Merge branch 'bpf-push-pop-helpers'

Alexei Starovoitov says:

====================
bpf: introduce bpf_skb_vlan_push/pop() helpers

Let TC+eBPF programs call skb_vlan_push/pop via helpers.

v1->v2:
- reworded commit log to better explain correctness of re-caching
  and fixed comparison of mixed endiannes (suggested by Eric)
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

arch/s390/net/bpf_jit_comp.c

@@ -973,6 +973,10 @@ static noinline int bpf_jit_insn(struct bpf_jit *jit, struct bpf_prog *fp, int i
 		 */
 		const u64 func = (u64)__bpf_call_base + imm;
 
+		if (bpf_helper_changes_skb_data((void *)func))
+			/* TODO reload skb->data, hlen */
+			return -1;
+
 		REG_SET_SEEN(BPF_REG_5);
 		jit->seen |= SEEN_FUNC;
 		/* lg %w1,<d(imm)>(%l) */

arch/x86/net/bpf_jit_comp.c

@@ -315,6 +315,26 @@ static void emit_bpf_tail_call(u8 **pprog)
 	*pprog = prog;
 }
 
+
+static void emit_load_skb_data_hlen(u8 **pprog)
+{
+	u8 *prog = *pprog;
+	int cnt = 0;
+
+	/* r9d = skb->len - skb->data_len (headlen)
+	 * r10 = skb->data
+	 */
+	/* mov %r9d, off32(%rdi) */
+	EMIT3_off32(0x44, 0x8b, 0x8f, offsetof(struct sk_buff, len));
+
+	/* sub %r9d, off32(%rdi) */
+	EMIT3_off32(0x44, 0x2b, 0x8f, offsetof(struct sk_buff, data_len));
+
+	/* mov %r10, off32(%rdi) */
+	EMIT3_off32(0x4c, 0x8b, 0x97, offsetof(struct sk_buff, data));
+	*pprog = prog;
+}
+
 static int do_jit(struct bpf_prog *bpf_prog, int *addrs, u8 *image,
 		  int oldproglen, struct jit_context *ctx)
 {
@@ -329,36 +349,8 @@ static int do_jit(struct bpf_prog *bpf_prog, int *addrs, u8 *image,
 
 	emit_prologue(&prog);
 
-	if (seen_ld_abs) {
-		/* r9d : skb->len - skb->data_len (headlen)
-		 * r10 : skb->data
-		 */
-		if (is_imm8(offsetof(struct sk_buff, len)))
-			/* mov %r9d, off8(%rdi) */
-			EMIT4(0x44, 0x8b, 0x4f,
-			      offsetof(struct sk_buff, len));
-		else
-			/* mov %r9d, off32(%rdi) */
-			EMIT3_off32(0x44, 0x8b, 0x8f,
-				    offsetof(struct sk_buff, len));
-
-		if (is_imm8(offsetof(struct sk_buff, data_len)))
-			/* sub %r9d, off8(%rdi) */
-			EMIT4(0x44, 0x2b, 0x4f,
-			      offsetof(struct sk_buff, data_len));
-		else
-			EMIT3_off32(0x44, 0x2b, 0x8f,
-				    offsetof(struct sk_buff, data_len));
-
-		if (is_imm8(offsetof(struct sk_buff, data)))
-			/* mov %r10, off8(%rdi) */
-			EMIT4(0x4c, 0x8b, 0x57,
-			      offsetof(struct sk_buff, data));
-		else
-			/* mov %r10, off32(%rdi) */
-			EMIT3_off32(0x4c, 0x8b, 0x97,
-				    offsetof(struct sk_buff, data));
-	}
+	if (seen_ld_abs)
+		emit_load_skb_data_hlen(&prog);
 
 	for (i = 0; i < insn_cnt; i++, insn++) {
 		const s32 imm32 = insn->imm;
@@ -367,6 +359,7 @@ static int do_jit(struct bpf_prog *bpf_prog, int *addrs, u8 *image,
 		u8 b1 = 0, b2 = 0, b3 = 0;
 		s64 jmp_offset;
 		u8 jmp_cond;
+		bool reload_skb_data;
 		int ilen;
 		u8 *func;
 
@@ -818,12 +811,18 @@ xadd:			if (is_imm8(insn->off))
 			func = (u8 *) __bpf_call_base + imm32;
 			jmp_offset = func - (image + addrs[i]);
 			if (seen_ld_abs) {
-				EMIT2(0x41, 0x52); /* push %r10 */
-				EMIT2(0x41, 0x51); /* push %r9 */
-				/* need to adjust jmp offset, since
-				 * pop %r9, pop %r10 take 4 bytes after call insn
-				 */
-				jmp_offset += 4;
+				reload_skb_data = bpf_helper_changes_skb_data(func);
+				if (reload_skb_data) {
+					EMIT1(0x57); /* push %rdi */
+					jmp_offset += 22; /* pop, mov, sub, mov */
+				} else {
+					EMIT2(0x41, 0x52); /* push %r10 */
+					EMIT2(0x41, 0x51); /* push %r9 */
+					/* need to adjust jmp offset, since
+					 * pop %r9, pop %r10 take 4 bytes after call insn
+					 */
+					jmp_offset += 4;
+				}
 			}
 			if (!imm32 || !is_simm32(jmp_offset)) {
 				pr_err("unsupported bpf func %d addr %p image %p\n",
@@ -832,8 +831,13 @@ xadd:			if (is_imm8(insn->off))
 			}
 			EMIT1_off32(0xE8, jmp_offset);
 			if (seen_ld_abs) {
-				EMIT2(0x41, 0x59); /* pop %r9 */
-				EMIT2(0x41, 0x5A); /* pop %r10 */
+				if (reload_skb_data) {
+					EMIT1(0x5F); /* pop %rdi */
+					emit_load_skb_data_hlen(&prog);
+				} else {
+					EMIT2(0x41, 0x59); /* pop %r9 */
+					EMIT2(0x41, 0x5A); /* pop %r10 */
+				}
 			}
 			break;
 

include/linux/bpf.h

@@ -192,5 +192,7 @@ extern const struct bpf_func_proto bpf_ktime_get_ns_proto;
 extern const struct bpf_func_proto bpf_get_current_pid_tgid_proto;
 extern const struct bpf_func_proto bpf_get_current_uid_gid_proto;
 extern const struct bpf_func_proto bpf_get_current_comm_proto;
+extern const struct bpf_func_proto bpf_skb_vlan_push_proto;
+extern const struct bpf_func_proto bpf_skb_vlan_pop_proto;
 
 #endif /* _LINUX_BPF_H */

include/linux/filter.h

@@ -411,6 +411,7 @@ void sk_filter_uncharge(struct sock *sk, struct sk_filter *fp);
 
 u64 __bpf_call_base(u64 r1, u64 r2, u64 r3, u64 r4, u64 r5);
 void bpf_int_jit_compile(struct bpf_prog *fp);
+bool bpf_helper_changes_skb_data(void *func);
 
 #ifdef CONFIG_BPF_JIT
 typedef void (*bpf_jit_fill_hole_t)(void *area, unsigned int size);

include/uapi/linux/bpf.h

@@ -256,6 +256,8 @@ enum bpf_func_id {
 	 * Return: classid if != 0
 	 */
 	BPF_FUNC_get_cgroup_classid,
+	BPF_FUNC_skb_vlan_push, /* bpf_skb_vlan_push(skb, vlan_proto, vlan_tci) */
+	BPF_FUNC_skb_vlan_pop,  /* bpf_skb_vlan_pop(skb) */
 	__BPF_FUNC_MAX_ID,
 };
 

kernel/bpf/core.c

@@ -177,6 +177,7 @@ noinline u64 __bpf_call_base(u64 r1, u64 r2, u64 r3, u64 r4, u64 r5)
 {
 	return 0;
 }
+EXPORT_SYMBOL_GPL(__bpf_call_base);
 
 /**
  *	__bpf_prog_run - run eBPF program on a given context

lib/test_bpf.c

@@ -18,6 +18,7 @@
 #include <linux/init.h>
 #include <linux/module.h>
 #include <linux/filter.h>
+#include <linux/bpf.h>
 #include <linux/skbuff.h>
 #include <linux/netdevice.h>
 #include <linux/if_vlan.h>
@@ -355,6 +356,81 @@ static int bpf_fill_ja(struct bpf_test *self)
 	return __bpf_fill_ja(self, 12, 9);
 }
 
+static int bpf_fill_ld_abs_get_processor_id(struct bpf_test *self)
+{
+	unsigned int len = BPF_MAXINSNS;
+	struct sock_filter *insn;
+	int i;
+
+	insn = kmalloc_array(len, sizeof(*insn), GFP_KERNEL);
+	if (!insn)
+		return -ENOMEM;
+
+	for (i = 0; i < len - 1; i += 2) {
+		insn[i] = __BPF_STMT(BPF_LD | BPF_B | BPF_ABS, 0);
+		insn[i + 1] = __BPF_STMT(BPF_LD | BPF_W | BPF_ABS,
+					 SKF_AD_OFF + SKF_AD_CPU);
+	}
+
+	insn[len - 1] = __BPF_STMT(BPF_RET | BPF_K, 0xbee);
+
+	self->u.ptr.insns = insn;
+	self->u.ptr.len = len;
+
+	return 0;
+}
+
+#define PUSH_CNT 68
+/* test: {skb->data[0], vlan_push} x 68 + {skb->data[0], vlan_pop} x 68 */
+static int bpf_fill_ld_abs_vlan_push_pop(struct bpf_test *self)
+{
+	unsigned int len = BPF_MAXINSNS;
+	struct bpf_insn *insn;
+	int i = 0, j, k = 0;
+
+	insn = kmalloc_array(len, sizeof(*insn), GFP_KERNEL);
+	if (!insn)
+		return -ENOMEM;
+
+	insn[i++] = BPF_MOV64_REG(R6, R1);
+loop:
+	for (j = 0; j < PUSH_CNT; j++) {
+		insn[i++] = BPF_LD_ABS(BPF_B, 0);
+		insn[i] = BPF_JMP_IMM(BPF_JNE, R0, 0x34, len - i - 2);
+		i++;
+		insn[i++] = BPF_MOV64_REG(R1, R6);
+		insn[i++] = BPF_MOV64_IMM(R2, 1);
+		insn[i++] = BPF_MOV64_IMM(R3, 2);
+		insn[i++] = BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0,
+					 bpf_skb_vlan_push_proto.func - __bpf_call_base);
+		insn[i] = BPF_JMP_IMM(BPF_JNE, R0, 0, len - i - 2);
+		i++;
+	}
+
+	for (j = 0; j < PUSH_CNT; j++) {
+		insn[i++] = BPF_LD_ABS(BPF_B, 0);
+		insn[i] = BPF_JMP_IMM(BPF_JNE, R0, 0x34, len - i - 2);
+		i++;
+		insn[i++] = BPF_MOV64_REG(R1, R6);
+		insn[i++] = BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0,
+					 bpf_skb_vlan_pop_proto.func - __bpf_call_base);
+		insn[i] = BPF_JMP_IMM(BPF_JNE, R0, 0, len - i - 2);
+		i++;
+	}
+	if (++k < 5)
+		goto loop;
+
+	for (; i < len - 1; i++)
+		insn[i] = BPF_ALU32_IMM(BPF_MOV, R0, 0xbef);
+
+	insn[len - 1] = BPF_EXIT_INSN();
+
+	self->u.ptr.insns = insn;
+	self->u.ptr.len = len;
+
+	return 0;
+}
+
 static struct bpf_test tests[] = {
 	{
 		"TAX",
@@ -4398,6 +4474,22 @@ static struct bpf_test tests[] = {
 		{ { 0, 0xababcbac } },
 		.fill_helper = bpf_fill_maxinsns11,
 	},
+	{
+		"BPF_MAXINSNS: ld_abs+get_processor_id",
+		{ },
+		CLASSIC,
+		{ },
+		{ { 1, 0xbee } },
+		.fill_helper = bpf_fill_ld_abs_get_processor_id,
+	},
+	{
+		"BPF_MAXINSNS: ld_abs+vlan_push/pop",
+		{ },
+		INTERNAL,
+		{ 0x34 },
+		{ { 1, 0xbef } },
+		.fill_helper = bpf_fill_ld_abs_vlan_push_pop,
+	},
 };
 
 static struct net_device dev;
@@ -4551,14 +4643,14 @@ static int __run_one(const struct bpf_prog *fp, const void *data,
 	u64 start, finish;
 	int ret = 0, i;
 
-	start = ktime_to_us(ktime_get());
+	start = ktime_get_ns();
 
 	for (i = 0; i < runs; i++)
 		ret = BPF_PROG_RUN(fp, data);
 
-	finish = ktime_to_us(ktime_get());
+	finish = ktime_get_ns();
 
-	*duration = (finish - start) * 1000ULL;
+	*duration = finish - start;
 	do_div(*duration, runs);
 
 	return ret;

net/core/filter.c

@@ -1437,6 +1437,52 @@ static const struct bpf_func_proto bpf_get_cgroup_classid_proto = {
 	.arg1_type      = ARG_PTR_TO_CTX,
 };
 
+static u64 bpf_skb_vlan_push(u64 r1, u64 r2, u64 vlan_tci, u64 r4, u64 r5)
+{
+	struct sk_buff *skb = (struct sk_buff *) (long) r1;
+	__be16 vlan_proto = (__force __be16) r2;
+
+	if (unlikely(vlan_proto != htons(ETH_P_8021Q) &&
+		     vlan_proto != htons(ETH_P_8021AD)))
+		vlan_proto = htons(ETH_P_8021Q);
+
+	return skb_vlan_push(skb, vlan_proto, vlan_tci);
+}
+
+const struct bpf_func_proto bpf_skb_vlan_push_proto = {
+	.func           = bpf_skb_vlan_push,
+	.gpl_only       = false,
+	.ret_type       = RET_INTEGER,
+	.arg1_type      = ARG_PTR_TO_CTX,
+	.arg2_type      = ARG_ANYTHING,
+	.arg3_type      = ARG_ANYTHING,
+};
+EXPORT_SYMBOL_GPL(bpf_skb_vlan_push_proto);
+
+static u64 bpf_skb_vlan_pop(u64 r1, u64 r2, u64 r3, u64 r4, u64 r5)
+{
+	struct sk_buff *skb = (struct sk_buff *) (long) r1;
+
+	return skb_vlan_pop(skb);
+}
+
+const struct bpf_func_proto bpf_skb_vlan_pop_proto = {
+	.func           = bpf_skb_vlan_pop,
+	.gpl_only       = false,
+	.ret_type       = RET_INTEGER,
+	.arg1_type      = ARG_PTR_TO_CTX,
+};
+EXPORT_SYMBOL_GPL(bpf_skb_vlan_pop_proto);
+
+bool bpf_helper_changes_skb_data(void *func)
+{
+	if (func == bpf_skb_vlan_push)
+		return true;
+	if (func == bpf_skb_vlan_pop)
+		return true;
+	return false;
+}
+
 static const struct bpf_func_proto *
 sk_filter_func_proto(enum bpf_func_id func_id)
 {
@@ -1476,6 +1522,10 @@ tc_cls_act_func_proto(enum bpf_func_id func_id)
 		return &bpf_clone_redirect_proto;
 	case BPF_FUNC_get_cgroup_classid:
 		return &bpf_get_cgroup_classid_proto;
+	case BPF_FUNC_skb_vlan_push:
+		return &bpf_skb_vlan_push_proto;
+	case BPF_FUNC_skb_vlan_pop:
+		return &bpf_skb_vlan_pop_proto;
 	default:
 		return sk_filter_func_proto(func_id);
 	}