Merge https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next

Andrii Nakryiko says:

====================
bpf-next 2021-12-10 v2

We've added 115 non-merge commits during the last 26 day(s) which contain
a total of 182 files changed, 5747 insertions(+), 2564 deletions(-).

The main changes are:

1) Various samples fixes, from Alexander Lobakin.

2) BPF CO-RE support in kernel and light skeleton, from Alexei Starovoitov.

3) A batch of new unified APIs for libbpf, logging improvements, version
querying, etc. Also a batch of old deprecations for old APIs and various
bug fixes, in preparation for libbpf 1.0, from Andrii Nakryiko.

4) BPF documentation reorganization and improvements, from Christoph Hellwig
and Dave Tucker.

5) Support for declarative initialization of BPF_MAP_TYPE_PROG_ARRAY in
libbpf, from Hengqi Chen.

6) Verifier log fixes, from Hou Tao.

7) Runtime-bounded loops support with bpf_loop() helper, from Joanne Koong.

8) Extend branch record capturing to all platforms that support it,
from Kajol Jain.

9) Light skeleton codegen improvements, from Kumar Kartikeya Dwivedi.

10) bpftool doc-generating script improvements, from Quentin Monnet.

11) Two libbpf v0.6 bug fixes, from Shuyi Cheng and Vincent Minet.

12) Deprecation warning fix for perf/bpf_counter, from Song Liu.

13) MAX_TAIL_CALL_CNT unification and MIPS build fix for libbpf,
from Tiezhu Yang.

14) BTF_KING_TYPE_TAG follow-up fixes, from Yonghong Song.

15) Selftests fixes and improvements, from Ilya Leoshkevich, Jean-Philippe
Brucker, Jiri Olsa, Maxim Mikityanskiy, Tirthendu Sarkar, Yucong Sun,
and others.

* https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next: (115 commits)
libbpf: Add "bool skipped" to struct bpf_map
libbpf: Fix typo in btf__dedup@LIBBPF_0.0.2 definition
bpftool: Switch bpf_object__load_xattr() to bpf_object__load()
selftests/bpf: Remove the only use of deprecated bpf_object__load_xattr()
selftests/bpf: Add test for libbpf's custom log_buf behavior
selftests/bpf: Replace all uses of bpf_load_btf() with bpf_btf_load()
libbpf: Deprecate bpf_object__load_xattr()
libbpf: Add per-program log buffer setter and getter
libbpf: Preserve kernel error code and remove kprobe prog type guessing
libbpf: Improve logging around BPF program loading
libbpf: Allow passing user log setting through bpf_object_open_opts
libbpf: Allow passing preallocated log_buf when loading BTF into kernel
libbpf: Add OPTS-based bpf_btf_load() API
libbpf: Fix bpf_prog_load() log_buf logic for log_level 0
samples/bpf: Remove unneeded variable
bpf: Remove redundant assignment to pointer t
selftests/bpf: Fix a compilation warning
perf/bpf_counter: Use bpf_map_create instead of bpf_create_map
samples: bpf: Fix 'unknown warning group' build warning on Clang
samples: bpf: Fix xdp_sample_user.o linking with Clang
...
====================

Link: https://lore.kernel.org/r/[e-mail]
Signed-off-by: Jakub Kicinski <[e-mail]>

Documentation/bpf/btf.rst

···
 =====================
 
 1. Introduction
-***************
+===============
 
 BTF (BPF Type Format) is the metadata format which encodes the debug info
 related to BPF program/map. The name BTF was used initially to describe data
···
 .. _BTF_Type_String:
 
 2. BTF Type and String Encoding
-*******************************
+===============================
 
 The file ``include/uapi/linux/btf.h`` provides high-level definition of how
 types/strings are encoded.
···
 generated.
 
 2.1 String Encoding
-===================
+-------------------
 
 The first string in the string section must be a null string. The rest of
 string table is a concatenation of other null-terminated strings.
 
 2.2 Type Encoding
-=================
+-----------------
 
 The type id ``0`` is reserved for ``void`` type. The type section is parsed
 sequentially and type id is assigned to each recognized type starting from id
···
  * ``type``: the type with ``btf_type_tag`` attribute
 
 3. BTF Kernel API
-*****************
+=================
 
 The following bpf syscall command involves BTF:
    * BPF_BTF_LOAD: load a blob of BTF data into kernel
···
 
 
 3.1 BPF_BTF_LOAD
-================
+----------------
 
 Load a blob of BTF data into kernel. A blob of data, described in
 :ref:`BTF_Type_String`, can be directly loaded into the kernel. A ``btf_fd``
 is returned to a userspace.
 
 3.2 BPF_MAP_CREATE
-==================
+------------------
 
 A map can be created with ``btf_fd`` and specified key/value type id.::
 
···
 .. _BPF_Prog_Load:
 
 3.3 BPF_PROG_LOAD
-=================
+-----------------
 
 During prog_load, func_info and line_info can be passed to kernel with proper
 values for the following attributes:
···
     #define BPF_LINE_INFO_LINE_COL(line_col)        ((line_col) & 0x3ff)
 
 3.4 BPF_{PROG,MAP}_GET_NEXT_ID
-==============================
+------------------------------
 
 In kernel, every loaded program, map or btf has a unique id. The id won't
 change during the lifetime of a program, map, or btf.
···
 inspection tool can inspect all programs and maps.
 
 3.5 BPF_{PROG,MAP}_GET_FD_BY_ID
-===============================
+-------------------------------
 
 An introspection tool cannot use id to get details about program or maps.
 A file descriptor needs to be obtained first for reference-counting purpose.
 
 3.6 BPF_OBJ_GET_INFO_BY_FD
-==========================
+--------------------------
 
 Once a program/map fd is acquired, an introspection tool can get the detailed
 information from kernel about this fd, some of which are BTF-related. For
···
 bpf byte codes, and jited_line_info.
 
 3.7 BPF_BTF_GET_FD_BY_ID
-========================
+------------------------
 
 With ``btf_id`` obtained in ``bpf_map_info`` and ``bpf_prog_info``, bpf
 syscall command BPF_BTF_GET_FD_BY_ID can retrieve a btf fd. Then, with
···
 func signatures and line info, along with byte/jit codes.
 
 4. ELF File Format Interface
-****************************
+============================
 
 4.1 .BTF section
-================
+----------------
 
 The .BTF section contains type and string data. The format of this section is
 same as the one describe in :ref:`BTF_Type_String`.
···
 .. _BTF_Ext_Section:
 
 4.2 .BTF.ext section
-====================
+--------------------
 
 The .BTF.ext section encodes func_info and line_info which needs loader
 manipulation before loading into the kernel.
···
 beginning of section (``btf_ext_info_sec->sec_name_off``).
 
 4.2 .BTF_ids section
-====================
+--------------------
 
 The .BTF_ids section encodes BTF ID values that are used within the kernel.
 
···
 resolved during the linking phase of kernel build by ``resolve_btfids`` tool.
 
 5. Using BTF
-************
+============
 
 5.1 bpftool map pretty print
-============================
+----------------------------
 
 With BTF, the map key/value can be printed based on fields rather than simply
 raw bytes. This is especially valuable for large structure or if your data
···
       ]
 
 5.2 bpftool prog dump
-=====================
+---------------------
 
 The following is an example showing how func_info and line_info can help prog
 dump with better kernel symbol names, function prototypes and line
···
     [...]
 
 5.3 Verifier Log
-================
+----------------
 
 The following is an example of how line_info can help debugging verification
 failure.::
···
         R2 offset is outside of the packet
 
 6. BTF Generation
-*****************
+=================
 
 You need latest pahole
 
···
             .long   8206                    # Line 8 Col 14
 
 7. Testing
-**********
+==========
 
 Kernel bpf selftest `test_btf.c` provides extensive set of BTF-related tests.

Documentation/bpf/faq.rst

···
+================================
+Frequently asked questions (FAQ)
+================================
+
+Two sets of Questions and Answers (Q&A) are maintained.
+
+.. toctree::
+   :maxdepth: 1
+
+   bpf_design_QA
+   bpf_devel_QA

Documentation/bpf/helpers.rst

···
+Helper functions
+================
+
+* `bpf-helpers(7)`_ maintains a list of helpers available to eBPF programs.
+
+.. Links
+.. _bpf-helpers(7): https://man7.org/linux/man-pages/man7/bpf-helpers.7.html

Documentation/bpf/index.rst

···
 This directory contains documentation for the BPF (Berkeley Packet
 Filter) facility, with a focus on the extended BPF version (eBPF).
 
-This kernel side documentation is still work in progress. The main
-textual documentation is (for historical reasons) described in
-:ref:`networking-filter`, which describe both classical and extended
-BPF instruction-set.
+This kernel side documentation is still work in progress.
 The Cilium project also maintains a `BPF and XDP Reference Guide`_
 that goes into great technical depth about the BPF Architecture.
 
-libbpf
-======
-
-Documentation/bpf/libbpf/index.rst is a userspace library for loading and interacting with bpf programs.
-
-BPF Type Format (BTF)
-=====================
-
 .. toctree::
    :maxdepth: 1
 
+   instruction-set
+   verifier
+   libbpf/index
    btf
-
-
-Frequently asked questions (FAQ)
-================================
-
-Two sets of Questions and Answers (Q&A) are maintained.
-
-.. toctree::
-   :maxdepth: 1
-
-   bpf_design_QA
-   bpf_devel_QA
-
-Syscall API
-===========
-
-The primary info for the bpf syscall is available in the `man-pages`_
-for `bpf(2)`_. For more information about the userspace API, see
-Documentation/userspace-api/ebpf/index.rst.
-
-Helper functions
-================
-
-* `bpf-helpers(7)`_ maintains a list of helpers available to eBPF programs.
-
-
-Program types
-=============
-
-.. toctree::
-   :maxdepth: 1
-
-   prog_cgroup_sockopt
-   prog_cgroup_sysctl
-   prog_flow_dissector
-   bpf_lsm
-   prog_sk_lookup
-
-
-Map types
-=========
-
-.. toctree::
-   :maxdepth: 1
-
-   map_cgroup_storage
-
-
-Testing and debugging BPF
-=========================
-
-.. toctree::
-   :maxdepth: 1
-
-   drgn
-   s390
-
-
-Licensing
-=========
-
-.. toctree::
-   :maxdepth: 1
-
+   faq
+   syscall_api
+   helpers
+   programs
+   maps
    bpf_licensing
+   test_debug
+   other
 
+.. only::  subproject and html
 
-Other
-=====
+   Indices
+   =======
 
-.. toctree::
-   :maxdepth: 1
-
-   ringbuf
-   llvm_reloc
+   * :ref:`genindex`
 
 .. Links:
-.. _networking-filter: ../networking/filter.rst
-.. _man-pages: https://www.kernel.org/doc/man-pages/
-.. _bpf(2): https://man7.org/linux/man-pages/man2/bpf.2.html
-.. _bpf-helpers(7): https://man7.org/linux/man-pages/man7/bpf-helpers.7.html
 .. _BPF and XDP Reference Guide: https://docs.cilium.io/en/latest/bpf/

Documentation/bpf/instruction-set.rst

···
+
+====================
+eBPF Instruction Set
+====================
+
+eBPF is designed to be JITed with one to one mapping, which can also open up
+the possibility for GCC/LLVM compilers to generate optimized eBPF code through
+an eBPF backend that performs almost as fast as natively compiled code.
+
+Some core changes of the eBPF format from classic BPF:
+
+- Number of registers increase from 2 to 10:
+
+  The old format had two registers A and X, and a hidden frame pointer. The
+  new layout extends this to be 10 internal registers and a read-only frame
+  pointer. Since 64-bit CPUs are passing arguments to functions via registers
+  the number of args from eBPF program to in-kernel function is restricted
+  to 5 and one register is used to accept return value from an in-kernel
+  function. Natively, x86_64 passes first 6 arguments in registers, aarch64/
+  sparcv9/mips64 have 7 - 8 registers for arguments; x86_64 has 6 callee saved
+  registers, and aarch64/sparcv9/mips64 have 11 or more callee saved registers.
+
+  Therefore, eBPF calling convention is defined as:
+
+    * R0	- return value from in-kernel function, and exit value for eBPF program
+    * R1 - R5	- arguments from eBPF program to in-kernel function
+    * R6 - R9	- callee saved registers that in-kernel function will preserve
+    * R10	- read-only frame pointer to access stack
+
+  Thus, all eBPF registers map one to one to HW registers on x86_64, aarch64,
+  etc, and eBPF calling convention maps directly to ABIs used by the kernel on
+  64-bit architectures.
+
+  On 32-bit architectures JIT may map programs that use only 32-bit arithmetic
+  and may let more complex programs to be interpreted.
+
+  R0 - R5 are scratch registers and eBPF program needs spill/fill them if
+  necessary across calls. Note that there is only one eBPF program (== one
+  eBPF main routine) and it cannot call other eBPF functions, it can only
+  call predefined in-kernel functions, though.
+
+- Register width increases from 32-bit to 64-bit:
+
+  Still, the semantics of the original 32-bit ALU operations are preserved
+  via 32-bit subregisters. All eBPF registers are 64-bit with 32-bit lower
+  subregisters that zero-extend into 64-bit if they are being written to.
+  That behavior maps directly to x86_64 and arm64 subregister definition, but
+  makes other JITs more difficult.
+
+  32-bit architectures run 64-bit eBPF programs via interpreter.
+  Their JITs may convert BPF programs that only use 32-bit subregisters into
+  native instruction set and let the rest being interpreted.
+
+  Operation is 64-bit, because on 64-bit architectures, pointers are also
+  64-bit wide, and we want to pass 64-bit values in/out of kernel functions,
+  so 32-bit eBPF registers would otherwise require to define register-pair
+  ABI, thus, there won't be able to use a direct eBPF register to HW register
+  mapping and JIT would need to do combine/split/move operations for every
+  register in and out of the function, which is complex, bug prone and slow.
+  Another reason is the use of atomic 64-bit counters.
+
+- Conditional jt/jf targets replaced with jt/fall-through:
+
+  While the original design has constructs such as ``if (cond) jump_true;
+  else jump_false;``, they are being replaced into alternative constructs like
+  ``if (cond) jump_true; /* else fall-through */``.
+
+- Introduces bpf_call insn and register passing convention for zero overhead
+  calls from/to other kernel functions:
+
+  Before an in-kernel function call, the eBPF program needs to
+  place function arguments into R1 to R5 registers to satisfy calling
+  convention, then the interpreter will take them from registers and pass
+  to in-kernel function. If R1 - R5 registers are mapped to CPU registers
+  that are used for argument passing on given architecture, the JIT compiler
+  doesn't need to emit extra moves. Function arguments will be in the correct
+  registers and BPF_CALL instruction will be JITed as single 'call' HW
+  instruction. This calling convention was picked to cover common call
+  situations without performance penalty.
+
+  After an in-kernel function call, R1 - R5 are reset to unreadable and R0 has
+  a return value of the function. Since R6 - R9 are callee saved, their state
+  is preserved across the call.
+
+  For example, consider three C functions::
+
+    u64 f1() { return (*_f2)(1); }
+    u64 f2(u64 a) { return f3(a + 1, a); }
+    u64 f3(u64 a, u64 b) { return a - b; }
+
+  GCC can compile f1, f3 into x86_64::
+
+    f1:
+	movl $1, %edi
+	movq _f2(%rip), %rax
+	jmp  *%rax
+    f3:
+	movq %rdi, %rax
+	subq %rsi, %rax
+	ret
+
+  Function f2 in eBPF may look like::
+
+    f2:
+	bpf_mov R2, R1
+	bpf_add R1, 1
+	bpf_call f3
+	bpf_exit
+
+  If f2 is JITed and the pointer stored to ``_f2``. The calls f1 -> f2 -> f3 and
+  returns will be seamless. Without JIT, __bpf_prog_run() interpreter needs to
+  be used to call into f2.
+
+  For practical reasons all eBPF programs have only one argument 'ctx' which is
+  already placed into R1 (e.g. on __bpf_prog_run() startup) and the programs
+  can call kernel functions with up to 5 arguments. Calls with 6 or more arguments
+  are currently not supported, but these restrictions can be lifted if necessary
+  in the future.
+
+  On 64-bit architectures all register map to HW registers one to one. For
+  example, x86_64 JIT compiler can map them as ...
+
+  ::
+
+    R0 - rax
+    R1 - rdi
+    R2 - rsi
+    R3 - rdx
+    R4 - rcx
+    R5 - r8
+    R6 - rbx
+    R7 - r13
+    R8 - r14
+    R9 - r15
+    R10 - rbp
+
+  ... since x86_64 ABI mandates rdi, rsi, rdx, rcx, r8, r9 for argument passing
+  and rbx, r12 - r15 are callee saved.
+
+  Then the following eBPF pseudo-program::
+
+    bpf_mov R6, R1 /* save ctx */
+    bpf_mov R2, 2
+    bpf_mov R3, 3
+    bpf_mov R4, 4
+    bpf_mov R5, 5
+    bpf_call foo
+    bpf_mov R7, R0 /* save foo() return value */
+    bpf_mov R1, R6 /* restore ctx for next call */
+    bpf_mov R2, 6
+    bpf_mov R3, 7
+    bpf_mov R4, 8
+    bpf_mov R5, 9
+    bpf_call bar
+    bpf_add R0, R7
+    bpf_exit
+
+  After JIT to x86_64 may look like::
+
+    push %rbp
+    mov %rsp,%rbp
+    sub $0x228,%rsp
+    mov %rbx,-0x228(%rbp)
+    mov %r13,-0x220(%rbp)
+    mov %rdi,%rbx
+    mov $0x2,%esi
+    mov $0x3,%edx
+    mov $0x4,%ecx
+    mov $0x5,%r8d
+    callq foo
+    mov %rax,%r13
+    mov %rbx,%rdi
+    mov $0x6,%esi
+    mov $0x7,%edx
+    mov $0x8,%ecx
+    mov $0x9,%r8d
+    callq bar
+    add %r13,%rax
+    mov -0x228(%rbp),%rbx
+    mov -0x220(%rbp),%r13
+    leaveq
+    retq
+
+  Which is in this example equivalent in C to::
+
+    u64 bpf_filter(u64 ctx)
+    {
+	return foo(ctx, 2, 3, 4, 5) + bar(ctx, 6, 7, 8, 9);
+    }
+
+  In-kernel functions foo() and bar() with prototype: u64 (*)(u64 arg1, u64
+  arg2, u64 arg3, u64 arg4, u64 arg5); will receive arguments in proper
+  registers and place their return value into ``%rax`` which is R0 in eBPF.
+  Prologue and epilogue are emitted by JIT and are implicit in the
+  interpreter. R0-R5 are scratch registers, so eBPF program needs to preserve
+  them across the calls as defined by calling convention.
+
+  For example the following program is invalid::
+
+    bpf_mov R1, 1
+    bpf_call foo
+    bpf_mov R0, R1
+    bpf_exit
+
+  After the call the registers R1-R5 contain junk values and cannot be read.
+  An in-kernel `eBPF verifier`_ is used to validate eBPF programs.
+
+Also in the new design, eBPF is limited to 4096 insns, which means that any
+program will terminate quickly and will only call a fixed number of kernel
+functions. Original BPF and eBPF are two operand instructions,
+which helps to do one-to-one mapping between eBPF insn and x86 insn during JIT.
+
+The input context pointer for invoking the interpreter function is generic,
+its content is defined by a specific use case. For seccomp register R1 points
+to seccomp_data, for converted BPF filters R1 points to a skb.
+
+A program, that is translated internally consists of the following elements::
+
+  op:16, jt:8, jf:8, k:32    ==>    op:8, dst_reg:4, src_reg:4, off:16, imm:32
+
+So far 87 eBPF instructions were implemented. 8-bit 'op' opcode field
+has room for new instructions. Some of them may use 16/24/32 byte encoding. New
+instructions must be multiple of 8 bytes to preserve backward compatibility.
+
+eBPF is a general purpose RISC instruction set. Not every register and
+every instruction are used during translation from original BPF to eBPF.
+For example, socket filters are not using ``exclusive add`` instruction, but
+tracing filters may do to maintain counters of events, for example. Register R9
+is not used by socket filters either, but more complex filters may be running
+out of registers and would have to resort to spill/fill to stack.
+
+eBPF can be used as a generic assembler for last step performance
+optimizations, socket filters and seccomp are using it as assembler. Tracing
+filters may use it as assembler to generate code from kernel. In kernel usage
+may not be bounded by security considerations, since generated eBPF code
+may be optimizing internal code path and not being exposed to the user space.
+Safety of eBPF can come from the `eBPF verifier`_. In such use cases as
+described, it may be used as safe instruction set.
+
+Just like the original BPF, eBPF runs within a controlled environment,
+is deterministic and the kernel can easily prove that. The safety of the program
+can be determined in two steps: first step does depth-first-search to disallow
+loops and other CFG validation; second step starts from the first insn and
+descends all possible paths. It simulates execution of every insn and observes
+the state change of registers and stack.
+
+eBPF opcode encoding
+====================
+
+eBPF is reusing most of the opcode encoding from classic to simplify conversion
+of classic BPF to eBPF. For arithmetic and jump instructions the 8-bit 'code'
+field is divided into three parts::
+
+  +----------------+--------+--------------------+
+  |   4 bits       |  1 bit |   3 bits           |
+  | operation code | source | instruction class  |
+  +----------------+--------+--------------------+
+  (MSB)                                      (LSB)
+
+Three LSB bits store instruction class which is one of:
+
+  ===================     ===============
+  Classic BPF classes     eBPF classes
+  ===================     ===============
+  BPF_LD    0x00          BPF_LD    0x00
+  BPF_LDX   0x01          BPF_LDX   0x01
+  BPF_ST    0x02          BPF_ST    0x02
+  BPF_STX   0x03          BPF_STX   0x03
+  BPF_ALU   0x04          BPF_ALU   0x04
+  BPF_JMP   0x05          BPF_JMP   0x05
+  BPF_RET   0x06          BPF_JMP32 0x06
+  BPF_MISC  0x07          BPF_ALU64 0x07
+  ===================     ===============
+
+When BPF_CLASS(code) == BPF_ALU or BPF_JMP, 4th bit encodes source operand ...
+
+    ::
+
+	BPF_K     0x00
+	BPF_X     0x08
+
+ * in classic BPF, this means::
+
+	BPF_SRC(code) == BPF_X - use register X as source operand
+	BPF_SRC(code) == BPF_K - use 32-bit immediate as source operand
+
+ * in eBPF, this means::
+
+	BPF_SRC(code) == BPF_X - use 'src_reg' register as source operand
+	BPF_SRC(code) == BPF_K - use 32-bit immediate as source operand
+
+... and four MSB bits store operation code.
+
+If BPF_CLASS(code) == BPF_ALU or BPF_ALU64 [ in eBPF ], BPF_OP(code) is one of::
+
+  BPF_ADD   0x00
+  BPF_SUB   0x10
+  BPF_MUL   0x20
+  BPF_DIV   0x30
+  BPF_OR    0x40
+  BPF_AND   0x50
+  BPF_LSH   0x60
+  BPF_RSH   0x70
+  BPF_NEG   0x80
+  BPF_MOD   0x90
+  BPF_XOR   0xa0
+  BPF_MOV   0xb0  /* eBPF only: mov reg to reg */
+  BPF_ARSH  0xc0  /* eBPF only: sign extending shift right */
+  BPF_END   0xd0  /* eBPF only: endianness conversion */
+
+If BPF_CLASS(code) == BPF_JMP or BPF_JMP32 [ in eBPF ], BPF_OP(code) is one of::
+
+  BPF_JA    0x00  /* BPF_JMP only */
+  BPF_JEQ   0x10
+  BPF_JGT   0x20
+  BPF_JGE   0x30
+  BPF_JSET  0x40
+  BPF_JNE   0x50  /* eBPF only: jump != */
+  BPF_JSGT  0x60  /* eBPF only: signed '>' */
+  BPF_JSGE  0x70  /* eBPF only: signed '>=' */
+  BPF_CALL  0x80  /* eBPF BPF_JMP only: function call */
+  BPF_EXIT  0x90  /* eBPF BPF_JMP only: function return */
+  BPF_JLT   0xa0  /* eBPF only: unsigned '<' */
+  BPF_JLE   0xb0  /* eBPF only: unsigned '<=' */
+  BPF_JSLT  0xc0  /* eBPF only: signed '<' */
+  BPF_JSLE  0xd0  /* eBPF only: signed '<=' */
+
+So BPF_ADD | BPF_X | BPF_ALU means 32-bit addition in both classic BPF
+and eBPF. There are only two registers in classic BPF, so it means A += X.
+In eBPF it means dst_reg = (u32) dst_reg + (u32) src_reg; similarly,
+BPF_XOR | BPF_K | BPF_ALU means A ^= imm32 in classic BPF and analogous
+src_reg = (u32) src_reg ^ (u32) imm32 in eBPF.
+
+Classic BPF is using BPF_MISC class to represent A = X and X = A moves.
+eBPF is using BPF_MOV | BPF_X | BPF_ALU code instead. Since there are no
+BPF_MISC operations in eBPF, the class 7 is used as BPF_ALU64 to mean
+exactly the same operations as BPF_ALU, but with 64-bit wide operands
+instead. So BPF_ADD | BPF_X | BPF_ALU64 means 64-bit addition, i.e.:
+dst_reg = dst_reg + src_reg
+
+Classic BPF wastes the whole BPF_RET class to represent a single ``ret``
+operation. Classic BPF_RET | BPF_K means copy imm32 into return register
+and perform function exit. eBPF is modeled to match CPU, so BPF_JMP | BPF_EXIT
+in eBPF means function exit only. The eBPF program needs to store return
+value into register R0 before doing a BPF_EXIT. Class 6 in eBPF is used as
+BPF_JMP32 to mean exactly the same operations as BPF_JMP, but with 32-bit wide
+operands for the comparisons instead.
+
+For load and store instructions the 8-bit 'code' field is divided as::
+
+  +--------+--------+-------------------+
+  | 3 bits | 2 bits |   3 bits          |
+  |  mode  |  size  | instruction class |
+  +--------+--------+-------------------+
+  (MSB)                             (LSB)
+
+Size modifier is one of ...
+
+::
+
+  BPF_W   0x00    /* word */
+  BPF_H   0x08    /* half word */
+  BPF_B   0x10    /* byte */
+  BPF_DW  0x18    /* eBPF only, double word */
+
+... which encodes size of load/store operation::
+
+ B  - 1 byte
+ H  - 2 byte
+ W  - 4 byte
+ DW - 8 byte (eBPF only)
+
+Mode modifier is one of::
+
+  BPF_IMM     0x00  /* used for 32-bit mov in classic BPF and 64-bit in eBPF */
+  BPF_ABS     0x20
+  BPF_IND     0x40
+  BPF_MEM     0x60
+  BPF_LEN     0x80  /* classic BPF only, reserved in eBPF */
+  BPF_MSH     0xa0  /* classic BPF only, reserved in eBPF */
+  BPF_ATOMIC  0xc0  /* eBPF only, atomic operations */
+
+eBPF has two non-generic instructions: (BPF_ABS | <size> | BPF_LD) and
+(BPF_IND | <size> | BPF_LD) which are used to access packet data.
+
+They had to be carried over from classic to have strong performance of
+socket filters running in eBPF interpreter. These instructions can only
+be used when interpreter context is a pointer to ``struct sk_buff`` and
+have seven implicit operands. Register R6 is an implicit input that must
+contain pointer to sk_buff. Register R0 is an implicit output which contains
+the data fetched from the packet. Registers R1-R5 are scratch registers
+and must not be used to store the data across BPF_ABS | BPF_LD or
+BPF_IND | BPF_LD instructions.
+
+These instructions have implicit program exit condition as well. When
+eBPF program is trying to access the data beyond the packet boundary,
+the interpreter will abort the execution of the program. JIT compilers
+therefore must preserve this property. src_reg and imm32 fields are
+explicit inputs to these instructions.
+
+For example::
+
+  BPF_IND | BPF_W | BPF_LD means:
+
+    R0 = ntohl(*(u32 *) (((struct sk_buff *) R6)->data + src_reg + imm32))
+    and R1 - R5 were scratched.
+
+Unlike classic BPF instruction set, eBPF has generic load/store operations::
+
+    BPF_MEM | <size> | BPF_STX:  *(size *) (dst_reg + off) = src_reg
+    BPF_MEM | <size> | BPF_ST:   *(size *) (dst_reg + off) = imm32
+    BPF_MEM | <size> | BPF_LDX:  dst_reg = *(size *) (src_reg + off)
+
+Where size is one of: BPF_B or BPF_H or BPF_W or BPF_DW.
+
+It also includes atomic operations, which use the immediate field for extra
+encoding::
+
+   .imm = BPF_ADD, .code = BPF_ATOMIC | BPF_W  | BPF_STX: lock xadd *(u32 *)(dst_reg + off16) += src_reg
+   .imm = BPF_ADD, .code = BPF_ATOMIC | BPF_DW | BPF_STX: lock xadd *(u64 *)(dst_reg + off16) += src_reg
+
+The basic atomic operations supported are::
+
+    BPF_ADD
+    BPF_AND
+    BPF_OR
+    BPF_XOR
+
+Each having equivalent semantics with the ``BPF_ADD`` example, that is: the
+memory location addresed by ``dst_reg + off`` is atomically modified, with
+``src_reg`` as the other operand. If the ``BPF_FETCH`` flag is set in the
+immediate, then these operations also overwrite ``src_reg`` with the
+value that was in memory before it was modified.
+
+The more special operations are::
+
+    BPF_XCHG
+
+This atomically exchanges ``src_reg`` with the value addressed by ``dst_reg +
+off``. ::
+
+    BPF_CMPXCHG
+
+This atomically compares the value addressed by ``dst_reg + off`` with
+``R0``. If they match it is replaced with ``src_reg``. In either case, the
+value that was there before is zero-extended and loaded back to ``R0``.
+
+Note that 1 and 2 byte atomic operations are not supported.
+
+Clang can generate atomic instructions by default when ``-mcpu=v3`` is
+enabled. If a lower version for ``-mcpu`` is set, the only atomic instruction
+Clang can generate is ``BPF_ADD`` *without* ``BPF_FETCH``. If you need to enable
+the atomics features, while keeping a lower ``-mcpu`` version, you can use
+``-Xclang -target-feature -Xclang +alu32``.
+
+You may encounter ``BPF_XADD`` - this is a legacy name for ``BPF_ATOMIC``,
+referring to the exclusive-add operation encoded when the immediate field is
+zero.
+
+eBPF has one 16-byte instruction: ``BPF_LD | BPF_DW | BPF_IMM`` which consists
+of two consecutive ``struct bpf_insn`` 8-byte blocks and interpreted as single
+instruction that loads 64-bit immediate value into a dst_reg.
+Classic BPF has similar instruction: ``BPF_LD | BPF_W | BPF_IMM`` which loads
+32-bit immediate value into a register.
+
+.. Links:
+.. _eBPF verifier: verifiers.rst

Documentation/bpf/libbpf/index.rst

···
 libbpf
 ======
 
-For API documentation see the `versioned API documentation site <https://libbpf.readthedocs.io/en/latest/api.html>`_.
-
 .. toctree::
    :maxdepth: 1
 
···
 
 This is documentation for libbpf, a userspace library for loading and
 interacting with bpf programs.
+
+For API documentation see the `versioned API documentation site <https://libbpf.readthedocs.io/en/latest/api.html>`_.
 
 All general BPF questions, including kernel functionality, libbpf APIs and
 their application, should be sent to [e-mail] mailing list.

Documentation/bpf/maps.rst

···
+
+=========
+eBPF maps
+=========
+
+'maps' is a generic storage of different types for sharing data between kernel
+and userspace.
+
+The maps are accessed from user space via BPF syscall, which has commands:
+
+- create a map with given type and attributes
+  ``map_fd = bpf(BPF_MAP_CREATE, union bpf_attr *attr, u32 size)``
+  using attr->map_type, attr->key_size, attr->value_size, attr->max_entries
+  returns process-local file descriptor or negative error
+
+- lookup key in a given map
+  ``err = bpf(BPF_MAP_LOOKUP_ELEM, union bpf_attr *attr, u32 size)``
+  using attr->map_fd, attr->key, attr->value
+  returns zero and stores found elem into value or negative error
+
+- create or update key/value pair in a given map
+  ``err = bpf(BPF_MAP_UPDATE_ELEM, union bpf_attr *attr, u32 size)``
+  using attr->map_fd, attr->key, attr->value
+  returns zero or negative error
+
+- find and delete element by key in a given map
+  ``err = bpf(BPF_MAP_DELETE_ELEM, union bpf_attr *attr, u32 size)``
+  using attr->map_fd, attr->key
+
+- to delete map: close(fd)
+  Exiting process will delete maps automatically
+
+userspace programs use this syscall to create/access maps that eBPF programs
+are concurrently updating.
+
+maps can have different types: hash, array, bloom filter, radix-tree, etc.
+
+The map is defined by:
+
+  - type
+  - max number of elements
+  - key size in bytes
+  - value size in bytes
+
+Map Types
+=========
+
+.. toctree::
+   :maxdepth: 1
+   :glob:
+
+   map_*

Documentation/bpf/other.rst

···
+=====
+Other
+=====
+
+.. toctree::
+   :maxdepth: 1
+
+   ringbuf
+   llvm_reloc

Documentation/bpf/programs.rst

···
+=============
+Program Types
+=============
+
+.. toctree::
+   :maxdepth: 1
+   :glob:
+
+   prog_*

Documentation/bpf/syscall_api.rst

···
+===========
+Syscall API
+===========
+
+The primary info for the bpf syscall is available in the `man-pages`_
+for `bpf(2)`_. For more information about the userspace API, see
+Documentation/userspace-api/ebpf/index.rst.
+
+.. Links:
+.. _man-pages: https://www.kernel.org/doc/man-pages/
+.. _bpf(2): https://man7.org/linux/man-pages/man2/bpf.2.html

Documentation/bpf/test_debug.rst

···
+=========================
+Testing and debugging BPF
+=========================
+
+.. toctree::
+   :maxdepth: 1
+
+   drgn
+   s390

Documentation/bpf/verifier.rst

···
+
+=============
+eBPF verifier
+=============
+
+The safety of the eBPF program is determined in two steps.
+
+First step does DAG check to disallow loops and other CFG validation.
+In particular it will detect programs that have unreachable instructions.
+(though classic BPF checker allows them)
+
+Second step starts from the first insn and descends all possible paths.
+It simulates execution of every insn and observes the state change of
+registers and stack.
+
+At the start of the program the register R1 contains a pointer to context
+and has type PTR_TO_CTX.
+If verifier sees an insn that does R2=R1, then R2 has now type
+PTR_TO_CTX as well and can be used on the right hand side of expression.
+If R1=PTR_TO_CTX and insn is R2=R1+R1, then R2=SCALAR_VALUE,
+since addition of two valid pointers makes invalid pointer.
+(In 'secure' mode verifier will reject any type of pointer arithmetic to make
+sure that kernel addresses don't leak to unprivileged users)
+
+If register was never written to, it's not readable::
+
+  bpf_mov R0 = R2
+  bpf_exit
+
+will be rejected, since R2 is unreadable at the start of the program.
+
+After kernel function call, R1-R5 are reset to unreadable and
+R0 has a return type of the function.
+
+Since R6-R9 are callee saved, their state is preserved across the call.
+
+::
+
+  bpf_mov R6 = 1
+  bpf_call foo
+  bpf_mov R0 = R6
+  bpf_exit
+
+is a correct program. If there was R1 instead of R6, it would have
+been rejected.
+
+load/store instructions are allowed only with registers of valid types, which
+are PTR_TO_CTX, PTR_TO_MAP, PTR_TO_STACK. They are bounds and alignment checked.
+For example::
+
+ bpf_mov R1 = 1
+ bpf_mov R2 = 2
+ bpf_xadd *(u32 *)(R1 + 3) += R2
+ bpf_exit
+
+will be rejected, since R1 doesn't have a valid pointer type at the time of
+execution of instruction bpf_xadd.
+
+At the start R1 type is PTR_TO_CTX (a pointer to generic ``struct bpf_context``)
+A callback is used to customize verifier to restrict eBPF program access to only
+certain fields within ctx structure with specified size and alignment.
+
+For example, the following insn::
+
+  bpf_ld R0 = *(u32 *)(R6 + 8)
+
+intends to load a word from address R6 + 8 and store it into R0
+If R6=PTR_TO_CTX, via is_valid_access() callback the verifier will know
+that offset 8 of size 4 bytes can be accessed for reading, otherwise
+the verifier will reject the program.
+If R6=PTR_TO_STACK, then access should be aligned and be within
+stack bounds, which are [-MAX_BPF_STACK, 0). In this example offset is 8,
+so it will fail verification, since it's out of bounds.
+
+The verifier will allow eBPF program to read data from stack only after
+it wrote into it.
+
+Classic BPF verifier does similar check with M[0-15] memory slots.
+For example::
+
+  bpf_ld R0 = *(u32 *)(R10 - 4)
+  bpf_exit
+
+is invalid program.
+Though R10 is correct read-only register and has type PTR_TO_STACK
+and R10 - 4 is within stack bounds, there were no stores into that location.
+
+Pointer register spill/fill is tracked as well, since four (R6-R9)
+callee saved registers may not be enough for some programs.
+
+Allowed function calls are customized with bpf_verifier_ops->get_func_proto()
+The eBPF verifier will check that registers match argument constraints.
+After the call register R0 will be set to return type of the function.
+
+Function calls is a main mechanism to extend functionality of eBPF programs.
+Socket filters may let programs to call one set of functions, whereas tracing
+filters may allow completely different set.
+
+If a function made accessible to eBPF program, it needs to be thought through
+from safety point of view. The verifier will guarantee that the function is
+called with valid arguments.
+
+seccomp vs socket filters have different security restrictions for classic BPF.
+Seccomp solves this by two stage verifier: classic BPF verifier is followed
+by seccomp verifier. In case of eBPF one configurable verifier is shared for
+all use cases.
+
+See details of eBPF verifier in kernel/bpf/verifier.c
+
+Register value tracking
+=======================
+
+In order to determine the safety of an eBPF program, the verifier must track
+the range of possible values in each register and also in each stack slot.
+This is done with ``struct bpf_reg_state``, defined in include/linux/
+bpf_verifier.h, which unifies tracking of scalar and pointer values.  Each
+register state has a type, which is either NOT_INIT (the register has not been
+written to), SCALAR_VALUE (some value which is not usable as a pointer), or a
+pointer type.  The types of pointers describe their base, as follows:
+
+
+    PTR_TO_CTX
+			Pointer to bpf_context.
+    CONST_PTR_TO_MAP
+			Pointer to struct bpf_map.  "Const" because arithmetic
+			on these pointers is forbidden.
+    PTR_TO_MAP_VALUE
+			Pointer to the value stored in a map element.
+    PTR_TO_MAP_VALUE_OR_NULL
+			Either a pointer to a map value, or NULL; map accesses
+			(see maps.rst) return this type, which becomes a
+			PTR_TO_MAP_VALUE when checked != NULL. Arithmetic on
+			these pointers is forbidden.
+    PTR_TO_STACK
+			Frame pointer.
+    PTR_TO_PACKET
+			skb->data.
+    PTR_TO_PACKET_END
+			skb->data + headlen; arithmetic forbidden.
+    PTR_TO_SOCKET
+			Pointer to struct bpf_sock_ops, implicitly refcounted.
+    PTR_TO_SOCKET_OR_NULL
+			Either a pointer to a socket, or NULL; socket lookup
+			returns this type, which becomes a PTR_TO_SOCKET when
+			checked != NULL. PTR_TO_SOCKET is reference-counted,
+			so programs must release the reference through the
+			socket release function before the end of the program.
+			Arithmetic on these pointers is forbidden.
+
+However, a pointer may be offset from this base (as a result of pointer
+arithmetic), and this is tracked in two parts: the 'fixed offset' and 'variable
+offset'.  The former is used when an exactly-known value (e.g. an immediate
+operand) is added to a pointer, while the latter is used for values which are
+not exactly known.  The variable offset is also used in SCALAR_VALUEs, to track
+the range of possible values in the register.
+
+The verifier's knowledge about the variable offset consists of:
+
+* minimum and maximum values as unsigned
+* minimum and maximum values as signed
+
+* knowledge of the values of individual bits, in the form of a 'tnum': a u64
+  'mask' and a u64 'value'.  1s in the mask represent bits whose value is unknown;
+  1s in the value represent bits known to be 1.  Bits known to be 0 have 0 in both
+  mask and value; no bit should ever be 1 in both.  For example, if a byte is read
+  into a register from memory, the register's top 56 bits are known zero, while
+  the low 8 are unknown - which is represented as the tnum (0x0; 0xff).  If we
+  then OR this with 0x40, we get (0x40; 0xbf), then if we add 1 we get (0x0;
+  0x1ff), because of potential carries.
+
+Besides arithmetic, the register state can also be updated by conditional
+branches.  For instance, if a SCALAR_VALUE is compared > 8, in the 'true' branch
+it will have a umin_value (unsigned minimum value) of 9, whereas in the 'false'
+branch it will have a umax_value of 8.  A signed compare (with BPF_JSGT or
+BPF_JSGE) would instead update the signed minimum/maximum values.  Information
+from the signed and unsigned bounds can be combined; for instance if a value is
+first tested < 8 and then tested s> 4, the verifier will conclude that the value
+is also > 4 and s< 8, since the bounds prevent crossing the sign boundary.
+
+PTR_TO_PACKETs with a variable offset part have an 'id', which is common to all
+pointers sharing that same variable offset.  This is important for packet range
+checks: after adding a variable to a packet pointer register A, if you then copy
+it to another register B and then add a constant 4 to A, both registers will
+share the same 'id' but the A will have a fixed offset of +4.  Then if A is
+bounds-checked and found to be less than a PTR_TO_PACKET_END, the register B is
+now known to have a safe range of at least 4 bytes.  See 'Direct packet access',
+below, for more on PTR_TO_PACKET ranges.
+
+The 'id' field is also used on PTR_TO_MAP_VALUE_OR_NULL, common to all copies of
+the pointer returned from a map lookup.  This means that when one copy is
+checked and found to be non-NULL, all copies can become PTR_TO_MAP_VALUEs.
+As well as range-checking, the tracked information is also used for enforcing
+alignment of pointer accesses.  For instance, on most systems the packet pointer
+is 2 bytes after a 4-byte alignment.  If a program adds 14 bytes to that to jump
+over the Ethernet header, then reads IHL and addes (IHL * 4), the resulting
+pointer will have a variable offset known to be 4n+2 for some n, so adding the 2
+bytes (NET_IP_ALIGN) gives a 4-byte alignment and so word-sized accesses through
+that pointer are safe.
+The 'id' field is also used on PTR_TO_SOCKET and PTR_TO_SOCKET_OR_NULL, common
+to all copies of the pointer returned from a socket lookup. This has similar
+behaviour to the handling for PTR_TO_MAP_VALUE_OR_NULL->PTR_TO_MAP_VALUE, but
+it also handles reference tracking for the pointer. PTR_TO_SOCKET implicitly
+represents a reference to the corresponding ``struct sock``. To ensure that the
+reference is not leaked, it is imperative to NULL-check the reference and in
+the non-NULL case, and pass the valid reference to the socket release function.
+
+Direct packet access
+====================
+
+In cls_bpf and act_bpf programs the verifier allows direct access to the packet
+data via skb->data and skb->data_end pointers.
+Ex::
+
+    1:  r4 = *(u32 *)(r1 +80)  /* load skb->data_end */
+    2:  r3 = *(u32 *)(r1 +76)  /* load skb->data */
+    3:  r5 = r3
+    4:  r5 += 14
+    5:  if r5 > r4 goto pc+16
+    R1=ctx R3=pkt(id=0,off=0,r=14) R4=pkt_end R5=pkt(id=0,off=14,r=14) R10=fp
+    6:  r0 = *(u16 *)(r3 +12) /* access 12 and 13 bytes of the packet */
+
+this 2byte load from the packet is safe to do, since the program author
+did check ``if (skb->data + 14 > skb->data_end) goto err`` at insn #5 which
+means that in the fall-through case the register R3 (which points to skb->data)
+has at least 14 directly accessible bytes. The verifier marks it
+as R3=pkt(id=0,off=0,r=14).
+id=0 means that no additional variables were added to the register.
+off=0 means that no additional constants were added.
+r=14 is the range of safe access which means that bytes [R3, R3 + 14) are ok.
+Note that R5 is marked as R5=pkt(id=0,off=14,r=14). It also points
+to the packet data, but constant 14 was added to the register, so
+it now points to ``skb->data + 14`` and accessible range is [R5, R5 + 14 - 14)
+which is zero bytes.
+
+More complex packet access may look like::
+
+
+    R0=inv1 R1=ctx R3=pkt(id=0,off=0,r=14) R4=pkt_end R5=pkt(id=0,off=14,r=14) R10=fp
+    6:  r0 = *(u8 *)(r3 +7) /* load 7th byte from the packet */
+    7:  r4 = *(u8 *)(r3 +12)
+    8:  r4 *= 14
+    9:  r3 = *(u32 *)(r1 +76) /* load skb->data */
+    10:  r3 += r4
+    11:  r2 = r1
+    12:  r2 <<= 48
+    13:  r2 >>= 48
+    14:  r3 += r2
+    15:  r2 = r3
+    16:  r2 += 8
+    17:  r1 = *(u32 *)(r1 +80) /* load skb->data_end */
+    18:  if r2 > r1 goto pc+2
+    R0=inv(id=0,umax_value=255,var_off=(0x0; 0xff)) R1=pkt_end R2=pkt(id=2,off=8,r=8) R3=pkt(id=2,off=0,r=8) R4=inv(id=0,umax_value=3570,var_off=(0x0; 0xfffe)) R5=pkt(id=0,off=14,r=14) R10=fp
+    19:  r1 = *(u8 *)(r3 +4)
+
+The state of the register R3 is R3=pkt(id=2,off=0,r=8)
+id=2 means that two ``r3 += rX`` instructions were seen, so r3 points to some
+offset within a packet and since the program author did
+``if (r3 + 8 > r1) goto err`` at insn #18, the safe range is [R3, R3 + 8).
+The verifier only allows 'add'/'sub' operations on packet registers. Any other
+operation will set the register state to 'SCALAR_VALUE' and it won't be
+available for direct packet access.
+
+Operation ``r3 += rX`` may overflow and become less than original skb->data,
+therefore the verifier has to prevent that.  So when it sees ``r3 += rX``
+instruction and rX is more than 16-bit value, any subsequent bounds-check of r3
+against skb->data_end will not give us 'range' information, so attempts to read
+through the pointer will give "invalid access to packet" error.
+
+Ex. after insn ``r4 = *(u8 *)(r3 +12)`` (insn #7 above) the state of r4 is
+R4=inv(id=0,umax_value=255,var_off=(0x0; 0xff)) which means that upper 56 bits
+of the register are guaranteed to be zero, and nothing is known about the lower
+8 bits. After insn ``r4 *= 14`` the state becomes
+R4=inv(id=0,umax_value=3570,var_off=(0x0; 0xfffe)), since multiplying an 8-bit
+value by constant 14 will keep upper 52 bits as zero, also the least significant
+bit will be zero as 14 is even.  Similarly ``r2 >>= 48`` will make
+R2=inv(id=0,umax_value=65535,var_off=(0x0; 0xffff)), since the shift is not sign
+extending.  This logic is implemented in adjust_reg_min_max_vals() function,
+which calls adjust_ptr_min_max_vals() for adding pointer to scalar (or vice
+versa) and adjust_scalar_min_max_vals() for operations on two scalars.
+
+The end result is that bpf program author can access packet directly
+using normal C code as::
+
+  void *data = (void *)(long)skb->data;
+  void *data_end = (void *)(long)skb->data_end;
+  struct eth_hdr *eth = data;
+  struct iphdr *iph = data + sizeof(*eth);
+  struct udphdr *udp = data + sizeof(*eth) + sizeof(*iph);
+
+  if (data + sizeof(*eth) + sizeof(*iph) + sizeof(*udp) > data_end)
+	  return 0;
+  if (eth->h_proto != htons(ETH_P_IP))
+	  return 0;
+  if (iph->protocol != IPPROTO_UDP || iph->ihl != 5)
+	  return 0;
+  if (udp->dest == 53 || udp->source == 9)
+	  ...;
+
+which makes such programs easier to write comparing to LD_ABS insn
+and significantly faster.
+
+Pruning
+=======
+
+The verifier does not actually walk all possible paths through the program.  For
+each new branch to analyse, the verifier looks at all the states it's previously
+been in when at this instruction.  If any of them contain the current state as a
+subset, the branch is 'pruned' - that is, the fact that the previous state was
+accepted implies the current state would be as well.  For instance, if in the
+previous state, r1 held a packet-pointer, and in the current state, r1 holds a
+packet-pointer with a range as long or longer and at least as strict an
+alignment, then r1 is safe.  Similarly, if r2 was NOT_INIT before then it can't
+have been used by any path from that point, so any value in r2 (including
+another NOT_INIT) is safe.  The implementation is in the function regsafe().
+Pruning considers not only the registers but also the stack (and any spilled
+registers it may hold).  They must all be safe for the branch to be pruned.
+This is implemented in states_equal().
+
+Understanding eBPF verifier messages
+====================================
+
+The following are few examples of invalid eBPF programs and verifier error
+messages as seen in the log:
+
+Program with unreachable instructions::
+
+  static struct bpf_insn prog[] = {
+  BPF_EXIT_INSN(),
+  BPF_EXIT_INSN(),
+  };
+
+Error:
+
+  unreachable insn 1
+
+Program that reads uninitialized register::
+
+  BPF_MOV64_REG(BPF_REG_0, BPF_REG_2),
+  BPF_EXIT_INSN(),
+
+Error::
+
+  0: (bf) r0 = r2
+  R2 !read_ok
+
+Program that doesn't initialize R0 before exiting::
+
+  BPF_MOV64_REG(BPF_REG_2, BPF_REG_1),
+  BPF_EXIT_INSN(),
+
+Error::
+
+  0: (bf) r2 = r1
+  1: (95) exit
+  R0 !read_ok
+
+Program that accesses stack out of bounds::
+
+    BPF_ST_MEM(BPF_DW, BPF_REG_10, 8, 0),
+    BPF_EXIT_INSN(),
+
+Error::
+
+    0: (7a) *(u64 *)(r10 +8) = 0
+    invalid stack off=8 size=8
+
+Program that doesn't initialize stack before passing its address into function::
+
+  BPF_MOV64_REG(BPF_REG_2, BPF_REG_10),
+  BPF_ALU64_IMM(BPF_ADD, BPF_REG_2, -8),
+  BPF_LD_MAP_FD(BPF_REG_1, 0),
+  BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0, BPF_FUNC_map_lookup_elem),
+  BPF_EXIT_INSN(),
+
+Error::
+
+  0: (bf) r2 = r10
+  1: (07) r2 += -8
+  2: (b7) r1 = 0x0
+  3: (85) call 1
+  invalid indirect read from stack off -8+0 size 8
+
+Program that uses invalid map_fd=0 while calling to map_lookup_elem() function::
+
+  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_LD_MAP_FD(BPF_REG_1, 0),
+  BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0, BPF_FUNC_map_lookup_elem),
+  BPF_EXIT_INSN(),
+
+Error::
+
+  0: (7a) *(u64 *)(r10 -8) = 0
+  1: (bf) r2 = r10
+  2: (07) r2 += -8
+  3: (b7) r1 = 0x0
+  4: (85) call 1
+  fd 0 is not pointing to valid bpf_map
+
+Program that doesn't check return value of map_lookup_elem() before accessing
+map element::
+
+  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_LD_MAP_FD(BPF_REG_1, 0),
+  BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0, BPF_FUNC_map_lookup_elem),
+  BPF_ST_MEM(BPF_DW, BPF_REG_0, 0, 0),
+  BPF_EXIT_INSN(),
+
+Error::
+
+  0: (7a) *(u64 *)(r10 -8) = 0
+  1: (bf) r2 = r10
+  2: (07) r2 += -8
+  3: (b7) r1 = 0x0
+  4: (85) call 1
+  5: (7a) *(u64 *)(r0 +0) = 0
+  R0 invalid mem access 'map_value_or_null'
+
+Program that correctly checks map_lookup_elem() returned value for NULL, but
+accesses the memory with incorrect alignment::
+
+  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_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, 1),
+  BPF_ST_MEM(BPF_DW, BPF_REG_0, 4, 0),
+  BPF_EXIT_INSN(),
+
+Error::
+
+  0: (7a) *(u64 *)(r10 -8) = 0
+  1: (bf) r2 = r10
+  2: (07) r2 += -8
+  3: (b7) r1 = 1
+  4: (85) call 1
+  5: (15) if r0 == 0x0 goto pc+1
+   R0=map_ptr R10=fp
+  6: (7a) *(u64 *)(r0 +4) = 0
+  misaligned access off 4 size 8
+
+Program that correctly checks map_lookup_elem() returned value for NULL and
+accesses memory with correct alignment in one side of 'if' branch, but fails
+to do so in the other side of 'if' branch::
+
+  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_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, 2),
+  BPF_ST_MEM(BPF_DW, BPF_REG_0, 0, 0),
+  BPF_EXIT_INSN(),
+  BPF_ST_MEM(BPF_DW, BPF_REG_0, 0, 1),
+  BPF_EXIT_INSN(),
+
+Error::
+
+  0: (7a) *(u64 *)(r10 -8) = 0
+  1: (bf) r2 = r10
+  2: (07) r2 += -8
+  3: (b7) r1 = 1
+  4: (85) call 1
+  5: (15) if r0 == 0x0 goto pc+2
+   R0=map_ptr R10=fp
+  6: (7a) *(u64 *)(r0 +0) = 0
+  7: (95) exit
+
+  from 5 to 8: R0=imm0 R10=fp
+  8: (7a) *(u64 *)(r0 +0) = 1
+  R0 invalid mem access 'imm'
+
+Program that performs a socket lookup then sets the pointer to NULL without
+checking it::
+
+  BPF_MOV64_IMM(BPF_REG_2, 0),
+  BPF_STX_MEM(BPF_W, BPF_REG_10, BPF_REG_2, -8),
+  BPF_MOV64_REG(BPF_REG_2, BPF_REG_10),
+  BPF_ALU64_IMM(BPF_ADD, BPF_REG_2, -8),
+  BPF_MOV64_IMM(BPF_REG_3, 4),
+  BPF_MOV64_IMM(BPF_REG_4, 0),
+  BPF_MOV64_IMM(BPF_REG_5, 0),
+  BPF_EMIT_CALL(BPF_FUNC_sk_lookup_tcp),
+  BPF_MOV64_IMM(BPF_REG_0, 0),
+  BPF_EXIT_INSN(),
+
+Error::
+
+  0: (b7) r2 = 0
+  1: (63) *(u32 *)(r10 -8) = r2
+  2: (bf) r2 = r10
+  3: (07) r2 += -8
+  4: (b7) r3 = 4
+  5: (b7) r4 = 0
+  6: (b7) r5 = 0
+  7: (85) call bpf_sk_lookup_tcp#65
+  8: (b7) r0 = 0
+  9: (95) exit
+  Unreleased reference id=1, alloc_insn=7
+
+Program that performs a socket lookup but does not NULL-check the returned
+value::
+
+  BPF_MOV64_IMM(BPF_REG_2, 0),
+  BPF_STX_MEM(BPF_W, BPF_REG_10, BPF_REG_2, -8),
+  BPF_MOV64_REG(BPF_REG_2, BPF_REG_10),
+  BPF_ALU64_IMM(BPF_ADD, BPF_REG_2, -8),
+  BPF_MOV64_IMM(BPF_REG_3, 4),
+  BPF_MOV64_IMM(BPF_REG_4, 0),
+  BPF_MOV64_IMM(BPF_REG_5, 0),
+  BPF_EMIT_CALL(BPF_FUNC_sk_lookup_tcp),
+  BPF_EXIT_INSN(),
+
+Error::
+
+  0: (b7) r2 = 0
+  1: (63) *(u32 *)(r10 -8) = r2
+  2: (bf) r2 = r10
+  3: (07) r2 += -8
+  4: (b7) r3 = 4
+  5: (b7) r4 = 0
+  6: (b7) r5 = 0
+  7: (85) call bpf_sk_lookup_tcp#65
+  8: (95) exit
+  Unreleased reference id=1, alloc_insn=7

Documentation/networking/filter.rst

···
 Linux Socket Filtering aka Berkeley Packet Filter (BPF)
 =======================================================
 
+Notice
+------
+
+This file used to document the eBPF format and mechanisms even when not
+related to socket filtering.  The ../bpf/index.rst has more details
+on eBPF.
+
 Introduction
 ------------
 
···
 paragraphs is being used. However, the instruction set format is modelled
 closer to the underlying architecture to mimic native instruction sets, so
 that a better performance can be achieved (more details later). This new
-ISA is called 'eBPF' or 'internal BPF' interchangeably. (Note: eBPF which
+ISA is called eBPF.  See the ../bpf/index.rst for details.  (Note: eBPF which
 originates from [e]xtended BPF is not the same as BPF extensions! While
 eBPF is an ISA, BPF extensions date back to classic BPF's 'overloading'
 of BPF_LD | BPF_{B,H,W} | BPF_ABS instruction.)
-
-It is designed to be JITed with one to one mapping, which can also open up
-the possibility for GCC/LLVM compilers to generate optimized eBPF code through
-an eBPF backend that performs almost as fast as natively compiled code.
 
 The new instruction set was originally designed with the possible goal in
 mind to write programs in "restricted C" and compile into eBPF with a optional
···
 sparc64, arm32, riscv64, riscv32 perform JIT compilation from eBPF
 instruction set.
 
-Some core changes of the new internal format:
-
-- Number of registers increase from 2 to 10:
-
-  The old format had two registers A and X, and a hidden frame pointer. The
-  new layout extends this to be 10 internal registers and a read-only frame
-  pointer. Since 64-bit CPUs are passing arguments to functions via registers
-  the number of args from eBPF program to in-kernel function is restricted
-  to 5 and one register is used to accept return value from an in-kernel
-  function. Natively, x86_64 passes first 6 arguments in registers, aarch64/
-  sparcv9/mips64 have 7 - 8 registers for arguments; x86_64 has 6 callee saved
-  registers, and aarch64/sparcv9/mips64 have 11 or more callee saved registers.
-
-  Therefore, eBPF calling convention is defined as:
-
-    * R0	- return value from in-kernel function, and exit value for eBPF program
-    * R1 - R5	- arguments from eBPF program to in-kernel function
-    * R6 - R9	- callee saved registers that in-kernel function will preserve
-    * R10	- read-only frame pointer to access stack
-
-  Thus, all eBPF registers map one to one to HW registers on x86_64, aarch64,
-  etc, and eBPF calling convention maps directly to ABIs used by the kernel on
-  64-bit architectures.
-
-  On 32-bit architectures JIT may map programs that use only 32-bit arithmetic
-  and may let more complex programs to be interpreted.
-
-  R0 - R5 are scratch registers and eBPF program needs spill/fill them if
-  necessary across calls. Note that there is only one eBPF program (== one
-  eBPF main routine) and it cannot call other eBPF functions, it can only
-  call predefined in-kernel functions, though.
-
-- Register width increases from 32-bit to 64-bit:
-
-  Still, the semantics of the original 32-bit ALU operations are preserved
-  via 32-bit subregisters. All eBPF registers are 64-bit with 32-bit lower
-  subregisters that zero-extend into 64-bit if they are being written to.
-  That behavior maps directly to x86_64 and arm64 subregister definition, but
-  makes other JITs more difficult.
-
-  32-bit architectures run 64-bit internal BPF programs via interpreter.
-  Their JITs may convert BPF programs that only use 32-bit subregisters into
-  native instruction set and let the rest being interpreted.
-
-  Operation is 64-bit, because on 64-bit architectures, pointers are also
-  64-bit wide, and we want to pass 64-bit values in/out of kernel functions,
-  so 32-bit eBPF registers would otherwise require to define register-pair
-  ABI, thus, there won't be able to use a direct eBPF register to HW register
-  mapping and JIT would need to do combine/split/move operations for every
-  register in and out of the function, which is complex, bug prone and slow.
-  Another reason is the use of atomic 64-bit counters.
-
-- Conditional jt/jf targets replaced with jt/fall-through:
-
-  While the original design has constructs such as ``if (cond) jump_true;
-  else jump_false;``, they are being replaced into alternative constructs like
-  ``if (cond) jump_true; /* else fall-through */``.
-
-- Introduces bpf_call insn and register passing convention for zero overhead
-  calls from/to other kernel functions:
-
-  Before an in-kernel function call, the internal BPF program needs to
-  place function arguments into R1 to R5 registers to satisfy calling
-  convention, then the interpreter will take them from registers and pass
-  to in-kernel function. If R1 - R5 registers are mapped to CPU registers
-  that are used for argument passing on given architecture, the JIT compiler
-  doesn't need to emit extra moves. Function arguments will be in the correct
-  registers and BPF_CALL instruction will be JITed as single 'call' HW
-  instruction. This calling convention was picked to cover common call
-  situations without performance penalty.
-
-  After an in-kernel function call, R1 - R5 are reset to unreadable and R0 has
-  a return value of the function. Since R6 - R9 are callee saved, their state
-  is preserved across the call.
-
-  For example, consider three C functions::
-
-    u64 f1() { return (*_f2)(1); }
-    u64 f2(u64 a) { return f3(a + 1, a); }
-    u64 f3(u64 a, u64 b) { return a - b; }
-
-  GCC can compile f1, f3 into x86_64::
-
-    f1:
-	movl $1, %edi
-	movq _f2(%rip), %rax
-	jmp  *%rax
-    f3:
-	movq %rdi, %rax
-	subq %rsi, %rax
-	ret
-
-  Function f2 in eBPF may look like::
-
-    f2:
-	bpf_mov R2, R1
-	bpf_add R1, 1
-	bpf_call f3
-	bpf_exit
-
-  If f2 is JITed and the pointer stored to ``_f2``. The calls f1 -> f2 -> f3 and
-  returns will be seamless. Without JIT, __bpf_prog_run() interpreter needs to
-  be used to call into f2.
-
-  For practical reasons all eBPF programs have only one argument 'ctx' which is
-  already placed into R1 (e.g. on __bpf_prog_run() startup) and the programs
-  can call kernel functions with up to 5 arguments. Calls with 6 or more arguments
-  are currently not supported, but these restrictions can be lifted if necessary
-  in the future.
-
-  On 64-bit architectures all register map to HW registers one to one. For
-  example, x86_64 JIT compiler can map them as ...
-
-  ::
-
-    R0 - rax
-    R1 - rdi
-    R2 - rsi
-    R3 - rdx
-    R4 - rcx
-    R5 - r8
-    R6 - rbx
-    R7 - r13
-    R8 - r14
-    R9 - r15
-    R10 - rbp
-
-  ... since x86_64 ABI mandates rdi, rsi, rdx, rcx, r8, r9 for argument passing
-  and rbx, r12 - r15 are callee saved.
-
-  Then the following internal BPF pseudo-program::
-
-    bpf_mov R6, R1 /* save ctx */
-    bpf_mov R2, 2
-    bpf_mov R3, 3
-    bpf_mov R4, 4
-    bpf_mov R5, 5
-    bpf_call foo
-    bpf_mov R7, R0 /* save foo() return value */
-    bpf_mov R1, R6 /* restore ctx for next call */
-    bpf_mov R2, 6
-    bpf_mov R3, 7
-    bpf_mov R4, 8
-    bpf_mov R5, 9
-    bpf_call bar
-    bpf_add R0, R7
-    bpf_exit
-
-  After JIT to x86_64 may look like::
-
-    push %rbp
-    mov %rsp,%rbp
-    sub $0x228,%rsp
-    mov %rbx,-0x228(%rbp)
-    mov %r13,-0x220(%rbp)
-    mov %rdi,%rbx
-    mov $0x2,%esi
-    mov $0x3,%edx
-    mov $0x4,%ecx
-    mov $0x5,%r8d
-    callq foo
-    mov %rax,%r13
-    mov %rbx,%rdi
-    mov $0x6,%esi
-    mov $0x7,%edx
-    mov $0x8,%ecx
-    mov $0x9,%r8d
-    callq bar
-    add %r13,%rax
-    mov -0x228(%rbp),%rbx
-    mov -0x220(%rbp),%r13
-    leaveq
-    retq
-
-  Which is in this example equivalent in C to::
-
-    u64 bpf_filter(u64 ctx)
-    {
-	return foo(ctx, 2, 3, 4, 5) + bar(ctx, 6, 7, 8, 9);
-    }
-
-  In-kernel functions foo() and bar() with prototype: u64 (*)(u64 arg1, u64
-  arg2, u64 arg3, u64 arg4, u64 arg5); will receive arguments in proper
-  registers and place their return value into ``%rax`` which is R0 in eBPF.
-  Prologue and epilogue are emitted by JIT and are implicit in the
-  interpreter. R0-R5 are scratch registers, so eBPF program needs to preserve
-  them across the calls as defined by calling convention.
-
-  For example the following program is invalid::
-
-    bpf_mov R1, 1
-    bpf_call foo
-    bpf_mov R0, R1
-    bpf_exit
-
-  After the call the registers R1-R5 contain junk values and cannot be read.
-  An in-kernel eBPF verifier is used to validate internal BPF programs.
-
-Also in the new design, eBPF is limited to 4096 insns, which means that any
-program will terminate quickly and will only call a fixed number of kernel
-functions. Original BPF and the new format are two operand instructions,
-which helps to do one-to-one mapping between eBPF insn and x86 insn during JIT.
-
-The input context pointer for invoking the interpreter function is generic,
-its content is defined by a specific use case. For seccomp register R1 points
-to seccomp_data, for converted BPF filters R1 points to a skb.
-
-A program, that is translated internally consists of the following elements::
-
-  op:16, jt:8, jf:8, k:32    ==>    op:8, dst_reg:4, src_reg:4, off:16, imm:32
-
-So far 87 internal BPF instructions were implemented. 8-bit 'op' opcode field
-has room for new instructions. Some of them may use 16/24/32 byte encoding. New
-instructions must be multiple of 8 bytes to preserve backward compatibility.
-
-Internal BPF is a general purpose RISC instruction set. Not every register and
-every instruction are used during translation from original BPF to new format.
-For example, socket filters are not using ``exclusive add`` instruction, but
-tracing filters may do to maintain counters of events, for example. Register R9
-is not used by socket filters either, but more complex filters may be running
-out of registers and would have to resort to spill/fill to stack.
-
-Internal BPF can be used as a generic assembler for last step performance
-optimizations, socket filters and seccomp are using it as assembler. Tracing
-filters may use it as assembler to generate code from kernel. In kernel usage
-may not be bounded by security considerations, since generated internal BPF code
-may be optimizing internal code path and not being exposed to the user space.
-Safety of internal BPF can come from a verifier (TBD). In such use cases as
-described, it may be used as safe instruction set.
-
-Just like the original BPF, the new format runs within a controlled environment,
-is deterministic and the kernel can easily prove that. The safety of the program
-can be determined in two steps: first step does depth-first-search to disallow
-loops and other CFG validation; second step starts from the first insn and
-descends all possible paths. It simulates execution of every insn and observes
-the state change of registers and stack.
-
-eBPF opcode encoding
---------------------
-
-eBPF is reusing most of the opcode encoding from classic to simplify conversion
-of classic BPF to eBPF. For arithmetic and jump instructions the 8-bit 'code'
-field is divided into three parts::
-
-  +----------------+--------+--------------------+
-  |   4 bits       |  1 bit |   3 bits           |
-  | operation code | source | instruction class  |
-  +----------------+--------+--------------------+
-  (MSB)                                      (LSB)
-
-Three LSB bits store instruction class which is one of:
-
-  ===================     ===============
-  Classic BPF classes     eBPF classes
-  ===================     ===============
-  BPF_LD    0x00          BPF_LD    0x00
-  BPF_LDX   0x01          BPF_LDX   0x01
-  BPF_ST    0x02          BPF_ST    0x02
-  BPF_STX   0x03          BPF_STX   0x03
-  BPF_ALU   0x04          BPF_ALU   0x04
-  BPF_JMP   0x05          BPF_JMP   0x05
-  BPF_RET   0x06          BPF_JMP32 0x06
-  BPF_MISC  0x07          BPF_ALU64 0x07
-  ===================     ===============
-
-When BPF_CLASS(code) == BPF_ALU or BPF_JMP, 4th bit encodes source operand ...
-
-    ::
-
-	BPF_K     0x00
-	BPF_X     0x08
-
- * in classic BPF, this means::
-
-	BPF_SRC(code) == BPF_X - use register X as source operand
-	BPF_SRC(code) == BPF_K - use 32-bit immediate as source operand
-
- * in eBPF, this means::
-
-	BPF_SRC(code) == BPF_X - use 'src_reg' register as source operand
-	BPF_SRC(code) == BPF_K - use 32-bit immediate as source operand
-
-... and four MSB bits store operation code.
-
-If BPF_CLASS(code) == BPF_ALU or BPF_ALU64 [ in eBPF ], BPF_OP(code) is one of::
-
-  BPF_ADD   0x00
-  BPF_SUB   0x10
-  BPF_MUL   0x20
-  BPF_DIV   0x30
-  BPF_OR    0x40
-  BPF_AND   0x50
-  BPF_LSH   0x60
-  BPF_RSH   0x70
-  BPF_NEG   0x80
-  BPF_MOD   0x90
-  BPF_XOR   0xa0
-  BPF_MOV   0xb0  /* eBPF only: mov reg to reg */
-  BPF_ARSH  0xc0  /* eBPF only: sign extending shift right */
-  BPF_END   0xd0  /* eBPF only: endianness conversion */
-
-If BPF_CLASS(code) == BPF_JMP or BPF_JMP32 [ in eBPF ], BPF_OP(code) is one of::
-
-  BPF_JA    0x00  /* BPF_JMP only */
-  BPF_JEQ   0x10
-  BPF_JGT   0x20
-  BPF_JGE   0x30
-  BPF_JSET  0x40
-  BPF_JNE   0x50  /* eBPF only: jump != */
-  BPF_JSGT  0x60  /* eBPF only: signed '>' */
-  BPF_JSGE  0x70  /* eBPF only: signed '>=' */
-  BPF_CALL  0x80  /* eBPF BPF_JMP only: function call */
-  BPF_EXIT  0x90  /* eBPF BPF_JMP only: function return */
-  BPF_JLT   0xa0  /* eBPF only: unsigned '<' */
-  BPF_JLE   0xb0  /* eBPF only: unsigned '<=' */
-  BPF_JSLT  0xc0  /* eBPF only: signed '<' */
-  BPF_JSLE  0xd0  /* eBPF only: signed '<=' */
-
-So BPF_ADD | BPF_X | BPF_ALU means 32-bit addition in both classic BPF
-and eBPF. There are only two registers in classic BPF, so it means A += X.
-In eBPF it means dst_reg = (u32) dst_reg + (u32) src_reg; similarly,
-BPF_XOR | BPF_K | BPF_ALU means A ^= imm32 in classic BPF and analogous
-src_reg = (u32) src_reg ^ (u32) imm32 in eBPF.
-
-Classic BPF is using BPF_MISC class to represent A = X and X = A moves.
-eBPF is using BPF_MOV | BPF_X | BPF_ALU code instead. Since there are no
-BPF_MISC operations in eBPF, the class 7 is used as BPF_ALU64 to mean
-exactly the same operations as BPF_ALU, but with 64-bit wide operands
-instead. So BPF_ADD | BPF_X | BPF_ALU64 means 64-bit addition, i.e.:
-dst_reg = dst_reg + src_reg
-
-Classic BPF wastes the whole BPF_RET class to represent a single ``ret``
-operation. Classic BPF_RET | BPF_K means copy imm32 into return register
-and perform function exit. eBPF is modeled to match CPU, so BPF_JMP | BPF_EXIT
-in eBPF means function exit only. The eBPF program needs to store return
-value into register R0 before doing a BPF_EXIT. Class 6 in eBPF is used as
-BPF_JMP32 to mean exactly the same operations as BPF_JMP, but with 32-bit wide
-operands for the comparisons instead.
-
-For load and store instructions the 8-bit 'code' field is divided as::
-
-  +--------+--------+-------------------+
-  | 3 bits | 2 bits |   3 bits          |
-  |  mode  |  size  | instruction class |
-  +--------+--------+-------------------+
-  (MSB)                             (LSB)
-
-Size modifier is one of ...
-
-::
-
-  BPF_W   0x00    /* word */
-  BPF_H   0x08    /* half word */
-  BPF_B   0x10    /* byte */
-  BPF_DW  0x18    /* eBPF only, double word */
-
-... which encodes size of load/store operation::
-
- B  - 1 byte
- H  - 2 byte
- W  - 4 byte
- DW - 8 byte (eBPF only)
-
-Mode modifier is one of::
-
-  BPF_IMM     0x00  /* used for 32-bit mov in classic BPF and 64-bit in eBPF */
-  BPF_ABS     0x20
-  BPF_IND     0x40
-  BPF_MEM     0x60
-  BPF_LEN     0x80  /* classic BPF only, reserved in eBPF */
-  BPF_MSH     0xa0  /* classic BPF only, reserved in eBPF */
-  BPF_ATOMIC  0xc0  /* eBPF only, atomic operations */
-
-eBPF has two non-generic instructions: (BPF_ABS | <size> | BPF_LD) and
-(BPF_IND | <size> | BPF_LD) which are used to access packet data.
-
-They had to be carried over from classic to have strong performance of
-socket filters running in eBPF interpreter. These instructions can only
-be used when interpreter context is a pointer to ``struct sk_buff`` and
-have seven implicit operands. Register R6 is an implicit input that must
-contain pointer to sk_buff. Register R0 is an implicit output which contains
-the data fetched from the packet. Registers R1-R5 are scratch registers
-and must not be used to store the data across BPF_ABS | BPF_LD or
-BPF_IND | BPF_LD instructions.
-
-These instructions have implicit program exit condition as well. When
-eBPF program is trying to access the data beyond the packet boundary,
-the interpreter will abort the execution of the program. JIT compilers
-therefore must preserve this property. src_reg and imm32 fields are
-explicit inputs to these instructions.
-
-For example::
-
-  BPF_IND | BPF_W | BPF_LD means:
-
-    R0 = ntohl(*(u32 *) (((struct sk_buff *) R6)->data + src_reg + imm32))
-    and R1 - R5 were scratched.
-
-Unlike classic BPF instruction set, eBPF has generic load/store operations::
-
-    BPF_MEM | <size> | BPF_STX:  *(size *) (dst_reg + off) = src_reg
-    BPF_MEM | <size> | BPF_ST:   *(size *) (dst_reg + off) = imm32
-    BPF_MEM | <size> | BPF_LDX:  dst_reg = *(size *) (src_reg + off)
-
-Where size is one of: BPF_B or BPF_H or BPF_W or BPF_DW.
-
-It also includes atomic operations, which use the immediate field for extra
-encoding::
-
-   .imm = BPF_ADD, .code = BPF_ATOMIC | BPF_W  | BPF_STX: lock xadd *(u32 *)(dst_reg + off16) += src_reg
-   .imm = BPF_ADD, .code = BPF_ATOMIC | BPF_DW | BPF_STX: lock xadd *(u64 *)(dst_reg + off16) += src_reg
-
-The basic atomic operations supported are::
-
-    BPF_ADD
-    BPF_AND
-    BPF_OR
-    BPF_XOR
-
-Each having equivalent semantics with the ``BPF_ADD`` example, that is: the
-memory location addresed by ``dst_reg + off`` is atomically modified, with
-``src_reg`` as the other operand. If the ``BPF_FETCH`` flag is set in the
-immediate, then these operations also overwrite ``src_reg`` with the
-value that was in memory before it was modified.
-
-The more special operations are::
-
-    BPF_XCHG
-
-This atomically exchanges ``src_reg`` with the value addressed by ``dst_reg +
-off``. ::
-
-    BPF_CMPXCHG
-
-This atomically compares the value addressed by ``dst_reg + off`` with
-``R0``. If they match it is replaced with ``src_reg``. In either case, the
-value that was there before is zero-extended and loaded back to ``R0``.
-
-Note that 1 and 2 byte atomic operations are not supported.
-
-Clang can generate atomic instructions by default when ``-mcpu=v3`` is
-enabled. If a lower version for ``-mcpu`` is set, the only atomic instruction
-Clang can generate is ``BPF_ADD`` *without* ``BPF_FETCH``. If you need to enable
-the atomics features, while keeping a lower ``-mcpu`` version, you can use
-``-Xclang -target-feature -Xclang +alu32``.
-
-You may encounter ``BPF_XADD`` - this is a legacy name for ``BPF_ATOMIC``,
-referring to the exclusive-add operation encoded when the immediate field is
-zero.
-
-eBPF has one 16-byte instruction: ``BPF_LD | BPF_DW | BPF_IMM`` which consists
-of two consecutive ``struct bpf_insn`` 8-byte blocks and interpreted as single
-instruction that loads 64-bit immediate value into a dst_reg.
-Classic BPF has similar instruction: ``BPF_LD | BPF_W | BPF_IMM`` which loads
-32-bit immediate value into a register.
-
-eBPF verifier
--------------
-The safety of the eBPF program is determined in two steps.
-
-First step does DAG check to disallow loops and other CFG validation.
-In particular it will detect programs that have unreachable instructions.
-(though classic BPF checker allows them)
-
-Second step starts from the first insn and descends all possible paths.
-It simulates execution of every insn and observes the state change of
-registers and stack.
-
-At the start of the program the register R1 contains a pointer to context
-and has type PTR_TO_CTX.
-If verifier sees an insn that does R2=R1, then R2 has now type
-PTR_TO_CTX as well and can be used on the right hand side of expression.
-If R1=PTR_TO_CTX and insn is R2=R1+R1, then R2=SCALAR_VALUE,
-since addition of two valid pointers makes invalid pointer.
-(In 'secure' mode verifier will reject any type of pointer arithmetic to make
-sure that kernel addresses don't leak to unprivileged users)
-
-If register was never written to, it's not readable::
-
-  bpf_mov R0 = R2
-  bpf_exit
-
-will be rejected, since R2 is unreadable at the start of the program.
-
-After kernel function call, R1-R5 are reset to unreadable and
-R0 has a return type of the function.
-
-Since R6-R9 are callee saved, their state is preserved across the call.
-
-::
-
-  bpf_mov R6 = 1
-  bpf_call foo
-  bpf_mov R0 = R6
-  bpf_exit
-
-is a correct program. If there was R1 instead of R6, it would have
-been rejected.
-
-load/store instructions are allowed only with registers of valid types, which
-are PTR_TO_CTX, PTR_TO_MAP, PTR_TO_STACK. They are bounds and alignment checked.
-For example::
-
- bpf_mov R1 = 1
- bpf_mov R2 = 2
- bpf_xadd *(u32 *)(R1 + 3) += R2
- bpf_exit
-
-will be rejected, since R1 doesn't have a valid pointer type at the time of
-execution of instruction bpf_xadd.
-
-At the start R1 type is PTR_TO_CTX (a pointer to generic ``struct bpf_context``)
-A callback is used to customize verifier to restrict eBPF program access to only
-certain fields within ctx structure with specified size and alignment.
-
-For example, the following insn::
-
-  bpf_ld R0 = *(u32 *)(R6 + 8)
-
-intends to load a word from address R6 + 8 and store it into R0
-If R6=PTR_TO_CTX, via is_valid_access() callback the verifier will know
-that offset 8 of size 4 bytes can be accessed for reading, otherwise
-the verifier will reject the program.
-If R6=PTR_TO_STACK, then access should be aligned and be within
-stack bounds, which are [-MAX_BPF_STACK, 0). In this example offset is 8,
-so it will fail verification, since it's out of bounds.
-
-The verifier will allow eBPF program to read data from stack only after
-it wrote into it.
-
-Classic BPF verifier does similar check with M[0-15] memory slots.
-For example::
-
-  bpf_ld R0 = *(u32 *)(R10 - 4)
-  bpf_exit
-
-is invalid program.
-Though R10 is correct read-only register and has type PTR_TO_STACK
-and R10 - 4 is within stack bounds, there were no stores into that location.
-
-Pointer register spill/fill is tracked as well, since four (R6-R9)
-callee saved registers may not be enough for some programs.
-
-Allowed function calls are customized with bpf_verifier_ops->get_func_proto()
-The eBPF verifier will check that registers match argument constraints.
-After the call register R0 will be set to return type of the function.
-
-Function calls is a main mechanism to extend functionality of eBPF programs.
-Socket filters may let programs to call one set of functions, whereas tracing
-filters may allow completely different set.
-
-If a function made accessible to eBPF program, it needs to be thought through
-from safety point of view. The verifier will guarantee that the function is
-called with valid arguments.
-
-seccomp vs socket filters have different security restrictions for classic BPF.
-Seccomp solves this by two stage verifier: classic BPF verifier is followed
-by seccomp verifier. In case of eBPF one configurable verifier is shared for
-all use cases.
-
-See details of eBPF verifier in kernel/bpf/verifier.c
-
-Register value tracking
------------------------
-In order to determine the safety of an eBPF program, the verifier must track
-the range of possible values in each register and also in each stack slot.
-This is done with ``struct bpf_reg_state``, defined in include/linux/
-bpf_verifier.h, which unifies tracking of scalar and pointer values.  Each
-register state has a type, which is either NOT_INIT (the register has not been
-written to), SCALAR_VALUE (some value which is not usable as a pointer), or a
-pointer type.  The types of pointers describe their base, as follows:
-
-
-    PTR_TO_CTX
-			Pointer to bpf_context.
-    CONST_PTR_TO_MAP
-			Pointer to struct bpf_map.  "Const" because arithmetic
-			on these pointers is forbidden.
-    PTR_TO_MAP_VALUE
-			Pointer to the value stored in a map element.
-    PTR_TO_MAP_VALUE_OR_NULL
-			Either a pointer to a map value, or NULL; map accesses
-			(see section 'eBPF maps', below) return this type,
-			which becomes a PTR_TO_MAP_VALUE when checked != NULL.
-			Arithmetic on these pointers is forbidden.
-    PTR_TO_STACK
-			Frame pointer.
-    PTR_TO_PACKET
-			skb->data.
-    PTR_TO_PACKET_END
-			skb->data + headlen; arithmetic forbidden.
-    PTR_TO_SOCKET
-			Pointer to struct bpf_sock_ops, implicitly refcounted.
-    PTR_TO_SOCKET_OR_NULL
-			Either a pointer to a socket, or NULL; socket lookup
-			returns this type, which becomes a PTR_TO_SOCKET when
-			checked != NULL. PTR_TO_SOCKET is reference-counted,
-			so programs must release the reference through the
-			socket release function before the end of the program.
-			Arithmetic on these pointers is forbidden.
-
-However, a pointer may be offset from this base (as a result of pointer
-arithmetic), and this is tracked in two parts: the 'fixed offset' and 'variable
-offset'.  The former is used when an exactly-known value (e.g. an immediate
-operand) is added to a pointer, while the latter is used for values which are
-not exactly known.  The variable offset is also used in SCALAR_VALUEs, to track
-the range of possible values in the register.
-
-The verifier's knowledge about the variable offset consists of:
-
-* minimum and maximum values as unsigned
-* minimum and maximum values as signed
-
-* knowledge of the values of individual bits, in the form of a 'tnum': a u64
-  'mask' and a u64 'value'.  1s in the mask represent bits whose value is unknown;
-  1s in the value represent bits known to be 1.  Bits known to be 0 have 0 in both
-  mask and value; no bit should ever be 1 in both.  For example, if a byte is read
-  into a register from memory, the register's top 56 bits are known zero, while
-  the low 8 are unknown - which is represented as the tnum (0x0; 0xff).  If we
-  then OR this with 0x40, we get (0x40; 0xbf), then if we add 1 we get (0x0;
-  0x1ff), because of potential carries.
-
-Besides arithmetic, the register state can also be updated by conditional
-branches.  For instance, if a SCALAR_VALUE is compared > 8, in the 'true' branch
-it will have a umin_value (unsigned minimum value) of 9, whereas in the 'false'
-branch it will have a umax_value of 8.  A signed compare (with BPF_JSGT or
-BPF_JSGE) would instead update the signed minimum/maximum values.  Information
-from the signed and unsigned bounds can be combined; for instance if a value is
-first tested < 8 and then tested s> 4, the verifier will conclude that the value
-is also > 4 and s< 8, since the bounds prevent crossing the sign boundary.
-
-PTR_TO_PACKETs with a variable offset part have an 'id', which is common to all
-pointers sharing that same variable offset.  This is important for packet range
-checks: after adding a variable to a packet pointer register A, if you then copy
-it to another register B and then add a constant 4 to A, both registers will
-share the same 'id' but the A will have a fixed offset of +4.  Then if A is
-bounds-checked and found to be less than a PTR_TO_PACKET_END, the register B is
-now known to have a safe range of at least 4 bytes.  See 'Direct packet access',
-below, for more on PTR_TO_PACKET ranges.
-
-The 'id' field is also used on PTR_TO_MAP_VALUE_OR_NULL, common to all copies of
-the pointer returned from a map lookup.  This means that when one copy is
-checked and found to be non-NULL, all copies can become PTR_TO_MAP_VALUEs.
-As well as range-checking, the tracked information is also used for enforcing
-alignment of pointer accesses.  For instance, on most systems the packet pointer
-is 2 bytes after a 4-byte alignment.  If a program adds 14 bytes to that to jump
-over the Ethernet header, then reads IHL and addes (IHL * 4), the resulting
-pointer will have a variable offset known to be 4n+2 for some n, so adding the 2
-bytes (NET_IP_ALIGN) gives a 4-byte alignment and so word-sized accesses through
-that pointer are safe.
-The 'id' field is also used on PTR_TO_SOCKET and PTR_TO_SOCKET_OR_NULL, common
-to all copies of the pointer returned from a socket lookup. This has similar
-behaviour to the handling for PTR_TO_MAP_VALUE_OR_NULL->PTR_TO_MAP_VALUE, but
-it also handles reference tracking for the pointer. PTR_TO_SOCKET implicitly
-represents a reference to the corresponding ``struct sock``. To ensure that the
-reference is not leaked, it is imperative to NULL-check the reference and in
-the non-NULL case, and pass the valid reference to the socket release function.
-
-Direct packet access
---------------------
-In cls_bpf and act_bpf programs the verifier allows direct access to the packet
-data via skb->data and skb->data_end pointers.
-Ex::
-
-    1:  r4 = *(u32 *)(r1 +80)  /* load skb->data_end */
-    2:  r3 = *(u32 *)(r1 +76)  /* load skb->data */
-    3:  r5 = r3
-    4:  r5 += 14
-    5:  if r5 > r4 goto pc+16
-    R1=ctx R3=pkt(id=0,off=0,r=14) R4=pkt_end R5=pkt(id=0,off=14,r=14) R10=fp
-    6:  r0 = *(u16 *)(r3 +12) /* access 12 and 13 bytes of the packet */
-
-this 2byte load from the packet is safe to do, since the program author
-did check ``if (skb->data + 14 > skb->data_end) goto err`` at insn #5 which
-means that in the fall-through case the register R3 (which points to skb->data)
-has at least 14 directly accessible bytes. The verifier marks it
-as R3=pkt(id=0,off=0,r=14).
-id=0 means that no additional variables were added to the register.
-off=0 means that no additional constants were added.
-r=14 is the range of safe access which means that bytes [R3, R3 + 14) are ok.
-Note that R5 is marked as R5=pkt(id=0,off=14,r=14). It also points
-to the packet data, but constant 14 was added to the register, so
-it now points to ``skb->data + 14`` and accessible range is [R5, R5 + 14 - 14)
-which is zero bytes.
-
-More complex packet access may look like::
-
-
-    R0=inv1 R1=ctx R3=pkt(id=0,off=0,r=14) R4=pkt_end R5=pkt(id=0,off=14,r=14) R10=fp
-    6:  r0 = *(u8 *)(r3 +7) /* load 7th byte from the packet */
-    7:  r4 = *(u8 *)(r3 +12)
-    8:  r4 *= 14
-    9:  r3 = *(u32 *)(r1 +76) /* load skb->data */
-    10:  r3 += r4
-    11:  r2 = r1
-    12:  r2 <<= 48
-    13:  r2 >>= 48
-    14:  r3 += r2
-    15:  r2 = r3
-    16:  r2 += 8
-    17:  r1 = *(u32 *)(r1 +80) /* load skb->data_end */
-    18:  if r2 > r1 goto pc+2
-    R0=inv(id=0,umax_value=255,var_off=(0x0; 0xff)) R1=pkt_end R2=pkt(id=2,off=8,r=8) R3=pkt(id=2,off=0,r=8) R4=inv(id=0,umax_value=3570,var_off=(0x0; 0xfffe)) R5=pkt(id=0,off=14,r=14) R10=fp
-    19:  r1 = *(u8 *)(r3 +4)
-
-The state of the register R3 is R3=pkt(id=2,off=0,r=8)
-id=2 means that two ``r3 += rX`` instructions were seen, so r3 points to some
-offset within a packet and since the program author did
-``if (r3 + 8 > r1) goto err`` at insn #18, the safe range is [R3, R3 + 8).
-The verifier only allows 'add'/'sub' operations on packet registers. Any other
-operation will set the register state to 'SCALAR_VALUE' and it won't be
-available for direct packet access.
-
-Operation ``r3 += rX`` may overflow and become less than original skb->data,
-therefore the verifier has to prevent that.  So when it sees ``r3 += rX``
-instruction and rX is more than 16-bit value, any subsequent bounds-check of r3
-against skb->data_end will not give us 'range' information, so attempts to read
-through the pointer will give "invalid access to packet" error.
-
-Ex. after insn ``r4 = *(u8 *)(r3 +12)`` (insn #7 above) the state of r4 is
-R4=inv(id=0,umax_value=255,var_off=(0x0; 0xff)) which means that upper 56 bits
-of the register are guaranteed to be zero, and nothing is known about the lower
-8 bits. After insn ``r4 *= 14`` the state becomes
-R4=inv(id=0,umax_value=3570,var_off=(0x0; 0xfffe)), since multiplying an 8-bit
-value by constant 14 will keep upper 52 bits as zero, also the least significant
-bit will be zero as 14 is even.  Similarly ``r2 >>= 48`` will make
-R2=inv(id=0,umax_value=65535,var_off=(0x0; 0xffff)), since the shift is not sign
-extending.  This logic is implemented in adjust_reg_min_max_vals() function,
-which calls adjust_ptr_min_max_vals() for adding pointer to scalar (or vice
-versa) and adjust_scalar_min_max_vals() for operations on two scalars.
-
-The end result is that bpf program author can access packet directly
-using normal C code as::
-
-  void *data = (void *)(long)skb->data;
-  void *data_end = (void *)(long)skb->data_end;
-  struct eth_hdr *eth = data;
-  struct iphdr *iph = data + sizeof(*eth);
-  struct udphdr *udp = data + sizeof(*eth) + sizeof(*iph);
-
-  if (data + sizeof(*eth) + sizeof(*iph) + sizeof(*udp) > data_end)
-	  return 0;
-  if (eth->h_proto != htons(ETH_P_IP))
-	  return 0;
-  if (iph->protocol != IPPROTO_UDP || iph->ihl != 5)
-	  return 0;
-  if (udp->dest == 53 || udp->source == 9)
-	  ...;
-
-which makes such programs easier to write comparing to LD_ABS insn
-and significantly faster.
-
-eBPF maps
----------
-'maps' is a generic storage of different types for sharing data between kernel
-and userspace.
-
-The maps are accessed from user space via BPF syscall, which has commands:
-
-- create a map with given type and attributes
-  ``map_fd = bpf(BPF_MAP_CREATE, union bpf_attr *attr, u32 size)``
-  using attr->map_type, attr->key_size, attr->value_size, attr->max_entries
-  returns process-local file descriptor or negative error
-
-- lookup key in a given map
-  ``err = bpf(BPF_MAP_LOOKUP_ELEM, union bpf_attr *attr, u32 size)``
-  using attr->map_fd, attr->key, attr->value
-  returns zero and stores found elem into value or negative error
-
-- create or update key/value pair in a given map
-  ``err = bpf(BPF_MAP_UPDATE_ELEM, union bpf_attr *attr, u32 size)``
-  using attr->map_fd, attr->key, attr->value
-  returns zero or negative error
-
-- find and delete element by key in a given map
-  ``err = bpf(BPF_MAP_DELETE_ELEM, union bpf_attr *attr, u32 size)``
-  using attr->map_fd, attr->key
-
-- to delete map: close(fd)
-  Exiting process will delete maps automatically
-
-userspace programs use this syscall to create/access maps that eBPF programs
-are concurrently updating.
-
-maps can have different types: hash, array, bloom filter, radix-tree, etc.
-
-The map is defined by:
-
-  - type
-  - max number of elements
-  - key size in bytes
-  - value size in bytes
-
-Pruning
--------
-The verifier does not actually walk all possible paths through the program.  For
-each new branch to analyse, the verifier looks at all the states it's previously
-been in when at this instruction.  If any of them contain the current state as a
-subset, the branch is 'pruned' - that is, the fact that the previous state was
-accepted implies the current state would be as well.  For instance, if in the
-previous state, r1 held a packet-pointer, and in the current state, r1 holds a
-packet-pointer with a range as long or longer and at least as strict an
-alignment, then r1 is safe.  Similarly, if r2 was NOT_INIT before then it can't
-have been used by any path from that point, so any value in r2 (including
-another NOT_INIT) is safe.  The implementation is in the function regsafe().
-Pruning considers not only the registers but also the stack (and any spilled
-registers it may hold).  They must all be safe for the branch to be pruned.
-This is implemented in states_equal().
-
-Understanding eBPF verifier messages
-------------------------------------
-
-The following are few examples of invalid eBPF programs and verifier error
-messages as seen in the log:
-
-Program with unreachable instructions::
-
-  static struct bpf_insn prog[] = {
-  BPF_EXIT_INSN(),
-  BPF_EXIT_INSN(),
-  };
-
-Error:
-
-  unreachable insn 1
-
-Program that reads uninitialized register::
-
-  BPF_MOV64_REG(BPF_REG_0, BPF_REG_2),
-  BPF_EXIT_INSN(),
-
-Error::
-
-  0: (bf) r0 = r2
-  R2 !read_ok
-
-Program that doesn't initialize R0 before exiting::
-
-  BPF_MOV64_REG(BPF_REG_2, BPF_REG_1),
-  BPF_EXIT_INSN(),
-
-Error::
-
-  0: (bf) r2 = r1
-  1: (95) exit
-  R0 !read_ok
-
-Program that accesses stack out of bounds::
-
-    BPF_ST_MEM(BPF_DW, BPF_REG_10, 8, 0),
-    BPF_EXIT_INSN(),
-
-Error::
-
-    0: (7a) *(u64 *)(r10 +8) = 0
-    invalid stack off=8 size=8
-
-Program that doesn't initialize stack before passing its address into function::
-
-  BPF_MOV64_REG(BPF_REG_2, BPF_REG_10),
-  BPF_ALU64_IMM(BPF_ADD, BPF_REG_2, -8),
-  BPF_LD_MAP_FD(BPF_REG_1, 0),
-  BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0, BPF_FUNC_map_lookup_elem),
-  BPF_EXIT_INSN(),
-
-Error::
-
-  0: (bf) r2 = r10
-  1: (07) r2 += -8
-  2: (b7) r1 = 0x0
-  3: (85) call 1
-  invalid indirect read from stack off -8+0 size 8
-
-Program that uses invalid map_fd=0 while calling to map_lookup_elem() function::
-
-  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_LD_MAP_FD(BPF_REG_1, 0),
-  BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0, BPF_FUNC_map_lookup_elem),
-  BPF_EXIT_INSN(),
-
-Error::
-
-  0: (7a) *(u64 *)(r10 -8) = 0
-  1: (bf) r2 = r10
-  2: (07) r2 += -8
-  3: (b7) r1 = 0x0
-  4: (85) call 1
-  fd 0 is not pointing to valid bpf_map
-
-Program that doesn't check return value of map_lookup_elem() before accessing
-map element::
-
-  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_LD_MAP_FD(BPF_REG_1, 0),
-  BPF_RAW_INSN(BPF_JMP | BPF_CALL, 0, 0, 0, BPF_FUNC_map_lookup_elem),
-  BPF_ST_MEM(BPF_DW, BPF_REG_0, 0, 0),
-  BPF_EXIT_INSN(),
-
-Error::
-
-  0: (7a) *(u64 *)(r10 -8) = 0
-  1: (bf) r2 = r10
-  2: (07) r2 += -8
-  3: (b7) r1 = 0x0
-  4: (85) call 1
-  5: (7a) *(u64 *)(r0 +0) = 0
-  R0 invalid mem access 'map_value_or_null'
-
-Program that correctly checks map_lookup_elem() returned value for NULL, but
-accesses the memory with incorrect alignment::
-
-  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_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, 1),
-  BPF_ST_MEM(BPF_DW, BPF_REG_0, 4, 0),
-  BPF_EXIT_INSN(),
-
-Error::
-
-  0: (7a) *(u64 *)(r10 -8) = 0
-  1: (bf) r2 = r10
-  2: (07) r2 += -8
-  3: (b7) r1 = 1
-  4: (85) call 1
-  5: (15) if r0 == 0x0 goto pc+1
-   R0=map_ptr R10=fp
-  6: (7a) *(u64 *)(r0 +4) = 0
-  misaligned access off 4 size 8
-
-Program that correctly checks map_lookup_elem() returned value for NULL and
-accesses memory with correct alignment in one side of 'if' branch, but fails
-to do so in the other side of 'if' branch::
-
-  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_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, 2),
-  BPF_ST_MEM(BPF_DW, BPF_REG_0, 0, 0),
-  BPF_EXIT_INSN(),
-  BPF_ST_MEM(BPF_DW, BPF_REG_0, 0, 1),
-  BPF_EXIT_INSN(),
-
-Error::
-
-  0: (7a) *(u64 *)(r10 -8) = 0
-  1: (bf) r2 = r10
-  2: (07) r2 += -8
-  3: (b7) r1 = 1
-  4: (85) call 1
-  5: (15) if r0 == 0x0 goto pc+2
-   R0=map_ptr R10=fp
-  6: (7a) *(u64 *)(r0 +0) = 0
-  7: (95) exit
-
-  from 5 to 8: R0=imm0 R10=fp
-  8: (7a) *(u64 *)(r0 +0) = 1
-  R0 invalid mem access 'imm'
-
-Program that performs a socket lookup then sets the pointer to NULL without
-checking it::
-
-  BPF_MOV64_IMM(BPF_REG_2, 0),
-  BPF_STX_MEM(BPF_W, BPF_REG_10, BPF_REG_2, -8),
-  BPF_MOV64_REG(BPF_REG_2, BPF_REG_10),
-  BPF_ALU64_IMM(BPF_ADD, BPF_REG_2, -8),
-  BPF_MOV64_IMM(BPF_REG_3, 4),
-  BPF_MOV64_IMM(BPF_REG_4, 0),
-  BPF_MOV64_IMM(BPF_REG_5, 0),
-  BPF_EMIT_CALL(BPF_FUNC_sk_lookup_tcp),
-  BPF_MOV64_IMM(BPF_REG_0, 0),
-  BPF_EXIT_INSN(),
-
-Error::
-
-  0: (b7) r2 = 0
-  1: (63) *(u32 *)(r10 -8) = r2
-  2: (bf) r2 = r10
-  3: (07) r2 += -8
-  4: (b7) r3 = 4
-  5: (b7) r4 = 0
-  6: (b7) r5 = 0
-  7: (85) call bpf_sk_lookup_tcp#65
-  8: (b7) r0 = 0
-  9: (95) exit
-  Unreleased reference id=1, alloc_insn=7
-
-Program that performs a socket lookup but does not NULL-check the returned
-value::
-
-  BPF_MOV64_IMM(BPF_REG_2, 0),
-  BPF_STX_MEM(BPF_W, BPF_REG_10, BPF_REG_2, -8),
-  BPF_MOV64_REG(BPF_REG_2, BPF_REG_10),
-  BPF_ALU64_IMM(BPF_ADD, BPF_REG_2, -8),
-  BPF_MOV64_IMM(BPF_REG_3, 4),
-  BPF_MOV64_IMM(BPF_REG_4, 0),
-  BPF_MOV64_IMM(BPF_REG_5, 0),
-  BPF_EMIT_CALL(BPF_FUNC_sk_lookup_tcp),
-  BPF_EXIT_INSN(),
-
-Error::
-
-  0: (b7) r2 = 0
-  1: (63) *(u32 *)(r10 -8) = r2
-  2: (bf) r2 = r10
-  3: (07) r2 += -8
-  4: (b7) r3 = 4
-  5: (b7) r4 = 0
-  6: (b7) r5 = 0
-  7: (85) call bpf_sk_lookup_tcp#65
-  8: (95) exit
-  Unreleased reference id=1, alloc_insn=7
-
 Testing
 -------
 
 Next to the BPF toolchain, the kernel also ships a test module that contains
-various test cases for classic and internal BPF that can be executed against
+various test cases for classic and eBPF that can be executed against
 the BPF interpreter and JIT compiler. It can be found in lib/test_bpf.c and
 enabled via Kconfig::
 

MAINTAINERS

···
 R:	Brendan Jackman <[e-mail]>
 L:	[e-mail]
 S:	Maintained
-F:	Documentation/bpf/bpf_lsm.rst
+F:	Documentation/bpf/prog_lsm.rst
 F:	include/linux/bpf_lsm.h
 F:	kernel/bpf/bpf_lsm.c
 F:	security/bpf/

arch/arm/net/bpf_jit_32.c

···
 	[BPF_REG_9] = {STACK_OFFSET(BPF_R9_HI), STACK_OFFSET(BPF_R9_LO)},
 	/* Read only Frame Pointer to access Stack */
 	[BPF_REG_FP] = {STACK_OFFSET(BPF_FP_HI), STACK_OFFSET(BPF_FP_LO)},
-	/* Temporary Register for internal BPF JIT, can be used
+	/* Temporary Register for BPF JIT, can be used
 	 * for constant blindings and others.
 	 */
 	[TMP_REG_1] = {ARM_R7, ARM_R6},
···
 
 	/* tmp2[0] = array, tmp2[1] = index */
 
-	/* if (tail_call_cnt > MAX_TAIL_CALL_CNT)
+	/*
+	 * if (tail_call_cnt >= MAX_TAIL_CALL_CNT)
 	 *	goto out;
 	 * tail_call_cnt++;
 	 */
···
 	tc = arm_bpf_get_reg64(tcc, tmp, ctx);
 	emit(ARM_CMP_I(tc[0], hi), ctx);
 	_emit(ARM_COND_EQ, ARM_CMP_I(tc[1], lo), ctx);
-	_emit(ARM_COND_HI, ARM_B(jmp_offset), ctx);
+	_emit(ARM_COND_CS, ARM_B(jmp_offset), ctx);
 	emit(ARM_ADDS_I(tc[1], tc[1], 1), ctx);
 	emit(ARM_ADC_I(tc[0], tc[0], 0), ctx);
 	arm_bpf_put_reg64(tcc, tmp, ctx);

arch/arm64/net/bpf_jit_comp.c

···
 	[BPF_REG_9] = A64_R(22),
 	/* read-only frame pointer to access stack */
 	[BPF_REG_FP] = A64_R(25),
-	/* temporary registers for internal BPF JIT */
+	/* temporary registers for BPF JIT */
 	[TMP_REG_1] = A64_R(10),
 	[TMP_REG_2] = A64_R(11),
 	[TMP_REG_3] = A64_R(12),
···
 	emit(A64_CMP(0, r3, tmp), ctx);
 	emit(A64_B_(A64_COND_CS, jmp_offset), ctx);
 
-	/* if (tail_call_cnt > MAX_TAIL_CALL_CNT)
+	/*
+	 * if (tail_call_cnt >= MAX_TAIL_CALL_CNT)
 	 *     goto out;
 	 * tail_call_cnt++;
 	 */
 	emit_a64_mov_i64(tmp, MAX_TAIL_CALL_CNT, ctx);
 	emit(A64_CMP(1, tcc, tmp), ctx);
-	emit(A64_B_(A64_COND_HI, jmp_offset), ctx);
+	emit(A64_B_(A64_COND_CS, jmp_offset), ctx);
 	emit(A64_ADD_I(1, tcc, tcc, 1), ctx);
 
 	/* prog = array->ptrs[index];

arch/mips/net/bpf_jit_comp32.c

···
 	 * 16-byte area in the parent's stack frame. On a tail call, the
 	 * calling function jumps into the prologue after these instructions.
 	 */
-	emit(ctx, ori, MIPS_R_T9, MIPS_R_ZERO,
-	     min(MAX_TAIL_CALL_CNT + 1, 0xffff));
+	emit(ctx, ori, MIPS_R_T9, MIPS_R_ZERO, min(MAX_TAIL_CALL_CNT, 0xffff));
 	emit(ctx, sw, MIPS_R_T9, 0, MIPS_R_SP);
 
 	/*

arch/mips/net/bpf_jit_comp64.c

···
 	 * On a tail call, the calling function jumps into the prologue
 	 * after this instruction.
 	 */
-	emit(ctx, addiu, tc, MIPS_R_ZERO, min(MAX_TAIL_CALL_CNT + 1, 0xffff));
+	emit(ctx, ori, tc, MIPS_R_ZERO, min(MAX_TAIL_CALL_CNT, 0xffff));
 
 	/* === Entry-point for tail calls === */
 

arch/powerpc/net/bpf_jit_comp32.c

···
 	PPC_BCC(COND_GE, out);
 
 	/*
-	 * if (tail_call_cnt > MAX_TAIL_CALL_CNT)
+	 * if (tail_call_cnt >= MAX_TAIL_CALL_CNT)
 	 *   goto out;
 	 */
 	EMIT(PPC_RAW_CMPLWI(_R0, MAX_TAIL_CALL_CNT));
 	/* tail_call_cnt++; */
 	EMIT(PPC_RAW_ADDIC(_R0, _R0, 1));
-	PPC_BCC(COND_GT, out);
+	PPC_BCC(COND_GE, out);
 
 	/* prog = array->ptrs[index]; */
 	EMIT(PPC_RAW_RLWINM(_R3, b2p_index, 2, 0, 29));

arch/powerpc/net/bpf_jit_comp64.c

···
 	PPC_BCC(COND_GE, out);
 
 	/*
-	 * if (tail_call_cnt > MAX_TAIL_CALL_CNT)
+	 * if (tail_call_cnt >= MAX_TAIL_CALL_CNT)
 	 *   goto out;
 	 */
 	PPC_BPF_LL(b2p[TMP_REG_1], 1, bpf_jit_stack_tailcallcnt(ctx));
 	EMIT(PPC_RAW_CMPLWI(b2p[TMP_REG_1], MAX_TAIL_CALL_CNT));
-	PPC_BCC(COND_GT, out);
+	PPC_BCC(COND_GE, out);
 
 	/*
 	 * tail_call_cnt++;

arch/riscv/net/bpf_jit_comp32.c

···
 	emit_bcc(BPF_JGE, lo(idx_reg), RV_REG_T1, off, ctx);
 
 	/*
-	 * temp_tcc = tcc - 1;
-	 * if (tcc < 0)
+	 * if (--tcc < 0)
 	 *   goto out;
 	 */
-	emit(rv_addi(RV_REG_T1, RV_REG_TCC, -1), ctx);
+	emit(rv_addi(RV_REG_TCC, RV_REG_TCC, -1), ctx);
 	off = ninsns_rvoff(tc_ninsn - (ctx->ninsns - start_insn));
 	emit_bcc(BPF_JSLT, RV_REG_TCC, RV_REG_ZERO, off, ctx);
 
···
 	if (is_12b_check(off, insn))
 		return -1;
 	emit(rv_lw(RV_REG_T0, off, RV_REG_T0), ctx);
-	emit(rv_addi(RV_REG_TCC, RV_REG_T1, 0), ctx);
 	/* Epilogue jumps to *(t0 + 4). */
 	__build_epilogue(true, ctx);
 	return 0;

arch/riscv/net/bpf_jit_comp64.c

···
 	off = ninsns_rvoff(tc_ninsn - (ctx->ninsns - start_insn));
 	emit_branch(BPF_JGE, RV_REG_A2, RV_REG_T1, off, ctx);
 
-	/* if (TCC-- < 0)
+	/* if (--TCC < 0)
 	 *     goto out;
 	 */
-	emit_addi(RV_REG_T1, tcc, -1, ctx);
+	emit_addi(RV_REG_TCC, tcc, -1, ctx);
 	off = ninsns_rvoff(tc_ninsn - (ctx->ninsns - start_insn));
-	emit_branch(BPF_JSLT, tcc, RV_REG_ZERO, off, ctx);
+	emit_branch(BPF_JSLT, RV_REG_TCC, RV_REG_ZERO, off, ctx);
 
 	/* prog = array->ptrs[index];
 	 * if (!prog)
···
 	if (is_12b_check(off, insn))
 		return -1;
 	emit_ld(RV_REG_T3, off, RV_REG_T2, ctx);
-	emit_mv(RV_REG_TCC, RV_REG_T1, ctx);
 	__build_epilogue(true, ctx);
 	return 0;
 }

arch/s390/net/bpf_jit_comp.c

···
 				 jit->prg);
 
 		/*
-		 * if (tail_call_cnt++ > MAX_TAIL_CALL_CNT)
+		 * if (tail_call_cnt++ >= MAX_TAIL_CALL_CNT)
 		 *         goto out;
 		 */
 
···
 		EMIT4_IMM(0xa7080000, REG_W0, 1);
 		/* laal %w1,%w0,off(%r15) */
 		EMIT6_DISP_LH(0xeb000000, 0x00fa, REG_W1, REG_W0, REG_15, off);
-		/* clij %w1,MAX_TAIL_CALL_CNT,0x2,out */
+		/* clij %w1,MAX_TAIL_CALL_CNT-1,0x2,out */
 		patch_2_clij = jit->prg;
-		EMIT6_PCREL_RIEC(0xec000000, 0x007f, REG_W1, MAX_TAIL_CALL_CNT,
+		EMIT6_PCREL_RIEC(0xec000000, 0x007f, REG_W1, MAX_TAIL_CALL_CNT - 1,
 				 2, jit->prg);
 
 		/*

arch/sparc/net/bpf_jit_comp_64.c

···
 
 	[BPF_REG_AX] = G7,
 
-	/* temporary register for internal BPF JIT */
+	/* temporary register for BPF JIT */
 	[TMP_REG_1] = G1,
 	[TMP_REG_2] = G2,
 	[TMP_REG_3] = G3,
···
 	emit(LD32 | IMMED | RS1(SP) | S13(off) | RD(tmp), ctx);
 	emit_cmpi(tmp, MAX_TAIL_CALL_CNT, ctx);
 #define OFFSET2 13
-	emit_branch(BGU, ctx->idx, ctx->idx + OFFSET2, ctx);
+	emit_branch(BGEU, ctx->idx, ctx->idx + OFFSET2, ctx);
 	emit_nop(ctx);
 
 	emit_alu_K(ADD, tmp, 1, ctx);

arch/x86/net/bpf_jit_comp.c

···
 // SPDX-License-Identifier: GPL-2.0-only
 /*
- * bpf_jit_comp.c: BPF JIT compiler
+ * BPF JIT compiler
  *
  * Copyright (C) 2011-2013 Eric Dumazet ([e-mail])
- * Internal BPF Copyright (c) 2011-2014 PLUMgrid, http://plumgrid.com
+ * Copyright (c) 2011-2014 PLUMgrid, http://plumgrid.com
  */
 #include <linux/netdevice.h>
 #include <linux/filter.h>
···
  * ... bpf_tail_call(void *ctx, struct bpf_array *array, u64 index) ...
  *   if (index >= array->map.max_entries)
  *     goto out;
- *   if (++tail_call_cnt > MAX_TAIL_CALL_CNT)
+ *   if (tail_call_cnt++ >= MAX_TAIL_CALL_CNT)
  *     goto out;
  *   prog = array->ptrs[index];
  *   if (prog == NULL)
···
 	EMIT2(X86_JBE, offset);                   /* jbe out */
 
 	/*
-	 * if (tail_call_cnt > MAX_TAIL_CALL_CNT)
+	 * if (tail_call_cnt++ >= MAX_TAIL_CALL_CNT)
 	 *	goto out;
 	 */
 	EMIT2_off32(0x8B, 0x85, tcc_off);         /* mov eax, dword ptr [rbp - tcc_off] */
 	EMIT3(0x83, 0xF8, MAX_TAIL_CALL_CNT);     /* cmp eax, MAX_TAIL_CALL_CNT */
 
 	offset = ctx->tail_call_indirect_label - (prog + 2 - start);
-	EMIT2(X86_JA, offset);                    /* ja out */
+	EMIT2(X86_JAE, offset);                   /* jae out */
 	EMIT3(0x83, 0xC0, 0x01);                  /* add eax, 1 */
 	EMIT2_off32(0x89, 0x85, tcc_off);         /* mov dword ptr [rbp - tcc_off], eax */
 
···
 	int offset;
 
 	/*
-	 * if (tail_call_cnt > MAX_TAIL_CALL_CNT)
+	 * if (tail_call_cnt++ >= MAX_TAIL_CALL_CNT)
 	 *	goto out;
 	 */
 	EMIT2_off32(0x8B, 0x85, tcc_off);             /* mov eax, dword ptr [rbp - tcc_off] */
 	EMIT3(0x83, 0xF8, MAX_TAIL_CALL_CNT);         /* cmp eax, MAX_TAIL_CALL_CNT */
 
 	offset = ctx->tail_call_direct_label - (prog + 2 - start);
-	EMIT2(X86_JA, offset);                        /* ja out */
+	EMIT2(X86_JAE, offset);                       /* jae out */
 	EMIT3(0x83, 0xC0, 0x01);                      /* add eax, 1 */
 	EMIT2_off32(0x89, 0x85, tcc_off);             /* mov dword ptr [rbp - tcc_off], eax */
 

arch/x86/net/bpf_jit_comp32.c

···
 	EMIT2(IA32_JBE, jmp_label(jmp_label1, 2));
 
 	/*
-	 * if (tail_call_cnt > MAX_TAIL_CALL_CNT)
+	 * if (tail_call_cnt++ >= MAX_TAIL_CALL_CNT)
 	 *     goto out;
 	 */
 	lo = (u32)MAX_TAIL_CALL_CNT;
···
 	/* cmp ecx,lo */
 	EMIT3(0x83, add_1reg(0xF8, IA32_ECX), lo);
 
-	/* ja out */
+	/* jae out */
 	EMIT2(IA32_JAE, jmp_label(jmp_label1, 2));
 
 	/* add eax,0x1 */

include/linux/bpf.h

···
 };
 
 #define BPF_COMPLEXITY_LIMIT_INSNS      1000000 /* yes. 1M insns */
-#define MAX_TAIL_CALL_CNT 32
+#define MAX_TAIL_CALL_CNT 33
 
 #define BPF_F_ACCESS_MASK	(BPF_F_RDONLY |		\
 				 BPF_F_RDONLY_PROG |	\
···
 const struct btf_func_model *
 bpf_jit_find_kfunc_model(const struct bpf_prog *prog,
 			 const struct bpf_insn *insn);
+struct bpf_core_ctx {
+	struct bpf_verifier_log *log;
+	const struct btf *btf;
+};
+
+int bpf_core_apply(struct bpf_core_ctx *ctx, const struct bpf_core_relo *relo,
+		   int relo_idx, void *insn);
+
 #else /* !CONFIG_BPF_SYSCALL */
 static inline struct bpf_prog *bpf_prog_get(u32 ufd)
 {
···
 extern const struct bpf_func_proto bpf_sk_getsockopt_proto;
 extern const struct bpf_func_proto bpf_kallsyms_lookup_name_proto;
 extern const struct bpf_func_proto bpf_find_vma_proto;
+extern const struct bpf_func_proto bpf_loop_proto;
 
 const struct bpf_func_proto *tracing_prog_func_proto(
   enum bpf_func_id func_id, const struct bpf_prog *prog);

include/linux/bpf_verifier.h

···
 		 log->level == BPF_LOG_KERNEL);
 }
 
+static inline bool
+bpf_verifier_log_attr_valid(const struct bpf_verifier_log *log)
+{
+	return log->len_total >= 128 && log->len_total <= UINT_MAX >> 2 &&
+	       log->level && log->ubuf && !(log->level & ~BPF_LOG_MASK);
+}
+
 #define BPF_MAX_SUBPROGS 256
 
 struct bpf_subprog_info {

include/linux/btf.h

···
 	return BTF_INFO_KIND(t->info) == BTF_KIND_ENUM;
 }
 
+static inline bool str_is_empty(const char *s)
+{
+	return !s || !s[0];
+}
+
+static inline u16 btf_kind(const struct btf_type *t)
+{
+	return BTF_INFO_KIND(t->info);
+}
+
+static inline bool btf_is_enum(const struct btf_type *t)
+{
+	return btf_kind(t) == BTF_KIND_ENUM;
+}
+
+static inline bool btf_is_composite(const struct btf_type *t)
+{
+	u16 kind = btf_kind(t);
+
+	return kind == BTF_KIND_STRUCT || kind == BTF_KIND_UNION;
+}
+
+static inline bool btf_is_array(const struct btf_type *t)
+{
+	return btf_kind(t) == BTF_KIND_ARRAY;
+}
+
+static inline bool btf_is_int(const struct btf_type *t)
+{
+	return btf_kind(t) == BTF_KIND_INT;
+}
+
+static inline bool btf_is_ptr(const struct btf_type *t)
+{
+	return btf_kind(t) == BTF_KIND_PTR;
+}
+
+static inline u8 btf_int_offset(const struct btf_type *t)
+{
+	return BTF_INT_OFFSET(*(u32 *)(t + 1));
+}
+
+static inline u8 btf_int_encoding(const struct btf_type *t)
+{
+	return BTF_INT_ENCODING(*(u32 *)(t + 1));
+}
+
 static inline bool btf_type_is_scalar(const struct btf_type *t)
 {
 	return btf_type_is_int(t) || btf_type_is_enum(t);
···
 	return BTF_INFO_VLEN(t->info);
 }
 
+static inline u16 btf_vlen(const struct btf_type *t)
+{
+	return btf_type_vlen(t);
+}
+
 static inline u16 btf_func_linkage(const struct btf_type *t)
 {
 	return BTF_INFO_VLEN(t->info);
···
 	return BTF_INFO_KFLAG(t->info);
 }
 
-static inline u32 btf_member_bit_offset(const struct btf_type *struct_type,
-					const struct btf_member *member)
+static inline u32 __btf_member_bit_offset(const struct btf_type *struct_type,
+					  const struct btf_member *member)
 {
 	return btf_type_kflag(struct_type) ? BTF_MEMBER_BIT_OFFSET(member->offset)
 					   : member->offset;
 }
 
-static inline u32 btf_member_bitfield_size(const struct btf_type *struct_type,
-					   const struct btf_member *member)
+static inline u32 __btf_member_bitfield_size(const struct btf_type *struct_type,
+					     const struct btf_member *member)
 {
 	return btf_type_kflag(struct_type) ? BTF_MEMBER_BITFIELD_SIZE(member->offset)
 					   : 0;
 }
 
+static inline struct btf_member *btf_members(const struct btf_type *t)
+{
+	return (struct btf_member *)(t + 1);
+}
+
+static inline u32 btf_member_bit_offset(const struct btf_type *t, u32 member_idx)
+{
+	const struct btf_member *m = btf_members(t) + member_idx;
+
+	return __btf_member_bit_offset(t, m);
+}
+
+static inline u32 btf_member_bitfield_size(const struct btf_type *t, u32 member_idx)
+{
+	const struct btf_member *m = btf_members(t) + member_idx;
+
+	return __btf_member_bitfield_size(t, m);
+}
+
 static inline const struct btf_member *btf_type_member(const struct btf_type *t)
 {
 	return (const struct btf_member *)(t + 1);
+}
+
+static inline struct btf_array *btf_array(const struct btf_type *t)
+{
+	return (struct btf_array *)(t + 1);
+}
+
+static inline struct btf_enum *btf_enum(const struct btf_type *t)
+{
+	return (struct btf_enum *)(t + 1);
 }
 
 static inline const struct btf_var_secinfo *btf_type_var_secinfo(

include/uapi/linux/bpf.h

···
 			/* or valid module BTF object fd or 0 to attach to vmlinux */
 			__u32		attach_btf_obj_fd;
 		};
-		__u32		:32;		/* pad */
+		__u32		core_relo_cnt;	/* number of bpf_core_relo */
 		__aligned_u64	fd_array;	/* array of FDs */
+		__aligned_u64	core_relos;
+		__u32		core_relo_rec_size; /* sizeof(struct bpf_core_relo) */
 	};
 
 	struct { /* anonymous struct used by BPF_OBJ_* commands */
···
  * 		if the maximum number of tail calls has been reached for this
  * 		chain of programs. This limit is defined in the kernel by the
  * 		macro **MAX_TAIL_CALL_CNT** (not accessible to user space),
- * 		which is currently set to 32.
+ *		which is currently set to 33.
  * 	Return
  * 		0 on success, or a negative error in case of failure.
  *
···
  *		**-ENOENT** if *task->mm* is NULL, or no vma contains *addr*.
  *		**-EBUSY** if failed to try lock mmap_lock.
  *		**-EINVAL** for invalid **flags**.
+ *
+ * long bpf_loop(u32 nr_loops, void *callback_fn, void *callback_ctx, u64 flags)
+ *	Description
+ *		For **nr_loops**, call **callback_fn** function
+ *		with **callback_ctx** as the context parameter.
+ *		The **callback_fn** should be a static function and
+ *		the **callback_ctx** should be a pointer to the stack.
+ *		The **flags** is used to control certain aspects of the helper.
+ *		Currently, the **flags** must be 0. Currently, nr_loops is
+ *		limited to 1 << 23 (~8 million) loops.
+ *
+ *		long (\*callback_fn)(u32 index, void \*ctx);
+ *
+ *		where **index** is the current index in the loop. The index
+ *		is zero-indexed.
+ *
+ *		If **callback_fn** returns 0, the helper will continue to the next
+ *		loop. If return value is 1, the helper will skip the rest of
+ *		the loops and return. Other return values are not used now,
+ *		and will be rejected by the verifier.
+ *
+ *	Return
+ *		The number of loops performed, **-EINVAL** for invalid **flags**,
+ *		**-E2BIG** if **nr_loops** exceeds the maximum number of loops.
  */
 #define __BPF_FUNC_MAPPER(FN)		\
 	FN(unspec),			\
···
 	FN(skc_to_unix_sock),		\
 	FN(kallsyms_lookup_name),	\
 	FN(find_vma),			\
+	FN(loop),			\
 	/* */
 
 /* integer value in 'imm' field of BPF_CALL instruction selects which helper
···
 	BTF_F_NONAME	=	(1ULL << 1),
 	BTF_F_PTR_RAW	=	(1ULL << 2),
 	BTF_F_ZERO	=	(1ULL << 3),
+};
+
+/* bpf_core_relo_kind encodes which aspect of captured field/type/enum value
+ * has to be adjusted by relocations. It is emitted by llvm and passed to
+ * libbpf and later to the kernel.
+ */
+enum bpf_core_relo_kind {
+	BPF_CORE_FIELD_BYTE_OFFSET = 0,      /* field byte offset */
+	BPF_CORE_FIELD_BYTE_SIZE = 1,        /* field size in bytes */
+	BPF_CORE_FIELD_EXISTS = 2,           /* field existence in target kernel */
+	BPF_CORE_FIELD_SIGNED = 3,           /* field signedness (0 - unsigned, 1 - signed) */
+	BPF_CORE_FIELD_LSHIFT_U64 = 4,       /* bitfield-specific left bitshift */
+	BPF_CORE_FIELD_RSHIFT_U64 = 5,       /* bitfield-specific right bitshift */
+	BPF_CORE_TYPE_ID_LOCAL = 6,          /* type ID in local BPF object */
+	BPF_CORE_TYPE_ID_TARGET = 7,         /* type ID in target kernel */
+	BPF_CORE_TYPE_EXISTS = 8,            /* type existence in target kernel */
+	BPF_CORE_TYPE_SIZE = 9,              /* type size in bytes */
+	BPF_CORE_ENUMVAL_EXISTS = 10,        /* enum value existence in target kernel */
+	BPF_CORE_ENUMVAL_VALUE = 11,         /* enum value integer value */
+};
+
+/*
+ * "struct bpf_core_relo" is used to pass relocation data form LLVM to libbpf
+ * and from libbpf to the kernel.
+ *
+ * CO-RE relocation captures the following data:
+ * - insn_off - instruction offset (in bytes) within a BPF program that needs
+ *   its insn->imm field to be relocated with actual field info;
+ * - type_id - BTF type ID of the "root" (containing) entity of a relocatable
+ *   type or field;
+ * - access_str_off - offset into corresponding .BTF string section. String
+ *   interpretation depends on specific relocation kind:
+ *     - for field-based relocations, string encodes an accessed field using
+ *       a sequence of field and array indices, separated by colon (:). It's
+ *       conceptually very close to LLVM's getelementptr ([0]) instruction's
+ *       arguments for identifying offset to a field.
+ *     - for type-based relocations, strings is expected to be just "0";
+ *     - for enum value-based relocations, string contains an index of enum
+ *       value within its enum type;
+ * - kind - one of enum bpf_core_relo_kind;
+ *
+ * Example:
+ *   struct sample {
+ *       int a;
+ *       struct {
+ *           int b[10];
+ *       };
+ *   };
+ *
+ *   struct sample *s = ...;
+ *   int *x = &s->a;     // encoded as "0:0" (a is field #0)
+ *   int *y = &s->b[5];  // encoded as "0:1:0:5" (anon struct is field #1,
+ *                       // b is field #0 inside anon struct, accessing elem #5)
+ *   int *z = &s[10]->b; // encoded as "10:1" (ptr is used as an array)
+ *
+ * type_id for all relocs in this example will capture BTF type id of
+ * `struct sample`.
+ *
+ * Such relocation is emitted when using __builtin_preserve_access_index()
+ * Clang built-in, passing expression that captures field address, e.g.:
+ *
+ * bpf_probe_read(&dst, sizeof(dst),
+ *		  __builtin_preserve_access_index(&src->a.b.c));
+ *
+ * In this case Clang will emit field relocation recording necessary data to
+ * be able to find offset of embedded `a.b.c` field within `src` struct.
+ *
+ * [0] https://llvm.org/docs/LangRef.html#getelementptr-instruction
+ */
+struct bpf_core_relo {
+	__u32 insn_off;
+	__u32 type_id;
+	__u32 access_str_off;
+	enum bpf_core_relo_kind kind;
 };
 
 #endif /* _UAPI__LINUX_BPF_H__ */

kernel/bpf/Makefile

···
 obj-${CONFIG_BPF_LSM} += bpf_lsm.o
 endif
 obj-$(CONFIG_BPF_PRELOAD) += preload/
+
+obj-$(CONFIG_BPF_SYSCALL) += relo_core.o
+$(obj)/relo_core.o: $(srctree)/tools/lib/bpf/relo_core.c FORCE
+	$(call if_changed_rule,cc_o_c)

kernel/bpf/bpf_iter.c

···
 	.arg3_type	= ARG_PTR_TO_STACK_OR_NULL,
 	.arg4_type	= ARG_ANYTHING,
 };
+
+/* maximum number of loops */
+#define MAX_LOOPS	BIT(23)
+
+BPF_CALL_4(bpf_loop, u32, nr_loops, void *, callback_fn, void *, callback_ctx,
+	   u64, flags)
+{
+	bpf_callback_t callback = (bpf_callback_t)callback_fn;
+	u64 ret;
+	u32 i;
+
+	if (flags)
+		return -EINVAL;
+	if (nr_loops > MAX_LOOPS)
+		return -E2BIG;
+
+	for (i = 0; i < nr_loops; i++) {
+		ret = callback((u64)i, (u64)(long)callback_ctx, 0, 0, 0);
+		/* return value: 0 - continue, 1 - stop and return */
+		if (ret)
+			return i + 1;
+	}
+
+	return i;
+}
+
+const struct bpf_func_proto bpf_loop_proto = {
+	.func		= bpf_loop,
+	.gpl_only	= false,
+	.ret_type	= RET_INTEGER,
+	.arg1_type	= ARG_ANYTHING,
+	.arg2_type	= ARG_PTR_TO_FUNC,
+	.arg3_type	= ARG_PTR_TO_STACK_OR_NULL,
+	.arg4_type	= ARG_ANYTHING,
+};

kernel/bpf/bpf_struct_ops.c

···
 				break;
 			}
 
-			if (btf_member_bitfield_size(t, member)) {
+			if (__btf_member_bitfield_size(t, member)) {
 				pr_warn("bit field member %s in struct %s is not supported\n",
 					mname, st_ops->name);
 				break;
···
 	const struct btf_type *mtype;
 
 	for_each_member(i, t, member) {
-		moff = btf_member_bit_offset(t, member) / 8;
+		moff = __btf_member_bit_offset(t, member) / 8;
 		if (moff > prev_mend &&
 		    memchr_inv(data + prev_mend, 0, moff - prev_mend))
 			return -EINVAL;
···
 		struct bpf_prog *prog;
 		u32 moff;
 
-		moff = btf_member_bit_offset(t, member) / 8;
+		moff = __btf_member_bit_offset(t, member) / 8;
 		ptype = btf_type_resolve_ptr(btf_vmlinux, member->type, NULL);
 		if (ptype == module_type) {
 			if (*(void **)(udata + moff))

kernel/bpf/btf.c

···
 #include <linux/kobject.h>
 #include <linux/sysfs.h>
 #include <net/sock.h>
+#include "../tools/lib/bpf/relo_core.h"
 
 /* BTF (BPF Type Format) is the meta data format which describes
  * the data types of BPF program/map.  Hence, it basically focus
···
 	const char *ptr_suffix = &ptr_suffixes[strlen(ptr_suffixes)];
 	const char *name = NULL, *prefix = "", *parens = "";
 	const struct btf_member *m = show->state.member;
-	const struct btf_type *t = show->state.type;
+	const struct btf_type *t;
 	const struct btf_array *array;
 	u32 id = show->state.type_id;
 	const char *member = NULL;
···
 			return -EINVAL;
 		}
 
-		offset = btf_member_bit_offset(t, member);
+		offset = __btf_member_bit_offset(t, member);
 		if (is_union && offset) {
 			btf_verifier_log_member(env, t, member,
 						"Invalid member bits_offset");
···
 		if (off != -ENOENT)
 			/* only one such field is allowed */
 			return -E2BIG;
-		off = btf_member_bit_offset(t, member);
+		off = __btf_member_bit_offset(t, member);
 		if (off % 8)
 			/* valid C code cannot generate such BTF */
 			return -EINVAL;
···
 
 		btf_show_start_member(show, member);
 
-		member_offset = btf_member_bit_offset(t, member);
-		bitfield_size = btf_member_bitfield_size(t, member);
+		member_offset = __btf_member_bit_offset(t, member);
+		bitfield_size = __btf_member_bitfield_size(t, member);
 		bytes_offset = BITS_ROUNDDOWN_BYTES(member_offset);
 		bits8_offset = BITS_PER_BYTE_MASKED(member_offset);
 		if (bitfield_size) {
···
 		log->len_total = log_size;
 
 		/* log attributes have to be sane */
-		if (log->len_total < 128 || log->len_total > UINT_MAX >> 8 ||
-		    !log->level || !log->ubuf) {
+		if (!bpf_verifier_log_attr_valid(log)) {
 			err = -EINVAL;
 			goto errout;
 		}
···
 		if (array_elem->nelems != 0)
 			goto error;
 
-		moff = btf_member_bit_offset(t, member) / 8;
+		moff = __btf_member_bit_offset(t, member) / 8;
 		if (off < moff)
 			goto error;
 
···
 
 	for_each_member(i, t, member) {
 		/* offset of the field in bytes */
-		moff = btf_member_bit_offset(t, member) / 8;
+		moff = __btf_member_bit_offset(t, member) / 8;
 		if (off + size <= moff)
 			/* won't find anything, field is already too far */
 			break;
 
-		if (btf_member_bitfield_size(t, member)) {
-			u32 end_bit = btf_member_bit_offset(t, member) +
-				btf_member_bitfield_size(t, member);
+		if (__btf_member_bitfield_size(t, member)) {
+			u32 end_bit = __btf_member_bit_offset(t, member) +
+				__btf_member_bitfield_size(t, member);
 
 			/* off <= moff instead of off == moff because clang
 			 * does not generate a BTF member for anonymous
···
 	return len;
 }
 
+static void purge_cand_cache(struct btf *btf);
+
 static int btf_module_notify(struct notifier_block *nb, unsigned long op,
 			     void *module)
 {
···
 			goto out;
 		}
 
+		purge_cand_cache(NULL);
 		mutex_lock(&btf_module_mutex);
 		btf_mod->module = module;
 		btf_mod->btf = btf;
···
 			list_del(&btf_mod->list);
 			if (btf_mod->sysfs_attr)
 				sysfs_remove_bin_file(btf_kobj, btf_mod->sysfs_attr);
+			purge_cand_cache(btf_mod->btf);
 			btf_put(btf_mod->btf);
 			kfree(btf_mod->sysfs_attr);
 			kfree(btf_mod);
···
 DEFINE_KFUNC_BTF_ID_LIST(prog_test_kfunc_list);
 
 #endif
+
+int bpf_core_types_are_compat(const struct btf *local_btf, __u32 local_id,
+			      const struct btf *targ_btf, __u32 targ_id)
+{
+	return -EOPNOTSUPP;
+}
+
+static bool bpf_core_is_flavor_sep(const char *s)
+{
+	/* check X___Y name pattern, where X and Y are not underscores */
+	return s[0] != '_' &&				      /* X */
+	       s[1] == '_' && s[2] == '_' && s[3] == '_' &&   /* ___ */
+	       s[4] != '_';				      /* Y */
+}
+
+size_t bpf_core_essential_name_len(const char *name)
+{
+	size_t n = strlen(name);
+	int i;
+
+	for (i = n - 5; i >= 0; i--) {
+		if (bpf_core_is_flavor_sep(name + i))
+			return i + 1;
+	}
+	return n;
+}
+
+struct bpf_cand_cache {
+	const char *name;
+	u32 name_len;
+	u16 kind;
+	u16 cnt;
+	struct {
+		const struct btf *btf;
+		u32 id;
+	} cands[];
+};
+
+static void bpf_free_cands(struct bpf_cand_cache *cands)
+{
+	if (!cands->cnt)
+		/* empty candidate array was allocated on stack */
+		return;
+	kfree(cands);
+}
+
+static void bpf_free_cands_from_cache(struct bpf_cand_cache *cands)
+{
+	kfree(cands->name);
+	kfree(cands);
+}
+
+#define VMLINUX_CAND_CACHE_SIZE 31
+static struct bpf_cand_cache *vmlinux_cand_cache[VMLINUX_CAND_CACHE_SIZE];
+
+#define MODULE_CAND_CACHE_SIZE 31
+static struct bpf_cand_cache *module_cand_cache[MODULE_CAND_CACHE_SIZE];
+
+static DEFINE_MUTEX(cand_cache_mutex);
+
+static void __print_cand_cache(struct bpf_verifier_log *log,
+			       struct bpf_cand_cache **cache,
+			       int cache_size)
+{
+	struct bpf_cand_cache *cc;
+	int i, j;
+
+	for (i = 0; i < cache_size; i++) {
+		cc = cache[i];
+		if (!cc)
+			continue;
+		bpf_log(log, "[%d]%s(", i, cc->name);
+		for (j = 0; j < cc->cnt; j++) {
+			bpf_log(log, "%d", cc->cands[j].id);
+			if (j < cc->cnt - 1)
+				bpf_log(log, " ");
+		}
+		bpf_log(log, "), ");
+	}
+}
+
+static void print_cand_cache(struct bpf_verifier_log *log)
+{
+	mutex_lock(&cand_cache_mutex);
+	bpf_log(log, "vmlinux_cand_cache:");
+	__print_cand_cache(log, vmlinux_cand_cache, VMLINUX_CAND_CACHE_SIZE);
+	bpf_log(log, "\nmodule_cand_cache:");
+	__print_cand_cache(log, module_cand_cache, MODULE_CAND_CACHE_SIZE);
+	bpf_log(log, "\n");
+	mutex_unlock(&cand_cache_mutex);
+}
+
+static u32 hash_cands(struct bpf_cand_cache *cands)
+{
+	return jhash(cands->name, cands->name_len, 0);
+}
+
+static struct bpf_cand_cache *check_cand_cache(struct bpf_cand_cache *cands,
+					       struct bpf_cand_cache **cache,
+					       int cache_size)
+{
+	struct bpf_cand_cache *cc = cache[hash_cands(cands) % cache_size];
+
+	if (cc && cc->name_len == cands->name_len &&
+	    !strncmp(cc->name, cands->name, cands->name_len))
+		return cc;
+	return NULL;
+}
+
+static size_t sizeof_cands(int cnt)
+{
+	return offsetof(struct bpf_cand_cache, cands[cnt]);
+}
+
+static struct bpf_cand_cache *populate_cand_cache(struct bpf_cand_cache *cands,
+						  struct bpf_cand_cache **cache,
+						  int cache_size)
+{
+	struct bpf_cand_cache **cc = &cache[hash_cands(cands) % cache_size], *new_cands;
+
+	if (*cc) {
+		bpf_free_cands_from_cache(*cc);
+		*cc = NULL;
+	}
+	new_cands = kmalloc(sizeof_cands(cands->cnt), GFP_KERNEL);
+	if (!new_cands) {
+		bpf_free_cands(cands);
+		return ERR_PTR(-ENOMEM);
+	}
+	memcpy(new_cands, cands, sizeof_cands(cands->cnt));
+	/* strdup the name, since it will stay in cache.
+	 * the cands->name points to strings in prog's BTF and the prog can be unloaded.
+	 */
+	new_cands->name = kmemdup_nul(cands->name, cands->name_len, GFP_KERNEL);
+	bpf_free_cands(cands);
+	if (!new_cands->name) {
+		kfree(new_cands);
+		return ERR_PTR(-ENOMEM);
+	}
+	*cc = new_cands;
+	return new_cands;
+}
+
+#ifdef CONFIG_DEBUG_INFO_BTF_MODULES
+static void __purge_cand_cache(struct btf *btf, struct bpf_cand_cache **cache,
+			       int cache_size)
+{
+	struct bpf_cand_cache *cc;
+	int i, j;
+
+	for (i = 0; i < cache_size; i++) {
+		cc = cache[i];
+		if (!cc)
+			continue;
+		if (!btf) {
+			/* when new module is loaded purge all of module_cand_cache,
+			 * since new module might have candidates with the name
+			 * that matches cached cands.
+			 */
+			bpf_free_cands_from_cache(cc);
+			cache[i] = NULL;
+			continue;
+		}
+		/* when module is unloaded purge cache entries
+		 * that match module's btf
+		 */
+		for (j = 0; j < cc->cnt; j++)
+			if (cc->cands[j].btf == btf) {
+				bpf_free_cands_from_cache(cc);
+				cache[i] = NULL;
+				break;
+			}
+	}
+
+}
+
+static void purge_cand_cache(struct btf *btf)
+{
+	mutex_lock(&cand_cache_mutex);
+	__purge_cand_cache(btf, module_cand_cache, MODULE_CAND_CACHE_SIZE);
+	mutex_unlock(&cand_cache_mutex);
+}
+#endif
+
+static struct bpf_cand_cache *
+bpf_core_add_cands(struct bpf_cand_cache *cands, const struct btf *targ_btf,
+		   int targ_start_id)
+{
+	struct bpf_cand_cache *new_cands;
+	const struct btf_type *t;
+	const char *targ_name;
+	size_t targ_essent_len;
+	int n, i;
+
+	n = btf_nr_types(targ_btf);
+	for (i = targ_start_id; i < n; i++) {
+		t = btf_type_by_id(targ_btf, i);
+		if (btf_kind(t) != cands->kind)
+			continue;
+
+		targ_name = btf_name_by_offset(targ_btf, t->name_off);
+		if (!targ_name)
+			continue;
+
+		/* the resched point is before strncmp to make sure that search
+		 * for non-existing name will have a chance to schedule().
+		 */
+		cond_resched();
+
+		if (strncmp(cands->name, targ_name, cands->name_len) != 0)
+			continue;
+
+		targ_essent_len = bpf_core_essential_name_len(targ_name);
+		if (targ_essent_len != cands->name_len)
+			continue;
+
+		/* most of the time there is only one candidate for a given kind+name pair */
+		new_cands = kmalloc(sizeof_cands(cands->cnt + 1), GFP_KERNEL);
+		if (!new_cands) {
+			bpf_free_cands(cands);
+			return ERR_PTR(-ENOMEM);
+		}
+
+		memcpy(new_cands, cands, sizeof_cands(cands->cnt));
+		bpf_free_cands(cands);
+		cands = new_cands;
+		cands->cands[cands->cnt].btf = targ_btf;
+		cands->cands[cands->cnt].id = i;
+		cands->cnt++;
+	}
+	return cands;
+}
+
+static struct bpf_cand_cache *
+bpf_core_find_cands(struct bpf_core_ctx *ctx, u32 local_type_id)
+{
+	struct bpf_cand_cache *cands, *cc, local_cand = {};
+	const struct btf *local_btf = ctx->btf;
+	const struct btf_type *local_type;
+	const struct btf *main_btf;
+	size_t local_essent_len;
+	struct btf *mod_btf;
+	const char *name;
+	int id;
+
+	main_btf = bpf_get_btf_vmlinux();
+	if (IS_ERR(main_btf))
+		return (void *)main_btf;
+
+	local_type = btf_type_by_id(local_btf, local_type_id);
+	if (!local_type)
+		return ERR_PTR(-EINVAL);
+
+	name = btf_name_by_offset(local_btf, local_type->name_off);
+	if (str_is_empty(name))
+		return ERR_PTR(-EINVAL);
+	local_essent_len = bpf_core_essential_name_len(name);
+
+	cands = &local_cand;
+	cands->name = name;
+	cands->kind = btf_kind(local_type);
+	cands->name_len = local_essent_len;
+
+	cc = check_cand_cache(cands, vmlinux_cand_cache, VMLINUX_CAND_CACHE_SIZE);
+	/* cands is a pointer to stack here */
+	if (cc) {
+		if (cc->cnt)
+			return cc;
+		goto check_modules;
+	}
+
+	/* Attempt to find target candidates in vmlinux BTF first */
+	cands = bpf_core_add_cands(cands, main_btf, 1);
+	if (IS_ERR(cands))
+		return cands;
+
+	/* cands is a pointer to kmalloced memory here if cands->cnt > 0 */
+
+	/* populate cache even when cands->cnt == 0 */
+	cc = populate_cand_cache(cands, vmlinux_cand_cache, VMLINUX_CAND_CACHE_SIZE);
+	if (IS_ERR(cc))
+		return cc;
+
+	/* if vmlinux BTF has any candidate, don't go for module BTFs */
+	if (cc->cnt)
+		return cc;
+
+check_modules:
+	/* cands is a pointer to stack here and cands->cnt == 0 */
+	cc = check_cand_cache(cands, module_cand_cache, MODULE_CAND_CACHE_SIZE);
+	if (cc)
+		/* if cache has it return it even if cc->cnt == 0 */
+		return cc;
+
+	/* If candidate is not found in vmlinux's BTF then search in module's BTFs */
+	spin_lock_bh(&btf_idr_lock);
+	idr_for_each_entry(&btf_idr, mod_btf, id) {
+		if (!btf_is_module(mod_btf))
+			continue;
+		/* linear search could be slow hence unlock/lock
+		 * the IDR to avoiding holding it for too long
+		 */
+		btf_get(mod_btf);
+		spin_unlock_bh(&btf_idr_lock);
+		cands = bpf_core_add_cands(cands, mod_btf, btf_nr_types(main_btf));
+		if (IS_ERR(cands)) {
+			btf_put(mod_btf);
+			return cands;
+		}
+		spin_lock_bh(&btf_idr_lock);
+		btf_put(mod_btf);
+	}
+	spin_unlock_bh(&btf_idr_lock);
+	/* cands is a pointer to kmalloced memory here if cands->cnt > 0
+	 * or pointer to stack if cands->cnd == 0.
+	 * Copy it into the cache even when cands->cnt == 0 and
+	 * return the result.
+	 */
+	return populate_cand_cache(cands, module_cand_cache, MODULE_CAND_CACHE_SIZE);
+}
+
+int bpf_core_apply(struct bpf_core_ctx *ctx, const struct bpf_core_relo *relo,
+		   int relo_idx, void *insn)
+{
+	bool need_cands = relo->kind != BPF_CORE_TYPE_ID_LOCAL;
+	struct bpf_core_cand_list cands = {};
+	struct bpf_core_spec *specs;
+	int err;
+
+	/* ~4k of temp memory necessary to convert LLVM spec like "0:1:0:5"
+	 * into arrays of btf_ids of struct fields and array indices.
+	 */
+	specs = kcalloc(3, sizeof(*specs), GFP_KERNEL);
+	if (!specs)
+		return -ENOMEM;
+
+	if (need_cands) {
+		struct bpf_cand_cache *cc;
+		int i;
+
+		mutex_lock(&cand_cache_mutex);
+		cc = bpf_core_find_cands(ctx, relo->type_id);
+		if (IS_ERR(cc)) {
+			bpf_log(ctx->log, "target candidate search failed for %d\n",
+				relo->type_id);
+			err = PTR_ERR(cc);
+			goto out;
+		}
+		if (cc->cnt) {
+			cands.cands = kcalloc(cc->cnt, sizeof(*cands.cands), GFP_KERNEL);
+			if (!cands.cands) {
+				err = -ENOMEM;
+				goto out;
+			}
+		}
+		for (i = 0; i < cc->cnt; i++) {
+			bpf_log(ctx->log,
+				"CO-RE relocating %s %s: found target candidate [%d]\n",
+				btf_kind_str[cc->kind], cc->name, cc->cands[i].id);
+			cands.cands[i].btf = cc->cands[i].btf;
+			cands.cands[i].id = cc->cands[i].id;
+		}
+		cands.len = cc->cnt;
+		/* cand_cache_mutex needs to span the cache lookup and
+		 * copy of btf pointer into bpf_core_cand_list,
+		 * since module can be unloaded while bpf_core_apply_relo_insn
+		 * is working with module's btf.
+		 */
+	}
+
+	err = bpf_core_apply_relo_insn((void *)ctx->log, insn, relo->insn_off / 8,
+				       relo, relo_idx, ctx->btf, &cands, specs);
+out:
+	kfree(specs);
+	if (need_cands) {
+		kfree(cands.cands);
+		mutex_unlock(&cand_cache_mutex);
+		if (ctx->log->level & BPF_LOG_LEVEL2)
+			print_cand_cache(ctx->log);
+	}
+	return err;
+}

kernel/bpf/core.c

···
 
 		if (unlikely(index >= array->map.max_entries))
 			goto out;
-		if (unlikely(tail_call_cnt > MAX_TAIL_CALL_CNT))
+
+		if (unlikely(tail_call_cnt >= MAX_TAIL_CALL_CNT))
 			goto out;
 
 		tail_call_cnt++;
···
 
 /**
  *	bpf_prog_select_runtime - select exec runtime for BPF program
- *	@fp: bpf_prog populated with internal BPF program
+ *	@fp: bpf_prog populated with BPF program
  *	@err: pointer to error variable
  *
  * Try to JIT eBPF program, if JIT is not available, use interpreter.
···
 	}
 }
 
-/* Free internal BPF program */
 void bpf_prog_free(struct bpf_prog *fp)
 {
 	struct bpf_prog_aux *aux = fp->aux;

kernel/bpf/helpers.c

···
 		return &bpf_ringbuf_query_proto;
 	case BPF_FUNC_for_each_map_elem:
 		return &bpf_for_each_map_elem_proto;
+	case BPF_FUNC_loop:
+		return &bpf_loop_proto;
 	default:
 		break;
 	}

kernel/bpf/syscall.c

···
 }
 
 /* last field in 'union bpf_attr' used by this command */
-#define	BPF_PROG_LOAD_LAST_FIELD fd_array
+#define	BPF_PROG_LOAD_LAST_FIELD core_relo_rec_size
 
 static int bpf_prog_load(union bpf_attr *attr, bpfptr_t uattr)
 {
···
 	.gpl_only	= false,
 	.ret_type	= RET_INTEGER,
 	.arg1_type	= ARG_PTR_TO_MEM,
-	.arg2_type	= ARG_CONST_SIZE,
+	.arg2_type	= ARG_CONST_SIZE_OR_ZERO,
 	.arg3_type	= ARG_ANYTHING,
 	.arg4_type	= ARG_PTR_TO_LONG,
 };

kernel/bpf/verifier.c

···
 	WARN_ONCE(n >= BPF_VERIFIER_TMP_LOG_SIZE - 1,
 		  "verifier log line truncated - local buffer too short\n");
 
-	n = min(log->len_total - log->len_used - 1, n);
-	log->kbuf[n] = '\0';
-
 	if (log->level == BPF_LOG_KERNEL) {
-		pr_err("BPF:%s\n", log->kbuf);
+		bool newline = n > 0 && log->kbuf[n - 1] == '\n';
+
+		pr_err("BPF: %s%s", log->kbuf, newline ? "" : "\n");
 		return;
 	}
+
+	n = min(log->len_total - log->len_used - 1, n);
+	log->kbuf[n] = '\0';
 	if (!copy_to_user(log->ubuf + log->len_used, log->kbuf, n + 1))
 		log->len_used += n;
 	else
···
 	return 0;
 }
 
+static int set_loop_callback_state(struct bpf_verifier_env *env,
+				   struct bpf_func_state *caller,
+				   struct bpf_func_state *callee,
+				   int insn_idx)
+{
+	/* bpf_loop(u32 nr_loops, void *callback_fn, void *callback_ctx,
+	 *	    u64 flags);
+	 * callback_fn(u32 index, void *callback_ctx);
+	 */
+	callee->regs[BPF_REG_1].type = SCALAR_VALUE;
+	callee->regs[BPF_REG_2] = caller->regs[BPF_REG_3];
+
+	/* unused */
+	__mark_reg_not_init(env, &callee->regs[BPF_REG_3]);
+	__mark_reg_not_init(env, &callee->regs[BPF_REG_4]);
+	__mark_reg_not_init(env, &callee->regs[BPF_REG_5]);
+
+	callee->in_callback_fn = true;
+	return 0;
+}
+
 static int set_timer_callback_state(struct bpf_verifier_env *env,
 				    struct bpf_func_state *caller,
 				    struct bpf_func_state *callee,
···
 			return err;
 	}
 
-	if (func_id == BPF_FUNC_tail_call) {
-		err = check_reference_leak(env);
-		if (err) {
-			verbose(env, "tail_call would lead to reference leak\n");
-			return err;
-		}
-	} else if (is_release_function(func_id)) {
+	if (is_release_function(func_id)) {
 		err = release_reference(env, meta.ref_obj_id);
 		if (err) {
 			verbose(env, "func %s#%d reference has not been acquired before\n",
···
 
 	regs = cur_regs(env);
 
-	/* check that flags argument in get_local_storage(map, flags) is 0,
-	 * this is required because get_local_storage() can't return an error.
-	 */
-	if (func_id == BPF_FUNC_get_local_storage &&
-	    !register_is_null(&regs[BPF_REG_2])) {
-		verbose(env, "get_local_storage() doesn't support non-zero flags\n");
-		return -EINVAL;
-	}
-
-	if (func_id == BPF_FUNC_for_each_map_elem) {
+	switch (func_id) {
+	case BPF_FUNC_tail_call:
+		err = check_reference_leak(env);
+		if (err) {
+			verbose(env, "tail_call would lead to reference leak\n");
+			return err;
+		}
+		break;
+	case BPF_FUNC_get_local_storage:
+		/* check that flags argument in get_local_storage(map, flags) is 0,
+		 * this is required because get_local_storage() can't return an error.
+		 */
+		if (!register_is_null(&regs[BPF_REG_2])) {
+			verbose(env, "get_local_storage() doesn't support non-zero flags\n");
+			return -EINVAL;
+		}
+		break;
+	case BPF_FUNC_for_each_map_elem:
 		err = __check_func_call(env, insn, insn_idx_p, meta.subprogno,
 					set_map_elem_callback_state);
-		if (err < 0)
-			return -EINVAL;
-	}
-
-	if (func_id == BPF_FUNC_timer_set_callback) {
+		break;
+	case BPF_FUNC_timer_set_callback:
 		err = __check_func_call(env, insn, insn_idx_p, meta.subprogno,
 					set_timer_callback_state);
-		if (err < 0)
-			return -EINVAL;
-	}
-
-	if (func_id == BPF_FUNC_find_vma) {
+		break;
+	case BPF_FUNC_find_vma:
 		err = __check_func_call(env, insn, insn_idx_p, meta.subprogno,
 					set_find_vma_callback_state);
-		if (err < 0)
-			return -EINVAL;
+		break;
+	case BPF_FUNC_snprintf:
+		err = check_bpf_snprintf_call(env, regs);
+		break;
+	case BPF_FUNC_loop:
+		err = __check_func_call(env, insn, insn_idx_p, meta.subprogno,
+					set_loop_callback_state);
+		break;
 	}
 
-	if (func_id == BPF_FUNC_snprintf) {
-		err = check_bpf_snprintf_call(env, regs);
-		if (err < 0)
-			return err;
-	}
+	if (err)
+		return err;
 
 	/* reset caller saved regs */
 	for (i = 0; i < CALLER_SAVED_REGS; i++) {
···
 	return err;
 }
 
+#define MIN_CORE_RELO_SIZE	sizeof(struct bpf_core_relo)
+#define MAX_CORE_RELO_SIZE	MAX_FUNCINFO_REC_SIZE
+
+static int check_core_relo(struct bpf_verifier_env *env,
+			   const union bpf_attr *attr,
+			   bpfptr_t uattr)
+{
+	u32 i, nr_core_relo, ncopy, expected_size, rec_size;
+	struct bpf_core_relo core_relo = {};
+	struct bpf_prog *prog = env->prog;
+	const struct btf *btf = prog->aux->btf;
+	struct bpf_core_ctx ctx = {
+		.log = &env->log,
+		.btf = btf,
+	};
+	bpfptr_t u_core_relo;
+	int err;
+
+	nr_core_relo = attr->core_relo_cnt;
+	if (!nr_core_relo)
+		return 0;
+	if (nr_core_relo > INT_MAX / sizeof(struct bpf_core_relo))
+		return -EINVAL;
+
+	rec_size = attr->core_relo_rec_size;
+	if (rec_size < MIN_CORE_RELO_SIZE ||
+	    rec_size > MAX_CORE_RELO_SIZE ||
+	    rec_size % sizeof(u32))
+		return -EINVAL;
+
+	u_core_relo = make_bpfptr(attr->core_relos, uattr.is_kernel);
+	expected_size = sizeof(struct bpf_core_relo);
+	ncopy = min_t(u32, expected_size, rec_size);
+
+	/* Unlike func_info and line_info, copy and apply each CO-RE
+	 * relocation record one at a time.
+	 */
+	for (i = 0; i < nr_core_relo; i++) {
+		/* future proofing when sizeof(bpf_core_relo) changes */
+		err = bpf_check_uarg_tail_zero(u_core_relo, expected_size, rec_size);
+		if (err) {
+			if (err == -E2BIG) {
+				verbose(env, "nonzero tailing record in core_relo");
+				if (copy_to_bpfptr_offset(uattr,
+							  offsetof(union bpf_attr, core_relo_rec_size),
+							  &expected_size, sizeof(expected_size)))
+					err = -EFAULT;
+			}
+			break;
+		}
+
+		if (copy_from_bpfptr(&core_relo, u_core_relo, ncopy)) {
+			err = -EFAULT;
+			break;
+		}
+
+		if (core_relo.insn_off % 8 || core_relo.insn_off / 8 >= prog->len) {
+			verbose(env, "Invalid core_relo[%u].insn_off:%u prog->len:%u\n",
+				i, core_relo.insn_off, prog->len);
+			err = -EINVAL;
+			break;
+		}
+
+		err = bpf_core_apply(&ctx, &core_relo, i,
+				     &prog->insnsi[core_relo.insn_off / 8]);
+		if (err)
+			break;
+		bpfptr_add(&u_core_relo, rec_size);
+	}
+	return err;
+}
+
 static int check_btf_info(struct bpf_verifier_env *env,
 			  const union bpf_attr *attr,
 			  bpfptr_t uattr)
···
 		return err;
 
 	err = check_btf_line(env, attr, uattr);
+	if (err)
+		return err;
+
+	err = check_core_relo(env, attr, uattr);
 	if (err)
 		return err;
 
···
 		log->ubuf = (char __user *) (unsigned long) attr->log_buf;
 		log->len_total = attr->log_size;
 
-		ret = -EINVAL;
 		/* log attributes have to be sane */
-		if (log->len_total < 128 || log->len_total > UINT_MAX >> 2 ||
-		    !log->level || !log->ubuf || log->level & ~BPF_LOG_MASK)
+		if (!bpf_verifier_log_attr_valid(log)) {
+			ret = -EINVAL;
 			goto err_unlock;
+		}
 	}
 
 	if (IS_ERR(btf_vmlinux)) {

kernel/trace/bpf_trace.c

···
 BPF_CALL_4(bpf_read_branch_records, struct bpf_perf_event_data_kern *, ctx,
 	   void *, buf, u32, size, u64, flags)
 {
-#ifndef CONFIG_X86
-	return -ENOENT;
-#else
 	static const u32 br_entry_size = sizeof(struct perf_branch_entry);
 	struct perf_branch_stack *br_stack = ctx->data->br_stack;
 	u32 to_copy;
···
 		return -EINVAL;
 
 	if (unlikely(!br_stack))
-		return -EINVAL;
+		return -ENOENT;
 
 	if (flags & BPF_F_GET_BRANCH_RECORDS_SIZE)
 		return br_stack->nr * br_entry_size;
···
 	memcpy(buf, br_stack->entries, to_copy);
 
 	return to_copy;
-#endif
 }
 
 static const struct bpf_func_proto bpf_read_branch_records_proto = {

lib/test_bpf.c

···
 			BPF_EXIT_INSN(),
 		},
 		.flags = FLAG_NEED_STATE | FLAG_RESULT_IN_STATE,
-		.result = (MAX_TAIL_CALL_CNT + 1 + 1) * MAX_TESTRUNS,
+		.result = (MAX_TAIL_CALL_CNT + 1) * MAX_TESTRUNS,
 	},
 	{
 		"Tail call count preserved across function calls",
···
 		},
 		.stack_depth = 8,
 		.flags = FLAG_NEED_STATE | FLAG_RESULT_IN_STATE,
-		.result = (MAX_TAIL_CALL_CNT + 1 + 1) * MAX_TESTRUNS,
+		.result = (MAX_TAIL_CALL_CNT + 1) * MAX_TESTRUNS,
 	},
 	{
 		"Tail call error path, NULL target",

net/core/filter.c

···
 	int err, new_len, old_len = fp->len;
 	bool seen_ld_abs = false;
 
-	/* We are free to overwrite insns et al right here as it
-	 * won't be used at this point in time anymore internally
-	 * after the migration to the internal BPF instruction
-	 * representation.
+	/* We are free to overwrite insns et al right here as it won't be used at
+	 * this point in time anymore internally after the migration to the eBPF
+	 * instruction representation.
 	 */
 	BUILD_BUG_ON(sizeof(struct sock_filter) !=
 		     sizeof(struct bpf_insn));
···
 	 */
 	bpf_jit_compile(fp);
 
-	/* JIT compiler couldn't process this filter, so do the
-	 * internal BPF translation for the optimized interpreter.
+	/* JIT compiler couldn't process this filter, so do the eBPF translation
+	 * for the optimized interpreter.
 	 */
 	if (!fp->jited)
 		fp = bpf_migrate_filter(fp);

net/ipv4/bpf_tcp_ca.c

···
 	t = bpf_tcp_congestion_ops.type;
 	m = &btf_type_member(t)[midx];
 
-	return btf_member_bit_offset(t, m) / 8;
+	return __btf_member_bit_offset(t, m) / 8;
 }
 
 static const struct bpf_func_proto *
···
 	utcp_ca = (const struct tcp_congestion_ops *)udata;
 	tcp_ca = (struct tcp_congestion_ops *)kdata;
 
-	moff = btf_member_bit_offset(t, member) / 8;
+	moff = __btf_member_bit_offset(t, member) / 8;
 	switch (moff) {
 	case offsetof(struct tcp_congestion_ops, flags):
 		if (utcp_ca->flags & ~TCP_CONG_MASK)
···
 static int bpf_tcp_ca_check_member(const struct btf_type *t,
 				   const struct btf_member *member)
 {
-	if (is_unsupported(btf_member_bit_offset(t, member) / 8))
+	if (is_unsupported(__btf_member_bit_offset(t, member) / 8))
 		return -ENOTSUPP;
 	return 0;
 }

samples/bpf/Makefile

···
 endif
 
 TPROGS_LDLIBS			+= $(LIBBPF) -lelf -lz
+TPROGLDLIBS_xdp_monitor		+= -lm
+TPROGLDLIBS_xdp_redirect	+= -lm
+TPROGLDLIBS_xdp_redirect_cpu	+= -lm
+TPROGLDLIBS_xdp_redirect_map	+= -lm
+TPROGLDLIBS_xdp_redirect_map_multi += -lm
 TPROGLDLIBS_tracex4		+= -lrt
 TPROGLDLIBS_trace_output	+= -lrt
 TPROGLDLIBS_map_perf_test	+= -lrt
···
 $(src)/*.c: verify_target_bpf $(LIBBPF)
 
 libbpf_hdrs: $(LIBBPF)
-$(obj)/$(TRACE_HELPERS): | libbpf_hdrs
+$(obj)/$(TRACE_HELPERS) $(obj)/$(CGROUP_HELPERS) $(obj)/$(XDP_SAMPLE): | libbpf_hdrs
 
 .PHONY: libbpf_hdrs
 
···
 $(obj)/hbm_out_kern.o: $(src)/hbm.h $(src)/hbm_kern.h
 $(obj)/hbm.o: $(src)/hbm.h
 $(obj)/hbm_edt_kern.o: $(src)/hbm.h $(src)/hbm_kern.h
+
+# Override includes for xdp_sample_user.o because $(srctree)/usr/include in
+# TPROGS_CFLAGS causes conflicts
+XDP_SAMPLE_CFLAGS += -Wall -O2 \
+		     -I$(src)/../../tools/include \
+		     -I$(src)/../../tools/include/uapi \
+		     -I$(LIBBPF_INCLUDE) \
+		     -I$(src)/../../tools/testing/selftests/bpf
+
+$(obj)/$(XDP_SAMPLE): TPROGS_CFLAGS = $(XDP_SAMPLE_CFLAGS)
+$(obj)/$(XDP_SAMPLE): $(src)/xdp_sample_user.h $(src)/xdp_sample_shared.h
 
 -include $(BPF_SAMPLES_PATH)/Makefile.target
 

samples/bpf/Makefile.target

···
       cmd_tprog-cobjs	= $(CC) $(tprogc_flags) -c -o $@ $<
 $(tprog-cobjs): $(obj)/%.o: $(src)/%.c FORCE
 	$(call if_changed_dep,tprog-cobjs)
-
-# Override includes for xdp_sample_user.o because $(srctree)/usr/include in
-# TPROGS_CFLAGS causes conflicts
-XDP_SAMPLE_CFLAGS += -Wall -O2 -lm \
-		     -I./tools/include \
-		     -I./tools/include/uapi \
-		     -I./tools/lib \
-		     -I./tools/testing/selftests/bpf
-$(obj)/xdp_sample_user.o: $(src)/xdp_sample_user.c \
-	$(src)/xdp_sample_user.h $(src)/xdp_sample_shared.h
-	$(CC) $(XDP_SAMPLE_CFLAGS) -c -o $@ $<

samples/bpf/cookie_uid_helper_example.c

···
 
 static void maps_create(void)
 {
-	map_fd = bpf_create_map(BPF_MAP_TYPE_HASH, sizeof(uint32_t),
-				sizeof(struct stats), 100, 0);
+	map_fd = bpf_map_create(BPF_MAP_TYPE_HASH, NULL, sizeof(uint32_t),
+				sizeof(struct stats), 100, NULL);
 	if (map_fd < 0)
 		error(1, errno, "map create failed!\n");
 }
···
 				offsetof(struct __sk_buff, len)),
 		BPF_EXIT_INSN(),
 	};
-	prog_fd = bpf_load_program(BPF_PROG_TYPE_SOCKET_FILTER, prog,
-					ARRAY_SIZE(prog), "GPL", 0,
-					log_buf, sizeof(log_buf));
+	LIBBPF_OPTS(bpf_prog_load_opts, opts,
+		.log_buf = log_buf,
+		.log_size = sizeof(log_buf),
+	);
+
+	prog_fd = bpf_prog_load(BPF_PROG_TYPE_SOCKET_FILTER, NULL, "GPL",
+				prog, ARRAY_SIZE(prog), &opts);
 	if (prog_fd < 0)
 		error(1, errno, "failed to load prog\n%s\n", log_buf);
 }

samples/bpf/fds_example.c

···
 	printf("       -h          Display this help.\n");
 }
 
-static int bpf_map_create(void)
-{
-	return bpf_create_map(BPF_MAP_TYPE_ARRAY, sizeof(uint32_t),
-			      sizeof(uint32_t), 1024, 0);
-}
-
 static int bpf_prog_create(const char *object)
 {
 	static struct bpf_insn insns[] = {
···
 	};
 	size_t insns_cnt = sizeof(insns) / sizeof(struct bpf_insn);
 	struct bpf_object *obj;
-	int prog_fd;
+	int err;
 
 	if (object) {
-		assert(!bpf_prog_load(object, BPF_PROG_TYPE_UNSPEC,
-				      &obj, &prog_fd));
-		return prog_fd;
+		obj = bpf_object__open_file(object, NULL);
+		assert(!libbpf_get_error(obj));
+		err = bpf_object__load(obj);
+		assert(!err);
+		return bpf_program__fd(bpf_object__next_program(obj, NULL));
 	} else {
-		return bpf_load_program(BPF_PROG_TYPE_SOCKET_FILTER,
-					insns, insns_cnt, "GPL", 0,
-					bpf_log_buf, BPF_LOG_BUF_SIZE);
+		LIBBPF_OPTS(bpf_prog_load_opts, opts,
+			.log_buf = bpf_log_buf,
+			.log_size = BPF_LOG_BUF_SIZE,
+		);
+
+		return bpf_prog_load(BPF_PROG_TYPE_SOCKET_FILTER, NULL, "GPL",
+				     insns, insns_cnt, &opts);
 	}
 }
 
···
 	int fd, ret;
 
 	if (flags & BPF_F_PIN) {
-		fd = bpf_map_create();
+		fd = bpf_map_create(BPF_MAP_TYPE_ARRAY, NULL, sizeof(uint32_t),
+				    sizeof(uint32_t), 1024, NULL);
 		printf("bpf: map fd:%d (%s)\n", fd, strerror(errno));
 		assert(fd > 0);
 

samples/bpf/lwt_len_hist_kern.c

···
 #include <uapi/linux/in.h>
 #include <bpf/bpf_helpers.h>
 
-# define printk(fmt, ...)						\
-		({							\
-			char ____fmt[] = fmt;				\
-			bpf_trace_printk(____fmt, sizeof(____fmt),	\
-				     ##__VA_ARGS__);			\
-		})
-
 struct bpf_elf_map {
 	__u32 type;
 	__u32 size_key;

samples/bpf/map_perf_test_user.c

···
 		 */
 		int outer_fd = map_fd[array_of_lru_hashs_idx];
 		unsigned int mycpu, mynode;
+		LIBBPF_OPTS(bpf_map_create_opts, opts,
+			.map_flags = BPF_F_NUMA_NODE,
+		);
 
 		assert(cpu < MAX_NR_CPUS);
 
 		ret = syscall(__NR_getcpu, &mycpu, &mynode, NULL);
 		assert(!ret);
 
+		opts.numa_node = mynode;
 		inner_lru_map_fds[cpu] =
-			bpf_create_map_node(BPF_MAP_TYPE_LRU_HASH,
-					    test_map_names[INNER_LRU_HASH_PREALLOC],
-					    sizeof(uint32_t),
-					    sizeof(long),
-					    inner_lru_hash_size, 0,
-					    mynode);
+			bpf_map_create(BPF_MAP_TYPE_LRU_HASH,
+				       test_map_names[INNER_LRU_HASH_PREALLOC],
+				       sizeof(uint32_t),
+				       sizeof(long),
+				       inner_lru_hash_size, &opts);
 		if (inner_lru_map_fds[cpu] == -1) {
 			printf("cannot create BPF_MAP_TYPE_LRU_HASH %s(%d)\n",
 			       strerror(errno), errno);

samples/bpf/sock_example.c

···
 	int sock = -1, map_fd, prog_fd, i, key;
 	long long value = 0, tcp_cnt, udp_cnt, icmp_cnt;
 
-	map_fd = bpf_create_map(BPF_MAP_TYPE_ARRAY, sizeof(key), sizeof(value),
-				256, 0);
+	map_fd = bpf_map_create(BPF_MAP_TYPE_ARRAY, NULL, sizeof(key), sizeof(value),
+				256, NULL);
 	if (map_fd < 0) {
 		printf("failed to create map '%s'\n", strerror(errno));
 		goto cleanup;
···
 		BPF_EXIT_INSN(),
 	};
 	size_t insns_cnt = sizeof(prog) / sizeof(struct bpf_insn);
+	LIBBPF_OPTS(bpf_prog_load_opts, opts,
+		.log_buf = bpf_log_buf,
+		.log_size = BPF_LOG_BUF_SIZE,
+	);
 
-	prog_fd = bpf_load_program(BPF_PROG_TYPE_SOCKET_FILTER, prog, insns_cnt,
-				   "GPL", 0, bpf_log_buf, BPF_LOG_BUF_SIZE);
+	prog_fd = bpf_prog_load(BPF_PROG_TYPE_SOCKET_FILTER, NULL, "GPL",
+				prog, insns_cnt, &opts);
 	if (prog_fd < 0) {
 		printf("failed to load prog '%s'\n", strerror(errno));
 		goto cleanup;

samples/bpf/sockex1_user.c

···
 int main(int ac, char **argv)
 {
 	struct bpf_object *obj;
+	struct bpf_program *prog;
 	int map_fd, prog_fd;
 	char filename[256];
-	int i, sock;
+	int i, sock, err;
 	FILE *f;
 
 	snprintf(filename, sizeof(filename), "%s_kern.o", argv[0]);
 
-	if (bpf_prog_load(filename, BPF_PROG_TYPE_SOCKET_FILTER,
-			  &obj, &prog_fd))
+	obj = bpf_object__open_file(filename, NULL);
+	if (libbpf_get_error(obj))
 		return 1;
 
+	prog = bpf_object__next_program(obj, NULL);
+	bpf_program__set_type(prog, BPF_PROG_TYPE_SOCKET_FILTER);
+
+	err = bpf_object__load(obj);
+	if (err)
+		return 1;
+
+	prog_fd = bpf_program__fd(prog);
 	map_fd = bpf_object__find_map_fd_by_name(obj, "my_map");
 
 	sock = open_raw_sock("lo");

samples/bpf/sockex2_user.c

···
 
 int main(int ac, char **argv)
 {
+	struct bpf_program *prog;
 	struct bpf_object *obj;
 	int map_fd, prog_fd;
 	char filename[256];
-	int i, sock;
+	int i, sock, err;
 	FILE *f;
 
 	snprintf(filename, sizeof(filename), "%s_kern.o", argv[0]);
-
-	if (bpf_prog_load(filename, BPF_PROG_TYPE_SOCKET_FILTER,
-			  &obj, &prog_fd))
+	obj = bpf_object__open_file(filename, NULL);
+	if (libbpf_get_error(obj))
 		return 1;
 
+	prog = bpf_object__next_program(obj, NULL);
+	bpf_program__set_type(prog, BPF_PROG_TYPE_SOCKET_FILTER);
+
+	err = bpf_object__load(obj);
+	if (err)
+		return 1;
+
+	prog_fd = bpf_program__fd(prog);
 	map_fd = bpf_object__find_map_fd_by_name(obj, "hash_map");
 
 	sock = open_raw_sock("lo");

samples/bpf/test_cgrp2_array_pin.c

···
 	}
 
 	if (create_array) {
-		array_fd = bpf_create_map(BPF_MAP_TYPE_CGROUP_ARRAY,
+		array_fd = bpf_map_create(BPF_MAP_TYPE_CGROUP_ARRAY, NULL,
 					  sizeof(uint32_t), sizeof(uint32_t),
-					  1, 0);
+					  1, NULL);
 		if (array_fd < 0) {
 			fprintf(stderr,
 				"bpf_create_map(BPF_MAP_TYPE_CGROUP_ARRAY,...): %s(%d)\n",

samples/bpf/test_cgrp2_attach.c

···
 		BPF_EXIT_INSN(),
 	};
 	size_t insns_cnt = sizeof(prog) / sizeof(struct bpf_insn);
+	LIBBPF_OPTS(bpf_prog_load_opts, opts,
+		.log_buf = bpf_log_buf,
+		.log_size = BPF_LOG_BUF_SIZE,
+	);
 
-	return bpf_load_program(BPF_PROG_TYPE_CGROUP_SKB,
-				prog, insns_cnt, "GPL", 0,
-				bpf_log_buf, BPF_LOG_BUF_SIZE);
+	return bpf_prog_load(BPF_PROG_TYPE_CGROUP_SKB, NULL, "GPL",
+			     prog, insns_cnt, &opts);
 }
 
 static int usage(const char *argv0)
···
 	int prog_fd, map_fd, ret, key;
 	long long pkt_cnt, byte_cnt;
 
-	map_fd = bpf_create_map(BPF_MAP_TYPE_ARRAY,
+	map_fd = bpf_map_create(BPF_MAP_TYPE_ARRAY, NULL,
 				sizeof(key), sizeof(byte_cnt),
-				256, 0);
+				256, NULL);
 	if (map_fd < 0) {
 		printf("Failed to create map: '%s'\n", strerror(errno));
 		return EXIT_FAILURE;

samples/bpf/test_cgrp2_sock.c

···
 		BPF_MOV64_IMM(BPF_REG_2, offsetof(struct bpf_sock, priority)),
 		BPF_STX_MEM(BPF_W, BPF_REG_1, BPF_REG_3, offsetof(struct bpf_sock, priority)),
 	};
+	LIBBPF_OPTS(bpf_prog_load_opts, opts,
+		.log_buf = bpf_log_buf,
+		.log_size = BPF_LOG_BUF_SIZE,
+	);
 
 	struct bpf_insn *prog;
 	size_t insns_cnt;
···
 
 	insns_cnt /= sizeof(struct bpf_insn);
 
-	ret = bpf_load_program(BPF_PROG_TYPE_CGROUP_SOCK, prog, insns_cnt,
-				"GPL", 0, bpf_log_buf, BPF_LOG_BUF_SIZE);
+	ret = bpf_prog_load(BPF_PROG_TYPE_CGROUP_SOCK, NULL, "GPL",
+			    prog, insns_cnt, &opts);
 
 	free(prog);
 

samples/bpf/test_lru_dist.c

···
 static void pfect_lru_init(struct pfect_lru *lru, unsigned int lru_size,
 			   unsigned int nr_possible_elems)
 {
-	lru->map_fd = bpf_create_map(BPF_MAP_TYPE_HASH,
+	lru->map_fd = bpf_map_create(BPF_MAP_TYPE_HASH, NULL,
 				     sizeof(unsigned long long),
 				     sizeof(struct pfect_lru_node *),
-				     nr_possible_elems, 0);
+				     nr_possible_elems, NULL);
 	assert(lru->map_fd != -1);
 
 	lru->free_nodes = malloc(lru_size * sizeof(struct pfect_lru_node));
···
 
 static int create_map(int map_type, int map_flags, unsigned int size)
 {
+	LIBBPF_OPTS(bpf_map_create_opts, opts,
+		.map_flags = map_flags,
+	);
 	int map_fd;
 
-	map_fd = bpf_create_map(map_type, sizeof(unsigned long long),
-				sizeof(unsigned long long), size, map_flags);
+	map_fd = bpf_map_create(map_type, NULL, sizeof(unsigned long long),
+				sizeof(unsigned long long), size, &opts);
 
 	if (map_fd == -1)
 		perror("bpf_create_map");

samples/bpf/trace_output_user.c

···
 
 int main(int argc, char **argv)
 {
-	struct perf_buffer_opts pb_opts = {};
 	struct bpf_link *link = NULL;
 	struct bpf_program *prog;
 	struct perf_buffer *pb;
···
 		goto cleanup;
 	}
 
-	pb_opts.sample_cb = print_bpf_output;
-	pb = perf_buffer__new(map_fd, 8, &pb_opts);
+	pb = perf_buffer__new(map_fd, 8, print_bpf_output, NULL, NULL, NULL);
 	ret = libbpf_get_error(pb);
 	if (ret) {
 		printf("failed to setup perf_buffer: %d\n", ret);

samples/bpf/xdp_redirect_cpu.bpf.c

···
 	void *data     = (void *)(long)ctx->data;
 	struct iphdr *iph = data + nh_off;
 	struct udphdr *udph;
-	u16 dport;
 
 	if (iph + 1 > data_end)
 		return 0;
···
 	if (udph + 1 > data_end)
 		return 0;
 
-	dport = bpf_ntohs(udph->dest);
-	return dport;
+	return bpf_ntohs(udph->dest);
 }
 
 static __always_inline

samples/bpf/xdp_sample_pkts_user.c

···
 
 int main(int argc, char **argv)
 {
-	struct bpf_prog_load_attr prog_load_attr = {
-		.prog_type	= BPF_PROG_TYPE_XDP,
-	};
-	struct perf_buffer_opts pb_opts = {};
 	const char *optstr = "FS";
 	int prog_fd, map_fd, opt;
+	struct bpf_program *prog;
 	struct bpf_object *obj;
 	struct bpf_map *map;
 	char filename[256];
···
 	}
 
 	snprintf(filename, sizeof(filename), "%s_kern.o", argv[0]);
-	prog_load_attr.file = filename;
 
-	if (bpf_prog_load_xattr(&prog_load_attr, &obj, &prog_fd))
+	obj = bpf_object__open_file(filename, NULL);
+	if (libbpf_get_error(obj))
 		return 1;
 
-	if (!prog_fd) {
-		printf("bpf_prog_load_xattr: %s\n", strerror(errno));
+	prog = bpf_object__next_program(obj, NULL);
+	bpf_program__set_type(prog, BPF_PROG_TYPE_XDP);
+
+	err = bpf_object__load(obj);
+	if (err)
 		return 1;
-	}
+
+	prog_fd = bpf_program__fd(prog);
 
 	map = bpf_object__next_map(obj, NULL);
 	if (!map) {
···
 		return 1;
 	}
 
-	pb_opts.sample_cb = print_bpf_output;
-	pb = perf_buffer__new(map_fd, 8, &pb_opts);
+	pb = perf_buffer__new(map_fd, 8, print_bpf_output, NULL, NULL, NULL);
 	err = libbpf_get_error(pb);
 	if (err) {
 		perror("perf_buffer setup failed");

samples/bpf/xdp_sample_user.h

···
 int get_mac_addr(int ifindex, void *mac_addr);
 
 #pragma GCC diagnostic push
+#ifndef __clang__
 #pragma GCC diagnostic ignored "-Wstringop-truncation"
+#endif
 __attribute__((unused))
 static inline char *safe_strncpy(char *dst, const char *src, size_t size)
 {

samples/bpf/xdpsock_ctrl_proc.c

···
 #include <bpf/xsk.h>
 #include "xdpsock.h"
 
+/* libbpf APIs for AF_XDP are deprecated starting from v0.7 */
+#pragma GCC diagnostic ignored "-Wdeprecated-declarations"
+
 static const char *opt_if = "";
 
 static struct option long_options[] = {

samples/bpf/xdpsock_user.c

···
 #include <bpf/bpf.h>
 #include "xdpsock.h"
 
+/* libbpf APIs for AF_XDP are deprecated starting from v0.7 */
+#pragma GCC diagnostic ignored "-Wdeprecated-declarations"
+
 #ifndef SOL_XDP
 #define SOL_XDP 283
 #endif

samples/bpf/xsk_fwd.c

···
 #include <bpf/xsk.h>
 #include <bpf/bpf.h>
 
+/* libbpf APIs for AF_XDP are deprecated starting from v0.7 */
+#pragma GCC diagnostic ignored "-Wdeprecated-declarations"
+
 #define ARRAY_SIZE(x) (sizeof(x) / sizeof((x)[0]))
 
 typedef __u64 u64;

tools/bpf/bpftool/Documentation/Makefile

···
 man8: $(DOC_MAN8)
 
 RST2MAN_DEP := $(shell command -v rst2man 2>/dev/null)
-RST2MAN_OPTS += --verbose
+RST2MAN_OPTS += --verbose --strip-comments
 
 list_pages = $(sort $(basename $(filter-out $(1),$(MAN8_RST))))
 see_also = $(subst " ",, \

tools/bpf/bpftool/Documentation/bpftool-btf.rst

···
+.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+
 ================
 bpftool-btf
 ================
···
 
 :Manual section: 8
 
+.. include:: substitutions.rst
+
 SYNOPSIS
 ========
 
 	**bpftool** [*OPTIONS*] **btf** *COMMAND*
 
-	*OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | {**-d** | **--debug** } |
-	{ **-B** | **--base-btf** } }
+	*OPTIONS* := { |COMMON_OPTIONS| | { **-B** | **--base-btf** } }
 
 	*COMMANDS* := { **dump** | **help** }
 

tools/bpf/bpftool/Documentation/bpftool-cgroup.rst

···
+.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+
 ================
 bpftool-cgroup
 ================
···
 
 :Manual section: 8
 
+.. include:: substitutions.rst
+
 SYNOPSIS
 ========
 
 	**bpftool** [*OPTIONS*] **cgroup** *COMMAND*
 
-	*OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } |
-	{ **-f** | **--bpffs** } }
+	*OPTIONS* := { |COMMON_OPTIONS| | { **-f** | **--bpffs** } }
 
 	*COMMANDS* :=
 	{ **show** | **list** | **tree** | **attach** | **detach** | **help** }

tools/bpf/bpftool/Documentation/bpftool-feature.rst

···
+.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+
 ===============
 bpftool-feature
 ===============
···
 
 :Manual section: 8
 
+.. include:: substitutions.rst
+
 SYNOPSIS
 ========
 
 	**bpftool** [*OPTIONS*] **feature** *COMMAND*
 
-	*OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } }
+	*OPTIONS* := { |COMMON_OPTIONS| }
 
 	*COMMANDS* := { **probe** | **help** }
 

tools/bpf/bpftool/Documentation/bpftool-gen.rst

···
+.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+
 ================
 bpftool-gen
 ================
···
 
 :Manual section: 8
 
+.. include:: substitutions.rst
+
 SYNOPSIS
 ========
 
 	**bpftool** [*OPTIONS*] **gen** *COMMAND*
 
-	*OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } |
-	{ **-L** | **--use-loader** } }
+	*OPTIONS* := { |COMMON_OPTIONS| | { **-L** | **--use-loader** } }
 
 	*COMMAND* := { **object** | **skeleton** | **help** }
 

tools/bpf/bpftool/Documentation/bpftool-iter.rst

···
+.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+
 ============
 bpftool-iter
 ============
···
 
 :Manual section: 8
 
+.. include:: substitutions.rst
+
 SYNOPSIS
 ========
 
 	**bpftool** [*OPTIONS*] **iter** *COMMAND*
 
-	*OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } }
+	*OPTIONS* := { |COMMON_OPTIONS| }
 
 	*COMMANDS* := { **pin** | **help** }
 

tools/bpf/bpftool/Documentation/bpftool-link.rst

···
+.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+
 ================
 bpftool-link
 ================
···
 
 :Manual section: 8
 
+.. include:: substitutions.rst
+
 SYNOPSIS
 ========
 
 	**bpftool** [*OPTIONS*] **link** *COMMAND*
 
-	*OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } |
-	{ **-f** | **--bpffs** } | { **-n** | **--nomount** } }
+	*OPTIONS* := { |COMMON_OPTIONS| | { **-f** | **--bpffs** } | { **-n** | **--nomount** } }
 
 	*COMMANDS* := { **show** | **list** | **pin** | **help** }
 

tools/bpf/bpftool/Documentation/bpftool-map.rst

···
+.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+
 ================
 bpftool-map
 ================
···
 
 :Manual section: 8
 
+.. include:: substitutions.rst
+
 SYNOPSIS
 ========
 
 	**bpftool** [*OPTIONS*] **map** *COMMAND*
 
-	*OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } |
-	{ **-f** | **--bpffs** } | { **-n** | **--nomount** } }
+	*OPTIONS* := { |COMMON_OPTIONS| | { **-f** | **--bpffs** } | { **-n** | **--nomount** } }
 
 	*COMMANDS* :=
 	{ **show** | **list** | **create** | **dump** | **update** | **lookup** | **getnext** |

tools/bpf/bpftool/Documentation/bpftool-net.rst

···
+.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+
 ================
 bpftool-net
 ================
···
 
 :Manual section: 8
 
+.. include:: substitutions.rst
+
 SYNOPSIS
 ========
 
 	**bpftool** [*OPTIONS*] **net** *COMMAND*
 
-	*OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } }
+	*OPTIONS* := { |COMMON_OPTIONS| }
 
 	*COMMANDS* :=
 	{ **show** | **list** | **attach** | **detach** | **help** }

tools/bpf/bpftool/Documentation/bpftool-perf.rst

···
+.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+
 ================
 bpftool-perf
 ================
···
 
 :Manual section: 8
 
+.. include:: substitutions.rst
+
 SYNOPSIS
 ========
 
 	**bpftool** [*OPTIONS*] **perf** *COMMAND*
 
-	*OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } }
+	*OPTIONS* := { |COMMON_OPTIONS| }
 
 	*COMMANDS* :=
 	{ **show** | **list** | **help** }

tools/bpf/bpftool/Documentation/bpftool-prog.rst

···
+.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+
 ================
 bpftool-prog
 ================
···
 
 :Manual section: 8
 
+.. include:: substitutions.rst
+
 SYNOPSIS
 ========
 
 	**bpftool** [*OPTIONS*] **prog** *COMMAND*
 
-	*OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } |
+	*OPTIONS* := { |COMMON_OPTIONS| |
 	{ **-f** | **--bpffs** } | { **-m** | **--mapcompat** } | { **-n** | **--nomount** } |
 	{ **-L** | **--use-loader** } }
 

tools/bpf/bpftool/Documentation/bpftool-struct_ops.rst

···
+.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+
 ==================
 bpftool-struct_ops
 ==================
···
 
 :Manual section: 8
 
+.. include:: substitutions.rst
+
 SYNOPSIS
 ========
 
 	**bpftool** [*OPTIONS*] **struct_ops** *COMMAND*
 
-	*OPTIONS* := { { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } }
+	*OPTIONS* := { |COMMON_OPTIONS| }
 
 	*COMMANDS* :=
 	{ **show** | **list** | **dump** | **register** | **unregister** | **help** }

tools/bpf/bpftool/Documentation/bpftool.rst

···
+.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+
 ================
 BPFTOOL
 ================
···
 -------------------------------------------------------------------------------
 
 :Manual section: 8
+
+.. include:: substitutions.rst
 
 SYNOPSIS
 ========
···
 
 	*OBJECT* := { **map** | **program** | **cgroup** | **perf** | **net** | **feature** }
 
-	*OPTIONS* := { { **-V** | **--version** } |
-	{ **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } }
+	*OPTIONS* := { { **-V** | **--version** } | |COMMON_OPTIONS| }
 
 	*MAP-COMMANDS* :=
 	{ **show** | **list** | **create** | **dump** | **update** | **lookup** | **getnext** |

tools/bpf/bpftool/Documentation/common_options.rst

···
+.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+
 -h, --help
 	  Print short help message (similar to **bpftool help**).
 

tools/bpf/bpftool/Documentation/substitutions.rst

···
+.. SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+
+.. |COMMON_OPTIONS| replace:: { **-j** | **--json** } [{ **-p** | **--pretty** }] | { **-d** | **--debug** } | { **-l** | **--legacy** }

tools/bpf/bpftool/gen.c

···
 
 static int gen_trace(struct bpf_object *obj, const char *obj_name, const char *header_guard)
 {
-	struct bpf_object_load_attr load_attr = {};
 	DECLARE_LIBBPF_OPTS(gen_loader_opts, opts);
 	struct bpf_map *map;
 	char ident[256];
···
 	if (err)
 		return err;
 
-	load_attr.obj = obj;
-	if (verifier_logs)
-		/* log_level1 + log_level2 + stats, but not stable UAPI */
-		load_attr.log_level = 1 + 2 + 4;
-
-	err = bpf_object__load_xattr(&load_attr);
+	err = bpf_object__load(obj);
 	if (err) {
 		p_err("failed to load object file");
 		goto out;
···
 	if (obj_name[0] == '\0')
 		get_obj_name(obj_name, file);
 	opts.object_name = obj_name;
+	if (verifier_logs)
+		/* log_level1 + log_level2 + stats, but not stable UAPI */
+		opts.kernel_log_level = 1 + 2 + 4;
 	obj = bpf_object__open_mem(obj_data, file_sz, &opts);
 	err = libbpf_get_error(obj);
 	if (err) {

tools/bpf/bpftool/main.c

···
 		jsonw_name(json_wtr, "features");
 		jsonw_start_object(json_wtr);	/* features */
 		jsonw_bool_field(json_wtr, "libbfd", has_libbfd);
+		jsonw_bool_field(json_wtr, "libbpf_strict", !legacy_libbpf);
 		jsonw_bool_field(json_wtr, "skeletons", has_skeletons);
 		jsonw_end_object(json_wtr);	/* features */
 
···
 		printf("features:");
 		if (has_libbfd) {
 			printf(" libbfd");
+			nb_features++;
+		}
+		if (!legacy_libbpf) {
+			printf("%s libbpf_strict", nb_features++ ? "," : "");
 			nb_features++;
 		}
 		if (has_skeletons)
···
 		{ "legacy",	no_argument,	NULL,	'l' },
 		{ 0 }
 	};
+	bool version_requested = false;
 	int opt, ret;
 
 	last_do_help = do_help;
···
 				  options, NULL)) >= 0) {
 		switch (opt) {
 		case 'V':
-			return do_version(argc, argv);
+			version_requested = true;
+			break;
 		case 'h':
 			return do_help(argc, argv);
 		case 'p':
···
 	argv += optind;
 	if (argc < 0)
 		usage();
+
+	if (version_requested)
+		return do_version(argc, argv);
 
 	ret = cmd_select(cmds, argc, argv, do_help);
 

tools/bpf/bpftool/map.c

···
 
 static int do_create(int argc, char **argv)
 {
-	struct bpf_create_map_attr attr = { NULL, };
+	LIBBPF_OPTS(bpf_map_create_opts, attr);
+	enum bpf_map_type map_type = BPF_MAP_TYPE_UNSPEC;
+	__u32 key_size = 0, value_size = 0, max_entries = 0;
+	const char *map_name = NULL;
 	const char *pinfile;
 	int err = -1, fd;
 
···
 		if (is_prefix(*argv, "type")) {
 			NEXT_ARG();
 
-			if (attr.map_type) {
+			if (map_type) {
 				p_err("map type already specified");
 				goto exit;
 			}
 
-			attr.map_type = map_type_from_str(*argv);
-			if ((int)attr.map_type < 0) {
+			map_type = map_type_from_str(*argv);
+			if ((int)map_type < 0) {
 				p_err("unrecognized map type: %s", *argv);
 				goto exit;
 			}
 			NEXT_ARG();
 		} else if (is_prefix(*argv, "name")) {
 			NEXT_ARG();
-			attr.name = GET_ARG();
+			map_name = GET_ARG();
 		} else if (is_prefix(*argv, "key")) {
-			if (parse_u32_arg(&argc, &argv, &attr.key_size,
+			if (parse_u32_arg(&argc, &argv, &key_size,
 					  "key size"))
 				goto exit;
 		} else if (is_prefix(*argv, "value")) {
-			if (parse_u32_arg(&argc, &argv, &attr.value_size,
+			if (parse_u32_arg(&argc, &argv, &value_size,
 					  "value size"))
 				goto exit;
 		} else if (is_prefix(*argv, "entries")) {
-			if (parse_u32_arg(&argc, &argv, &attr.max_entries,
+			if (parse_u32_arg(&argc, &argv, &max_entries,
 					  "max entries"))
 				goto exit;
 		} else if (is_prefix(*argv, "flags")) {
···
 		}
 	}
 
-	if (!attr.name) {
+	if (!map_name) {
 		p_err("map name not specified");
 		goto exit;
 	}
 
 	set_max_rlimit();
 
-	fd = bpf_create_map_xattr(&attr);
+	fd = bpf_map_create(map_type, map_name, key_size, value_size, max_entries, &attr);
 	if (fd < 0) {
 		p_err("map create failed: %s", strerror(errno));
 		goto exit;

tools/bpf/bpftool/prog.c

···
 	DECLARE_LIBBPF_OPTS(bpf_object_open_opts, open_opts,
 		.relaxed_maps = relaxed_maps,
 	);
-	struct bpf_object_load_attr load_attr = { 0 };
 	enum bpf_attach_type expected_attach_type;
 	struct map_replace *map_replace = NULL;
 	struct bpf_program *prog = NULL, *pos;
···
 
 	set_max_rlimit();
 
+	if (verifier_logs)
+		/* log_level1 + log_level2 + stats, but not stable UAPI */
+		open_opts.kernel_log_level = 1 + 2 + 4;
+
 	obj = bpf_object__open_file(file, &open_opts);
 	if (libbpf_get_error(obj)) {
 		p_err("failed to open object file");
···
 		goto err_close_obj;
 	}
 
-	load_attr.obj = obj;
-	if (verifier_logs)
-		/* log_level1 + log_level2 + stats, but not stable UAPI */
-		load_attr.log_level = 1 + 2 + 4;
-
-	err = bpf_object__load_xattr(&load_attr);
+	err = bpf_object__load(obj);
 	if (err) {
 		p_err("failed to load object file");
 		goto err_close_obj;
···
 					     sizeof(struct bpf_prog_desc));
 	int log_buf_sz = (1u << 24) - 1;
 	int err, fds_before, fd_delta;
-	char *log_buf;
+	char *log_buf = NULL;
 
 	ctx = alloca(ctx_sz);
 	memset(ctx, 0, ctx_sz);
 	ctx->sz = ctx_sz;
-	ctx->log_level = 1;
-	ctx->log_size = log_buf_sz;
-	log_buf = malloc(log_buf_sz);
-	if (!log_buf)
-		return -ENOMEM;
-	ctx->log_buf = (long) log_buf;
+	if (verifier_logs) {
+		ctx->log_level = 1 + 2 + 4;
+		ctx->log_size = log_buf_sz;
+		log_buf = malloc(log_buf_sz);
+		if (!log_buf)
+			return -ENOMEM;
+		ctx->log_buf = (long) log_buf;
+	}
 	opts.ctx = ctx;
 	opts.data = gen->data;
 	opts.data_sz = gen->data_sz;
···
 	fds_before = count_open_fds();
 	err = bpf_load_and_run(&opts);
 	fd_delta = count_open_fds() - fds_before;
-	if (err < 0) {
+	if (err < 0 || verifier_logs) {
 		fprintf(stderr, "err %d\n%s\n%s", err, opts.errstr, log_buf);
-		if (fd_delta)
+		if (fd_delta && err < 0)
 			fprintf(stderr, "loader prog leaked %d FDs\n",
 				fd_delta);
 	}
···
 {
 	DECLARE_LIBBPF_OPTS(bpf_object_open_opts, open_opts);
 	DECLARE_LIBBPF_OPTS(gen_loader_opts, gen);
-	struct bpf_object_load_attr load_attr = {};
 	struct bpf_object *obj;
 	const char *file;
 	int err = 0;
···
 	if (!REQ_ARGS(1))
 		return -1;
 	file = GET_ARG();
+
+	if (verifier_logs)
+		/* log_level1 + log_level2 + stats, but not stable UAPI */
+		open_opts.kernel_log_level = 1 + 2 + 4;
 
 	obj = bpf_object__open_file(file, &open_opts);
 	if (libbpf_get_error(obj)) {
···
 	if (err)
 		goto err_close_obj;
 
-	load_attr.obj = obj;
-	if (verifier_logs)
-		/* log_level1 + log_level2 + stats, but not stable UAPI */
-		load_attr.log_level = 1 + 2 + 4;
-
-	err = bpf_object__load_xattr(&load_attr);
+	err = bpf_object__load(obj);
 	if (err) {
 		p_err("failed to load object file");
 		goto err_close_obj;

tools/bpf/bpftool/struct_ops.c

···
 
 static int do_register(int argc, char **argv)
 {
-	struct bpf_object_load_attr load_attr = {};
+	LIBBPF_OPTS(bpf_object_open_opts, open_opts);
 	const struct bpf_map_def *def;
 	struct bpf_map_info info = {};
 	__u32 info_len = sizeof(info);
···
 
 	file = GET_ARG();
 
-	obj = bpf_object__open(file);
+	if (verifier_logs)
+		/* log_level1 + log_level2 + stats, but not stable UAPI */
+		open_opts.kernel_log_level = 1 + 2 + 4;
+
+	obj = bpf_object__open_file(file, &open_opts);
 	if (libbpf_get_error(obj))
 		return -1;
 
 	set_max_rlimit();
 
-	load_attr.obj = obj;
-	if (verifier_logs)
-		/* log_level1 + log_level2 + stats, but not stable UAPI */
-		load_attr.log_level = 1 + 2 + 4;
-
-	if (bpf_object__load_xattr(&load_attr)) {
+	if (bpf_object__load(obj)) {
 		bpf_object__close(obj);
 		return -1;
 	}

tools/bpf/resolve_btfids/main.c

···
 	return NULL;
 }
 
-static struct btf_id*
+static struct btf_id *
 btf_id__add(struct rb_root *root, char *name, bool unique)
 {
 	struct rb_node **p = &root->rb_node;
···
 	if (obj.efile.idlist_shndx == -1 ||
 	    obj.efile.symbols_shndx == -1) {
 		pr_debug("Cannot find .BTF_ids or symbols sections, nothing to do\n");
-		return 0;
+		err = 0;
+		goto out;
 	}
 
 	if (symbols_collect(&obj))

tools/build/feature/test-bpf.c

···
 #  define __NR_bpf 349
 # elif defined(__s390__)
 #  define __NR_bpf 351
+# elif defined(__mips__) && defined(_ABIO32)
+#  define __NR_bpf 4355
+# elif defined(__mips__) && defined(_ABIN32)
+#  define __NR_bpf 6319
+# elif defined(__mips__) && defined(_ABI64)
+#  define __NR_bpf 5315
 # else
 #  error __NR_bpf not defined. libbpf does not support your arch.
 # endif

tools/include/uapi/linux/bpf.h

···
 			/* or valid module BTF object fd or 0 to attach to vmlinux */
 			__u32		attach_btf_obj_fd;
 		};
-		__u32		:32;		/* pad */
+		__u32		core_relo_cnt;	/* number of bpf_core_relo */
 		__aligned_u64	fd_array;	/* array of FDs */
+		__aligned_u64	core_relos;
+		__u32		core_relo_rec_size; /* sizeof(struct bpf_core_relo) */
 	};
 
 	struct { /* anonymous struct used by BPF_OBJ_* commands */
···
  * 		if the maximum number of tail calls has been reached for this
  * 		chain of programs. This limit is defined in the kernel by the
  * 		macro **MAX_TAIL_CALL_CNT** (not accessible to user space),
- * 		which is currently set to 32.
+ *		which is currently set to 33.
  * 	Return
  * 		0 on success, or a negative error in case of failure.
  *
···
  *		**-ENOENT** if *task->mm* is NULL, or no vma contains *addr*.
  *		**-EBUSY** if failed to try lock mmap_lock.
  *		**-EINVAL** for invalid **flags**.
+ *
+ * long bpf_loop(u32 nr_loops, void *callback_fn, void *callback_ctx, u64 flags)
+ *	Description
+ *		For **nr_loops**, call **callback_fn** function
+ *		with **callback_ctx** as the context parameter.
+ *		The **callback_fn** should be a static function and
+ *		the **callback_ctx** should be a pointer to the stack.
+ *		The **flags** is used to control certain aspects of the helper.
+ *		Currently, the **flags** must be 0. Currently, nr_loops is
+ *		limited to 1 << 23 (~8 million) loops.
+ *
+ *		long (\*callback_fn)(u32 index, void \*ctx);
+ *
+ *		where **index** is the current index in the loop. The index
+ *		is zero-indexed.
+ *
+ *		If **callback_fn** returns 0, the helper will continue to the next
+ *		loop. If return value is 1, the helper will skip the rest of
+ *		the loops and return. Other return values are not used now,
+ *		and will be rejected by the verifier.
+ *
+ *	Return
+ *		The number of loops performed, **-EINVAL** for invalid **flags**,
+ *		**-E2BIG** if **nr_loops** exceeds the maximum number of loops.
  */
 #define __BPF_FUNC_MAPPER(FN)		\
 	FN(unspec),			\
···
 	FN(skc_to_unix_sock),		\
 	FN(kallsyms_lookup_name),	\
 	FN(find_vma),			\
+	FN(loop),			\
 	/* */
 
 /* integer value in 'imm' field of BPF_CALL instruction selects which helper
···
 	BTF_F_NONAME	=	(1ULL << 1),
 	BTF_F_PTR_RAW	=	(1ULL << 2),
 	BTF_F_ZERO	=	(1ULL << 3),
+};
+
+/* bpf_core_relo_kind encodes which aspect of captured field/type/enum value
+ * has to be adjusted by relocations. It is emitted by llvm and passed to
+ * libbpf and later to the kernel.
+ */
+enum bpf_core_relo_kind {
+	BPF_CORE_FIELD_BYTE_OFFSET = 0,      /* field byte offset */
+	BPF_CORE_FIELD_BYTE_SIZE = 1,        /* field size in bytes */
+	BPF_CORE_FIELD_EXISTS = 2,           /* field existence in target kernel */
+	BPF_CORE_FIELD_SIGNED = 3,           /* field signedness (0 - unsigned, 1 - signed) */
+	BPF_CORE_FIELD_LSHIFT_U64 = 4,       /* bitfield-specific left bitshift */
+	BPF_CORE_FIELD_RSHIFT_U64 = 5,       /* bitfield-specific right bitshift */
+	BPF_CORE_TYPE_ID_LOCAL = 6,          /* type ID in local BPF object */
+	BPF_CORE_TYPE_ID_TARGET = 7,         /* type ID in target kernel */
+	BPF_CORE_TYPE_EXISTS = 8,            /* type existence in target kernel */
+	BPF_CORE_TYPE_SIZE = 9,              /* type size in bytes */
+	BPF_CORE_ENUMVAL_EXISTS = 10,        /* enum value existence in target kernel */
+	BPF_CORE_ENUMVAL_VALUE = 11,         /* enum value integer value */
+};
+
+/*
+ * "struct bpf_core_relo" is used to pass relocation data form LLVM to libbpf
+ * and from libbpf to the kernel.
+ *
+ * CO-RE relocation captures the following data:
+ * - insn_off - instruction offset (in bytes) within a BPF program that needs
+ *   its insn->imm field to be relocated with actual field info;
+ * - type_id - BTF type ID of the "root" (containing) entity of a relocatable
+ *   type or field;
+ * - access_str_off - offset into corresponding .BTF string section. String
+ *   interpretation depends on specific relocation kind:
+ *     - for field-based relocations, string encodes an accessed field using
+ *       a sequence of field and array indices, separated by colon (:). It's
+ *       conceptually very close to LLVM's getelementptr ([0]) instruction's
+ *       arguments for identifying offset to a field.
+ *     - for type-based relocations, strings is expected to be just "0";
+ *     - for enum value-based relocations, string contains an index of enum
+ *       value within its enum type;
+ * - kind - one of enum bpf_core_relo_kind;
+ *
+ * Example:
+ *   struct sample {
+ *       int a;
+ *       struct {
+ *           int b[10];
+ *       };
+ *   };
+ *
+ *   struct sample *s = ...;
+ *   int *x = &s->a;     // encoded as "0:0" (a is field #0)
+ *   int *y = &s->b[5];  // encoded as "0:1:0:5" (anon struct is field #1,
+ *                       // b is field #0 inside anon struct, accessing elem #5)
+ *   int *z = &s[10]->b; // encoded as "10:1" (ptr is used as an array)
+ *
+ * type_id for all relocs in this example will capture BTF type id of
+ * `struct sample`.
+ *
+ * Such relocation is emitted when using __builtin_preserve_access_index()
+ * Clang built-in, passing expression that captures field address, e.g.:
+ *
+ * bpf_probe_read(&dst, sizeof(dst),
+ *		  __builtin_preserve_access_index(&src->a.b.c));
+ *
+ * In this case Clang will emit field relocation recording necessary data to
+ * be able to find offset of embedded `a.b.c` field within `src` struct.
+ *
+ * [0] https://llvm.org/docs/LangRef.html#getelementptr-instruction
+ */
+struct bpf_core_relo {
+	__u32 insn_off;
+	__u32 type_id;
+	__u32 access_str_off;
+	enum bpf_core_relo_kind kind;
 };
 
 #endif /* _UAPI__LINUX_BPF_H__ */

tools/lib/bpf/bpf.c

···
 #  define __NR_bpf 351
 # elif defined(__arc__)
 #  define __NR_bpf 280
+# elif defined(__mips__) && defined(_ABIO32)
+#  define __NR_bpf 4355
+# elif defined(__mips__) && defined(_ABIN32)
+#  define __NR_bpf 6319
+# elif defined(__mips__) && defined(_ABI64)
+#  define __NR_bpf 5315
 # else
 #  error __NR_bpf not defined. libbpf does not support your arch.
 # endif
···
 	return fd;
 }
 
-int libbpf__bpf_create_map_xattr(const struct bpf_create_map_params *create_attr)
+int bpf_map_create(enum bpf_map_type map_type,
+		   const char *map_name,
+		   __u32 key_size,
+		   __u32 value_size,
+		   __u32 max_entries,
+		   const struct bpf_map_create_opts *opts)
 {
+	const size_t attr_sz = offsetofend(union bpf_attr, map_extra);
 	union bpf_attr attr;
 	int fd;
 
-	memset(&attr, '\0', sizeof(attr));
+	memset(&attr, 0, attr_sz);
 
-	attr.map_type = create_attr->map_type;
-	attr.key_size = create_attr->key_size;
-	attr.value_size = create_attr->value_size;
-	attr.max_entries = create_attr->max_entries;
-	attr.map_flags = create_attr->map_flags;
-	if (create_attr->name)
-		memcpy(attr.map_name, create_attr->name,
-		       min(strlen(create_attr->name), BPF_OBJ_NAME_LEN - 1));
-	attr.numa_node = create_attr->numa_node;
-	attr.btf_fd = create_attr->btf_fd;
-	attr.btf_key_type_id = create_attr->btf_key_type_id;
-	attr.btf_value_type_id = create_attr->btf_value_type_id;
-	attr.map_ifindex = create_attr->map_ifindex;
-	if (attr.map_type == BPF_MAP_TYPE_STRUCT_OPS)
-		attr.btf_vmlinux_value_type_id =
-			create_attr->btf_vmlinux_value_type_id;
-	else
-		attr.inner_map_fd = create_attr->inner_map_fd;
-	attr.map_extra = create_attr->map_extra;
+	if (!OPTS_VALID(opts, bpf_map_create_opts))
+		return libbpf_err(-EINVAL);
 
-	fd = sys_bpf_fd(BPF_MAP_CREATE, &attr, sizeof(attr));
+	attr.map_type = map_type;
+	if (map_name)
+		strncat(attr.map_name, map_name, sizeof(attr.map_name) - 1);
+	attr.key_size = key_size;
+	attr.value_size = value_size;
+	attr.max_entries = max_entries;
+
+	attr.btf_fd = OPTS_GET(opts, btf_fd, 0);
+	attr.btf_key_type_id = OPTS_GET(opts, btf_key_type_id, 0);
+	attr.btf_value_type_id = OPTS_GET(opts, btf_value_type_id, 0);
+	attr.btf_vmlinux_value_type_id = OPTS_GET(opts, btf_vmlinux_value_type_id, 0);
+
+	attr.inner_map_fd = OPTS_GET(opts, inner_map_fd, 0);
+	attr.map_flags = OPTS_GET(opts, map_flags, 0);
+	attr.map_extra = OPTS_GET(opts, map_extra, 0);
+	attr.numa_node = OPTS_GET(opts, numa_node, 0);
+	attr.map_ifindex = OPTS_GET(opts, map_ifindex, 0);
+
+	fd = sys_bpf_fd(BPF_MAP_CREATE, &attr, attr_sz);
 	return libbpf_err_errno(fd);
 }
 
 int bpf_create_map_xattr(const struct bpf_create_map_attr *create_attr)
 {
-	struct bpf_create_map_params p = {};
+	LIBBPF_OPTS(bpf_map_create_opts, p);
 
-	p.map_type = create_attr->map_type;
-	p.key_size = create_attr->key_size;
-	p.value_size = create_attr->value_size;
-	p.max_entries = create_attr->max_entries;
 	p.map_flags = create_attr->map_flags;
-	p.name = create_attr->name;
 	p.numa_node = create_attr->numa_node;
 	p.btf_fd = create_attr->btf_fd;
 	p.btf_key_type_id = create_attr->btf_key_type_id;
 	p.btf_value_type_id = create_attr->btf_value_type_id;
 	p.map_ifindex = create_attr->map_ifindex;
-	if (p.map_type == BPF_MAP_TYPE_STRUCT_OPS)
-		p.btf_vmlinux_value_type_id =
-			create_attr->btf_vmlinux_value_type_id;
+	if (create_attr->map_type == BPF_MAP_TYPE_STRUCT_OPS)
+		p.btf_vmlinux_value_type_id = create_attr->btf_vmlinux_value_type_id;
 	else
 		p.inner_map_fd = create_attr->inner_map_fd;
 
-	return libbpf__bpf_create_map_xattr(&p);
+	return bpf_map_create(create_attr->map_type, create_attr->name,
+			      create_attr->key_size, create_attr->value_size,
+			      create_attr->max_entries, &p);
 }
 
 int bpf_create_map_node(enum bpf_map_type map_type, const char *name,
 			int key_size, int value_size, int max_entries,
 			__u32 map_flags, int node)
 {
-	struct bpf_create_map_attr map_attr = {};
+	LIBBPF_OPTS(bpf_map_create_opts, opts);
 
-	map_attr.name = name;
-	map_attr.map_type = map_type;
-	map_attr.map_flags = map_flags;
-	map_attr.key_size = key_size;
-	map_attr.value_size = value_size;
-	map_attr.max_entries = max_entries;
+	opts.map_flags = map_flags;
 	if (node >= 0) {
-		map_attr.numa_node = node;
-		map_attr.map_flags |= BPF_F_NUMA_NODE;
+		opts.numa_node = node;
+		opts.map_flags |= BPF_F_NUMA_NODE;
 	}
 
-	return bpf_create_map_xattr(&map_attr);
+	return bpf_map_create(map_type, name, key_size, value_size, max_entries, &opts);
 }
 
 int bpf_create_map(enum bpf_map_type map_type, int key_size,
 		   int value_size, int max_entries, __u32 map_flags)
 {
-	struct bpf_create_map_attr map_attr = {};
+	LIBBPF_OPTS(bpf_map_create_opts, opts, .map_flags = map_flags);
 
-	map_attr.map_type = map_type;
-	map_attr.map_flags = map_flags;
-	map_attr.key_size = key_size;
-	map_attr.value_size = value_size;
-	map_attr.max_entries = max_entries;
-
-	return bpf_create_map_xattr(&map_attr);
+	return bpf_map_create(map_type, NULL, key_size, value_size, max_entries, &opts);
 }
 
 int bpf_create_map_name(enum bpf_map_type map_type, const char *name,
 			int key_size, int value_size, int max_entries,
 			__u32 map_flags)
 {
-	struct bpf_create_map_attr map_attr = {};
+	LIBBPF_OPTS(bpf_map_create_opts, opts, .map_flags = map_flags);
 
-	map_attr.name = name;
-	map_attr.map_type = map_type;
-	map_attr.map_flags = map_flags;
-	map_attr.key_size = key_size;
-	map_attr.value_size = value_size;
-	map_attr.max_entries = max_entries;
-
-	return bpf_create_map_xattr(&map_attr);
+	return bpf_map_create(map_type, name, key_size, value_size, max_entries, &opts);
 }
 
 int bpf_create_map_in_map_node(enum bpf_map_type map_type, const char *name,
 			       int key_size, int inner_map_fd, int max_entries,
 			       __u32 map_flags, int node)
 {
-	union bpf_attr attr;
-	int fd;
+	LIBBPF_OPTS(bpf_map_create_opts, opts);
 
-	memset(&attr, '\0', sizeof(attr));
-
-	attr.map_type = map_type;
-	attr.key_size = key_size;
-	attr.value_size = 4;
-	attr.inner_map_fd = inner_map_fd;
-	attr.max_entries = max_entries;
-	attr.map_flags = map_flags;
-	if (name)
-		memcpy(attr.map_name, name,
-		       min(strlen(name), BPF_OBJ_NAME_LEN - 1));
-
+	opts.inner_map_fd = inner_map_fd;
+	opts.map_flags = map_flags;
 	if (node >= 0) {
-		attr.map_flags |= BPF_F_NUMA_NODE;
-		attr.numa_node = node;
+		opts.map_flags |= BPF_F_NUMA_NODE;
+		opts.numa_node = node;
 	}
 
-	fd = sys_bpf_fd(BPF_MAP_CREATE, &attr, sizeof(attr));
-	return libbpf_err_errno(fd);
+	return bpf_map_create(map_type, name, key_size, 4, max_entries, &opts);
 }
 
 int bpf_create_map_in_map(enum bpf_map_type map_type, const char *name,
 			  int key_size, int inner_map_fd, int max_entries,
 			  __u32 map_flags)
 {
-	return bpf_create_map_in_map_node(map_type, name, key_size,
-					  inner_map_fd, max_entries, map_flags,
-					  -1);
+	LIBBPF_OPTS(bpf_map_create_opts, opts,
+		.inner_map_fd = inner_map_fd,
+		.map_flags = map_flags,
+	);
+
+	return bpf_map_create(map_type, name, key_size, 4, max_entries, &opts);
 }
 
 static void *
···
 	if (log_level && !log_buf)
 		return libbpf_err(-EINVAL);
 
-	attr.log_level = log_level;
-	attr.log_buf = ptr_to_u64(log_buf);
-	attr.log_size = log_size;
-
 	func_info_rec_size = OPTS_GET(opts, func_info_rec_size, 0);
 	func_info = OPTS_GET(opts, func_info, NULL);
 	attr.func_info_rec_size = func_info_rec_size;
···
 	attr.line_info_cnt = OPTS_GET(opts, line_info_cnt, 0);
 
 	attr.fd_array = ptr_to_u64(OPTS_GET(opts, fd_array, NULL));
+
+	if (log_level) {
+		attr.log_buf = ptr_to_u64(log_buf);
+		attr.log_size = log_size;
+		attr.log_level = log_level;
+	}
 
 	fd = sys_bpf_prog_load(&attr, sizeof(attr), attempts);
 	if (fd >= 0)
···
 			goto done;
 	}
 
-	if (log_level || !log_buf)
-		goto done;
+	if (log_level == 0 && log_buf) {
+		/* log_level == 0 with non-NULL log_buf requires retrying on error
+		 * with log_level == 1 and log_buf/log_buf_size set, to get details of
+		 * failure
+		 */
+		attr.log_buf = ptr_to_u64(log_buf);
+		attr.log_size = log_size;
+		attr.log_level = 1;
 
-	/* Try again with log */
-	log_buf[0] = 0;
-	attr.log_buf = ptr_to_u64(log_buf);
-	attr.log_size = log_size;
-	attr.log_level = 1;
-
-	fd = sys_bpf_prog_load(&attr, sizeof(attr), attempts);
+		fd = sys_bpf_prog_load(&attr, sizeof(attr), attempts);
+	}
 done:
 	/* free() doesn't affect errno, so we don't need to restore it */
 	free(finfo);
···
 	return libbpf_err_errno(fd);
 }
 
-int bpf_load_btf(const void *btf, __u32 btf_size, char *log_buf, __u32 log_buf_size,
-		 bool do_log)
+int bpf_btf_load(const void *btf_data, size_t btf_size, const struct bpf_btf_load_opts *opts)
 {
-	union bpf_attr attr = {};
+	const size_t attr_sz = offsetofend(union bpf_attr, btf_log_level);
+	union bpf_attr attr;
+	char *log_buf;
+	size_t log_size;
+	__u32 log_level;
 	int fd;
 
-	attr.btf = ptr_to_u64(btf);
+	memset(&attr, 0, attr_sz);
+
+	if (!OPTS_VALID(opts, bpf_btf_load_opts))
+		return libbpf_err(-EINVAL);
+
+	log_buf = OPTS_GET(opts, log_buf, NULL);
+	log_size = OPTS_GET(opts, log_size, 0);
+	log_level = OPTS_GET(opts, log_level, 0);
+
+	if (log_size > UINT_MAX)
+		return libbpf_err(-EINVAL);
+	if (log_size && !log_buf)
+		return libbpf_err(-EINVAL);
+
+	attr.btf = ptr_to_u64(btf_data);
 	attr.btf_size = btf_size;
+	/* log_level == 0 and log_buf != NULL means "try loading without
+	 * log_buf, but retry with log_buf and log_level=1 on error", which is
+	 * consistent across low-level and high-level BTF and program loading
+	 * APIs within libbpf and provides a sensible behavior in practice
+	 */
+	if (log_level) {
+		attr.btf_log_buf = ptr_to_u64(log_buf);
+		attr.btf_log_size = (__u32)log_size;
+		attr.btf_log_level = log_level;
+	}
+
+	fd = sys_bpf_fd(BPF_BTF_LOAD, &attr, attr_sz);
+	if (fd < 0 && log_buf && log_level == 0) {
+		attr.btf_log_buf = ptr_to_u64(log_buf);
+		attr.btf_log_size = (__u32)log_size;
+		attr.btf_log_level = 1;
+		fd = sys_bpf_fd(BPF_BTF_LOAD, &attr, attr_sz);
+	}
+	return libbpf_err_errno(fd);
+}
+
+int bpf_load_btf(const void *btf, __u32 btf_size, char *log_buf, __u32 log_buf_size, bool do_log)
+{
+	LIBBPF_OPTS(bpf_btf_load_opts, opts);
+	int fd;
 
 retry:
 	if (do_log && log_buf && log_buf_size) {
-		attr.btf_log_level = 1;
-		attr.btf_log_size = log_buf_size;
-		attr.btf_log_buf = ptr_to_u64(log_buf);
+		opts.log_buf = log_buf;
+		opts.log_size = log_buf_size;
+		opts.log_level = 1;
 	}
 
-	fd = sys_bpf_fd(BPF_BTF_LOAD, &attr, sizeof(attr));
-
+	fd = bpf_btf_load(btf, btf_size, &opts);
 	if (fd < 0 && !do_log && log_buf && log_buf_size) {
 		do_log = true;
 		goto retry;

tools/lib/bpf/bpf.h

···
 extern "C" {
 #endif
 
+struct bpf_map_create_opts {
+	size_t sz; /* size of this struct for forward/backward compatibility */
+
+	__u32 btf_fd;
+	__u32 btf_key_type_id;
+	__u32 btf_value_type_id;
+	__u32 btf_vmlinux_value_type_id;
+
+	__u32 inner_map_fd;
+	__u32 map_flags;
+	__u64 map_extra;
+
+	__u32 numa_node;
+	__u32 map_ifindex;
+};
+#define bpf_map_create_opts__last_field map_ifindex
+
+LIBBPF_API int bpf_map_create(enum bpf_map_type map_type,
+			      const char *map_name,
+			      __u32 key_size,
+			      __u32 value_size,
+			      __u32 max_entries,
+			      const struct bpf_map_create_opts *opts);
+
 struct bpf_create_map_attr {
 	const char *name;
 	enum bpf_map_type map_type;
···
 	};
 };
 
-LIBBPF_API int
-bpf_create_map_xattr(const struct bpf_create_map_attr *create_attr);
+LIBBPF_DEPRECATED_SINCE(0, 7, "use bpf_map_create() instead")
+LIBBPF_API int bpf_create_map_xattr(const struct bpf_create_map_attr *create_attr);
+LIBBPF_DEPRECATED_SINCE(0, 7, "use bpf_map_create() instead")
 LIBBPF_API int bpf_create_map_node(enum bpf_map_type map_type, const char *name,
 				   int key_size, int value_size,
 				   int max_entries, __u32 map_flags, int node);
+LIBBPF_DEPRECATED_SINCE(0, 7, "use bpf_map_create() instead")
 LIBBPF_API int bpf_create_map_name(enum bpf_map_type map_type, const char *name,
 				   int key_size, int value_size,
 				   int max_entries, __u32 map_flags);
+LIBBPF_DEPRECATED_SINCE(0, 7, "use bpf_map_create() instead")
 LIBBPF_API int bpf_create_map(enum bpf_map_type map_type, int key_size,
 			      int value_size, int max_entries, __u32 map_flags);
+LIBBPF_DEPRECATED_SINCE(0, 7, "use bpf_map_create() instead")
 LIBBPF_API int bpf_create_map_in_map_node(enum bpf_map_type map_type,
 					  const char *name, int key_size,
 					  int inner_map_fd, int max_entries,
 					  __u32 map_flags, int node);
+LIBBPF_DEPRECATED_SINCE(0, 7, "use bpf_map_create() instead")
 LIBBPF_API int bpf_create_map_in_map(enum bpf_map_type map_type,
 				     const char *name, int key_size,
 				     int inner_map_fd, int max_entries,
···
 /* Flags to direct loading requirements */
 #define MAPS_RELAX_COMPAT	0x01
 
-/* Recommend log buffer size */
+/* Recommended log buffer size */
 #define BPF_LOG_BUF_SIZE (UINT32_MAX >> 8) /* verifier maximum in kernels <= 5.1 */
+
 LIBBPF_DEPRECATED_SINCE(0, 7, "use bpf_prog_load() instead")
 LIBBPF_API int bpf_load_program_xattr(const struct bpf_load_program_attr *load_attr,
 				      char *log_buf, size_t log_buf_sz);
···
 				  const char *license, __u32 kern_version,
 				  char *log_buf, size_t log_buf_sz,
 				  int log_level);
+
+struct bpf_btf_load_opts {
+	size_t sz; /* size of this struct for forward/backward compatibility */
+
+	/* kernel log options */
+	char *log_buf;
+	__u32 log_level;
+	__u32 log_size;
+};
+#define bpf_btf_load_opts__last_field log_size
+
+LIBBPF_API int bpf_btf_load(const void *btf_data, size_t btf_size,
+			    const struct bpf_btf_load_opts *opts);
+
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_btf_load() instead")
+LIBBPF_API int bpf_load_btf(const void *btf, __u32 btf_size, char *log_buf,
+			    __u32 log_buf_size, bool do_log);
 
 LIBBPF_API int bpf_map_update_elem(int fd, const void *key, const void *value,
 				   __u64 flags);
···
 			      __u32 query_flags, __u32 *attach_flags,
 			      __u32 *prog_ids, __u32 *prog_cnt);
 LIBBPF_API int bpf_raw_tracepoint_open(const char *name, int prog_fd);
-LIBBPF_API int bpf_load_btf(const void *btf, __u32 btf_size, char *log_buf,
-			    __u32 log_buf_size, bool do_log);
 LIBBPF_API int bpf_task_fd_query(int pid, int fd, __u32 flags, char *buf,
 				 __u32 *buf_len, __u32 *prog_id, __u32 *fd_type,
 				 __u64 *probe_offset, __u64 *probe_addr);

tools/lib/bpf/bpf_gen_internal.h

···
 	int error;
 	struct ksym_relo_desc *relos;
 	int relo_cnt;
+	struct bpf_core_relo *core_relos;
+	int core_relo_cnt;
 	char attach_target[128];
 	int attach_kind;
 	struct ksym_desc *ksyms;
···
 int bpf_gen__finish(struct bpf_gen *gen, int nr_progs, int nr_maps);
 void bpf_gen__free(struct bpf_gen *gen);
 void bpf_gen__load_btf(struct bpf_gen *gen, const void *raw_data, __u32 raw_size);
-void bpf_gen__map_create(struct bpf_gen *gen, struct bpf_create_map_params *map_attr, int map_idx);
+void bpf_gen__map_create(struct bpf_gen *gen,
+			 enum bpf_map_type map_type, const char *map_name,
+			 __u32 key_size, __u32 value_size, __u32 max_entries,
+			 struct bpf_map_create_opts *map_attr, int map_idx);
 void bpf_gen__prog_load(struct bpf_gen *gen,
 			enum bpf_prog_type prog_type, const char *prog_name,
 			const char *license, struct bpf_insn *insns, size_t insn_cnt,
···
 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, bool is_weak,
 			    bool is_typeless, int kind, int insn_idx);
+void bpf_gen__record_relo_core(struct bpf_gen *gen, const struct bpf_core_relo *core_relo);
+void bpf_gen__populate_outer_map(struct bpf_gen *gen, int outer_map_idx, int key, int inner_map_idx);
 
 #endif

tools/lib/bpf/btf.c

···
 }
 
 /* internal helper returning non-const pointer to a type */
-struct btf_type *btf_type_by_id(struct btf *btf, __u32 type_id)
+struct btf_type *btf_type_by_id(const struct btf *btf, __u32 type_id)
 {
 	if (type_id == 0)
 		return &btf_void;
···
 		case BTF_KIND_RESTRICT:
 		case BTF_KIND_VAR:
 		case BTF_KIND_DECL_TAG:
+		case BTF_KIND_TYPE_TAG:
 			type_id = t->type;
 			break;
 		case BTF_KIND_ARRAY:
···
 
 static void *btf_get_raw_data(const struct btf *btf, __u32 *size, bool swap_endian);
 
-int btf__load_into_kernel(struct btf *btf)
+int btf_load_into_kernel(struct btf *btf, char *log_buf, size_t log_sz, __u32 log_level)
 {
-	__u32 log_buf_size = 0, raw_size;
-	char *log_buf = NULL;
+	LIBBPF_OPTS(bpf_btf_load_opts, opts);
+	__u32 buf_sz = 0, raw_size;
+	char *buf = NULL, *tmp;
 	void *raw_data;
 	int err = 0;
 
 	if (btf->fd >= 0)
 		return libbpf_err(-EEXIST);
+	if (log_sz && !log_buf)
+		return libbpf_err(-EINVAL);
 
-retry_load:
-	if (log_buf_size) {
-		log_buf = malloc(log_buf_size);
-		if (!log_buf)
-			return libbpf_err(-ENOMEM);
-
-		*log_buf = 0;
-	}
-
+	/* cache native raw data representation */
 	raw_data = btf_get_raw_data(btf, &raw_size, false);
 	if (!raw_data) {
 		err = -ENOMEM;
 		goto done;
 	}
-	/* cache native raw data representation */
 	btf->raw_size = raw_size;
 	btf->raw_data = raw_data;
 
-	btf->fd = bpf_load_btf(raw_data, raw_size, log_buf, log_buf_size, false);
-	if (btf->fd < 0) {
-		if (!log_buf || errno == ENOSPC) {
-			log_buf_size = max((__u32)BPF_LOG_BUF_SIZE,
-					   log_buf_size << 1);
-			free(log_buf);
-			goto retry_load;
+retry_load:
+	/* if log_level is 0, we won't provide log_buf/log_size to the kernel,
+	 * initially. Only if BTF loading fails, we bump log_level to 1 and
+	 * retry, using either auto-allocated or custom log_buf. This way
+	 * non-NULL custom log_buf provides a buffer just in case, but hopes
+	 * for successful load and no need for log_buf.
+	 */
+	if (log_level) {
+		/* if caller didn't provide custom log_buf, we'll keep
+		 * allocating our own progressively bigger buffers for BTF
+		 * verification log
+		 */
+		if (!log_buf) {
+			buf_sz = max((__u32)BPF_LOG_BUF_SIZE, buf_sz * 2);
+			tmp = realloc(buf, buf_sz);
+			if (!tmp) {
+				err = -ENOMEM;
+				goto done;
+			}
+			buf = tmp;
+			buf[0] = '\0';
 		}
 
+		opts.log_buf = log_buf ? log_buf : buf;
+		opts.log_size = log_buf ? log_sz : buf_sz;
+		opts.log_level = log_level;
+	}
+
+	btf->fd = bpf_btf_load(raw_data, raw_size, &opts);
+	if (btf->fd < 0) {
+		/* time to turn on verbose mode and try again */
+		if (log_level == 0) {
+			log_level = 1;
+			goto retry_load;
+		}
+		/* only retry if caller didn't provide custom log_buf, but
+		 * make sure we can never overflow buf_sz
+		 */
+		if (!log_buf && errno == ENOSPC && buf_sz <= UINT_MAX / 2)
+			goto retry_load;
+
 		err = -errno;
-		pr_warn("Error loading BTF: %s(%d)\n", strerror(errno), errno);
-		if (*log_buf)
-			pr_warn("%s\n", log_buf);
-		goto done;
+		pr_warn("BTF loading error: %d\n", err);
+		/* don't print out contents of custom log_buf */
+		if (!log_buf && buf[0])
+			pr_warn("-- BEGIN BTF LOAD LOG ---\n%s\n-- END BTF LOAD LOG --\n", buf);
 	}
 
 done:
-	free(log_buf);
+	free(buf);
 	return libbpf_err(err);
 }
+
+int btf__load_into_kernel(struct btf *btf)
+{
+	return btf_load_into_kernel(btf, NULL, 0, 0);
+}
+
 int btf__load(struct btf *) __attribute__((alias("btf__load_into_kernel")));
 
 int btf__fd(const struct btf *btf)
···
 	free(btf_ext);
 }
 
-struct btf_ext *btf_ext__new(__u8 *data, __u32 size)
+struct btf_ext *btf_ext__new(const __u8 *data, __u32 size)
 {
 	struct btf_ext *btf_ext;
 	int err;
-
-	err = btf_ext_parse_hdr(data, size);
-	if (err)
-		return libbpf_err_ptr(err);
 
 	btf_ext = calloc(1, sizeof(struct btf_ext));
 	if (!btf_ext)
···
 		goto done;
 	}
 	memcpy(btf_ext->data, data, size);
+
+	err = btf_ext_parse_hdr(btf_ext->data, size);
+	if (err)
+		goto done;
 
 	if (btf_ext->hdr->hdr_len < offsetofend(struct btf_ext_header, line_info_len)) {
 		err = -EINVAL;
···
 	return libbpf_err(err);
 }
 
-COMPAT_VERSION(bpf__dedup_deprecated, btf__dedup, LIBBPF_0.0.2)
+COMPAT_VERSION(btf__dedup_deprecated, btf__dedup, LIBBPF_0.0.2)
 int btf__dedup_deprecated(struct btf *btf, struct btf_ext *btf_ext, const void *unused_opts)
 {
 	LIBBPF_OPTS(btf_dedup_opts, opts, .btf_ext = btf_ext);
···
 }
 
 /*
- * Check structural compatibility of two FUNC_PROTOs, ignoring referenced type
- * IDs. This check is performed during type graph equivalence check and
+ * Check structural compatibility of two STRUCTs/UNIONs, ignoring referenced
+ * type IDs. This check is performed during type graph equivalence check and
  * referenced types equivalence is checked separately.
  */
 static bool btf_shallow_equal_struct(struct btf_type *t1, struct btf_type *t2)
···
 	return btf_equal_array(t1, t2);
 }
 
+/* Check if given two types are identical STRUCT/UNION definitions */
+static bool btf_dedup_identical_structs(struct btf_dedup *d, __u32 id1, __u32 id2)
+{
+	const struct btf_member *m1, *m2;
+	struct btf_type *t1, *t2;
+	int n, i;
+
+	t1 = btf_type_by_id(d->btf, id1);
+	t2 = btf_type_by_id(d->btf, id2);
+
+	if (!btf_is_composite(t1) || btf_kind(t1) != btf_kind(t2))
+		return false;
+
+	if (!btf_shallow_equal_struct(t1, t2))
+		return false;
+
+	m1 = btf_members(t1);
+	m2 = btf_members(t2);
+	for (i = 0, n = btf_vlen(t1); i < n; i++, m1++, m2++) {
+		if (m1->type != m2->type)
+			return false;
+	}
+	return true;
+}
+
 /*
  * Check equivalence of BTF type graph formed by candidate struct/union (we'll
  * call it "candidate graph" in this description for brevity) to a type graph
···
 
 	hypot_type_id = d->hypot_map[canon_id];
 	if (hypot_type_id <= BTF_MAX_NR_TYPES) {
+		if (hypot_type_id == cand_id)
+			return 1;
 		/* In some cases compiler will generate different DWARF types
 		 * for *identical* array type definitions and use them for
 		 * different fields within the *same* struct. This breaks type
···
 		 * types within a single CU. So work around that by explicitly
 		 * allowing identical array types here.
 		 */
-		return hypot_type_id == cand_id ||
-		       btf_dedup_identical_arrays(d, hypot_type_id, cand_id);
+		if (btf_dedup_identical_arrays(d, hypot_type_id, cand_id))
+			return 1;
+		/* It turns out that similar situation can happen with
+		 * struct/union sometimes, sigh... Handle the case where
+		 * structs/unions are exactly the same, down to the referenced
+		 * type IDs. Anything more complicated (e.g., if referenced
+		 * types are different, but equivalent) is *way more*
+		 * complicated and requires a many-to-many equivalence mapping.
+		 */
+		if (btf_dedup_identical_structs(d, hypot_type_id, cand_id))
+			return 1;
+		return 0;
 	}
 
 	if (btf_dedup_hypot_map_add(d, canon_id, cand_id))
···
 	case BTF_KIND_PTR:
 	case BTF_KIND_TYPEDEF:
 	case BTF_KIND_FUNC:
+	case BTF_KIND_TYPE_TAG:
 		if (cand_type->info != canon_type->info)
 			return 0;
 		return btf_dedup_is_equiv(d, cand_type->type, canon_type->type);

tools/lib/bpf/btf.h

···
 				    __u32 expected_value_size,
 				    __u32 *key_type_id, __u32 *value_type_id);
 
-LIBBPF_API struct btf_ext *btf_ext__new(__u8 *data, __u32 size);
+LIBBPF_API struct btf_ext *btf_ext__new(const __u8 *data, __u32 size);
 LIBBPF_API void btf_ext__free(struct btf_ext *btf_ext);
 LIBBPF_API const void *btf_ext__get_raw_data(const struct btf_ext *btf_ext,
 					     __u32 *size);

tools/lib/bpf/btf_dump.c

···
 				   __u8 bits_offset,
 				   __u8 bit_sz)
 {
-	int size, err;
+	int size, err = 0;
 
 	size = btf_dump_type_data_check_overflow(d, t, id, data, bits_offset);
 	if (size < 0)

tools/lib/bpf/gen_loader.c

···
 }
 
 void bpf_gen__map_create(struct bpf_gen *gen,
-			 struct bpf_create_map_params *map_attr, int map_idx)
+			 enum bpf_map_type map_type,
+			 const char *map_name,
+			 __u32 key_size, __u32 value_size, __u32 max_entries,
+			 struct bpf_map_create_opts *map_attr, int map_idx)
 {
-	int attr_size = offsetofend(union bpf_attr, btf_vmlinux_value_type_id);
+	int attr_size = offsetofend(union bpf_attr, map_extra);
 	bool close_inner_map_fd = false;
 	int map_create_attr, idx;
 	union bpf_attr attr;
 
 	memset(&attr, 0, attr_size);
-	attr.map_type = map_attr->map_type;
-	attr.key_size = map_attr->key_size;
-	attr.value_size = map_attr->value_size;
+	attr.map_type = map_type;
+	attr.key_size = key_size;
+	attr.value_size = value_size;
 	attr.map_flags = map_attr->map_flags;
 	attr.map_extra = map_attr->map_extra;
-	memcpy(attr.map_name, map_attr->name,
-	       min((unsigned)strlen(map_attr->name), BPF_OBJ_NAME_LEN - 1));
+	if (map_name)
+		memcpy(attr.map_name, map_name,
+		       min((unsigned)strlen(map_name), BPF_OBJ_NAME_LEN - 1));
 	attr.numa_node = map_attr->numa_node;
 	attr.map_ifindex = map_attr->map_ifindex;
-	attr.max_entries = map_attr->max_entries;
-	switch (attr.map_type) {
-	case BPF_MAP_TYPE_PERF_EVENT_ARRAY:
-	case BPF_MAP_TYPE_CGROUP_ARRAY:
-	case BPF_MAP_TYPE_STACK_TRACE:
-	case BPF_MAP_TYPE_ARRAY_OF_MAPS:
-	case BPF_MAP_TYPE_HASH_OF_MAPS:
-	case BPF_MAP_TYPE_DEVMAP:
-	case BPF_MAP_TYPE_DEVMAP_HASH:
-	case BPF_MAP_TYPE_CPUMAP:
-	case BPF_MAP_TYPE_XSKMAP:
-	case BPF_MAP_TYPE_SOCKMAP:
-	case BPF_MAP_TYPE_SOCKHASH:
-	case BPF_MAP_TYPE_QUEUE:
-	case BPF_MAP_TYPE_STACK:
-	case BPF_MAP_TYPE_RINGBUF:
-		break;
-	default:
-		attr.btf_key_type_id = map_attr->btf_key_type_id;
-		attr.btf_value_type_id = map_attr->btf_value_type_id;
-	}
+	attr.max_entries = max_entries;
+	attr.btf_key_type_id = map_attr->btf_key_type_id;
+	attr.btf_value_type_id = map_attr->btf_value_type_id;
 
 	pr_debug("gen: map_create: %s idx %d type %d value_type_id %d\n",
-		 attr.map_name, map_idx, map_attr->map_type, attr.btf_value_type_id);
+		 attr.map_name, map_idx, map_type, attr.btf_value_type_id);
 
 	map_create_attr = add_data(gen, &attr, attr_size);
 	if (attr.btf_value_type_id)
···
 	/* emit MAP_CREATE command */
 	emit_sys_bpf(gen, BPF_MAP_CREATE, map_create_attr, attr_size);
 	debug_ret(gen, "map_create %s idx %d type %d value_size %d value_btf_id %d",
-		  attr.map_name, map_idx, map_attr->map_type, attr.value_size,
+		  attr.map_name, map_idx, map_type, value_size,
 		  attr.btf_value_type_id);
 	emit_check_err(gen);
 	/* remember map_fd in the stack, if successful */
···
 		return;
 	}
 	kdesc->off = btf_fd_idx;
-	/* set a default value for imm */
+	/* jump to success case */
+	emit(gen, BPF_JMP_IMM(BPF_JSGE, BPF_REG_7, 0, 3));
+	/* set value for imm, off as 0 */
 	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));
+	emit(gen, BPF_ST_MEM(BPF_H, BPF_REG_8, offsetof(struct bpf_insn, off), 0));
+	/* skip success case for ret < 0 */
+	emit(gen, BPF_JMP_IMM(BPF_JA, 0, 0, 10));
 	/* 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)));
+	/* obtain fd in BPF_REG_9 */
+	emit(gen, BPF_MOV64_REG(BPF_REG_9, BPF_REG_7));
+	emit(gen, BPF_ALU64_IMM(BPF_RSH, BPF_REG_9, 32));
+	/* jump to fd_array store if fd denotes module BTF */
+	emit(gen, BPF_JMP_IMM(BPF_JNE, BPF_REG_9, 0, 2));
+	/* set the default value for off */
+	emit(gen, BPF_ST_MEM(BPF_H, BPF_REG_8, offsetof(struct bpf_insn, off), 0));
+	/* skip BTF fd store for vmlinux BTF */
+	emit(gen, BPF_JMP_IMM(BPF_JA, 0, 0, 4));
 	/* 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:
···
 			       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));
-		emit(gen, BPF_LDX_MEM(BPF_W, BPF_REG_9, BPF_REG_8, offsetof(struct bpf_insn, imm)));
-		/* jump over src_reg adjustment if imm is not 0 */
-		emit(gen, BPF_JMP_IMM(BPF_JNE, BPF_REG_9, 0, 3));
+		/* jump over src_reg adjustment if imm is not 0, reuse BPF_REG_0 from move_blob2blob */
+		emit(gen, BPF_JMP_IMM(BPF_JNE, BPF_REG_0, 0, 3));
 		goto clear_src_reg;
 	}
 	/* remember insn offset, so we can copy BTF ID and FD later */
···
 	emit_bpf_find_by_name_kind(gen, relo);
 	if (!relo->is_weak)
 		emit_check_err(gen);
-	/* set default values as 0 */
+	/* jump to success case */
+	emit(gen, BPF_JMP_IMM(BPF_JSGE, BPF_REG_7, 0, 3));
+	/* set values for insn[insn_idx].imm, insn[insn_idx + 1].imm as 0 */
 	emit(gen, BPF_ST_MEM(BPF_W, BPF_REG_8, offsetof(struct bpf_insn, imm), 0));
 	emit(gen, BPF_ST_MEM(BPF_W, BPF_REG_8, sizeof(struct bpf_insn) + offsetof(struct bpf_insn, imm), 0));
-	/* skip success case stores if ret < 0 */
-	emit(gen, BPF_JMP_IMM(BPF_JSLT, BPF_REG_7, 0, 4));
+	/* skip success case for ret < 0 */
+	emit(gen, BPF_JMP_IMM(BPF_JA, 0, 0, 4));
 	/* 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)));
 	/* 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)));
+	/* skip src_reg adjustment */
 	emit(gen, BPF_JMP_IMM(BPF_JSGE, BPF_REG_7, 0, 3));
 clear_src_reg:
 	/* clear bpf_object__relocate_data's src_reg assignment, otherwise we get a verifier failure */
···
 	emit(gen, BPF_STX_MEM(BPF_B, BPF_REG_8, BPF_REG_9, offsetofend(struct bpf_insn, code)));
 
 	emit_ksym_relo_log(gen, relo, kdesc->ref);
+}
+
+void bpf_gen__record_relo_core(struct bpf_gen *gen,
+			       const struct bpf_core_relo *core_relo)
+{
+	struct bpf_core_relo *relos;
+
+	relos = libbpf_reallocarray(gen->core_relos, gen->core_relo_cnt + 1, sizeof(*relos));
+	if (!relos) {
+		gen->error = -ENOMEM;
+		return;
+	}
+	gen->core_relos = relos;
+	relos += gen->core_relo_cnt;
+	memcpy(relos, core_relo, sizeof(*relos));
+	gen->core_relo_cnt++;
 }
 
 static void emit_relo(struct bpf_gen *gen, struct ksym_relo_desc *relo, int insns)
···
 		emit_relo(gen, gen->relos + i, insns);
 }
 
+static void cleanup_core_relo(struct bpf_gen *gen)
+{
+	if (!gen->core_relo_cnt)
+		return;
+	free(gen->core_relos);
+	gen->core_relo_cnt = 0;
+	gen->core_relos = NULL;
+}
+
 static void cleanup_relos(struct bpf_gen *gen, int insns)
 {
 	int i, insn;
···
 		gen->relo_cnt = 0;
 		gen->relos = NULL;
 	}
+	cleanup_core_relo(gen);
 }
 
 void bpf_gen__prog_load(struct bpf_gen *gen,
···
 			const char *license, struct bpf_insn *insns, size_t insn_cnt,
 			struct bpf_prog_load_opts *load_attr, int prog_idx)
 {
-	int attr_size = offsetofend(union bpf_attr, fd_array);
-	int prog_load_attr, license_off, insns_off, func_info, line_info;
+	int prog_load_attr, license_off, insns_off, func_info, line_info, core_relos;
+	int attr_size = offsetofend(union bpf_attr, core_relo_rec_size);
 	union bpf_attr attr;
 
 	memset(&attr, 0, attr_size);
-	pr_debug("gen: prog_load: type %d insns_cnt %zd\n", prog_type, insn_cnt);
+	pr_debug("gen: prog_load: type %d insns_cnt %zd progi_idx %d\n",
+		 prog_type, insn_cnt, prog_idx);
 	/* add license string to blob of bytes */
 	license_off = add_data(gen, license, strlen(license) + 1);
 	/* add insns to blob of bytes */
···
 	line_info = add_data(gen, load_attr->line_info,
 			     attr.line_info_cnt * attr.line_info_rec_size);
 
+	attr.core_relo_rec_size = sizeof(struct bpf_core_relo);
+	attr.core_relo_cnt = gen->core_relo_cnt;
+	core_relos = add_data(gen, gen->core_relos,
+			     attr.core_relo_cnt * attr.core_relo_rec_size);
+
 	memcpy(attr.prog_name, prog_name,
 	       min((unsigned)strlen(prog_name), BPF_OBJ_NAME_LEN - 1));
 	prog_load_attr = add_data(gen, &attr, attr_size);
···
 
 	/* 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 with a pointer to core_relos */
+	emit_rel_store(gen, attr_field(prog_load_attr, core_relos), core_relos);
 
 	/* 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);
···
 	debug_ret(gen, "prog_load %s insn_cnt %d", attr.prog_name, attr.insn_cnt);
 	/* successful or not, close btf module FDs used in extern ksyms and attach_btf_obj_fd */
 	cleanup_relos(gen, insns_off);
-	if (gen->attach_kind)
+	if (gen->attach_kind) {
 		emit_sys_close_blob(gen,
 				    attr_field(prog_load_attr, attach_btf_obj_fd));
+		gen->attach_kind = 0;
+	}
 	emit_check_err(gen);
 	/* remember prog_fd in the stack, if successful */
 	emit(gen, BPF_STX_MEM(BPF_W, BPF_REG_10, BPF_REG_7,
···
 	/* emit MAP_UPDATE_ELEM command */
 	emit_sys_bpf(gen, BPF_MAP_UPDATE_ELEM, map_update_attr, attr_size);
 	debug_ret(gen, "update_elem idx %d value_size %d", map_idx, value_size);
+	emit_check_err(gen);
+}
+
+void bpf_gen__populate_outer_map(struct bpf_gen *gen, int outer_map_idx, int slot,
+				 int inner_map_idx)
+{
+	int attr_size = offsetofend(union bpf_attr, flags);
+	int map_update_attr, key;
+	union bpf_attr attr;
+
+	memset(&attr, 0, attr_size);
+	pr_debug("gen: populate_outer_map: outer %d key %d inner %d\n",
+		 outer_map_idx, slot, inner_map_idx);
+
+	key = add_data(gen, &slot, sizeof(slot));
+
+	map_update_attr = add_data(gen, &attr, attr_size);
+	move_blob2blob(gen, attr_field(map_update_attr, map_fd), 4,
+		       blob_fd_array_off(gen, outer_map_idx));
+	emit_rel_store(gen, attr_field(map_update_attr, key), key);
+	emit_rel_store(gen, attr_field(map_update_attr, value),
+		       blob_fd_array_off(gen, inner_map_idx));
+
+	/* emit MAP_UPDATE_ELEM command */
+	emit_sys_bpf(gen, BPF_MAP_UPDATE_ELEM, map_update_attr, attr_size);
+	debug_ret(gen, "populate_outer_map outer %d key %d inner %d",
+		  outer_map_idx, slot, inner_map_idx);
 	emit_check_err(gen);
 }
 

tools/lib/bpf/libbpf.c

···
 	return 0;
 }
 
+__u32 libbpf_major_version(void)
+{
+	return LIBBPF_MAJOR_VERSION;
+}
+
+__u32 libbpf_minor_version(void)
+{
+	return LIBBPF_MINOR_VERSION;
+}
+
+const char *libbpf_version_string(void)
+{
+#define __S(X) #X
+#define _S(X) __S(X)
+	return  "v" _S(LIBBPF_MAJOR_VERSION) "." _S(LIBBPF_MINOR_VERSION);
+#undef _S
+#undef __S
+}
+
 enum kern_feature_id {
 	/* v4.14: kernel support for program & map names. */
 	FEAT_PROG_NAME,
···
 	RELO_EXTERN_VAR,
 	RELO_EXTERN_FUNC,
 	RELO_SUBPROG_ADDR,
+	RELO_CORE,
 };
 
 struct reloc_desc {
 	enum reloc_type type;
 	int insn_idx;
-	int map_idx;
-	int sym_off;
+	union {
+		const struct bpf_core_relo *core_relo; /* used when type == RELO_CORE */
+		struct {
+			int map_idx;
+			int sym_off;
+		};
+	};
 };
 
 struct bpf_sec_def;
···
 
 	struct reloc_desc *reloc_desc;
 	int nr_reloc;
-	int log_level;
+
+	/* BPF verifier log settings */
+	char *log_buf;
+	size_t log_size;
+	__u32 log_level;
 
 	struct {
 		int nr;
···
 	char *pin_path;
 	bool pinned;
 	bool reused;
+	bool skipped;
 	__u64 map_extra;
 };
 
···
 	size_t btf_module_cnt;
 	size_t btf_module_cap;
 
+	/* optional log settings passed to BPF_BTF_LOAD and BPF_PROG_LOAD commands */
+	char *log_buf;
+	size_t log_size;
+	__u32 log_level;
+
 	void *priv;
 	bpf_object_clear_priv_t clear_priv;
 
···
 
 	prog->instances.fds = NULL;
 	prog->instances.nr = -1;
+
+	/* inherit object's log_level */
+	prog->log_level = obj->log_level;
 
 	prog->sec_name = strdup(sec_name);
 	if (!prog->sec_name)
···
 			map_def->parts |= MAP_DEF_VALUE_SIZE | MAP_DEF_VALUE_TYPE;
 		}
 		else if (strcmp(name, "values") == 0) {
+			bool is_map_in_map = bpf_map_type__is_map_in_map(map_def->map_type);
+			bool is_prog_array = map_def->map_type == BPF_MAP_TYPE_PROG_ARRAY;
+			const char *desc = is_map_in_map ? "map-in-map inner" : "prog-array value";
 			char inner_map_name[128];
 			int err;
 
···
 					map_name, name);
 				return -EINVAL;
 			}
-			if (!bpf_map_type__is_map_in_map(map_def->map_type)) {
-				pr_warn("map '%s': should be map-in-map.\n",
+			if (!is_map_in_map && !is_prog_array) {
+				pr_warn("map '%s': should be map-in-map or prog-array.\n",
 					map_name);
 				return -ENOTSUP;
 			}
···
 			map_def->value_size = 4;
 			t = btf__type_by_id(btf, m->type);
 			if (!t) {
-				pr_warn("map '%s': map-in-map inner type [%d] not found.\n",
-					map_name, m->type);
+				pr_warn("map '%s': %s type [%d] not found.\n",
+					map_name, desc, m->type);
 				return -EINVAL;
 			}
 			if (!btf_is_array(t) || btf_array(t)->nelems) {
-				pr_warn("map '%s': map-in-map inner spec is not a zero-sized array.\n",
-					map_name);
+				pr_warn("map '%s': %s spec is not a zero-sized array.\n",
+					map_name, desc);
 				return -EINVAL;
 			}
 			t = skip_mods_and_typedefs(btf, btf_array(t)->type, NULL);
 			if (!btf_is_ptr(t)) {
-				pr_warn("map '%s': map-in-map inner def is of unexpected kind %s.\n",
-					map_name, btf_kind_str(t));
+				pr_warn("map '%s': %s def is of unexpected kind %s.\n",
+					map_name, desc, btf_kind_str(t));
 				return -EINVAL;
 			}
 			t = skip_mods_and_typedefs(btf, t->type, NULL);
+			if (is_prog_array) {
+				if (!btf_is_func_proto(t)) {
+					pr_warn("map '%s': prog-array value def is of unexpected kind %s.\n",
+						map_name, btf_kind_str(t));
+					return -EINVAL;
+				}
+				continue;
+			}
 			if (!btf_is_struct(t)) {
 				pr_warn("map '%s': map-in-map inner def is of unexpected kind %s.\n",
 					map_name, btf_kind_str(t));
···
 		 */
 		btf__set_fd(kern_btf, 0);
 	} else {
-		err = btf__load_into_kernel(kern_btf);
+		/* currently BPF_BTF_LOAD only supports log_level 1 */
+		err = btf_load_into_kernel(kern_btf, obj->log_buf, obj->log_size,
+					   obj->log_level ? 1 : 0);
 	}
 	if (sanitize) {
 		if (!err) {
···
 
 	/* sort BPF programs by section name and in-section instruction offset
 	 * for faster search */
-	qsort(obj->programs, obj->nr_programs, sizeof(*obj->programs), cmp_progs);
+	if (obj->nr_programs)
+		qsort(obj->programs, obj->nr_programs, sizeof(*obj->programs), cmp_progs);
 
 	return bpf_object__init_btf(obj, btf_data, btf_ext_data);
 }
···
 
 static int probe_kern_global_data(void)
 {
-	struct bpf_create_map_attr map_attr;
 	char *cp, errmsg[STRERR_BUFSIZE];
 	struct bpf_insn insns[] = {
 		BPF_LD_MAP_VALUE(BPF_REG_1, 0, 16),
···
 	};
 	int ret, map, insn_cnt = ARRAY_SIZE(insns);
 
-	memset(&map_attr, 0, sizeof(map_attr));
-	map_attr.map_type = BPF_MAP_TYPE_ARRAY;
-	map_attr.key_size = sizeof(int);
-	map_attr.value_size = 32;
-	map_attr.max_entries = 1;
-
-	map = bpf_create_map_xattr(&map_attr);
+	map = bpf_map_create(BPF_MAP_TYPE_ARRAY, NULL, sizeof(int), 32, 1, NULL);
 	if (map < 0) {
 		ret = -errno;
 		cp = libbpf_strerror_r(ret, errmsg, sizeof(errmsg));
···
 
 static int probe_kern_array_mmap(void)
 {
-	struct bpf_create_map_attr attr = {
-		.map_type = BPF_MAP_TYPE_ARRAY,
-		.map_flags = BPF_F_MMAPABLE,
-		.key_size = sizeof(int),
-		.value_size = sizeof(int),
-		.max_entries = 1,
-	};
+	LIBBPF_OPTS(bpf_map_create_opts, opts, .map_flags = BPF_F_MMAPABLE);
+	int fd;
 
-	return probe_fd(bpf_create_map_xattr(&attr));
+	fd = bpf_map_create(BPF_MAP_TYPE_ARRAY, NULL, sizeof(int), sizeof(int), 1, &opts);
+	return probe_fd(fd);
 }
 
 static int probe_kern_exp_attach_type(void)
···
 
 static int probe_prog_bind_map(void)
 {
-	struct bpf_create_map_attr map_attr;
 	char *cp, errmsg[STRERR_BUFSIZE];
 	struct bpf_insn insns[] = {
 		BPF_MOV64_IMM(BPF_REG_0, 0),
···
 	};
 	int ret, map, prog, insn_cnt = ARRAY_SIZE(insns);
 
-	memset(&map_attr, 0, sizeof(map_attr));
-	map_attr.map_type = BPF_MAP_TYPE_ARRAY;
-	map_attr.key_size = sizeof(int);
-	map_attr.value_size = 32;
-	map_attr.max_entries = 1;
-
-	map = bpf_create_map_xattr(&map_attr);
+	map = bpf_map_create(BPF_MAP_TYPE_ARRAY, NULL, sizeof(int), 32, 1, NULL);
 	if (map < 0) {
 		ret = -errno;
 		cp = libbpf_strerror_r(ret, errmsg, sizeof(errmsg));
···
 
 static int bpf_object__create_map(struct bpf_object *obj, struct bpf_map *map, bool is_inner)
 {
-	struct bpf_create_map_params create_attr;
+	LIBBPF_OPTS(bpf_map_create_opts, create_attr);
 	struct bpf_map_def *def = &map->def;
+	const char *map_name = NULL;
+	__u32 max_entries;
 	int err = 0;
 
-	memset(&create_attr, 0, sizeof(create_attr));
-
 	if (kernel_supports(obj, FEAT_PROG_NAME))
-		create_attr.name = map->name;
+		map_name = map->name;
 	create_attr.map_ifindex = map->map_ifindex;
-	create_attr.map_type = def->type;
 	create_attr.map_flags = def->map_flags;
-	create_attr.key_size = def->key_size;
-	create_attr.value_size = def->value_size;
 	create_attr.numa_node = map->numa_node;
 	create_attr.map_extra = map->map_extra;
 
···
 			return nr_cpus;
 		}
 		pr_debug("map '%s': setting size to %d\n", map->name, nr_cpus);
-		create_attr.max_entries = nr_cpus;
+		max_entries = nr_cpus;
 	} else {
-		create_attr.max_entries = def->max_entries;
+		max_entries = def->max_entries;
 	}
 
 	if (bpf_map__is_struct_ops(map))
-		create_attr.btf_vmlinux_value_type_id =
-			map->btf_vmlinux_value_type_id;
+		create_attr.btf_vmlinux_value_type_id = map->btf_vmlinux_value_type_id;
 
-	create_attr.btf_fd = 0;
-	create_attr.btf_key_type_id = 0;
-	create_attr.btf_value_type_id = 0;
 	if (obj->btf && btf__fd(obj->btf) >= 0 && !bpf_map_find_btf_info(obj, map)) {
 		create_attr.btf_fd = btf__fd(obj->btf);
 		create_attr.btf_key_type_id = map->btf_key_type_id;
···
 	}
 
 	if (obj->gen_loader) {
-		bpf_gen__map_create(obj->gen_loader, &create_attr, is_inner ? -1 : map - obj->maps);
+		bpf_gen__map_create(obj->gen_loader, def->type, map_name,
+				    def->key_size, def->value_size, max_entries,
+				    &create_attr, is_inner ? -1 : map - obj->maps);
 		/* Pretend to have valid FD to pass various fd >= 0 checks.
 		 * This fd == 0 will not be used with any syscall and will be reset to -1 eventually.
 		 */
 		map->fd = 0;
 	} else {
-		map->fd = libbpf__bpf_create_map_xattr(&create_attr);
+		map->fd = bpf_map_create(def->type, map_name,
+					 def->key_size, def->value_size,
+					 max_entries, &create_attr);
 	}
 	if (map->fd < 0 && (create_attr.btf_key_type_id ||
 			    create_attr.btf_value_type_id)) {
···
 		create_attr.btf_value_type_id = 0;
 		map->btf_key_type_id = 0;
 		map->btf_value_type_id = 0;
-		map->fd = libbpf__bpf_create_map_xattr(&create_attr);
+		map->fd = bpf_map_create(def->type, map_name,
+					 def->key_size, def->value_size,
+					 max_entries, &create_attr);
 	}
 
 	err = map->fd < 0 ? -errno : 0;
···
 	return err;
 }
 
-static int init_map_slots(struct bpf_object *obj, struct bpf_map *map)
+static int init_map_in_map_slots(struct bpf_object *obj, struct bpf_map *map)
 {
 	const struct bpf_map *targ_map;
 	unsigned int i;
···
 
 		targ_map = map->init_slots[i];
 		fd = bpf_map__fd(targ_map);
+
 		if (obj->gen_loader) {
-			pr_warn("// TODO map_update_elem: idx %td key %d value==map_idx %td\n",
-				map - obj->maps, i, targ_map - obj->maps);
-			return -ENOTSUP;
+			bpf_gen__populate_outer_map(obj->gen_loader,
+						    map - obj->maps, i,
+						    targ_map - obj->maps);
 		} else {
 			err = bpf_map_update_elem(map->fd, &i, &fd, 0);
 		}
 		if (err) {
 			err = -errno;
 			pr_warn("map '%s': failed to initialize slot [%d] to map '%s' fd=%d: %d\n",
-				map->name, i, targ_map->name,
-				fd, err);
+				map->name, i, targ_map->name, fd, err);
 			return err;
 		}
 		pr_debug("map '%s': slot [%d] set to map '%s' fd=%d\n",
···
 	zfree(&map->init_slots);
 	map->init_slots_sz = 0;
 
+	return 0;
+}
+
+static int init_prog_array_slots(struct bpf_object *obj, struct bpf_map *map)
+{
+	const struct bpf_program *targ_prog;
+	unsigned int i;
+	int fd, err;
+
+	if (obj->gen_loader)
+		return -ENOTSUP;
+
+	for (i = 0; i < map->init_slots_sz; i++) {
+		if (!map->init_slots[i])
+			continue;
+
+		targ_prog = map->init_slots[i];
+		fd = bpf_program__fd(targ_prog);
+
+		err = bpf_map_update_elem(map->fd, &i, &fd, 0);
+		if (err) {
+			err = -errno;
+			pr_warn("map '%s': failed to initialize slot [%d] to prog '%s' fd=%d: %d\n",
+				map->name, i, targ_prog->name, fd, err);
+			return err;
+		}
+		pr_debug("map '%s': slot [%d] set to prog '%s' fd=%d\n",
+			 map->name, i, targ_prog->name, fd);
+	}
+
+	zfree(&map->init_slots);
+	map->init_slots_sz = 0;
+
+	return 0;
+}
+
+static int bpf_object_init_prog_arrays(struct bpf_object *obj)
+{
+	struct bpf_map *map;
+	int i, err;
+
+	for (i = 0; i < obj->nr_maps; i++) {
+		map = &obj->maps[i];
+
+		if (!map->init_slots_sz || map->def.type != BPF_MAP_TYPE_PROG_ARRAY)
+			continue;
+
+		err = init_prog_array_slots(obj, map);
+		if (err < 0) {
+			zclose(map->fd);
+			return err;
+		}
+	}
 	return 0;
 }
 
···
 
 	for (i = 0; i < obj->nr_maps; i++) {
 		map = &obj->maps[i];
+
+		/* To support old kernels, we skip creating global data maps
+		 * (.rodata, .data, .kconfig, etc); later on, during program
+		 * loading, if we detect that at least one of the to-be-loaded
+		 * programs is referencing any global data map, we'll error
+		 * out with program name and relocation index logged.
+		 * This approach allows to accommodate Clang emitting
+		 * unnecessary .rodata.str1.1 sections for string literals,
+		 * but also it allows to have CO-RE applications that use
+		 * global variables in some of BPF programs, but not others.
+		 * If those global variable-using programs are not loaded at
+		 * runtime due to bpf_program__set_autoload(prog, false),
+		 * bpf_object loading will succeed just fine even on old
+		 * kernels.
+		 */
+		if (bpf_map__is_internal(map) &&
+		    !kernel_supports(obj, FEAT_GLOBAL_DATA)) {
+			map->skipped = true;
+			continue;
+		}
 
 		retried = false;
 retry:
···
 				}
 			}
 
-			if (map->init_slots_sz) {
-				err = init_map_slots(obj, map);
+			if (map->init_slots_sz && map->def.type != BPF_MAP_TYPE_PROG_ARRAY) {
+				err = init_map_in_map_slots(obj, map);
 				if (err < 0) {
 					zclose(map->fd);
 					goto err_out;
···
 			      struct bpf_core_cand_list *cands)
 {
 	struct bpf_core_cand *new_cands, *cand;
-	const struct btf_type *t;
-	const char *targ_name;
+	const struct btf_type *t, *local_t;
+	const char *targ_name, *local_name;
 	size_t targ_essent_len;
 	int n, i;
+
+	local_t = btf__type_by_id(local_cand->btf, local_cand->id);
+	local_name = btf__str_by_offset(local_cand->btf, local_t->name_off);
 
 	n = btf__type_cnt(targ_btf);
 	for (i = targ_start_id; i < n; i++) {
 		t = btf__type_by_id(targ_btf, i);
-		if (btf_kind(t) != btf_kind(local_cand->t))
+		if (btf_kind(t) != btf_kind(local_t))
 			continue;
 
 		targ_name = btf__name_by_offset(targ_btf, t->name_off);
···
 		if (targ_essent_len != local_essent_len)
 			continue;
 
-		if (strncmp(local_cand->name, targ_name, local_essent_len) != 0)
+		if (strncmp(local_name, targ_name, local_essent_len) != 0)
 			continue;
 
 		pr_debug("CO-RE relocating [%d] %s %s: found target candidate [%d] %s %s in [%s]\n",
-			 local_cand->id, btf_kind_str(local_cand->t),
-			 local_cand->name, i, btf_kind_str(t), targ_name,
+			 local_cand->id, btf_kind_str(local_t),
+			 local_name, i, btf_kind_str(t), targ_name,
 			 targ_btf_name);
 		new_cands = libbpf_reallocarray(cands->cands, cands->len + 1,
 					      sizeof(*cands->cands));
···
 
 		cand = &new_cands[cands->len];
 		cand->btf = targ_btf;
-		cand->t = t;
-		cand->name = targ_name;
 		cand->id = i;
 
 		cands->cands = new_cands;
···
 	struct bpf_core_cand local_cand = {};
 	struct bpf_core_cand_list *cands;
 	const struct btf *main_btf;
+	const struct btf_type *local_t;
+	const char *local_name;
 	size_t local_essent_len;
 	int err, i;
 
 	local_cand.btf = local_btf;
-	local_cand.t = btf__type_by_id(local_btf, local_type_id);
-	if (!local_cand.t)
+	local_cand.id = local_type_id;
+	local_t = btf__type_by_id(local_btf, local_type_id);
+	if (!local_t)
 		return ERR_PTR(-EINVAL);
 
-	local_cand.name = btf__name_by_offset(local_btf, local_cand.t->name_off);
-	if (str_is_empty(local_cand.name))
+	local_name = btf__name_by_offset(local_btf, local_t->name_off);
+	if (str_is_empty(local_name))
 		return ERR_PTR(-EINVAL);
-	local_essent_len = bpf_core_essential_name_len(local_cand.name);
+	local_essent_len = bpf_core_essential_name_len(local_name);
 
 	cands = calloc(1, sizeof(*cands));
 	if (!cands)
···
 	return (void *)(uintptr_t)x;
 }
 
+static int record_relo_core(struct bpf_program *prog,
+			    const struct bpf_core_relo *core_relo, int insn_idx)
+{
+	struct reloc_desc *relos, *relo;
+
+	relos = libbpf_reallocarray(prog->reloc_desc,
+				    prog->nr_reloc + 1, sizeof(*relos));
+	if (!relos)
+		return -ENOMEM;
+	relo = &relos[prog->nr_reloc];
+	relo->type = RELO_CORE;
+	relo->insn_idx = insn_idx;
+	relo->core_relo = core_relo;
+	prog->reloc_desc = relos;
+	prog->nr_reloc++;
+	return 0;
+}
+
 static int bpf_core_apply_relo(struct bpf_program *prog,
 			       const struct bpf_core_relo *relo,
 			       int relo_idx,
 			       const struct btf *local_btf,
 			       struct hashmap *cand_cache)
 {
+	struct bpf_core_spec specs_scratch[3] = {};
 	const void *type_key = u32_as_hash_key(relo->type_id);
 	struct bpf_core_cand_list *cands = NULL;
 	const char *prog_name = prog->name;
···
 		return -EINVAL;
 
 	if (prog->obj->gen_loader) {
-		pr_warn("// TODO core_relo: prog %td insn[%d] %s kind %d\n",
+		const char *spec_str = btf__name_by_offset(local_btf, relo->access_str_off);
+
+		pr_debug("record_relo_core: prog %td insn[%d] %s %s %s final insn_idx %d\n",
 			prog - prog->obj->programs, relo->insn_off / 8,
-			local_name, relo->kind);
-		return -ENOTSUP;
+			btf_kind_str(local_type), local_name, spec_str, insn_idx);
+		return record_relo_core(prog, relo, insn_idx);
 	}
 
-	if (relo->kind != BPF_TYPE_ID_LOCAL &&
+	if (relo->kind != BPF_CORE_TYPE_ID_LOCAL &&
 	    !hashmap__find(cand_cache, type_key, (void **)&cands)) {
 		cands = bpf_core_find_cands(prog->obj, local_btf, local_id);
 		if (IS_ERR(cands)) {
···
 		}
 	}
 
-	return bpf_core_apply_relo_insn(prog_name, insn, insn_idx, relo, relo_idx, local_btf, cands);
+	return bpf_core_apply_relo_insn(prog_name, insn, insn_idx, relo,
+					relo_idx, local_btf, cands, specs_scratch);
 }
 
 static int
···
 				insn[0].src_reg = BPF_PSEUDO_MAP_IDX_VALUE;
 				insn[0].imm = relo->map_idx;
 			} else {
+				const struct bpf_map *map = &obj->maps[relo->map_idx];
+
+				if (map->skipped) {
+					pr_warn("prog '%s': relo #%d: kernel doesn't support global data\n",
+						prog->name, i);
+					return -ENOTSUP;
+				}
 				insn[0].src_reg = BPF_PSEUDO_MAP_VALUE;
 				insn[0].imm = obj->maps[relo->map_idx].fd;
 			}
···
 			break;
 		case RELO_CALL:
 			/* handled already */
+			break;
+		case RELO_CORE:
+			/* will be handled by bpf_program_record_relos() */
 			break;
 		default:
 			pr_warn("prog '%s': relo #%d: bad relo type %d\n",
···
 
 static struct reloc_desc *find_prog_insn_relo(const struct bpf_program *prog, size_t insn_idx)
 {
+	if (!prog->nr_reloc)
+		return NULL;
 	return bsearch(&insn_idx, prog->reloc_desc, prog->nr_reloc,
 		       sizeof(*prog->reloc_desc), cmp_relo_by_insn_idx);
 }
···
 	relos = libbpf_reallocarray(main_prog->reloc_desc, new_cnt, sizeof(*relos));
 	if (!relos)
 		return -ENOMEM;
-	memcpy(relos + main_prog->nr_reloc, subprog->reloc_desc,
-	       sizeof(*relos) * subprog->nr_reloc);
+	if (subprog->nr_reloc)
+		memcpy(relos + main_prog->nr_reloc, subprog->reloc_desc,
+		       sizeof(*relos) * subprog->nr_reloc);
 
 	for (i = main_prog->nr_reloc; i < new_cnt; i++)
 		relos[i].insn_idx += subprog->sub_insn_off;
···
 	}
 }
 
+static int cmp_relocs(const void *_a, const void *_b)
+{
+	const struct reloc_desc *a = _a;
+	const struct reloc_desc *b = _b;
+
+	if (a->insn_idx != b->insn_idx)
+		return a->insn_idx < b->insn_idx ? -1 : 1;
+
+	/* no two relocations should have the same insn_idx, but ... */
+	if (a->type != b->type)
+		return a->type < b->type ? -1 : 1;
+
+	return 0;
+}
+
+static void bpf_object__sort_relos(struct bpf_object *obj)
+{
+	int i;
+
+	for (i = 0; i < obj->nr_programs; i++) {
+		struct bpf_program *p = &obj->programs[i];
+
+		if (!p->nr_reloc)
+			continue;
+
+		qsort(p->reloc_desc, p->nr_reloc, sizeof(*p->reloc_desc), cmp_relocs);
+	}
+}
+
 static int
 bpf_object__relocate(struct bpf_object *obj, const char *targ_btf_path)
 {
···
 				err);
 			return err;
 		}
+		if (obj->gen_loader)
+			bpf_object__sort_relos(obj);
 	}
 
 	/* Before relocating calls pre-process relocations and mark
···
 		 */
 		if (prog_is_subprog(obj, prog))
 			continue;
+		if (!prog->load)
+			continue;
 
 		err = bpf_object__relocate_calls(obj, prog);
 		if (err) {
···
 	for (i = 0; i < obj->nr_programs; i++) {
 		prog = &obj->programs[i];
 		if (prog_is_subprog(obj, prog))
+			continue;
+		if (!prog->load)
 			continue;
 		err = bpf_object__relocate_data(obj, prog);
 		if (err) {
···
 	int i, j, nrels, new_sz;
 	const struct btf_var_secinfo *vi = NULL;
 	const struct btf_type *sec, *var, *def;
-	struct bpf_map *map = NULL, *targ_map;
+	struct bpf_map *map = NULL, *targ_map = NULL;
+	struct bpf_program *targ_prog = NULL;
+	bool is_prog_array, is_map_in_map;
 	const struct btf_member *member;
-	const char *name, *mname;
+	const char *name, *mname, *type;
 	unsigned int moff;
 	Elf64_Sym *sym;
 	Elf64_Rel *rel;
···
 			return -LIBBPF_ERRNO__FORMAT;
 		}
 		name = elf_sym_str(obj, sym->st_name) ?: "<?>";
-		if (sym->st_shndx != obj->efile.btf_maps_shndx) {
-			pr_warn(".maps relo #%d: '%s' isn't a BTF-defined map\n",
-				i, name);
-			return -LIBBPF_ERRNO__RELOC;
-		}
 
 		pr_debug(".maps relo #%d: for %zd value %zd rel->r_offset %zu name %d ('%s')\n",
 			 i, (ssize_t)(rel->r_info >> 32), (size_t)sym->st_value,
···
 			return -EINVAL;
 		}
 
-		if (!bpf_map_type__is_map_in_map(map->def.type))
-			return -EINVAL;
-		if (map->def.type == BPF_MAP_TYPE_HASH_OF_MAPS &&
-		    map->def.key_size != sizeof(int)) {
-			pr_warn(".maps relo #%d: hash-of-maps '%s' should have key size %zu.\n",
-				i, map->name, sizeof(int));
+		is_map_in_map = bpf_map_type__is_map_in_map(map->def.type);
+		is_prog_array = map->def.type == BPF_MAP_TYPE_PROG_ARRAY;
+		type = is_map_in_map ? "map" : "prog";
+		if (is_map_in_map) {
+			if (sym->st_shndx != obj->efile.btf_maps_shndx) {
+				pr_warn(".maps relo #%d: '%s' isn't a BTF-defined map\n",
+					i, name);
+				return -LIBBPF_ERRNO__RELOC;
+			}
+			if (map->def.type == BPF_MAP_TYPE_HASH_OF_MAPS &&
+			    map->def.key_size != sizeof(int)) {
+				pr_warn(".maps relo #%d: hash-of-maps '%s' should have key size %zu.\n",
+					i, map->name, sizeof(int));
+				return -EINVAL;
+			}
+			targ_map = bpf_object__find_map_by_name(obj, name);
+			if (!targ_map) {
+				pr_warn(".maps relo #%d: '%s' isn't a valid map reference\n",
+					i, name);
+				return -ESRCH;
+			}
+		} else if (is_prog_array) {
+			targ_prog = bpf_object__find_program_by_name(obj, name);
+			if (!targ_prog) {
+				pr_warn(".maps relo #%d: '%s' isn't a valid program reference\n",
+					i, name);
+				return -ESRCH;
+			}
+			if (targ_prog->sec_idx != sym->st_shndx ||
+			    targ_prog->sec_insn_off * 8 != sym->st_value ||
+			    prog_is_subprog(obj, targ_prog)) {
+				pr_warn(".maps relo #%d: '%s' isn't an entry-point program\n",
+					i, name);
+				return -LIBBPF_ERRNO__RELOC;
+			}
+		} else {
 			return -EINVAL;
 		}
-
-		targ_map = bpf_object__find_map_by_name(obj, name);
-		if (!targ_map)
-			return -ESRCH;
 
 		var = btf__type_by_id(obj->btf, vi->type);
 		def = skip_mods_and_typedefs(obj->btf, var->type, NULL);
···
 			       (new_sz - map->init_slots_sz) * host_ptr_sz);
 			map->init_slots_sz = new_sz;
 		}
-		map->init_slots[moff] = targ_map;
+		map->init_slots[moff] = is_map_in_map ? (void *)targ_map : (void *)targ_prog;
 
-		pr_debug(".maps relo #%d: map '%s' slot [%d] points to map '%s'\n",
-			 i, map->name, moff, name);
+		pr_debug(".maps relo #%d: map '%s' slot [%d] points to %s '%s'\n",
+			 i, map->name, moff, type, name);
 	}
-
-	return 0;
-}
-
-static int cmp_relocs(const void *_a, const void *_b)
-{
-	const struct reloc_desc *a = _a;
-	const struct reloc_desc *b = _b;
-
-	if (a->insn_idx != b->insn_idx)
-		return a->insn_idx < b->insn_idx ? -1 : 1;
-
-	/* no two relocations should have the same insn_idx, but ... */
-	if (a->type != b->type)
-		return a->type < b->type ? -1 : 1;
 
 	return 0;
 }
···
 			return err;
 	}
 
-	for (i = 0; i < obj->nr_programs; i++) {
-		struct bpf_program *p = &obj->programs[i];
-
-		if (!p->nr_reloc)
-			continue;
-
-		qsort(p->reloc_desc, p->nr_reloc, sizeof(*p->reloc_desc), cmp_relocs);
-	}
+	bpf_object__sort_relos(obj);
 	return 0;
 }
 
···
 	const char *prog_name = NULL;
 	char *cp, errmsg[STRERR_BUFSIZE];
 	size_t log_buf_size = 0;
-	char *log_buf = NULL;
+	char *log_buf = NULL, *tmp;
 	int btf_fd, ret, err;
+	bool own_log_buf = true;
+	__u32 log_level = prog->log_level;
 
 	if (prog->type == BPF_PROG_TYPE_UNSPEC) {
 		/*
···
 	load_attr.expected_attach_type = prog->expected_attach_type;
 	if (kernel_supports(obj, FEAT_PROG_NAME))
 		prog_name = prog->name;
-	load_attr.attach_btf_id = prog->attach_btf_id;
 	load_attr.attach_prog_fd = prog->attach_prog_fd;
 	load_attr.attach_btf_obj_fd = prog->attach_btf_obj_fd;
 	load_attr.attach_btf_id = prog->attach_btf_id;
···
 		load_attr.line_info_rec_size = prog->line_info_rec_size;
 		load_attr.line_info_cnt = prog->line_info_cnt;
 	}
-	load_attr.log_level = prog->log_level;
+	load_attr.log_level = log_level;
 	load_attr.prog_flags = prog->prog_flags;
 	load_attr.fd_array = obj->fd_array;
 
···
 		*prog_fd = -1;
 		return 0;
 	}
-retry_load:
-	if (log_buf_size) {
-		log_buf = malloc(log_buf_size);
-		if (!log_buf)
-			return -ENOMEM;
 
-		*log_buf = 0;
+retry_load:
+	/* if log_level is zero, we don't request logs initiallly even if
+	 * custom log_buf is specified; if the program load fails, then we'll
+	 * bump log_level to 1 and use either custom log_buf or we'll allocate
+	 * our own and retry the load to get details on what failed
+	 */
+	if (log_level) {
+		if (prog->log_buf) {
+			log_buf = prog->log_buf;
+			log_buf_size = prog->log_size;
+			own_log_buf = false;
+		} else if (obj->log_buf) {
+			log_buf = obj->log_buf;
+			log_buf_size = obj->log_size;
+			own_log_buf = false;
+		} else {
+			log_buf_size = max((size_t)BPF_LOG_BUF_SIZE, log_buf_size * 2);
+			tmp = realloc(log_buf, log_buf_size);
+			if (!tmp) {
+				ret = -ENOMEM;
+				goto out;
+			}
+			log_buf = tmp;
+			log_buf[0] = '\0';
+			own_log_buf = true;
+		}
 	}
 
 	load_attr.log_buf = log_buf;
 	load_attr.log_size = log_buf_size;
-	ret = bpf_prog_load(prog->type, prog_name, license, insns, insns_cnt, &load_attr);
+	load_attr.log_level = log_level;
 
+	ret = bpf_prog_load(prog->type, prog_name, license, insns, insns_cnt, &load_attr);
 	if (ret >= 0) {
-		if (log_buf && load_attr.log_level)
-			pr_debug("verifier log:\n%s", log_buf);
+		if (log_level && own_log_buf) {
+			pr_debug("prog '%s': -- BEGIN PROG LOAD LOG --\n%s-- END PROG LOAD LOG --\n",
+				 prog->name, log_buf);
+		}
 
 		if (obj->has_rodata && kernel_supports(obj, FEAT_PROG_BIND_MAP)) {
 			struct bpf_map *map;
···
 
 				if (bpf_prog_bind_map(ret, bpf_map__fd(map), NULL)) {
 					cp = libbpf_strerror_r(errno, errmsg, sizeof(errmsg));
-					pr_warn("prog '%s': failed to bind .rodata map: %s\n",
-						prog->name, cp);
+					pr_warn("prog '%s': failed to bind map '%s': %s\n",
+						prog->name, map->real_name, cp);
 					/* Don't fail hard if can't bind rodata. */
 				}
 			}
···
 		goto out;
 	}
 
-	if (!log_buf || errno == ENOSPC) {
-		log_buf_size = max((size_t)BPF_LOG_BUF_SIZE,
-				   log_buf_size << 1);
-
-		free(log_buf);
+	if (log_level == 0) {
+		log_level = 1;
 		goto retry_load;
 	}
-	ret = errno ? -errno : -LIBBPF_ERRNO__LOAD;
+	/* On ENOSPC, increase log buffer size and retry, unless custom
+	 * log_buf is specified.
+	 * Be careful to not overflow u32, though. Kernel's log buf size limit
+	 * isn't part of UAPI so it can always be bumped to full 4GB. So don't
+	 * multiply by 2 unless we are sure we'll fit within 32 bits.
+	 * Currently, we'll get -EINVAL when we reach (UINT_MAX >> 2).
+	 */
+	if (own_log_buf && errno == ENOSPC && log_buf_size <= UINT_MAX / 2)
+		goto retry_load;
+
+	ret = -errno;
 	cp = libbpf_strerror_r(errno, errmsg, sizeof(errmsg));
-	pr_warn("load bpf program failed: %s\n", cp);
+	pr_warn("prog '%s': BPF program load failed: %s\n", prog->name, cp);
 	pr_perm_msg(ret);
 
-	if (log_buf && log_buf[0] != '\0') {
-		ret = -LIBBPF_ERRNO__VERIFY;
-		pr_warn("-- BEGIN DUMP LOG ---\n");
-		pr_warn("\n%s\n", log_buf);
-		pr_warn("-- END LOG --\n");
-	} else if (insns_cnt >= BPF_MAXINSNS) {
-		pr_warn("Program too large (%d insns), at most %d insns\n",
-			insns_cnt, BPF_MAXINSNS);
-		ret = -LIBBPF_ERRNO__PROG2BIG;
-	} else if (prog->type != BPF_PROG_TYPE_KPROBE) {
-		/* Wrong program type? */
-		int fd;
-
-		load_attr.expected_attach_type = 0;
-		load_attr.log_buf = NULL;
-		load_attr.log_size = 0;
-		fd = bpf_prog_load(BPF_PROG_TYPE_KPROBE, prog_name, license,
-				   insns, insns_cnt, &load_attr);
-		if (fd >= 0) {
-			close(fd);
-			ret = -LIBBPF_ERRNO__PROGTYPE;
-			goto out;
-		}
+	if (own_log_buf && log_buf && log_buf[0] != '\0') {
+		pr_warn("prog '%s': -- BEGIN PROG LOAD LOG --\n%s-- END PROG LOAD LOG --\n",
+			prog->name, log_buf);
+	}
+	if (insns_cnt >= BPF_MAXINSNS) {
+		pr_warn("prog '%s': program too large (%d insns), at most %d insns\n",
+			prog->name, insns_cnt, BPF_MAXINSNS);
 	}
 
 out:
-	free(log_buf);
+	if (own_log_buf)
+		free(log_buf);
 	return ret;
 }
 
-static int bpf_program__record_externs(struct bpf_program *prog)
+static int bpf_program_record_relos(struct bpf_program *prog)
 {
 	struct bpf_object *obj = prog->obj;
 	int i;
···
 					       ext->is_weak, false, BTF_KIND_FUNC,
 					       relo->insn_idx);
 			break;
+		case RELO_CORE: {
+			struct bpf_core_relo cr = {
+				.insn_off = relo->insn_idx * 8,
+				.type_id = relo->core_relo->type_id,
+				.access_str_off = relo->core_relo->access_str_off,
+				.kind = relo->core_relo->kind,
+			};
+
+			bpf_gen__record_relo_core(obj->gen_loader, &cr);
+			break;
+		}
 		default:
 			continue;
 		}
···
 				prog->name, prog->instances.nr);
 		}
 		if (obj->gen_loader)
-			bpf_program__record_externs(prog);
+			bpf_program_record_relos(prog);
 		err = bpf_object_load_prog_instance(obj, prog,
 						    prog->insns, prog->insns_cnt,
 						    license, kern_ver, &fd);
···
 	return 0;
 }
 
-static struct bpf_object *
-__bpf_object__open(const char *path, const void *obj_buf, size_t obj_buf_sz,
-		   const struct bpf_object_open_opts *opts)
+static struct bpf_object *bpf_object_open(const char *path, const void *obj_buf, size_t obj_buf_sz,
+					  const struct bpf_object_open_opts *opts)
 {
 	const char *obj_name, *kconfig, *btf_tmp_path;
 	struct bpf_object *obj;
 	char tmp_name[64];
 	int err;
+	char *log_buf;
+	size_t log_size;
+	__u32 log_level;
 
 	if (elf_version(EV_CURRENT) == EV_NONE) {
 		pr_warn("failed to init libelf for %s\n",
···
 		pr_debug("loading object '%s' from buffer\n", obj_name);
 	}
 
+	log_buf = OPTS_GET(opts, kernel_log_buf, NULL);
+	log_size = OPTS_GET(opts, kernel_log_size, 0);
+	log_level = OPTS_GET(opts, kernel_log_level, 0);
+	if (log_size > UINT_MAX)
+		return ERR_PTR(-EINVAL);
+	if (log_size && !log_buf)
+		return ERR_PTR(-EINVAL);
+
 	obj = bpf_object__new(path, obj_buf, obj_buf_sz, obj_name);
 	if (IS_ERR(obj))
 		return obj;
+
+	obj->log_buf = log_buf;
+	obj->log_size = log_size;
+	obj->log_level = log_level;
 
 	btf_tmp_path = OPTS_GET(opts, btf_custom_path, NULL);
 	if (btf_tmp_path) {
···
 		return NULL;
 
 	pr_debug("loading %s\n", attr->file);
-	return __bpf_object__open(attr->file, NULL, 0, &opts);
+	return bpf_object_open(attr->file, NULL, 0, &opts);
 }
 
 struct bpf_object *bpf_object__open_xattr(struct bpf_object_open_attr *attr)
···
 
 	pr_debug("loading %s\n", path);
 
-	return libbpf_ptr(__bpf_object__open(path, NULL, 0, opts));
+	return libbpf_ptr(bpf_object_open(path, NULL, 0, opts));
 }
 
 struct bpf_object *
···
 	if (!obj_buf || obj_buf_sz == 0)
 		return libbpf_err_ptr(-EINVAL);
 
-	return libbpf_ptr(__bpf_object__open(NULL, obj_buf, obj_buf_sz, opts));
+	return libbpf_ptr(bpf_object_open(NULL, obj_buf, obj_buf_sz, opts));
 }
 
 struct bpf_object *
···
 	if (!obj_buf || obj_buf_sz == 0)
 		return errno = EINVAL, NULL;
 
-	return libbpf_ptr(__bpf_object__open(NULL, obj_buf, obj_buf_sz, &opts));
+	return libbpf_ptr(bpf_object_open(NULL, obj_buf, obj_buf_sz, &opts));
 }
 
 static int bpf_object_unload(struct bpf_object *obj)
···
 	bpf_object__for_each_map(m, obj) {
 		if (!bpf_map__is_internal(m))
 			continue;
-		if (!kernel_supports(obj, FEAT_GLOBAL_DATA)) {
-			pr_warn("kernel doesn't support global data\n");
-			return -ENOTSUP;
-		}
 		if (!kernel_supports(obj, FEAT_ARRAY_MMAP))
 			m->def.map_flags ^= BPF_F_MMAPABLE;
 	}
···
 	return 0;
 }
 
-int bpf_object__load_xattr(struct bpf_object_load_attr *attr)
+static int bpf_object_load(struct bpf_object *obj, int extra_log_level, const char *target_btf_path)
 {
-	struct bpf_object *obj;
 	int err, i;
 
-	if (!attr)
-		return libbpf_err(-EINVAL);
-	obj = attr->obj;
 	if (!obj)
 		return libbpf_err(-EINVAL);
 
···
 	}
 
 	if (obj->gen_loader)
-		bpf_gen__init(obj->gen_loader, attr->log_level, obj->nr_programs, obj->nr_maps);
+		bpf_gen__init(obj->gen_loader, extra_log_level, obj->nr_programs, obj->nr_maps);
 
 	err = bpf_object__probe_loading(obj);
 	err = err ? : bpf_object__load_vmlinux_btf(obj, false);
···
 	err = err ? : bpf_object__sanitize_maps(obj);
 	err = err ? : bpf_object__init_kern_struct_ops_maps(obj);
 	err = err ? : bpf_object__create_maps(obj);
-	err = err ? : bpf_object__relocate(obj, obj->btf_custom_path ? : attr->target_btf_path);
-	err = err ? : bpf_object__load_progs(obj, attr->log_level);
+	err = err ? : bpf_object__relocate(obj, obj->btf_custom_path ? : target_btf_path);
+	err = err ? : bpf_object__load_progs(obj, extra_log_level);
+	err = err ? : bpf_object_init_prog_arrays(obj);
 
 	if (obj->gen_loader) {
 		/* reset FDs */
···
 	return libbpf_err(err);
 }
 
+int bpf_object__load_xattr(struct bpf_object_load_attr *attr)
+{
+	return bpf_object_load(attr->obj, attr->log_level, attr->target_btf_path);
+}
+
 int bpf_object__load(struct bpf_object *obj)
 {
-	struct bpf_object_load_attr attr = {
-		.obj = obj,
-	};
-
-	return bpf_object__load_xattr(&attr);
+	return bpf_object_load(obj, 0, NULL);
 }
 
 static int make_parent_dir(const char *path)
···
 	bpf_object__for_each_map(map, obj) {
 		char *pin_path = NULL;
 		char buf[PATH_MAX];
+
+		if (map->skipped)
+			continue;
 
 		if (path) {
 			int len;
···
 	return prog->prog_flags;
 }
 
-int bpf_program__set_extra_flags(struct bpf_program *prog, __u32 extra_flags)
+int bpf_program__set_flags(struct bpf_program *prog, __u32 flags)
 {
 	if (prog->obj->loaded)
 		return libbpf_err(-EBUSY);
 
-	prog->prog_flags |= extra_flags;
+	prog->prog_flags = flags;
+	return 0;
+}
+
+__u32 bpf_program__log_level(const struct bpf_program *prog)
+{
+	return prog->log_level;
+}
+
+int bpf_program__set_log_level(struct bpf_program *prog, __u32 log_level)
+{
+	if (prog->obj->loaded)
+		return libbpf_err(-EBUSY);
+
+	prog->log_level = log_level;
+	return 0;
+}
+
+const char *bpf_program__log_buf(const struct bpf_program *prog, size_t *log_size)
+{
+	*log_size = prog->log_size;
+	return prog->log_buf;
+}
+
+int bpf_program__set_log_buf(struct bpf_program *prog, char *log_buf, size_t log_size)
+{
+	if (log_size && !log_buf)
+		return -EINVAL;
+	if (prog->log_size > UINT_MAX)
+		return -EINVAL;
+	if (prog->obj->loaded)
+		return -EBUSY;
+
+	prog->log_buf = log_buf;
+	prog->log_size = log_size;
 	return 0;
 }
 

tools/lib/bpf/libbpf.h

···
 extern "C" {
 #endif
 
+LIBBPF_API __u32 libbpf_major_version(void);
+LIBBPF_API __u32 libbpf_minor_version(void);
+LIBBPF_API const char *libbpf_version_string(void);
+
 enum libbpf_errno {
 	__LIBBPF_ERRNO__START = 4000,
 
···
 	 * struct_ops, etc) will need actual kernel BTF at /sys/kernel/btf/vmlinux.
 	 */
 	const char *btf_custom_path;
+	/* Pointer to a buffer for storing kernel logs for applicable BPF
+	 * commands. Valid kernel_log_size has to be specified as well and are
+	 * passed-through to bpf() syscall. Keep in mind that kernel might
+	 * fail operation with -ENOSPC error if provided buffer is too small
+	 * to contain entire log output.
+	 * See the comment below for kernel_log_level for interaction between
+	 * log_buf and log_level settings.
+	 *
+	 * If specified, this log buffer will be passed for:
+	 *   - each BPF progral load (BPF_PROG_LOAD) attempt, unless overriden
+	 *     with bpf_program__set_log() on per-program level, to get
+	 *     BPF verifier log output.
+	 *   - during BPF object's BTF load into kernel (BPF_BTF_LOAD) to get
+	 *     BTF sanity checking log.
+	 *
+	 * Each BPF command (BPF_BTF_LOAD or BPF_PROG_LOAD) will overwrite
+	 * previous contents, so if you need more fine-grained control, set
+	 * per-program buffer with bpf_program__set_log_buf() to preserve each
+	 * individual program's verification log. Keep using kernel_log_buf
+	 * for BTF verification log, if necessary.
+	 */
+	char *kernel_log_buf;
+	size_t kernel_log_size;
+	/*
+	 * Log level can be set independently from log buffer. Log_level=0
+	 * means that libbpf will attempt loading BTF or program without any
+	 * logging requested, but will retry with either its own or custom log
+	 * buffer, if provided, and log_level=1 on any error.
+	 * And vice versa, setting log_level>0 will request BTF or prog
+	 * loading with verbose log from the first attempt (and as such also
+	 * for successfully loaded BTF or program), and the actual log buffer
+	 * could be either libbpf's own auto-allocated log buffer, if
+	 * kernel_log_buffer is NULL, or user-provided custom kernel_log_buf.
+	 * If user didn't provide custom log buffer, libbpf will emit captured
+	 * logs through its print callback.
+	 */
+	__u32 kernel_log_level;
+
+	size_t :0;
 };
-#define bpf_object_open_opts__last_field btf_custom_path
+#define bpf_object_open_opts__last_field kernel_log_level
 
 LIBBPF_API struct bpf_object *bpf_object__open(const char *path);
+
+/**
+ * @brief **bpf_object__open_file()** creates a bpf_object by opening
+ * the BPF ELF object file pointed to by the passed path and loading it
+ * into memory.
+ * @param path BPF object file path
+ * @param opts options for how to load the bpf object, this parameter is
+ * optional and can be set to NULL
+ * @return pointer to the new bpf_object; or NULL is returned on error,
+ * error code is stored in errno
+ */
 LIBBPF_API struct bpf_object *
 bpf_object__open_file(const char *path, const struct bpf_object_open_opts *opts);
+
+/**
+ * @brief **bpf_object__open_mem()** creates a bpf_object by reading
+ * the BPF objects raw bytes from a memory buffer containing a valid
+ * BPF ELF object file.
+ * @param obj_buf pointer to the buffer containing ELF file bytes
+ * @param obj_buf_sz number of bytes in the buffer
+ * @param opts options for how to load the bpf object
+ * @return pointer to the new bpf_object; or NULL is returned on error,
+ * error code is stored in errno
+ */
 LIBBPF_API struct bpf_object *
 bpf_object__open_mem(const void *obj_buf, size_t obj_buf_sz,
 		     const struct bpf_object_open_opts *opts);
···
 
 /* Load/unload object into/from kernel */
 LIBBPF_API int bpf_object__load(struct bpf_object *obj);
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_object__load() instead")
 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);
···
 };
 #define bpf_uprobe_opts__last_field retprobe
 
+/**
+ * @brief **bpf_program__attach_uprobe()** attaches a BPF program
+ * to the userspace function which is found by binary path and
+ * offset. You can optionally specify a particular proccess to attach
+ * to. You can also optionally attach the program to the function
+ * exit instead of entry.
+ *
+ * @param prog BPF program to attach
+ * @param retprobe Attach to function exit
+ * @param pid Process ID to attach the uprobe to, 0 for self (own process),
+ * -1 for all processes
+ * @param binary_path Path to binary that contains the function symbol
+ * @param func_offset Offset within the binary of the function symbol
+ * @return Reference to the newly created BPF link; or NULL is returned on error,
+ * error code is stored in errno
+ */
 LIBBPF_API struct bpf_link *
 bpf_program__attach_uprobe(const struct bpf_program *prog, bool retprobe,
 			   pid_t pid, const char *binary_path,
 			   size_t func_offset);
+
+/**
+ * @brief **bpf_program__attach_uprobe_opts()** is just like
+ * bpf_program__attach_uprobe() except with a options struct
+ * for various configurations.
+ *
+ * @param prog BPF program to attach
+ * @param pid Process ID to attach the uprobe to, 0 for self (own process),
+ * -1 for all processes
+ * @param binary_path Path to binary that contains the function symbol
+ * @param func_offset Offset within the binary of the function symbol
+ * @param opts Options for altering program attachment
+ * @return Reference to the newly created BPF link; or NULL is returned on error,
+ * error code is stored in errno
+ */
 LIBBPF_API struct bpf_link *
 bpf_program__attach_uprobe_opts(const struct bpf_program *prog, pid_t pid,
 				const char *binary_path, size_t func_offset,
···
 				      enum bpf_attach_type type);
 
 LIBBPF_API __u32 bpf_program__flags(const struct bpf_program *prog);
-LIBBPF_API int bpf_program__set_extra_flags(struct bpf_program *prog, __u32 extra_flags);
+LIBBPF_API int bpf_program__set_flags(struct bpf_program *prog, __u32 flags);
+
+/* Per-program log level and log buffer getters/setters.
+ * See bpf_object_open_opts comments regarding log_level and log_buf
+ * interactions.
+ */
+LIBBPF_API __u32 bpf_program__log_level(const struct bpf_program *prog);
+LIBBPF_API int bpf_program__set_log_level(struct bpf_program *prog, __u32 log_level);
+LIBBPF_API const char *bpf_program__log_buf(const struct bpf_program *prog, size_t *log_size);
+LIBBPF_API int bpf_program__set_log_buf(struct bpf_program *prog, char *log_buf, size_t log_size);
 
 LIBBPF_API int
 bpf_program__set_attach_target(struct bpf_program *prog, int attach_prog_fd,
···
 	int prog_flags;
 };
 
+LIBBPF_DEPRECATED_SINCE(0, 8, "use bpf_object__open() and bpf_object__load() instead")
 LIBBPF_API int bpf_prog_load_xattr(const struct bpf_prog_load_attr *attr,
 				   struct bpf_object **pobj, int *prog_fd);
 LIBBPF_DEPRECATED_SINCE(0, 7, "use bpf_object__open() and bpf_object__load() instead")
···
 	struct bpf_object **obj;
 
 	int map_cnt;
-	int map_skel_sz; /* sizeof(struct bpf_skeleton_map) */
+	int map_skel_sz; /* sizeof(struct bpf_map_skeleton) */
 	struct bpf_map_skeleton *maps;
 
 	int prog_cnt;
-	int prog_skel_sz; /* sizeof(struct bpf_skeleton_prog) */
+	int prog_skel_sz; /* sizeof(struct bpf_prog_skeleton) */
 	struct bpf_prog_skeleton *progs;
 };
 

tools/lib/bpf/libbpf.map

···
 	global:
 		bpf_map__map_extra;
 		bpf_map__set_map_extra;
+		bpf_map_create;
 		bpf_object__next_map;
 		bpf_object__next_program;
 		bpf_object__prev_map;
···
 		bpf_program__flags;
 		bpf_program__insn_cnt;
 		bpf_program__insns;
-		bpf_program__set_extra_flags;
+		bpf_program__set_flags;
 		btf__add_btf;
 		btf__add_decl_tag;
 		btf__add_type_tag;
···
 		btf__type_cnt;
 		btf_dump__new;
 		btf_dump__new_deprecated;
+		libbpf_major_version;
+		libbpf_minor_version;
+		libbpf_version_string;
 		perf_buffer__new;
 		perf_buffer__new_deprecated;
 		perf_buffer__new_raw;
 		perf_buffer__new_raw_deprecated;
 } LIBBPF_0.5.0;
+
+LIBBPF_0.7.0 {
+	global:
+		bpf_btf_load;
+		bpf_program__log_buf;
+		bpf_program__log_level;
+		bpf_program__set_log_buf;
+		bpf_program__set_log_level;
+};

tools/lib/bpf/libbpf_common.h

···
 #else
 #define __LIBBPF_MARK_DEPRECATED_0_7(X)
 #endif
+#if __LIBBPF_CURRENT_VERSION_GEQ(0, 8)
+#define __LIBBPF_MARK_DEPRECATED_0_8(X) X
+#else
+#define __LIBBPF_MARK_DEPRECATED_0_8(X)
+#endif
 
 /* This set of internal macros allows to do "function overloading" based on
  * number of arguments provided by used in backwards-compatible way during the

tools/lib/bpf/libbpf_internal.h

···
 struct btf;
 struct btf_type;
 
-struct btf_type *btf_type_by_id(struct btf *btf, __u32 type_id);
+struct btf_type *btf_type_by_id(const struct btf *btf, __u32 type_id);
 const char *btf_kind_str(const struct btf_type *t);
 const struct btf_type *skip_mods_and_typedefs(const struct btf *btf, __u32 id, __u32 *res_id);
 
···
 int parse_cpu_mask_file(const char *fcpu, bool **mask, int *mask_sz);
 int libbpf__load_raw_btf(const char *raw_types, size_t types_len,
 			 const char *str_sec, size_t str_len);
-
-struct bpf_create_map_params {
-	const char *name;
-	enum bpf_map_type map_type;
-	__u32 map_flags;
-	__u32 key_size;
-	__u32 value_size;
-	__u32 max_entries;
-	__u32 numa_node;
-	__u32 btf_fd;
-	__u32 btf_key_type_id;
-	__u32 btf_value_type_id;
-	__u32 map_ifindex;
-	union {
-		__u32 inner_map_fd;
-		__u32 btf_vmlinux_value_type_id;
-	};
-	__u64 map_extra;
-};
-
-int libbpf__bpf_create_map_xattr(const struct bpf_create_map_params *create_attr);
+int btf_load_into_kernel(struct btf *btf, char *log_buf, size_t log_sz, __u32 log_level);
 
 struct btf *btf_get_from_fd(int btf_fd, struct btf *base_btf);
 void btf_get_kernel_prefix_kind(enum bpf_attach_type attach_type,

tools/lib/bpf/libbpf_probes.c

···
 	memcpy(raw_btf + hdr.hdr_len, raw_types, hdr.type_len);
 	memcpy(raw_btf + hdr.hdr_len + hdr.type_len, str_sec, hdr.str_len);
 
-	btf_fd = bpf_load_btf(raw_btf, btf_len, NULL, 0, false);
+	btf_fd = bpf_btf_load(raw_btf, btf_len, NULL);
 
 	free(raw_btf);
 	return btf_fd;
···
 {
 	int key_size, value_size, max_entries, map_flags;
 	__u32 btf_key_type_id = 0, btf_value_type_id = 0;
-	struct bpf_create_map_attr attr = {};
 	int fd = -1, btf_fd = -1, fd_inner;
 
 	key_size	= sizeof(__u32);
···
 
 	if (map_type == BPF_MAP_TYPE_ARRAY_OF_MAPS ||
 	    map_type == BPF_MAP_TYPE_HASH_OF_MAPS) {
+		LIBBPF_OPTS(bpf_map_create_opts, opts);
+
 		/* TODO: probe for device, once libbpf has a function to create
 		 * map-in-map for offload
 		 */
 		if (ifindex)
 			return false;
 
-		fd_inner = bpf_create_map(BPF_MAP_TYPE_HASH,
-					  sizeof(__u32), sizeof(__u32), 1, 0);
+		fd_inner = bpf_map_create(BPF_MAP_TYPE_HASH, NULL,
+					  sizeof(__u32), sizeof(__u32), 1, NULL);
 		if (fd_inner < 0)
 			return false;
-		fd = bpf_create_map_in_map(map_type, NULL, sizeof(__u32),
-					   fd_inner, 1, 0);
+
+		opts.inner_map_fd = fd_inner;
+		fd = bpf_map_create(map_type, NULL, sizeof(__u32), sizeof(__u32), 1, &opts);
 		close(fd_inner);
 	} else {
+		LIBBPF_OPTS(bpf_map_create_opts, opts);
+
 		/* Note: No other restriction on map type probes for offload */
-		attr.map_type = map_type;
-		attr.key_size = key_size;
-		attr.value_size = value_size;
-		attr.max_entries = max_entries;
-		attr.map_flags = map_flags;
-		attr.map_ifindex = ifindex;
+		opts.map_flags = map_flags;
+		opts.map_ifindex = ifindex;
 		if (btf_fd >= 0) {
-			attr.btf_fd = btf_fd;
-			attr.btf_key_type_id = btf_key_type_id;
-			attr.btf_value_type_id = btf_value_type_id;
+			opts.btf_fd = btf_fd;
+			opts.btf_key_type_id = btf_key_type_id;
+			opts.btf_value_type_id = btf_value_type_id;
 		}
 
-		fd = bpf_create_map_xattr(&attr);
+		fd = bpf_map_create(map_type, NULL, key_size, value_size, max_entries, &opts);
 	}
 	if (fd >= 0)
 		close(fd);

tools/lib/bpf/libbpf_version.h

···
 #define __LIBBPF_VERSION_H
 
 #define LIBBPF_MAJOR_VERSION 0
-#define LIBBPF_MINOR_VERSION 6
+#define LIBBPF_MINOR_VERSION 7
 
 #endif /* __LIBBPF_VERSION_H */

tools/lib/bpf/linker.c

···
 	}
 	free(linker->secs);
 
+	free(linker->glob_syms);
 	free(linker);
 }
 
···
 static int linker_append_elf_relos(struct bpf_linker *linker, struct src_obj *obj)
 {
 	struct src_sec *src_symtab = &obj->secs[obj->symtab_sec_idx];
-	struct dst_sec *dst_symtab = &linker->secs[linker->symtab_sec_idx];
+	struct dst_sec *dst_symtab;
 	int i, err;
 
 	for (i = 1; i < obj->sec_cnt; i++) {
···
 			pr_warn("sections %s are not compatible\n", src_sec->sec_name);
 			return -1;
 		}
+
+		/* add_dst_sec() above could have invalidated linker->secs */
+		dst_symtab = &linker->secs[linker->symtab_sec_idx];
 
 		/* shdr->sh_link points to SYMTAB */
 		dst_sec->shdr->sh_link = linker->symtab_sec_idx;

tools/lib/bpf/relo_core.c

···
 // SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause)
 /* Copyright (c) 2019 Facebook */
 
+#ifdef __KERNEL__
+#include <linux/bpf.h>
+#include <linux/btf.h>
+#include <linux/string.h>
+#include <linux/bpf_verifier.h>
+#include "relo_core.h"
+
+static const char *btf_kind_str(const struct btf_type *t)
+{
+	return btf_type_str(t);
+}
+
+static bool is_ldimm64_insn(struct bpf_insn *insn)
+{
+	return insn->code == (BPF_LD | BPF_IMM | BPF_DW);
+}
+
+static const struct btf_type *
+skip_mods_and_typedefs(const struct btf *btf, u32 id, u32 *res_id)
+{
+	return btf_type_skip_modifiers(btf, id, res_id);
+}
+
+static const char *btf__name_by_offset(const struct btf *btf, u32 offset)
+{
+	return btf_name_by_offset(btf, offset);
+}
+
+static s64 btf__resolve_size(const struct btf *btf, u32 type_id)
+{
+	const struct btf_type *t;
+	int size;
+
+	t = btf_type_by_id(btf, type_id);
+	t = btf_resolve_size(btf, t, &size);
+	if (IS_ERR(t))
+		return PTR_ERR(t);
+	return size;
+}
+
+enum libbpf_print_level {
+	LIBBPF_WARN,
+	LIBBPF_INFO,
+	LIBBPF_DEBUG,
+};
+
+#undef pr_warn
+#undef pr_info
+#undef pr_debug
+#define pr_warn(fmt, log, ...)	bpf_log((void *)log, fmt, "", ##__VA_ARGS__)
+#define pr_info(fmt, log, ...)	bpf_log((void *)log, fmt, "", ##__VA_ARGS__)
+#define pr_debug(fmt, log, ...)	bpf_log((void *)log, fmt, "", ##__VA_ARGS__)
+#define libbpf_print(level, fmt, ...)	bpf_log((void *)prog_name, fmt, ##__VA_ARGS__)
+#else
 #include <stdio.h>
 #include <string.h>
 #include <errno.h>
···
 #include "btf.h"
 #include "str_error.h"
 #include "libbpf_internal.h"
-
-#define BPF_CORE_SPEC_MAX_LEN 64
-
-/* represents BPF CO-RE field or array element accessor */
-struct bpf_core_accessor {
-	__u32 type_id;		/* struct/union type or array element type */
-	__u32 idx;		/* field index or array index */
-	const char *name;	/* field name or NULL for array accessor */
-};
-
-struct bpf_core_spec {
-	const struct btf *btf;
-	/* high-level spec: named fields and array indices only */
-	struct bpf_core_accessor spec[BPF_CORE_SPEC_MAX_LEN];
-	/* original unresolved (no skip_mods_or_typedefs) root type ID */
-	__u32 root_type_id;
-	/* CO-RE relocation kind */
-	enum bpf_core_relo_kind relo_kind;
-	/* high-level spec length */
-	int len;
-	/* raw, low-level spec: 1-to-1 with accessor spec string */
-	int raw_spec[BPF_CORE_SPEC_MAX_LEN];
-	/* raw spec length */
-	int raw_len;
-	/* field bit offset represented by spec */
-	__u32 bit_offset;
-};
+#endif
 
 static bool is_flex_arr(const struct btf *btf,
 			const struct bpf_core_accessor *acc,
···
 		return false;
 
 	/* has to be the last member of enclosing struct */
-	t = btf__type_by_id(btf, acc->type_id);
+	t = btf_type_by_id(btf, acc->type_id);
 	return acc->idx == btf_vlen(t) - 1;
 }
 
 static const char *core_relo_kind_str(enum bpf_core_relo_kind kind)
 {
 	switch (kind) {
-	case BPF_FIELD_BYTE_OFFSET: return "byte_off";
-	case BPF_FIELD_BYTE_SIZE: return "byte_sz";
-	case BPF_FIELD_EXISTS: return "field_exists";
-	case BPF_FIELD_SIGNED: return "signed";
-	case BPF_FIELD_LSHIFT_U64: return "lshift_u64";
-	case BPF_FIELD_RSHIFT_U64: return "rshift_u64";
-	case BPF_TYPE_ID_LOCAL: return "local_type_id";
-	case BPF_TYPE_ID_TARGET: return "target_type_id";
-	case BPF_TYPE_EXISTS: return "type_exists";
-	case BPF_TYPE_SIZE: return "type_size";
-	case BPF_ENUMVAL_EXISTS: return "enumval_exists";
-	case BPF_ENUMVAL_VALUE: return "enumval_value";
+	case BPF_CORE_FIELD_BYTE_OFFSET: return "byte_off";
+	case BPF_CORE_FIELD_BYTE_SIZE: return "byte_sz";
+	case BPF_CORE_FIELD_EXISTS: return "field_exists";
+	case BPF_CORE_FIELD_SIGNED: return "signed";
+	case BPF_CORE_FIELD_LSHIFT_U64: return "lshift_u64";
+	case BPF_CORE_FIELD_RSHIFT_U64: return "rshift_u64";
+	case BPF_CORE_TYPE_ID_LOCAL: return "local_type_id";
+	case BPF_CORE_TYPE_ID_TARGET: return "target_type_id";
+	case BPF_CORE_TYPE_EXISTS: return "type_exists";
+	case BPF_CORE_TYPE_SIZE: return "type_size";
+	case BPF_CORE_ENUMVAL_EXISTS: return "enumval_exists";
+	case BPF_CORE_ENUMVAL_VALUE: return "enumval_value";
 	default: return "unknown";
 	}
 }
···
 static bool core_relo_is_field_based(enum bpf_core_relo_kind kind)
 {
 	switch (kind) {
-	case BPF_FIELD_BYTE_OFFSET:
-	case BPF_FIELD_BYTE_SIZE:
-	case BPF_FIELD_EXISTS:
-	case BPF_FIELD_SIGNED:
-	case BPF_FIELD_LSHIFT_U64:
-	case BPF_FIELD_RSHIFT_U64:
+	case BPF_CORE_FIELD_BYTE_OFFSET:
+	case BPF_CORE_FIELD_BYTE_SIZE:
+	case BPF_CORE_FIELD_EXISTS:
+	case BPF_CORE_FIELD_SIGNED:
+	case BPF_CORE_FIELD_LSHIFT_U64:
+	case BPF_CORE_FIELD_RSHIFT_U64:
 		return true;
 	default:
 		return false;
···
 static bool core_relo_is_type_based(enum bpf_core_relo_kind kind)
 {
 	switch (kind) {
-	case BPF_TYPE_ID_LOCAL:
-	case BPF_TYPE_ID_TARGET:
-	case BPF_TYPE_EXISTS:
-	case BPF_TYPE_SIZE:
+	case BPF_CORE_TYPE_ID_LOCAL:
+	case BPF_CORE_TYPE_ID_TARGET:
+	case BPF_CORE_TYPE_EXISTS:
+	case BPF_CORE_TYPE_SIZE:
 		return true;
 	default:
 		return false;
···
 static bool core_relo_is_enumval_based(enum bpf_core_relo_kind kind)
 {
 	switch (kind) {
-	case BPF_ENUMVAL_EXISTS:
-	case BPF_ENUMVAL_VALUE:
+	case BPF_CORE_ENUMVAL_EXISTS:
+	case BPF_CORE_ENUMVAL_VALUE:
 		return true;
 	default:
 		return false;
···
  * Enum value-based relocations (ENUMVAL_EXISTS/ENUMVAL_VALUE) use access
  * string to specify enumerator's value index that need to be relocated.
  */
-static int bpf_core_parse_spec(const struct btf *btf,
+static int bpf_core_parse_spec(const char *prog_name, const struct btf *btf,
 			       __u32 type_id,
 			       const char *spec_str,
 			       enum bpf_core_relo_kind relo_kind,
···
 				return sz;
 			spec->bit_offset += access_idx * sz * 8;
 		} else {
-			pr_warn("relo for [%u] %s (at idx %d) captures type [%d] of unexpected kind %s\n",
-				type_id, spec_str, i, id, btf_kind_str(t));
+			pr_warn("prog '%s': relo for [%u] %s (at idx %d) captures type [%d] of unexpected kind %s\n",
+				prog_name, type_id, spec_str, i, id, btf_kind_str(t));
 			return -EINVAL;
 		}
 	}
···
 		targ_id = btf_array(targ_type)->type;
 		goto recur;
 	default:
-		pr_warn("unexpected kind %d relocated, local [%d], target [%d]\n",
-			btf_kind(local_type), local_id, targ_id);
 		return 0;
 	}
 }
···
 		return 0;
 
 	local_id = local_acc->type_id;
-	local_type = btf__type_by_id(local_btf, local_id);
+	local_type = btf_type_by_id(local_btf, local_id);
 	local_member = btf_members(local_type) + local_acc->idx;
 	local_name = btf__name_by_offset(local_btf, local_member->name_off);
 
···
 
 	*field_sz = 0;
 
-	if (relo->kind == BPF_FIELD_EXISTS) {
+	if (relo->kind == BPF_CORE_FIELD_EXISTS) {
 		*val = spec ? 1 : 0;
 		return 0;
 	}
···
 		return -EUCLEAN; /* request instruction poisoning */
 
 	acc = &spec->spec[spec->len - 1];
-	t = btf__type_by_id(spec->btf, acc->type_id);
+	t = btf_type_by_id(spec->btf, acc->type_id);
 
 	/* a[n] accessor needs special handling */
 	if (!acc->name) {
-		if (relo->kind == BPF_FIELD_BYTE_OFFSET) {
+		if (relo->kind == BPF_CORE_FIELD_BYTE_OFFSET) {
 			*val = spec->bit_offset / 8;
 			/* remember field size for load/store mem size */
 			sz = btf__resolve_size(spec->btf, acc->type_id);
···
 				return -EINVAL;
 			*field_sz = sz;
 			*type_id = acc->type_id;
-		} else if (relo->kind == BPF_FIELD_BYTE_SIZE) {
+		} else if (relo->kind == BPF_CORE_FIELD_BYTE_SIZE) {
 			sz = btf__resolve_size(spec->btf, acc->type_id);
 			if (sz < 0)
 				return -EINVAL;
···
 		*validate = !bitfield;
 
 	switch (relo->kind) {
-	case BPF_FIELD_BYTE_OFFSET:
+	case BPF_CORE_FIELD_BYTE_OFFSET:
 		*val = byte_off;
 		if (!bitfield) {
 			*field_sz = byte_sz;
 			*type_id = field_type_id;
 		}
 		break;
-	case BPF_FIELD_BYTE_SIZE:
+	case BPF_CORE_FIELD_BYTE_SIZE:
 		*val = byte_sz;
 		break;
-	case BPF_FIELD_SIGNED:
+	case BPF_CORE_FIELD_SIGNED:
 		/* enums will be assumed unsigned */
 		*val = btf_is_enum(mt) ||
 		       (btf_int_encoding(mt) & BTF_INT_SIGNED);
 		if (validate)
 			*validate = true; /* signedness is never ambiguous */
 		break;
-	case BPF_FIELD_LSHIFT_U64:
+	case BPF_CORE_FIELD_LSHIFT_U64:
 #if __BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__
 		*val = 64 - (bit_off + bit_sz - byte_off  * 8);
 #else
 		*val = (8 - byte_sz) * 8 + (bit_off - byte_off * 8);
 #endif
 		break;
-	case BPF_FIELD_RSHIFT_U64:
+	case BPF_CORE_FIELD_RSHIFT_U64:
 		*val = 64 - bit_sz;
 		if (validate)
 			*validate = true; /* right shift is never ambiguous */
 		break;
-	case BPF_FIELD_EXISTS:
+	case BPF_CORE_FIELD_EXISTS:
 	default:
 		return -EOPNOTSUPP;
 	}
···
 	}
 
 	switch (relo->kind) {
-	case BPF_TYPE_ID_TARGET:
+	case BPF_CORE_TYPE_ID_TARGET:
 		*val = spec->root_type_id;
 		break;
-	case BPF_TYPE_EXISTS:
+	case BPF_CORE_TYPE_EXISTS:
 		*val = 1;
 		break;
-	case BPF_TYPE_SIZE:
+	case BPF_CORE_TYPE_SIZE:
 		sz = btf__resolve_size(spec->btf, spec->root_type_id);
 		if (sz < 0)
 			return -EINVAL;
 		*val = sz;
 		break;
-	case BPF_TYPE_ID_LOCAL:
-	/* BPF_TYPE_ID_LOCAL is handled specially and shouldn't get here */
+	case BPF_CORE_TYPE_ID_LOCAL:
+	/* BPF_CORE_TYPE_ID_LOCAL is handled specially and shouldn't get here */
 	default:
 		return -EOPNOTSUPP;
 	}
···
 	const struct btf_enum *e;
 
 	switch (relo->kind) {
-	case BPF_ENUMVAL_EXISTS:
+	case BPF_CORE_ENUMVAL_EXISTS:
 		*val = spec ? 1 : 0;
 		break;
-	case BPF_ENUMVAL_VALUE:
+	case BPF_CORE_ENUMVAL_VALUE:
 		if (!spec)
 			return -EUCLEAN; /* request instruction poisoning */
-		t = btf__type_by_id(spec->btf, spec->spec[0].type_id);
+		t = btf_type_by_id(spec->btf, spec->spec[0].type_id);
 		e = btf_enum(t) + spec->spec[0].idx;
 		*val = e->val;
 		break;
···
 		if (res->orig_sz != res->new_sz) {
 			const struct btf_type *orig_t, *new_t;
 
-			orig_t = btf__type_by_id(local_spec->btf, res->orig_type_id);
-			new_t = btf__type_by_id(targ_spec->btf, res->new_type_id);
+			orig_t = btf_type_by_id(local_spec->btf, res->orig_type_id);
+			new_t = btf_type_by_id(targ_spec->btf, res->new_type_id);
 
 			/* There are two use cases in which it's safe to
 			 * adjust load/store's mem size:
···
  * [<type-id>] (<type-name>) + <raw-spec> => <offset>@<spec>,
  * where <spec> is a C-syntax view of recorded field access, e.g.: x.a[3].b
  */
-static void bpf_core_dump_spec(int level, const struct bpf_core_spec *spec)
+static void bpf_core_dump_spec(const char *prog_name, int level, const struct bpf_core_spec *spec)
 {
 	const struct btf_type *t;
 	const struct btf_enum *e;
···
 	int i;
 
 	type_id = spec->root_type_id;
-	t = btf__type_by_id(spec->btf, type_id);
+	t = btf_type_by_id(spec->btf, type_id);
 	s = btf__name_by_offset(spec->btf, t->name_off);
 
 	libbpf_print(level, "[%u] %s %s", type_id, btf_kind_str(t), str_is_empty(s) ? "<anon>" : s);
···
 			     const struct bpf_core_relo *relo,
 			     int relo_idx,
 			     const struct btf *local_btf,
-			     struct bpf_core_cand_list *cands)
+			     struct bpf_core_cand_list *cands,
+			     struct bpf_core_spec *specs_scratch)
 {
-	struct bpf_core_spec local_spec, cand_spec, targ_spec = {};
+	struct bpf_core_spec *local_spec = &specs_scratch[0];
+	struct bpf_core_spec *cand_spec = &specs_scratch[1];
+	struct bpf_core_spec *targ_spec = &specs_scratch[2];
 	struct bpf_core_relo_res cand_res, targ_res;
 	const struct btf_type *local_type;
 	const char *local_name;
···
 	int i, j, err;
 
 	local_id = relo->type_id;
-	local_type = btf__type_by_id(local_btf, local_id);
-	if (!local_type)
-		return -EINVAL;
-
+	local_type = btf_type_by_id(local_btf, local_id);
 	local_name = btf__name_by_offset(local_btf, local_type->name_off);
 	if (!local_name)
 		return -EINVAL;
···
 	if (str_is_empty(spec_str))
 		return -EINVAL;
 
-	err = bpf_core_parse_spec(local_btf, local_id, spec_str, relo->kind, &local_spec);
+	err = bpf_core_parse_spec(prog_name, local_btf, local_id, spec_str,
+				  relo->kind, local_spec);
 	if (err) {
 		pr_warn("prog '%s': relo #%d: parsing [%d] %s %s + %s failed: %d\n",
 			prog_name, relo_idx, local_id, btf_kind_str(local_type),
···
 
 	pr_debug("prog '%s': relo #%d: kind <%s> (%d), spec is ", prog_name,
 		 relo_idx, core_relo_kind_str(relo->kind), relo->kind);
-	bpf_core_dump_spec(LIBBPF_DEBUG, &local_spec);
+	bpf_core_dump_spec(prog_name, LIBBPF_DEBUG, local_spec);
 	libbpf_print(LIBBPF_DEBUG, "\n");
 
 	/* TYPE_ID_LOCAL relo is special and doesn't need candidate search */
-	if (relo->kind == BPF_TYPE_ID_LOCAL) {
+	if (relo->kind == BPF_CORE_TYPE_ID_LOCAL) {
 		targ_res.validate = true;
 		targ_res.poison = false;
-		targ_res.orig_val = local_spec.root_type_id;
-		targ_res.new_val = local_spec.root_type_id;
+		targ_res.orig_val = local_spec->root_type_id;
+		targ_res.new_val = local_spec->root_type_id;
 		goto patch_insn;
 	}
 
···
 
 
 	for (i = 0, j = 0; i < cands->len; i++) {
-		err = bpf_core_spec_match(&local_spec, cands->cands[i].btf,
-					  cands->cands[i].id, &cand_spec);
+		err = bpf_core_spec_match(local_spec, cands->cands[i].btf,
+					  cands->cands[i].id, cand_spec);
 		if (err < 0) {
 			pr_warn("prog '%s': relo #%d: error matching candidate #%d ",
 				prog_name, relo_idx, i);
-			bpf_core_dump_spec(LIBBPF_WARN, &cand_spec);
+			bpf_core_dump_spec(prog_name, LIBBPF_WARN, cand_spec);
 			libbpf_print(LIBBPF_WARN, ": %d\n", err);
 			return err;
 		}
 
 		pr_debug("prog '%s': relo #%d: %s candidate #%d ", prog_name,
 			 relo_idx, err == 0 ? "non-matching" : "matching", i);
-		bpf_core_dump_spec(LIBBPF_DEBUG, &cand_spec);
+		bpf_core_dump_spec(prog_name, LIBBPF_DEBUG, cand_spec);
 		libbpf_print(LIBBPF_DEBUG, "\n");
 
 		if (err == 0)
 			continue;
 
-		err = bpf_core_calc_relo(prog_name, relo, relo_idx, &local_spec, &cand_spec, &cand_res);
+		err = bpf_core_calc_relo(prog_name, relo, relo_idx, local_spec, cand_spec, &cand_res);
 		if (err)
 			return err;
 
 		if (j == 0) {
 			targ_res = cand_res;
-			targ_spec = cand_spec;
-		} else if (cand_spec.bit_offset != targ_spec.bit_offset) {
+			*targ_spec = *cand_spec;
+		} else if (cand_spec->bit_offset != targ_spec->bit_offset) {
 			/* if there are many field relo candidates, they
 			 * should all resolve to the same bit offset
 			 */
 			pr_warn("prog '%s': relo #%d: field offset ambiguity: %u != %u\n",
-				prog_name, relo_idx, cand_spec.bit_offset,
-				targ_spec.bit_offset);
+				prog_name, relo_idx, cand_spec->bit_offset,
+				targ_spec->bit_offset);
 			return -EINVAL;
 		} else if (cand_res.poison != targ_res.poison || cand_res.new_val != targ_res.new_val) {
 			/* all candidates should result in the same relocation
···
 	}
 
 	/*
-	 * For BPF_FIELD_EXISTS relo or when used BPF program has field
+	 * For BPF_CORE_FIELD_EXISTS relo or when used BPF program has field
 	 * existence checks or kernel version/config checks, it's expected
 	 * that we might not find any candidates. In this case, if field
 	 * wasn't found in any candidate, the list of candidates shouldn't
···
 			 prog_name, relo_idx);
 
 		/* calculate single target relo result explicitly */
-		err = bpf_core_calc_relo(prog_name, relo, relo_idx, &local_spec, NULL, &targ_res);
+		err = bpf_core_calc_relo(prog_name, relo, relo_idx, local_spec, NULL, &targ_res);
 		if (err)
 			return err;
 	}

tools/lib/bpf/relo_core.h

···
 #ifndef __RELO_CORE_H
 #define __RELO_CORE_H
 
-/* bpf_core_relo_kind encodes which aspect of captured field/type/enum value
- * has to be adjusted by relocations.
- */
-enum bpf_core_relo_kind {
-	BPF_FIELD_BYTE_OFFSET = 0,	/* field byte offset */
-	BPF_FIELD_BYTE_SIZE = 1,	/* field size in bytes */
-	BPF_FIELD_EXISTS = 2,		/* field existence in target kernel */
-	BPF_FIELD_SIGNED = 3,		/* field signedness (0 - unsigned, 1 - signed) */
-	BPF_FIELD_LSHIFT_U64 = 4,	/* bitfield-specific left bitshift */
-	BPF_FIELD_RSHIFT_U64 = 5,	/* bitfield-specific right bitshift */
-	BPF_TYPE_ID_LOCAL = 6,		/* type ID in local BPF object */
-	BPF_TYPE_ID_TARGET = 7,		/* type ID in target kernel */
-	BPF_TYPE_EXISTS = 8,		/* type existence in target kernel */
-	BPF_TYPE_SIZE = 9,		/* type size in bytes */
-	BPF_ENUMVAL_EXISTS = 10,	/* enum value existence in target kernel */
-	BPF_ENUMVAL_VALUE = 11,		/* enum value integer value */
-};
-
-/* The minimum bpf_core_relo checked by the loader
- *
- * CO-RE relocation captures the following data:
- * - insn_off - instruction offset (in bytes) within a BPF program that needs
- *   its insn->imm field to be relocated with actual field info;
- * - type_id - BTF type ID of the "root" (containing) entity of a relocatable
- *   type or field;
- * - access_str_off - offset into corresponding .BTF string section. String
- *   interpretation depends on specific relocation kind:
- *     - for field-based relocations, string encodes an accessed field using
- *     a sequence of field and array indices, separated by colon (:). It's
- *     conceptually very close to LLVM's getelementptr ([0]) instruction's
- *     arguments for identifying offset to a field.
- *     - for type-based relocations, strings is expected to be just "0";
- *     - for enum value-based relocations, string contains an index of enum
- *     value within its enum type;
- *
- * Example to provide a better feel.
- *
- *   struct sample {
- *       int a;
- *       struct {
- *           int b[10];
- *       };
- *   };
- *
- *   struct sample *s = ...;
- *   int x = &s->a;     // encoded as "0:0" (a is field #0)
- *   int y = &s->b[5];  // encoded as "0:1:0:5" (anon struct is field #1,
- *                      // b is field #0 inside anon struct, accessing elem #5)
- *   int z = &s[10]->b; // encoded as "10:1" (ptr is used as an array)
- *
- * type_id for all relocs in this example  will capture BTF type id of
- * `struct sample`.
- *
- * Such relocation is emitted when using __builtin_preserve_access_index()
- * Clang built-in, passing expression that captures field address, e.g.:
- *
- * bpf_probe_read(&dst, sizeof(dst),
- *		  __builtin_preserve_access_index(&src->a.b.c));
- *
- * In this case Clang will emit field relocation recording necessary data to
- * be able to find offset of embedded `a.b.c` field within `src` struct.
- *
- *   [0] https://llvm.org/docs/LangRef.html#getelementptr-instruction
- */
-struct bpf_core_relo {
-	__u32   insn_off;
-	__u32   type_id;
-	__u32   access_str_off;
-	enum bpf_core_relo_kind kind;
-};
+#include <linux/bpf.h>
 
 struct bpf_core_cand {
 	const struct btf *btf;
-	const struct btf_type *t;
-	const char *name;
 	__u32 id;
 };
 
···
 	int len;
 };
 
+#define BPF_CORE_SPEC_MAX_LEN 64
+
+/* represents BPF CO-RE field or array element accessor */
+struct bpf_core_accessor {
+	__u32 type_id;		/* struct/union type or array element type */
+	__u32 idx;		/* field index or array index */
+	const char *name;	/* field name or NULL for array accessor */
+};
+
+struct bpf_core_spec {
+	const struct btf *btf;
+	/* high-level spec: named fields and array indices only */
+	struct bpf_core_accessor spec[BPF_CORE_SPEC_MAX_LEN];
+	/* original unresolved (no skip_mods_or_typedefs) root type ID */
+	__u32 root_type_id;
+	/* CO-RE relocation kind */
+	enum bpf_core_relo_kind relo_kind;
+	/* high-level spec length */
+	int len;
+	/* raw, low-level spec: 1-to-1 with accessor spec string */
+	int raw_spec[BPF_CORE_SPEC_MAX_LEN];
+	/* raw spec length */
+	int raw_len;
+	/* field bit offset represented by spec */
+	__u32 bit_offset;
+};
+
 int bpf_core_apply_relo_insn(const char *prog_name,
 			     struct bpf_insn *insn, int insn_idx,
 			     const struct bpf_core_relo *relo, int relo_idx,
 			     const struct btf *local_btf,
-			     struct bpf_core_cand_list *cands);
+			     struct bpf_core_cand_list *cands,
+			     struct bpf_core_spec *specs_scratch);
 int bpf_core_types_are_compat(const struct btf *local_btf, __u32 local_id,
 			      const struct btf *targ_btf, __u32 targ_id);
 

tools/lib/bpf/skel_internal.h

···
 #include <sys/syscall.h>
 #include <sys/mman.h>
 
+#ifndef __NR_bpf
+# if defined(__mips__) && defined(_ABIO32)
+#  define __NR_bpf 4355
+# elif defined(__mips__) && defined(_ABIN32)
+#  define __NR_bpf 6319
+# elif defined(__mips__) && defined(_ABI64)
+#  define __NR_bpf 5315
+# endif
+#endif
+
 /* This file is a base header for auto-generated *.lskel.h files.
  * Its contents will change and may become part of auto-generation in the future.
  *
···
 	int map_fd = -1, prog_fd = -1, key = 0, err;
 	union bpf_attr attr;
 
-	map_fd = bpf_create_map_name(BPF_MAP_TYPE_ARRAY, "__loader.map", 4,
-				     opts->data_sz, 1, 0);
+	map_fd = bpf_map_create(BPF_MAP_TYPE_ARRAY, "__loader.map", 4, opts->data_sz, 1, NULL);
 	if (map_fd < 0) {
 		opts->errstr = "failed to create loader map";
 		err = -errno;

tools/lib/bpf/xsk.c

···
 #include "libbpf_internal.h"
 #include "xsk.h"
 
+/* entire xsk.h and xsk.c is going away in libbpf 1.0, so ignore all internal
+ * uses of deprecated APIs
+ */
+#pragma GCC diagnostic ignored "-Wdeprecated-declarations"
+
 #ifndef SOL_XDP
  #define SOL_XDP 283
 #endif
···
 static enum xsk_prog get_xsk_prog(void)
 {
 	enum xsk_prog detected = XSK_PROG_FALLBACK;
-	struct bpf_create_map_attr map_attr;
 	__u32 size_out, retval, duration;
 	char data_in = 0, data_out;
 	struct bpf_insn insns[] = {
···
 	};
 	int prog_fd, map_fd, ret, insn_cnt = ARRAY_SIZE(insns);
 
-	memset(&map_attr, 0, sizeof(map_attr));
-	map_attr.map_type = BPF_MAP_TYPE_XSKMAP;
-	map_attr.key_size = sizeof(int);
-	map_attr.value_size = sizeof(int);
-	map_attr.max_entries = 1;
-
-	map_fd = bpf_create_map_xattr(&map_attr);
+	map_fd = bpf_map_create(BPF_MAP_TYPE_XSKMAP, NULL, sizeof(int), sizeof(int), 1, NULL);
 	if (map_fd < 0)
 		return detected;
 
···
 	if (max_queues < 0)
 		return max_queues;
 
-	fd = bpf_create_map_name(BPF_MAP_TYPE_XSKMAP, "xsks_map",
-				 sizeof(int), sizeof(int), max_queues, 0);
+	fd = bpf_map_create(BPF_MAP_TYPE_XSKMAP, "xsks_map",
+			    sizeof(int), sizeof(int), max_queues, NULL);
 	if (fd < 0)
 		return fd;
 

tools/perf/tests/bpf.c

···
 		return err;
 	}
 
+/* temporarily disable libbpf deprecation warnings */
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wdeprecated-declarations"
 	err = bpf_load_program(BPF_PROG_TYPE_KPROBE, insns,
 			       ARRAY_SIZE(insns),
 			       license, kver_int, NULL, 0);
+#pragma GCC diagnostic pop
 	if (err < 0) {
 		pr_err("Missing basic BPF support, skip this test: %s\n",
 		       strerror(errno));

tools/perf/util/bpf-loader.c

···
 
 #include <internal/xyarray.h>
 
+/* temporarily disable libbpf deprecation warnings */
+#pragma GCC diagnostic ignored "-Wdeprecated-declarations"
+
 static int libbpf_perf_print(enum libbpf_print_level level __attribute__((unused)),
 			      const char *fmt, va_list args)
 {

tools/perf/util/bpf_counter.c

···
 		(map_info.value_size == sizeof(struct perf_event_attr_map_entry));
 }
 
+int __weak
+bpf_map_create(enum bpf_map_type map_type,
+	       const char *map_name __maybe_unused,
+	       __u32 key_size,
+	       __u32 value_size,
+	       __u32 max_entries,
+	       const struct bpf_map_create_opts *opts __maybe_unused)
+{
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wdeprecated-declarations"
+	return bpf_create_map(map_type, key_size, value_size, max_entries, 0);
+#pragma GCC diagnostic pop
+}
+
 static int bperf_lock_attr_map(struct target *target)
 {
 	char path[PATH_MAX];
···
 	}
 
 	if (access(path, F_OK)) {
-		map_fd = bpf_create_map(BPF_MAP_TYPE_HASH,
+		map_fd = bpf_map_create(BPF_MAP_TYPE_HASH, NULL,
 					sizeof(struct perf_event_attr),
 					sizeof(struct perf_event_attr_map_entry),
-					ATTR_MAP_SIZE, 0);
+					ATTR_MAP_SIZE, NULL);
 		if (map_fd < 0)
 			return -1;
 

tools/testing/selftests/bpf/Makefile

···
 
 $(TEST_GEN_PROGS) $(TEST_GEN_PROGS_EXTENDED): $(BPFOBJ)
 
-$(OUTPUT)/test_dev_cgroup: cgroup_helpers.c testing_helpers.o
-$(OUTPUT)/test_skb_cgroup_id_user: cgroup_helpers.c testing_helpers.o
-$(OUTPUT)/test_sock: cgroup_helpers.c testing_helpers.o
-$(OUTPUT)/test_sock_addr: cgroup_helpers.c testing_helpers.o
-$(OUTPUT)/test_sockmap: cgroup_helpers.c testing_helpers.o
-$(OUTPUT)/test_tcpnotify_user: cgroup_helpers.c trace_helpers.c testing_helpers.o
-$(OUTPUT)/get_cgroup_id_user: cgroup_helpers.c testing_helpers.o
-$(OUTPUT)/test_cgroup_storage: cgroup_helpers.c testing_helpers.o
-$(OUTPUT)/test_sock_fields: cgroup_helpers.c testing_helpers.o
-$(OUTPUT)/test_sysctl: cgroup_helpers.c testing_helpers.o
-$(OUTPUT)/test_tag: testing_helpers.o
-$(OUTPUT)/test_lirc_mode2_user: testing_helpers.o
-$(OUTPUT)/xdping: testing_helpers.o
-$(OUTPUT)/flow_dissector_load: testing_helpers.o
-$(OUTPUT)/test_maps: testing_helpers.o
-$(OUTPUT)/test_verifier: testing_helpers.o
+CGROUP_HELPERS	:= $(OUTPUT)/cgroup_helpers.o
+TESTING_HELPERS	:= $(OUTPUT)/testing_helpers.o
+TRACE_HELPERS	:= $(OUTPUT)/trace_helpers.o
+
+$(OUTPUT)/test_dev_cgroup: $(CGROUP_HELPERS) $(TESTING_HELPERS)
+$(OUTPUT)/test_skb_cgroup_id_user: $(CGROUP_HELPERS) $(TESTING_HELPERS)
+$(OUTPUT)/test_sock: $(CGROUP_HELPERS) $(TESTING_HELPERS)
+$(OUTPUT)/test_sock_addr: $(CGROUP_HELPERS) $(TESTING_HELPERS)
+$(OUTPUT)/test_sockmap: $(CGROUP_HELPERS) $(TESTING_HELPERS)
+$(OUTPUT)/test_tcpnotify_user: $(CGROUP_HELPERS) $(TESTING_HELPERS) $(TRACE_HELPERS)
+$(OUTPUT)/get_cgroup_id_user: $(CGROUP_HELPERS) $(TESTING_HELPERS)
+$(OUTPUT)/test_cgroup_storage: $(CGROUP_HELPERS) $(TESTING_HELPERS)
+$(OUTPUT)/test_sock_fields: $(CGROUP_HELPERS) $(TESTING_HELPERS)
+$(OUTPUT)/test_sysctl: $(CGROUP_HELPERS) $(TESTING_HELPERS)
+$(OUTPUT)/test_tag: $(TESTING_HELPERS)
+$(OUTPUT)/test_lirc_mode2_user: $(TESTING_HELPERS)
+$(OUTPUT)/xdping: $(TESTING_HELPERS)
+$(OUTPUT)/flow_dissector_load: $(TESTING_HELPERS)
+$(OUTPUT)/test_maps: $(TESTING_HELPERS)
+$(OUTPUT)/test_verifier: $(TESTING_HELPERS)
 
 BPFTOOL ?= $(DEFAULT_BPFTOOL)
 $(DEFAULT_BPFTOOL): $(wildcard $(BPFTOOLDIR)/*.[ch] $(BPFTOOLDIR)/Makefile)    \
···
 		linked_vars.skel.h linked_maps.skel.h
 
 LSKELS := kfunc_call_test.c fentry_test.c fexit_test.c fexit_sleep.c \
-	test_ringbuf.c atomics.c trace_printk.c trace_vprintk.c
+	test_ringbuf.c atomics.c trace_printk.c trace_vprintk.c \
+	map_ptr_kern.c core_kern.c
 # Generate both light skeleton and libbpf skeleton for these
-LSKELS_EXTRA := test_ksyms_module.c test_ksyms_weak.c
+LSKELS_EXTRA := test_ksyms_module.c test_ksyms_weak.c kfunc_call_test_subprog.c
 SKEL_BLACKLIST += $$(LSKELS)
 
 test_static_linked.skel.h-deps := test_static_linked1.o test_static_linked2.o
···
 $(OUTPUT)/bench_ringbufs.o: $(OUTPUT)/ringbuf_bench.skel.h \
 			    $(OUTPUT)/perfbuf_bench.skel.h
 $(OUTPUT)/bench_bloom_filter_map.o: $(OUTPUT)/bloom_filter_bench.skel.h
+$(OUTPUT)/bench_bpf_loop.o: $(OUTPUT)/bpf_loop_bench.skel.h
 $(OUTPUT)/bench.o: bench.h testing_helpers.h $(BPFOBJ)
 $(OUTPUT)/bench: LDLIBS += -lm
-$(OUTPUT)/bench: $(OUTPUT)/bench.o $(OUTPUT)/testing_helpers.o \
+$(OUTPUT)/bench: $(OUTPUT)/bench.o \
+		 $(TESTING_HELPERS) \
+		 $(TRACE_HELPERS) \
 		 $(OUTPUT)/bench_count.o \
 		 $(OUTPUT)/bench_rename.o \
 		 $(OUTPUT)/bench_trigger.o \
 		 $(OUTPUT)/bench_ringbufs.o \
-		 $(OUTPUT)/bench_bloom_filter_map.o
+		 $(OUTPUT)/bench_bloom_filter_map.o \
+		 $(OUTPUT)/bench_bpf_loop.o
 	$(call msg,BINARY,,$@)
 	$(Q)$(CC) $(LDFLAGS) $(filter %.a %.o,$^) $(LDLIBS) -o $@
 

tools/testing/selftests/bpf/bench.c

···
 	       total_ops_mean, total_ops_stddev);
 }
 
+void ops_report_progress(int iter, struct bench_res *res, long delta_ns)
+{
+	double hits_per_sec, hits_per_prod;
+
+	hits_per_sec = res->hits / 1000000.0 / (delta_ns / 1000000000.0);
+	hits_per_prod = hits_per_sec / env.producer_cnt;
+
+	printf("Iter %3d (%7.3lfus): ", iter, (delta_ns - 1000000000) / 1000.0);
+
+	printf("hits %8.3lfM/s (%7.3lfM/prod)\n", hits_per_sec, hits_per_prod);
+}
+
+void ops_report_final(struct bench_res res[], int res_cnt)
+{
+	double hits_mean = 0.0, hits_stddev = 0.0;
+	int i;
+
+	for (i = 0; i < res_cnt; i++)
+		hits_mean += res[i].hits / 1000000.0 / (0.0 + res_cnt);
+
+	if (res_cnt > 1)  {
+		for (i = 0; i < res_cnt; i++)
+			hits_stddev += (hits_mean - res[i].hits / 1000000.0) *
+				       (hits_mean - res[i].hits / 1000000.0) /
+				       (res_cnt - 1.0);
+
+		hits_stddev = sqrt(hits_stddev);
+	}
+	printf("Summary: throughput %8.3lf \u00B1 %5.3lf M ops/s (%7.3lfM ops/prod), ",
+	       hits_mean, hits_stddev, hits_mean / env.producer_cnt);
+	printf("latency %8.3lf ns/op\n", 1000.0 / hits_mean * env.producer_cnt);
+}
+
 const char *argp_program_version = "benchmark";
 const char *argp_program_bug_address = "<[e-mail]>";
 const char argp_program_doc[] =
···
 
 extern struct argp bench_ringbufs_argp;
 extern struct argp bench_bloom_map_argp;
+extern struct argp bench_bpf_loop_argp;
 
 static const struct argp_child bench_parsers[] = {
 	{ &bench_ringbufs_argp, 0, "Ring buffers benchmark", 0 },
 	{ &bench_bloom_map_argp, 0, "Bloom filter map benchmark", 0 },
+	{ &bench_bpf_loop_argp, 0, "bpf_loop helper benchmark", 0 },
 	{},
 };
 
···
 extern const struct bench bench_trig_fentry;
 extern const struct bench bench_trig_fentry_sleep;
 extern const struct bench bench_trig_fmodret;
+extern const struct bench bench_trig_uprobe_base;
+extern const struct bench bench_trig_uprobe_with_nop;
+extern const struct bench bench_trig_uretprobe_with_nop;
+extern const struct bench bench_trig_uprobe_without_nop;
+extern const struct bench bench_trig_uretprobe_without_nop;
 extern const struct bench bench_rb_libbpf;
 extern const struct bench bench_rb_custom;
 extern const struct bench bench_pb_libbpf;
···
 extern const struct bench bench_bloom_false_positive;
 extern const struct bench bench_hashmap_without_bloom;
 extern const struct bench bench_hashmap_with_bloom;
+extern const struct bench bench_bpf_loop;
 
 static const struct bench *benchs[] = {
 	&bench_count_global,
···
 	&bench_trig_fentry,
 	&bench_trig_fentry_sleep,
 	&bench_trig_fmodret,
+	&bench_trig_uprobe_base,
+	&bench_trig_uprobe_with_nop,
+	&bench_trig_uretprobe_with_nop,
+	&bench_trig_uprobe_without_nop,
+	&bench_trig_uretprobe_without_nop,
 	&bench_rb_libbpf,
 	&bench_rb_custom,
 	&bench_pb_libbpf,
···
 	&bench_bloom_false_positive,
 	&bench_hashmap_without_bloom,
 	&bench_hashmap_with_bloom,
+	&bench_bpf_loop,
 };
 
 static void setup_benchmark()

tools/testing/selftests/bpf/bench.h

···
 void hits_drops_report_final(struct bench_res res[], int res_cnt);
 void false_hits_report_progress(int iter, struct bench_res *res, long delta_ns);
 void false_hits_report_final(struct bench_res res[], int res_cnt);
+void ops_report_progress(int iter, struct bench_res *res, long delta_ns);
+void ops_report_final(struct bench_res res[], int res_cnt);
 
 static inline __u64 get_time_ns() {
 	struct timespec t;

tools/testing/selftests/bpf/benchs/bench_bpf_loop.c

···
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2021 Facebook */
+
+#include <argp.h>
+#include "bench.h"
+#include "bpf_loop_bench.skel.h"
+
+/* BPF triggering benchmarks */
+static struct ctx {
+	struct bpf_loop_bench *skel;
+} ctx;
+
+static struct {
+	__u32 nr_loops;
+} args = {
+	.nr_loops = 10,
+};
+
+enum {
+	ARG_NR_LOOPS = 4000,
+};
+
+static const struct argp_option opts[] = {
+	{ "nr_loops", ARG_NR_LOOPS, "nr_loops", 0,
+		"Set number of loops for the bpf_loop helper"},
+	{},
+};
+
+static error_t parse_arg(int key, char *arg, struct argp_state *state)
+{
+	switch (key) {
+	case ARG_NR_LOOPS:
+		args.nr_loops = strtol(arg, NULL, 10);
+		break;
+	default:
+		return ARGP_ERR_UNKNOWN;
+	}
+
+	return 0;
+}
+
+/* exported into benchmark runner */
+const struct argp bench_bpf_loop_argp = {
+	.options = opts,
+	.parser = parse_arg,
+};
+
+static void validate(void)
+{
+	if (env.consumer_cnt != 1) {
+		fprintf(stderr, "benchmark doesn't support multi-consumer!\n");
+		exit(1);
+	}
+}
+
+static void *producer(void *input)
+{
+	while (true)
+		/* trigger the bpf program */
+		syscall(__NR_getpgid);
+
+	return NULL;
+}
+
+static void *consumer(void *input)
+{
+	return NULL;
+}
+
+static void measure(struct bench_res *res)
+{
+	res->hits = atomic_swap(&ctx.skel->bss->hits, 0);
+}
+
+static void setup(void)
+{
+	struct bpf_link *link;
+
+	setup_libbpf();
+
+	ctx.skel = bpf_loop_bench__open_and_load();
+	if (!ctx.skel) {
+		fprintf(stderr, "failed to open skeleton\n");
+		exit(1);
+	}
+
+	link = bpf_program__attach(ctx.skel->progs.benchmark);
+	if (!link) {
+		fprintf(stderr, "failed to attach program!\n");
+		exit(1);
+	}
+
+	ctx.skel->bss->nr_loops = args.nr_loops;
+}
+
+const struct bench bench_bpf_loop = {
+	.name = "bpf-loop",
+	.validate = validate,
+	.setup = setup,
+	.producer_thread = producer,
+	.consumer_thread = consumer,
+	.measure = measure,
+	.report_progress = ops_report_progress,
+	.report_final = ops_report_final,
+};

tools/testing/selftests/bpf/benchs/bench_trigger.c

···
 /* Copyright (c) 2020 Facebook */
 #include "bench.h"
 #include "trigger_bench.skel.h"
+#include "trace_helpers.h"
 
 /* BPF triggering benchmarks */
 static struct trigger_ctx {
···
 	return NULL;
 }
 
+/* make sure call is not inlined and not avoided by compiler, so __weak and
+ * inline asm volatile in the body of the function
+ *
+ * There is a performance difference between uprobing at nop location vs other
+ * instructions. So use two different targets, one of which starts with nop
+ * and another doesn't.
+ *
+ * GCC doesn't generate stack setup preample for these functions due to them
+ * having no input arguments and doing nothing in the body.
+ */
+__weak void uprobe_target_with_nop(void)
+{
+	asm volatile ("nop");
+}
+
+__weak void uprobe_target_without_nop(void)
+{
+	asm volatile ("");
+}
+
+static void *uprobe_base_producer(void *input)
+{
+	while (true) {
+		uprobe_target_with_nop();
+		atomic_inc(&base_hits.value);
+	}
+	return NULL;
+}
+
+static void *uprobe_producer_with_nop(void *input)
+{
+	while (true)
+		uprobe_target_with_nop();
+	return NULL;
+}
+
+static void *uprobe_producer_without_nop(void *input)
+{
+	while (true)
+		uprobe_target_without_nop();
+	return NULL;
+}
+
+static void usetup(bool use_retprobe, bool use_nop)
+{
+	size_t uprobe_offset;
+	ssize_t base_addr;
+	struct bpf_link *link;
+
+	setup_libbpf();
+
+	ctx.skel = trigger_bench__open_and_load();
+	if (!ctx.skel) {
+		fprintf(stderr, "failed to open skeleton\n");
+		exit(1);
+	}
+
+	base_addr = get_base_addr();
+	if (use_nop)
+		uprobe_offset = get_uprobe_offset(&uprobe_target_with_nop, base_addr);
+	else
+		uprobe_offset = get_uprobe_offset(&uprobe_target_without_nop, base_addr);
+
+	link = bpf_program__attach_uprobe(ctx.skel->progs.bench_trigger_uprobe,
+					  use_retprobe,
+					  -1 /* all PIDs */,
+					  "/proc/self/exe",
+					  uprobe_offset);
+	if (!link) {
+		fprintf(stderr, "failed to attach uprobe!\n");
+		exit(1);
+	}
+	ctx.skel->links.bench_trigger_uprobe = link;
+}
+
+static void uprobe_setup_with_nop()
+{
+	usetup(false, true);
+}
+
+static void uretprobe_setup_with_nop()
+{
+	usetup(true, true);
+}
+
+static void uprobe_setup_without_nop()
+{
+	usetup(false, false);
+}
+
+static void uretprobe_setup_without_nop()
+{
+	usetup(true, false);
+}
+
 const struct bench bench_trig_base = {
 	.name = "trig-base",
 	.validate = trigger_validate,
···
 	.validate = trigger_validate,
 	.setup = trigger_fmodret_setup,
 	.producer_thread = trigger_producer,
+	.consumer_thread = trigger_consumer,
+	.measure = trigger_measure,
+	.report_progress = hits_drops_report_progress,
+	.report_final = hits_drops_report_final,
+};
+
+const struct bench bench_trig_uprobe_base = {
+	.name = "trig-uprobe-base",
+	.setup = NULL, /* no uprobe/uretprobe is attached */
+	.producer_thread = uprobe_base_producer,
+	.consumer_thread = trigger_consumer,
+	.measure = trigger_base_measure,
+	.report_progress = hits_drops_report_progress,
+	.report_final = hits_drops_report_final,
+};
+
+const struct bench bench_trig_uprobe_with_nop = {
+	.name = "trig-uprobe-with-nop",
+	.setup = uprobe_setup_with_nop,
+	.producer_thread = uprobe_producer_with_nop,
+	.consumer_thread = trigger_consumer,
+	.measure = trigger_measure,
+	.report_progress = hits_drops_report_progress,
+	.report_final = hits_drops_report_final,
+};
+
+const struct bench bench_trig_uretprobe_with_nop = {
+	.name = "trig-uretprobe-with-nop",
+	.setup = uretprobe_setup_with_nop,
+	.producer_thread = uprobe_producer_with_nop,
+	.consumer_thread = trigger_consumer,
+	.measure = trigger_measure,
+	.report_progress = hits_drops_report_progress,
+	.report_final = hits_drops_report_final,
+};
+
+const struct bench bench_trig_uprobe_without_nop = {
+	.name = "trig-uprobe-without-nop",
+	.setup = uprobe_setup_without_nop,
+	.producer_thread = uprobe_producer_without_nop,
+	.consumer_thread = trigger_consumer,
+	.measure = trigger_measure,
+	.report_progress = hits_drops_report_progress,
+	.report_final = hits_drops_report_final,
+};
+
+const struct bench bench_trig_uretprobe_without_nop = {
+	.name = "trig-uretprobe-without-nop",
+	.setup = uretprobe_setup_without_nop,
+	.producer_thread = uprobe_producer_without_nop,
 	.consumer_thread = trigger_consumer,
 	.measure = trigger_measure,
 	.report_progress = hits_drops_report_progress,

tools/testing/selftests/bpf/benchs/run_bench_bpf_loop.sh

···
+#!/bin/bash
+# SPDX-License-Identifier: GPL-2.0
+
+source ./benchs/run_common.sh
+
+set -eufo pipefail
+
+for t in 1 4 8 12 16; do
+for i in 10 100 500 1000 5000 10000 50000 100000 500000 1000000; do
+subtitle "nr_loops: $i, nr_threads: $t"
+	summarize_ops "bpf_loop: " \
+	    "$($RUN_BENCH -p $t --nr_loops $i bpf-loop)"
+	printf "\n"
+done
+done

tools/testing/selftests/bpf/benchs/run_common.sh

···
 	echo "$*" | sed -E "s/.*Percentage\s=\s+([0-9]+\.[0-9]+).*/\1/"
 }
 
+function ops()
+{
+	echo -n "throughput: "
+	echo -n "$*" | sed -E "s/.*throughput\s+([0-9]+\.[0-9]+ ± [0-9]+\.[0-9]+\sM\sops\/s).*/\1/"
+	echo -n -e ", latency: "
+	echo "$*" | sed -E "s/.*latency\s+([0-9]+\.[0-9]+\sns\/op).*/\1/"
+}
+
 function total()
 {
 	echo "$*" | sed -E "s/.*total operations\s+([0-9]+\.[0-9]+ ± [0-9]+\.[0-9]+M\/s).*/\1/"
···
 	bench="$1"
 	summary=$(echo $2 | tail -n1)
 	printf "%-20s %s%%\n" "$bench" "$(percentage $summary)"
+}
+
+function summarize_ops()
+{
+	bench="$1"
+	summary=$(echo $2 | tail -n1)
+	printf "%-20s %s\n" "$bench" "$(ops $summary)"
 }
 
 function summarize_total()

tools/testing/selftests/bpf/map_tests/array_map_batch_ops.c

···
 
 static void __test_map_lookup_and_update_batch(bool is_pcpu)
 {
-	struct bpf_create_map_attr xattr = {
-		.name = "array_map",
-		.map_type = is_pcpu ? BPF_MAP_TYPE_PERCPU_ARRAY :
-				      BPF_MAP_TYPE_ARRAY,
-		.key_size = sizeof(int),
-		.value_size = sizeof(__s64),
-	};
 	int map_fd, *keys, *visited;
 	__u32 count, total, total_success;
 	const __u32 max_entries = 10;
···
 		.flags = 0,
 	);
 
-	xattr.max_entries = max_entries;
-	map_fd = bpf_create_map_xattr(&xattr);
+	map_fd = bpf_map_create(is_pcpu ? BPF_MAP_TYPE_PERCPU_ARRAY : BPF_MAP_TYPE_ARRAY,
+				"array_map", sizeof(int), sizeof(__s64), max_entries, NULL);
 	CHECK(map_fd == -1,
-	      "bpf_create_map_xattr()", "error:%s\n", strerror(errno));
+	      "bpf_map_create()", "error:%s\n", strerror(errno));
 
 	value_size = sizeof(__s64);
 	if (is_pcpu)

tools/testing/selftests/bpf/map_tests/htab_map_batch_ops.c

···
 	int err, step, value_size;
 	bool nospace_err;
 	void *values;
-	struct bpf_create_map_attr xattr = {
-		.name = "hash_map",
-		.map_type = is_pcpu ? BPF_MAP_TYPE_PERCPU_HASH :
-			    BPF_MAP_TYPE_HASH,
-		.key_size = sizeof(int),
-		.value_size = sizeof(int),
-	};
 	DECLARE_LIBBPF_OPTS(bpf_map_batch_opts, opts,
 		.elem_flags = 0,
 		.flags = 0,
 	);
 
-	xattr.max_entries = max_entries;
-	map_fd = bpf_create_map_xattr(&xattr);
+	map_fd = bpf_map_create(is_pcpu ? BPF_MAP_TYPE_PERCPU_HASH : BPF_MAP_TYPE_HASH,
+				"hash_map", sizeof(int), sizeof(int), max_entries, NULL);
 	CHECK(map_fd == -1,
-	      "bpf_create_map_xattr()", "error:%s\n", strerror(errno));
+	      "bpf_map_create()", "error:%s\n", strerror(errno));
 
 	value_size = is_pcpu ? sizeof(value) : sizeof(int);
 	keys = malloc(max_entries * sizeof(int));

tools/testing/selftests/bpf/map_tests/lpm_trie_map_batch_ops.c

···
 
 void test_lpm_trie_map_batch_ops(void)
 {
-	struct bpf_create_map_attr xattr = {
-		.name = "lpm_trie_map",
-		.map_type = BPF_MAP_TYPE_LPM_TRIE,
-		.key_size = sizeof(struct test_lpm_key),
-		.value_size = sizeof(int),
-		.map_flags = BPF_F_NO_PREALLOC,
-	};
+	LIBBPF_OPTS(bpf_map_create_opts, create_opts, .map_flags = BPF_F_NO_PREALLOC);
 	struct test_lpm_key *keys, key;
 	int map_fd, *values, *visited;
 	__u32 step, count, total, total_success;
···
 		.flags = 0,
 	);
 
-	xattr.max_entries = max_entries;
-	map_fd = bpf_create_map_xattr(&xattr);
-	CHECK(map_fd == -1, "bpf_create_map_xattr()", "error:%s\n",
+	map_fd = bpf_map_create(BPF_MAP_TYPE_LPM_TRIE, "lpm_trie_map",
+				sizeof(struct test_lpm_key), sizeof(int),
+				max_entries, &create_opts);
+	CHECK(map_fd == -1, "bpf_map_create()", "error:%s\n",
 	      strerror(errno));
 
 	keys = malloc(max_entries * sizeof(struct test_lpm_key));

tools/testing/selftests/bpf/map_tests/sk_storage_map.c

···
 #include <test_btf.h>
 #include <test_maps.h>
 
-static struct bpf_create_map_attr xattr = {
-	.name = "sk_storage_map",
-	.map_type = BPF_MAP_TYPE_SK_STORAGE,
-	.map_flags = BPF_F_NO_PREALLOC,
-	.max_entries = 0,
-	.key_size = 4,
-	.value_size = 8,
+static struct bpf_map_create_opts map_opts = {
+	.sz = sizeof(map_opts),
 	.btf_key_type_id = 1,
 	.btf_value_type_id = 3,
 	.btf_fd = -1,
+	.map_flags = BPF_F_NO_PREALLOC,
 };
 
 static unsigned int nr_sk_threads_done;
···
 	memcpy(raw_btf + sizeof(btf_hdr) + sizeof(btf_raw_types),
 	       btf_str_sec, sizeof(btf_str_sec));
 
-	return bpf_load_btf(raw_btf, sizeof(raw_btf), 0, 0, 0);
+	return bpf_btf_load(raw_btf, sizeof(raw_btf), NULL);
 }
 
 static int create_sk_storage_map(void)
···
 	btf_fd = load_btf();
 	CHECK(btf_fd == -1, "bpf_load_btf", "btf_fd:%d errno:%d\n",
 	      btf_fd, errno);
-	xattr.btf_fd = btf_fd;
+	map_opts.btf_fd = btf_fd;
 
-	map_fd = bpf_create_map_xattr(&xattr);
-	xattr.btf_fd = -1;
+	map_fd = bpf_map_create(BPF_MAP_TYPE_SK_STORAGE, "sk_storage_map", 4, 8, 0, &map_opts);
+	map_opts.btf_fd = -1;
 	close(btf_fd);
 	CHECK(map_fd == -1,
-	      "bpf_create_map_xattr()", "errno:%d\n", errno);
+	      "bpf_map_create()", "errno:%d\n", errno);
 
 	return map_fd;
 }
···
 		int cnt;
 		int lock;
 	} value = { .cnt = 0xeB9f, .lock = 0, }, lookup_value;
-	struct bpf_create_map_attr bad_xattr;
+	struct bpf_map_create_opts bad_xattr;
 	int btf_fd, map_fd, sk_fd, err;
 
 	btf_fd = load_btf();
 	CHECK(btf_fd == -1, "bpf_load_btf", "btf_fd:%d errno:%d\n",
 	      btf_fd, errno);
-	xattr.btf_fd = btf_fd;
+	map_opts.btf_fd = btf_fd;
 
 	sk_fd = socket(AF_INET6, SOCK_STREAM, 0);
 	CHECK(sk_fd == -1, "socket()", "sk_fd:%d errno:%d\n",
 	      sk_fd, errno);
 
-	map_fd = bpf_create_map_xattr(&xattr);
-	CHECK(map_fd == -1, "bpf_create_map_xattr(good_xattr)",
+	map_fd = bpf_map_create(BPF_MAP_TYPE_SK_STORAGE, "sk_storage_map", 4, 8, 0, &map_opts);
+	CHECK(map_fd == -1, "bpf_map_create(good_xattr)",
 	      "map_fd:%d errno:%d\n", map_fd, errno);
 
 	/* Add new elem */
···
 	CHECK(!err || errno != ENOENT, "bpf_map_delete_elem()",
 	      "err:%d errno:%d\n", err, errno);
 
-	memcpy(&bad_xattr, &xattr, sizeof(xattr));
+	memcpy(&bad_xattr, &map_opts, sizeof(map_opts));
 	bad_xattr.btf_key_type_id = 0;
-	err = bpf_create_map_xattr(&bad_xattr);
-	CHECK(!err || errno != EINVAL, "bap_create_map_xattr(bad_xattr)",
+	err = bpf_map_create(BPF_MAP_TYPE_SK_STORAGE, "sk_storage_map", 4, 8, 0, &bad_xattr);
+	CHECK(!err || errno != EINVAL, "bpf_map_create(bad_xattr)",
 	      "err:%d errno:%d\n", err, errno);
 
-	memcpy(&bad_xattr, &xattr, sizeof(xattr));
+	memcpy(&bad_xattr, &map_opts, sizeof(map_opts));
 	bad_xattr.btf_key_type_id = 3;
-	err = bpf_create_map_xattr(&bad_xattr);
-	CHECK(!err || errno != EINVAL, "bap_create_map_xattr(bad_xattr)",
+	err = bpf_map_create(BPF_MAP_TYPE_SK_STORAGE, "sk_storage_map", 4, 8, 0, &bad_xattr);
+	CHECK(!err || errno != EINVAL, "bpf_map_create(bad_xattr)",
 	      "err:%d errno:%d\n", err, errno);
 
-	memcpy(&bad_xattr, &xattr, sizeof(xattr));
-	bad_xattr.max_entries = 1;
-	err = bpf_create_map_xattr(&bad_xattr);
-	CHECK(!err || errno != EINVAL, "bap_create_map_xattr(bad_xattr)",
+	err = bpf_map_create(BPF_MAP_TYPE_SK_STORAGE, "sk_storage_map", 4, 8, 1, &map_opts);
+	CHECK(!err || errno != EINVAL, "bpf_map_create(bad_xattr)",
 	      "err:%d errno:%d\n", err, errno);
 
-	memcpy(&bad_xattr, &xattr, sizeof(xattr));
+	memcpy(&bad_xattr, &map_opts, sizeof(map_opts));
 	bad_xattr.map_flags = 0;
-	err = bpf_create_map_xattr(&bad_xattr);
+	err = bpf_map_create(BPF_MAP_TYPE_SK_STORAGE, "sk_storage_map", 4, 8, 0, &bad_xattr);
 	CHECK(!err || errno != EINVAL, "bap_create_map_xattr(bad_xattr)",
 	      "err:%d errno:%d\n", err, errno);
 
-	xattr.btf_fd = -1;
+	map_opts.btf_fd = -1;
 	close(btf_fd);
 	close(map_fd);
 	close(sk_fd);

tools/testing/selftests/bpf/prog_tests/atomics.c

···
 	prog_fd = skel->progs.cmpxchg.prog_fd;
 	err = bpf_prog_test_run(prog_fd, 1, NULL, 0,
 				NULL, NULL, &retval, &duration);
-	if (CHECK(err || retval, "test_run add",
+	if (CHECK(err || retval, "test_run cmpxchg",
 		  "err %d errno %d retval %d duration %d\n", err, errno, retval, duration))
 		goto cleanup;
 
···
 	prog_fd = skel->progs.xchg.prog_fd;
 	err = bpf_prog_test_run(prog_fd, 1, NULL, 0,
 				NULL, NULL, &retval, &duration);
-	if (CHECK(err || retval, "test_run add",
+	if (CHECK(err || retval, "test_run xchg",
 		  "err %d errno %d retval %d duration %d\n", err, errno, retval, duration))
 		goto cleanup;
 

tools/testing/selftests/bpf/prog_tests/bloom_filter_map.c

···
 
 static void test_fail_cases(void)
 {
+	LIBBPF_OPTS(bpf_map_create_opts, opts);
 	__u32 value;
 	int fd, err;
 
 	/* Invalid key size */
-	fd = bpf_create_map(BPF_MAP_TYPE_BLOOM_FILTER, 4, sizeof(value), 100, 0);
-	if (!ASSERT_LT(fd, 0, "bpf_create_map bloom filter invalid key size"))
+	fd = bpf_map_create(BPF_MAP_TYPE_BLOOM_FILTER, NULL, 4, sizeof(value), 100, NULL);
+	if (!ASSERT_LT(fd, 0, "bpf_map_create bloom filter invalid key size"))
 		close(fd);
 
 	/* Invalid value size */
-	fd = bpf_create_map(BPF_MAP_TYPE_BLOOM_FILTER, 0, 0, 100, 0);
-	if (!ASSERT_LT(fd, 0, "bpf_create_map bloom filter invalid value size 0"))
+	fd = bpf_map_create(BPF_MAP_TYPE_BLOOM_FILTER, NULL, 0, 0, 100, NULL);
+	if (!ASSERT_LT(fd, 0, "bpf_map_create bloom filter invalid value size 0"))
 		close(fd);
 
 	/* Invalid max entries size */
-	fd = bpf_create_map(BPF_MAP_TYPE_BLOOM_FILTER, 0, sizeof(value), 0, 0);
-	if (!ASSERT_LT(fd, 0, "bpf_create_map bloom filter invalid max entries size"))
+	fd = bpf_map_create(BPF_MAP_TYPE_BLOOM_FILTER, NULL, 0, sizeof(value), 0, NULL);
+	if (!ASSERT_LT(fd, 0, "bpf_map_create bloom filter invalid max entries size"))
 		close(fd);
 
 	/* Bloom filter maps do not support BPF_F_NO_PREALLOC */
-	fd = bpf_create_map(BPF_MAP_TYPE_BLOOM_FILTER, 0, sizeof(value), 100,
-			    BPF_F_NO_PREALLOC);
-	if (!ASSERT_LT(fd, 0, "bpf_create_map bloom filter invalid flags"))
+	opts.map_flags = BPF_F_NO_PREALLOC;
+	fd = bpf_map_create(BPF_MAP_TYPE_BLOOM_FILTER, NULL, 0, sizeof(value), 100, &opts);
+	if (!ASSERT_LT(fd, 0, "bpf_map_create bloom filter invalid flags"))
 		close(fd);
 
-	fd = bpf_create_map(BPF_MAP_TYPE_BLOOM_FILTER, 0, sizeof(value), 100, 0);
-	if (!ASSERT_GE(fd, 0, "bpf_create_map bloom filter"))
+	fd = bpf_map_create(BPF_MAP_TYPE_BLOOM_FILTER, NULL, 0, sizeof(value), 100, NULL);
+	if (!ASSERT_GE(fd, 0, "bpf_map_create bloom filter"))
 		return;
 
 	/* Test invalid flags */
···
 
 static void test_success_cases(void)
 {
+	LIBBPF_OPTS(bpf_map_create_opts, opts);
 	char value[11];
 	int fd, err;
 
 	/* Create a map */
-	fd = bpf_create_map(BPF_MAP_TYPE_BLOOM_FILTER, 0, sizeof(value), 100,
-			    BPF_F_ZERO_SEED | BPF_F_NUMA_NODE);
-	if (!ASSERT_GE(fd, 0, "bpf_create_map bloom filter success case"))
+	opts.map_flags = BPF_F_ZERO_SEED | BPF_F_NUMA_NODE;
+	fd = bpf_map_create(BPF_MAP_TYPE_BLOOM_FILTER, NULL, 0, sizeof(value), 100, &opts);
+	if (!ASSERT_GE(fd, 0, "bpf_map_create bloom filter success case"))
 		return;
 
 	/* Add a value to the bloom filter */
···
 	struct bpf_link *link;
 
 	/* Create a bloom filter map that will be used as the inner map */
-	inner_map_fd = bpf_create_map(BPF_MAP_TYPE_BLOOM_FILTER, 0, sizeof(*rand_vals),
-				      nr_rand_vals, 0);
-	if (!ASSERT_GE(inner_map_fd, 0, "bpf_create_map bloom filter inner map"))
+	inner_map_fd = bpf_map_create(BPF_MAP_TYPE_BLOOM_FILTER, NULL, 0, sizeof(*rand_vals),
+				      nr_rand_vals, NULL);
+	if (!ASSERT_GE(inner_map_fd, 0, "bpf_map_create bloom filter inner map"))
 		return;
 
 	for (i = 0; i < nr_rand_vals; i++) {

tools/testing/selftests/bpf/prog_tests/bpf_iter.c

···
 	 * fills seq_file buffer and then the other will trigger
 	 * overflow and needs restart.
 	 */
-	map1_fd = bpf_create_map(BPF_MAP_TYPE_ARRAY, 4, 8, 1, 0);
-	if (CHECK(map1_fd < 0, "bpf_create_map",
+	map1_fd = bpf_map_create(BPF_MAP_TYPE_ARRAY, NULL, 4, 8, 1, NULL);
+	if (CHECK(map1_fd < 0, "bpf_map_create",
 		  "map_creation failed: %s\n", strerror(errno)))
 		goto out;
-	map2_fd = bpf_create_map(BPF_MAP_TYPE_ARRAY, 4, 8, 1, 0);
-	if (CHECK(map2_fd < 0, "bpf_create_map",
+	map2_fd = bpf_map_create(BPF_MAP_TYPE_ARRAY, NULL, 4, 8, 1, NULL);
+	if (CHECK(map2_fd < 0, "bpf_map_create",
 		  "map_creation failed: %s\n", strerror(errno)))
 		goto free_map1;
 
···
 		goto out;
 
 	/* Read CMP_BUFFER_SIZE (1kB) from bpf_iter. Read in small chunks
-	 * to trigger seq_file corner cases. The expected output is much
-	 * longer than 1kB, so the while loop will terminate.
+	 * to trigger seq_file corner cases.
 	 */
 	len = 0;
 	while (len < CMP_BUFFER_SIZE) {
 		err = read_fd_into_buffer(iter_fd, task_vma_output + len,
 					  min(read_size, CMP_BUFFER_SIZE - len));
+		if (!err)
+			break;
 		if (CHECK(err < 0, "read_iter_fd", "read_iter_fd failed\n"))
 			goto out;
 		len += err;

tools/testing/selftests/bpf/prog_tests/bpf_loop.c

···
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2021 Facebook */
+
+#include <test_progs.h>
+#include <network_helpers.h>
+#include "bpf_loop.skel.h"
+
+static void check_nr_loops(struct bpf_loop *skel)
+{
+	struct bpf_link *link;
+
+	link = bpf_program__attach(skel->progs.test_prog);
+	if (!ASSERT_OK_PTR(link, "link"))
+		return;
+
+	/* test 0 loops */
+	skel->bss->nr_loops = 0;
+
+	usleep(1);
+
+	ASSERT_EQ(skel->bss->nr_loops_returned, skel->bss->nr_loops,
+		  "0 loops");
+
+	/* test 500 loops */
+	skel->bss->nr_loops = 500;
+
+	usleep(1);
+
+	ASSERT_EQ(skel->bss->nr_loops_returned, skel->bss->nr_loops,
+		  "500 loops");
+	ASSERT_EQ(skel->bss->g_output, (500 * 499) / 2, "g_output");
+
+	/* test exceeding the max limit */
+	skel->bss->nr_loops = -1;
+
+	usleep(1);
+
+	ASSERT_EQ(skel->bss->err, -E2BIG, "over max limit");
+
+	bpf_link__destroy(link);
+}
+
+static void check_callback_fn_stop(struct bpf_loop *skel)
+{
+	struct bpf_link *link;
+
+	link = bpf_program__attach(skel->progs.test_prog);
+	if (!ASSERT_OK_PTR(link, "link"))
+		return;
+
+	/* testing that loop is stopped when callback_fn returns 1 */
+	skel->bss->nr_loops = 400;
+	skel->data->stop_index = 50;
+
+	usleep(1);
+
+	ASSERT_EQ(skel->bss->nr_loops_returned, skel->data->stop_index + 1,
+		  "nr_loops_returned");
+	ASSERT_EQ(skel->bss->g_output, (50 * 49) / 2,
+		  "g_output");
+
+	bpf_link__destroy(link);
+}
+
+static void check_null_callback_ctx(struct bpf_loop *skel)
+{
+	struct bpf_link *link;
+
+	/* check that user is able to pass in a null callback_ctx */
+	link = bpf_program__attach(skel->progs.prog_null_ctx);
+	if (!ASSERT_OK_PTR(link, "link"))
+		return;
+
+	skel->bss->nr_loops = 10;
+
+	usleep(1);
+
+	ASSERT_EQ(skel->bss->nr_loops_returned, skel->bss->nr_loops,
+		  "nr_loops_returned");
+
+	bpf_link__destroy(link);
+}
+
+static void check_invalid_flags(struct bpf_loop *skel)
+{
+	struct bpf_link *link;
+
+	/* check that passing in non-zero flags returns -EINVAL */
+	link = bpf_program__attach(skel->progs.prog_invalid_flags);
+	if (!ASSERT_OK_PTR(link, "link"))
+		return;
+
+	usleep(1);
+
+	ASSERT_EQ(skel->bss->err, -EINVAL, "err");
+
+	bpf_link__destroy(link);
+}
+
+static void check_nested_calls(struct bpf_loop *skel)
+{
+	__u32 nr_loops = 100, nested_callback_nr_loops = 4;
+	struct bpf_link *link;
+
+	/* check that nested calls are supported */
+	link = bpf_program__attach(skel->progs.prog_nested_calls);
+	if (!ASSERT_OK_PTR(link, "link"))
+		return;
+
+	skel->bss->nr_loops = nr_loops;
+	skel->bss->nested_callback_nr_loops = nested_callback_nr_loops;
+
+	usleep(1);
+
+	ASSERT_EQ(skel->bss->nr_loops_returned, nr_loops * nested_callback_nr_loops
+		  * nested_callback_nr_loops, "nr_loops_returned");
+	ASSERT_EQ(skel->bss->g_output, (4 * 3) / 2 * nested_callback_nr_loops
+		* nr_loops, "g_output");
+
+	bpf_link__destroy(link);
+}
+
+void test_bpf_loop(void)
+{
+	struct bpf_loop *skel;
+
+	skel = bpf_loop__open_and_load();
+	if (!ASSERT_OK_PTR(skel, "bpf_loop__open_and_load"))
+		return;
+
+	skel->bss->pid = getpid();
+
+	if (test__start_subtest("check_nr_loops"))
+		check_nr_loops(skel);
+	if (test__start_subtest("check_callback_fn_stop"))
+		check_callback_fn_stop(skel);
+	if (test__start_subtest("check_null_callback_ctx"))
+		check_null_callback_ctx(skel);
+	if (test__start_subtest("check_invalid_flags"))
+		check_invalid_flags(skel);
+	if (test__start_subtest("check_nested_calls"))
+		check_nested_calls(skel);
+
+	bpf_loop__destroy(skel);
+}

tools/testing/selftests/bpf/prog_tests/bpf_tcp_ca.c

···
 static int libbpf_debug_print(enum libbpf_print_level level,
 			      const char *format, va_list args)
 {
-	char *log_buf;
+	const char *log_buf;
 
 	if (level != LIBBPF_WARN ||
-	    strcmp(format, "libbpf: \n%s\n")) {
+	    !strstr(format, "-- BEGIN PROG LOAD LOG --")) {
 		vprintf(format, args);
 		return 0;
 	}
 
+	/* skip prog_name */
+	va_arg(args, char *);
 	log_buf = va_arg(args, char *);
 	if (!log_buf)
 		goto out;

tools/testing/selftests/bpf/prog_tests/bpf_verif_scale.c

···
 
 static int check_load(const char *file, enum bpf_prog_type type)
 {
-	struct bpf_prog_load_attr attr;
 	struct bpf_object *obj = NULL;
-	int err, prog_fd;
+	struct bpf_program *prog;
+	int err;
 
-	memset(&attr, 0, sizeof(struct bpf_prog_load_attr));
-	attr.file = file;
-	attr.prog_type = type;
-	attr.log_level = 4 | extra_prog_load_log_flags;
-	attr.prog_flags = BPF_F_TEST_RND_HI32;
-	err = bpf_prog_load_xattr(&attr, &obj, &prog_fd);
+	obj = bpf_object__open_file(file, NULL);
+	err = libbpf_get_error(obj);
+	if (err)
+		return err;
+
+	prog = bpf_object__next_program(obj, NULL);
+	if (!prog) {
+		err = -ENOENT;
+		goto err_out;
+	}
+
+	bpf_program__set_type(prog, type);
+	bpf_program__set_flags(prog, BPF_F_TEST_RND_HI32);
+	bpf_program__set_log_level(prog, 4 | extra_prog_load_log_flags);
+
+	err = bpf_object__load(obj);
+
+err_out:
 	bpf_object__close(obj);
 	return err;
 }
···
 	scale_test("pyperf600.o", BPF_PROG_TYPE_RAW_TRACEPOINT, false);
 }
 
+void test_verif_scale_pyperf600_bpf_loop(void)
+{
+	/* use the bpf_loop helper*/
+	scale_test("pyperf600_bpf_loop.o", BPF_PROG_TYPE_RAW_TRACEPOINT, false);
+}
+
 void test_verif_scale_pyperf600_nounroll()
 {
 	/* no unroll at all.
···
 	 * ~350k processed_insns
 	 */
 	scale_test("strobemeta.o", BPF_PROG_TYPE_RAW_TRACEPOINT, false);
+}
+
+void test_verif_scale_strobemeta_bpf_loop(void)
+{
+	/* use the bpf_loop helper*/
+	scale_test("strobemeta_bpf_loop.o", BPF_PROG_TYPE_RAW_TRACEPOINT, false);
 }
 
 void test_verif_scale_strobemeta_nounroll1()

tools/testing/selftests/bpf/prog_tests/btf.c

···
 	return raw_btf;
 }
 
+static int load_raw_btf(const void *raw_data, size_t raw_size)
+{
+	LIBBPF_OPTS(bpf_btf_load_opts, opts);
+	int btf_fd;
+
+	if (always_log) {
+		opts.log_buf = btf_log_buf,
+		opts.log_size = BTF_LOG_BUF_SIZE,
+		opts.log_level = 1;
+	}
+
+	btf_fd = bpf_btf_load(raw_data, raw_size, &opts);
+	if (btf_fd < 0 && !always_log) {
+		opts.log_buf = btf_log_buf,
+		opts.log_size = BTF_LOG_BUF_SIZE,
+		opts.log_level = 1;
+		btf_fd = bpf_btf_load(raw_data, raw_size, &opts);
+	}
+
+	return btf_fd;
+}
+
 static void do_test_raw(unsigned int test_num)
 {
 	struct btf_raw_test *test = &raw_tests[test_num - 1];
-	struct bpf_create_map_attr create_attr = {};
+	LIBBPF_OPTS(bpf_map_create_opts, opts);
 	int map_fd = -1, btf_fd = -1;
 	unsigned int raw_btf_size;
 	struct btf_header *hdr;
···
 	hdr->str_len = (int)hdr->str_len + test->str_len_delta;
 
 	*btf_log_buf = '\0';
-	btf_fd = bpf_load_btf(raw_btf, raw_btf_size,
-			      btf_log_buf, BTF_LOG_BUF_SIZE,
-			      always_log);
+	btf_fd = load_raw_btf(raw_btf, raw_btf_size);
 	free(raw_btf);
 
 	err = ((btf_fd < 0) != test->btf_load_err);
 	if (CHECK(err, "btf_fd:%d test->btf_load_err:%u",
 		  btf_fd, test->btf_load_err) ||
 	    CHECK(test->err_str && !strstr(btf_log_buf, test->err_str),
-		  "expected err_str:%s", test->err_str)) {
+		  "expected err_str:%s\n", test->err_str)) {
 		err = -1;
 		goto done;
 	}
···
 	if (err || btf_fd < 0)
 		goto done;
 
-	create_attr.name = test->map_name;
-	create_attr.map_type = test->map_type;
-	create_attr.key_size = test->key_size;
-	create_attr.value_size = test->value_size;
-	create_attr.max_entries = test->max_entries;
-	create_attr.btf_fd = btf_fd;
-	create_attr.btf_key_type_id = test->key_type_id;
-	create_attr.btf_value_type_id = test->value_type_id;
-
-	map_fd = bpf_create_map_xattr(&create_attr);
+	opts.btf_fd = btf_fd;
+	opts.btf_key_type_id = test->key_type_id;
+	opts.btf_value_type_id = test->value_type_id;
+	map_fd = bpf_map_create(test->map_type, test->map_name,
+				test->key_size, test->value_size, test->max_entries, &opts);
 
 	err = ((map_fd < 0) != test->map_create_err);
 	CHECK(err, "map_fd:%d test->map_create_err:%u",
···
 		goto done;
 	}
 
-	btf_fd = bpf_load_btf(raw_btf, raw_btf_size,
-			      btf_log_buf, BTF_LOG_BUF_SIZE,
-			      always_log);
+	btf_fd = load_raw_btf(raw_btf, raw_btf_size);
 	if (CHECK(btf_fd < 0, "errno:%d", errno)) {
 		err = -1;
 		goto done;
···
 static int test_btf_id(unsigned int test_num)
 {
 	const struct btf_get_info_test *test = &get_info_tests[test_num - 1];
-	struct bpf_create_map_attr create_attr = {};
+	LIBBPF_OPTS(bpf_map_create_opts, opts);
 	uint8_t *raw_btf = NULL, *user_btf[2] = {};
 	int btf_fd[2] = {-1, -1}, map_fd = -1;
 	struct bpf_map_info map_info = {};
···
 		info[i].btf_size = raw_btf_size;
 	}
 
-	btf_fd[0] = bpf_load_btf(raw_btf, raw_btf_size,
-				 btf_log_buf, BTF_LOG_BUF_SIZE,
-				 always_log);
+	btf_fd[0] = load_raw_btf(raw_btf, raw_btf_size);
 	if (CHECK(btf_fd[0] < 0, "errno:%d", errno)) {
 		err = -1;
 		goto done;
···
 	}
 
 	/* Test btf members in struct bpf_map_info */
-	create_attr.name = "test_btf_id";
-	create_attr.map_type = BPF_MAP_TYPE_ARRAY;
-	create_attr.key_size = sizeof(int);
-	create_attr.value_size = sizeof(unsigned int);
-	create_attr.max_entries = 4;
-	create_attr.btf_fd = btf_fd[0];
-	create_attr.btf_key_type_id = 1;
-	create_attr.btf_value_type_id = 2;
-
-	map_fd = bpf_create_map_xattr(&create_attr);
+	opts.btf_fd = btf_fd[0];
+	opts.btf_key_type_id = 1;
+	opts.btf_value_type_id = 2;
+	map_fd = bpf_map_create(BPF_MAP_TYPE_ARRAY, "test_btf_id",
+				sizeof(int), sizeof(int), 4, &opts);
 	if (CHECK(map_fd < 0, "errno:%d", errno)) {
 		err = -1;
 		goto done;
···
 		goto done;
 	}
 
-	btf_fd = bpf_load_btf(raw_btf, raw_btf_size,
-			      btf_log_buf, BTF_LOG_BUF_SIZE,
-			      always_log);
+	btf_fd = load_raw_btf(raw_btf, raw_btf_size);
 	if (CHECK(btf_fd <= 0, "errno:%d", errno)) {
 		err = -1;
 		goto done;
···
 {
 	const struct btf_raw_test *test = &pprint_test_template[test_num];
 	enum pprint_mapv_kind_t mapv_kind = test->mapv_kind;
-	struct bpf_create_map_attr create_attr = {};
+	LIBBPF_OPTS(bpf_map_create_opts, opts);
 	bool ordered_map, lossless_map, percpu_map;
 	int err, ret, num_cpus, rounded_value_size;
 	unsigned int key, nr_read_elems;
···
 		return;
 
 	*btf_log_buf = '\0';
-	btf_fd = bpf_load_btf(raw_btf, raw_btf_size,
-			      btf_log_buf, BTF_LOG_BUF_SIZE,
-			      always_log);
+	btf_fd = load_raw_btf(raw_btf, raw_btf_size);
 	free(raw_btf);
 
-	if (CHECK(btf_fd < 0, "errno:%d", errno)) {
+	if (CHECK(btf_fd < 0, "errno:%d\n", errno)) {
 		err = -1;
 		goto done;
 	}
 
-	create_attr.name = test->map_name;
-	create_attr.map_type = test->map_type;
-	create_attr.key_size = test->key_size;
-	create_attr.value_size = test->value_size;
-	create_attr.max_entries = test->max_entries;
-	create_attr.btf_fd = btf_fd;
-	create_attr.btf_key_type_id = test->key_type_id;
-	create_attr.btf_value_type_id = test->value_type_id;
-
-	map_fd = bpf_create_map_xattr(&create_attr);
+	opts.btf_fd = btf_fd;
+	opts.btf_key_type_id = test->key_type_id;
+	opts.btf_value_type_id = test->value_type_id;
+	map_fd = bpf_map_create(test->map_type, test->map_name,
+				test->key_size, test->value_size, test->max_entries, &opts);
 	if (CHECK(map_fd < 0, "errno:%d", errno)) {
 		err = -1;
 		goto done;
···
 		return;
 
 	*btf_log_buf = '\0';
-	btf_fd = bpf_load_btf(raw_btf, raw_btf_size,
-			      btf_log_buf, BTF_LOG_BUF_SIZE,
-			      always_log);
+	btf_fd = load_raw_btf(raw_btf, raw_btf_size);
 	free(raw_btf);
 
 	if (CHECK(btf_fd < 0, "invalid btf_fd errno:%d", errno)) {
···
 			BTF_END_RAW,
 		},
 		BTF_STR_SEC("\0tag1"),
+	},
+},
+{
+	.descr = "dedup: btf_type_tag #5, struct",
+	.input = {
+		.raw_types = {
+			BTF_TYPE_INT_ENC(0, BTF_INT_SIGNED, 0, 32, 4),				/* [1] */
+			BTF_TYPE_TAG_ENC(NAME_NTH(1), 1),					/* [2] */
+			BTF_TYPE_ENC(NAME_NTH(2), BTF_INFO_ENC(BTF_KIND_STRUCT, 1, 1), 4),	/* [3] */
+			BTF_MEMBER_ENC(NAME_NTH(3), 2, BTF_MEMBER_OFFSET(0, 0)),
+			BTF_TYPE_TAG_ENC(NAME_NTH(1), 1),					/* [4] */
+			BTF_TYPE_ENC(NAME_NTH(2), BTF_INFO_ENC(BTF_KIND_STRUCT, 1, 1), 4),	/* [5] */
+			BTF_MEMBER_ENC(NAME_NTH(3), 4, BTF_MEMBER_OFFSET(0, 0)),
+			BTF_END_RAW,
+		},
+		BTF_STR_SEC("\0tag1\0t\0m"),
+	},
+	.expect = {
+		.raw_types = {
+			BTF_TYPE_INT_ENC(0, BTF_INT_SIGNED, 0, 32, 4),				/* [1] */
+			BTF_TYPE_TAG_ENC(NAME_NTH(1), 1),					/* [2] */
+			BTF_TYPE_ENC(NAME_NTH(2), BTF_INFO_ENC(BTF_KIND_STRUCT, 1, 1), 4),	/* [3] */
+			BTF_MEMBER_ENC(NAME_NTH(3), 2, BTF_MEMBER_OFFSET(0, 0)),
+			BTF_END_RAW,
+		},
+		BTF_STR_SEC("\0tag1\0t\0m"),
 	},
 },
 

tools/testing/selftests/bpf/prog_tests/btf_dedup_split.c

···
 	btf__free(btf1);
 }
 
+static void btf_add_dup_struct_in_cu(struct btf *btf, int start_id)
+{
+#define ID(n) (start_id + n)
+	btf__set_pointer_size(btf, 8); /* enforce 64-bit arch */
+
+	btf__add_int(btf, "int", 4, BTF_INT_SIGNED);    /* [1] int */
+
+	btf__add_struct(btf, "s", 8);                   /* [2] struct s { */
+	btf__add_field(btf, "a", ID(3), 0, 0);          /*      struct anon a; */
+	btf__add_field(btf, "b", ID(4), 0, 0);          /*      struct anon b; */
+							/* } */
+
+	btf__add_struct(btf, "(anon)", 8);              /* [3] struct anon { */
+	btf__add_field(btf, "f1", ID(1), 0, 0);         /*      int f1; */
+	btf__add_field(btf, "f2", ID(1), 32, 0);        /*      int f2; */
+							/* } */
+
+	btf__add_struct(btf, "(anon)", 8);              /* [4] struct anon { */
+	btf__add_field(btf, "f1", ID(1), 0, 0);         /*      int f1; */
+	btf__add_field(btf, "f2", ID(1), 32, 0);        /*      int f2; */
+							/* } */
+#undef ID
+}
+
+static void test_split_dup_struct_in_cu()
+{
+	struct btf *btf1, *btf2 = NULL;
+	int err;
+
+	/* generate the base data.. */
+	btf1 = btf__new_empty();
+	if (!ASSERT_OK_PTR(btf1, "empty_main_btf"))
+		return;
+
+	btf_add_dup_struct_in_cu(btf1, 0);
+
+	VALIDATE_RAW_BTF(
+			btf1,
+			"[1] INT 'int' size=4 bits_offset=0 nr_bits=32 encoding=SIGNED",
+			"[2] STRUCT 's' size=8 vlen=2\n"
+			"\t'a' type_id=3 bits_offset=0\n"
+			"\t'b' type_id=4 bits_offset=0",
+			"[3] STRUCT '(anon)' size=8 vlen=2\n"
+			"\t'f1' type_id=1 bits_offset=0\n"
+			"\t'f2' type_id=1 bits_offset=32",
+			"[4] STRUCT '(anon)' size=8 vlen=2\n"
+			"\t'f1' type_id=1 bits_offset=0\n"
+			"\t'f2' type_id=1 bits_offset=32");
+
+	/* ..dedup them... */
+	err = btf__dedup(btf1, NULL);
+	if (!ASSERT_OK(err, "btf_dedup"))
+		goto cleanup;
+
+	VALIDATE_RAW_BTF(
+			btf1,
+			"[1] INT 'int' size=4 bits_offset=0 nr_bits=32 encoding=SIGNED",
+			"[2] STRUCT 's' size=8 vlen=2\n"
+			"\t'a' type_id=3 bits_offset=0\n"
+			"\t'b' type_id=3 bits_offset=0",
+			"[3] STRUCT '(anon)' size=8 vlen=2\n"
+			"\t'f1' type_id=1 bits_offset=0\n"
+			"\t'f2' type_id=1 bits_offset=32");
+
+	/* and add the same data on top of it */
+	btf2 = btf__new_empty_split(btf1);
+	if (!ASSERT_OK_PTR(btf2, "empty_split_btf"))
+		goto cleanup;
+
+	btf_add_dup_struct_in_cu(btf2, 3);
+
+	VALIDATE_RAW_BTF(
+			btf2,
+			"[1] INT 'int' size=4 bits_offset=0 nr_bits=32 encoding=SIGNED",
+			"[2] STRUCT 's' size=8 vlen=2\n"
+			"\t'a' type_id=3 bits_offset=0\n"
+			"\t'b' type_id=3 bits_offset=0",
+			"[3] STRUCT '(anon)' size=8 vlen=2\n"
+			"\t'f1' type_id=1 bits_offset=0\n"
+			"\t'f2' type_id=1 bits_offset=32",
+			"[4] INT 'int' size=4 bits_offset=0 nr_bits=32 encoding=SIGNED",
+			"[5] STRUCT 's' size=8 vlen=2\n"
+			"\t'a' type_id=6 bits_offset=0\n"
+			"\t'b' type_id=7 bits_offset=0",
+			"[6] STRUCT '(anon)' size=8 vlen=2\n"
+			"\t'f1' type_id=4 bits_offset=0\n"
+			"\t'f2' type_id=4 bits_offset=32",
+			"[7] STRUCT '(anon)' size=8 vlen=2\n"
+			"\t'f1' type_id=4 bits_offset=0\n"
+			"\t'f2' type_id=4 bits_offset=32");
+
+	err = btf__dedup(btf2, NULL);
+	if (!ASSERT_OK(err, "btf_dedup"))
+		goto cleanup;
+
+	/* after dedup it should match the original data */
+	VALIDATE_RAW_BTF(
+			btf2,
+			"[1] INT 'int' size=4 bits_offset=0 nr_bits=32 encoding=SIGNED",
+			"[2] STRUCT 's' size=8 vlen=2\n"
+			"\t'a' type_id=3 bits_offset=0\n"
+			"\t'b' type_id=3 bits_offset=0",
+			"[3] STRUCT '(anon)' size=8 vlen=2\n"
+			"\t'f1' type_id=1 bits_offset=0\n"
+			"\t'f2' type_id=1 bits_offset=32");
+
+cleanup:
+	btf__free(btf2);
+	btf__free(btf1);
+}
+
 void test_btf_dedup_split()
 {
 	if (test__start_subtest("split_simple"))
···
 		test_split_struct_duped();
 	if (test__start_subtest("split_fwd_resolve"))
 		test_split_fwd_resolve();
+	if (test__start_subtest("split_dup_struct_in_cu"))
+		test_split_dup_struct_in_cu();
 }

tools/testing/selftests/bpf/prog_tests/btf_dump.c

···
 				   char *str)
 {
 #ifdef __SIZEOF_INT128__
-	__int128 i = 0xffffffffffffffff;
+	unsigned __int128 i = 0xffffffffffffffff;
 
 	/* this dance is required because we cannot directly initialize
 	 * a 128-bit value to anything larger than a 64-bit value.
···
 	/* overflow bpf_sock_ops struct with final element nonzero/zero.
 	 * Regardless of the value of the final field, we don't have all the
 	 * data we need to display it, so we should trigger an overflow.
-	 * In other words oveflow checking should trump "is field zero?"
+	 * In other words overflow checking should trump "is field zero?"
 	 * checks because if we've overflowed, it shouldn't matter what the
 	 * field is - we can't trust its value so shouldn't display it.
 	 */

tools/testing/selftests/bpf/prog_tests/cgroup_attach_multi.c

···
 	int cgroup_storage_fd, percpu_cgroup_storage_fd;
 
 	if (map_fd < 0)
-		map_fd = bpf_create_map(BPF_MAP_TYPE_ARRAY, 4, 8, 1, 0);
+		map_fd = bpf_map_create(BPF_MAP_TYPE_ARRAY, NULL, 4, 8, 1, NULL);
 	if (map_fd < 0) {
 		printf("failed to create map '%s'\n", strerror(errno));
 		return -1;
 	}
 
-	cgroup_storage_fd = bpf_create_map(BPF_MAP_TYPE_CGROUP_STORAGE,
-				sizeof(struct bpf_cgroup_storage_key), 8, 0, 0);
+	cgroup_storage_fd = bpf_map_create(BPF_MAP_TYPE_CGROUP_STORAGE, NULL,
+				sizeof(struct bpf_cgroup_storage_key), 8, 0, NULL);
 	if (cgroup_storage_fd < 0) {
 		printf("failed to create map '%s'\n", strerror(errno));
 		return -1;
 	}
 
-	percpu_cgroup_storage_fd = bpf_create_map(
-		BPF_MAP_TYPE_PERCPU_CGROUP_STORAGE,
-		sizeof(struct bpf_cgroup_storage_key), 8, 0, 0);
+	percpu_cgroup_storage_fd = bpf_map_create(
+		BPF_MAP_TYPE_PERCPU_CGROUP_STORAGE, NULL,
+		sizeof(struct bpf_cgroup_storage_key), 8, 0, NULL);
 	if (percpu_cgroup_storage_fd < 0) {
 		printf("failed to create map '%s'\n", strerror(errno));
 		return -1;

tools/testing/selftests/bpf/prog_tests/connect_force_port.c

···
 	bool v4 = family == AF_INET;
 	__u16 expected_local_port = v4 ? 22222 : 22223;
 	__u16 expected_peer_port = 60000;
-	struct bpf_prog_load_attr attr = {
-		.file = v4 ? "./connect_force_port4.o" :
-			     "./connect_force_port6.o",
-	};
 	struct bpf_program *prog;
 	struct bpf_object *obj;
-	int xlate_fd, fd, err;
+	const char *obj_file = v4 ? "connect_force_port4.o" : "connect_force_port6.o";
+	int fd, err;
 	__u32 duration = 0;
 
-	err = bpf_prog_load_xattr(&attr, &obj, &xlate_fd);
-	if (err) {
-		log_err("Failed to load BPF object");
+	obj = bpf_object__open_file(obj_file, NULL);
+	if (!ASSERT_OK_PTR(obj, "bpf_obj_open"))
 		return -1;
+
+	err = bpf_object__load(obj);
+	if (!ASSERT_OK(err, "bpf_obj_load")) {
+		err = -EIO;
+		goto close_bpf_object;
 	}
 
 	prog = bpf_object__find_program_by_title(obj, v4 ?

tools/testing/selftests/bpf/prog_tests/core_kern.c

···
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2021 Facebook */
+
+#include "test_progs.h"
+#include "core_kern.lskel.h"
+
+void test_core_kern_lskel(void)
+{
+	struct core_kern_lskel *skel;
+
+	skel = core_kern_lskel__open_and_load();
+	ASSERT_OK_PTR(skel, "open_and_load");
+	core_kern_lskel__destroy(skel);
+}

tools/testing/selftests/bpf/prog_tests/core_reloc.c

···
 		data = mmap_data;
 
 		memset(mmap_data, 0, sizeof(*data));
-		memcpy(data->in, test_case->input, test_case->input_len);
+		if (test_case->input_len)
+			memcpy(data->in, test_case->input, test_case->input_len);
 		data->my_pid_tgid = my_pid_tgid;
 
 		link = bpf_program__attach_raw_tracepoint(prog, tp_name);

tools/testing/selftests/bpf/prog_tests/get_stack_raw_tp.c

···
 {
 	bool good_kern_stack = false, good_user_stack = false;
 	const char *nonjit_func = "___bpf_prog_run";
-	struct get_stack_trace_t *e = data;
+	/* perfbuf-submitted data is 4-byte aligned, but we need 8-byte
+	 * alignment, so copy data into a local variable, for simplicity
+	 */
+	struct get_stack_trace_t e;
 	int i, num_stack;
 	static __u64 cnt;
 	struct ksym *ks;
 
 	cnt++;
+
+	memset(&e, 0, sizeof(e));
+	memcpy(&e, data, size <= sizeof(e) ? size : sizeof(e));
 
 	if (size < sizeof(struct get_stack_trace_t)) {
 		__u64 *raw_data = data;
···
 			good_user_stack = true;
 		}
 	} else {
-		num_stack = e->kern_stack_size / sizeof(__u64);
+		num_stack = e.kern_stack_size / sizeof(__u64);
 		if (env.jit_enabled) {
 			good_kern_stack = num_stack > 0;
 		} else {
 			for (i = 0; i < num_stack; i++) {
-				ks = ksym_search(e->kern_stack[i]);
+				ks = ksym_search(e.kern_stack[i]);
 				if (ks && (strcmp(ks->name, nonjit_func) == 0)) {
 					good_kern_stack = true;
 					break;
 				}
 			}
 		}
-		if (e->user_stack_size > 0 && e->user_stack_buildid_size > 0)
+		if (e.user_stack_size > 0 && e.user_stack_buildid_size > 0)
 			good_user_stack = true;
 	}
 

tools/testing/selftests/bpf/prog_tests/kfree_skb.c

···
 // SPDX-License-Identifier: GPL-2.0
 #include <test_progs.h>
 #include <network_helpers.h>
+#include "kfree_skb.skel.h"
 
 struct meta {
 	int ifindex;
···
 		.ctx_in = &skb,
 		.ctx_size_in = sizeof(skb),
 	};
-	struct bpf_prog_load_attr attr = {
-		.file = "./kfree_skb.o",
-	};
-
-	struct bpf_link *link = NULL, *link_fentry = NULL, *link_fexit = NULL;
-	struct bpf_map *perf_buf_map, *global_data;
-	struct bpf_program *prog, *fentry, *fexit;
-	struct bpf_object *obj, *obj2 = NULL;
+	struct kfree_skb *skel = NULL;
+	struct bpf_link *link;
+	struct bpf_object *obj;
 	struct perf_buffer *pb = NULL;
-	int err, kfree_skb_fd;
+	int err;
 	bool passed = false;
 	__u32 duration = 0;
 	const int zero = 0;
···
 	if (CHECK(err, "prog_load sched cls", "err %d errno %d\n", err, errno))
 		return;
 
-	err = bpf_prog_load_xattr(&attr, &obj2, &kfree_skb_fd);
-	if (CHECK(err, "prog_load raw tp", "err %d errno %d\n", err, errno))
+	skel = kfree_skb__open_and_load();
+	if (!ASSERT_OK_PTR(skel, "kfree_skb_skel"))
 		goto close_prog;
 
-	prog = bpf_object__find_program_by_title(obj2, "tp_btf/kfree_skb");
-	if (CHECK(!prog, "find_prog", "prog kfree_skb not found\n"))
-		goto close_prog;
-	fentry = bpf_object__find_program_by_title(obj2, "fentry/eth_type_trans");
-	if (CHECK(!fentry, "find_prog", "prog eth_type_trans not found\n"))
-		goto close_prog;
-	fexit = bpf_object__find_program_by_title(obj2, "fexit/eth_type_trans");
-	if (CHECK(!fexit, "find_prog", "prog eth_type_trans not found\n"))
-		goto close_prog;
-
-	global_data = bpf_object__find_map_by_name(obj2, ".bss");
-	if (CHECK(!global_data, "find global data", "not found\n"))
-		goto close_prog;
-
-	link = bpf_program__attach_raw_tracepoint(prog, NULL);
+	link = bpf_program__attach_raw_tracepoint(skel->progs.trace_kfree_skb, NULL);
 	if (!ASSERT_OK_PTR(link, "attach_raw_tp"))
 		goto close_prog;
-	link_fentry = bpf_program__attach_trace(fentry);
-	if (!ASSERT_OK_PTR(link_fentry, "attach fentry"))
-		goto close_prog;
-	link_fexit = bpf_program__attach_trace(fexit);
-	if (!ASSERT_OK_PTR(link_fexit, "attach fexit"))
-		goto close_prog;
+	skel->links.trace_kfree_skb = link;
 
-	perf_buf_map = bpf_object__find_map_by_name(obj2, "perf_buf_map");
-	if (CHECK(!perf_buf_map, "find_perf_buf_map", "not found\n"))
+	link = bpf_program__attach_trace(skel->progs.fentry_eth_type_trans);
+	if (!ASSERT_OK_PTR(link, "attach fentry"))
 		goto close_prog;
+	skel->links.fentry_eth_type_trans = link;
+
+	link = bpf_program__attach_trace(skel->progs.fexit_eth_type_trans);
+	if (!ASSERT_OK_PTR(link, "attach fexit"))
+		goto close_prog;
+	skel->links.fexit_eth_type_trans = link;
 
 	/* set up perf buffer */
-	pb = perf_buffer__new(bpf_map__fd(perf_buf_map), 1,
+	pb = perf_buffer__new(bpf_map__fd(skel->maps.perf_buf_map), 1,
 			      on_sample, NULL, &passed, NULL);
 	if (!ASSERT_OK_PTR(pb, "perf_buf__new"))
 		goto close_prog;
···
 	 */
 	ASSERT_TRUE(passed, "passed");
 
-	err = bpf_map_lookup_elem(bpf_map__fd(global_data), &zero, test_ok);
+	err = bpf_map_lookup_elem(bpf_map__fd(skel->maps.bss), &zero, test_ok);
 	if (CHECK(err, "get_result",
 		  "failed to get output data: %d\n", err))
 		goto close_prog;
···
 	CHECK_FAIL(!test_ok[0] || !test_ok[1]);
 close_prog:
 	perf_buffer__free(pb);
-	bpf_link__destroy(link);
-	bpf_link__destroy(link_fentry);
-	bpf_link__destroy(link_fexit);
 	bpf_object__close(obj);
-	bpf_object__close(obj2);
+	kfree_skb__destroy(skel);
 }

tools/testing/selftests/bpf/prog_tests/kfunc_call.c

···
 #include <network_helpers.h>
 #include "kfunc_call_test.lskel.h"
 #include "kfunc_call_test_subprog.skel.h"
+#include "kfunc_call_test_subprog.lskel.h"
 
 static void test_main(void)
 {
···
 	kfunc_call_test_subprog__destroy(skel);
 }
 
+static void test_subprog_lskel(void)
+{
+	struct kfunc_call_test_subprog_lskel *skel;
+	int prog_fd, retval, err;
+
+	skel = kfunc_call_test_subprog_lskel__open_and_load();
+	if (!ASSERT_OK_PTR(skel, "skel"))
+		return;
+
+	prog_fd = skel->progs.kfunc_call_test1.prog_fd;
+	err = bpf_prog_test_run(prog_fd, 1, &pkt_v4, sizeof(pkt_v4),
+				NULL, NULL, (__u32 *)&retval, NULL);
+	ASSERT_OK(err, "bpf_prog_test_run(test1)");
+	ASSERT_EQ(retval, 10, "test1-retval");
+	ASSERT_NEQ(skel->data->active_res, -1, "active_res");
+	ASSERT_EQ(skel->data->sk_state_res, BPF_TCP_CLOSE, "sk_state_res");
+
+	kfunc_call_test_subprog_lskel__destroy(skel);
+}
+
 void test_kfunc_call(void)
 {
 	if (test__start_subtest("main"))
···
 
 	if (test__start_subtest("subprog"))
 		test_subprog();
+
+	if (test__start_subtest("subprog_lskel"))
+		test_subprog_lskel();
 }

tools/testing/selftests/bpf/prog_tests/legacy_printk.c

···
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2021 Facebook */
+#include <test_progs.h>
+#include "test_legacy_printk.skel.h"
+
+static int execute_one_variant(bool legacy)
+{
+	struct test_legacy_printk *skel;
+	int err, zero = 0, my_pid = getpid(), res, map_fd;
+
+	skel = test_legacy_printk__open();
+	if (!ASSERT_OK_PTR(skel, "skel_open"))
+		return -errno;
+
+	bpf_program__set_autoload(skel->progs.handle_legacy, legacy);
+	bpf_program__set_autoload(skel->progs.handle_modern, !legacy);
+
+	err = test_legacy_printk__load(skel);
+	/* no ASSERT_OK, we expect one of two variants can fail here */
+	if (err)
+		goto err_out;
+
+	if (legacy) {
+		map_fd = bpf_map__fd(skel->maps.my_pid_map);
+		err = bpf_map_update_elem(map_fd, &zero, &my_pid, BPF_ANY);
+		if (!ASSERT_OK(err, "my_pid_map_update"))
+			goto err_out;
+		err = bpf_map_lookup_elem(map_fd, &zero, &res);
+	} else {
+		skel->bss->my_pid_var = my_pid;
+	}
+
+	err = test_legacy_printk__attach(skel);
+	if (!ASSERT_OK(err, "skel_attach"))
+		goto err_out;
+
+	usleep(1); /* trigger */
+
+	if (legacy) {
+		map_fd = bpf_map__fd(skel->maps.res_map);
+		err = bpf_map_lookup_elem(map_fd, &zero, &res);
+		if (!ASSERT_OK(err, "res_map_lookup"))
+			goto err_out;
+	} else {
+		res = skel->bss->res_var;
+	}
+
+	if (!ASSERT_GT(res, 0, "res")) {
+		err = -EINVAL;
+		goto err_out;
+	}
+
+err_out:
+	test_legacy_printk__destroy(skel);
+	return err;
+}
+
+void test_legacy_printk(void)
+{
+	/* legacy variant should work everywhere */
+	ASSERT_OK(execute_one_variant(true /* legacy */), "legacy_case");
+
+	/* execute modern variant, can fail the load on old kernels */
+	execute_one_variant(false);
+}

tools/testing/selftests/bpf/prog_tests/log_buf.c

···
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2021 Facebook */
+
+#include <test_progs.h>
+#include <bpf/btf.h>
+
+#include "test_log_buf.skel.h"
+
+static size_t libbpf_log_pos;
+static char libbpf_log_buf[1024 * 1024];
+static bool libbpf_log_error;
+
+static int libbpf_print_cb(enum libbpf_print_level level, const char *fmt, va_list args)
+{
+	int emitted_cnt;
+	size_t left_cnt;
+
+	left_cnt = sizeof(libbpf_log_buf) - libbpf_log_pos;
+	emitted_cnt = vsnprintf(libbpf_log_buf + libbpf_log_pos, left_cnt, fmt, args);
+
+	if (emitted_cnt < 0 || emitted_cnt + 1 > left_cnt) {
+		libbpf_log_error = true;
+		return 0;
+	}
+
+	libbpf_log_pos += emitted_cnt;
+	return 0;
+}
+
+static void obj_load_log_buf(void)
+{
+	libbpf_print_fn_t old_print_cb = libbpf_set_print(libbpf_print_cb);
+	LIBBPF_OPTS(bpf_object_open_opts, opts);
+	const size_t log_buf_sz = 1024 * 1024;
+	struct test_log_buf* skel;
+	char *obj_log_buf, *good_log_buf, *bad_log_buf;
+	int err;
+
+	obj_log_buf = malloc(3 * log_buf_sz);
+	if (!ASSERT_OK_PTR(obj_log_buf, "obj_log_buf"))
+		return;
+
+	good_log_buf = obj_log_buf + log_buf_sz;
+	bad_log_buf = obj_log_buf + 2 * log_buf_sz;
+	obj_log_buf[0] = good_log_buf[0] = bad_log_buf[0] = '\0';
+
+	opts.kernel_log_buf = obj_log_buf;
+	opts.kernel_log_size = log_buf_sz;
+	opts.kernel_log_level = 4; /* for BTF this will turn into 1 */
+
+	/* In the first round every prog has its own log_buf, so libbpf logs
+	 * don't have program failure logs
+	 */
+	skel = test_log_buf__open_opts(&opts);
+	if (!ASSERT_OK_PTR(skel, "skel_open"))
+		goto cleanup;
+
+	/* set very verbose level for good_prog so we always get detailed logs */
+	bpf_program__set_log_buf(skel->progs.good_prog, good_log_buf, log_buf_sz);
+	bpf_program__set_log_level(skel->progs.good_prog, 2);
+
+	bpf_program__set_log_buf(skel->progs.bad_prog, bad_log_buf, log_buf_sz);
+	/* log_level 0 with custom log_buf means that verbose logs are not
+	 * requested if program load is successful, but libbpf should retry
+	 * with log_level 1 on error and put program's verbose load log into
+	 * custom log_buf
+	 */
+	bpf_program__set_log_level(skel->progs.bad_prog, 0);
+
+	err = test_log_buf__load(skel);
+	if (!ASSERT_ERR(err, "unexpected_load_success"))
+		goto cleanup;
+
+	ASSERT_FALSE(libbpf_log_error, "libbpf_log_error");
+
+	/* there should be no prog loading log because we specified per-prog log buf */
+	ASSERT_NULL(strstr(libbpf_log_buf, "-- BEGIN PROG LOAD LOG --"), "unexp_libbpf_log");
+	ASSERT_OK_PTR(strstr(libbpf_log_buf, "prog 'bad_prog': BPF program load failed"),
+		      "libbpf_log_not_empty");
+	ASSERT_OK_PTR(strstr(obj_log_buf, "DATASEC license"), "obj_log_not_empty");
+	ASSERT_OK_PTR(strstr(good_log_buf, "0: R1=ctx(id=0,off=0,imm=0) R10=fp0"),
+		      "good_log_verbose");
+	ASSERT_OK_PTR(strstr(bad_log_buf, "invalid access to map value, value_size=16 off=16000 size=4"),
+		      "bad_log_not_empty");
+
+	if (env.verbosity > VERBOSE_NONE) {
+		printf("LIBBPF LOG:   \n=================\n%s=================\n", libbpf_log_buf);
+		printf("OBJ LOG:      \n=================\n%s=================\n", obj_log_buf);
+		printf("GOOD_PROG LOG:\n=================\n%s=================\n", good_log_buf);
+		printf("BAD_PROG  LOG:\n=================\n%s=================\n", bad_log_buf);
+	}
+
+	/* reset everything */
+	test_log_buf__destroy(skel);
+	obj_log_buf[0] = good_log_buf[0] = bad_log_buf[0] = '\0';
+	libbpf_log_buf[0] = '\0';
+	libbpf_log_pos = 0;
+	libbpf_log_error = false;
+
+	/* In the second round we let bad_prog's failure be logged through print callback */
+	opts.kernel_log_buf = NULL; /* let everything through into print callback */
+	opts.kernel_log_size = 0;
+	opts.kernel_log_level = 1;
+
+	skel = test_log_buf__open_opts(&opts);
+	if (!ASSERT_OK_PTR(skel, "skel_open"))
+		goto cleanup;
+
+	/* set normal verbose level for good_prog to check log_level is taken into account */
+	bpf_program__set_log_buf(skel->progs.good_prog, good_log_buf, log_buf_sz);
+	bpf_program__set_log_level(skel->progs.good_prog, 1);
+
+	err = test_log_buf__load(skel);
+	if (!ASSERT_ERR(err, "unexpected_load_success"))
+		goto cleanup;
+
+	ASSERT_FALSE(libbpf_log_error, "libbpf_log_error");
+
+	/* this time prog loading error should be logged through print callback */
+	ASSERT_OK_PTR(strstr(libbpf_log_buf, "libbpf: prog 'bad_prog': -- BEGIN PROG LOAD LOG --"),
+		      "libbpf_log_correct");
+	ASSERT_STREQ(obj_log_buf, "", "obj_log__empty");
+	ASSERT_STREQ(good_log_buf, "processed 4 insns (limit 1000000) max_states_per_insn 0 total_states 0 peak_states 0 mark_read 0\n",
+		     "good_log_ok");
+	ASSERT_STREQ(bad_log_buf, "", "bad_log_empty");
+
+	if (env.verbosity > VERBOSE_NONE) {
+		printf("LIBBPF LOG:   \n=================\n%s=================\n", libbpf_log_buf);
+		printf("OBJ LOG:      \n=================\n%s=================\n", obj_log_buf);
+		printf("GOOD_PROG LOG:\n=================\n%s=================\n", good_log_buf);
+		printf("BAD_PROG  LOG:\n=================\n%s=================\n", bad_log_buf);
+	}
+
+cleanup:
+	free(obj_log_buf);
+	test_log_buf__destroy(skel);
+	libbpf_set_print(old_print_cb);
+}
+
+static void bpf_prog_load_log_buf(void)
+{
+	const struct bpf_insn good_prog_insns[] = {
+		BPF_MOV64_IMM(BPF_REG_0, 0),
+		BPF_EXIT_INSN(),
+	};
+	const size_t good_prog_insn_cnt = sizeof(good_prog_insns) / sizeof(struct bpf_insn);
+	const struct bpf_insn bad_prog_insns[] = {
+		BPF_EXIT_INSN(),
+	};
+	size_t bad_prog_insn_cnt = sizeof(bad_prog_insns) / sizeof(struct bpf_insn);
+	LIBBPF_OPTS(bpf_prog_load_opts, opts);
+	const size_t log_buf_sz = 1024 * 1024;
+	char *log_buf;
+	int fd = -1;
+
+	log_buf = malloc(log_buf_sz);
+	if (!ASSERT_OK_PTR(log_buf, "log_buf_alloc"))
+		return;
+	opts.log_buf = log_buf;
+	opts.log_size = log_buf_sz;
+
+	/* with log_level == 0 log_buf shoud stay empty for good prog */
+	log_buf[0] = '\0';
+	opts.log_level = 0;
+	fd = bpf_prog_load(BPF_PROG_TYPE_SOCKET_FILTER, "good_prog", "GPL",
+			   good_prog_insns, good_prog_insn_cnt, &opts);
+	ASSERT_STREQ(log_buf, "", "good_log_0");
+	ASSERT_GE(fd, 0, "good_fd1");
+	if (fd >= 0)
+		close(fd);
+	fd = -1;
+
+	/* log_level == 2 should always fill log_buf, even for good prog */
+	log_buf[0] = '\0';
+	opts.log_level = 2;
+	fd = bpf_prog_load(BPF_PROG_TYPE_SOCKET_FILTER, "good_prog", "GPL",
+			   good_prog_insns, good_prog_insn_cnt, &opts);
+	ASSERT_OK_PTR(strstr(log_buf, "0: R1=ctx(id=0,off=0,imm=0) R10=fp0"), "good_log_2");
+	ASSERT_GE(fd, 0, "good_fd2");
+	if (fd >= 0)
+		close(fd);
+	fd = -1;
+
+	/* log_level == 0 should fill log_buf for bad prog */
+	log_buf[0] = '\0';
+	opts.log_level = 0;
+	fd = bpf_prog_load(BPF_PROG_TYPE_SOCKET_FILTER, "bad_prog", "GPL",
+			   bad_prog_insns, bad_prog_insn_cnt, &opts);
+	ASSERT_OK_PTR(strstr(log_buf, "R0 !read_ok"), "bad_log_0");
+	ASSERT_LT(fd, 0, "bad_fd");
+	if (fd >= 0)
+		close(fd);
+	fd = -1;
+
+	free(log_buf);
+}
+
+static void bpf_btf_load_log_buf(void)
+{
+	LIBBPF_OPTS(bpf_btf_load_opts, opts);
+	const size_t log_buf_sz = 1024 * 1024;
+	const void *raw_btf_data;
+	__u32 raw_btf_size;
+	struct btf *btf;
+	char *log_buf;
+	int fd = -1;
+
+	btf = btf__new_empty();
+	if (!ASSERT_OK_PTR(btf, "empty_btf"))
+		return;
+
+	ASSERT_GT(btf__add_int(btf, "int", 4, 0), 0, "int_type");
+
+	raw_btf_data = btf__raw_data(btf, &raw_btf_size);
+	if (!ASSERT_OK_PTR(raw_btf_data, "raw_btf_data_good"))
+		goto cleanup;
+
+	log_buf = malloc(log_buf_sz);
+	if (!ASSERT_OK_PTR(log_buf, "log_buf_alloc"))
+		goto cleanup;
+	opts.log_buf = log_buf;
+	opts.log_size = log_buf_sz;
+
+	/* with log_level == 0 log_buf shoud stay empty for good BTF */
+	log_buf[0] = '\0';
+	opts.log_level = 0;
+	fd = bpf_btf_load(raw_btf_data, raw_btf_size, &opts);
+	ASSERT_STREQ(log_buf, "", "good_log_0");
+	ASSERT_GE(fd, 0, "good_fd1");
+	if (fd >= 0)
+		close(fd);
+	fd = -1;
+
+	/* log_level == 2 should always fill log_buf, even for good BTF */
+	log_buf[0] = '\0';
+	opts.log_level = 2;
+	fd = bpf_btf_load(raw_btf_data, raw_btf_size, &opts);
+	printf("LOG_BUF: %s\n", log_buf);
+	ASSERT_OK_PTR(strstr(log_buf, "magic: 0xeb9f"), "good_log_2");
+	ASSERT_GE(fd, 0, "good_fd2");
+	if (fd >= 0)
+		close(fd);
+	fd = -1;
+
+	/* make BTF bad, add pointer pointing to non-existing type */
+	ASSERT_GT(btf__add_ptr(btf, 100), 0, "bad_ptr_type");
+
+	raw_btf_data = btf__raw_data(btf, &raw_btf_size);
+	if (!ASSERT_OK_PTR(raw_btf_data, "raw_btf_data_bad"))
+		goto cleanup;
+
+	/* log_level == 0 should fill log_buf for bad BTF */
+	log_buf[0] = '\0';
+	opts.log_level = 0;
+	fd = bpf_btf_load(raw_btf_data, raw_btf_size, &opts);
+	printf("LOG_BUF: %s\n", log_buf);
+	ASSERT_OK_PTR(strstr(log_buf, "[2] PTR (anon) type_id=100 Invalid type_id"), "bad_log_0");
+	ASSERT_LT(fd, 0, "bad_fd");
+	if (fd >= 0)
+		close(fd);
+	fd = -1;
+
+cleanup:
+	free(log_buf);
+	btf__free(btf);
+}
+
+void test_log_buf(void)
+{
+	if (test__start_subtest("obj_load_log_buf"))
+		obj_load_log_buf();
+	if (test__start_subtest("bpf_prog_load_log_buf"))
+		bpf_prog_load_log_buf();
+	if (test__start_subtest("bpf_btf_load_log_buf"))
+		bpf_btf_load_log_buf();
+}

tools/testing/selftests/bpf/prog_tests/map_ptr.c

···
 #include <test_progs.h>
 #include <network_helpers.h>
 
-#include "map_ptr_kern.skel.h"
+#include "map_ptr_kern.lskel.h"
 
 void test_map_ptr(void)
 {
-	struct map_ptr_kern *skel;
+	struct map_ptr_kern_lskel *skel;
 	__u32 duration = 0, retval;
 	char buf[128];
 	int err;
 	int page_size = getpagesize();
 
-	skel = map_ptr_kern__open();
+	skel = map_ptr_kern_lskel__open();
 	if (!ASSERT_OK_PTR(skel, "skel_open"))
 		return;
 
-	err = bpf_map__set_max_entries(skel->maps.m_ringbuf, page_size);
-	if (!ASSERT_OK(err, "bpf_map__set_max_entries"))
-		goto cleanup;
+	skel->maps.m_ringbuf.max_entries = page_size;
 
-	err = map_ptr_kern__load(skel);
+	err = map_ptr_kern_lskel__load(skel);
 	if (!ASSERT_OK(err, "skel_load"))
 		goto cleanup;
 
 	skel->bss->page_size = page_size;
 
-	err = bpf_prog_test_run(bpf_program__fd(skel->progs.cg_skb), 1, &pkt_v4,
+	err = bpf_prog_test_run(skel->progs.cg_skb.prog_fd, 1, &pkt_v4,
 				sizeof(pkt_v4), buf, NULL, &retval, NULL);
 
 	if (CHECK(err, "test_run", "err=%d errno=%d\n", err, errno))
···
 		goto cleanup;
 
 cleanup:
-	map_ptr_kern__destroy(skel);
+	map_ptr_kern_lskel__destroy(skel);
 }

tools/testing/selftests/bpf/prog_tests/pinning.c

···
 		goto out;
 	}
 
-	map_fd = bpf_create_map(BPF_MAP_TYPE_ARRAY, sizeof(__u32),
-				sizeof(__u64), 1, 0);
+	map_fd = bpf_map_create(BPF_MAP_TYPE_ARRAY, NULL, sizeof(__u32),
+				sizeof(__u64), 1, NULL);
 	if (CHECK(map_fd < 0, "create pinmap manually", "fd %d\n", map_fd))
 		goto out;
 

tools/testing/selftests/bpf/prog_tests/prog_array_init.c

···
+/* SPDX-License-Identifier: GPL-2.0 */
+/* Copyright (c) 2021 Hengqi Chen */
+
+#include <test_progs.h>
+#include "test_prog_array_init.skel.h"
+
+void test_prog_array_init(void)
+{
+	struct test_prog_array_init *skel;
+	int err;
+
+	skel = test_prog_array_init__open();
+	if (!ASSERT_OK_PTR(skel, "could not open BPF object"))
+		return;
+
+	skel->rodata->my_pid = getpid();
+
+	err = test_prog_array_init__load(skel);
+	if (!ASSERT_OK(err, "could not load BPF object"))
+		goto cleanup;
+
+	skel->links.entry = bpf_program__attach_raw_tracepoint(skel->progs.entry, "sys_enter");
+	if (!ASSERT_OK_PTR(skel->links.entry, "could not attach BPF program"))
+		goto cleanup;
+
+	usleep(1);
+
+	ASSERT_EQ(skel->bss->value, 42, "unexpected value");
+
+cleanup:
+	test_prog_array_init__destroy(skel);
+}

tools/testing/selftests/bpf/prog_tests/queue_stack_map.c

···
 	int i, err, prog_fd, map_in_fd, map_out_fd;
 	char file[32], buf[128];
 	struct bpf_object *obj;
-	struct iphdr *iph = (void *)buf + sizeof(struct ethhdr);
+	struct iphdr iph;
 
 	/* Fill test values to be used */
 	for (i = 0; i < MAP_SIZE; i++)
···
 
 		err = bpf_prog_test_run(prog_fd, 1, &pkt_v4, sizeof(pkt_v4),
 					buf, &size, &retval, &duration);
-		if (err || retval || size != sizeof(pkt_v4) ||
-		    iph->daddr != val)
+		if (err || retval || size != sizeof(pkt_v4))
+			break;
+		memcpy(&iph, buf + sizeof(struct ethhdr), sizeof(iph));
+		if (iph.daddr != val)
 			break;
 	}
 
-	CHECK(err || retval || size != sizeof(pkt_v4) || iph->daddr != val,
+	CHECK(err || retval || size != sizeof(pkt_v4) || iph.daddr != val,
 	      "bpf_map_pop_elem",
 	      "err %d errno %d retval %d size %d iph->daddr %u\n",
-	      err, errno, retval, size, iph->daddr);
+	      err, errno, retval, size, iph.daddr);
 
 	/* Queue is empty, program should return TC_ACT_SHOT */
 	err = bpf_prog_test_run(prog_fd, 1, &pkt_v4, sizeof(pkt_v4),

tools/testing/selftests/bpf/prog_tests/ringbuf_multi.c

···
 	if (CHECK(err != 0, "bpf_map__set_max_entries", "bpf_map__set_max_entries failed\n"))
 		goto cleanup;
 
-	proto_fd = bpf_create_map(BPF_MAP_TYPE_RINGBUF, 0, 0, page_size, 0);
-	if (CHECK(proto_fd < 0, "bpf_create_map", "bpf_create_map failed\n"))
+	proto_fd = bpf_map_create(BPF_MAP_TYPE_RINGBUF, NULL, 0, 0, page_size, NULL);
+	if (CHECK(proto_fd < 0, "bpf_map_create", "bpf_map_create failed\n"))
 		goto cleanup;
 
 	err = bpf_map__set_inner_map_fd(skel->maps.ringbuf_hash, proto_fd);

tools/testing/selftests/bpf/prog_tests/select_reuseport.c

···
 
 static int create_maps(enum bpf_map_type inner_type)
 {
-	struct bpf_create_map_attr attr = {};
+	LIBBPF_OPTS(bpf_map_create_opts, opts);
 
 	inner_map_type = inner_type;
 
 	/* Creating reuseport_array */
-	attr.name = "reuseport_array";
-	attr.map_type = inner_type;
-	attr.key_size = sizeof(__u32);
-	attr.value_size = sizeof(__u32);
-	attr.max_entries = REUSEPORT_ARRAY_SIZE;
-
-	reuseport_array = bpf_create_map_xattr(&attr);
+	reuseport_array = bpf_map_create(inner_type, "reuseport_array",
+					 sizeof(__u32), sizeof(__u32), REUSEPORT_ARRAY_SIZE, NULL);
 	RET_ERR(reuseport_array < 0, "creating reuseport_array",
 		"reuseport_array:%d errno:%d\n", reuseport_array, errno);
 
 	/* Creating outer_map */
-	attr.name = "outer_map";
-	attr.map_type = BPF_MAP_TYPE_ARRAY_OF_MAPS;
-	attr.key_size = sizeof(__u32);
-	attr.value_size = sizeof(__u32);
-	attr.max_entries = 1;
-	attr.inner_map_fd = reuseport_array;
-	outer_map = bpf_create_map_xattr(&attr);
+	opts.inner_map_fd = reuseport_array;
+	outer_map = bpf_map_create(BPF_MAP_TYPE_ARRAY_OF_MAPS, "outer_map",
+				   sizeof(__u32), sizeof(__u32), 1, &opts);
 	RET_ERR(outer_map < 0, "creating outer_map",
 		"outer_map:%d errno:%d\n", outer_map, errno);
 

tools/testing/selftests/bpf/prog_tests/sockmap_basic.c

···
 	if (CHECK_FAIL(s < 0))
 		return;
 
-	map = bpf_create_map(map_type, sizeof(int), sizeof(int), 1, 0);
+	map = bpf_map_create(map_type, NULL, sizeof(int), sizeof(int), 1, NULL);
 	if (CHECK_FAIL(map < 0)) {
-		perror("bpf_create_map");
+		perror("bpf_cmap_create");
 		goto out;
 	}
 

tools/testing/selftests/bpf/prog_tests/sockmap_ktls.c

···
 	char test_name[MAX_TEST_NAME];
 	int map;
 
-	map = bpf_create_map(map_type, sizeof(int), sizeof(int), 1, 0);
+	map = bpf_map_create(map_type, NULL, sizeof(int), sizeof(int), 1, NULL);
 	if (CHECK_FAIL(map < 0)) {
 		perror("bpf_map_create");
 		return;

tools/testing/selftests/bpf/prog_tests/sockmap_listen.c

···
 	if (s < 0)
 		return;
 
-	mapfd = bpf_create_map(BPF_MAP_TYPE_SOCKMAP, sizeof(key),
-			       sizeof(value32), 1, 0);
+	mapfd = bpf_map_create(BPF_MAP_TYPE_SOCKMAP, NULL, sizeof(key),
+			       sizeof(value32), 1, NULL);
 	if (mapfd < 0) {
 		FAIL_ERRNO("map_create");
 		goto close;

tools/testing/selftests/bpf/prog_tests/sockopt_inherit.c

···
 
 static void run_test(int cgroup_fd)
 {
-	struct bpf_prog_load_attr attr = {
-		.file = "./sockopt_inherit.o",
-	};
 	int server_fd = -1, client_fd;
 	struct bpf_object *obj;
 	void *server_err;
 	pthread_t tid;
-	int ignored;
 	int err;
 
-	err = bpf_prog_load_xattr(&attr, &obj, &ignored);
-	if (CHECK_FAIL(err))
+	obj = bpf_object__open_file("sockopt_inherit.o", NULL);
+	if (!ASSERT_OK_PTR(obj, "obj_open"))
 		return;
+
+	err = bpf_object__load(obj);
+	if (!ASSERT_OK(err, "obj_load"))
+		goto close_bpf_object;
 
 	err = prog_attach(obj, cgroup_fd, "cgroup/getsockopt");
 	if (CHECK_FAIL(err))

tools/testing/selftests/bpf/prog_tests/sockopt_multi.c

···
 
 void test_sockopt_multi(void)
 {
-	struct bpf_prog_load_attr attr = {
-		.file = "./sockopt_multi.o",
-	};
 	int cg_parent = -1, cg_child = -1;
 	struct bpf_object *obj = NULL;
 	int sock_fd = -1;
 	int err = -1;
-	int ignored;
 
 	cg_parent = test__join_cgroup("/parent");
 	if (CHECK_FAIL(cg_parent < 0))
···
 	if (CHECK_FAIL(cg_child < 0))
 		goto out;
 
-	err = bpf_prog_load_xattr(&attr, &obj, &ignored);
-	if (CHECK_FAIL(err))
+	obj = bpf_object__open_file("sockopt_multi.o", NULL);
+	if (!ASSERT_OK_PTR(obj, "obj_load"))
+		goto out;
+
+	err = bpf_object__load(obj);
+	if (!ASSERT_OK(err, "obj_load"))
 		goto out;
 
 	sock_fd = socket(AF_INET, SOCK_STREAM, 0);

tools/testing/selftests/bpf/prog_tests/tcp_rtt.c

···
 #include <test_progs.h>
 #include "cgroup_helpers.h"
 #include "network_helpers.h"
+#include "tcp_rtt.skel.h"
 
 struct tcp_rtt_storage {
 	__u32 invoked;
···
 
 static int run_test(int cgroup_fd, int server_fd)
 {
-	struct bpf_prog_load_attr attr = {
-		.prog_type = BPF_PROG_TYPE_SOCK_OPS,
-		.file = "./tcp_rtt.o",
-		.expected_attach_type = BPF_CGROUP_SOCK_OPS,
-	};
-	struct bpf_object *obj;
-	struct bpf_map *map;
+	struct tcp_rtt *skel;
 	int client_fd;
 	int prog_fd;
 	int map_fd;
 	int err;
 
-	err = bpf_prog_load_xattr(&attr, &obj, &prog_fd);
-	if (err) {
-		log_err("Failed to load BPF object");
+	skel = tcp_rtt__open_and_load();
+	if (!ASSERT_OK_PTR(skel, "skel_open_load"))
 		return -1;
-	}
 
-	map = bpf_object__next_map(obj, NULL);
-	map_fd = bpf_map__fd(map);
+	map_fd = bpf_map__fd(skel->maps.socket_storage_map);
+	prog_fd = bpf_program__fd(skel->progs._sockops);
 
 	err = bpf_prog_attach(prog_fd, cgroup_fd, BPF_CGROUP_SOCK_OPS, 0);
 	if (err) {
···
 	close(client_fd);
 
 close_bpf_object:
-	bpf_object__close(obj);
+	tcp_rtt__destroy(skel);
 	return err;
 }
 

tools/testing/selftests/bpf/prog_tests/test_bpffs.c

···
 	fd = open(file, 0);
 	if (fd < 0)
 		return -1;
-	while ((len = read(fd, buf, sizeof(buf))) > 0)
+	while ((len = read(fd, buf, sizeof(buf))) > 0) {
+		buf[sizeof(buf) - 1] = '\0';
 		if (strstr(buf, "iter")) {
 			close(fd);
 			return 0;
 		}
+	}
 	close(fd);
 	return -1;
 }
···
 	if (!ASSERT_OK(err, "creating " TDIR "/fs1/b"))
 		goto out;
 
-	map = bpf_create_map(BPF_MAP_TYPE_ARRAY, 4, 4, 1, 0);
+	map = bpf_map_create(BPF_MAP_TYPE_ARRAY, NULL, 4, 4, 1, NULL);
 	if (!ASSERT_GT(map, 0, "create_map(ARRAY)"))
 		goto out;
 	err = bpf_obj_pin(map, TDIR "/fs1/c");

tools/testing/selftests/bpf/prog_tests/test_global_funcs.c

···
 
 static int check_load(const char *file)
 {
-	struct bpf_prog_load_attr attr;
 	struct bpf_object *obj = NULL;
-	int err, prog_fd;
+	struct bpf_program *prog;
+	int err;
 
-	memset(&attr, 0, sizeof(struct bpf_prog_load_attr));
-	attr.file = file;
-	attr.prog_type = BPF_PROG_TYPE_UNSPEC;
-	attr.log_level = extra_prog_load_log_flags;
-	attr.prog_flags = BPF_F_TEST_RND_HI32;
 	found = false;
-	err = bpf_prog_load_xattr(&attr, &obj, &prog_fd);
+
+	obj = bpf_object__open_file(file, NULL);
+	err = libbpf_get_error(obj);
+	if (err)
+		return err;
+
+	prog = bpf_object__next_program(obj, NULL);
+	if (!prog) {
+		err = -ENOENT;
+		goto err_out;
+	}
+
+	bpf_program__set_flags(prog, BPF_F_TEST_RND_HI32);
+	bpf_program__set_log_level(prog, extra_prog_load_log_flags);
+
+	err = bpf_object__load(obj);
+
+err_out:
 	bpf_object__close(obj);
 	return err;
 }

tools/testing/selftests/bpf/prog_tests/xdp.c

···
 	const char *file = "./test_xdp.o";
 	struct bpf_object *obj;
 	char buf[128];
-	struct ipv6hdr *iph6 = (void *)buf + sizeof(struct ethhdr);
-	struct iphdr *iph = (void *)buf + sizeof(struct ethhdr);
+	struct ipv6hdr iph6;
+	struct iphdr iph;
 	__u32 duration, retval, size;
 	int err, prog_fd, map_fd;
 
···
 
 	err = bpf_prog_test_run(prog_fd, 1, &pkt_v4, sizeof(pkt_v4),
 				buf, &size, &retval, &duration);
-
+	memcpy(&iph, buf + sizeof(struct ethhdr), sizeof(iph));
 	CHECK(err || retval != XDP_TX || size != 74 ||
-	      iph->protocol != IPPROTO_IPIP, "ipv4",
+	      iph.protocol != IPPROTO_IPIP, "ipv4",
 	      "err %d errno %d retval %d size %d\n",
 	      err, errno, retval, size);
 
 	err = bpf_prog_test_run(prog_fd, 1, &pkt_v6, sizeof(pkt_v6),
 				buf, &size, &retval, &duration);
+	memcpy(&iph6, buf + sizeof(struct ethhdr), sizeof(iph6));
 	CHECK(err || retval != XDP_TX || size != 114 ||
-	      iph6->nexthdr != IPPROTO_IPV6, "ipv6",
+	      iph6.nexthdr != IPPROTO_IPV6, "ipv6",
 	      "err %d errno %d retval %d size %d\n",
 	      err, errno, retval, size);
 out:

tools/testing/selftests/bpf/prog_tests/xdp_bonding.c

···
 		.h_dest = BOND2_MAC,
 		.h_proto = htons(ETH_P_IP),
 	};
-	uint8_t buf[128] = {};
-	struct iphdr *iph = (struct iphdr *)(buf + sizeof(eh));
-	struct udphdr *uh = (struct udphdr *)(buf + sizeof(eh) + sizeof(*iph));
+	struct iphdr iph = {};
+	struct udphdr uh = {};
+	uint8_t buf[128];
 	int i, s = -1;
 	int ifindex;
 
···
 	if (!ASSERT_GT(ifindex, 0, "get bond1 ifindex"))
 		goto err;
 
-	memcpy(buf, &eh, sizeof(eh));
-	iph->ihl = 5;
-	iph->version = 4;
-	iph->tos = 16;
-	iph->id = 1;
-	iph->ttl = 64;
-	iph->protocol = IPPROTO_UDP;
-	iph->saddr = 1;
-	iph->daddr = 2;
-	iph->tot_len = htons(sizeof(buf) - ETH_HLEN);
-	iph->check = 0;
+	iph.ihl = 5;
+	iph.version = 4;
+	iph.tos = 16;
+	iph.id = 1;
+	iph.ttl = 64;
+	iph.protocol = IPPROTO_UDP;
+	iph.saddr = 1;
+	iph.daddr = 2;
+	iph.tot_len = htons(sizeof(buf) - ETH_HLEN);
+	iph.check = 0;
 
 	for (i = 1; i <= NPACKETS; i++) {
 		int n;
···
 		};
 
 		/* vary the UDP destination port for even distribution with roundrobin/xor modes */
-		uh->dest++;
+		uh.dest++;
 
 		if (vary_dst_ip)
-			iph->daddr++;
+			iph.daddr++;
+
+		/* construct a packet */
+		memcpy(buf, &eh, sizeof(eh));
+		memcpy(buf + sizeof(eh), &iph, sizeof(iph));
+		memcpy(buf + sizeof(eh) + sizeof(iph), &uh, sizeof(uh));
 
 		n = sendto(s, buf, sizeof(buf), 0, (struct sockaddr *)&saddr_ll, sizeof(saddr_ll));
 		if (!ASSERT_EQ(n, sizeof(buf), "sendto"))

tools/testing/selftests/bpf/prog_tests/xdp_bpf2bpf.c

···
 	char buf[128];
 	int err, pkt_fd, map_fd;
 	bool passed = false;
-	struct iphdr *iph = (void *)buf + sizeof(struct ethhdr);
+	struct iphdr iph;
 	struct iptnl_info value4 = {.family = AF_INET};
 	struct test_xdp *pkt_skel = NULL;
 	struct test_xdp_bpf2bpf *ftrace_skel = NULL;
···
 	/* Run test program */
 	err = bpf_prog_test_run(pkt_fd, 1, &pkt_v4, sizeof(pkt_v4),
 				buf, &size, &retval, &duration);
-
+	memcpy(&iph, buf + sizeof(struct ethhdr), sizeof(iph));
 	if (CHECK(err || retval != XDP_TX || size != 74 ||
-		  iph->protocol != IPPROTO_IPIP, "ipv4",
+		  iph.protocol != IPPROTO_IPIP, "ipv4",
 		  "err %d errno %d retval %d size %d\n",
 		  err, errno, retval, size))
 		goto out;

tools/testing/selftests/bpf/progs/bpf_loop.c

···
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2021 Facebook */
+
+#include "vmlinux.h"
+#include <bpf/bpf_helpers.h>
+
+char _license[] SEC("license") = "GPL";
+
+struct callback_ctx {
+	int output;
+};
+
+/* These should be set by the user program */
+u32 nested_callback_nr_loops;
+u32 stop_index = -1;
+u32 nr_loops;
+int pid;
+
+/* Making these global variables so that the userspace program
+ * can verify the output through the skeleton
+ */
+int nr_loops_returned;
+int g_output;
+int err;
+
+static int callback(__u32 index, void *data)
+{
+	struct callback_ctx *ctx = data;
+
+	if (index >= stop_index)
+		return 1;
+
+	ctx->output += index;
+
+	return 0;
+}
+
+static int empty_callback(__u32 index, void *data)
+{
+	return 0;
+}
+
+static int nested_callback2(__u32 index, void *data)
+{
+	nr_loops_returned += bpf_loop(nested_callback_nr_loops, callback, data, 0);
+
+	return 0;
+}
+
+static int nested_callback1(__u32 index, void *data)
+{
+	bpf_loop(nested_callback_nr_loops, nested_callback2, data, 0);
+	return 0;
+}
+
+SEC("fentry/__x64_sys_nanosleep")
+int test_prog(void *ctx)
+{
+	struct callback_ctx data = {};
+
+	if (bpf_get_current_pid_tgid() >> 32 != pid)
+		return 0;
+
+	nr_loops_returned = bpf_loop(nr_loops, callback, &data, 0);
+
+	if (nr_loops_returned < 0)
+		err = nr_loops_returned;
+	else
+		g_output = data.output;
+
+	return 0;
+}
+
+SEC("fentry/__x64_sys_nanosleep")
+int prog_null_ctx(void *ctx)
+{
+	if (bpf_get_current_pid_tgid() >> 32 != pid)
+		return 0;
+
+	nr_loops_returned = bpf_loop(nr_loops, empty_callback, NULL, 0);
+
+	return 0;
+}
+
+SEC("fentry/__x64_sys_nanosleep")
+int prog_invalid_flags(void *ctx)
+{
+	struct callback_ctx data = {};
+
+	if (bpf_get_current_pid_tgid() >> 32 != pid)
+		return 0;
+
+	err = bpf_loop(nr_loops, callback, &data, 1);
+
+	return 0;
+}
+
+SEC("fentry/__x64_sys_nanosleep")
+int prog_nested_calls(void *ctx)
+{
+	struct callback_ctx data = {};
+
+	if (bpf_get_current_pid_tgid() >> 32 != pid)
+		return 0;
+
+	nr_loops_returned = 0;
+	bpf_loop(nr_loops, nested_callback1, &data, 0);
+
+	g_output = data.output;
+
+	return 0;
+}

tools/testing/selftests/bpf/progs/bpf_loop_bench.c

···
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2021 Facebook */
+
+#include "vmlinux.h"
+#include <bpf/bpf_helpers.h>
+
+char _license[] SEC("license") = "GPL";
+
+u32 nr_loops;
+long hits;
+
+static int empty_callback(__u32 index, void *data)
+{
+	return 0;
+}
+
+SEC("fentry/__x64_sys_getpgid")
+int benchmark(void *ctx)
+{
+	for (int i = 0; i < 1000; i++) {
+		bpf_loop(nr_loops, empty_callback, NULL, 0);
+
+		__sync_add_and_fetch(&hits, nr_loops);
+	}
+	return 0;
+}

tools/testing/selftests/bpf/progs/core_kern.c

···
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2021 Facebook */
+#include "vmlinux.h"
+
+#include <bpf/bpf_helpers.h>
+#include <bpf/bpf_tracing.h>
+#include <bpf/bpf_core_read.h>
+
+#define ATTR __always_inline
+#include "test_jhash.h"
+
+struct {
+	__uint(type, BPF_MAP_TYPE_ARRAY);
+	__type(key, u32);
+	__type(value, u32);
+	__uint(max_entries, 256);
+} array1 SEC(".maps");
+
+struct {
+	__uint(type, BPF_MAP_TYPE_ARRAY);
+	__type(key, u32);
+	__type(value, u32);
+	__uint(max_entries, 256);
+} array2 SEC(".maps");
+
+static __noinline int randmap(int v, const struct net_device *dev)
+{
+	struct bpf_map *map = (struct bpf_map *)&array1;
+	int key = bpf_get_prandom_u32() & 0xff;
+	int *val;
+
+	if (bpf_get_prandom_u32() & 1)
+		map = (struct bpf_map *)&array2;
+
+	val = bpf_map_lookup_elem(map, &key);
+	if (val)
+		*val = bpf_get_prandom_u32() + v + dev->mtu;
+
+	return 0;
+}
+
+SEC("tp_btf/xdp_devmap_xmit")
+int BPF_PROG(tp_xdp_devmap_xmit_multi, const struct net_device
+	     *from_dev, const struct net_device *to_dev, int sent, int drops,
+	     int err)
+{
+	return randmap(from_dev->ifindex, from_dev);
+}
+
+SEC("fentry/eth_type_trans")
+int BPF_PROG(fentry_eth_type_trans, struct sk_buff *skb,
+	     struct net_device *dev, unsigned short protocol)
+{
+	return randmap(dev->ifindex + skb->len, dev);
+}
+
+SEC("fexit/eth_type_trans")
+int BPF_PROG(fexit_eth_type_trans, struct sk_buff *skb,
+	     struct net_device *dev, unsigned short protocol)
+{
+	return randmap(dev->ifindex + skb->len, dev);
+}
+
+volatile const int never;
+
+struct __sk_bUfF /* it will not exist in vmlinux */ {
+	int len;
+} __attribute__((preserve_access_index));
+
+struct bpf_testmod_test_read_ctx /* it exists in bpf_testmod */ {
+	size_t len;
+} __attribute__((preserve_access_index));
+
+SEC("tc")
+int balancer_ingress(struct __sk_buff *ctx)
+{
+	void *data_end = (void *)(long)ctx->data_end;
+	void *data = (void *)(long)ctx->data;
+	void *ptr;
+	int ret = 0, nh_off, i = 0;
+
+	nh_off = 14;
+
+	/* pragma unroll doesn't work on large loops */
+#define C do { \
+	ptr = data + i; \
+	if (ptr + nh_off > data_end) \
+		break; \
+	ctx->tc_index = jhash(ptr, nh_off, ctx->cb[0] + i++); \
+	if (never) { \
+		/* below is a dead code with unresolvable CO-RE relo */ \
+		i += ((struct __sk_bUfF *)ctx)->len; \
+		/* this CO-RE relo may or may not resolve
+		 * depending on whether bpf_testmod is loaded.
+		 */ \
+		i += ((struct bpf_testmod_test_read_ctx *)ctx)->len; \
+	} \
+	} while (0);
+#define C30 C;C;C;C;C;C;C;C;C;C;C;C;C;C;C;C;C;C;C;C;C;C;C;C;C;C;C;C;C;C;
+	C30;C30;C30; /* 90 calls */
+	return 0;
+}
+
+char LICENSE[] SEC("license") = "GPL";

tools/testing/selftests/bpf/progs/map_ptr_kern.c

···
 	return 1;
 }
 
+#define INNER_MAX_ENTRIES 1234
+
 struct inner_map {
 	__uint(type, BPF_MAP_TYPE_ARRAY);
-	__uint(max_entries, 1);
+	__uint(max_entries, INNER_MAX_ENTRIES);
 	__type(key, __u32);
 	__type(value, __u32);
 } inner_map SEC(".maps");
···
 	__type(value, __u32);
 	__array(values, struct {
 		__uint(type, BPF_MAP_TYPE_ARRAY);
-		__uint(max_entries, 1);
+		__uint(max_entries, INNER_MAX_ENTRIES);
 		__type(key, __u32);
 		__type(value, __u32);
 	});
···
 {
 	struct bpf_array *array_of_maps = (struct bpf_array *)&m_array_of_maps;
 	struct bpf_map *map = (struct bpf_map *)&m_array_of_maps;
+	struct bpf_array *inner_map;
+	int key = 0;
 
 	VERIFY(check_default(&array_of_maps->map, map));
+	inner_map = bpf_map_lookup_elem(array_of_maps, &key);
+	VERIFY(inner_map != 0);
+	VERIFY(inner_map->map.max_entries == INNER_MAX_ENTRIES);
 
 	return 1;
 }
···
 {
 	struct bpf_htab *hash_of_maps = (struct bpf_htab *)&m_hash_of_maps;
 	struct bpf_map *map = (struct bpf_map *)&m_hash_of_maps;
+	struct bpf_htab *inner_map;
+	int key = 2;
 
 	VERIFY(check_default(&hash_of_maps->map, map));
+	inner_map = bpf_map_lookup_elem(hash_of_maps, &key);
+	VERIFY(inner_map != 0);
+	VERIFY(inner_map->map.max_entries == INNER_MAX_ENTRIES);
 
 	return 1;
 }

tools/testing/selftests/bpf/progs/pyperf.h

···
 	__uint(value_size, sizeof(long long) * 127);
 } stackmap SEC(".maps");
 
+#ifdef USE_BPF_LOOP
+struct process_frame_ctx {
+	int cur_cpu;
+	int32_t *symbol_counter;
+	void *frame_ptr;
+	FrameData *frame;
+	PidData *pidData;
+	Symbol *sym;
+	Event *event;
+	bool done;
+};
+
+#define barrier_var(var) asm volatile("" : "=r"(var) : "0"(var))
+
+static int process_frame_callback(__u32 i, struct process_frame_ctx *ctx)
+{
+	int zero = 0;
+	void *frame_ptr = ctx->frame_ptr;
+	PidData *pidData = ctx->pidData;
+	FrameData *frame = ctx->frame;
+	int32_t *symbol_counter = ctx->symbol_counter;
+	int cur_cpu = ctx->cur_cpu;
+	Event *event = ctx->event;
+	Symbol *sym = ctx->sym;
+
+	if (frame_ptr && get_frame_data(frame_ptr, pidData, frame, sym)) {
+		int32_t new_symbol_id = *symbol_counter * 64 + cur_cpu;
+		int32_t *symbol_id = bpf_map_lookup_elem(&symbolmap, sym);
+
+		if (!symbol_id) {
+			bpf_map_update_elem(&symbolmap, sym, &zero, 0);
+			symbol_id = bpf_map_lookup_elem(&symbolmap, sym);
+			if (!symbol_id) {
+				ctx->done = true;
+				return 1;
+			}
+		}
+		if (*symbol_id == new_symbol_id)
+			(*symbol_counter)++;
+
+		barrier_var(i);
+		if (i >= STACK_MAX_LEN)
+			return 1;
+
+		event->stack[i] = *symbol_id;
+
+		event->stack_len = i + 1;
+		frame_ptr = frame->f_back;
+	}
+	return 0;
+}
+#endif /* USE_BPF_LOOP */
+
 #ifdef GLOBAL_FUNC
 __noinline
 #elif defined(SUBPROGS)
···
 		int32_t* symbol_counter = bpf_map_lookup_elem(&symbolmap, &sym);
 		if (symbol_counter == NULL)
 			return 0;
+#ifdef USE_BPF_LOOP
+	struct process_frame_ctx ctx = {
+		.cur_cpu = cur_cpu,
+		.symbol_counter = symbol_counter,
+		.frame_ptr = frame_ptr,
+		.frame = &frame,
+		.pidData = pidData,
+		.sym = &sym,
+		.event = event,
+	};
+
+	bpf_loop(STACK_MAX_LEN, process_frame_callback, &ctx, 0);
+	if (ctx.done)
+		return 0;
+#else
 #ifdef NO_UNROLL
 #pragma clang loop unroll(disable)
 #else
 #pragma clang loop unroll(full)
-#endif
+#endif /* NO_UNROLL */
 		/* Unwind python stack */
 		for (int i = 0; i < STACK_MAX_LEN; ++i) {
 			if (frame_ptr && get_frame_data(frame_ptr, pidData, &frame, &sym)) {
···
 				frame_ptr = frame.f_back;
 			}
 		}
+#endif /* USE_BPF_LOOP */
 		event->stack_complete = frame_ptr == NULL;
 	} else {
 		event->stack_complete = 1;

tools/testing/selftests/bpf/progs/pyperf600_bpf_loop.c

···
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2021 Facebook */
+
+#define STACK_MAX_LEN 600
+#define USE_BPF_LOOP
+#include "pyperf.h"

tools/testing/selftests/bpf/progs/strobemeta.h

···
 	return payload;
 }
 
+#ifdef USE_BPF_LOOP
+enum read_type {
+	READ_INT_VAR,
+	READ_MAP_VAR,
+	READ_STR_VAR,
+};
+
+struct read_var_ctx {
+	struct strobemeta_payload *data;
+	void *tls_base;
+	struct strobemeta_cfg *cfg;
+	void *payload;
+	/* value gets mutated */
+	struct strobe_value_generic *value;
+	enum read_type type;
+};
+
+static int read_var_callback(__u32 index, struct read_var_ctx *ctx)
+{
+	switch (ctx->type) {
+	case READ_INT_VAR:
+		if (index >= STROBE_MAX_INTS)
+			return 1;
+		read_int_var(ctx->cfg, index, ctx->tls_base, ctx->value, ctx->data);
+		break;
+	case READ_MAP_VAR:
+		if (index >= STROBE_MAX_MAPS)
+			return 1;
+		ctx->payload = read_map_var(ctx->cfg, index, ctx->tls_base,
+					    ctx->value, ctx->data, ctx->payload);
+		break;
+	case READ_STR_VAR:
+		if (index >= STROBE_MAX_STRS)
+			return 1;
+		ctx->payload += read_str_var(ctx->cfg, index, ctx->tls_base,
+					     ctx->value, ctx->data, ctx->payload);
+		break;
+	}
+	return 0;
+}
+#endif /* USE_BPF_LOOP */
+
 /*
  * read_strobe_meta returns NULL, if no metadata was read; otherwise returns
  * pointer to *right after* payload ends
···
 	 */
 	tls_base = (void *)task;
 
+#ifdef USE_BPF_LOOP
+	struct read_var_ctx ctx = {
+		.cfg = cfg,
+		.tls_base = tls_base,
+		.value = &value,
+		.data = data,
+		.payload = payload,
+	};
+	int err;
+
+	ctx.type = READ_INT_VAR;
+	err = bpf_loop(STROBE_MAX_INTS, read_var_callback, &ctx, 0);
+	if (err != STROBE_MAX_INTS)
+		return NULL;
+
+	ctx.type = READ_STR_VAR;
+	err = bpf_loop(STROBE_MAX_STRS, read_var_callback, &ctx, 0);
+	if (err != STROBE_MAX_STRS)
+		return NULL;
+
+	ctx.type = READ_MAP_VAR;
+	err = bpf_loop(STROBE_MAX_MAPS, read_var_callback, &ctx, 0);
+	if (err != STROBE_MAX_MAPS)
+		return NULL;
+#else
 #ifdef NO_UNROLL
 #pragma clang loop unroll(disable)
 #else
 #pragma unroll
-#endif
+#endif /* NO_UNROLL */
 	for (int i = 0; i < STROBE_MAX_INTS; ++i) {
 		read_int_var(cfg, i, tls_base, &value, data);
 	}
···
 #pragma clang loop unroll(disable)
 #else
 #pragma unroll
-#endif
+#endif /* NO_UNROLL */
 	for (int i = 0; i < STROBE_MAX_STRS; ++i) {
 		payload += read_str_var(cfg, i, tls_base, &value, data, payload);
 	}
···
 #pragma clang loop unroll(disable)
 #else
 #pragma unroll
-#endif
+#endif /* NO_UNROLL */
 	for (int i = 0; i < STROBE_MAX_MAPS; ++i) {
 		payload = read_map_var(cfg, i, tls_base, &value, data, payload);
 	}
+#endif /* USE_BPF_LOOP */
+
 	/*
 	 * return pointer right after end of payload, so it's possible to
 	 * calculate exact amount of useful data that needs to be sent

tools/testing/selftests/bpf/progs/strobemeta_bpf_loop.c

···
+// SPDX-License-Identifier: (LGPL-2.1 OR BSD-2-Clause)
+/* Copyright (c) 2021 Facebook */
+
+#define STROBE_MAX_INTS 2
+#define STROBE_MAX_STRS 25
+#define STROBE_MAX_MAPS 100
+#define STROBE_MAX_MAP_ENTRIES 20
+#define USE_BPF_LOOP
+#include "strobemeta.h"

tools/testing/selftests/bpf/progs/test_ksyms_weak.c

···
 	/* tests existing symbols. */
 	rq = (struct rq *)bpf_per_cpu_ptr(&runqueues, 0);
 	if (rq)
-		out__existing_typed = 0;
+		out__existing_typed = rq->cpu;
 	out__existing_typeless = (__u64)&bpf_prog_active;
 
 	/* tests non-existent symbols. */

tools/testing/selftests/bpf/progs/test_legacy_printk.c

···
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2021 Facebook */
+
+#include <linux/bpf.h>
+#define BPF_NO_GLOBAL_DATA
+#include <bpf/bpf_helpers.h>
+
+char LICENSE[] SEC("license") = "GPL";
+
+struct {
+	__uint(type, BPF_MAP_TYPE_ARRAY);
+	__type(key, int);
+	__type(value, int);
+	__uint(max_entries, 1);
+} my_pid_map SEC(".maps");
+
+struct {
+	__uint(type, BPF_MAP_TYPE_ARRAY);
+	__type(key, int);
+	__type(value, int);
+	__uint(max_entries, 1);
+} res_map SEC(".maps");
+
+volatile int my_pid_var = 0;
+volatile int res_var = 0;
+
+SEC("tp/raw_syscalls/sys_enter")
+int handle_legacy(void *ctx)
+{
+	int zero = 0, *my_pid, cur_pid, *my_res;
+
+	my_pid = bpf_map_lookup_elem(&my_pid_map, &zero);
+	if (!my_pid)
+		return 1;
+
+	cur_pid = bpf_get_current_pid_tgid() >> 32;
+	if (cur_pid != *my_pid)
+		return 1;
+
+	my_res = bpf_map_lookup_elem(&res_map, &zero);
+	if (!my_res)
+		return 1;
+
+	if (*my_res == 0)
+		/* use bpf_printk() in combination with BPF_NO_GLOBAL_DATA to
+		 * force .rodata.str1.1 section that previously caused
+		 * problems on old kernels due to libbpf always tried to
+		 * create a global data map for it
+		 */
+		bpf_printk("Legacy-case bpf_printk test, pid %d\n", cur_pid);
+	*my_res = 1;
+
+	return *my_res;
+}
+
+SEC("tp/raw_syscalls/sys_enter")
+int handle_modern(void *ctx)
+{
+	int zero = 0, cur_pid;
+
+	cur_pid = bpf_get_current_pid_tgid() >> 32;
+	if (cur_pid != my_pid_var)
+		return 1;
+
+	if (res_var == 0)
+		/* we need bpf_printk() to validate libbpf logic around unused
+		 * global maps and legacy kernels; see comment in handle_legacy()
+		 */
+		bpf_printk("Modern-case bpf_printk test, pid %d\n", cur_pid);
+	res_var = 1;
+
+	return res_var;
+}

tools/testing/selftests/bpf/progs/test_log_buf.c

···
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright (c) 2021 Facebook */
+
+#include <linux/bpf.h>
+#include <bpf/bpf_helpers.h>
+
+int a[4];
+const volatile int off = 4000;
+
+SEC("raw_tp/sys_enter")
+int good_prog(const void *ctx)
+{
+	a[0] = (int)(long)ctx;
+	return a[1];
+}
+
+SEC("raw_tp/sys_enter")
+int bad_prog(const void *ctx)
+{
+	/* out of bounds access */
+	return a[off];
+}
+
+char _license[] SEC("license") = "GPL";

tools/testing/selftests/bpf/progs/test_prog_array_init.c

···
+/* SPDX-License-Identifier: GPL-2.0 */
+/* Copyright (c) 2021 Hengqi Chen */
+
+#include "vmlinux.h"
+#include <bpf/bpf_helpers.h>
+#include <bpf/bpf_tracing.h>
+
+const volatile pid_t my_pid = 0;
+int value = 0;
+
+SEC("raw_tp/sys_enter")
+int tailcall_1(void *ctx)
+{
+	value = 42;
+	return 0;
+}
+
+struct {
+	__uint(type, BPF_MAP_TYPE_PROG_ARRAY);
+	__uint(max_entries, 2);
+	__uint(key_size, sizeof(__u32));
+	__array(values, int (void *));
+} prog_array_init SEC(".maps") = {
+	.values = {
+		[1] = (void *)&tailcall_1,
+	},
+};
+
+SEC("raw_tp/sys_enter")
+int entry(void *ctx)
+{
+	pid_t pid = bpf_get_current_pid_tgid() >> 32;
+
+	if (pid != my_pid)
+		return 0;
+
+	bpf_tail_call(ctx, &prog_array_init, 1);
+	return 0;
+}

tools/testing/selftests/bpf/progs/test_verif_scale2.c

···
 // SPDX-License-Identifier: GPL-2.0
 // Copyright (c) 2019 Facebook
-#include <linux/bpf.h>
+#include "vmlinux.h"
 #include <bpf/bpf_helpers.h>
 #define ATTR __always_inline
 #include "test_jhash.h"
 
-SEC("scale90_inline")
+SEC("tc")
 int balancer_ingress(struct __sk_buff *ctx)
 {
 	void *data_end = (void *)(long)ctx->data_end;

tools/testing/selftests/bpf/progs/trigger_bench.c

···
 	__sync_add_and_fetch(&hits, 1);
 	return -22;
 }
+
+SEC("uprobe/self/uprobe_target")
+int bench_trigger_uprobe(void *ctx)
+{
+	__sync_add_and_fetch(&hits, 1);
+	return 0;
+}

tools/testing/selftests/bpf/test_bpftool_synctypes.py

···
 
 LINUX_ROOT = os.path.abspath(os.path.join(__file__,
     os.pardir, os.pardir, os.pardir, os.pardir, os.pardir))
-BPFTOOL_DIR = os.path.join(LINUX_ROOT, 'tools/bpf/bpftool')
+BPFTOOL_DIR = os.getenv('BPFTOOL_DIR',
+    os.path.join(LINUX_ROOT, 'tools/bpf/bpftool'))
+BPFTOOL_BASHCOMP_DIR = os.getenv('BPFTOOL_BASHCOMP_DIR',
+    os.path.join(BPFTOOL_DIR, 'bash-completion'))
+BPFTOOL_DOC_DIR = os.getenv('BPFTOOL_DOC_DIR',
+    os.path.join(BPFTOOL_DIR, 'Documentation'))
+INCLUDE_DIR = os.getenv('INCLUDE_DIR',
+    os.path.join(LINUX_ROOT, 'tools/include'))
+
 retval = 0
 
 class BlockParser(object):
···
         end_marker = re.compile('}\\\\n')
         return self.__get_description_list(start_marker, pattern, end_marker)
 
-    def default_options(self):
-        """
-        Return the default options contained in HELP_SPEC_OPTIONS
-        """
-        return { '-j', '--json', '-p', '--pretty', '-d', '--debug' }
-
     def get_bashcomp_list(self, block_name):
         """
         Search for and parse a list of type names from a variable in bash
···
     defined in children classes.
     """
     def get_options(self):
-        return self.default_options().union(self.get_help_list_macro('HELP_SPEC_OPTIONS'))
+        return self.get_help_list_macro('HELP_SPEC_OPTIONS')
+
+class MainHeaderFileExtractor(SourceFileExtractor):
+    """
+    An extractor for bpftool's main.h
+    """
+    filename = os.path.join(BPFTOOL_DIR, 'main.h')
+
+    def get_common_options(self):
+        """
+        Parse the list of common options in main.h (options that apply to all
+        commands), which looks to the lists of options in other source files
+        but has different start and end markers:
+
+            "OPTIONS := { {-j|--json} [{-p|--pretty}] | {-d|--debug} | {-l|--legacy}"
+
+        Return a set containing all options, such as:
+
+            {'-p', '-d', '--legacy', '--pretty', '--debug', '--json', '-l', '-j'}
+        """
+        start_marker = re.compile(f'"OPTIONS :=')
+        pattern = re.compile('([\w-]+) ?(?:\||}[ }\]"])')
+        end_marker = re.compile('#define')
+
+        parser = InlineListParser(self.reader)
+        parser.search_block(start_marker)
+        return parser.parse(pattern, end_marker)
+
+class ManSubstitutionsExtractor(SourceFileExtractor):
+    """
+    An extractor for substitutions.rst
+    """
+    filename = os.path.join(BPFTOOL_DOC_DIR, 'substitutions.rst')
+
+    def get_common_options(self):
+        """
+        Parse the list of common options in substitutions.rst (options that
+        apply to all commands).
+
+        Return a set containing all options, such as:
+
+            {'-p', '-d', '--legacy', '--pretty', '--debug', '--json', '-l', '-j'}
+        """
+        start_marker = re.compile('\|COMMON_OPTIONS\| replace:: {')
+        pattern = re.compile('\*\*([\w/-]+)\*\*')
+        end_marker = re.compile('}$')
+
+        parser = InlineListParser(self.reader)
+        parser.search_block(start_marker)
+        return parser.parse(pattern, end_marker)
 
 class ProgFileExtractor(SourceFileExtractor):
     """
···
     """
     An extractor for the UAPI BPF header.
     """
-    filename = os.path.join(LINUX_ROOT, 'tools/include/uapi/linux/bpf.h')
+    filename = os.path.join(INCLUDE_DIR, 'uapi/linux/bpf.h')
 
     def get_prog_types(self):
         return self.get_enum('bpf_prog_type')
···
     """
     An extractor for bpftool-prog.rst.
     """
-    filename = os.path.join(BPFTOOL_DIR, 'Documentation/bpftool-prog.rst')
+    filename = os.path.join(BPFTOOL_DOC_DIR, 'bpftool-prog.rst')
 
     def get_attach_types(self):
         return self.get_rst_list('ATTACH_TYPE')
···
     """
     An extractor for bpftool-map.rst.
     """
-    filename = os.path.join(BPFTOOL_DIR, 'Documentation/bpftool-map.rst')
+    filename = os.path.join(BPFTOOL_DOC_DIR, 'bpftool-map.rst')
 
     def get_map_types(self):
         return self.get_rst_list('TYPE')
···
     """
     An extractor for bpftool-cgroup.rst.
     """
-    filename = os.path.join(BPFTOOL_DIR, 'Documentation/bpftool-cgroup.rst')
+    filename = os.path.join(BPFTOOL_DOC_DIR, 'bpftool-cgroup.rst')
 
     def get_attach_types(self):
         return self.get_rst_list('ATTACH_TYPE')
···
     """
     An extractor for bpftool's bash completion file.
     """
-    filename = os.path.join(BPFTOOL_DIR, 'bash-completion/bpftool')
+    filename = os.path.join(BPFTOOL_BASHCOMP_DIR, 'bpftool')
 
     def get_prog_attach_types(self):
         return self.get_bashcomp_list('BPFTOOL_PROG_ATTACH_TYPES')
···
         help_cmd_options = source_info.get_options()
         source_info.close()
 
-        man_cmd_info = ManGenericExtractor(os.path.join('Documentation', 'bpftool-' + cmd + '.rst'))
+        man_cmd_info = ManGenericExtractor(os.path.join(BPFTOOL_DOC_DIR, 'bpftool-' + cmd + '.rst'))
         man_cmd_options = man_cmd_info.get_options()
         man_cmd_info.close()
 
···
     help_main_options = source_main_info.get_options()
     source_main_info.close()
 
-    man_main_info = ManGenericExtractor(os.path.join('Documentation', 'bpftool.rst'))
+    man_main_info = ManGenericExtractor(os.path.join(BPFTOOL_DOC_DIR, 'bpftool.rst'))
     man_main_options = man_main_info.get_options()
     man_main_info.close()
 
     verify(help_main_options, man_main_options,
             f'Comparing {source_main_info.filename} (do_help() OPTIONS) and {man_main_info.filename} (OPTIONS):')
+
+    # Compare common options (options that apply to all commands)
+
+    main_hdr_info = MainHeaderFileExtractor()
+    source_common_options = main_hdr_info.get_common_options()
+    main_hdr_info.close()
+
+    man_substitutions = ManSubstitutionsExtractor()
+    man_common_options = man_substitutions.get_common_options()
+    man_substitutions.close()
+
+    verify(source_common_options, man_common_options,
+            f'Comparing common options from {main_hdr_info.filename} (HELP_SPEC_OPTIONS) and {man_substitutions.filename}:')
 
     sys.exit(retval)
 

tools/testing/selftests/bpf/test_cgroup_storage.c

···
 		goto err;
 	}
 
-	map_fd = bpf_create_map(BPF_MAP_TYPE_CGROUP_STORAGE, sizeof(key),
-				sizeof(value), 0, 0);
+	map_fd = bpf_map_create(BPF_MAP_TYPE_CGROUP_STORAGE, NULL, sizeof(key),
+				sizeof(value), 0, NULL);
 	if (map_fd < 0) {
 		printf("Failed to create map: %s\n", strerror(errno));
 		goto out;
 	}
 
-	percpu_map_fd = bpf_create_map(BPF_MAP_TYPE_PERCPU_CGROUP_STORAGE,
-				       sizeof(key), sizeof(value), 0, 0);
+	percpu_map_fd = bpf_map_create(BPF_MAP_TYPE_PERCPU_CGROUP_STORAGE, NULL,
+				       sizeof(key), sizeof(value), 0, NULL);
 	if (percpu_map_fd < 0) {
 		printf("Failed to create map: %s\n", strerror(errno));
 		goto out;

tools/testing/selftests/bpf/test_lpm_map.c

···
 
 static void test_lpm_map(int keysize)
 {
+	LIBBPF_OPTS(bpf_map_create_opts, opts, .map_flags = BPF_F_NO_PREALLOC);
 	size_t i, j, n_matches, n_matches_after_delete, n_nodes, n_lookups;
 	struct tlpm_node *t, *list = NULL;
 	struct bpf_lpm_trie_key *key;
···
 	key = alloca(sizeof(*key) + keysize);
 	memset(key, 0, sizeof(*key) + keysize);
 
-	map = bpf_create_map(BPF_MAP_TYPE_LPM_TRIE,
+	map = bpf_map_create(BPF_MAP_TYPE_LPM_TRIE, NULL,
 			     sizeof(*key) + keysize,
 			     keysize + 1,
 			     4096,
-			     BPF_F_NO_PREALLOC);
+			     &opts);
 	assert(map >= 0);
 
 	for (i = 0; i < n_nodes; ++i) {
···
 
 static void test_lpm_ipaddr(void)
 {
+	LIBBPF_OPTS(bpf_map_create_opts, opts, .map_flags = BPF_F_NO_PREALLOC);
 	struct bpf_lpm_trie_key *key_ipv4;
 	struct bpf_lpm_trie_key *key_ipv6;
 	size_t key_size_ipv4;
···
 	key_ipv4 = alloca(key_size_ipv4);
 	key_ipv6 = alloca(key_size_ipv6);
 
-	map_fd_ipv4 = bpf_create_map(BPF_MAP_TYPE_LPM_TRIE,
+	map_fd_ipv4 = bpf_map_create(BPF_MAP_TYPE_LPM_TRIE, NULL,
 				     key_size_ipv4, sizeof(value),
-				     100, BPF_F_NO_PREALLOC);
+				     100, &opts);
 	assert(map_fd_ipv4 >= 0);
 
-	map_fd_ipv6 = bpf_create_map(BPF_MAP_TYPE_LPM_TRIE,
+	map_fd_ipv6 = bpf_map_create(BPF_MAP_TYPE_LPM_TRIE, NULL,
 				     key_size_ipv6, sizeof(value),
-				     100, BPF_F_NO_PREALLOC);
+				     100, &opts);
 	assert(map_fd_ipv6 >= 0);
 
 	/* Fill data some IPv4 and IPv6 address ranges */
···
 
 static void test_lpm_delete(void)
 {
+	LIBBPF_OPTS(bpf_map_create_opts, opts, .map_flags = BPF_F_NO_PREALLOC);
 	struct bpf_lpm_trie_key *key;
 	size_t key_size;
 	int map_fd;
···
 	key_size = sizeof(*key) + sizeof(__u32);
 	key = alloca(key_size);
 
-	map_fd = bpf_create_map(BPF_MAP_TYPE_LPM_TRIE,
+	map_fd = bpf_map_create(BPF_MAP_TYPE_LPM_TRIE, NULL,
 				key_size, sizeof(value),
-				100, BPF_F_NO_PREALLOC);
+				100, &opts);
 	assert(map_fd >= 0);
 
 	/* Add nodes:
···
 
 static void test_lpm_get_next_key(void)
 {
+	LIBBPF_OPTS(bpf_map_create_opts, opts, .map_flags = BPF_F_NO_PREALLOC);
 	struct bpf_lpm_trie_key *key_p, *next_key_p;
 	size_t key_size;
 	__u32 value = 0;
···
 	key_p = alloca(key_size);
 	next_key_p = alloca(key_size);
 
-	map_fd = bpf_create_map(BPF_MAP_TYPE_LPM_TRIE, key_size, sizeof(value),
-				100, BPF_F_NO_PREALLOC);
+	map_fd = bpf_map_create(BPF_MAP_TYPE_LPM_TRIE, NULL, key_size, sizeof(value), 100, &opts);
 	assert(map_fd >= 0);
 
 	/* empty tree. get_next_key should return ENOENT */
···
 
 static void test_lpm_multi_thread(void)
 {
+	LIBBPF_OPTS(bpf_map_create_opts, opts, .map_flags = BPF_F_NO_PREALLOC);
 	struct lpm_mt_test_info info[4];
 	size_t key_size, value_size;
 	pthread_t thread_id[4];
···
 	/* create a trie */
 	value_size = sizeof(__u32);
 	key_size = sizeof(struct bpf_lpm_trie_key) + value_size;
-	map_fd = bpf_create_map(BPF_MAP_TYPE_LPM_TRIE, key_size, value_size,
-				100, BPF_F_NO_PREALLOC);
+	map_fd = bpf_map_create(BPF_MAP_TYPE_LPM_TRIE, NULL, key_size, value_size, 100, &opts);
 
 	/* create 4 threads to test update, delete, lookup and get_next_key */
 	setup_lpm_mt_test_info(&info[0], map_fd);

tools/testing/selftests/bpf/test_lru_map.c

···
 
 static int create_map(int map_type, int map_flags, unsigned int size)
 {
+	LIBBPF_OPTS(bpf_map_create_opts, opts, .map_flags = map_flags);
 	int map_fd;
 
-	map_fd = bpf_create_map(map_type, sizeof(unsigned long long),
-				sizeof(unsigned long long), size, map_flags);
+	map_fd = bpf_map_create(map_type, NULL,  sizeof(unsigned long long),
+				sizeof(unsigned long long), size, &opts);
 
 	if (map_fd == -1)
-		perror("bpf_create_map");
+		perror("bpf_map_create");
 
 	return map_fd;
 }
···
 static int bpf_map_lookup_elem_with_ref_bit(int fd, unsigned long long key,
 					    void *value)
 {
-	struct bpf_create_map_attr map;
 	struct bpf_insn insns[] = {
 		BPF_LD_MAP_VALUE(BPF_REG_9, 0, 0),
 		BPF_LD_MAP_FD(BPF_REG_1, fd),
···
 	int mfd, pfd, ret, zero = 0;
 	__u32 retval = 0;
 
-	memset(&map, 0, sizeof(map));
-	map.map_type = BPF_MAP_TYPE_ARRAY;
-	map.key_size = sizeof(int);
-	map.value_size = sizeof(unsigned long long);
-	map.max_entries = 1;
-
-	mfd = bpf_create_map_xattr(&map);
+	mfd = bpf_map_create(BPF_MAP_TYPE_ARRAY, NULL, sizeof(int), sizeof(__u64), 1, NULL);
 	if (mfd < 0)
 		return -1;
 

tools/testing/selftests/bpf/test_maps.c

···
 
 static int skips;
 
-static int map_flags;
+static struct bpf_map_create_opts map_opts = { .sz = sizeof(map_opts) };
 
 static void test_hashmap(unsigned int task, void *data)
 {
 	long long key, next_key, first_key, value;
 	int fd;
 
-	fd = bpf_create_map(BPF_MAP_TYPE_HASH, sizeof(key), sizeof(value),
-			    2, map_flags);
+	fd = bpf_map_create(BPF_MAP_TYPE_HASH, NULL, sizeof(key), sizeof(value), 2, &map_opts);
 	if (fd < 0) {
 		printf("Failed to create hashmap '%s'!\n", strerror(errno));
 		exit(1);
···
 
 	for (i = 1; i <= 512; i <<= 1)
 		for (j = 1; j <= 1 << 18; j <<= 1) {
-			fd = bpf_create_map(BPF_MAP_TYPE_HASH, i, j,
-					    2, map_flags);
+			fd = bpf_map_create(BPF_MAP_TYPE_HASH, NULL, i, j, 2, &map_opts);
 			if (fd < 0) {
 				if (errno == ENOMEM)
 					return;
···
 	int expected_key_mask = 0;
 	int fd, i;
 
-	fd = bpf_create_map(BPF_MAP_TYPE_PERCPU_HASH, sizeof(key),
-			    sizeof(bpf_percpu(value, 0)), 2, map_flags);
+	fd = bpf_map_create(BPF_MAP_TYPE_PERCPU_HASH, NULL, sizeof(key),
+			    sizeof(bpf_percpu(value, 0)), 2, &map_opts);
 	if (fd < 0) {
 		printf("Failed to create hashmap '%s'!\n", strerror(errno));
 		exit(1);
···
 	int i, fd, ret;
 	long long key, value;
 
-	fd = bpf_create_map(BPF_MAP_TYPE_HASH, sizeof(key), sizeof(value),
-			    max_entries, map_flags);
+	fd = bpf_map_create(BPF_MAP_TYPE_HASH, NULL, sizeof(key), sizeof(value),
+			    max_entries, &map_opts);
 	CHECK(fd < 0,
 	      "failed to create hashmap",
-	      "err: %s, flags: 0x%x\n", strerror(errno), map_flags);
+	      "err: %s, flags: 0x%x\n", strerror(errno), map_opts.map_flags);
 
 	for (i = 0; i < max_entries; i++) {
 		key = i; value = key;
···
 	int i, first, second, old_flags;
 	long long key, next_first, next_second;
 
-	old_flags = map_flags;
-	map_flags |= BPF_F_ZERO_SEED;
+	old_flags = map_opts.map_flags;
+	map_opts.map_flags |= BPF_F_ZERO_SEED;
 
 	first = helper_fill_hashmap(3);
 	second = helper_fill_hashmap(3);
···
 		key = next_first;
 	}
 
-	map_flags = old_flags;
+	map_opts.map_flags = old_flags;
 	close(first);
 	close(second);
 }
···
 	int key, next_key, fd;
 	long long value;
 
-	fd = bpf_create_map(BPF_MAP_TYPE_ARRAY, sizeof(key), sizeof(value),
-			    2, 0);
+	fd = bpf_map_create(BPF_MAP_TYPE_ARRAY, NULL, sizeof(key), sizeof(value), 2, NULL);
 	if (fd < 0) {
 		printf("Failed to create arraymap '%s'!\n", strerror(errno));
 		exit(1);
···
 	BPF_DECLARE_PERCPU(long, values);
 	int key, next_key, fd, i;
 
-	fd = bpf_create_map(BPF_MAP_TYPE_PERCPU_ARRAY, sizeof(key),
-			    sizeof(bpf_percpu(values, 0)), 2, 0);
+	fd = bpf_map_create(BPF_MAP_TYPE_PERCPU_ARRAY, NULL, sizeof(key),
+			    sizeof(bpf_percpu(values, 0)), 2, NULL);
 	if (fd < 0) {
 		printf("Failed to create arraymap '%s'!\n", strerror(errno));
 		exit(1);
···
 	unsigned int nr_keys = 2000;
 	int key, fd, i;
 
-	fd = bpf_create_map(BPF_MAP_TYPE_PERCPU_ARRAY, sizeof(key),
-			    sizeof(bpf_percpu(values, 0)), nr_keys, 0);
+	fd = bpf_map_create(BPF_MAP_TYPE_PERCPU_ARRAY, NULL, sizeof(key),
+			    sizeof(bpf_percpu(values, 0)), nr_keys, NULL);
 	if (fd < 0) {
 		printf("Failed to create per-cpu arraymap '%s'!\n",
 		       strerror(errno));
···
 	int fd;
 	__u32 key, value;
 
-	fd = bpf_create_map(BPF_MAP_TYPE_DEVMAP, sizeof(key), sizeof(value),
-			    2, 0);
+	fd = bpf_map_create(BPF_MAP_TYPE_DEVMAP, NULL, sizeof(key), sizeof(value), 2, NULL);
 	if (fd < 0) {
 		printf("Failed to create devmap '%s'!\n", strerror(errno));
 		exit(1);
···
 	int fd;
 	__u32 key, value;
 
-	fd = bpf_create_map(BPF_MAP_TYPE_DEVMAP_HASH, sizeof(key), sizeof(value),
-			    2, 0);
+	fd = bpf_map_create(BPF_MAP_TYPE_DEVMAP_HASH, NULL, sizeof(key), sizeof(value), 2, NULL);
 	if (fd < 0) {
 		printf("Failed to create devmap_hash '%s'!\n", strerror(errno));
 		exit(1);
···
 		vals[i] = rand();
 
 	/* Invalid key size */
-	fd = bpf_create_map(BPF_MAP_TYPE_QUEUE, 4, sizeof(val), MAP_SIZE,
-			    map_flags);
+	fd = bpf_map_create(BPF_MAP_TYPE_QUEUE, NULL, 4, sizeof(val), MAP_SIZE, &map_opts);
 	assert(fd < 0 && errno == EINVAL);
 
-	fd = bpf_create_map(BPF_MAP_TYPE_QUEUE, 0, sizeof(val), MAP_SIZE,
-			    map_flags);
+	fd = bpf_map_create(BPF_MAP_TYPE_QUEUE, NULL, 0, sizeof(val), MAP_SIZE, &map_opts);
 	/* Queue map does not support BPF_F_NO_PREALLOC */
-	if (map_flags & BPF_F_NO_PREALLOC) {
+	if (map_opts.map_flags & BPF_F_NO_PREALLOC) {
 		assert(fd < 0 && errno == EINVAL);
 		return;
 	}
···
 		vals[i] = rand();
 
 	/* Invalid key size */
-	fd = bpf_create_map(BPF_MAP_TYPE_STACK, 4, sizeof(val), MAP_SIZE,
-			    map_flags);
+	fd = bpf_map_create(BPF_MAP_TYPE_STACK, NULL, 4, sizeof(val), MAP_SIZE, &map_opts);
 	assert(fd < 0 && errno == EINVAL);
 
-	fd = bpf_create_map(BPF_MAP_TYPE_STACK, 0, sizeof(val), MAP_SIZE,
-			    map_flags);
+	fd = bpf_map_create(BPF_MAP_TYPE_STACK, NULL, 0, sizeof(val), MAP_SIZE, &map_opts);
 	/* Stack map does not support BPF_F_NO_PREALLOC */
-	if (map_flags & BPF_F_NO_PREALLOC) {
+	if (map_opts.map_flags & BPF_F_NO_PREALLOC) {
 		assert(fd < 0 && errno == EINVAL);
 		return;
 	}
···
 	}
 
 	/* Test sockmap with connected sockets */
-	fd = bpf_create_map(BPF_MAP_TYPE_SOCKMAP,
+	fd = bpf_map_create(BPF_MAP_TYPE_SOCKMAP, NULL,
 			    sizeof(key), sizeof(value),
-			    6, 0);
+			    6, NULL);
 	if (fd < 0) {
 		if (!bpf_probe_map_type(BPF_MAP_TYPE_SOCKMAP, 0)) {
 			printf("%s SKIP (unsupported map type BPF_MAP_TYPE_SOCKMAP)\n",
···
 
 	obj = bpf_object__open(MAPINMAP_PROG);
 
-	fd = bpf_create_map(BPF_MAP_TYPE_HASH, sizeof(int), sizeof(int),
-			    2, 0);
+	fd = bpf_map_create(BPF_MAP_TYPE_HASH, NULL, sizeof(int), sizeof(int), 2, NULL);
 	if (fd < 0) {
 		printf("Failed to create hashmap '%s'!\n", strerror(errno));
 		exit(1);
···
 	} key;
 	int fd, i, value;
 
-	fd = bpf_create_map(BPF_MAP_TYPE_HASH, sizeof(key), sizeof(value),
-			    MAP_SIZE, map_flags);
+	fd = bpf_map_create(BPF_MAP_TYPE_HASH, NULL, sizeof(key), sizeof(value),
+			    MAP_SIZE, &map_opts);
 	if (fd < 0) {
 		printf("Failed to create large map '%s'!\n", strerror(errno));
 		exit(1);
···
 	int i, fd, key = 0, value = 0;
 	int data[2];
 
-	fd = bpf_create_map(BPF_MAP_TYPE_HASH, sizeof(key), sizeof(value),
-			    MAP_SIZE, map_flags);
+	fd = bpf_map_create(BPF_MAP_TYPE_HASH, NULL, sizeof(key), sizeof(value),
+			    MAP_SIZE, &map_opts);
 	if (fd < 0) {
 		printf("Failed to create map for parallel test '%s'!\n",
 		       strerror(errno));
···
 static void test_map_rdonly(void)
 {
 	int fd, key = 0, value = 0;
+	__u32 old_flags;
 
-	fd = bpf_create_map(BPF_MAP_TYPE_HASH, sizeof(key), sizeof(value),
-			    MAP_SIZE, map_flags | BPF_F_RDONLY);
+	old_flags = map_opts.map_flags;
+	map_opts.map_flags |= BPF_F_RDONLY;
+	fd = bpf_map_create(BPF_MAP_TYPE_HASH, NULL, sizeof(key), sizeof(value),
+			    MAP_SIZE, &map_opts);
+	map_opts.map_flags = old_flags;
 	if (fd < 0) {
 		printf("Failed to create map for read only test '%s'!\n",
 		       strerror(errno));
···
 static void test_map_wronly_hash(void)
 {
 	int fd, key = 0, value = 0;
+	__u32 old_flags;
 
-	fd = bpf_create_map(BPF_MAP_TYPE_HASH, sizeof(key), sizeof(value),
-			    MAP_SIZE, map_flags | BPF_F_WRONLY);
+	old_flags = map_opts.map_flags;
+	map_opts.map_flags |= BPF_F_WRONLY;
+	fd = bpf_map_create(BPF_MAP_TYPE_HASH, NULL, sizeof(key), sizeof(value),
+			    MAP_SIZE, &map_opts);
+	map_opts.map_flags = old_flags;
 	if (fd < 0) {
 		printf("Failed to create map for write only test '%s'!\n",
 		       strerror(errno));
···
 static void test_map_wronly_stack_or_queue(enum bpf_map_type map_type)
 {
 	int fd, value = 0;
+	__u32 old_flags;
+
 
 	assert(map_type == BPF_MAP_TYPE_QUEUE ||
 	       map_type == BPF_MAP_TYPE_STACK);
-	fd = bpf_create_map(map_type, 0, sizeof(value), MAP_SIZE,
-			    map_flags | BPF_F_WRONLY);
+	old_flags = map_opts.map_flags;
+	map_opts.map_flags |= BPF_F_WRONLY;
+	fd = bpf_map_create(map_type, NULL, 0, sizeof(value), MAP_SIZE, &map_opts);
+	map_opts.map_flags = old_flags;
 	/* Stack/Queue maps do not support BPF_F_NO_PREALLOC */
-	if (map_flags & BPF_F_NO_PREALLOC) {
+	if (map_opts.map_flags & BPF_F_NO_PREALLOC) {
 		assert(fd < 0 && errno == EINVAL);
 		return;
 	}
···
 	__u32 fds_idx = 0;
 	int fd;
 
-	map_fd = bpf_create_map(BPF_MAP_TYPE_REUSEPORT_SOCKARRAY,
-				sizeof(__u32), sizeof(__u64), array_size, 0);
+	map_fd = bpf_map_create(BPF_MAP_TYPE_REUSEPORT_SOCKARRAY, NULL,
+				sizeof(__u32), sizeof(__u64), array_size, NULL);
 	CHECK(map_fd < 0, "reuseport array create",
 	      "map_fd:%d, errno:%d\n", map_fd, errno);
 
···
 	close(map_fd);
 
 	/* Test 32 bit fd */
-	map_fd = bpf_create_map(BPF_MAP_TYPE_REUSEPORT_SOCKARRAY,
-				sizeof(__u32), sizeof(__u32), array_size, 0);
+	map_fd = bpf_map_create(BPF_MAP_TYPE_REUSEPORT_SOCKARRAY, NULL,
+				sizeof(__u32), sizeof(__u32), array_size, NULL);
 	CHECK(map_fd < 0, "reuseport array create",
 	      "map_fd:%d, errno:%d\n", map_fd, errno);
 	prepare_reuseport_grp(SOCK_STREAM, map_fd, sizeof(__u32), &fd64,
···
 
 	libbpf_set_strict_mode(LIBBPF_STRICT_ALL);
 
-	map_flags = 0;
+	map_opts.map_flags = 0;
 	run_all_tests();
 
-	map_flags = BPF_F_NO_PREALLOC;
+	map_opts.map_flags = BPF_F_NO_PREALLOC;
 	run_all_tests();
 
 #define DEFINE_TEST(name) test_##name();

tools/testing/selftests/bpf/test_progs.c

···
 #include <prog_tests/tests.h>
 #undef DEFINE_TEST
 };
-const int prog_test_cnt = ARRAY_SIZE(prog_test_defs);
+static const int prog_test_cnt = ARRAY_SIZE(prog_test_defs);
 
 const char *argp_program_version = "test_progs 0.1";
 const char *argp_program_bug_address = "<[e-mail]>";
-const char argp_program_doc[] = "BPF selftests test runner";
+static const char argp_program_doc[] = "BPF selftests test runner";
 
 enum ARG_KEYS {
 	ARG_TEST_NUM = 'n',
···
 {
 	struct dispatch_data *data = ctx;
 	int sock_fd;
-	FILE *log_fd = NULL;
+	FILE *log_fp = NULL;
 
 	sock_fd = data->sock_fd;
 
···
 
 			/* collect all logs */
 			if (msg_test_done.test_done.have_log) {
-				log_fd = open_memstream(&result->log_buf, &result->log_cnt);
-				if (!log_fd)
+				log_fp = open_memstream(&result->log_buf, &result->log_cnt);
+				if (!log_fp)
 					goto error;
 
 				while (true) {
···
 					if (msg_log.type != MSG_TEST_LOG)
 						goto error;
 
-					fprintf(log_fd, "%s", msg_log.test_log.log_buf);
+					fprintf(log_fp, "%s", msg_log.test_log.log_buf);
 					if (msg_log.test_log.is_last)
 						break;
 				}
-				fclose(log_fd);
-				log_fd = NULL;
+				fclose(log_fp);
+				log_fp = NULL;
 			}
 			/* output log */
 			{
···
 	if (env.debug)
 		fprintf(stderr, "[%d]: Protocol/IO error: %s.\n", data->worker_id, strerror(errno));
 
-	if (log_fd)
-		fclose(log_fd);
+	if (log_fp)
+		fclose(log_fp);
 done:
 	{
 		struct msg msg_exit;
···
 		env.sub_succ_cnt += result->sub_succ_cnt;
 	}
 
+	print_all_error_logs();
+
 	fprintf(stdout, "Summary: %d/%d PASSED, %d SKIPPED, %d FAILED\n",
 		env.succ_cnt, env.sub_succ_cnt, env.skip_cnt, env.fail_cnt);
-
-	print_all_error_logs();
 
 	/* reap all workers */
 	for (i = 0; i < env.workers; i++) {
···
 	if (env.list_test_names)
 		goto out;
 
+	print_all_error_logs();
+
 	fprintf(stdout, "Summary: %d/%d PASSED, %d SKIPPED, %d FAILED\n",
 		env.succ_cnt, env.sub_succ_cnt, env.skip_cnt, env.fail_cnt);
-
-	print_all_error_logs();
 
 	close(env.saved_netns_fd);
 out:

tools/testing/selftests/bpf/test_sock_addr.c

···
 
 static int load_path(const struct sock_addr_test *test, const char *path)
 {
-	struct bpf_prog_load_attr attr;
 	struct bpf_object *obj;
-	int prog_fd;
+	struct bpf_program *prog;
+	int err;
 
-	memset(&attr, 0, sizeof(struct bpf_prog_load_attr));
-	attr.file = path;
-	attr.prog_type = BPF_PROG_TYPE_CGROUP_SOCK_ADDR;
-	attr.expected_attach_type = test->expected_attach_type;
-	attr.prog_flags = BPF_F_TEST_RND_HI32;
-
-	if (bpf_prog_load_xattr(&attr, &obj, &prog_fd)) {
-		if (test->expected_result != LOAD_REJECT)
-			log_err(">>> Loading program (%s) error.\n", path);
+	obj = bpf_object__open_file(path, NULL);
+	err = libbpf_get_error(obj);
+	if (err) {
+		log_err(">>> Opening BPF object (%s) error.\n", path);
 		return -1;
 	}
 
-	return prog_fd;
+	prog = bpf_object__next_program(obj, NULL);
+	if (!prog)
+		goto err_out;
+
+	bpf_program__set_type(prog, BPF_PROG_TYPE_CGROUP_SOCK_ADDR);
+	bpf_program__set_expected_attach_type(prog, test->expected_attach_type);
+	bpf_program__set_flags(prog, BPF_F_TEST_RND_HI32);
+
+	err = bpf_object__load(obj);
+	if (err) {
+		if (test->expected_result != LOAD_REJECT)
+			log_err(">>> Loading program (%s) error.\n", path);
+		goto err_out;
+	}
+
+	return bpf_program__fd(prog);
+err_out:
+	bpf_object__close(obj);
+	return -1;
 }
 
 static int bind4_prog_load(const struct sock_addr_test *test)

tools/testing/selftests/bpf/test_tag.c

···
 
 int main(void)
 {
+	LIBBPF_OPTS(bpf_map_create_opts, opts, .map_flags = BPF_F_NO_PREALLOC);
 	uint32_t tests = 0;
 	int i, fd_map;
 
-	fd_map = bpf_create_map(BPF_MAP_TYPE_HASH, sizeof(int),
-				sizeof(int), 1, BPF_F_NO_PREALLOC);
+	fd_map = bpf_map_create(BPF_MAP_TYPE_HASH, NULL, sizeof(int),
+				sizeof(int), 1, &opts);
 	assert(fd_map > 0);
 
 	for (i = 0; i < 5; i++) {

tools/testing/selftests/bpf/test_verifier.c

···
 			uint32_t size_value, uint32_t max_elem,
 			uint32_t extra_flags)
 {
+	LIBBPF_OPTS(bpf_map_create_opts, opts);
 	int fd;
 
-	fd = bpf_create_map(type, size_key, size_value, max_elem,
-			    (type == BPF_MAP_TYPE_HASH ?
-			     BPF_F_NO_PREALLOC : 0) | extra_flags);
+	opts.map_flags = (type == BPF_MAP_TYPE_HASH ? BPF_F_NO_PREALLOC : 0) | extra_flags;
+	fd = bpf_map_create(type, NULL, size_key, size_value, max_elem, &opts);
 	if (fd < 0) {
 		if (skip_unsupported_map(type))
 			return -1;
···
 {
 	int mfd, p1fd, p2fd, p3fd;
 
-	mfd = bpf_create_map(BPF_MAP_TYPE_PROG_ARRAY, sizeof(int),
-			     sizeof(int), max_elem, 0);
+	mfd = bpf_map_create(BPF_MAP_TYPE_PROG_ARRAY, NULL, sizeof(int),
+			     sizeof(int), max_elem, NULL);
 	if (mfd < 0) {
 		if (skip_unsupported_map(BPF_MAP_TYPE_PROG_ARRAY))
 			return -1;
···
 
 static int create_map_in_map(void)
 {
+	LIBBPF_OPTS(bpf_map_create_opts, opts);
 	int inner_map_fd, outer_map_fd;
 
-	inner_map_fd = bpf_create_map(BPF_MAP_TYPE_ARRAY, sizeof(int),
-				      sizeof(int), 1, 0);
+	inner_map_fd = bpf_map_create(BPF_MAP_TYPE_ARRAY, NULL, sizeof(int),
+				      sizeof(int), 1, NULL);
 	if (inner_map_fd < 0) {
 		if (skip_unsupported_map(BPF_MAP_TYPE_ARRAY))
 			return -1;
···
 		return inner_map_fd;
 	}
 
-	outer_map_fd = bpf_create_map_in_map(BPF_MAP_TYPE_ARRAY_OF_MAPS, NULL,
-					     sizeof(int), inner_map_fd, 1, 0);
+	opts.inner_map_fd = inner_map_fd;
+	outer_map_fd = bpf_map_create(BPF_MAP_TYPE_ARRAY_OF_MAPS, NULL,
+				      sizeof(int), sizeof(int), 1, &opts);
 	if (outer_map_fd < 0) {
 		if (skip_unsupported_map(BPF_MAP_TYPE_ARRAY_OF_MAPS))
 			return -1;
···
 		BPF_MAP_TYPE_CGROUP_STORAGE;
 	int fd;
 
-	fd = bpf_create_map(type, sizeof(struct bpf_cgroup_storage_key),
-			    TEST_DATA_LEN, 0, 0);
+	fd = bpf_map_create(type, NULL, sizeof(struct bpf_cgroup_storage_key),
+			    TEST_DATA_LEN, 0, NULL);
 	if (fd < 0) {
 		if (skip_unsupported_map(type))
 			return -1;
···
 	memcpy(ptr, btf_str_sec, hdr.str_len);
 	ptr += hdr.str_len;
 
-	btf_fd = bpf_load_btf(raw_btf, ptr - raw_btf, 0, 0, 0);
+	btf_fd = bpf_btf_load(raw_btf, ptr - raw_btf, NULL);
 	free(raw_btf);
 	if (btf_fd < 0)
 		return -1;
···
 
 static int create_map_spin_lock(void)
 {
-	struct bpf_create_map_attr attr = {
-		.name = "test_map",
-		.map_type = BPF_MAP_TYPE_ARRAY,
-		.key_size = 4,
-		.value_size = 8,
-		.max_entries = 1,
+	LIBBPF_OPTS(bpf_map_create_opts, opts,
 		.btf_key_type_id = 1,
 		.btf_value_type_id = 3,
-	};
+	);
 	int fd, btf_fd;
 
 	btf_fd = load_btf();
 	if (btf_fd < 0)
 		return -1;
-	attr.btf_fd = btf_fd;
-	fd = bpf_create_map_xattr(&attr);
+	opts.btf_fd = btf_fd;
+	fd = bpf_map_create(BPF_MAP_TYPE_ARRAY, "test_map", 4, 8, 1, &opts);
 	if (fd < 0)
 		printf("Failed to create map with spin_lock\n");
 	return fd;
···
 
 static int create_sk_storage_map(void)
 {
-	struct bpf_create_map_attr attr = {
-		.name = "test_map",
-		.map_type = BPF_MAP_TYPE_SK_STORAGE,
-		.key_size = 4,
-		.value_size = 8,
-		.max_entries = 0,
+	LIBBPF_OPTS(bpf_map_create_opts, opts,
 		.map_flags = BPF_F_NO_PREALLOC,
 		.btf_key_type_id = 1,
 		.btf_value_type_id = 3,
-	};
+	);
 	int fd, btf_fd;
 
 	btf_fd = load_btf();
 	if (btf_fd < 0)
 		return -1;
-	attr.btf_fd = btf_fd;
-	fd = bpf_create_map_xattr(&attr);
-	close(attr.btf_fd);
+	opts.btf_fd = btf_fd;
+	fd = bpf_map_create(BPF_MAP_TYPE_SK_STORAGE, "test_map", 4, 8, 0, &opts);
+	close(opts.btf_fd);
 	if (fd < 0)
 		printf("Failed to create sk_storage_map\n");
 	return fd;

tools/testing/selftests/bpf/testing_helpers.c

···
 int bpf_prog_test_load(const char *file, enum bpf_prog_type type,
 		       struct bpf_object **pobj, int *prog_fd)
 {
-	struct bpf_object_load_attr attr = {};
+	LIBBPF_OPTS(bpf_object_open_opts, opts,
+		.kernel_log_level = extra_prog_load_log_flags,
+	);
 	struct bpf_object *obj;
 	struct bpf_program *prog;
+	__u32 flags;
 	int err;
 
-	obj = bpf_object__open(file);
+	obj = bpf_object__open_file(file, &opts);
 	if (!obj)
 		return -errno;
 
···
 	if (type != BPF_PROG_TYPE_UNSPEC)
 		bpf_program__set_type(prog, type);
 
-	bpf_program__set_extra_flags(prog, BPF_F_TEST_RND_HI32);
+	flags = bpf_program__flags(prog) | BPF_F_TEST_RND_HI32;
+	bpf_program__set_flags(prog, flags);
 
-	attr.obj = obj;
-	attr.log_level = extra_prog_load_log_flags;
-	err = bpf_object__load_xattr(&attr);
+	err = bpf_object__load(obj);
 	if (err)
 		goto err_out;
 

tools/testing/selftests/bpf/vmtest.sh

···
 set -u
 set -e
 
-# This script currently only works for x86_64, as
-# it is based on the VM image used by the BPF CI which is
-# x86_64.
-QEMU_BINARY="${QEMU_BINARY:="qemu-system-x86_64"}"
-X86_BZIMAGE="arch/x86/boot/bzImage"
+# This script currently only works for x86_64 and s390x, as
+# it is based on the VM image used by the BPF CI, which is
+# available only for these architectures.
+ARCH="$(uname -m)"
+case "${ARCH}" in
+s390x)
+	QEMU_BINARY=qemu-system-s390x
+	QEMU_CONSOLE="ttyS1"
+	QEMU_FLAGS=(-smp 2)
+	BZIMAGE="arch/s390/boot/compressed/vmlinux"
+	;;
+x86_64)
+	QEMU_BINARY=qemu-system-x86_64
+	QEMU_CONSOLE="ttyS0,115200"
+	QEMU_FLAGS=(-cpu host -smp 8)
+	BZIMAGE="arch/x86/boot/bzImage"
+	;;
+*)
+	echo "Unsupported architecture"
+	exit 1
+	;;
+esac
 DEFAULT_COMMAND="./test_progs"
 MOUNT_DIR="mnt"
 ROOTFS_IMAGE="root.img"
 OUTPUT_DIR="$HOME/.bpf_selftests"
-KCONFIG_URL="https://raw.githubusercontent.com/libbpf/libbpf/master/travis-ci/vmtest/configs/latest.config"
-KCONFIG_API_URL="https://api.github.com/repos/libbpf/libbpf/contents/travis-ci/vmtest/configs/latest.config"
+KCONFIG_URL="https://raw.githubusercontent.com/libbpf/libbpf/master/travis-ci/vmtest/configs/config-latest.${ARCH}"
+KCONFIG_API_URL="https://api.github.com/repos/libbpf/libbpf/contents/travis-ci/vmtest/configs/config-latest.${ARCH}"
 INDEX_URL="https://raw.githubusercontent.com/libbpf/libbpf/master/travis-ci/vmtest/configs/INDEX"
 NUM_COMPILE_JOBS="$(nproc)"
 LOG_FILE_BASE="$(date +"bpf_selftests.%Y-%m-%d_%H-%M-%S")"
···
 {
 	{
 	for file in "${!URLS[@]}"; do
-		if [[ $file =~ ^libbpf-vmtest-rootfs-(.*)\.tar\.zst$ ]]; then
+		if [[ $file =~ ^"${ARCH}"/libbpf-vmtest-rootfs-(.*)\.tar\.zst$ ]]; then
 			echo "${BASH_REMATCH[1]}"
 		fi
 	done
···
 		exit 1
 	fi
 
-	download "libbpf-vmtest-rootfs-$rootfsversion.tar.zst" |
+	download "${ARCH}/libbpf-vmtest-rootfs-$rootfsversion.tar.zst" |
 		zstd -d | sudo tar -C "$dir" -x
 }
 
···
 		-nodefaults \
 		-display none \
 		-serial mon:stdio \
-		-cpu host \
+		"${qemu_flags[@]}" \
 		-enable-kvm \
-		-smp 8 \
 		-m 4G \
 		-drive file="${rootfs_img}",format=raw,index=1,media=disk,if=virtio,cache=none \
 		-kernel "${kernel_bzimage}" \
-		-append "root=/dev/vda rw console=ttyS0,115200"
+		-append "root=/dev/vda rw console=${QEMU_CONSOLE}"
 }
 
 copy_logs()
···
 	local kernel_checkout=$(realpath "${script_dir}"/../../../../)
 	# By default the script searches for the kernel in the checkout directory but
 	# it also obeys environment variables O= and KBUILD_OUTPUT=
-	local kernel_bzimage="${kernel_checkout}/${X86_BZIMAGE}"
+	local kernel_bzimage="${kernel_checkout}/${BZIMAGE}"
 	local command="${DEFAULT_COMMAND}"
 	local update_image="no"
 	local exit_command="poweroff -f"
···
 		if is_rel_path "${O}"; then
 			O="$(realpath "${PWD}/${O}")"
 		fi
-		kernel_bzimage="${O}/${X86_BZIMAGE}"
+		kernel_bzimage="${O}/${BZIMAGE}"
 		make_command="${make_command} O=${O}"
 	elif [[ "${KBUILD_OUTPUT:=""}" != "" ]]; then
 		if is_rel_path "${KBUILD_OUTPUT}"; then
 			KBUILD_OUTPUT="$(realpath "${PWD}/${KBUILD_OUTPUT}")"
 		fi
-		kernel_bzimage="${KBUILD_OUTPUT}/${X86_BZIMAGE}"
+		kernel_bzimage="${KBUILD_OUTPUT}/${BZIMAGE}"
 		make_command="${make_command} KBUILD_OUTPUT=${KBUILD_OUTPUT}"
 	fi
 

tools/testing/selftests/bpf/xdp_redirect_multi.c

···
 {
 	int prog_fd, group_all, mac_map;
 	struct bpf_program *ingress_prog, *egress_prog;
-	struct bpf_prog_load_attr prog_load_attr = {
-		.prog_type = BPF_PROG_TYPE_UNSPEC,
-	};
-	int i, ret, opt, egress_prog_fd = 0;
+	int i, err, ret, opt, egress_prog_fd = 0;
 	struct bpf_devmap_val devmap_val;
 	bool attach_egress_prog = false;
 	unsigned char mac_addr[6];
···
 	printf("\n");
 
 	snprintf(filename, sizeof(filename), "%s_kern.o", argv[0]);
-	prog_load_attr.file = filename;
-
-	if (bpf_prog_load_xattr(&prog_load_attr, &obj, &prog_fd))
+	obj = bpf_object__open_file(filename, NULL);
+	err = libbpf_get_error(obj);
+	if (err)
 		goto err_out;
+	err = bpf_object__load(obj);
+	if (err)
+		goto err_out;
+	prog_fd = bpf_program__fd(bpf_object__next_program(obj, NULL));
 
 	if (attach_egress_prog)
 		group_all = bpf_object__find_map_fd_by_name(obj, "map_egress");

tools/testing/selftests/bpf/xdpxceiver.c

···
 #include "xdpxceiver.h"
 #include "../kselftest.h"
 
+/* AF_XDP APIs were moved into libxdp and marked as deprecated in libbpf.
+ * Until xdpxceiver is either moved or re-writed into libxdp, suppress
+ * deprecation warnings in this file
+ */
+#pragma GCC diagnostic ignored "-Wdeprecated-declarations"
+
 static const char *MAC1 = "\x00\x0A\x56\x9E\xEE\x62";
 static const char *MAC2 = "\x00\x0A\x56\x9E\xEE\x61";
 static const char *IP1 = "192.168.100.162";
···
 	void *bufs;
 
 	bufs = mmap(NULL, mmap_sz, PROT_READ | PROT_WRITE,
-		    MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE | MAP_HUGETLB, -1, 0);
+		    MAP_PRIVATE | MAP_ANONYMOUS | MAP_HUGETLB, -1, 0);
 	if (bufs == MAP_FAILED)
 		return false;
 
···
 		testapp_invalid_desc(test);
 		break;
 	case TEST_TYPE_UNALIGNED_INV_DESC:
+		if (!hugepages_present(test->ifobj_tx)) {
+			ksft_test_result_skip("No 2M huge pages present.\n");
+			return;
+		}
 		test_spec_set_name(test, "UNALIGNED_INV_DESC");
 		test->ifobj_tx->umem->unaligned_mode = true;
 		test->ifobj_rx->umem->unaligned_mode = true;