64_REG(BPF_ADD, BPF_REG_1, BPF_REG_0),
+	BPF_LDX_MEM(BPF_B, BPF_REG_0, BPF_REG_1, 0),
+	BPF_MOV64_IMM(BPF_REG_0, 1),
+	BPF_EXIT_INSN(),
+	},
+	.fixup_map_hash_16b = { 5 },
+	.fixup_map_array_48b = { 8 },
+	.result_unpriv = REJECT,
+	.errstr_unpriv = "R1 tried to add from different maps, paths or scalars",
+	.result = ACCEPT,
+	.retval = 1,
+},
+{
+	"map access: known scalar += value_ptr const vs unknown",
+	.insns = {
+	BPF_LDX_MEM(BPF_W, BPF_REG_0, BPF_REG_1,
+		    offsetof(struct __sk_buff, len)),
+	BPF_ST_MEM(BPF_DW, BPF_REG_10, -8, 0),
+	BPF_MOV64_REG(BPF_REG_2, BPF_REG_10),
+	BPF_ALU64_IMM(BPF_ADD, BPF_REG_2, -8),
+	BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 1, 3),
+	BPF_LD_MAP_FD(BPF_REG_1, 0),
+	BPF_JMP_IMM(BPF_JNE, BPF_REG_0, 1, 2),
+	BPF_LD_MAP_FD(BPF_REG_1, 0),
+	BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0, BPF_FUNC_map_lookup_elem),
+	BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 0, 9),
+	BPF_LDX_MEM(BPF_B, BPF_REG_4, BPF_REG_0, 0),
+	BPF_JMP_IMM(BPF_JEQ, BPF_REG_4, 1, 2),
+	BPF_MOV64_IMM(BPF_REG_1, 3),
+	BPF_JMP_IMM(BPF_JA, 0, 0, 3),
+	BPF_MOV64_IMM(BPF_REG_1, 6),
+	BPF_ALU64_IMM(BPF_NEG, BPF_REG_1, 0),
+	BPF_ALU64_IMM(BPF_AND, BPF_REG_1, 0x7),
+	BPF_ALU64_REG(BPF_ADD, BPF_REG_1, BPF_REG_0),
+	BPF_LDX_MEM(BPF_B, BPF_REG_0, BPF_REG_1, 0),
+	BPF_MOV64_IMM(BPF_REG_0, 1),
+	BPF_EXIT_INSN(),
+	},
+	.fixup_map_hash_16b = { 5 },
+	.fixup_map_array_48b = { 8 },
+	.result_unpriv = REJECT,
+	.errstr_unpriv = "R1 tried to add from different maps, paths or scalars",
+	.result = ACCEPT,
+	.retval = 1,
+},
+{
+	"map access: known scalar += value_ptr const vs const (ne)",
+	.insns = {
+	BPF_LDX_MEM(BPF_W, BPF_REG_0, BPF_REG_1,
+		    offsetof(struct __sk_buff, len)),
+	BPF_ST_MEM(BPF_DW, BPF_REG_10, -8, 0),
+	BPF_MOV64_REG(BPF_REG_2, BPF_REG_10),
+	BPF_ALU64_IMM(BPF_ADD, BPF_REG_2, -8),
+	BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 1, 3),
+	BPF_LD_MAP_FD(BPF_REG_1, 0),
+	BPF_JMP_IMM(BPF_JNE, BPF_REG_0, 1, 2),
+	BPF_LD_MAP_FD(BPF_REG_1, 0),
+	BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0, BPF_FUNC_map_lookup_elem),
+	BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 0, 7),
+	BPF_LDX_MEM(BPF_B, BPF_REG_4, BPF_REG_0, 0),
+	BPF_JMP_IMM(BPF_JEQ, BPF_REG_4, 1, 2),
+	BPF_MOV64_IMM(BPF_REG_1, 3),
+	BPF_JMP_IMM(BPF_JA, 0, 0, 1),
+	BPF_MOV64_IMM(BPF_REG_1, 5),
+	BPF_ALU64_REG(BPF_ADD, BPF_REG_1, BPF_REG_0),
+	BPF_LDX_MEM(BPF_B, BPF_REG_0, BPF_REG_1, 0),
+	BPF_MOV64_IMM(BPF_REG_0, 1),
+	BPF_EXIT_INSN(),
+	},
+	.fixup_map_hash_16b = { 5 },
+	.fixup_map_array_48b = { 8 },
+	.result_unpriv = REJECT,
+	.errstr_unpriv = "R1 tried to add from different maps, paths or scalars",
+	.result = ACCEPT,
+	.retval = 1,
+},
+{
+	"map access: known scalar += value_ptr const vs const (eq)",
+	.insns = {
+	BPF_LDX_MEM(BPF_W, BPF_REG_0, BPF_REG_1,
+		    offsetof(struct __sk_buff, len)),
+	BPF_ST_MEM(BPF_DW, BPF_REG_10, -8, 0),
+	BPF_MOV64_REG(BPF_REG_2, BPF_REG_10),
+	BPF_ALU64_IMM(BPF_ADD, BPF_REG_2, -8),
+	BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 1, 3),
+	BPF_LD_MAP_FD(BPF_REG_1, 0),
+	BPF_JMP_IMM(BPF_JNE, BPF_REG_0, 1, 2),
+	BPF_LD_MAP_FD(BPF_REG_1, 0),
+	BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0, BPF_FUNC_map_lookup_elem),
+	BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 0, 7),
+	BPF_LDX_MEM(BPF_B, BPF_REG_4, BPF_REG_0, 0),
+	BPF_JMP_IMM(BPF_JEQ, BPF_REG_4, 1, 2),
+	BPF_MOV64_IMM(BPF_REG_1, 5),
+	BPF_JMP_IMM(BPF_JA, 0, 0, 1),
+	BPF_MOV64_IMM(BPF_REG_1, 5),
+	BPF_ALU64_REG(BPF_ADD, BPF_REG_1, BPF_REG_0),
+	BPF_LDX_MEM(BPF_B, BPF_REG_0, BPF_REG_1, 0),
+	BPF_MOV64_IMM(BPF_REG_0, 1),
+	BPF_EXIT_INSN(),
+	},
+	.fixup_map_hash_16b = { 5 },
+	.fixup_map_array_48b = { 8 },
+	.result = ACCEPT,
+	.retval = 1,
+},
+{
+	"map access: known scalar += value_ptr unknown vs unknown (eq)",
+	.insns = {
+	BPF_LDX_MEM(BPF_W, BPF_REG_0, BPF_REG_1,
+		    offsetof(struct __sk_buff, len)),
+	BPF_ST_MEM(BPF_DW, BPF_REG_10, -8, 0),
+	BPF_MOV64_REG(BPF_REG_2, BPF_REG_10),
+	BPF_ALU64_IMM(BPF_ADD, BPF_REG_2, -8),
+	BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 1, 3),
+	BPF_LD_MAP_FD(BPF_REG_1, 0),
+	BPF_JMP_IMM(BPF_JNE, BPF_REG_0, 1, 2),
+	BPF_LD_MAP_FD(BPF_REG_1, 0),
+	BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0, BPF_FUNC_map_lookup_elem),
+	BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 0, 11),
+	BPF_LDX_MEM(BPF_B, BPF_REG_4, BPF_REG_0, 0),
+	BPF_JMP_IMM(BPF_JEQ, BPF_REG_4, 1, 4),
+	BPF_MOV64_IMM(BPF_REG_1, 6),
+	BPF_ALU64_IMM(BPF_NEG, BPF_REG_1, 0),
+	BPF_ALU64_IMM(BPF_AND, BPF_REG_1, 0x7),
+	BPF_JMP_IMM(BPF_JA, 0, 0, 3),
+	BPF_MOV64_IMM(BPF_REG_1, 6),
+	BPF_ALU64_IMM(BPF_NEG, BPF_REG_1, 0),
+	BPF_ALU64_IMM(BPF_AND, BPF_REG_1, 0x7),
+	BPF_ALU64_REG(BPF_ADD, BPF_REG_1, BPF_REG_0),
+	BPF_LDX_MEM(BPF_B, BPF_REG_0, BPF_REG_1, 0),
+	BPF_MOV64_IMM(BPF_REG_0, 1),
+	BPF_EXIT_INSN(),
+	},
+	.fixup_map_hash_16b = { 5 },
+	.fixup_map_array_48b = { 8 },
+	.result = ACCEPT,
+	.retval = 1,
+},
+{
+	"map access: known scalar += value_ptr unknown vs unknown (lt)",
+	.insns = {
+	BPF_LDX_MEM(BPF_W, BPF_REG_0, BPF_REG_1,
+		    offsetof(struct __sk_buff, len)),
+	BPF_ST_MEM(BPF_DW, BPF_REG_10, -8, 0),
+	BPF_MOV64_REG(BPF_REG_2, BPF_REG_10),
+	BPF_ALU64_IMM(BPF_ADD, BPF_REG_2, -8),
+	BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 1, 3),
+	BPF_LD_MAP_FD(BPF_REG_1, 0),
+	BPF_JMP_IMM(BPF_JNE, BPF_REG_0, 1, 2),
+	BPF_LD_MAP_FD(BPF_REG_1, 0),
+	BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0, BPF_FUNC_map_lookup_elem),
+	BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 0, 11),
+	BPF_LDX_MEM(BPF_B, BPF_REG_4, BPF_REG_0, 0),
+	BPF_JMP_IMM(BPF_JEQ, BPF_REG_4, 1, 4),
+	BPF_MOV64_IMM(BPF_REG_1, 6),
+	BPF_ALU64_IMM(BPF_NEG, BPF_REG_1, 0),
+	BPF_ALU64_IMM(BPF_AND, BPF_REG_1, 0x3),
+	BPF_JMP_IMM(BPF_JA, 0, 0, 3),
+	BPF_MOV64_IMM(BPF_REG_1, 6),
+	BPF_ALU64_IMM(BPF_NEG, BPF_REG_1, 0),
+	BPF_ALU64_IMM(BPF_AND, BPF_REG_1, 0x7),
+	BPF_ALU64_REG(BPF_ADD, BPF_REG_1, BPF_REG_0),
+	BPF_LDX_MEM(BPF_B, BPF_REG_0, BPF_REG_1, 0),
+	BPF_MOV64_IMM(BPF_REG_0, 1),
+	BPF_EXIT_INSN(),
+	},
+	.fixup_map_hash_16b = { 5 },
+	.fixup_map_array_48b = { 8 },
+	.result_unpriv = REJECT,
+	.errstr_unpriv = "R1 tried to add from different maps, paths or scalars",
+	.result = ACCEPT,
+	.retval = 1,
+},
+{
+	"map access: known scalar += value_ptr unknown vs unknown (gt)",
+	.insns = {
+	BPF_LDX_MEM(BPF_W, BPF_REG_0, BPF_REG_1,
+		    offsetof(struct __sk_buff, len)),
+	BPF_ST_MEM(BPF_DW, BPF_REG_10, -8, 0),
+	BPF_MOV64_REG(BPF_REG_2, BPF_REG_10),
+	BPF_ALU64_IMM(BPF_ADD, BPF_REG_2, -8),
+	BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 1, 3),
+	BPF_LD_MAP_FD(BPF_REG_1, 0),
+	BPF_JMP_IMM(BPF_JNE, BPF_REG_0, 1, 2),
+	BPF_LD_MAP_FD(BPF_REG_1, 0),
+	BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0, BPF_FUNC_map_lookup_elem),
+	BPF_JMP_IMM(BPF_JEQ, BPF_REG_0, 0, 11),
+	BPF_LDX_MEM(BPF_B, BPF_REG_4, BPF_REG_0, 0),
+	BPF_JMP_IMM(BPF_JEQ, BPF_REG_4, 1, 4),
+	BPF_MOV64_IMM(BPF_REG_1, 6),
+	BPF_ALU64_IMM(BPF_NEG, BPF_REG_1, 0),
+	BPF_ALU64_IMM(BPF_AND, BPF_REG_1, 0x7),
+	BPF_JMP_IMM(BPF_JA, 0, 0, 3),
+	BPF_MOV64_IMM(BPF_REG_1, 6),
+	BPF_ALU64_IMM(BPF_NEG, BPF_REG_1, 0),
+	BPF_ALU64_IMM(BPF_AND, BPF_REG_1, 0x3),
+	BPF_ALU64_REG(BPF_ADD, BPF_REG_1, BPF_REG_0),
+	BPF_LDX_MEM(BPF_B, BPF_REG_0, BPF_REG_1, 0),
+	BPF_MOV64_IMM(BPF_REG_0, 1),
+	BPF_EXIT_INSN(),
+	},
+	.fixup_map_hash_16b = { 5 },
+	.fixup_map_array_48b = { 8 },
+	.result_unpriv = REJECT,
+	.errstr_unpriv = "R1 tried to add from different maps, paths or scalars",
+	.result = ACCEPT,
+	.retval = 1,
+},
 {
 	"map access: known scalar += value_ptr from different maps",
 	.insns = {