Some new verifier tests that hit some important gaps in the parameter
space for atomic ops.

There are already exhaustive tests for the JIT part in
lib/test_bpf.c, but these exercise the verifier too.
Signed-off-by: Brendan Jackman <jackmanb@google.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20211015093318.1273686-1-jackmanb@google.com

Ilya Leoshkevich says:

====================

This series along with [1] and [2] fixes all the failures in the
btf_dump testsuite currently present on s390, in particular:

* [1] fixes intermittent build bug causing "failed to encode tag ..."
  * error messages.
* [2] fixes missing VAR entries on s390.
* Patch 1 disables Intel-specific code in a testcase.
* Patch 2 fixes an endianness-related bug.
* Patch 3 fixes an alignment-related bug.
* Patch 4 improves overly pessimistic alignment handling.

[1] https://lore.kernel.org/bpf/20211012022521.399302-1-iii@linux.ibm.com/
[2] https://lore.kernel.org/bpf/20211012022637.399365-1-iii@linux.ibm.com/

v1: https://lore.kernel.org/bpf/20211012023218.399568-1-iii@linux.ibm.com/
v1 -> v2:
- Remove redundant local variables, use t->size directly instead.

Best regards,
Ilya
====================
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>

Non-aligned integers are dumped as bitfields, which is supported for at
most 64-bit integers. Fix by using the same trick as
btf_dump_float_data(): copy non-aligned values to the local buffer.
Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211013160902.428340-4-iii@linux.ibm.com

On big-endian arches not only bytes, but also bits are numbered in
reverse order (see e.g. S/390 ELF ABI Supplement, but this is also true
for other big-endian arches as well).
Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211013160902.428340-3-iii@linux.ibm.com

cpu_number exists only on Intel and aarch64, so skip the test involing
it on other arches. An alternative would be to replace it with an
exported non-ifdefed primitive-typed percpu variable from the common
code, but there appears to be none.
Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211013160902.428340-2-iii@linux.ibm.com

The coccinelle check report:
"./samples/bpf/xdp_redirect_cpu_user.c:397:32-38:
ERROR: application of sizeof to pointer"
Using the "strlen" to fix it.
Reported-by: Zeal Robot <zealci@zte.com.cn>
Signed-off-by: David Yang <davidcomponentone@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211012111649.983253-1-davidcomponentone@gmail.com

The header is no longer needed since the event_pipe implementation
was updated to rely on libbpf's perf_buffer. This makes bpftool free of
dependencies to perf files, and we can update the Makefile accordingly.

Fixes: 9b190f18 ("tools/bpftool: switch map event_pipe to libbpf's perf_buffer")
Signed-off-by: Quentin Monnet <quentin@isovalent.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211020094826.16046-1-quentin@isovalent.com

Fix following checkincludes.pl warning:
./scripts/checkincludes.pl tools/testing/selftests/bpf/cgroup_helpers.c
tools/testing/selftests/bpf/cgroup_helpers.c: unistd.h is included more
than once.
Signed-off-by: Wan Jiabing <wanjiabing@vivo.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Link: https://lore.kernel.org/bpf/20211012023231.19911-1-wanjiabing@vivo.com

kernel/bpf/preload/Makefile was recently updated to have it install
libbpf's headers locally instead of pulling them from tools/lib/bpf. But
two items still need to be addressed.

First, the local .gitignore file was not adjusted to ignore the files
generated in the new kernel/bpf/preload/libbpf output directory.

Second, the "clean-files" target is now incorrect. The old artefacts
names were not removed from the target, while the new ones were added
incorrectly. This is because "clean-files" expects names relative to
$(obj), but we passed the absolute path instead. This results in the
output and header-destination directories for libbpf (and their
contents) not being removed from kernel/bpf/preload on "make clean" from
the root of the repository.

This commit fixes both issues. Note that $(userprogs) needs not be added
to "clean-files", because the cleaning infrastructure already accounts
for it.

Cleaning the files properly also prevents make from printing the
following message, for builds coming after a "make clean":
"make[4]: Nothing to be done for 'install_headers'."

v2: Simplify the "clean-files" target.

Fixes: bf607917 ("bpf: preload: Install libbpf headers when building")
Signed-off-by: Quentin Monnet <quentin@isovalent.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Link: https://lore.kernel.org/bpf/20211020094647.15564-1-quentin@isovalent.com

In preparation for bpf_program__get_prog_info_linear deprecation, move
the single use in libbpf to call bpf_obj_get_info_by_fd directly.
Signed-off-by: Dave Marchevsky <davemarchevsky@fb.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211011082031.4148337-2-davemarchevsky@fb.com

The helper function returns a pointer that in the failure case encodes
an error in the struct btf pointer. The current code lead to Coverity
warning about the use of the invalid pointer:

 *** CID 1507963:  Memory - illegal accesses  (USE_AFTER_FREE)
 /kernel/bpf/verifier.c: 1788 in find_kfunc_desc_btf()
 1782                          return ERR_PTR(-EINVAL);
 1783                  }
 1784
 1785                  kfunc_btf = __find_kfunc_desc_btf(env, offset, btf_modp);
 1786                  if (IS_ERR_OR_NULL(kfunc_btf)) {
 1787                          verbose(env, "cannot find module BTF for func_id %u\n", func_id);
 >>>      CID 1507963:  Memory - illegal accesses  (USE_AFTER_FREE)
 >>>      Using freed pointer "kfunc_btf".
 1788                          return kfunc_btf ?: ERR_PTR(-ENOENT);
 1789                  }
 1790                  return kfunc_btf;
 1791          }
 1792          return btf_vmlinux ?: ERR_PTR(-ENOENT);
 1793     }

Daniel suggested the use of ERR_CAST so that the intended use is clear
to Coverity, but on closer look it seems that we never return NULL from
the helper. Andrii noted that since __find_kfunc_desc_btf already logs
errors for all cases except btf_get_by_fd, it is much easier to add
logging for that and remove the IS_ERR check altogether, returning
directly from it.
Suggested-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211009040900.803436-1-memxor@gmail.com

Quentin Monnet says:

====================

This set contains one fix for bpftool's Makefile, to make sure that the
headers internal to libbpf are installed properly even if we add more
headers to the relevant Makefile variable in the future (although we'd like
to avoid that if possible).

The other patches aim at cleaning up the output from the Makefile, in
particular when running the command "make" another time after bpftool is
built.
====================
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>

One of bpftool's object files depends on zlib. To make sure we do not
attempt to build that object when the library is not available, commit
d66fa3c7 ("tools: bpftool: add feature check for zlib") introduced a
feature check to detect whether zlib is present.

This check comes as a rule for which the target ("zdep") is a
nonexistent file (phony target), which means that the Makefile always
attempts to rebuild it. It is mostly harmless. However, one side effect
is that, on running again once bpftool is already built, make considers
that "something" (the recipe for zdep) was executed, and does not print
the usual message "make: Nothing to be done for 'all'", which is a
user-friendly indicator that the build went fine.

Before, with some level of debugging information:

    $ make --debug=m
    [...]
    Reading makefiles...

    Auto-detecting system features:
    ...                        libbfd: [ on  ]
    ...        disassembler-four-args: [ on  ]
    ...                          zlib: [ on  ]
    ...                        libcap: [ on  ]
    ...               clang-bpf-co-re: [ on  ]

    Updating makefiles....
    Updating goal targets....
     File 'all' does not exist.
           File 'zdep' does not exist.
          Must remake target 'zdep'.
     File 'all' does not exist.
    Must remake target 'all'.
    Successfully remade target file 'all'.

After the patch:

    $ make --debug=m
    [...]

    Auto-detecting system features:
    ...                        libbfd: [ on  ]
    ...        disassembler-four-args: [ on  ]
    ...                          zlib: [ on  ]
    ...                        libcap: [ on  ]
    ...               clang-bpf-co-re: [ on  ]

    Updating makefiles....
    Updating goal targets....
     File 'all' does not exist.
    Must remake target 'all'.
    Successfully remade target file 'all'.
    make: Nothing to be done for 'all'.

(Note the last line, which is not part of make's debug information.)
Signed-off-by: Quentin Monnet <quentin@isovalent.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Link: https://lore.kernel.org/bpf/20211009210341.6291-4-quentin@isovalent.com

In bpftool's Makefile, libbpf has a FORCE dependency, to make sure we
rebuild it in case its source files changed. Let's instead make the
rebuild depend on the source files directly, through a call to the
"$(wildcard ...)" function. This avoids descending into libbpf's
directory if there is nothing to update.

Do the same for the bootstrap libbpf version.

This results in a slightly faster operation and less verbose output when
running make a second time in bpftool's directory.

Before:

    Auto-detecting system features:
    ...                        libbfd: [ on  ]
    ...        disassembler-four-args: [ on  ]
    ...                          zlib: [ on  ]
    ...                        libcap: [ on  ]
    ...               clang-bpf-co-re: [ on  ]

    make[1]: Entering directory '/root/dev/linux/tools/lib/bpf'
    make[1]: Entering directory '/root/dev/linux/tools/lib/bpf'
    make[1]: Nothing to be done for 'install_headers'.
    make[1]: Leaving directory '/root/dev/linux/tools/lib/bpf'
    make[1]: Leaving directory '/root/dev/linux/tools/lib/bpf'

After:

    Auto-detecting system features:
    ...                        libbfd: [ on  ]
    ...        disassembler-four-args: [ on  ]
    ...                          zlib: [ on  ]
    ...                        libcap: [ on  ]
    ...               clang-bpf-co-re: [ on  ]

Other ways to clean up the output could be to pass the "-s" option, or
to redirect the output to >/dev/null, when calling make recursively to
descend into libbpf's directory. However, this would suppress some
useful output if something goes wrong during the build. A better
alternative would be to pass "--no-print-directory" to the recursive
make, but that would still leave us with some noise for
"install_headers". Skipping the descent into libbpf's directory if no
source file has changed works best, and seems the most logical option
overall.
Reported-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Quentin Monnet <quentin@isovalent.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Link: https://lore.kernel.org/bpf/20211009210341.6291-3-quentin@isovalent.com

We recently updated bpftool's Makefile to make it install the headers
from libbpf, instead of pulling them directly from libbpf's directory.
There is also an additional header, internal to libbpf, that needs be
installed. The way that bpftool's Makefile installs that particular
header is currently correct, but would break if we were to modify
$(LIBBPF_INTERNAL_HDRS) to make it point to more than one header.

Use a static pattern rule instead, so that the Makefile can withstand
the addition of other headers to install.

The objective is simply to make the Makefile more robust. It should
_not_ be read as an invitation to import more internal headers from
libbpf into bpftool.

Fixes: f012ade1 ("bpftool: Install libbpf headers instead of including the dir")
Reported-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Quentin Monnet <quentin@isovalent.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Link: https://lore.kernel.org/bpf/20211009210341.6291-2-quentin@isovalent.com

Although relying on some definitions from the netlink.h and if_link.h
headers copied into tools/include/uapi/linux/, libbpf does not need
those headers to stay entirely up-to-date with their original versions,
and the warnings emitted by the Makefile when it detects a difference
are usually just noise. Let's remove those warnings.
Suggested-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Quentin Monnet <quentin@isovalent.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211010002528.9772-1-quentin@isovalent.com

Patch set [1] introduced BTF_KIND_TAG to allow tagging
declarations for struct/union, struct/union field, var, func
and func arguments and these tags will be encoded into
dwarf. They are also encoded to btf by llvm for the bpf target.

After BTF_KIND_TAG is introduced, we intended to use it
for kernel __user attributes. But kernel __user is actually
a type attribute. Upstream and internal discussion showed
it is not a good idea to mix declaration attribute and
type attribute. So we proposed to introduce btf_type_tag
as a type attribute and existing btf_tag renamed to
btf_decl_tag ([2]).

This patch renamed BTF_KIND_TAG to BTF_KIND_DECL_TAG and some
other declarations with *_tag to *_decl_tag to make it clear
the tag is for declaration. In the future, BTF_KIND_TYPE_TAG
might be introduced per [3].

 [1] https://lore.kernel.org/bpf/20210914223004.244411-1-yhs@fb.com/
 [2] https://reviews.llvm.org/D111588
 [3] https://reviews.llvm.org/D111199

Fixes: b5ea834d ("bpf: Support for new btf kind BTF_KIND_TAG")
Fixes: 5b84bd10 ("libbpf: Add support for BTF_KIND_TAG")
Fixes: 5c07f2fe ("bpftool: Add support for BTF_KIND_TAG")
Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20211012164838.3345699-1-yhs@fb.com

In emit_tail_call() of bpf_jit_comp32.c, "blez t2" (t2 <= 0) is
not consistent with the comment "t2 < 0", update the comment to
keep consistency.
Signed-off-by: Tiezhu Yang <yangtiezhu@loongson.cn>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Johan Almbladh <johan.almbladh@anyfinetworks.com>
Link: https://lore.kernel.org/bpf/1633915150-13220-3-git-send-email-yangtiezhu@loongson.cn

The config options MIPS_CBPF_JIT and MIPS_EBPF_JIT are useless, remove
them in arch/mips/Kconfig, and then modify arch/mips/net/Makefile.
Signed-off-by: Tiezhu Yang <yangtiezhu@loongson.cn>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Johan Almbladh <johan.almbladh@anyfinetworks.com>
Link: https://lore.kernel.org/bpf/1633915150-13220-2-git-send-email-yangtiezhu@loongson.cn

The verifier tests added in commit c48e51c8 ("bpf: selftests: Add
selftests for module kfunc support") fail on s390, since the JIT does
not support calling kernel functions. This is most likely an issue for
all the other non-Intel arches, as well as on Intel with
!CONFIG_DEBUG_INFO_BTF or !CONFIG_BPF_JIT.

Trying to check for messages from all the possible add_kfunc_call()
failure cases in test_verifier looks pointless, so do a much simpler
thing instead: just like it's already done in do_prog_test_run(), skip
the tests that fail to load with ENOTSUPP.
Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Song Liu <songliubraving@fb.com>
Link: https://lore.kernel.org/bpf/20211007173329.381754-1-iii@linux.ibm.com

Yucong Sun says:

====================

This patch series adds "-j" parelell execution to test_progs, with "--debug" to
display server/worker communications. Also, some Tests that often fails in
parallel are marked as serial test, and it will run in sequence after parallel
execution is done.

This patch series also adds a error summary after all tests execution finished.

V6 -> V5:
  * adding error summary logic for non parallel mode too.
  * changed how serial tests are implemented, use main process instead of worker 0.
  * fixed a dozen broken test when running in parallel.

V5 -> V4:
  * change to SOCK_SEQPACKET for close notification.
  * move all debug output to "--debug" mode
  * output log as test finish, and all error logs again after summary line.
  * variable naming / style changes
  * adds serial_test_name() to replace serial test lists.
====================
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>

Change tests that often fails in parallel execution mode to serial.
Signed-off-by: Yucong Sun <sunyucong@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211006185619.364369-15-fallentree@fb.com

bpf_get_current_pid_tgid() returns u64, whose upper 32 bits are the same
as userspace getpid() return value.
Signed-off-by: Yucong Sun <sunyucong@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211006185619.364369-13-fallentree@fb.com

This make atomics test able to run in parallel with other tests.
Signed-off-by: Yucong Sun <sunyucong@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211006185619.364369-11-fallentree@fb.com

This patch change cgroup_v1v2 use a different port, avoid conflict with
other tests.
Signed-off-by: Yucong Sun <sunyucong@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211006185619.364369-8-fallentree@fb.com

In parallel execution mode, this test now need to use atomic operation
to avoid race condition.
Signed-off-by: Yucong Sun <sunyucong@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211006185619.364369-7-fallentree@fb.com

This patch make each worker use a unique cgroup base directory, thus
allowing tests that uses cgroups to run concurrently.
Signed-off-by: Yucong Sun <sunyucong@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211006185619.364369-5-fallentree@fb.com

This patch allows tests to define serial_test_name() instead of
test_name(), and this will make test_progs execute those in sequence
after all other tests finished executing concurrently.
Signed-off-by: Yucong Sun <sunyucong@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211006185619.364369-3-fallentree@fb.com

This patch adds "-j" mode to test_progs, executing tests in multiple
process.  "-j" mode is optional, and works with all existing test
selection mechanism, as well as "-v", "-l" etc.

In "-j" mode, main process use UDS/SEQPACKET to communicate to each forked
worker, commanding it to run tests and collect logs. After all tests are
finished, a summary is printed. main process use multiple competing
threads to dispatch work to worker, trying to keep them all busy.

The test status will be printed as soon as it is finished, if there are
error logs, it will be printed after the final summary line.

By specifying "--debug", additional debug information on server/worker
communication will be printed.

Example output:
  > ./test_progs -n 15-20 -j
  [   12.801730] bpf_testmod: loading out-of-tree module taints kernel.
  Launching 8 workers.
  #20 btf_split:OK
  #16 btf_endian:OK
  #18 btf_module:OK
  #17 btf_map_in_map:OK
  #19 btf_skc_cls_ingress:OK
  #15 btf_dump:OK
  Summary: 6/20 PASSED, 0 SKIPPED, 0 FAILED
Signed-off-by: Yucong Sun <sunyucong@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211006185619.364369-2-fallentree@fb.com

Hou Tao says:

====================

From: Hou Tao <houtao1@huawei.com>

Hi,

The patchset series supports writable context for bare tracepoint.

The main idea comes from patchset "writable contexts for bpf raw
tracepoints" [1], but it only supports normal tracepoint with
associated trace event under tracefs. Now we have one use case
in which we add bare tracepoint in VFS layer, and update
file::f_mode for specific files. The reason using bare tracepoint
is that it doesn't form a ABI and we can change it freely. So
add support for it in BPF.

Comments are always welcome.

[1]: https://lore.kernel.org/lkml/20190426184951.21812-1-mmullins@fb.com

Change log:
v5:
 * rebased on bpf-next
 * patch 1: add Acked-by tag
 * patch 2: handle invalid section name, make prefixes array being const

v4: https://www.spinics.net/lists/bpf/msg47021.html
 * rebased on bpf-next
 * update patch 2 to add support for writable raw tracepoint attachment
   in attach_raw_tp().
 * update patch 3 to add Acked-by tag

v3: https://www.spinics.net/lists/bpf/msg46824.html
  * use raw_tp.w instead of raw_tp_writable as section
    name of writable tp
  * use ASSERT_XXX() instead of CHECK()
  * define a common macro for "/sys/kernel/bpf_testmod"

v2: https://www.spinics.net/lists/bpf/msg46356.html
  * rebase on bpf-next tree
  * address comments from Yonghong Song
  * rename bpf_testmode_test_writable_ctx::ret as early_ret to reflect
    its purpose better.

v1: https://www.spinics.net/lists/bpf/msg46221.html
====================
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>

Add a writable bare tracepoint in bpf_testmod module, and
trigger its calling when reading /sys/kernel/bpf_testmod
with a specific buffer length. The reading will return
the value in writable context if the early return flag
is enabled in writable context.
Signed-off-by: Hou Tao <houtao1@huawei.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211004094857.30868-4-hotforest@gmail.com

Program on writable tracepoint is BPF_PROG_TYPE_RAW_TRACEPOINT_WRITABLE,
but its attachment is the same as BPF_PROG_TYPE_RAW_TRACEPOINT.
Signed-off-by: Hou Tao <houtao1@huawei.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211004094857.30868-3-hotforest@gmail.com

Commit 9df1c28b ("bpf: add writable context for raw tracepoints")
supports writable context for tracepoint, but it misses the support
for bare tracepoint which has no associated trace event.

Bare tracepoint is defined by DECLARE_TRACE(), so adding a corresponding
DECLARE_TRACE_WRITABLE() macro to generate a definition in __bpf_raw_tp_map
section for bare tracepoint in a similar way to DEFINE_TRACE_WRITABLE().
Signed-off-by: Hou Tao <houtao1@huawei.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211004094857.30868-2-hotforest@gmail.com

Quentin Monnet says:

====================

Libbpf is used at several locations in the repository. Most of the time,
the tools relying on it build the library in its own directory, and include
the headers from there. This works, but this is not the cleanest approach.
It generates objects outside of the directory of the tool which is being
built, and it also increases the risk that developers include a header file
internal to libbpf, which is not supposed to be exposed to user
applications.

This set adjusts all involved Makefiles to make sure that libbpf is built
locally (with respect to the tool's directory or provided build directory),
and by ensuring that "make install_headers" is run from libbpf's Makefile
to export user headers properly.

This comes at a cost: given that the libbpf was so far mostly compiled in
its own directory by the different components using it, compiling it once
would be enough for all those components. With the new approach, each
component compiles its own version. To mitigate this cost, efforts were
made to reuse the compiled library when possible:

- Make the bpftool version in samples/bpf reuse the library previously
  compiled for the selftests.
- Make the bpftool version in BPF selftests reuse the library previously
  compiled for the selftests.
- Similarly, make resolve_btfids in BPF selftests reuse the same compiled
  library.
- Similarly, make runqslower in BPF selftests reuse the same compiled
  library; and make it rely on the bpftool version also compiled from the
  selftests (instead of compiling its own version).
- runqslower, when compiled independently, needs its own version of
  bpftool: make them share the same compiled libbpf.

As a result:

- Compiling the samples/bpf should compile libbpf just once.
- Compiling the BPF selftests should compile libbpf just once.
- Compiling the kernel (with BTF support) should now lead to compiling
  libbpf twice: one for resolve_btfids, one for kernel/bpf/preload.
- Compiling runqslower individually should compile libbpf just once. Same
  thing for bpftool, resolve_btfids, and kernel/bpf/preload/iterators.

(Not accounting for the boostrap version of libbpf required by bpftool,
which was already placed under a dedicated .../boostrap/libbpf/ directory,
and for which the count remains unchanged.)

A few commits in the series also contain drive-by clean-up changes for
bpftool includes, samples/bpf/.gitignore, or test_bpftool_build.sh. Please
refer to individual commit logs for details.

v4:
  - Make the "libbpf_hdrs" dependency an order-only dependency in
    kernel/bpf/preload/Makefile, samples/bpf/Makefile, and
    tools/bpf/runqslower/Makefile. This is to avoid to unconditionally
    recompile the targets.
  - samples/bpf/.gitignore: prefix objects with a "/" to mark that we
    ignore them when at the root of the samples/bpf/ directory.
  - libbpf: add a commit to make "install_headers" depend on the header
    files, to avoid exporting again if the sources are older than the
    targets. This ensures that applications relying on those headers are
    not rebuilt unnecessarily.
  - bpftool: uncouple the copy of nlattr.h from libbpf target, to have it
    depend on the source header itself. By avoiding to reinstall this
    header every time, we avoid unnecessary builds of bpftool.
  - samples/bpf: Add a new commit to remove the FORCE dependency for
    libbpf, and replace it with a "$(wildcard ...)" on the .c/.h files in
    libbpf's directory. This is to avoid always recompiling libbpf/bpftool.
  - Adjust prefixes in commit subjects.

v3:
  - Remove order-only dependencies on $(LIBBPF_INCLUDE) (or equivalent)
    directories, given that they are created by libbpf's Makefile.
  - Add libbpf as a dependency for bpftool/resolve_btfids/runqslower when
    they are supposed to reuse a libbpf compiled previously. This is to
    avoid having several libbpf versions being compiled simultaneously in
    the same directory with parallel builds. Even if this didn't show up
    during tests, let's remain on the safe side.
  - kernel/bpf/preload/Makefile: Rename libbpf-hdrs (dash) dependency as
    libbpf_hdrs.
  - samples/bpf/.gitignore: Add bpftool/
  - samples/bpf/Makefile: Change "/bin/rm -rf" to "$(RM) -r".
  - samples/bpf/Makefile: Add missing slashes for $(LIBBPF_OUTPUT) and
    $(LIBBPF_DESTDIR) when buildling bpftool
  - samples/bpf/Makefile: Add a dependency to libbpf's headers for
    $(TRACE_HELPERS).
  - bpftool's Makefile: Use $(LIBBPF) instead of equivalent (but longer)
    $(LIBBPF_OUTPUT)libbpf.a
  - BPF iterators' Makefile: build bpftool in .output/bpftool (instead of
    .output/), add and clean up variables.
  - runqslower's Makefile: Add an explicit dependency on libbpf's headers
    to several objects. The dependency is not required (libbpf should have
    been compiled and so the headers exported through other dependencies
    for those targets), but they better mark the logical dependency and
    should help if exporting the headers changed in the future.
  - New commit to add an "install-bin" target to bpftool, to avoid
    installing bash completion when buildling BPF iterators and selftests.

v2: Declare an additional dependency on libbpf's headers for
    iterators/iterators.o in kernel/preload/Makefile to make sure that
    these headers are exported before we compile the object file (and not
    just before we link it).
====================
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>

With "make install", bpftool installs its binary and its bash completion
file. Usually, this is what we want. But a few components in the kernel
repository (namely, BPF iterators and selftests) also install bpftool
locally before using it. In such a case, bash completion is not
necessary and is just a useless build artifact.

Let's add an "install-bin" target to bpftool, to offer a way to install
the binary only.
Signed-off-by: Quentin Monnet <quentin@isovalent.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211007194438.34443-13-quentin@isovalent.com

The script test_bpftool_build.sh attempts to build bpftool in the
various supported ways, to make sure nothing breaks.

One of those ways is to run "make tools/bpf" from the root of the kernel
repository. This command builds bpftool, along with the other tools
under tools/bpf, and runqslower in particular. After running the
command and upon a successful bpftool build, the script attempts to
cleanup the generated objects. However, after building with this target
and in the case of runqslower, the files are not cleaned up as expected.

This is because the "tools/bpf" target sets $(OUTPUT) to
.../tools/bpf/runqslower/ when building the tool, causing the object
files to be placed directly under the runqslower directory. But when
running "cd tools/bpf; make clean", the value for $(OUTPUT) is set to
".output" (relative to the runqslower directory) by runqslower's
Makefile, and this is where the Makefile looks for files to clean up.

We cannot easily fix in the root Makefile (where "tools/bpf" is defined)
or in tools/scripts/Makefile.include (setting $(OUTPUT)), where changing
the way the output variables are passed would likely have consequences
elsewhere. We could change runqslower's Makefile to build in the
repository instead of in a dedicated ".output/", but doing so just to
accommodate a test script doesn't sound great. Instead, let's just make
sure that we clean up runqslower properly by adding the correct command
to the script.

This will attempt to clean runqslower twice: the first try with command
"cd tools/bpf; make clean" will search for tools/bpf/runqslower/.output
and fail to clean it (but will still clean the other tools, in
particular bpftool), the second one (added in this commit) sets the
$(OUTPUT) variable like for building with the "tool/bpf" target and
should succeed.
Signed-off-by: Quentin Monnet <quentin@isovalent.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211007194438.34443-12-quentin@isovalent.com

In samples/bpf/Makefile, libbpf has a FORCE dependency that force it to
be rebuilt. I read this as a way to keep the library up-to-date, given
that we do not have, in samples/bpf, a list of the source files for
libbpf itself. However, a better approach would be to use the
"$(wildcard ...)" function from make, and to have libbpf depend on all
the .c and .h files in its directory. This is what samples/bpf/Makefile
does for bpftool, and also what the BPF selftests' Makefile does for
libbpf.

Let's update the Makefile to avoid rebuilding libbpf all the time (and
bpftool on top of it).
Signed-off-by: Quentin Monnet <quentin@isovalent.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211007194438.34443-11-quentin@isovalent.com

API headers from libbpf should not be accessed directly from the source
directory. Instead, they should be exported with "make install_headers".
Make sure that samples/bpf/Makefile installs the headers properly when
building.

The object compiled from and exported by libbpf are now placed into a
subdirectory of sample/bpf/ instead of remaining in tools/lib/bpf/. We
attempt to remove this directory on "make clean". However, the "clean"
target re-enters the samples/bpf/ directory from the root of the
repository ("$(MAKE) -C ../../ M=$(CURDIR) clean"), in such a way that
$(srctree) and $(src) are not defined, making it impossible to use
$(LIBBPF_OUTPUT) and $(LIBBPF_DESTDIR) in the recipe. So we only attempt
to clean $(CURDIR)/libbpf, which is the default value.

Add a dependency on libbpf's headers for the $(TRACE_HELPERS).

We also change the output directory for bpftool, to place the generated
objects under samples/bpf/bpftool/ instead of building in bpftool's
directory directly. Doing so, we make sure bpftool reuses the libbpf
library previously compiled and installed.
Signed-off-by: Quentin Monnet <quentin@isovalent.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211007194438.34443-10-quentin@isovalent.com

Update samples/bpf/.gitignore to ignore files generated when building
the samples. Add:

  - vmlinux.h
  - the generated skeleton files (*.skel.h)
  - the samples/bpf/libbpf/ and .../bpftool/ directories, in preparation
    of a future commit which introduces a local output directory for
    building libbpf and bpftool.
Signed-off-by: Quentin Monnet <quentin@isovalent.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211007194438.34443-9-quentin@isovalent.com

API headers from libbpf should not be accessed directly from the
library's source directory. Instead, they should be exported with "make
install_headers". Let's make sure that bpf/preload/iterators/Makefile
installs the headers properly when building.
Signed-off-by: Quentin Monnet <quentin@isovalent.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20211007194438.34443-8-quentin@isovalent.com