While compiling natively, the host's cflags and ldflags are equal to
ones used from HOSTCFLAGS and HOSTLDFLAGS. When cross compiling it
should have own, used for target arch. While verification, for arm,
arm64 and x86_64 the following flags were used always:

-Wall -O2
-fomit-frame-pointer
-Wmissing-prototypes
-Wstrict-prototypes

So, add them as they were verified and used before adding
Makefile.target and lets omit "-fomit-frame-pointer" as were proposed
while review, as no sense in such optimization for samples.
Signed-off-by: NIvan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20191011002808.28206-10-ivan.khoronzhuk@linaro.org

The main reason for that - HOSTCC and CC have different aims.
HOSTCC is used to build programs running on host, that can
cross-comple target programs with CC. It was tested for arm and arm64
cross compilation, based on linaro toolchain, but should work for
others.

So, in order to split cross compilation (CC) with host build (HOSTCC),
lets base samples on Makefile.target. It allows to cross-compile
samples/bpf programs with CC while auxialry tools running on host
built with HOSTCC.
Signed-off-by: NIvan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20191011002808.28206-9-ivan.khoronzhuk@linaro.org

The Makefile.target is added only and will be used in
sample/bpf/Makefile later in order to switch cross-compiling to CC
from HOSTCC environment.

The HOSTCC is supposed to build binaries and tools running on the host
afterwards, in order to simplify build or so, like "fixdep" or else.
In case of cross compiling "fixdep" is executed on host when the rest
samples should run on target arch. In order to build binaries for
target arch with CC and tools running on host with HOSTCC, lets add
Makefile.target for simplicity, having definition and routines similar
to ones, used in script/Makefile.host. This allows later add
cross-compilation to samples/bpf with minimum changes.

The tprog stands for target programs built with CC.

Makefile.target contains only stuff needed for samples/bpf, potentially
can be reused later and now needed only for unblocking tricky
samples/bpf cross compilation.
Signed-off-by: NIvan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20191011002808.28206-8-ivan.khoronzhuk@linaro.org

Drop inclusion for bpf_load -I$(objtree)/usr/include as it is
included for all objects anyway, with above line:
KBUILD_HOSTCFLAGS += -I$(objtree)/usr/include
Signed-off-by: NIvan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>
Acked-by: NAndrii Nakryiko <andriin@fb.com>
Link: https://lore.kernel.org/bpf/20191011002808.28206-7-ivan.khoronzhuk@linaro.org

For arm, -D__LINUX_ARM_ARCH__=X is min version used as instruction
set selector and is absolutely required while parsing some parts of
headers. It's present in KBUILD_CFLAGS but not in autoconf.h, so let's
retrieve it from and add to programs cflags. In another case errors
like "SMP is not supported" for armv7 and bunch of other errors are
issued resulting to incorrect final object.
Signed-off-by: NIvan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20191011002808.28206-6-ivan.khoronzhuk@linaro.org

It can overlap with CFLAGS used for libraries built with gcc if
not now then in next patches. Correct it here for simplicity.
Signed-off-by: NIvan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>
Acked-by: NAndrii Nakryiko <andriin@fb.com>
Link: https://lore.kernel.org/bpf/20191011002808.28206-5-ivan.khoronzhuk@linaro.org

For cross compiling the target triple can be inherited from
cross-compile prefix as it's done in CLANG_FLAGS from kernel makefile.
So copy-paste this decision from kernel Makefile.
Signed-off-by: NIvan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>
Acked-by: NAndrii Nakryiko <andriin@fb.com>
Link: https://lore.kernel.org/bpf/20191011002808.28206-4-ivan.khoronzhuk@linaro.org

Don't list userspace "cookie_uid_helper_example" object in list for
bpf objects.

'always' target is used for listing bpf programs, but
'cookie_uid_helper_example.o' is a user space ELF file, and covered
by rule `per_socket_stats_example`, so shouldn't be in 'always'.
Let us remove `always += cookie_uid_helper_example.o`, which avoids
breaking cross compilation due to mismatched includes.
Signed-off-by: NIvan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>
Acked-by: NAndrii Nakryiko <andriin@fb.com>
Link: https://lore.kernel.org/bpf/20191011002808.28206-3-ivan.khoronzhuk@linaro.org

echo should be replaced with echo -e to handle '\n' correctly, but
instead, replace it with printf as some systems can't handle echo -e.
Signed-off-by: NIvan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>
Acked-by: NAndrii Nakryiko <andriin@fb.com>
Link: https://lore.kernel.org/bpf/20191011002808.28206-2-ivan.khoronzhuk@linaro.org

Old GCC versions are producing invalid typedef for __gnuc_va_list
pointing to void. Special-case this and emit valid:

typedef __builtin_va_list __gnuc_va_list;
Reported-by: NJohn Fastabend <john.fastabend@gmail.com>
Signed-off-by: NAndrii Nakryiko <andriin@fb.com>
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Acked-by: NMartin KaFai Lau <kafai@fb.com>
Acked-by: NJohn Fastabend <john.fastabend@gmail.com>
Link: https://lore.kernel.org/bpf/20191011032901.452042-1-andriin@fb.com

Existing BPF_CORE_READ() macro generates slightly suboptimal code. If
there are intermediate pointers to be read, initial source pointer is
going to be assigned into a temporary variable and then temporary
variable is going to be uniformly used as a "source" pointer for all
intermediate pointer reads. Schematically (ignoring all the type casts),
BPF_CORE_READ(s, a, b, c) is expanded into:
({
	const void *__t = src;
	bpf_probe_read(&__t, sizeof(*__t), &__t->a);
	bpf_probe_read(&__t, sizeof(*__t), &__t->b);

	typeof(s->a->b->c) __r;
	bpf_probe_read(&__r, sizeof(*__r), &__t->c);
})

This initial `__t = src` makes calls more uniform, but causes slightly
less optimal register usage sometimes when compiled with Clang. This can
cascase into, e.g., more register spills.

This patch fixes this issue by generating more optimal sequence:
({
	const void *__t;
	bpf_probe_read(&__t, sizeof(*__t), &src->a); /* <-- src here */
	bpf_probe_read(&__t, sizeof(*__t), &__t->b);

	typeof(s->a->b->c) __r;
	bpf_probe_read(&__r, sizeof(*__r), &__t->c);
})

Fixes: 7db3822a ("libbpf: Add BPF_CORE_READ/BPF_CORE_READ_INTO helpers")
Signed-off-by: NAndrii Nakryiko <andriin@fb.com>
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20191011023847.275936-1-andriin@fb.com

Fix typo 'boolian' into 'boolean'.
Signed-off-by: NAnton Ivanov <anton.ivanov@cambridgegreys.com>
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20191011084303.28418-1-anton.ivanov@cambridgegreys.com

Fix "warning: cast to pointer from integer of different size" when
casting u64 addr to void *.

Fixes: a23740ec ("bpf: Track contents of read-only maps as scalars")
Reported-by: Nkbuild test robot <lkp@intel.com>
Signed-off-by: NAndrii Nakryiko <andriin@fb.com>
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Acked-by: NMartin KaFai Lau <kafai@fb.com>
Link: https://lore.kernel.org/bpf/20191011172053.2980619-1-andriin@fb.com

Make sure a new flow dissector program can be attached to replace the old
one with a single syscall. Also check that attaching the same program twice
is prohibited.
Signed-off-by: NJakub Sitnicki <jakub@cloudflare.com>
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Reviewed-by: NStanislav Fomichev <sdf@google.com>
Acked-by: NMartin KaFai Lau <kafai@fb.com>
Link: https://lore.kernel.org/bpf/20191011082946.22695-3-jakub@cloudflare.com

It is currently not possible to detach the flow dissector program and
attach a new one in an atomic fashion, that is with a single syscall.
Attempts to do so will be met with EEXIST error.

This makes updates to flow dissector program hard. Traffic steering that
relies on BPF-powered flow dissection gets disrupted while old program has
been already detached but the new one has not been attached yet.

There is also a window of opportunity to attach a flow dissector to a
non-root namespace while updating the root flow dissector, thus blocking
the update.

Lastly, the behavior is inconsistent with cgroup BPF programs, which can be
replaced with a single bpf(BPF_PROG_ATTACH, ...) syscall without any
restrictions.

Allow attaching a new flow dissector program when another one is already
present with a restriction that it can't be the same program.
Signed-off-by: NJakub Sitnicki <jakub@cloudflare.com>
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Reviewed-by: NStanislav Fomichev <sdf@google.com>
Acked-by: NMartin KaFai Lau <kafai@fb.com>
Link: https://lore.kernel.org/bpf/20191011082946.22695-2-jakub@cloudflare.com

Do not risk spanning these small structures on two cache lines.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20191011181140.2898-1-edumazet@google.com

Andrii Nakryiko says:

====================
With BPF maps supporting direct map access (currently, array_map w/ single
element, used for global data) that are read-only both from system call and
BPF side, it's possible for BPF verifier to track its contents as known
constants.

Now it's possible for user-space control app to pre-initialize read-only map
(e.g., for .rodata section) with user-provided flags and parameters and rely
on BPF verifier to detect and eliminate dead code resulting from specific
combination of input parameters.

v1->v2:
- BPF_F_RDONLY means nothing, stick to just map->frozen (Daniel);
- stick to passing just offset into map_direct_value_addr (Martin).
====================
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>

Add tests checking that verifier does proper constant propagation for
read-only maps. If constant propagation didn't work, skipp_loop and
part_loop BPF programs would be rejected due to BPF verifier otherwise
not being able to prove they ever complete. With constant propagation,
though, they are succesfully validated as properly terminating loops.
Signed-off-by: NAndrii Nakryiko <andriin@fb.com>
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20191009201458.2679171-3-andriin@fb.com

Maps that are read-only both from BPF program side and user space side
have their contents constant, so verifier can track referenced values
precisely and use that knowledge for dead code elimination, branch
pruning, etc. This patch teaches BPF verifier how to do this.
Signed-off-by: NAndrii Nakryiko <andriin@fb.com>
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20191009201458.2679171-2-andriin@fb.com

Fix typo in struct xpd_md, generated from bpf_helpers_doc.py, which is
causing compilation warnings for programs using bpf_helpers.h

Fixes: 7a387bed ("scripts/bpf: teach bpf_helpers_doc.py to dump BPF helper definitions")
Signed-off-by: NAndrii Nakryiko <andriin@fb.com>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20191010042534.290562-1-andriin@fb.com

'struct xdp_umem_reg' has 4 bytes of padding at the end that makes
valgrind complain about passing uninitialized stack memory to the
syscall:

  Syscall param socketcall.setsockopt() points to uninitialised byte(s)
    at 0x4E7AB7E: setsockopt (in /usr/lib64/libc-2.29.so)
    by 0x4BDE035: xsk_umem__create@@LIBBPF_0.0.4 (xsk.c:172)
  Uninitialised value was created by a stack allocation
    at 0x4BDDEBA: xsk_umem__create@@LIBBPF_0.0.4 (xsk.c:140)

Padding bytes appeared after introducing of a new 'flags' field.
memset() is required to clear them.

Fixes: 10d30e30 ("libbpf: add flags to umem config")
Signed-off-by: NIlya Maximets <i.maximets@ovn.org>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>
Acked-by: NAndrii Nakryiko <andriin@fb.com>
Link: https://lore.kernel.org/bpf/20191009164929.17242-1-i.maximets@ovn.org

Andrii Nakryiko says:

====================
Fix BTF-to-C logic of handling padding at the end of a struct. Fix existing
test that should have captured this. Also move test_btf_dump into a test_progs
test to leverage common infrastructure.
====================
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>

Existing padding test case for btf_dump has a good test that was
supposed to test padding generation at the end of a struct, but its
expected output was specified incorrectly. Fix this.

Fixes: 2d2a3ad8 ("selftests/bpf: add btf_dump BTF-to-C conversion tests")
Reported-by: NJohn Fastabend <john.fastabend@gmail.com>
Signed-off-by: NAndrii Nakryiko <andriin@fb.com>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20191008231009.2991130-4-andriin@fb.com

Convert test_btf_dump into a part of test_progs, instead of
a stand-alone test binary.
Signed-off-by: NAndrii Nakryiko <andriin@fb.com>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20191008231009.2991130-3-andriin@fb.com

Fix a case where explicit padding at the end of a struct is necessary
due to non-standart alignment requirements of fields (which BTF doesn't
capture explicitly).

Fixes: 351131b5 ("libbpf: add btf_dump API for BTF-to-C conversion")
Reported-by: NJohn Fastabend <john.fastabend@gmail.com>
Signed-off-by: NAndrii Nakryiko <andriin@fb.com>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>
Tested-by: NJohn Fastabend <john.fastabend@gmail.com>
Link: https://lore.kernel.org/bpf/20191008231009.2991130-2-andriin@fb.com

Andrii Nakryiko says:

====================
This patch set makes bpf_helpers.h and bpf_endian.h a part of libbpf itself
for consumption by user BPF programs, not just selftests. It also splits off
tracing helpers into bpf_tracing.h, which also becomes part of libbpf. Some of
the legacy stuff (BPF_ANNOTATE_KV_PAIR, load_{byte,half,word}, bpf_map_def
with unsupported fields, etc, is extracted into selftests-only bpf_legacy.h.
All the selftests and samples are switched to use libbpf's headers and
selftests' ones are removed.

As part of this patch set we also add BPF_CORE_READ variadic macros, that are
simplifying BPF CO-RE reads, especially the ones that have to follow few
pointers. E.g., what in non-BPF world (and when using BCC) would be:

int x = s->a->b.c->d; /* s, a, and b.c are pointers */

Today would have to be written using explicit bpf_probe_read() calls as:

  void *t;
  int x;
  bpf_probe_read(&t, sizeof(t), s->a);
  bpf_probe_read(&t, sizeof(t), ((struct b *)t)->b.c);
  bpf_probe_read(&x, sizeof(x), ((struct c *)t)->d);

This is super inconvenient and distracts from program logic a lot. Now, with
added BPF_CORE_READ() macros, you can write the above as:

  int x = BPF_CORE_READ(s, a, b.c, d);

Up to 9 levels of pointer chasing are supported, which should be enough for
any practical purpose, hopefully, without adding too much boilerplate macro
definitions (though there is admittedly some, given how variadic and recursive
C macro have to be implemented).

There is also BPF_CORE_READ_INTO() variant, which relies on caller to allocate
space for result:

  int x;
  BPF_CORE_READ_INTO(&x, s, a, b.c, d);

Result of last bpf_probe_read() call in the chain of calls is the result of
BPF_CORE_READ_INTO(). If any intermediate bpf_probe_read() aall fails, then
all the subsequent ones will fail too, so this is sufficient to know whether
overall "operation" succeeded or not. No short-circuiting of bpf_probe_read()s
is done, though.

BPF_CORE_READ_STR_INTO() is added as well, which differs from
BPF_CORE_READ_INTO() only in that last bpf_probe_read() call (to read final
field after chasing pointers) is replaced with bpf_probe_read_str(). Result of
bpf_probe_read_str() is returned as a result of BPF_CORE_READ_STR_INTO() macro
itself, so that applications can track return code and/or length of read
string.

Patch set outline:
- patch #1 undoes previously added GCC-specific bpf-helpers.h include;
- patch #2 splits off legacy stuff we don't want to carry over;
- patch #3 adjusts CO-RE reloc tests to avoid subsequent naming conflict with
  BPF_CORE_READ;
- patch #4 splits off bpf_tracing.h;
- patch #5 moves bpf_{helpers,endian,tracing}.h and bpf_helper_defs.h
  generation into libbpf and adjusts Makefiles to include libbpf for header
  search;
- patch #6 adds variadic BPF_CORE_READ() macro family, as described above;
- patch #7 adds tests to verify all possible levels of pointer nestedness for
  BPF_CORE_READ(), as well as correctness test for BPF_CORE_READ_STR_INTO().

v4->v5:
- move BPF_CORE_READ() stuff into bpf_core_read.h header (Alexei);

v3->v4:
- rebase on latest bpf-next master;
- bpf_helper_defs.h generation is moved into libbpf's Makefile;

v2->v3:
- small formatting fixes and macro () fixes (Song);

v1->v2:
- fix CO-RE reloc tests before bpf_helpers.h move (Song);
- split off legacy stuff we don't want to carry over (Daniel, Toke);
- split off bpf_tracing.h (Daniel);
- fix samples/bpf build (assuming other fixes are applied);
- switch remaining maps either to bpf_map_def_legacy or BTF-defined maps;
====================
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>

Validate BPF_CORE_READ correctness and handling of up to 9 levels of
nestedness using cyclic task->(group_leader->)*->tgid chains.

Also add a test of maximum-dpeth BPF_CORE_READ_STR_INTO() macro.
Signed-off-by: NAndrii Nakryiko <andriin@fb.com>
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Acked-by: NJohn Fastabend <john.fastabend@gmail.com>
Acked-by: NSong Liu <songliubraving@fb.com>
Link: https://lore.kernel.org/bpf/20191008175942.1769476-8-andriin@fb.com

Add few macros simplifying BCC-like multi-level probe reads, while also
emitting CO-RE relocations for each read.
Signed-off-by: NAndrii Nakryiko <andriin@fb.com>
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Acked-by: NJohn Fastabend <john.fastabend@gmail.com>
Acked-by: NSong Liu <songliubraving@fb.com>
Link: https://lore.kernel.org/bpf/20191008175942.1769476-7-andriin@fb.com

Move bpf_helpers.h, bpf_tracing.h, and bpf_endian.h into libbpf. Move
bpf_helper_defs.h generation into libbpf's Makefile. Ensure all those
headers are installed along the other libbpf headers. Also, adjust
selftests and samples include path to include libbpf now.
Signed-off-by: NAndrii Nakryiko <andriin@fb.com>
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Acked-by: NSong Liu <songliubraving@fb.com>
Link: https://lore.kernel.org/bpf/20191008175942.1769476-6-andriin@fb.com

Split-off PT_REGS-related helpers into bpf_tracing.h header. Adjust
selftests and samples to include it where necessary.
Signed-off-by: NAndrii Nakryiko <andriin@fb.com>
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Acked-by: NJohn Fastabend <john.fastabend@gmail.com>
Acked-by: NSong Liu <songliubraving@fb.com>
Link: https://lore.kernel.org/bpf/20191008175942.1769476-5-andriin@fb.com

To allow adding a variadic BPF_CORE_READ macro with slightly different
syntax and semantics, define CORE_READ in CO-RE reloc tests, which is
a thin wrapper around low-level bpf_core_read() macro, which in turn is
just a wrapper around bpf_probe_read().
Signed-off-by: NAndrii Nakryiko <andriin@fb.com>
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Acked-by: NJohn Fastabend <john.fastabend@gmail.com>
Acked-by: NSong Liu <songliubraving@fb.com>
Link: https://lore.kernel.org/bpf/20191008175942.1769476-4-andriin@fb.com

Split off few legacy things from bpf_helpers.h into separate
bpf_legacy.h file:
- load_{byte|half|word};
- remove extra inner_idx and numa_node fields from bpf_map_def and
  introduce bpf_map_def_legacy for use in samples;
- move BPF_ANNOTATE_KV_PAIR into bpf_legacy.h.

Adjust samples and selftests accordingly by either including
bpf_legacy.h and using bpf_map_def_legacy, or switching to BTF-defined
maps altogether.
Signed-off-by: NAndrii Nakryiko <andriin@fb.com>
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Acked-by: NJohn Fastabend <john.fastabend@gmail.com>
Acked-by: NSong Liu <songliubraving@fb.com>
Link: https://lore.kernel.org/bpf/20191008175942.1769476-3-andriin@fb.com

Having GCC provide its own bpf-helper.h is not the right approach and is
going to be changed. Undo bpf_helpers.h change before moving
bpf_helpers.h into libbpf.
Signed-off-by: NAndrii Nakryiko <andriin@fb.com>
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Acked-by: NSong Liu <songliubraving@fb.com>
Acked-by: NIlya Leoshkevich <iii@linux.ibm.com>
Acked-by: NJohn Fastabend <john.fastabend@gmail.com>
Link: https://lore.kernel.org/bpf/20191008175942.1769476-2-andriin@fb.com

Currently, at xdp_adjust_tail_kern.c, MAX_PCKT_SIZE is limited
to 600. To make this size flexible, static global variable
'max_pcktsz' is added.

By updating new packet size from the user space, xdp_adjust_tail_kern.o
will use this value as a new max packet size.

This static global variable can be accesible from .data section with
bpf_object__find_map* from user space, since it is considered as
internal map (accessible with .bss/.data/.rodata suffix).

If no '-P <MAX_PCKT_SIZE>' option is used, the size of maximum packet
will be 600 as a default.

For clarity, change the helper to fetch map from 'bpf_map__next'
to 'bpf_object__find_map_fd_by_name'. Also, changed the way to
test prog_fd, map_fd from '!= 0' to '< 0', since fd could be 0
when stdin is closed.
Signed-off-by: NDaniel T. Lee <danieltimlee@gmail.com>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>
Acked-by: NAndrii Nakryiko <andriin@fb.com>
Link: https://lore.kernel.org/bpf/20191007172117.3916-1-danieltimlee@gmail.com

Stanislav Fomichev says:

====================
While having a per-net-ns flow dissector programs is convenient for
testing, security-wise it's better to have only one vetted global
flow dissector implementation.

Let's have a convention that when BPF flow dissector is installed
in the root namespace, child namespaces can't override it.

The intended use-case is to attach global BPF flow dissector
early from the init scripts/systemd. Attaching global dissector
is prohibited if some non-root namespace already has flow dissector
attached. Also, attaching to non-root namespace is prohibited
when there is flow dissector attached to the root namespace.

v3:
* drop extra check and empty line (Andrii Nakryiko)

v2:
* EPERM -> EEXIST (Song Liu)
* Make sure we don't have dissector attached to non-root namespaces
  when attaching the global one (Andrii Nakryiko)
====================
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>

Make sure non-root namespaces get an error if root flow dissector is
attached.

Cc: Petar Penkov <ppenkov@google.com>
Acked-by: NSong Liu <songliubraving@fb.com>
Signed-off-by: NStanislav Fomichev <sdf@google.com>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>

Always use init_net flow dissector BPF program if it's attached and fall
back to the per-net namespace one. Also, deny installing new programs if
there is already one attached to the root namespace.
Users can still detach their BPF programs, but can't attach any
new ones (-EEXIST).

Cc: Petar Penkov <ppenkov@google.com>
Acked-by: NAndrii Nakryiko <andriin@fb.com>
Acked-by: NSong Liu <songliubraving@fb.com>
Signed-off-by: NStanislav Fomichev <sdf@google.com>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>

Fix spelling mistake.
Signed-off-by: NAnton Ivanov <anton.ivanov@cambridgegreys.com>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>
Acked-by: NSong Liu <songliubraving@fb.com>
Link: https://lore.kernel.org/bpf/20191007082636.14686-1-anton.ivanov@cambridgegreys.com

As part of libbpf in 5e61f270 ("libbpf: stop enforcing kern_version,
populate it for users") non-LIBBPF_API __bpf_object__open_xattr() API
was removed from libbpf.h header. This broke bpftool, which relied on
that function. This patch fixes the build by switching to newly added
bpf_object__open_file() which provides the same capabilities, but is
official and future-proof API.

v1->v2:
- fix prog_type shadowing (Stanislav).

Fixes: 5e61f270 ("libbpf: stop enforcing kern_version, populate it for users")
Reported-by: NStanislav Fomichev <sdf@google.com>
Signed-off-by: NAndrii Nakryiko <andriin@fb.com>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>
Reviewed-by: NStanislav Fomichev <sdf@google.com>
Link: https://lore.kernel.org/bpf/20191007225604.2006146-1-andriin@fb.com

Current Makefile dependency chain is not strict enough and allows
test_attach_probe.o to be built before test_progs's
prog_test/attach_probe.o is built, which leads to assembler complaining
about missing included binary.

This patch is a minimal fix to fix this issue by enforcing that
test_attach_probe.o (BPF object file) is built before
prog_tests/attach_probe.c is attempted to be compiled.

Fixes: 928ca75e ("selftests/bpf: switch tests to new bpf_object__open_{file, mem}() APIs")
Signed-off-by: NAndrii Nakryiko <andriin@fb.com>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20191007204149.1575990-1-andriin@fb.com