bpf_object_open_opts.attach_prog_fd makes a pretty strong assumption
that bpf_object contains either only single freplace BPF program or all
of BPF programs in BPF object are freplaces intended to replace
different subprograms of the same target BPF program. This seems both
a bit confusing, too assuming, and limiting.
We've had bpf_program__set_attach_target() API which allows more
fine-grained control over this, on a per-program level. As such, mark
open_opts.attach_prog_fd as deprecated starting from v0.7, so that we
have one more universal way of setting freplace targets. With previous
change to allow NULL attach_func_name argument, and especially combined
with BPF skeleton, arguable bpf_program__set_attach_target() is a more
convenient and explicit API as well.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Yonghong Song <yhs@fb.com>
Link: https://lore.kernel.org/bpf/20210916015836.1248906-7-andrii@kernel.org
Switch fexit_bpf2bpf selftest to bpf_program__set_attach_target()
instead of using bpf_object_open_opts.attach_prog_fd, which is going to
be deprecated. These changes also demonstrate the new mode of
set_attach_target() in which it allows NULL when the target is BPF
program (attach_prog_fd != 0).
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Yonghong Song <yhs@fb.com>
Link: https://lore.kernel.org/bpf/20210916015836.1248906-6-andrii@kernel.org
Allow to use bpf_program__set_attach_target to only set target attach
program FD, while letting libbpf to use target attach function name from
SEC() definition. This might be useful for some scenarios where
bpf_object contains multiple related freplace BPF programs intended to
replace different sub-programs in target BPF program. In such case all
programs will have the same attach_prog_fd, but different
attach_func_name. It's convenient to specify such target function names
declaratively in SEC() definitions, but attach_prog_fd is a dynamic
runtime setting.
To simplify such scenario, allow bpf_program__set_attach_target() to
delay BTF ID resolution till the BPF program load time by providing NULL
attach_func_name. In that case the behavior will be similar to using
bpf_object_open_opts.attach_prog_fd (which is marked deprecated since
v0.7), but has the benefit of allowing more control by user in what is
attached to what. Such setup allows having BPF programs attached to
different target attach_prog_fd with target functions still declaratively
recorded in BPF source code in SEC() definitions.
Selftests changes in the next patch should make this more obvious.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Yonghong Song <yhs@fb.com>
Link: https://lore.kernel.org/bpf/20210916015836.1248906-5-andrii@kernel.org
It's relevant and hasn't been doing anything for a long while now.
Deprecated it.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Yonghong Song <yhs@fb.com>
Link: https://lore.kernel.org/bpf/20210916015836.1248906-4-andrii@kernel.org
relaxed_core_relocs option hasn't had any effect for a while now, stop
specifying it. Next patch marks it as deprecated.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Yonghong Song <yhs@fb.com>
Link: https://lore.kernel.org/bpf/20210916015836.1248906-3-andrii@kernel.org
Don't perform another search for sec_def inside
libbpf_find_attach_btf_id(), as each recognized bpf_program already has
prog->sec_def set.
Also remove unnecessary NULL check for prog->sec_name, as it can never
be NULL.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Yonghong Song <yhs@fb.com>
Link: https://lore.kernel.org/bpf/20210916015836.1248906-2-andrii@kernel.org
BPF programs run with migration disabled regardless of preemption, as
they are protected by migrate_disable(). Update the uapi documentation
accordingly.
Signed-off-by: Matteo Croce <mcroce@microsoft.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Yonghong Song <yhs@fb.com>
Link: https://lore.kernel.org/bpf/20210914235400.59427-1-mcroce@linux.microsoft.com
This adds comments above five functions in btf.h which document
their uses. These comments are of a format that doxygen and sphinx
can pick up and render. These are rendered by libbpf.readthedocs.org
Signed-off-by: Grant Seltzer <grantseltzer@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20210915021951.117186-1-grantseltzer@gmail.com
Commit c240ba2878 ("selftests/bpf: Add a test with a bpf
program with btf_tag attributes") added btf_tag selftest
to test BTF_KIND_TAG generation from C source code, and to
test kernel validation of generated BTF types.
But if an old clang (clang 13 or earlier) is used, the
following compiler warning may be seen:
progs/tag.c:23:20: warning: unknown attribute 'btf_tag' ignored
and the test itself is marked OK. The compiler warning is bad
and the test itself shouldn't be marked OK.
This patch added the check for btf_tag attribute support.
If btf_tag is not supported by the clang, the attribute will
not be used in the code and the test will be marked as skipped.
For example, with clang 13:
./test_progs -t btf_tag
#21 btf_tag:SKIP
Summary: 1/0 PASSED, 1 SKIPPED, 0 FAILED
The selftests/README.rst is updated to clarify when the btf_tag
test may be skipped.
Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20210915061036.2577971-1-yhs@fb.com
Yonghong Song says:
====================
LLVM14 added support for a new C attribute ([1])
__attribute__((btf_tag("arbitrary_str")))
This attribute will be emitted to dwarf ([2]) and pahole
will convert it to BTF. Or for bpf target, this
attribute will be emitted to BTF directly ([3], [4]).
The attribute is intended to provide additional
information for
- struct/union type or struct/union member
- static/global variables
- static/global function or function parameter.
This new attribute can be used to add attributes
to kernel codes, e.g., pre- or post- conditions,
allow/deny info, or any other info in which only
the kernel is interested. Such attributes will
be processed by clang frontend and emitted to
dwarf, converting to BTF by pahole. Ultimiately
the verifier can use these information for
verification purpose.
The new attribute can also be used for bpf
programs, e.g., tagging with __user attributes
for function parameters, specifying global
function preconditions, etc. Such information
may help verifier to detect user program
bugs.
After this series, pahole dwarf->btf converter
will be enhanced to support new llvm tag
for btf_tag attribute. With pahole support,
we will then try to add a few real use case,
e.g., __user/__rcu tagging, allow/deny list,
some kernel function precondition, etc,
in the kernel.
In the rest of the series, Patches 1-2 had
kernel support. Patches 3-4 added
libbpf support. Patch 5 added bpftool
support. Patches 6-10 added various selftests.
Patch 11 added documentation for the new kind.
[1] https://reviews.llvm.org/D106614
[2] https://reviews.llvm.org/D106621
[3] https://reviews.llvm.org/D106622
[4] https://reviews.llvm.org/D109560
Changelog:
v2 -> v3:
- put NR_BTF_KINDS and BTF_KIND_MAX into enum as well
- check component_idx earlier (check_meta stage) in kernel
- add more tests
- fix misc nits
v1 -> v2:
- BTF ELF format changed in llvm ([4] above),
so cross-board change to use the new format.
- Clarified in commit message that BTF_KIND_TAG
is not emitted by bpftool btf dump format c.
- Fix various comments from Andrii.
====================
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Add a bpf program with btf_tag attributes. The program is
loaded successfully with the kernel. With the command
bpftool btf dump file ./tag.o
the following dump shows that tags are properly encoded:
[8] STRUCT 'key_t' size=12 vlen=3
'a' type_id=2 bits_offset=0
'b' type_id=2 bits_offset=32
'c' type_id=2 bits_offset=64
[9] TAG 'tag1' type_id=8 component_id=-1
[10] TAG 'tag2' type_id=8 component_id=-1
[11] TAG 'tag1' type_id=8 component_id=1
[12] TAG 'tag2' type_id=8 component_id=1
...
[21] FUNC_PROTO '(anon)' ret_type_id=2 vlen=1
'x' type_id=2
[22] FUNC 'foo' type_id=21 linkage=static
[23] TAG 'tag1' type_id=22 component_id=0
[24] TAG 'tag2' type_id=22 component_id=0
[25] TAG 'tag1' type_id=22 component_id=-1
[26] TAG 'tag2' type_id=22 component_id=-1
...
[29] VAR 'total' type_id=27, linkage=global
[30] TAG 'tag1' type_id=29 component_id=-1
[31] TAG 'tag2' type_id=29 component_id=-1
If an old clang compiler, which does not support btf_tag attribute,
is used, these btf_tag attributes will be silently ignored.
Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20210914223058.248949-1-yhs@fb.com
Add unit tests for BTF_KIND_TAG deduplication for
- struct and struct member
- variable
- func and func argument
Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20210914223052.248535-1-yhs@fb.com
Test good and bad variants of BTF_KIND_TAG encoding.
Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20210914223047.248223-1-yhs@fb.com
BTF_KIND_TAG ELF format has a component_idx which might have value -1.
test_btf may confuse it with common_type.name as NAME_NTH checkes
high 16bit to be 0xffff. Change NAME_NTH high 16bit check to be
0xfffe so it won't confuse with component_idx.
Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20210914223041.248009-1-yhs@fb.com
Added bpftool support to dump BTF_KIND_TAG information.
The new bpftool will be used in later patches to dump
btf in the test bpf program object file.
Currently, the tags are not emitted with
bpftool btf dump file <path> format c
and they are silently ignored. The tag information is
mostly used in the kernel for verification purpose and the kernel
uses its own btf to check. With adding these tags
to vmlinux.h, tags will be encoded in program's btf but
they will not be used by the kernel, at least for now.
So let us delay adding these tags to format C header files
until there is a real need.
Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20210914223031.246951-1-yhs@fb.com
Add BTF_KIND_TAG support for parsing and dedup.
Also added sanitization for BTF_KIND_TAG. If BTF_KIND_TAG is not
supported in the kernel, sanitize it to INTs.
Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20210914223025.246687-1-yhs@fb.com
This patch renames functions btf_{hash,equal}_int() to
btf_{hash,equal}_int_tag() so they can be reused for
BTF_KIND_TAG support. There is no functionality change for
this patch.
Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20210914223020.245829-1-yhs@fb.com
LLVM14 added support for a new C attribute ([1])
__attribute__((btf_tag("arbitrary_str")))
This attribute will be emitted to dwarf ([2]) and pahole
will convert it to BTF. Or for bpf target, this
attribute will be emitted to BTF directly ([3], [4]).
The attribute is intended to provide additional
information for
- struct/union type or struct/union member
- static/global variables
- static/global function or function parameter.
For linux kernel, the btf_tag can be applied
in various places to specify user pointer,
function pre- or post- condition, function
allow/deny in certain context, etc. Such information
will be encoded in vmlinux BTF and can be used
by verifier.
The btf_tag can also be applied to bpf programs
to help global verifiable functions, e.g.,
specifying preconditions, etc.
This patch added basic parsing and checking support
in kernel for new BTF_KIND_TAG kind.
[1] https://reviews.llvm.org/D106614
[2] https://reviews.llvm.org/D106621
[3] https://reviews.llvm.org/D106622
[4] https://reviews.llvm.org/D109560
Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20210914223015.245546-1-yhs@fb.com
Change BTF_KIND_* macros to enums so they are encoded in dwarf and
appear in vmlinux.h. This will make it easier for bpf programs
to use these constants without macro definitions.
Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20210914223009.245307-1-yhs@fb.com
List all possible test_progs flavors explicitly to avoid accidentally
ignoring valid source code files. In this case, test_progs.c was still
ignored after recent 809ed84de8 ("selftests/bpf: Whitelist test_progs.h
from .gitignore") fix that added exception only for test_progs.h.
Fixes: 74b5a5968f ("selftests/bpf: Replace test_progs and test_maps w/ general rule")
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Yonghong Song <yhs@fb.com>
Link: https://lore.kernel.org/bpf/20210914162228.3995740-1-andrii@kernel.org
IMUL allows for multiple operands and saving and storing rax/rdx is no
longer needed. Signedness of the operands doesn't matter here because
the we only keep the lower 32/64 bit of the product for 32/64 bit
multiplications.
Signed-off-by: Jie Meng <jmeng@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20210913211337.1564014-1-jmeng@fb.com
Andrii Nakryiko says:
====================
This small patch set performs internal refactorings around libbpf BPF program
ELF section definitions' handling. This is preparatory changes for further
changes around making libbpf BPF program section handling more strict but also
pluggable and customizable, as part of the libbpf 1.0 effort. See individual
patches for details.
====================
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Remove almost all the code that explicitly iterated BPF program section
definitions in favor of using find_sec_def(). The only remaining user of
section_defs is libbpf_get_type_names that has to iterate all of them to
construct its result.
Having one internal API entry point for section definitions will
simplify further refactorings around libbpf's program section
definitions parsing.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Link: https://lore.kernel.org/bpf/20210914014733.2768-5-andrii@kernel.org
Remove the need to explicitly pass bpf_sec_def for auto-attachable BPF
programs, as it is already recorded at bpf_object__open() time for all
recognized type of BPF programs. This further reduces number of explicit
calls to find_sec_def(), simplifying further refactorings.
No functional changes are done by this patch.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Link: https://lore.kernel.org/bpf/20210914014733.2768-4-andrii@kernel.org
Refactor bpf_object__open() sequencing to perform BPF program type
detection based on SEC() definitions before we get to relocations
collection. This allows to have more information about BPF program by
the time we get to, say, struct_ops relocation gathering. This,
subsequently, simplifies struct_ops logic and removes the need to
perform extra find_sec_def() resolution.
With this patch libbpf will require all struct_ops BPF programs to be
marked with SEC("struct_ops") or SEC("struct_ops/xxx") annotations.
Real-world applications are already doing that through something like
selftests's BPF_STRUCT_OPS() macro. This change streamlines libbpf's
internal handling of SEC() definitions and is in the sprit of
upcoming libbpf-1.0 section strictness changes ([0]).
[0] https://github.com/libbpf/libbpf/wiki/Libbpf:-the-road-to-v1.0#stricter-and-more-uniform-bpf-program-section-name-sec-handling
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Link: https://lore.kernel.org/bpf/20210914014733.2768-3-andrii@kernel.org
Update struct_ops selftests to always specify "struct_ops" section
prefix. Libbpf will require a proper BPF program type set in the next
patch, so this prevents tests breaking.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Link: https://lore.kernel.org/bpf/20210914014733.2768-2-andrii@kernel.org
Allow kprobe tracepoint events creation through legacy interface, as the
kprobe dynamic PMUs support, used by default, was only created in v4.17.
Store legacy kprobe name in struct bpf_perf_link, instead of creating
a new "subclass" off of bpf_perf_link. This is ok as it's just two new
fields, which are also going to be reused for legacy uprobe support in
follow up patches.
Signed-off-by: Rafael David Tinoco <rafaeldtinoco@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20210912064844.3181742-1-rafaeldtinoco@gmail.com
Turn previously auto-generated libbpf_version.h header into a normal
header file. This prevents various tricky Makefile integration issues,
simplifies the overall build process, but also allows to further extend
it with some more versioning-related APIs in the future.
To prevent accidental out-of-sync versions as defined by libbpf.map and
libbpf_version.h, Makefile checks their consistency at build time.
Simultaneously with this change bump libbpf.map to v0.6.
Also undo adding libbpf's output directory into include path for
kernel/bpf/preload, bpftool, and resolve_btfids, which is not necessary
because libbpf_version.h is just a normal header like any other.
Fixes: 0b46b75505 ("libbpf: Add LIBBPF_DEPRECATED_SINCE macro for scheduling API deprecations")
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20210913222309.3220849-1-andrii@kernel.org
The tailcall_3 test program uses bpf_tail_call_static() where the JIT
would patch a direct jump. Add a new tailcall_6 test program replicating
exactly the same test just ensuring that bpf_tail_call() uses a map
index where the verifier cannot make assumptions this time.
In other words, this will now cover both on x86-64 JIT, meaning, JIT
images with emit_bpf_tail_call_direct() emission as well as JIT images
with emit_bpf_tail_call_indirect() emission.
# echo 1 > /proc/sys/net/core/bpf_jit_enable
# ./test_progs -t tailcalls
#136/1 tailcalls/tailcall_1:OK
#136/2 tailcalls/tailcall_2:OK
#136/3 tailcalls/tailcall_3:OK
#136/4 tailcalls/tailcall_4:OK
#136/5 tailcalls/tailcall_5:OK
#136/6 tailcalls/tailcall_6:OK
#136/7 tailcalls/tailcall_bpf2bpf_1:OK
#136/8 tailcalls/tailcall_bpf2bpf_2:OK
#136/9 tailcalls/tailcall_bpf2bpf_3:OK
#136/10 tailcalls/tailcall_bpf2bpf_4:OK
#136/11 tailcalls/tailcall_bpf2bpf_5:OK
#136 tailcalls:OK
Summary: 1/11 PASSED, 0 SKIPPED, 0 FAILED
# echo 0 > /proc/sys/net/core/bpf_jit_enable
# ./test_progs -t tailcalls
#136/1 tailcalls/tailcall_1:OK
#136/2 tailcalls/tailcall_2:OK
#136/3 tailcalls/tailcall_3:OK
#136/4 tailcalls/tailcall_4:OK
#136/5 tailcalls/tailcall_5:OK
#136/6 tailcalls/tailcall_6:OK
[...]
For interpreter, the tailcall_1-6 tests are passing as well. The later
tailcall_bpf2bpf_* are failing due lack of bpf2bpf + tailcall support
in interpreter, so this is expected.
Also, manual inspection shows that both loaded programs from tailcall_3
and tailcall_6 test case emit the expected opcodes:
* tailcall_3 disasm, emit_bpf_tail_call_direct():
[...]
b: push %rax
c: push %rbx
d: push %r13
f: mov %rdi,%rbx
12: movabs $0xffff8d3f5afb0200,%r13
1c: mov %rbx,%rdi
1f: mov %r13,%rsi
22: xor %edx,%edx _
24: mov -0x4(%rbp),%eax | limit check
2a: cmp $0x20,%eax |
2d: ja 0x0000000000000046 |
2f: add $0x1,%eax |
32: mov %eax,-0x4(%rbp) |_
38: nopl 0x0(%rax,%rax,1)
3d: pop %r13
3f: pop %rbx
40: pop %rax
41: jmpq 0xffffffffffffe377
[...]
* tailcall_6 disasm, emit_bpf_tail_call_indirect():
[...]
47: movabs $0xffff8d3f59143a00,%rsi
51: mov %edx,%edx
53: cmp %edx,0x24(%rsi)
56: jbe 0x0000000000000093 _
58: mov -0x4(%rbp),%eax | limit check
5e: cmp $0x20,%eax |
61: ja 0x0000000000000093 |
63: add $0x1,%eax |
66: mov %eax,-0x4(%rbp) |_
6c: mov 0x110(%rsi,%rdx,8),%rcx
74: test %rcx,%rcx
77: je 0x0000000000000093
79: pop %rax
7a: mov 0x30(%rcx),%rcx
7e: add $0xb,%rcx
82: callq 0x000000000000008e
87: pause
89: lfence
8c: jmp 0x0000000000000087
8e: mov %rcx,(%rsp)
92: retq
[...]
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Tested-by: Tiezhu Yang <yangtiezhu@loongson.cn>
Acked-by: Yonghong Song <yhs@fb.com>
Acked-by: Johan Almbladh <johan.almbladh@anyfinetworks.com>
Acked-by: Paul Chaignon <paul@cilium.io>
Link: https://lore.kernel.org/bpf/CAM1=_QRyRVCODcXo_Y6qOm1iT163HoiSj8U2pZ8Rj3hzMTT=HQ@mail.gmail.com
Link: https://lore.kernel.org/bpf/20210910091900.16119-1-daniel@iogearbox.net
Song Liu says:
====================
Changes v6 => v7:
1. Improve/fix intel_pmu_snapshot_branch_stack() logic. (Peter).
Changes v5 => v6:
1. Add local_irq_save/restore to intel_pmu_snapshot_branch_stack. (Peter)
2. Remove buf and size check in bpf_get_branch_snapshot, move flags check
to later fo the function. (Peter, Andrii)
3. Revise comments for bpf_get_branch_snapshot in bpf.h (Andrii)
Changes v4 => v5:
1. Modify perf_snapshot_branch_stack_t to save some memcpy. (Andrii)
2. Minor fixes in selftests. (Andrii)
Changes v3 => v4:
1. Do not reshuffle intel_pmu_disable_all(). Use some inline to save LBR
entries. (Peter)
2. Move static_call(perf_snapshot_branch_stack) to the helper. (Alexei)
3. Add argument flags to bpf_get_branch_snapshot. (Andrii)
4. Make MAX_BRANCH_SNAPSHOT an enum (Andrii). And rename it as
PERF_MAX_BRANCH_SNAPSHOT
5. Make bpf_get_branch_snapshot similar to bpf_read_branch_records.
(Andrii)
6. Move the test target function to bpf_testmod. Updated kallsyms_find_next
to work properly with modules. (Andrii)
Changes v2 => v3:
1. Fix the use of static_call. (Peter)
2. Limit the use to perfmon version >= 2. (Peter)
3. Modify intel_pmu_snapshot_branch_stack() to use intel_pmu_disable_all
and intel_pmu_enable_all().
Changes v1 => v2:
1. Rename the helper as bpf_get_branch_snapshot;
2. Fix/simplify the use of static_call;
3. Instead of percpu variables, let intel_pmu_snapshot_branch_stack output
branch records to an output argument of type perf_branch_snapshot.
Branch stack can be very useful in understanding software events. For
example, when a long function, e.g. sys_perf_event_open, returns an errno,
it is not obvious why the function failed. Branch stack could provide very
helpful information in this type of scenarios.
This set adds support to read branch stack with a new BPF helper
bpf_get_branch_trace(). Currently, this is only supported in Intel systems.
It is also possible to support the same feaure for PowerPC.
The hardware that records the branch stace is not stopped automatically on
software events. Therefore, it is necessary to stop it in software soon.
Otherwise, the hardware buffers/registers will be flushed. One of the key
design consideration in this set is to minimize the number of branch record
entries between the event triggers and the hardware recorder is stopped.
Based on this goal, current design is different from the discussions in
original RFC [1]:
1) Static call is used when supported, to save function pointer
dereference;
2) intel_pmu_lbr_disable_all is used instead of perf_pmu_disable(),
because the latter uses about 10 entries before stopping LBR.
With current code, on Intel CPU, LBR is stopped after 7 branch entries
after fexit triggers:
ID: 0 from bpf_get_branch_snapshot+18 to intel_pmu_snapshot_branch_stack+0
ID: 1 from __brk_limit+477143934 to bpf_get_branch_snapshot+0
ID: 2 from __brk_limit+477192263 to __brk_limit+477143880 # trampoline
ID: 3 from __bpf_prog_enter+34 to __brk_limit+477192251
ID: 4 from migrate_disable+60 to __bpf_prog_enter+9
ID: 5 from __bpf_prog_enter+4 to migrate_disable+0
ID: 6 from bpf_testmod_loop_test+20 to __bpf_prog_enter+0
ID: 7 from bpf_testmod_loop_test+20 to bpf_testmod_loop_test+13
ID: 8 from bpf_testmod_loop_test+20 to bpf_testmod_loop_test+13
...
[1] https://lore.kernel.org/bpf/20210818012937.2522409-1-songliubraving@fb.com/
====================
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
This test uses bpf_get_branch_snapshot from a fexit program. The test uses
a target function (bpf_testmod_loop_test) and compares the record against
kallsyms. If there isn't enough record matching kallsyms, the test fails.
Signed-off-by: Song Liu <songliubraving@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Link: https://lore.kernel.org/bpf/20210910183352.3151445-4-songliubraving@fb.com
Introduce bpf_get_branch_snapshot(), which allows tracing pogram to get
branch trace from hardware (e.g. Intel LBR). To use the feature, the
user need to create perf_event with proper branch_record filtering
on each cpu, and then calls bpf_get_branch_snapshot in the bpf function.
On Intel CPUs, VLBR event (raw event 0x1b00) can be use for this.
Signed-off-by: Song Liu <songliubraving@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20210910183352.3151445-3-songliubraving@fb.com
The typical way to access branch record (e.g. Intel LBR) is via hardware
perf_event. For CPUs with FREEZE_LBRS_ON_PMI support, PMI could capture
reliable LBR. On the other hand, LBR could also be useful in non-PMI
scenario. For example, in kretprobe or bpf fexit program, LBR could
provide a lot of information on what happened with the function. Add API
to use branch record for software use.
Note that, when the software event triggers, it is necessary to stop the
branch record hardware asap. Therefore, static_call is used to remove some
branch instructions in this process.
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Song Liu <songliubraving@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/bpf/20210910183352.3151445-2-songliubraving@fb.com
BPF programs may want to know hardware timestamps if NIC supports
such timestamping.
Expose this data as hwtstamp field of __sk_buff the same way as
gso_segs/gso_size. This field could be accessed from the same
programs as tstamp field, but it's read-only field. Explicit test
to deny access to padding data is added to bpf_skb_is_valid_access.
Also update BPF_PROG_TEST_RUN tests of the feature.
Signed-off-by: Vadim Fedorenko <vfedorenko@novek.ru>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Link: https://lore.kernel.org/bpf/20210909220409.8804-2-vfedorenko@novek.ru
Magnus Karlsson says:
====================
This patch set facilitates adding new tests as well as describing
existing ones in the xsk selftests suite and adds 3 new test suites at
the end. The idea is to isolate the run-time that executes the test
from the actual implementation of the test. Today, implementing a test
amounts to adding test specific if-statements all around the run-time,
which is not scalable or amenable for reuse. This patch set instead
introduces a test specification that is the only thing that a test
fills in. The run-time then gets this specification and acts upon it
completely unaware of what test it is executing. This way, we can get
rid of all test specific if-statements from the run-time and the
implementation of the test can be contained in a single function. This
hopefully makes it easier to add tests and for users to understand
what the test accomplishes.
As a recap of what the run-time does: each test is based on the
run-time launching two threads and connecting a veth link between the
two threads. Each thread opens an AF_XDP socket on that veth interface
and one of them sends traffic that the other one receives and
validates. Each thread has its own umem. Note that this behavior is
not changed by this patch set.
A test specification consists of several items. Most importantly:
* Two packet streams. One for Tx thread that specifies what traffic to
send and one for the Rx thread that specifies what that thread
should receive. If it receives exactly what is specified, the test
passes, otherwise it fails. A packet stream can also specify what
buffers in the umem that should be used by the Rx and Tx threads.
* What kind of AF_XDP sockets it should create and bind to what
interfaces
* How many times it should repeat the socket creation and destruction
* The name of the test
The interface for the test spec is the following:
void test_spec_init(struct test_spec *test, struct ifobject *ifobj_tx,
struct ifobject *ifobj_rx, enum test_mode mode);
/* Reset everything but the interface specifications and the mode */
void test_spec_reset(struct test_spec *test);
void test_spec_set_name(struct test_spec *test, const char *name);
Packet streams have the following interfaces:
struct pkt *pkt_stream_get_pkt(struct pkt_stream *pkt_stream, u32 pkt_nb)
struct pkt *pkt_stream_get_next_rx_pkt(struct pkt_stream *pkt_stream)
struct pkt_stream *pkt_stream_generate(struct xsk_umem_info *umem,
u32 nb_pkts, u32 pkt_len);
void pkt_stream_delete(struct pkt_stream *pkt_stream);
struct pkt_stream *pkt_stream_clone(struct xsk_umem_info *umem,
struct pkt_stream *pkt_stream);
/* Replaces all packets in the stream*/
void pkt_stream_replace(struct test_spec *test, u32 nb_pkts, u32 pkt_len);
/* Replaces every other packet in the stream */
void pkt_stream_replace_half(struct test_spec *test, u32 pkt_len, u32 offset);
/* For creating custom made packet streams */
void pkt_stream_generate_custom(struct test_spec *test, struct pkt *pkts,
u32 nb_pkts);
/* Restores the default packet stream */
void pkt_stream_restore_default(struct test_spec *test);
A test can then then in the most basic case described like this
(provided the test specification has been created before calling the
function):
static bool testapp_aligned(struct test_spec *test)
{
test_spec_set_name(test, "RUN_TO_COMPLETION");
testapp_validate_traffic(test);
}
Running the same test in unaligned mode would then look like this:
static bool testapp_unaligned(struct test_spec *test)
{
if (!hugepages_present(test->ifobj_tx)) {
ksft_test_result_skip("No 2M huge pages present.\n");
return false;
}
test_spec_set_name(test, "UNALIGNED_MODE");
test->ifobj_tx->umem->unaligned_mode = true;
test->ifobj_rx->umem->unaligned_mode = true;
/* Let half of the packets straddle a buffer boundrary */
pkt_stream_replace_half(test, PKT_SIZE,
XSK_UMEM__DEFAULT_FRAME_SIZE - 32);
/* Populate fill ring with addresses in the packet stream */
test->ifobj_rx->pkt_stream->use_addr_for_fill = true;
testapp_validate_traffic(test);
pkt_stream_restore_default(test);
return true;
}
3 of the last 4 patches in the set add 3 new test suites, one for
unaligned mode, one for testing the rejection of tricky invalid
descriptors plus the acceptance of some valid ones in the Tx ring, and
one for testing 2K frame sizes (the default is 4K).
What is left to do for follow-up patches:
* Convert the statistics tests to the new framework.
* Implement a way of registering new tests without having the enum
test_type. Once this has been done (together with the previous
bullet), all the test types can be dropped from the header
file. This means that we should be able to add tests by just writing
a single function with a new test specification, which is one of the
goals.
* Introduce functions for manipulating parts of the test or interface
spec instead of direct manipulations such as
test->ifobj_rx->pkt_stream->use_addr_for_fill = true; which is kind
of awkward.
* Move the run-time and its interface to its own .c and .h files. Then
we can have all the tests in a separate file.
* Better error reporting if a test fails. Today it does not state what
test fails and might not continue execute the rest of the tests due
to this failure. Failures are not propagated upwards through the
functions so a failed test will also be a passed test, which messes
up the stats counting. This needs to be changed.
* Add option to run specific test instead of all of them
* Introduce pacing of sent packets so that they are never dropped
by the receiver even if it is stalled for some reason. If you run
the current tests on a heavily loaded system, they might fail in SKB
mode due to packets being dropped by the driver on Tx. Though I have
never seen it, it might happen.
v1 -> v2:
* Fixed a number of spelling errors [Maciej]
* Fixed use after free bug in pkt_stream_replace() [Maciej]
* pkt_stream_set -> pkt_stream_generate_custom [Maciej]
* Fixed formatting problem in testapp_invalid_desc() [Maciej]
====================
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Add tests for 2K frame size. Both a standard send and receive test and
one testing for invalid descriptors when the frame size is 2K.
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Link: https://lore.kernel.org/bpf/20210907071928.9750-21-magnus.karlsson@gmail.com
Add tests for invalid xsk descriptors in the Tx ring. A number of
handcrafted nasty invalid descriptors are created and submitted to the
tx ring to check that they are validated correctly. Corner case valid
ones are also sent. The tests are run for both aligned and unaligned
mode.
pkt_stream_set() is introduced to be able to create a hand-crafted
packet stream where every single packet is specified in detail.
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Link: https://lore.kernel.org/bpf/20210907071928.9750-20-magnus.karlsson@gmail.com
Eliminate a test specific if-statement for the RX_FILL_EMTPY stats
test that is present in the test runner. We can do this as we now have
the use_addr_for_fill option. Just create and empty Rx packet stream
and indicated that the test runner should use the addresses in that to
populate the fill ring. As there are no packets in the stream, the
fill ring will be empty and we will get the error stats that we want
to test.
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Link: https://lore.kernel.org/bpf/20210907071928.9750-19-magnus.karlsson@gmail.com
Add a test for unaligned mode in which packet buffers can be placed
anywhere within the umem. Some packets are made to straddle page
boundaries in order to check for correctness. On the Tx side, buffers
are now allocated according to the addresses found in the packet
stream. Thus, the placement of buffers can be controlled with the
boolean use_addr_for_fill in the packet stream.
One new pkt_stream interface is introduced: pkt_stream_replace_half()
that replaces every other packet in the default packet stream with the
specified new packet. The constant DEFAULT_OFFSET is also
introduced. It specifies at what offset from the start of a chunk a Tx
packet is placed by the sending thread. This is just to be able to
test that it is possible to send packets at an offset not equal to
zero.
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Link: https://lore.kernel.org/bpf/20210907071928.9750-18-magnus.karlsson@gmail.com
Introduce the concept of a default packet stream that is the set of
packets sent by most tests. Then add the ability to replace it for a
test that would like to send or receive something else through the use
of the function pkt_stream_replace() and then restored with
pkt_stream_restore_default(). These are then used to convert the
STAT_TEST_TX_INVALID to use these new APIs.
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Link: https://lore.kernel.org/bpf/20210907071928.9750-17-magnus.karlsson@gmail.com
Allow for invalid packets to be sent. These are verified by the Rx
thread not to be received. Or put in another way, if they are
received, the test will fail. This feature will be used to eliminate
an if statement for a stats test and will also be used by other tests
in later patches. The previous code could only deal with valid
packets.
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Link: https://lore.kernel.org/bpf/20210907071928.9750-16-magnus.karlsson@gmail.com
Remove the MAX_SOCKS define as it always will be one for the forseable
future and the code does not work for any other case anyway.
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Link: https://lore.kernel.org/bpf/20210907071928.9750-15-magnus.karlsson@gmail.com
Make the pthread_t variables local scope instead of global. No reason
for them to be global.
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Link: https://lore.kernel.org/bpf/20210907071928.9750-14-magnus.karlsson@gmail.com
Make xdp_flags and bind_flags local instead of global by moving them
into the interface object. These flags decide if the socket should be
created in SKB mode or in DRV mode and therefore they are sticky and
will survive a test_spec_reset. Since every test is first run in SKB
mode then in DRV mode, this change only happens once. With this
change, the configured_mode global variable can also be
erradicated. The first test_spec_init() also becomes superfluous and
can be eliminated.
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Link: https://lore.kernel.org/bpf/20210907071928.9750-13-magnus.karlsson@gmail.com
Add the ability in the test specification to specify numbers of
sockets to create. The default is one socket. This is then used to
remove test specific if-statements around the bpf_res tests.
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Link: https://lore.kernel.org/bpf/20210907071928.9750-12-magnus.karlsson@gmail.com
Replace the second_step global variable with a test specification
variable called total_steps that a test can be set to indicate how
many times the packet stream should be sent without reinitializing any
sockets. This eliminates test specific code in the test runner around
the bidirectional test.
The total_steps variable is 1 by default as most tests only need a
single round of packets.
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Link: https://lore.kernel.org/bpf/20210907071928.9750-11-magnus.karlsson@gmail.com
Introduce rx_on and tx_on in the ifobject so that we can describe if
the thread should create a socket with only tx, rx, or both. This
eliminates some test specific if statements from the code. We can also
eliminate the flow vector structure now as this is fully specified
by the tx_on and rx_on variables.
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Link: https://lore.kernel.org/bpf/20210907071928.9750-10-magnus.karlsson@gmail.com
