Merge remote-tracking branch 'upstream/master'

.github/workflows/pahole.yml

@@ -14,27 +14,7 @@ jobs:
     steps:
       - uses: actions/checkout@v2
       - uses: ./.github/actions/setup
-      - name: Get current pahole sha
-        id: current_sha
-        run:
-          git ls-remote https://git.kernel.org/pub/scm/devel/pahole/pahole.git $STAGING | awk '{print "::set-output name=sha::" $1}'
-      - name: Get latest result for this sha
-        id: latest
-        uses: pat-s/always-upload-cache@v2
-        with:
-          path: last_tested_pahole
-          key: ${{ steps.current_sha.outputs.sha }}
-      - name: Return cached test result
-        run: exit `cat last_tested_pahole || echo 1` # if file is empty that mean previous run timed out of canceled, returning failure
-        if: steps.latest.outputs.cache-hit == 'true'
       - uses: ./.github/actions/vmtest
         with:
           kernel: LATEST
           pahole: $STAGING
-        if: steps.latest.outputs.cache-hit != 'true'
-      - name: Save success
-        run: echo 0 > last_tested_pahole
-        if: steps.latest.outputs.cache-hit != 'true'
-      - name: Save failure
-        run: echo 1 > last_tested_pahole
-        if: ${{ failure() && steps.latest.outputs.cache-hit != 'true' }}

BPF-CHECKPOINT-COMMIT

@@ -1 +1 @@
-3776f3517ed94d40ff0e3851d7ce2ce17b63099f
+8d6c414cd2fb74aa6812e9bfec6178f8246c4f3a

CHECKPOINT-COMMIT

@@ -1 +1 @@
-d20b41115ad53293201cc07ee429a38740cb056b
+5319255b8df9271474bc9027cabf82253934f28d

docs/api.rst

@@ -12,40 +12,40 @@ libbpf.h
 --------
 .. doxygenfile:: libbpf.h
    :project: libbpf
-   :sections: func define public-type
+   :sections: func define public-type enum
 
 bpf.h
 -----
 .. doxygenfile:: bpf.h
    :project: libbpf
-   :sections: func define public-type
+   :sections: func define public-type enum
 
 btf.h
 -----
 .. doxygenfile:: btf.h
    :project: libbpf
-   :sections: func define public-type
+   :sections: func define public-type enum
 
 xsk.h
 -----
 .. doxygenfile:: xsk.h
    :project: libbpf
-   :sections: func define public-type
+   :sections: func define public-type enum
 
 bpf_tracing.h
 -------------
 .. doxygenfile:: bpf_tracing.h
    :project: libbpf
-   :sections: func define public-type
+   :sections: func define public-type enum
 
 bpf_core_read.h
 ---------------
 .. doxygenfile:: bpf_core_read.h
    :project: libbpf
-   :sections: func define public-type
+   :sections: func define public-type enum
 
 bpf_endian.h
 ------------
 .. doxygenfile:: bpf_endian.h
    :project: libbpf
-   :sections: func define public-type
+   :sections: func define public-type enum

docs/libbpf_naming_convention.rst

@@ -150,6 +150,46 @@ mirror of the mainline's version of libbpf for a stand-alone build.
 However, all changes to libbpf's code base must be upstreamed through
 the mainline kernel tree.
 
+
+API documentation convention
+============================
+
+The libbpf API is documented via comments above definitions in
+header files. These comments can be rendered by doxygen and sphinx
+for well organized html output. This section describes the
+convention in which these comments should be formated.
+
+Here is an example from btf.h:
+
+.. code-block:: c
+
+        /**
+         * @brief **btf__new()** creates a new instance of a BTF object from the raw
+         * bytes of an ELF's BTF section
+         * @param data raw bytes
+         * @param size number of bytes passed in `data`
+         * @return new BTF object instance which has to be eventually freed with
+         * **btf__free()**
+         *
+         * On error, error-code-encoded-as-pointer is returned, not a NULL. To extract
+         * error code from such a pointer `libbpf_get_error()` should be used. If
+         * `libbpf_set_strict_mode(LIBBPF_STRICT_CLEAN_PTRS)` is enabled, NULL is
+         * returned on error instead. In both cases thread-local `errno` variable is
+         * always set to error code as well.
+         */
+
+The comment must start with a block comment of the form '/\*\*'.
+
+The documentation always starts with a @brief directive. This line is a short
+description about this API. It starts with the name of the API, denoted in bold
+like so: **api_name**. Please include an open and close parenthesis if this is a
+function. Follow with the short description of the API. A longer form description
+can be added below the last directive, at the bottom of the comment.
+
+Parameters are denoted with the @param directive, there should be one for each
+parameter. If this is a function with a non-void return, use the @return directive
+to document it.
+
 License
 -------------------
 

include/uapi/linux/bpf.h

@@ -84,7 +84,7 @@ struct bpf_lpm_trie_key {
 
 struct bpf_cgroup_storage_key {
 	__u64	cgroup_inode_id;	/* cgroup inode id */
-	__u32	attach_type;		/* program attach type */
+	__u32	attach_type;		/* program attach type (enum bpf_attach_type) */
 };
 
 union bpf_iter_link_info {
@@ -1629,7 +1629,7 @@ union bpf_attr {
  * u32 bpf_get_smp_processor_id(void)
  * 	Description
  * 		Get the SMP (symmetric multiprocessing) processor id. Note that
- * 		all programs run with preemption disabled, which means that the
+ * 		all programs run with migration disabled, which means that the
  * 		SMP processor id is stable during all the execution of the
  * 		program.
  * 	Return
@@ -4046,7 +4046,7 @@ union bpf_attr {
  * 		arguments. The *data* are a **u64** array and corresponding format string
  * 		values are stored in the array. For strings and pointers where pointees
  * 		are accessed, only the pointer values are stored in the *data* array.
- * 		The *data_len* is the size of *data* in bytes.
+ * 		The *data_len* is the size of *data* in bytes - must be a multiple of 8.
  *
  *		Formats **%s**, **%p{i,I}{4,6}** requires to read kernel memory.
  *		Reading kernel memory may fail due to either invalid address or
@@ -4751,7 +4751,8 @@ union bpf_attr {
  *		Each format specifier in **fmt** corresponds to one u64 element
  *		in the **data** array. For strings and pointers where pointees
  *		are accessed, only the pointer values are stored in the *data*
- *		array. The *data_len* is the size of *data* in bytes.
+ *		array. The *data_len* is the size of *data* in bytes - must be
+ *		a multiple of 8.
  *
  *		Formats **%s** and **%p{i,I}{4,6}** require to read kernel
  *		memory. Reading kernel memory may fail due to either invalid
@@ -4871,6 +4872,43 @@ union bpf_attr {
  * 	Return
  *		Value specified by user at BPF link creation/attachment time
  *		or 0, if it was not specified.
+ *
+ * long bpf_task_pt_regs(struct task_struct *task)
+ *	Description
+ *		Get the struct pt_regs associated with **task**.
+ *	Return
+ *		A pointer to struct pt_regs.
+ *
+ * long bpf_get_branch_snapshot(void *entries, u32 size, u64 flags)
+ *	Description
+ *		Get branch trace from hardware engines like Intel LBR. The
+ *		hardware engine is stopped shortly after the helper is
+ *		called. Therefore, the user need to filter branch entries
+ *		based on the actual use case. To capture branch trace
+ *		before the trigger point of the BPF program, the helper
+ *		should be called at the beginning of the BPF program.
+ *
+ *		The data is stored as struct perf_branch_entry into output
+ *		buffer *entries*. *size* is the size of *entries* in bytes.
+ *		*flags* is reserved for now and must be zero.
+ *
+ *	Return
+ *		On success, number of bytes written to *buf*. On error, a
+ *		negative value.
+ *
+ *		**-EINVAL** if *flags* is not zero.
+ *
+ *		**-ENOENT** if architecture does not support branch records.
+ *
+ * long bpf_trace_vprintk(const char *fmt, u32 fmt_size, const void *data, u32 data_len)
+ *	Description
+ *		Behaves like **bpf_trace_printk**\ () helper, but takes an array of u64
+ *		to format and can handle more format args as a result.
+ *
+ *		Arguments are to be used as in **bpf_seq_printf**\ () helper.
+ *	Return
+ *		The number of bytes written to the buffer, or a negative error
+ *		in case of failure.
  */
 #define __BPF_FUNC_MAPPER(FN)		\
 	FN(unspec),			\
@@ -5048,6 +5086,9 @@ union bpf_attr {
 	FN(timer_cancel),		\
 	FN(get_func_ip),		\
 	FN(get_attach_cookie),		\
+	FN(task_pt_regs),		\
+	FN(get_branch_snapshot),	\
+	FN(trace_vprintk),		\
 	/* */
 
 /* integer value in 'imm' field of BPF_CALL instruction selects which helper
@@ -5277,6 +5318,8 @@ struct __sk_buff {
 	__u32 gso_segs;
 	__bpf_md_ptr(struct bpf_sock *, sk);
 	__u32 gso_size;
+	__u32 :32;		/* Padding, future use. */
+	__u64 hwtstamp;
 };
 
 struct bpf_tunnel_key {

include/uapi/linux/btf.h

@@ -43,7 +43,7 @@ struct btf_type {
 	 * "size" tells the size of the type it is describing.
 	 *
 	 * "type" is used by PTR, TYPEDEF, VOLATILE, CONST, RESTRICT,
-	 * FUNC, FUNC_PROTO and VAR.
+	 * FUNC, FUNC_PROTO, VAR and TAG.
 	 * "type" is a type_id referring to another type.
 	 */
 	union {
@@ -56,25 +56,29 @@ struct btf_type {
 #define BTF_INFO_VLEN(info)	((info) & 0xffff)
 #define BTF_INFO_KFLAG(info)	((info) >> 31)
 
-#define BTF_KIND_UNKN		0	/* Unknown	*/
-#define BTF_KIND_INT		1	/* Integer	*/
-#define BTF_KIND_PTR		2	/* Pointer	*/
-#define BTF_KIND_ARRAY		3	/* Array	*/
-#define BTF_KIND_STRUCT		4	/* Struct	*/
-#define BTF_KIND_UNION		5	/* Union	*/
-#define BTF_KIND_ENUM		6	/* Enumeration	*/
-#define BTF_KIND_FWD		7	/* Forward	*/
-#define BTF_KIND_TYPEDEF	8	/* Typedef	*/
-#define BTF_KIND_VOLATILE	9	/* Volatile	*/
-#define BTF_KIND_CONST		10	/* Const	*/
-#define BTF_KIND_RESTRICT	11	/* Restrict	*/
-#define BTF_KIND_FUNC		12	/* Function	*/
-#define BTF_KIND_FUNC_PROTO	13	/* Function Proto	*/
-#define BTF_KIND_VAR		14	/* Variable	*/
-#define BTF_KIND_DATASEC	15	/* Section	*/
-#define BTF_KIND_FLOAT		16	/* Floating point	*/
-#define BTF_KIND_MAX		BTF_KIND_FLOAT
-#define NR_BTF_KINDS		(BTF_KIND_MAX + 1)
+enum {
+	BTF_KIND_UNKN		= 0,	/* Unknown	*/
+	BTF_KIND_INT		= 1,	/* Integer	*/
+	BTF_KIND_PTR		= 2,	/* Pointer	*/
+	BTF_KIND_ARRAY		= 3,	/* Array	*/
+	BTF_KIND_STRUCT		= 4,	/* Struct	*/
+	BTF_KIND_UNION		= 5,	/* Union	*/
+	BTF_KIND_ENUM		= 6,	/* Enumeration	*/
+	BTF_KIND_FWD		= 7,	/* Forward	*/
+	BTF_KIND_TYPEDEF	= 8,	/* Typedef	*/
+	BTF_KIND_VOLATILE	= 9,	/* Volatile	*/
+	BTF_KIND_CONST		= 10,	/* Const	*/
+	BTF_KIND_RESTRICT	= 11,	/* Restrict	*/
+	BTF_KIND_FUNC		= 12,	/* Function	*/
+	BTF_KIND_FUNC_PROTO	= 13,	/* Function Proto	*/
+	BTF_KIND_VAR		= 14,	/* Variable	*/
+	BTF_KIND_DATASEC	= 15,	/* Section	*/
+	BTF_KIND_FLOAT		= 16,	/* Floating point	*/
+	BTF_KIND_TAG		= 17,	/* Tag */
+
+	NR_BTF_KINDS,
+	BTF_KIND_MAX		= NR_BTF_KINDS - 1,
+};
 
 /* For some specific BTF_KIND, "struct btf_type" is immediately
  * followed by extra data.
@@ -170,4 +174,15 @@ struct btf_var_secinfo {
 	__u32	size;
 };
 
+/* BTF_KIND_TAG is followed by a single "struct btf_tag" to describe
+ * additional information related to the tag applied location.
+ * If component_idx == -1, the tag is applied to a struct, union,
+ * variable or function. Otherwise, it is applied to a struct/union
+ * member or a func argument, and component_idx indicates which member
+ * or argument (0 ... vlen-1).
+ */
+struct btf_tag {
+       __s32   component_idx;
+};
+
 #endif /* _UAPI__LINUX_BTF_H__ */

include/uapi/linux/if_link.h

@@ -230,6 +230,7 @@ enum {
 	IFLA_INET6_ICMP6STATS,	/* statistics (icmpv6)		*/
 	IFLA_INET6_TOKEN,	/* device token			*/
 	IFLA_INET6_ADDR_GEN_MODE, /* implicit address generator mode */
+	IFLA_INET6_RA_MTU,	/* mtu carried in the RA message */
 	__IFLA_INET6_MAX
 };
 

src/Makefile

@@ -51,7 +51,7 @@ endif
 
 HEADERS := bpf.h libbpf.h btf.h libbpf_common.h libbpf_legacy.h xsk.h \
 	   bpf_helpers.h bpf_helper_defs.h bpf_tracing.h \
-	   bpf_endian.h bpf_core_read.h skel_internal.h
+	   bpf_endian.h bpf_core_read.h skel_internal.h libbpf_version.h
 UAPI_HEADERS := $(addprefix $(TOPDIR)/include/uapi/linux/,\
 			    bpf.h bpf_common.h btf.h)
 

src/bpf.c

@@ -264,6 +264,7 @@ int libbpf__bpf_prog_load(const struct bpf_prog_load_params *load_attr)
 	attr.line_info_rec_size = load_attr->line_info_rec_size;
 	attr.line_info_cnt = load_attr->line_info_cnt;
 	attr.line_info = ptr_to_u64(load_attr->line_info);
+	attr.fd_array = ptr_to_u64(load_attr->fd_array);
 
 	if (load_attr->name)
 		memcpy(attr.prog_name, load_attr->name,

src/bpf_gen_internal.h

@@ -7,6 +7,15 @@ struct ksym_relo_desc {
 	const char *name;
 	int kind;
 	int insn_idx;
+	bool is_weak;
+};
+
+struct ksym_desc {
+	const char *name;
+	int ref;
+	int kind;
+	int off;
+	int insn;
 };
 
 struct bpf_gen {
@@ -24,6 +33,10 @@ struct bpf_gen {
 	int relo_cnt;
 	char attach_target[128];
 	int attach_kind;
+	struct ksym_desc *ksyms;
+	__u32 nr_ksyms;
+	int fd_array;
+	int nr_fd_array;
 };
 
 void bpf_gen__init(struct bpf_gen *gen, int log_level);
@@ -36,6 +49,7 @@ void bpf_gen__prog_load(struct bpf_gen *gen, struct bpf_prog_load_params *load_a
 void bpf_gen__map_update_elem(struct bpf_gen *gen, int map_idx, void *value, __u32 value_size);
 void bpf_gen__map_freeze(struct bpf_gen *gen, int map_idx);
 void bpf_gen__record_attach_target(struct bpf_gen *gen, const char *name, enum bpf_attach_type type);
-void bpf_gen__record_extern(struct bpf_gen *gen, const char *name, int kind, int insn_idx);
+void bpf_gen__record_extern(struct bpf_gen *gen, const char *name, bool is_weak, int kind,
+			    int insn_idx);
 
 #endif

src/bpf_helper_defs.h

@@ -190,7 +190,7 @@ static __u32 (*bpf_get_prandom_u32)(void) = (void *) 7;
  * bpf_get_smp_processor_id
  *
  * 	Get the SMP (symmetric multiprocessing) processor id. Note that
- * 	all programs run with preemption disabled, which means that the
+ * 	all programs run with migration disabled, which means that the
  * 	SMP processor id is stable during all the execution of the
  * 	program.
  *
@@ -3012,7 +3012,7 @@ static __u64 (*bpf_ktime_get_boot_ns)(void) = (void *) 125;
  * 	arguments. The *data* are a **u64** array and corresponding format string
  * 	values are stored in the array. For strings and pointers where pointees
  * 	are accessed, only the pointer values are stored in the *data* array.
- * 	The *data_len* is the size of *data* in bytes.
+ * 	The *data_len* is the size of *data* in bytes - must be a multiple of 8.
  *
  * 	Formats **%s**, **%p{i,I}{4,6}** requires to read kernel memory.
  * 	Reading kernel memory may fail due to either invalid address or
@@ -3873,7 +3873,8 @@ static long (*bpf_for_each_map_elem)(void *map, void *callback_fn, void *callbac
  * 	Each format specifier in **fmt** corresponds to one u64 element
  * 	in the **data** array. For strings and pointers where pointees
  * 	are accessed, only the pointer values are stored in the *data*
- * 	array. The *data_len* is the size of *data* in bytes.
+ * 	array. The *data_len* is the size of *data* in bytes - must be
+ * 	a multiple of 8.
  *
  * 	Formats **%s** and **%p{i,I}{4,6}** require to read kernel
  * 	memory. Reading kernel memory may fail due to either invalid
@@ -4033,4 +4034,53 @@ static __u64 (*bpf_get_func_ip)(void *ctx) = (void *) 173;
  */
 static __u64 (*bpf_get_attach_cookie)(void *ctx) = (void *) 174;
 
+/*
+ * bpf_task_pt_regs
+ *
+ * 	Get the struct pt_regs associated with **task**.
+ *
+ * Returns
+ * 	A pointer to struct pt_regs.
+ */
+static long (*bpf_task_pt_regs)(struct task_struct *task) = (void *) 175;
+
+/*
+ * bpf_get_branch_snapshot
+ *
+ * 	Get branch trace from hardware engines like Intel LBR. The
+ * 	hardware engine is stopped shortly after the helper is
+ * 	called. Therefore, the user need to filter branch entries
+ * 	based on the actual use case. To capture branch trace
+ * 	before the trigger point of the BPF program, the helper
+ * 	should be called at the beginning of the BPF program.
+ *
+ * 	The data is stored as struct perf_branch_entry into output
+ * 	buffer *entries*. *size* is the size of *entries* in bytes.
+ * 	*flags* is reserved for now and must be zero.
+ *
+ *
+ * Returns
+ * 	On success, number of bytes written to *buf*. On error, a
+ * 	negative value.
+ *
+ * 	**-EINVAL** if *flags* is not zero.
+ *
+ * 	**-ENOENT** if architecture does not support branch records.
+ */
+static long (*bpf_get_branch_snapshot)(void *entries, __u32 size, __u64 flags) = (void *) 176;
+
+/*
+ * bpf_trace_vprintk
+ *
+ * 	Behaves like **bpf_trace_printk**\ () helper, but takes an array of u64
+ * 	to format and can handle more format args as a result.
+ *
+ * 	Arguments are to be used as in **bpf_seq_printf**\ () helper.
+ *
+ * Returns
+ * 	The number of bytes written to the buffer, or a negative error
+ * 	in case of failure.
+ */
+static long (*bpf_trace_vprintk)(const char *fmt, __u32 fmt_size, const void *data, __u32 data_len) = (void *) 177;
+
 

src/bpf_helpers.h

@@ -14,14 +14,6 @@
 #define __type(name, val) typeof(val) *name
 #define __array(name, val) typeof(val) *name[]
 
-/* Helper macro to print out debug messages */
-#define bpf_printk(fmt, ...)				\
-({							\
-	char ____fmt[] = fmt;				\
-	bpf_trace_printk(____fmt, sizeof(____fmt),	\
-			 ##__VA_ARGS__);		\
-})
-
 /*
  * Helper macro to place programs, maps, license in
  * different sections in elf_bpf file. Section names
@@ -224,4 +216,47 @@ enum libbpf_tristate {
 		     ___param, sizeof(___param));		\
 })
 
+#ifdef BPF_NO_GLOBAL_DATA
+#define BPF_PRINTK_FMT_MOD
+#else
+#define BPF_PRINTK_FMT_MOD static const
+#endif
+
+#define __bpf_printk(fmt, ...)				\
+({							\
+	BPF_PRINTK_FMT_MOD char ____fmt[] = fmt;	\
+	bpf_trace_printk(____fmt, sizeof(____fmt),	\
+			 ##__VA_ARGS__);		\
+})
+
+/*
+ * __bpf_vprintk wraps the bpf_trace_vprintk helper with variadic arguments
+ * instead of an array of u64.
+ */
+#define __bpf_vprintk(fmt, args...)				\
+({								\
+	static const char ___fmt[] = fmt;			\
+	unsigned long long ___param[___bpf_narg(args)];		\
+								\
+	_Pragma("GCC diagnostic push")				\
+	_Pragma("GCC diagnostic ignored \"-Wint-conversion\"")	\
+	___bpf_fill(___param, args);				\
+	_Pragma("GCC diagnostic pop")				\
+								\
+	bpf_trace_vprintk(___fmt, sizeof(___fmt),		\
+			  ___param, sizeof(___param));		\
+})
+
+/* Use __bpf_printk when bpf_printk call has 3 or fewer fmt args
+ * Otherwise use __bpf_vprintk
+ */
+#define ___bpf_pick_printk(...) \
+	___bpf_nth(_, ##__VA_ARGS__, __bpf_vprintk, __bpf_vprintk, __bpf_vprintk,	\
+		   __bpf_vprintk, __bpf_vprintk, __bpf_vprintk, __bpf_vprintk,		\
+		   __bpf_vprintk, __bpf_vprintk, __bpf_printk /*3*/, __bpf_printk /*2*/,\
+		   __bpf_printk /*1*/, __bpf_printk /*0*/)
+
+/* Helper macro to print out debug messages */
+#define bpf_printk(fmt, args...) ___bpf_pick_printk(args)(fmt, ##args)
+
 #endif

src/btf.c

@@ -189,12 +189,17 @@ int libbpf_ensure_mem(void **data, size_t *cap_cnt, size_t elem_sz, size_t need_
 	return 0;
 }
 
+static void *btf_add_type_offs_mem(struct btf *btf, size_t add_cnt)
+{
+	return libbpf_add_mem((void **)&btf->type_offs, &btf->type_offs_cap, sizeof(__u32),
+			      btf->nr_types, BTF_MAX_NR_TYPES, add_cnt);
+}
+
 static int btf_add_type_idx_entry(struct btf *btf, __u32 type_off)
 {
 	__u32 *p;
 
-	p = libbpf_add_mem((void **)&btf->type_offs, &btf->type_offs_cap, sizeof(__u32),
-			   btf->nr_types, BTF_MAX_NR_TYPES, 1);
+	p = btf_add_type_offs_mem(btf, 1);
 	if (!p)
 		return -ENOMEM;
 
@@ -304,6 +309,8 @@ static int btf_type_size(const struct btf_type *t)
 		return base_size + sizeof(struct btf_var);
 	case BTF_KIND_DATASEC:
 		return base_size + vlen * sizeof(struct btf_var_secinfo);
+	case BTF_KIND_TAG:
+		return base_size + sizeof(struct btf_tag);
 	default:
 		pr_debug("Unsupported BTF_KIND:%u\n", btf_kind(t));
 		return -EINVAL;
@@ -376,6 +383,9 @@ static int btf_bswap_type_rest(struct btf_type *t)
 			v->size = bswap_32(v->size);
 		}
 		return 0;
+	case BTF_KIND_TAG:
+		btf_tag(t)->component_idx = bswap_32(btf_tag(t)->component_idx);
+		return 0;
 	default:
 		pr_debug("Unsupported BTF_KIND:%u\n", btf_kind(t));
 		return -EINVAL;
@@ -586,6 +596,7 @@ __s64 btf__resolve_size(const struct btf *btf, __u32 type_id)
 		case BTF_KIND_CONST:
 		case BTF_KIND_RESTRICT:
 		case BTF_KIND_VAR:
+		case BTF_KIND_TAG:
 			type_id = t->type;
 			break;
 		case BTF_KIND_ARRAY:
@@ -689,15 +700,15 @@ __s32 btf__find_by_name(const struct btf *btf, const char *type_name)
 	return libbpf_err(-ENOENT);
 }
 
-__s32 btf__find_by_name_kind(const struct btf *btf, const char *type_name,
-			     __u32 kind)
+static __s32 btf_find_by_name_kind(const struct btf *btf, int start_id,
+				   const char *type_name, __u32 kind)
 {
 	__u32 i, nr_types = btf__get_nr_types(btf);
 
 	if (kind == BTF_KIND_UNKN || !strcmp(type_name, "void"))
 		return 0;
 
-	for (i = 1; i <= nr_types; i++) {
+	for (i = start_id; i <= nr_types; i++) {
 		const struct btf_type *t = btf__type_by_id(btf, i);
 		const char *name;
 
@@ -711,6 +722,18 @@ __s32 btf__find_by_name_kind(const struct btf *btf, const char *type_name,
 	return libbpf_err(-ENOENT);
 }
 
+__s32 btf__find_by_name_kind_own(const struct btf *btf, const char *type_name,
+				 __u32 kind)
+{
+	return btf_find_by_name_kind(btf, btf->start_id, type_name, kind);
+}
+
+__s32 btf__find_by_name_kind(const struct btf *btf, const char *type_name,
+			     __u32 kind)
+{
+	return btf_find_by_name_kind(btf, 1, type_name, kind);
+}
+
 static bool btf_is_modifiable(const struct btf *btf)
 {
 	return (void *)btf->hdr != btf->raw_data;
@@ -1685,6 +1708,111 @@ int btf__add_type(struct btf *btf, const struct btf *src_btf, const struct btf_t
 	return btf_commit_type(btf, sz);
 }
 
+static int btf_rewrite_type_ids(__u32 *type_id, void *ctx)
+{
+	struct btf *btf = ctx;
+
+	if (!*type_id) /* nothing to do for VOID references */
+		return 0;
+
+	/* we haven't updated btf's type count yet, so
+	 * btf->start_id + btf->nr_types - 1 is the type ID offset we should
+	 * add to all newly added BTF types
+	 */
+	*type_id += btf->start_id + btf->nr_types - 1;
+	return 0;
+}
+
+int btf__add_btf(struct btf *btf, const struct btf *src_btf)
+{
+	struct btf_pipe p = { .src = src_btf, .dst = btf };
+	int data_sz, sz, cnt, i, err, old_strs_len;
+	__u32 *off;
+	void *t;
+
+	/* appending split BTF isn't supported yet */
+	if (src_btf->base_btf)
+		return libbpf_err(-ENOTSUP);
+
+	/* deconstruct BTF, if necessary, and invalidate raw_data */
+	if (btf_ensure_modifiable(btf))
+		return libbpf_err(-ENOMEM);
+
+	/* remember original strings section size if we have to roll back
+	 * partial strings section changes
+	 */
+	old_strs_len = btf->hdr->str_len;
+
+	data_sz = src_btf->hdr->type_len;
+	cnt = btf__get_nr_types(src_btf);
+
+	/* pre-allocate enough memory for new types */
+	t = btf_add_type_mem(btf, data_sz);
+	if (!t)
+		return libbpf_err(-ENOMEM);
+
+	/* pre-allocate enough memory for type offset index for new types */
+	off = btf_add_type_offs_mem(btf, cnt);
+	if (!off)
+		return libbpf_err(-ENOMEM);
+
+	/* bulk copy types data for all types from src_btf */
+	memcpy(t, src_btf->types_data, data_sz);
+
+	for (i = 0; i < cnt; i++) {
+		sz = btf_type_size(t);
+		if (sz < 0) {
+			/* unlikely, has to be corrupted src_btf */
+			err = sz;
+			goto err_out;
+		}
+
+		/* fill out type ID to type offset mapping for lookups by type ID */
+		*off = t - btf->types_data;
+
+		/* add, dedup, and remap strings referenced by this BTF type */
+		err = btf_type_visit_str_offs(t, btf_rewrite_str, &p);
+		if (err)
+			goto err_out;
+
+		/* remap all type IDs referenced from this BTF type */
+		err = btf_type_visit_type_ids(t, btf_rewrite_type_ids, btf);
+		if (err)
+			goto err_out;
+
+		/* go to next type data and type offset index entry */
+		t += sz;
+		off++;
+	}
+
+	/* Up until now any of the copied type data was effectively invisible,
+	 * so if we exited early before this point due to error, BTF would be
+	 * effectively unmodified. There would be extra internal memory
+	 * pre-allocated, but it would not be available for querying.  But now
+	 * that we've copied and rewritten all the data successfully, we can
+	 * update type count and various internal offsets and sizes to
+	 * "commit" the changes and made them visible to the outside world.
+	 */
+	btf->hdr->type_len += data_sz;
+	btf->hdr->str_off += data_sz;
+	btf->nr_types += cnt;
+
+	/* return type ID of the first added BTF type */
+	return btf->start_id + btf->nr_types - cnt;
+err_out:
+	/* zero out preallocated memory as if it was just allocated with
+	 * libbpf_add_mem()
+	 */
+	memset(btf->types_data + btf->hdr->type_len, 0, data_sz);
+	memset(btf->strs_data + old_strs_len, 0, btf->hdr->str_len - old_strs_len);
+
+	/* and now restore original strings section size; types data size
+	 * wasn't modified, so doesn't need restoring, see big comment above */
+	btf->hdr->str_len = old_strs_len;
+
+	return libbpf_err(err);
+}
+
 /*
  * Append new BTF_KIND_INT type with:
  *   - *name* - non-empty, non-NULL type name;
@@ -2440,6 +2568,48 @@ int btf__add_datasec_var_info(struct btf *btf, int var_type_id, __u32 offset, __
 	return 0;
 }
 
+/*
+ * Append new BTF_KIND_TAG type with:
+ *   - *value* - non-empty/non-NULL string;
+ *   - *ref_type_id* - referenced type ID, it might not exist yet;
+ *   - *component_idx* - -1 for tagging reference type, otherwise struct/union
+ *     member or function argument index;
+ * Returns:
+ *   - >0, type ID of newly added BTF type;
+ *   - <0, on error.
+ */
+int btf__add_tag(struct btf *btf, const char *value, int ref_type_id,
+		 int component_idx)
+{
+	struct btf_type *t;
+	int sz, value_off;
+
+	if (!value || !value[0] || component_idx < -1)
+		return libbpf_err(-EINVAL);
+
+	if (validate_type_id(ref_type_id))
+		return libbpf_err(-EINVAL);
+
+	if (btf_ensure_modifiable(btf))
+		return libbpf_err(-ENOMEM);
+
+	sz = sizeof(struct btf_type) + sizeof(struct btf_tag);
+	t = btf_add_type_mem(btf, sz);
+	if (!t)
+		return libbpf_err(-ENOMEM);
+
+	value_off = btf__add_str(btf, value);
+	if (value_off < 0)
+		return value_off;
+
+	t->name_off = value_off;
+	t->info = btf_type_info(BTF_KIND_TAG, 0, false);
+	t->type = ref_type_id;
+	btf_tag(t)->component_idx = component_idx;
+
+	return btf_commit_type(btf, sz);
+}
+
 struct btf_ext_sec_setup_param {
 	__u32 off;
 	__u32 len;
@@ -3256,8 +3426,8 @@ static bool btf_equal_common(struct btf_type *t1, struct btf_type *t2)
 	       t1->size == t2->size;
 }
 
-/* Calculate type signature hash of INT. */
-static long btf_hash_int(struct btf_type *t)
+/* Calculate type signature hash of INT or TAG. */
+static long btf_hash_int_tag(struct btf_type *t)
 {
 	__u32 info = *(__u32 *)(t + 1);
 	long h;
@@ -3267,8 +3437,8 @@ static long btf_hash_int(struct btf_type *t)
 	return h;
 }
 
-/* Check structural equality of two INTs. */
-static bool btf_equal_int(struct btf_type *t1, struct btf_type *t2)
+/* Check structural equality of two INTs or TAGs. */
+static bool btf_equal_int_tag(struct btf_type *t1, struct btf_type *t2)
 {
 	__u32 info1, info2;
 
@@ -3535,7 +3705,8 @@ static int btf_dedup_prep(struct btf_dedup *d)
 			h = btf_hash_common(t);
 			break;
 		case BTF_KIND_INT:
-			h = btf_hash_int(t);
+		case BTF_KIND_TAG:
+			h = btf_hash_int_tag(t);
 			break;
 		case BTF_KIND_ENUM:
 			h = btf_hash_enum(t);
@@ -3590,14 +3761,15 @@ static int btf_dedup_prim_type(struct btf_dedup *d, __u32 type_id)
 	case BTF_KIND_FUNC_PROTO:
 	case BTF_KIND_VAR:
 	case BTF_KIND_DATASEC:
+	case BTF_KIND_TAG:
 		return 0;
 
 	case BTF_KIND_INT:
-		h = btf_hash_int(t);
+		h = btf_hash_int_tag(t);
 		for_each_dedup_cand(d, hash_entry, h) {
 			cand_id = (__u32)(long)hash_entry->value;
 			cand = btf_type_by_id(d->btf, cand_id);
-			if (btf_equal_int(t, cand)) {
+			if (btf_equal_int_tag(t, cand)) {
 				new_id = cand_id;
 				break;
 			}
@@ -3881,7 +4053,7 @@ static int btf_dedup_is_equiv(struct btf_dedup *d, __u32 cand_id,
 
 	switch (cand_kind) {
 	case BTF_KIND_INT:
-		return btf_equal_int(cand_type, canon_type);
+		return btf_equal_int_tag(cand_type, canon_type);
 
 	case BTF_KIND_ENUM:
 		if (d->opts.dont_resolve_fwds)
@@ -4210,6 +4382,23 @@ static int btf_dedup_ref_type(struct btf_dedup *d, __u32 type_id)
 		}
 		break;
 
+	case BTF_KIND_TAG:
+		ref_type_id = btf_dedup_ref_type(d, t->type);
+		if (ref_type_id < 0)
+			return ref_type_id;
+		t->type = ref_type_id;
+
+		h = btf_hash_int_tag(t);
+		for_each_dedup_cand(d, hash_entry, h) {
+			cand_id = (__u32)(long)hash_entry->value;
+			cand = btf_type_by_id(d->btf, cand_id);
+			if (btf_equal_int_tag(t, cand)) {
+				new_id = cand_id;
+				break;
+			}
+		}
+		break;
+
 	case BTF_KIND_ARRAY: {
 		struct btf_array *info = btf_array(t);
 
@@ -4482,6 +4671,7 @@ int btf_type_visit_type_ids(struct btf_type *t, type_id_visit_fn visit, void *ct
 	case BTF_KIND_TYPEDEF:
 	case BTF_KIND_FUNC:
 	case BTF_KIND_VAR:
+	case BTF_KIND_TAG:
 		return visit(&t->type, ctx);
 
 	case BTF_KIND_ARRAY: {

src/btf.h

@@ -1,5 +1,6 @@
 /* SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) */
 /* Copyright (c) 2018 Facebook */
+/*! \file */
 
 #ifndef __LIBBPF_BTF_H
 #define __LIBBPF_BTF_H
@@ -30,11 +31,80 @@ enum btf_endianness {
 	BTF_BIG_ENDIAN = 1,
 };
 
+/**
+ * @brief **btf__free()** frees all data of a BTF object
+ * @param btf BTF object to free
+ */
 LIBBPF_API void btf__free(struct btf *btf);
 
+/**
+ * @brief **btf__new()** creates a new instance of a BTF object from the raw
+ * bytes of an ELF's BTF section
+ * @param data raw bytes
+ * @param size number of bytes passed in `data`
+ * @return new BTF object instance which has to be eventually freed with
+ * **btf__free()**
+ *
+ * On error, error-code-encoded-as-pointer is returned, not a NULL. To extract
+ * error code from such a pointer `libbpf_get_error()` should be used. If
+ * `libbpf_set_strict_mode(LIBBPF_STRICT_CLEAN_PTRS)` is enabled, NULL is
+ * returned on error instead. In both cases thread-local `errno` variable is
+ * always set to error code as well.
+ */
 LIBBPF_API struct btf *btf__new(const void *data, __u32 size);
+
+/**
+ * @brief **btf__new_split()** create a new instance of a BTF object from the
+ * provided raw data bytes. It takes another BTF instance, **base_btf**, which
+ * serves as a base BTF, which is extended by types in a newly created BTF
+ * instance
+ * @param data raw bytes
+ * @param size length of raw bytes
+ * @param base_btf the base BTF object
+ * @return new BTF object instance which has to be eventually freed with
+ * **btf__free()**
+ *
+ * If *base_btf* is NULL, `btf__new_split()` is equivalent to `btf__new()` and
+ * creates non-split BTF.
+ *
+ * On error, error-code-encoded-as-pointer is returned, not a NULL. To extract
+ * error code from such a pointer `libbpf_get_error()` should be used. If
+ * `libbpf_set_strict_mode(LIBBPF_STRICT_CLEAN_PTRS)` is enabled, NULL is
+ * returned on error instead. In both cases thread-local `errno` variable is
+ * always set to error code as well.
+ */
 LIBBPF_API struct btf *btf__new_split(const void *data, __u32 size, struct btf *base_btf);
+
+/**
+ * @brief **btf__new_empty()** creates an empty BTF object.  Use
+ * `btf__add_*()` to populate such BTF object.
+ * @return new BTF object instance which has to be eventually freed with
+ * **btf__free()**
+ *
+ * On error, error-code-encoded-as-pointer is returned, not a NULL. To extract
+ * error code from such a pointer `libbpf_get_error()` should be used. If
+ * `libbpf_set_strict_mode(LIBBPF_STRICT_CLEAN_PTRS)` is enabled, NULL is
+ * returned on error instead. In both cases thread-local `errno` variable is
+ * always set to error code as well.
+ */
 LIBBPF_API struct btf *btf__new_empty(void);
+
+/**
+ * @brief **btf__new_empty_split()** creates an unpopulated BTF object from an
+ * ELF BTF section except with a base BTF on top of which split BTF should be
+ * based
+ * @return new BTF object instance which has to be eventually freed with
+ * **btf__free()**
+ *
+ * If *base_btf* is NULL, `btf__new_empty_split()` is equivalent to
+ * `btf__new_empty()` and creates non-split BTF.
+ *
+ * On error, error-code-encoded-as-pointer is returned, not a NULL. To extract
+ * error code from such a pointer `libbpf_get_error()` should be used. If
+ * `libbpf_set_strict_mode(LIBBPF_STRICT_CLEAN_PTRS)` is enabled, NULL is
+ * returned on error instead. In both cases thread-local `errno` variable is
+ * always set to error code as well.
+ */
 LIBBPF_API struct btf *btf__new_empty_split(struct btf *base_btf);
 
 LIBBPF_API struct btf *btf__parse(const char *path, struct btf_ext **btf_ext);
@@ -50,9 +120,11 @@ LIBBPF_API struct btf *libbpf_find_kernel_btf(void);
 
 LIBBPF_API struct btf *btf__load_from_kernel_by_id(__u32 id);
 LIBBPF_API struct btf *btf__load_from_kernel_by_id_split(__u32 id, struct btf *base_btf);
+LIBBPF_DEPRECATED_SINCE(0, 6, "use btf__load_from_kernel_by_id instead")
 LIBBPF_API int btf__get_from_id(__u32 id, struct btf **btf);
 
 LIBBPF_API int btf__finalize_data(struct bpf_object *obj, struct btf *btf);
+LIBBPF_DEPRECATED_SINCE(0, 6, "use btf__load_into_kernel instead")
 LIBBPF_API int btf__load(struct btf *btf);
 LIBBPF_API int btf__load_into_kernel(struct btf *btf);
 LIBBPF_API __s32 btf__find_by_name(const struct btf *btf,
@@ -101,6 +173,28 @@ LIBBPF_API int btf__find_str(struct btf *btf, const char *s);
 LIBBPF_API int btf__add_str(struct btf *btf, const char *s);
 LIBBPF_API int btf__add_type(struct btf *btf, const struct btf *src_btf,
 			     const struct btf_type *src_type);
+/**
+ * @brief **btf__add_btf()** appends all the BTF types from *src_btf* into *btf*
+ * @param btf BTF object which all the BTF types and strings are added to
+ * @param src_btf BTF object which all BTF types and referenced strings are copied from
+ * @return BTF type ID of the first appended BTF type, or negative error code
+ *
+ * **btf__add_btf()** can be used to simply and efficiently append the entire
+ * contents of one BTF object to another one. All the BTF type data is copied
+ * over, all referenced type IDs are adjusted by adding a necessary ID offset.
+ * Only strings referenced from BTF types are copied over and deduplicated, so
+ * if there were some unused strings in *src_btf*, those won't be copied over,
+ * which is consistent with the general string deduplication semantics of BTF
+ * writing APIs.
+ *
+ * If any error is encountered during this process, the contents of *btf* is
+ * left intact, which means that **btf__add_btf()** follows the transactional
+ * semantics and the operation as a whole is all-or-nothing.
+ *
+ * *src_btf* has to be non-split BTF, as of now copying types from split BTF
+ * is not supported and will result in -ENOTSUP error code returned.
+ */
+LIBBPF_API int btf__add_btf(struct btf *btf, const struct btf *src_btf);
 
 LIBBPF_API int btf__add_int(struct btf *btf, const char *name, size_t byte_sz, int encoding);
 LIBBPF_API int btf__add_float(struct btf *btf, const char *name, size_t byte_sz);
@@ -141,6 +235,10 @@ LIBBPF_API int btf__add_datasec(struct btf *btf, const char *name, __u32 byte_sz
 LIBBPF_API int btf__add_datasec_var_info(struct btf *btf, int var_type_id,
 					 __u32 offset, __u32 byte_sz);
 
+/* tag construction API */
+LIBBPF_API int btf__add_tag(struct btf *btf, const char *value, int ref_type_id,
+			    int component_idx);
+
 struct btf_dedup_opts {
 	unsigned int dedup_table_size;
 	bool dont_resolve_fwds;
@@ -328,6 +426,11 @@ static inline bool btf_is_float(const struct btf_type *t)
 	return btf_kind(t) == BTF_KIND_FLOAT;
 }
 
+static inline bool btf_is_tag(const struct btf_type *t)
+{
+	return btf_kind(t) == BTF_KIND_TAG;
+}
+
 static inline __u8 btf_int_encoding(const struct btf_type *t)
 {
 	return BTF_INT_ENCODING(*(__u32 *)(t + 1));
@@ -396,6 +499,12 @@ btf_var_secinfos(const struct btf_type *t)
 	return (struct btf_var_secinfo *)(t + 1);
 }
 
+struct btf_tag;
+static inline struct btf_tag *btf_tag(const struct btf_type *t)
+{
+	return (struct btf_tag *)(t + 1);
+}
+
 #ifdef __cplusplus
 } /* extern "C" */
 #endif

src/btf_dump.c

@@ -316,6 +316,7 @@ static int btf_dump_mark_referenced(struct btf_dump *d)
 		case BTF_KIND_TYPEDEF:
 		case BTF_KIND_FUNC:
 		case BTF_KIND_VAR:
+		case BTF_KIND_TAG:
 			d->type_states[t->type].referenced = 1;
 			break;
 
@@ -583,6 +584,7 @@ static int btf_dump_order_type(struct btf_dump *d, __u32 id, bool through_ptr)
 	case BTF_KIND_FUNC:
 	case BTF_KIND_VAR:
 	case BTF_KIND_DATASEC:
+	case BTF_KIND_TAG:
 		d->type_states[id].order_state = ORDERED;
 		return 0;
 
@@ -2215,6 +2217,7 @@ static int btf_dump_dump_type_data(struct btf_dump *d,
 	case BTF_KIND_FWD:
 	case BTF_KIND_FUNC:
 	case BTF_KIND_FUNC_PROTO:
+	case BTF_KIND_TAG:
 		err = btf_dump_unsupported_data(d, t, id);
 		break;
 	case BTF_KIND_INT:

src/gen_loader.c

@@ -5,6 +5,7 @@
 #include <string.h>
 #include <errno.h>
 #include <linux/filter.h>
+#include <sys/param.h>
 #include "btf.h"
 #include "bpf.h"
 #include "libbpf.h"
@@ -13,8 +14,10 @@
 #include "bpf_gen_internal.h"
 #include "skel_internal.h"
 
-#define MAX_USED_MAPS 64
-#define MAX_USED_PROGS 32
+#define MAX_USED_MAPS	64
+#define MAX_USED_PROGS	32
+#define MAX_KFUNC_DESCS 256
+#define MAX_FD_ARRAY_SZ (MAX_USED_PROGS + MAX_KFUNC_DESCS)
 
 /* The following structure describes the stack layout of the loader program.
  * In addition R6 contains the pointer to context.
@@ -29,7 +32,6 @@
  */
 struct loader_stack {
 	__u32 btf_fd;
-	__u32 map_fd[MAX_USED_MAPS];
 	__u32 prog_fd[MAX_USED_PROGS];
 	__u32 inner_map_fd;
 };
@@ -135,16 +137,56 @@ void bpf_gen__init(struct bpf_gen *gen, int log_level)
 
 static int add_data(struct bpf_gen *gen, const void *data, __u32 size)
 {
+	__u32 size8 = roundup(size, 8);
+	__u64 zero = 0;
 	void *prev;
 
-	if (realloc_data_buf(gen, size))
+	if (realloc_data_buf(gen, size8))
 		return 0;
 	prev = gen->data_cur;
-	memcpy(gen->data_cur, data, size);
-	gen->data_cur += size;
+	if (data) {
+		memcpy(gen->data_cur, data, size);
+		memcpy(gen->data_cur + size, &zero, size8 - size);
+	} else {
+		memset(gen->data_cur, 0, size8);
+	}
+	gen->data_cur += size8;
 	return prev - gen->data_start;
 }
 
+/* Get index for map_fd/btf_fd slot in reserved fd_array, or in data relative
+ * to start of fd_array. Caller can decide if it is usable or not.
+ */
+static int add_map_fd(struct bpf_gen *gen)
+{
+	if (!gen->fd_array)
+		gen->fd_array = add_data(gen, NULL, MAX_FD_ARRAY_SZ * sizeof(int));
+	if (gen->nr_maps == MAX_USED_MAPS) {
+		pr_warn("Total maps exceeds %d\n", MAX_USED_MAPS);
+		gen->error = -E2BIG;
+		return 0;
+	}
+	return gen->nr_maps++;
+}
+
+static int add_kfunc_btf_fd(struct bpf_gen *gen)
+{
+	int cur;
+
+	if (!gen->fd_array)
+		gen->fd_array = add_data(gen, NULL, MAX_FD_ARRAY_SZ * sizeof(int));
+	if (gen->nr_fd_array == MAX_KFUNC_DESCS) {
+		cur = add_data(gen, NULL, sizeof(int));
+		return (cur - gen->fd_array) / sizeof(int);
+	}
+	return MAX_USED_MAPS + gen->nr_fd_array++;
+}
+
+static int blob_fd_array_off(struct bpf_gen *gen, int index)
+{
+	return gen->fd_array + index * sizeof(int);
+}
+
 static int insn_bytes_to_bpf_size(__u32 sz)
 {
 	switch (sz) {
@@ -166,14 +208,22 @@ static void emit_rel_store(struct bpf_gen *gen, int off, int data)
 	emit(gen, BPF_STX_MEM(BPF_DW, BPF_REG_1, BPF_REG_0, 0));
 }
 
-/* *(u64 *)(blob + off) = (u64)(void *)(%sp + stack_off) */
-static void emit_rel_store_sp(struct bpf_gen *gen, int off, int stack_off)
+static void move_blob2blob(struct bpf_gen *gen, int off, int size, int blob_off)
 {
-	emit(gen, BPF_MOV64_REG(BPF_REG_0, BPF_REG_10));
-	emit(gen, BPF_ALU64_IMM(BPF_ADD, BPF_REG_0, stack_off));
+	emit2(gen, BPF_LD_IMM64_RAW_FULL(BPF_REG_2, BPF_PSEUDO_MAP_IDX_VALUE,
+					 0, 0, 0, blob_off));
+	emit(gen, BPF_LDX_MEM(insn_bytes_to_bpf_size(size), BPF_REG_0, BPF_REG_2, 0));
 	emit2(gen, BPF_LD_IMM64_RAW_FULL(BPF_REG_1, BPF_PSEUDO_MAP_IDX_VALUE,
 					 0, 0, 0, off));
-	emit(gen, BPF_STX_MEM(BPF_DW, BPF_REG_1, BPF_REG_0, 0));
+	emit(gen, BPF_STX_MEM(insn_bytes_to_bpf_size(size), BPF_REG_1, BPF_REG_0, 0));
+}
+
+static void move_blob2ctx(struct bpf_gen *gen, int ctx_off, int size, int blob_off)
+{
+	emit2(gen, BPF_LD_IMM64_RAW_FULL(BPF_REG_1, BPF_PSEUDO_MAP_IDX_VALUE,
+					 0, 0, 0, blob_off));
+	emit(gen, BPF_LDX_MEM(insn_bytes_to_bpf_size(size), BPF_REG_0, BPF_REG_1, 0));
+	emit(gen, BPF_STX_MEM(insn_bytes_to_bpf_size(size), BPF_REG_6, BPF_REG_0, ctx_off));
 }
 
 static void move_ctx2blob(struct bpf_gen *gen, int off, int size, int ctx_off,
@@ -321,11 +371,11 @@ int bpf_gen__finish(struct bpf_gen *gen)
 			       offsetof(struct bpf_prog_desc, prog_fd), 4,
 			       stack_off(prog_fd[i]));
 	for (i = 0; i < gen->nr_maps; i++)
-		move_stack2ctx(gen,
-			       sizeof(struct bpf_loader_ctx) +
-			       sizeof(struct bpf_map_desc) * i +
-			       offsetof(struct bpf_map_desc, map_fd), 4,
-			       stack_off(map_fd[i]));
+		move_blob2ctx(gen,
+			      sizeof(struct bpf_loader_ctx) +
+			      sizeof(struct bpf_map_desc) * i +
+			      offsetof(struct bpf_map_desc, map_fd), 4,
+			      blob_fd_array_off(gen, i));
 	emit(gen, BPF_MOV64_IMM(BPF_REG_0, 0));
 	emit(gen, BPF_EXIT_INSN());
 	pr_debug("gen: finish %d\n", gen->error);
@@ -385,7 +435,7 @@ void bpf_gen__map_create(struct bpf_gen *gen,
 {
 	int attr_size = offsetofend(union bpf_attr, btf_vmlinux_value_type_id);
 	bool close_inner_map_fd = false;
-	int map_create_attr;
+	int map_create_attr, idx;
 	union bpf_attr attr;
 
 	memset(&attr, 0, attr_size);
@@ -462,9 +512,11 @@ void bpf_gen__map_create(struct bpf_gen *gen,
 		gen->error = -EDOM; /* internal bug */
 		return;
 	} else {
-		emit(gen, BPF_STX_MEM(BPF_W, BPF_REG_10, BPF_REG_7,
-				      stack_off(map_fd[map_idx])));
-		gen->nr_maps++;
+		/* add_map_fd does gen->nr_maps++ */
+		idx = add_map_fd(gen);
+		emit2(gen, BPF_LD_IMM64_RAW_FULL(BPF_REG_1, BPF_PSEUDO_MAP_IDX_VALUE,
+						 0, 0, 0, blob_fd_array_off(gen, idx)));
+		emit(gen, BPF_STX_MEM(BPF_W, BPF_REG_1, BPF_REG_7, 0));
 	}
 	if (close_inner_map_fd)
 		emit_sys_close_stack(gen, stack_off(inner_map_fd));
@@ -506,8 +558,8 @@ static void emit_find_attach_target(struct bpf_gen *gen)
 	 */
 }
 
-void bpf_gen__record_extern(struct bpf_gen *gen, const char *name, int kind,
-			    int insn_idx)
+void bpf_gen__record_extern(struct bpf_gen *gen, const char *name, bool is_weak,
+			    int kind, int insn_idx)
 {
 	struct ksym_relo_desc *relo;
 
@@ -519,38 +571,192 @@ void bpf_gen__record_extern(struct bpf_gen *gen, const char *name, int kind,
 	gen->relos = relo;
 	relo += gen->relo_cnt;
 	relo->name = name;
+	relo->is_weak = is_weak;
 	relo->kind = kind;
 	relo->insn_idx = insn_idx;
 	gen->relo_cnt++;
 }
 
-static void emit_relo(struct bpf_gen *gen, struct ksym_relo_desc *relo, int insns)
+/* returns existing ksym_desc with ref incremented, or inserts a new one */
+static struct ksym_desc *get_ksym_desc(struct bpf_gen *gen, struct ksym_relo_desc *relo)
 {
-	int name, insn, len = strlen(relo->name) + 1;
+	struct ksym_desc *kdesc;
 
-	pr_debug("gen: emit_relo: %s at %d\n", relo->name, relo->insn_idx);
-	name = add_data(gen, relo->name, len);
+	for (int i = 0; i < gen->nr_ksyms; i++) {
+		if (!strcmp(gen->ksyms[i].name, relo->name)) {
+			gen->ksyms[i].ref++;
+			return &gen->ksyms[i];
+		}
+	}
+	kdesc = libbpf_reallocarray(gen->ksyms, gen->nr_ksyms + 1, sizeof(*kdesc));
+	if (!kdesc) {
+		gen->error = -ENOMEM;
+		return NULL;
+	}
+	gen->ksyms = kdesc;
+	kdesc = &gen->ksyms[gen->nr_ksyms++];
+	kdesc->name = relo->name;
+	kdesc->kind = relo->kind;
+	kdesc->ref = 1;
+	kdesc->off = 0;
+	kdesc->insn = 0;
+	return kdesc;
+}
 
+/* Overwrites BPF_REG_{0, 1, 2, 3, 4, 7}
+ * Returns result in BPF_REG_7
+ */
+static void emit_bpf_find_by_name_kind(struct bpf_gen *gen, struct ksym_relo_desc *relo)
+{
+	int name_off, len = strlen(relo->name) + 1;
+
+	name_off = add_data(gen, relo->name, len);
 	emit2(gen, BPF_LD_IMM64_RAW_FULL(BPF_REG_1, BPF_PSEUDO_MAP_IDX_VALUE,
-					 0, 0, 0, name));
+					 0, 0, 0, name_off));
 	emit(gen, BPF_MOV64_IMM(BPF_REG_2, len));
 	emit(gen, BPF_MOV64_IMM(BPF_REG_3, relo->kind));
 	emit(gen, BPF_MOV64_IMM(BPF_REG_4, 0));
 	emit(gen, BPF_EMIT_CALL(BPF_FUNC_btf_find_by_name_kind));
 	emit(gen, BPF_MOV64_REG(BPF_REG_7, BPF_REG_0));
 	debug_ret(gen, "find_by_name_kind(%s,%d)", relo->name, relo->kind);
+}
+
+/* Expects:
+ * BPF_REG_8 - pointer to instruction
+ *
+ * We need to reuse BTF fd for same symbol otherwise each relocation takes a new
+ * index, while kernel limits total kfunc BTFs to 256. For duplicate symbols,
+ * this would mean a new BTF fd index for each entry. By pairing symbol name
+ * with index, we get the insn->imm, insn->off pairing that kernel uses for
+ * kfunc_tab, which becomes the effective limit even though all of them may
+ * share same index in fd_array (such that kfunc_btf_tab has 1 element).
+ */
+static void emit_relo_kfunc_btf(struct bpf_gen *gen, struct ksym_relo_desc *relo, int insn)
+{
+	struct ksym_desc *kdesc;
+	int btf_fd_idx;
+
+	kdesc = get_ksym_desc(gen, relo);
+	if (!kdesc)
+		return;
+	/* try to copy from existing bpf_insn */
+	if (kdesc->ref > 1) {
+		move_blob2blob(gen, insn + offsetof(struct bpf_insn, imm), 4,
+			       kdesc->insn + offsetof(struct bpf_insn, imm));
+		move_blob2blob(gen, insn + offsetof(struct bpf_insn, off), 2,
+			       kdesc->insn + offsetof(struct bpf_insn, off));
+		goto log;
+	}
+	/* remember insn offset, so we can copy BTF ID and FD later */
+	kdesc->insn = insn;
+	emit_bpf_find_by_name_kind(gen, relo);
+	if (!relo->is_weak)
+		emit_check_err(gen);
+	/* get index in fd_array to store BTF FD at */
+	btf_fd_idx = add_kfunc_btf_fd(gen);
+	if (btf_fd_idx > INT16_MAX) {
+		pr_warn("BTF fd off %d for kfunc %s exceeds INT16_MAX, cannot process relocation\n",
+			btf_fd_idx, relo->name);
+		gen->error = -E2BIG;
+		return;
+	}
+	kdesc->off = btf_fd_idx;
+	/* set a default value for imm */
+	emit(gen, BPF_ST_MEM(BPF_W, BPF_REG_8, offsetof(struct bpf_insn, imm), 0));
+	/* skip success case store if ret < 0 */
+	emit(gen, BPF_JMP_IMM(BPF_JSLT, BPF_REG_7, 0, 1));
+	/* store btf_id into insn[insn_idx].imm */
+	emit(gen, BPF_STX_MEM(BPF_W, BPF_REG_8, BPF_REG_7, offsetof(struct bpf_insn, imm)));
+	/* load fd_array slot pointer */
+	emit2(gen, BPF_LD_IMM64_RAW_FULL(BPF_REG_0, BPF_PSEUDO_MAP_IDX_VALUE,
+					 0, 0, 0, blob_fd_array_off(gen, btf_fd_idx)));
+	/* skip store of BTF fd if ret < 0 */
+	emit(gen, BPF_JMP_IMM(BPF_JSLT, BPF_REG_7, 0, 3));
+	/* store BTF fd in slot */
+	emit(gen, BPF_MOV64_REG(BPF_REG_9, BPF_REG_7));
+	emit(gen, BPF_ALU64_IMM(BPF_RSH, BPF_REG_9, 32));
+	emit(gen, BPF_STX_MEM(BPF_W, BPF_REG_0, BPF_REG_9, 0));
+	/* set a default value for off */
+	emit(gen, BPF_ST_MEM(BPF_H, BPF_REG_8, offsetof(struct bpf_insn, off), 0));
+	/* skip insn->off store if ret < 0 */
+	emit(gen, BPF_JMP_IMM(BPF_JSLT, BPF_REG_7, 0, 2));
+	/* skip if vmlinux BTF */
+	emit(gen, BPF_JMP_IMM(BPF_JEQ, BPF_REG_9, 0, 1));
+	/* store index into insn[insn_idx].off */
+	emit(gen, BPF_ST_MEM(BPF_H, BPF_REG_8, offsetof(struct bpf_insn, off), btf_fd_idx));
+log:
+	if (!gen->log_level)
+		return;
+	emit(gen, BPF_LDX_MEM(BPF_W, BPF_REG_7, BPF_REG_8,
+			      offsetof(struct bpf_insn, imm)));
+	emit(gen, BPF_LDX_MEM(BPF_H, BPF_REG_9, BPF_REG_8,
+			      offsetof(struct bpf_insn, off)));
+	debug_regs(gen, BPF_REG_7, BPF_REG_9, " func (%s:count=%d): imm: %%d, off: %%d",
+		   relo->name, kdesc->ref);
+	emit2(gen, BPF_LD_IMM64_RAW_FULL(BPF_REG_0, BPF_PSEUDO_MAP_IDX_VALUE,
+					 0, 0, 0, blob_fd_array_off(gen, kdesc->off)));
+	emit(gen, BPF_LDX_MEM(BPF_W, BPF_REG_9, BPF_REG_0, 0));
+	debug_regs(gen, BPF_REG_9, -1, " func (%s:count=%d): btf_fd",
+		   relo->name, kdesc->ref);
+}
+
+/* Expects:
+ * BPF_REG_8 - pointer to instruction
+ */
+static void emit_relo_ksym_btf(struct bpf_gen *gen, struct ksym_relo_desc *relo, int insn)
+{
+	struct ksym_desc *kdesc;
+
+	kdesc = get_ksym_desc(gen, relo);
+	if (!kdesc)
+		return;
+	/* try to copy from existing ldimm64 insn */
+	if (kdesc->ref > 1) {
+		move_blob2blob(gen, insn + offsetof(struct bpf_insn, imm), 4,
+			       kdesc->insn + offsetof(struct bpf_insn, imm));
+		move_blob2blob(gen, insn + sizeof(struct bpf_insn) + offsetof(struct bpf_insn, imm), 4,
+			       kdesc->insn + sizeof(struct bpf_insn) + offsetof(struct bpf_insn, imm));
+		goto log;
+	}
+	/* remember insn offset, so we can copy BTF ID and FD later */
+	kdesc->insn = insn;
+	emit_bpf_find_by_name_kind(gen, relo);
 	emit_check_err(gen);
 	/* store btf_id into insn[insn_idx].imm */
-	insn = insns + sizeof(struct bpf_insn) * relo->insn_idx +
-		offsetof(struct bpf_insn, imm);
-	emit2(gen, BPF_LD_IMM64_RAW_FULL(BPF_REG_0, BPF_PSEUDO_MAP_IDX_VALUE,
-					 0, 0, 0, insn));
-	emit(gen, BPF_STX_MEM(BPF_W, BPF_REG_0, BPF_REG_7, 0));
-	if (relo->kind == BTF_KIND_VAR) {
-		/* store btf_obj_fd into insn[insn_idx + 1].imm */
-		emit(gen, BPF_ALU64_IMM(BPF_RSH, BPF_REG_7, 32));
-		emit(gen, BPF_STX_MEM(BPF_W, BPF_REG_0, BPF_REG_7,
-				      sizeof(struct bpf_insn)));
+	emit(gen, BPF_STX_MEM(BPF_W, BPF_REG_8, BPF_REG_7, offsetof(struct bpf_insn, imm)));
+	/* store btf_obj_fd into insn[insn_idx + 1].imm */
+	emit(gen, BPF_ALU64_IMM(BPF_RSH, BPF_REG_7, 32));
+	emit(gen, BPF_STX_MEM(BPF_W, BPF_REG_8, BPF_REG_7,
+			      sizeof(struct bpf_insn) + offsetof(struct bpf_insn, imm)));
+log:
+	if (!gen->log_level)
+		return;
+	emit(gen, BPF_LDX_MEM(BPF_W, BPF_REG_7, BPF_REG_8,
+			      offsetof(struct bpf_insn, imm)));
+	emit(gen, BPF_LDX_MEM(BPF_H, BPF_REG_9, BPF_REG_8, sizeof(struct bpf_insn) +
+			      offsetof(struct bpf_insn, imm)));
+	debug_regs(gen, BPF_REG_7, BPF_REG_9, " var (%s:count=%d): imm: %%d, fd: %%d",
+		   relo->name, kdesc->ref);
+}
+
+static void emit_relo(struct bpf_gen *gen, struct ksym_relo_desc *relo, int insns)
+{
+	int insn;
+
+	pr_debug("gen: emit_relo (%d): %s at %d\n", relo->kind, relo->name, relo->insn_idx);
+	insn = insns + sizeof(struct bpf_insn) * relo->insn_idx;
+	emit2(gen, BPF_LD_IMM64_RAW_FULL(BPF_REG_8, BPF_PSEUDO_MAP_IDX_VALUE, 0, 0, 0, insn));
+	switch (relo->kind) {
+	case BTF_KIND_VAR:
+		emit_relo_ksym_btf(gen, relo, insn);
+		break;
+	case BTF_KIND_FUNC:
+		emit_relo_kfunc_btf(gen, relo, insn);
+		break;
+	default:
+		pr_warn("Unknown relocation kind '%d'\n", relo->kind);
+		gen->error = -EDOM;
+		return;
 	}
 }
 
@@ -566,14 +772,22 @@ static void cleanup_relos(struct bpf_gen *gen, int insns)
 {
 	int i, insn;
 
-	for (i = 0; i < gen->relo_cnt; i++) {
-		if (gen->relos[i].kind != BTF_KIND_VAR)
-			continue;
-		/* close fd recorded in insn[insn_idx + 1].imm */
-		insn = insns +
-			sizeof(struct bpf_insn) * (gen->relos[i].insn_idx + 1) +
-			offsetof(struct bpf_insn, imm);
-		emit_sys_close_blob(gen, insn);
+	for (i = 0; i < gen->nr_ksyms; i++) {
+		if (gen->ksyms[i].kind == BTF_KIND_VAR) {
+			/* close fd recorded in insn[insn_idx + 1].imm */
+			insn = gen->ksyms[i].insn;
+			insn += sizeof(struct bpf_insn) + offsetof(struct bpf_insn, imm);
+			emit_sys_close_blob(gen, insn);
+		} else { /* BTF_KIND_FUNC */
+			emit_sys_close_blob(gen, blob_fd_array_off(gen, gen->ksyms[i].off));
+			if (gen->ksyms[i].off < MAX_FD_ARRAY_SZ)
+				gen->nr_fd_array--;
+		}
+	}
+	if (gen->nr_ksyms) {
+		free(gen->ksyms);
+		gen->nr_ksyms = 0;
+		gen->ksyms = NULL;
 	}
 	if (gen->relo_cnt) {
 		free(gen->relos);
@@ -632,9 +846,8 @@ void bpf_gen__prog_load(struct bpf_gen *gen,
 	/* populate union bpf_attr with a pointer to line_info */
 	emit_rel_store(gen, attr_field(prog_load_attr, line_info), line_info);
 
-	/* populate union bpf_attr fd_array with a pointer to stack where map_fds are saved */
-	emit_rel_store_sp(gen, attr_field(prog_load_attr, fd_array),
-			  stack_off(map_fd[0]));
+	/* populate union bpf_attr fd_array with a pointer to data where map_fds are saved */
+	emit_rel_store(gen, attr_field(prog_load_attr, fd_array), gen->fd_array);
 
 	/* populate union bpf_attr with user provided log details */
 	move_ctx2blob(gen, attr_field(prog_load_attr, log_level), 4,
@@ -701,8 +914,8 @@ void bpf_gen__map_update_elem(struct bpf_gen *gen, int map_idx, void *pvalue,
 	emit(gen, BPF_EMIT_CALL(BPF_FUNC_copy_from_user));
 
 	map_update_attr = add_data(gen, &attr, attr_size);
-	move_stack2blob(gen, attr_field(map_update_attr, map_fd), 4,
-			stack_off(map_fd[map_idx]));
+	move_blob2blob(gen, attr_field(map_update_attr, map_fd), 4,
+		       blob_fd_array_off(gen, map_idx));
 	emit_rel_store(gen, attr_field(map_update_attr, key), key);
 	emit_rel_store(gen, attr_field(map_update_attr, value), value);
 	/* emit MAP_UPDATE_ELEM command */
@@ -720,8 +933,8 @@ void bpf_gen__map_freeze(struct bpf_gen *gen, int map_idx)
 	memset(&attr, 0, attr_size);
 	pr_debug("gen: map_freeze: idx %d\n", map_idx);
 	map_freeze_attr = add_data(gen, &attr, attr_size);
-	move_stack2blob(gen, attr_field(map_freeze_attr, map_fd), 4,
-			stack_off(map_fd[map_idx]));
+	move_blob2blob(gen, attr_field(map_freeze_attr, map_fd), 4,
+		       blob_fd_array_off(gen, map_idx));
 	/* emit MAP_FREEZE command */
 	emit_sys_bpf(gen, BPF_MAP_FREEZE, map_freeze_attr, attr_size);
 	debug_ret(gen, "map_freeze");

src/libbpf.h

@@ -83,12 +83,15 @@ struct bpf_object_open_opts {
 	 * Non-relocatable instructions are replaced with invalid ones to
 	 * prevent accidental errors.
 	 * */
+	LIBBPF_DEPRECATED_SINCE(0, 6, "field has no effect")
 	bool relaxed_core_relocs;
 	/* maps that set the 'pinning' attribute in their definition will have
 	 * their pin_path attribute set to a file in this directory, and be
 	 * auto-pinned to that path on load; defaults to "/sys/fs/bpf".
 	 */
 	const char *pin_root_path;
+
+	LIBBPF_DEPRECATED_SINCE(0, 7, "use bpf_program__set_attach_target() on each individual bpf_program")
 	__u32 attach_prog_fd;
 	/* Additional kernel config content that augments and overrides
 	 * system Kconfig for CONFIG_xxx externs.
@@ -147,6 +150,7 @@ struct bpf_object_load_attr {
 /* Load/unload object into/from kernel */
 LIBBPF_API int bpf_object__load(struct bpf_object *obj);
 LIBBPF_API int bpf_object__load_xattr(struct bpf_object_load_attr *attr);
+LIBBPF_DEPRECATED_SINCE(0, 6, "bpf_object__unload() is deprecated, use bpf_object__close() instead")
 LIBBPF_API int bpf_object__unload(struct bpf_object *obj);
 
 LIBBPF_API const char *bpf_object__name(const struct bpf_object *obj);
@@ -186,16 +190,22 @@ LIBBPF_API int libbpf_find_vmlinux_btf_id(const char *name,
 
 /* Accessors of bpf_program */
 struct bpf_program;
-LIBBPF_API struct bpf_program *bpf_program__next(struct bpf_program *prog,
-						 const struct bpf_object *obj);
+LIBBPF_API LIBBPF_DEPRECATED_SINCE(0, 7, "use bpf_object__next_program() instead")
+struct bpf_program *bpf_program__next(struct bpf_program *prog,
+				      const struct bpf_object *obj);
+LIBBPF_API struct bpf_program *
+bpf_object__next_program(const struct bpf_object *obj, struct bpf_program *prog);
 
-#define bpf_object__for_each_program(pos, obj)		\
-	for ((pos) = bpf_program__next(NULL, (obj));	\
-	     (pos) != NULL;				\
-	     (pos) = bpf_program__next((pos), (obj)))
+#define bpf_object__for_each_program(pos, obj)			\
+	for ((pos) = bpf_object__next_program((obj), NULL);	\
+	     (pos) != NULL;					\
+	     (pos) = bpf_object__next_program((obj), (pos)))
 
-LIBBPF_API struct bpf_program *bpf_program__prev(struct bpf_program *prog,
-						 const struct bpf_object *obj);
+LIBBPF_API LIBBPF_DEPRECATED_SINCE(0, 7, "use bpf_object__prev_program() instead")
+struct bpf_program *bpf_program__prev(struct bpf_program *prog,
+				      const struct bpf_object *obj);
+LIBBPF_API struct bpf_program *
+bpf_object__prev_program(const struct bpf_object *obj, struct bpf_program *prog);
 
 typedef void (*bpf_program_clear_priv_t)(struct bpf_program *, void *);
 
@@ -243,7 +253,7 @@ LIBBPF_API int bpf_link__detach(struct bpf_link *link);
 LIBBPF_API int bpf_link__destroy(struct bpf_link *link);
 
 LIBBPF_API struct bpf_link *
-bpf_program__attach(struct bpf_program *prog);
+bpf_program__attach(const struct bpf_program *prog);
 
 struct bpf_perf_event_opts {
 	/* size of this struct, for forward/backward compatiblity */
@@ -254,10 +264,10 @@ struct bpf_perf_event_opts {
 #define bpf_perf_event_opts__last_field bpf_cookie
 
 LIBBPF_API struct bpf_link *
-bpf_program__attach_perf_event(struct bpf_program *prog, int pfd);
+bpf_program__attach_perf_event(const struct bpf_program *prog, int pfd);
 
 LIBBPF_API struct bpf_link *
-bpf_program__attach_perf_event_opts(struct bpf_program *prog, int pfd,
+bpf_program__attach_perf_event_opts(const struct bpf_program *prog, int pfd,
 				    const struct bpf_perf_event_opts *opts);
 
 struct bpf_kprobe_opts {
@@ -266,7 +276,7 @@ struct bpf_kprobe_opts {
 	/* custom user-provided value fetchable through bpf_get_attach_cookie() */
 	__u64 bpf_cookie;
 	/* function's offset to install kprobe to */
-	unsigned long offset;
+	size_t offset;
 	/* kprobe is return probe */
 	bool retprobe;
 	size_t :0;
@@ -274,10 +284,10 @@ struct bpf_kprobe_opts {
 #define bpf_kprobe_opts__last_field retprobe
 
 LIBBPF_API struct bpf_link *
-bpf_program__attach_kprobe(struct bpf_program *prog, bool retprobe,
+bpf_program__attach_kprobe(const struct bpf_program *prog, bool retprobe,
 			   const char *func_name);
 LIBBPF_API struct bpf_link *
-bpf_program__attach_kprobe_opts(struct bpf_program *prog,
+bpf_program__attach_kprobe_opts(const struct bpf_program *prog,
                                 const char *func_name,
                                 const struct bpf_kprobe_opts *opts);
 
@@ -297,11 +307,11 @@ struct bpf_uprobe_opts {
 #define bpf_uprobe_opts__last_field retprobe
 
 LIBBPF_API struct bpf_link *
-bpf_program__attach_uprobe(struct bpf_program *prog, bool retprobe,
+bpf_program__attach_uprobe(const struct bpf_program *prog, bool retprobe,
 			   pid_t pid, const char *binary_path,
 			   size_t func_offset);
 LIBBPF_API struct bpf_link *
-bpf_program__attach_uprobe_opts(struct bpf_program *prog, pid_t pid,
+bpf_program__attach_uprobe_opts(const struct bpf_program *prog, pid_t pid,
 				const char *binary_path, size_t func_offset,
 				const struct bpf_uprobe_opts *opts);
 
@@ -314,35 +324,35 @@ struct bpf_tracepoint_opts {
 #define bpf_tracepoint_opts__last_field bpf_cookie
 
 LIBBPF_API struct bpf_link *
-bpf_program__attach_tracepoint(struct bpf_program *prog,
+bpf_program__attach_tracepoint(const struct bpf_program *prog,
 			       const char *tp_category,
 			       const char *tp_name);
 LIBBPF_API struct bpf_link *
-bpf_program__attach_tracepoint_opts(struct bpf_program *prog,
+bpf_program__attach_tracepoint_opts(const struct bpf_program *prog,
 				    const char *tp_category,
 				    const char *tp_name,
 				    const struct bpf_tracepoint_opts *opts);
 
 LIBBPF_API struct bpf_link *
-bpf_program__attach_raw_tracepoint(struct bpf_program *prog,
+bpf_program__attach_raw_tracepoint(const struct bpf_program *prog,
 				   const char *tp_name);
 LIBBPF_API struct bpf_link *
-bpf_program__attach_trace(struct bpf_program *prog);
+bpf_program__attach_trace(const struct bpf_program *prog);
 LIBBPF_API struct bpf_link *
-bpf_program__attach_lsm(struct bpf_program *prog);
+bpf_program__attach_lsm(const struct bpf_program *prog);
 LIBBPF_API struct bpf_link *
-bpf_program__attach_cgroup(struct bpf_program *prog, int cgroup_fd);
+bpf_program__attach_cgroup(const struct bpf_program *prog, int cgroup_fd);
 LIBBPF_API struct bpf_link *
-bpf_program__attach_netns(struct bpf_program *prog, int netns_fd);
+bpf_program__attach_netns(const struct bpf_program *prog, int netns_fd);
 LIBBPF_API struct bpf_link *
-bpf_program__attach_xdp(struct bpf_program *prog, int ifindex);
+bpf_program__attach_xdp(const struct bpf_program *prog, int ifindex);
 LIBBPF_API struct bpf_link *
-bpf_program__attach_freplace(struct bpf_program *prog,
+bpf_program__attach_freplace(const struct bpf_program *prog,
 			     int target_fd, const char *attach_func_name);
 
 struct bpf_map;
 
-LIBBPF_API struct bpf_link *bpf_map__attach_struct_ops(struct bpf_map *map);
+LIBBPF_API struct bpf_link *bpf_map__attach_struct_ops(const struct bpf_map *map);
 
 struct bpf_iter_attach_opts {
 	size_t sz; /* size of this struct for forward/backward compatibility */
@@ -352,7 +362,7 @@ struct bpf_iter_attach_opts {
 #define bpf_iter_attach_opts__last_field link_info_len
 
 LIBBPF_API struct bpf_link *
-bpf_program__attach_iter(struct bpf_program *prog,
+bpf_program__attach_iter(const struct bpf_program *prog,
 			 const struct bpf_iter_attach_opts *opts);
 
 struct bpf_insn;
@@ -478,9 +488,13 @@ struct bpf_map_def {
 	unsigned int map_flags;
 };
 
-/*
- * The 'struct bpf_map' in include/linux/bpf.h is internal to the kernel,
- * so no need to worry about a name clash.
+/**
+ * @brief **bpf_object__find_map_by_name()** returns BPF map of
+ * the given name, if it exists within the passed BPF object
+ * @param obj BPF object
+ * @param name name of the BPF map
+ * @return BPF map instance, if such map exists within the BPF object;
+ * or NULL otherwise.
  */
 LIBBPF_API struct bpf_map *
 bpf_object__find_map_by_name(const struct bpf_object *obj, const char *name);
@@ -495,18 +509,28 @@ bpf_object__find_map_fd_by_name(const struct bpf_object *obj, const char *name);
 LIBBPF_API struct bpf_map *
 bpf_object__find_map_by_offset(struct bpf_object *obj, size_t offset);
 
+LIBBPF_API LIBBPF_DEPRECATED_SINCE(0, 7, "use bpf_object__next_map() instead")
+struct bpf_map *bpf_map__next(const struct bpf_map *map, const struct bpf_object *obj);
 LIBBPF_API struct bpf_map *
-bpf_map__next(const struct bpf_map *map, const struct bpf_object *obj);
+bpf_object__next_map(const struct bpf_object *obj, const struct bpf_map *map);
+
 #define bpf_object__for_each_map(pos, obj)		\
-	for ((pos) = bpf_map__next(NULL, (obj));	\
+	for ((pos) = bpf_object__next_map((obj), NULL);	\
 	     (pos) != NULL;				\
-	     (pos) = bpf_map__next((pos), (obj)))
+	     (pos) = bpf_object__next_map((obj), (pos)))
 #define bpf_map__for_each bpf_object__for_each_map
 
+LIBBPF_API LIBBPF_DEPRECATED_SINCE(0, 7, "use bpf_object__prev_map() instead")
+struct bpf_map *bpf_map__prev(const struct bpf_map *map, const struct bpf_object *obj);
 LIBBPF_API struct bpf_map *
-bpf_map__prev(const struct bpf_map *map, const struct bpf_object *obj);
+bpf_object__prev_map(const struct bpf_object *obj, const struct bpf_map *map);
 
-/* get/set map FD */
+/**
+ * @brief **bpf_map__fd()** gets the file descriptor of the passed
+ * BPF map
+ * @param map the BPF map instance
+ * @return the file descriptor; or -EINVAL in case of an error
+ */
 LIBBPF_API int bpf_map__fd(const struct bpf_map *map);
 LIBBPF_API int bpf_map__reuse_fd(struct bpf_map *map, int fd);
 /* get map definition */
@@ -547,6 +571,14 @@ LIBBPF_API int bpf_map__set_initial_value(struct bpf_map *map,
 					  const void *data, size_t size);
 LIBBPF_API const void *bpf_map__initial_value(struct bpf_map *map, size_t *psize);
 LIBBPF_API bool bpf_map__is_offload_neutral(const struct bpf_map *map);
+
+/**
+ * @brief **bpf_map__is_internal()** tells the caller whether or not the
+ * passed map is a special map created by libbpf automatically for things like
+ * global variables, __ksym externs, Kconfig values, etc
+ * @param map the bpf_map
+ * @return true, if the map is an internal map; false, otherwise
+ */
 LIBBPF_API bool bpf_map__is_internal(const struct bpf_map *map);
 LIBBPF_API int bpf_map__set_pin_path(struct bpf_map *map, const char *path);
 LIBBPF_API const char *bpf_map__get_pin_path(const struct bpf_map *map);
@@ -558,6 +590,38 @@ LIBBPF_API int bpf_map__unpin(struct bpf_map *map, const char *path);
 LIBBPF_API int bpf_map__set_inner_map_fd(struct bpf_map *map, int fd);
 LIBBPF_API struct bpf_map *bpf_map__inner_map(struct bpf_map *map);
 
+/**
+ * @brief **libbpf_get_error()** extracts the error code from the passed
+ * pointer
+ * @param ptr pointer returned from libbpf API function
+ * @return error code; or 0 if no error occured
+ *
+ * Many libbpf API functions which return pointers have logic to encode error
+ * codes as pointers, and do not return NULL. Meaning **libbpf_get_error()**
+ * should be used on the return value from these functions immediately after
+ * calling the API function, with no intervening calls that could clobber the
+ * `errno` variable. Consult the individual functions documentation to verify
+ * if this logic applies should be used.
+ *
+ * For these API functions, if `libbpf_set_strict_mode(LIBBPF_STRICT_CLEAN_PTRS)`
+ * is enabled, NULL is returned on error instead.
+ *
+ * If ptr is NULL, then errno should be already set by the failing
+ * API, because libbpf never returns NULL on success and it now always
+ * sets errno on error.
+ *
+ * Example usage:
+ *
+ *   struct perf_buffer *pb;
+ *
+ *   pb = perf_buffer__new(bpf_map__fd(obj->maps.events), PERF_BUFFER_PAGES, &opts);
+ *   err = libbpf_get_error(pb);
+ *   if (err) {
+ *	  pb = NULL;
+ *	  fprintf(stderr, "failed to open perf buffer: %d\n", err);
+ *	  goto cleanup;
+ *   }
+ */
 LIBBPF_API long libbpf_get_error(const void *ptr);
 
 struct bpf_prog_load_attr {
@@ -822,9 +886,10 @@ bpf_program__bpil_addr_to_offs(struct bpf_prog_info_linear *info_linear);
 LIBBPF_API void
 bpf_program__bpil_offs_to_addr(struct bpf_prog_info_linear *info_linear);
 
-/*
- * A helper function to get the number of possible CPUs before looking up
- * per-CPU maps. Negative errno is returned on failure.
+/**
+ * @brief **libbpf_num_possible_cpus()** is a helper function to get the
+ * number of possible CPUs that the host kernel supports and expects.
+ * @return number of possible CPUs; or error code on failure
  *
  * Example usage:
  *
@@ -834,7 +899,6 @@ bpf_program__bpil_offs_to_addr(struct bpf_prog_info_linear *info_linear);
  *     }
  *     long values[ncpus];
  *     bpf_map_lookup_elem(per_cpu_map_fd, key, values);
- *
  */
 LIBBPF_API int libbpf_num_possible_cpus(void);
 
@@ -854,7 +918,7 @@ struct bpf_object_skeleton {
 	size_t sz; /* size of this struct, for forward/backward compatibility */
 
 	const char *name;
-	void *data;
+	const void *data;
 	size_t data_sz;
 
 	struct bpf_object **obj;

src/libbpf.map

@@ -386,3 +386,13 @@ LIBBPF_0.5.0 {
 		btf_dump__dump_type_data;
 		libbpf_set_strict_mode;
 } LIBBPF_0.4.0;
+
+LIBBPF_0.6.0 {
+	global:
+		bpf_object__next_map;
+		bpf_object__next_program;
+		bpf_object__prev_map;
+		bpf_object__prev_program;
+		btf__add_btf;
+		btf__add_tag;
+} LIBBPF_0.5.0;

src/libbpf_common.h

@@ -10,6 +10,7 @@
 #define __LIBBPF_LIBBPF_COMMON_H
 
 #include <string.h>
+#include "libbpf_version.h"
 
 #ifndef LIBBPF_API
 #define LIBBPF_API __attribute__((visibility("default")))
@@ -17,6 +18,29 @@
 
 #define LIBBPF_DEPRECATED(msg) __attribute__((deprecated(msg)))
 
+/* Mark a symbol as deprecated when libbpf version is >= {major}.{minor} */
+#define LIBBPF_DEPRECATED_SINCE(major, minor, msg)			    \
+	__LIBBPF_MARK_DEPRECATED_ ## major ## _ ## minor		    \
+		(LIBBPF_DEPRECATED("libbpf v" # major "." # minor "+: " msg))
+
+#define __LIBBPF_CURRENT_VERSION_GEQ(major, minor)			    \
+	(LIBBPF_MAJOR_VERSION > (major) ||				    \
+	 (LIBBPF_MAJOR_VERSION == (major) && LIBBPF_MINOR_VERSION >= (minor)))
+
+/* Add checks for other versions below when planning deprecation of API symbols
+ * with the LIBBPF_DEPRECATED_SINCE macro.
+ */
+#if __LIBBPF_CURRENT_VERSION_GEQ(0, 6)
+#define __LIBBPF_MARK_DEPRECATED_0_6(X) X
+#else
+#define __LIBBPF_MARK_DEPRECATED_0_6(X)
+#endif
+#if __LIBBPF_CURRENT_VERSION_GEQ(0, 7)
+#define __LIBBPF_MARK_DEPRECATED_0_7(X) X
+#else
+#define __LIBBPF_MARK_DEPRECATED_0_7(X)
+#endif
+
 /* Helper macro to declare and initialize libbpf options struct
  *
  * This dance with uninitialized declaration, followed by memset to zero,

src/libbpf_internal.h

@@ -69,6 +69,8 @@
 #define BTF_VAR_SECINFO_ENC(type, offset, size) (type), (offset), (size)
 #define BTF_TYPE_FLOAT_ENC(name, sz) \
 	BTF_TYPE_ENC(name, BTF_INFO_ENC(BTF_KIND_FLOAT, 0, 0), sz)
+#define BTF_TYPE_TAG_ENC(value, type, component_idx) \
+	BTF_TYPE_ENC(value, BTF_INFO_ENC(BTF_KIND_TAG, 0, 0), type), (component_idx)
 
 #ifndef likely
 #define likely(x) __builtin_expect(!!(x), 1)
@@ -87,20 +89,40 @@
 	(offsetof(TYPE, FIELD) + sizeof(((TYPE *)0)->FIELD))
 #endif
 
+/* Check whether a string `str` has prefix `pfx`, regardless if `pfx` is
+ * a string literal known at compilation time or char * pointer known only at
+ * runtime.
+ */
+#define str_has_pfx(str, pfx) \
+	(strncmp(str, pfx, __builtin_constant_p(pfx) ? sizeof(pfx) - 1 : strlen(pfx)) == 0)
+
 /* Symbol versioning is different between static and shared library.
  * Properly versioned symbols are needed for shared library, but
  * only the symbol of the new version is needed for static library.
+ * Starting with GNU C 10, use symver attribute instead of .symver assembler
+ * directive, which works better with GCC LTO builds.
  */
-#ifdef SHARED
-# define COMPAT_VERSION(internal_name, api_name, version) \
+#if defined(SHARED) && defined(__GNUC__) && __GNUC__ >= 10
+
+#define DEFAULT_VERSION(internal_name, api_name, version) \
+	__attribute__((symver(#api_name "@@" #version)))
+#define COMPAT_VERSION(internal_name, api_name, version) \
+	__attribute__((symver(#api_name "@" #version)))
+
+#elif defined(SHARED)
+
+#define COMPAT_VERSION(internal_name, api_name, version) \
 	asm(".symver " #internal_name "," #api_name "@" #version);
-# define DEFAULT_VERSION(internal_name, api_name, version) \
+#define DEFAULT_VERSION(internal_name, api_name, version) \
 	asm(".symver " #internal_name "," #api_name "@@" #version);
-#else
-# define COMPAT_VERSION(internal_name, api_name, version)
-# define DEFAULT_VERSION(internal_name, api_name, version) \
+
+#else /* !SHARED */
+
+#define COMPAT_VERSION(internal_name, api_name, version)
+#define DEFAULT_VERSION(internal_name, api_name, version) \
 	extern typeof(internal_name) api_name \
 	__attribute__((alias(#internal_name)));
+
 #endif
 
 extern void libbpf_print(enum libbpf_print_level level,
@@ -276,6 +298,7 @@ struct bpf_prog_load_params {
 	__u32 log_level;
 	char *log_buf;
 	size_t log_buf_sz;
+	int *fd_array;
 };
 
 int libbpf__bpf_prog_load(const struct bpf_prog_load_params *load_attr);
@@ -386,6 +409,8 @@ int btf_type_visit_type_ids(struct btf_type *t, type_id_visit_fn visit, void *ct
 int btf_type_visit_str_offs(struct btf_type *t, str_off_visit_fn visit, void *ctx);
 int btf_ext_visit_type_ids(struct btf_ext *btf_ext, type_id_visit_fn visit, void *ctx);
 int btf_ext_visit_str_offs(struct btf_ext *btf_ext, str_off_visit_fn visit, void *ctx);
+__s32 btf__find_by_name_kind_own(const struct btf *btf, const char *type_name,
+				 __u32 kind);
 
 extern enum libbpf_strict_mode libbpf_mode;
 

src/libbpf_legacy.h

@@ -46,6 +46,15 @@ enum libbpf_strict_mode {
 	 */
 	LIBBPF_STRICT_DIRECT_ERRS = 0x02,
 
+	/*
+	 * Enforce strict BPF program section (SEC()) names.
+	 * E.g., while prefiously SEC("xdp_whatever") or SEC("perf_event_blah") were
+	 * allowed, with LIBBPF_STRICT_SEC_PREFIX this will become
+	 * unrecognized by libbpf and would have to be just SEC("xdp") and
+	 * SEC("xdp") and SEC("perf_event").
+	 */
+	LIBBPF_STRICT_SEC_NAME = 0x04,
+
 	__LIBBPF_STRICT_LAST,
 };
 

src/libbpf_version.h

@@ -0,0 +1,9 @@
+/* SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause) */
+/* Copyright (C) 2021 Facebook */
+#ifndef __LIBBPF_VERSION_H
+#define __LIBBPF_VERSION_H
+
+#define LIBBPF_MAJOR_VERSION 0
+#define LIBBPF_MINOR_VERSION 6
+
+#endif /* __LIBBPF_VERSION_H */

src/linker.c

@@ -1649,11 +1649,17 @@ static bool btf_is_non_static(const struct btf_type *t)
 static int find_glob_sym_btf(struct src_obj *obj, Elf64_Sym *sym, const char *sym_name,
 			     int *out_btf_sec_id, int *out_btf_id)
 {
-	int i, j, n = btf__get_nr_types(obj->btf), m, btf_id = 0;
+	int i, j, n, m, btf_id = 0;
 	const struct btf_type *t;
 	const struct btf_var_secinfo *vi;
 	const char *name;
 
+	if (!obj->btf) {
+		pr_warn("failed to find BTF info for object '%s'\n", obj->filename);
+		return -EINVAL;
+	}
+
+	n = btf__get_nr_types(obj->btf);
 	for (i = 1; i <= n; i++) {
 		t = btf__type_by_id(obj->btf, i);
 

src/skel_internal.h

@@ -105,10 +105,12 @@ static inline int bpf_load_and_run(struct bpf_load_and_run_opts *opts)
 	err = skel_sys_bpf(BPF_PROG_RUN, &attr, sizeof(attr));
 	if (err < 0 || (int)attr.test.retval < 0) {
 		opts->errstr = "failed to execute loader prog";
-		if (err < 0)
+		if (err < 0) {
 			err = -errno;
-		else
+		} else {
 			err = (int)attr.test.retval;
+			errno = -err;
+		}
 		goto out;
 	}
 	err = 0;

src/strset.c

@@ -88,6 +88,7 @@ void strset__free(struct strset *set)
 
 	hashmap__free(set->strs_hash);
 	free(set->strs_data);
+	free(set);
 }
 
 size_t strset__data_size(const struct strset *set)

src/xsk.c

@@ -281,6 +281,7 @@ out_mmap:
 	return err;
 }
 
+DEFAULT_VERSION(xsk_umem__create_v0_0_4, xsk_umem__create, LIBBPF_0.0.4)
 int xsk_umem__create_v0_0_4(struct xsk_umem **umem_ptr, void *umem_area,
 			    __u64 size, struct xsk_ring_prod *fill,
 			    struct xsk_ring_cons *comp,
@@ -345,6 +346,7 @@ struct xsk_umem_config_v1 {
 	__u32 frame_headroom;
 };
 
+COMPAT_VERSION(xsk_umem__create_v0_0_2, xsk_umem__create, LIBBPF_0.0.2)
 int xsk_umem__create_v0_0_2(struct xsk_umem **umem_ptr, void *umem_area,
 			    __u64 size, struct xsk_ring_prod *fill,
 			    struct xsk_ring_cons *comp,
@@ -358,8 +360,6 @@ int xsk_umem__create_v0_0_2(struct xsk_umem **umem_ptr, void *umem_area,
 	return xsk_umem__create_v0_0_4(umem_ptr, umem_area, size, fill, comp,
 					&config);
 }
-COMPAT_VERSION(xsk_umem__create_v0_0_2, xsk_umem__create, LIBBPF_0.0.2)
-DEFAULT_VERSION(xsk_umem__create_v0_0_4, xsk_umem__create, LIBBPF_0.0.4)
 
 static enum xsk_prog get_xsk_prog(void)
 {

travis-ci/vmtest/configs/blacklist/BLACKLIST-5.5.0

@@ -45,6 +45,7 @@ mmap			# 5.5 kernel is too permissive with re-mmaping
 modify_return		# fmod_ret support is missing
 module_attach		# module BTF support missing (v5.11+)
 netcnt
+netns_cookie		# v5.15+
 ns_current_pid_tgid	# bpf_get_ns_current_pid_tgid() helper is missing
 pe_preserve_elems	# v5.10+
 perf_branches		# bpf_read_branch_records() helper is missing
@@ -74,10 +75,12 @@ socket_cookie		# v5.12+
 sockmap_basic		# uses new socket fields, 5.8+
 sockmap_listen		# no listen socket supportin SOCKMAP
 sockopt_sk
+sockopt_qos_to_cc	# v5.15+
 stacktrace_build_id	# v5.9+
 stack_var_off		# v5.12+
 syscall			# v5.14+
 task_local_storage	# v5.12+
+task_pt_regs		# v5.15+
 tcp_hdr_options		# v5.10+, new TCP header options feature in BPF
 tcpbpf_user		# LINK_CREATE is missing
 tc_redirect		# v5.14+

travis-ci/vmtest/configs/blacklist/BLACKLIST-latest

@@ -1 +1,5 @@
 # TEMPORARY
+get_stack_raw_tp    # spams with kernel warnings until next bpf -> bpf-next merge
+stacktrace_build_id_nmi
+stacktrace_build_id
+task_fd_query_rawtp

travis-ci/vmtest/configs/whitelist/WHITELIST-5.5.0

@@ -0,0 +1,55 @@
+attach_probe
+autoload
+bpf_verif_scale
+cgroup_attach_autodetach
+cgroup_attach_override
+core_autosize
+core_extern
+core_read_macros
+core_reloc
+core_retro
+cpu_mask
+endian
+fexit_stress
+get_branch_snapshot
+get_stackid_cannot_attach
+global_data
+global_data_init
+global_func_args
+hashmap
+l4lb_all
+linked_funcs
+linked_maps
+map_lock
+obj_name
+perf_buffer
+perf_event_stackmap
+pinning
+pkt_md_access
+probe_user
+queue_stack_map
+raw_tp_writable_reject_nbd_invalid
+raw_tp_writable_test_run
+rdonly_maps
+section_names
+signal_pending
+skeleton
+sockmap_ktls
+sockopt
+sockopt_inherit
+sockopt_multi
+spinlock
+stacktrace_map
+stacktrace_map_raw_tp
+static_linked
+subprogs
+task_fd_query_rawtp
+task_fd_query_tp
+tc_bpf
+tcp_estats
+tcp_rtt
+tp_attach_query
+xdp
+xdp_info
+xdp_noinline
+xdp_perf

travis-ci/vmtest/run.sh

@@ -459,10 +459,10 @@ if kvm-ok ; then
 else
   accel="-cpu qemu64 -machine accel=tcg"
 fi
-qemu-system-x86_64 -nodefaults -display none -serial mon:stdio \
+qemu-system-x86_64 -nodefaults -display none -serial mon:stdio -no-reboot \
   ${accel} -smp "$(nproc)" -m 4G \
   -drive file="$IMG",format=raw,index=1,media=disk,if=virtio,cache=none \
-  -kernel "$vmlinuz" -append "root=/dev/vda rw console=ttyS0,115200$APPEND"
+  -kernel "$vmlinuz" -append "root=/dev/vda rw console=ttyS0,115200 kernel.panic=-1 $APPEND"
 sudo mount -o loop "$IMG" "$mnt"
 if exitstatus="$(cat "$mnt/exitstatus" 2>/dev/null)"; then
 	printf '\nTests exit status: %s\n' "$exitstatus" >&2

travis-ci/vmtest/run_selftests.sh

@@ -7,12 +7,12 @@ source $(cd $(dirname $0) && pwd)/helpers.sh
 test_progs() {
 	if [[ "${KERNEL}" != '4.9.0' ]]; then
 		travis_fold start test_progs "Testing test_progs"
-		./test_progs ${BLACKLIST:+-b$BLACKLIST} ${WHITELIST:+-t$WHITELIST}
+		./test_progs ${BLACKLIST:+-d$BLACKLIST} ${WHITELIST:+-a$WHITELIST}
 		travis_fold end test_progs
 	fi
 
 	travis_fold start test_progs-no_alu32 "Testing test_progs-no_alu32"
-	./test_progs-no_alu32 ${BLACKLIST:+-b$BLACKLIST} ${WHITELIST:+-t$WHITELIST}
+	./test_progs-no_alu32 ${BLACKLIST:+-d$BLACKLIST} ${WHITELIST:+-a$WHITELIST}
 	travis_fold end test_progs-no_alu32
 }