| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
* "BPF ADD" should be "BPF_ADD".
* "src" should be "src_reg" in several places. The latter is the field name
in the instruction. The former refers to the value of the register, or the
immediate.
* Add '' around field names in one sentence, for consistency with the rest
of the document.
Signed-off-by: Dave Thaler <dthaler1968@gmail.com>
Acked-by: David Vernet <void@manifault.com>
Link: https://lore.kernel.org/r/20240221173535.16601-1-dthaler1968@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
Eduard Zingerman says:
====================
selftests/bpf: reduce tcp_custom_syncookie verification complexity
Thread [0] discusses a fix for bpf_loop() handling bug.
That change makes tcp_custom_syncookie test too complex to verify.
The fix discussed in [0] would be sent via 'bpf' tree,
tcp_custom_syncookie test is not in 'bpf' tree yet.
As agreed in [0] I'm sending syncookie test update separately.
[0] https://lore.kernel.org/bpf/20240216150334.31937-1-eddyz87@gmail.com/
====================
Link: https://lore.kernel.org/r/20240222150300.14909-1-eddyz87@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
This commit updates tcp_custom_syncookie.c:tcp_parse_option() to use
explicit packet offset (ctx->off) for packet access instead of ever
moving pointer (ctx->ptr), this reduces verification complexity:
- the tcp_parse_option() is passed as a callback to bpf_loop();
- suppose a checkpoint is created each time at function entry;
- the ctx->ptr is tracked by verifier as PTR_TO_PACKET;
- the ctx->ptr is incremented in tcp_parse_option(),
thus umax_value field tracked for it is incremented as well;
- on each next iteration of tcp_parse_option()
checkpoint from a previous iteration can't be reused
for state pruning, because PTR_TO_PACKET registers are
considered equivalent only if old->umax_value >= cur->umax_value;
- on the other hand, the ctx->off is a SCALAR,
subject to widen_imprecise_scalars();
- it's exact bounds are eventually forgotten and it is tracked as
unknown scalar at entry to tcp_parse_option();
- hence checkpoints created at the start of the function eventually
converge.
The change is similar to one applied in [0] to xdp_synproxy_kern.c.
Comparing before and after with veristat yields following results:
File Insns (A) Insns (B) Insns (DIFF)
------------------------------- --------- --------- -----------------
test_tcp_custom_syncookie.bpf.o 466657 12423 -454234 (-97.34%)
[0] commit 977bc146d4eb ("selftests/bpf: track tcp payload offset as scalar in xdp_synproxy")
Acked-by: Yonghong Song <yonghong.song@linux.dev>
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/r/20240222150300.14909-2-eddyz87@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
Back in 2018 the commit be95a845cc44 ("bpf: avoid false sharing of map refcount
with max_entries") added ____cacheline_aligned to "struct bpf_map" to make sure
that fields like refcnt don't share a cache line with max_entries that is used
to bounds check map access. That was done to make spectre style attacks harder.
The main mitigation is done via code similar to array_index_nospec(), of course.
This was an additional precaution.
It increased the size of "struct bpf_map" a little, but it's affect on all
other maps (like array) is significant, since "struct bpf_map" is typically
the first member in other map types.
Undo this ____cacheline_aligned tag. Instead move freeze_mutex field around, so
that refcnt and max_entries are still in different cache lines.
The main effect is seen in sizeof(struct bpf_array) that reduces from 320
to 248 bytes.
BEFORE:
struct bpf_map {
const struct bpf_map_ops * ops; /* 0 8 */
...
char name[16]; /* 96 16 */
/* XXX 16 bytes hole, try to pack */
/* --- cacheline 2 boundary (128 bytes) --- */
atomic64_t refcnt __attribute__((__aligned__(64))); /* 128 8 */
...
/* size: 256, cachelines: 4, members: 30 */
/* sum members: 232, holes: 1, sum holes: 16 */
/* padding: 8 */
/* paddings: 1, sum paddings: 2 */
} __attribute__((__aligned__(64)));
struct bpf_array {
struct bpf_map map; /* 0 256 */
...
/* size: 320, cachelines: 5, members: 5 */
/* padding: 48 */
/* paddings: 1, sum paddings: 8 */
} __attribute__((__aligned__(64)));
AFTER:
struct bpf_map {
/* size: 232, cachelines: 4, members: 30 */
/* paddings: 1, sum paddings: 2 */
/* last cacheline: 40 bytes */
};
struct bpf_array {
/* size: 248, cachelines: 4, members: 5 */
/* last cacheline: 56 bytes */
};
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Yonghong Song <yonghong.song@linux.dev>
Link: https://lore.kernel.org/bpf/20240220235001.57411-1-alexei.starovoitov@gmail.com
|
|
The test of linking process creates several intermediate files.
Remove them once the build is over.
This reduces the number of files in selftests/bpf/ directory
from ~4400 to ~2600.
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20240220231102.49090-1-alexei.starovoitov@gmail.com
|
|
Commit f04a32b2c5b5 ("selftests/bpf: Do not use sign-file as testcase")
removed the TEST_CUSTOM_PROGS assignment, and removed it from being used
on TEST_GEN_FILES. Remove two leftovers from that cleanup. Found by
inspection.
Signed-off-by: Marcos Paulo de Souza <mpdesouza@suse.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Cc: Alexey Gladkov <legion@kernel.org>
Link: https://lore.kernel.org/bpf/20240216-bpf-selftests-custom-progs-v1-1-f7cf281a1fda@suse.com
|
|
With latest llvm19, I hit the following selftest failures with
$ ./test_progs -j
libbpf: prog 'on_event': BPF program load failed: Permission denied
libbpf: prog 'on_event': -- BEGIN PROG LOAD LOG --
combined stack size of 4 calls is 544. Too large
verification time 1344153 usec
stack depth 24+440+0+32
processed 51008 insns (limit 1000000) max_states_per_insn 19 total_states 1467 peak_states 303 mark_read 146
-- END PROG LOAD LOG --
libbpf: prog 'on_event': failed to load: -13
libbpf: failed to load object 'strobemeta_subprogs.bpf.o'
scale_test:FAIL:expect_success unexpected error: -13 (errno 13)
#498 verif_scale_strobemeta_subprogs:FAIL
The verifier complains too big of the combined stack size (544 bytes) which
exceeds the maximum stack limit 512. This is a regression from llvm19 ([1]).
In the above error log, the original stack depth is 24+440+0+32.
To satisfy interpreter's need, in verifier the stack depth is adjusted to
32+448+32+32=544 which exceeds 512, hence the error. The same adjusted
stack size is also used for jit case.
But the jitted codes could use smaller stack size.
$ egrep -r stack_depth | grep round_up
arm64/net/bpf_jit_comp.c: ctx->stack_size = round_up(prog->aux->stack_depth, 16);
loongarch/net/bpf_jit.c: bpf_stack_adjust = round_up(ctx->prog->aux->stack_depth, 16);
powerpc/net/bpf_jit_comp.c: cgctx.stack_size = round_up(fp->aux->stack_depth, 16);
riscv/net/bpf_jit_comp32.c: round_up(ctx->prog->aux->stack_depth, STACK_ALIGN);
riscv/net/bpf_jit_comp64.c: bpf_stack_adjust = round_up(ctx->prog->aux->stack_depth, 16);
s390/net/bpf_jit_comp.c: u32 stack_depth = round_up(fp->aux->stack_depth, 8);
sparc/net/bpf_jit_comp_64.c: stack_needed += round_up(stack_depth, 16);
x86/net/bpf_jit_comp.c: EMIT3_off32(0x48, 0x81, 0xEC, round_up(stack_depth, 8));
x86/net/bpf_jit_comp.c: int tcc_off = -4 - round_up(stack_depth, 8);
x86/net/bpf_jit_comp.c: round_up(stack_depth, 8));
x86/net/bpf_jit_comp.c: int tcc_off = -4 - round_up(stack_depth, 8);
x86/net/bpf_jit_comp.c: EMIT3_off32(0x48, 0x81, 0xC4, round_up(stack_depth, 8));
In the above, STACK_ALIGN in riscv/net/bpf_jit_comp32.c is defined as 16.
So stack is aligned in either 8 or 16, x86/s390 having 8-byte stack alignment and
the rest having 16-byte alignment.
This patch calculates total stack depth based on 16-byte alignment if jit is requested.
For the above failing case, the new stack size will be 32+448+0+32=512 and no verification
failure. llvm19 regression will be discussed separately in llvm upstream.
The verifier change caused three test failures as these tests compared messages
with stack size. More specifically,
- test_global_funcs/global_func1: fail with interpreter mode and success with jit mode.
Adjusted stack sizes so both jit and interpreter modes will fail.
- async_stack_depth/{pseudo_call_check, async_call_root_check}: since jit and interpreter
will calculate different stack sizes, the failure msg is adjusted to omit those
specific stack size numbers.
[1] https://lore.kernel.org/bpf/32bde0f0-1881-46c9-931a-673be566c61d@linux.dev/
Suggested-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Yonghong Song <yonghong.song@linux.dev>
Link: https://lore.kernel.org/r/20240214232951.4113094-1-yonghong.song@linux.dev
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
Verifier log avoids printing the same source code line multiple times
when a consecutive block of BPF assembly instructions are covered by the
same original (C) source code line. This greatly improves verifier log
legibility.
Unfortunately, this check is imperfect and in production applications it
quite often happens that verifier log will have multiple duplicated
source lines emitted, for no apparently good reason. E.g., this is
excerpt from a real-world BPF application (with register states omitted
for clarity):
BEFORE
======
; for (int i = 0; i < STROBE_MAX_MAP_ENTRIES; ++i) { @ strobemeta_probe.bpf.c:394
5369: (07) r8 += 2 ;
5370: (07) r7 += 16 ;
; for (int i = 0; i < STROBE_MAX_MAP_ENTRIES; ++i) { @ strobemeta_probe.bpf.c:394
5371: (07) r9 += 1 ;
5372: (79) r4 = *(u64 *)(r10 -32) ;
; for (int i = 0; i < STROBE_MAX_MAP_ENTRIES; ++i) { @ strobemeta_probe.bpf.c:394
5373: (55) if r9 != 0xf goto pc+2
; if (i >= map->cnt) @ strobemeta_probe.bpf.c:396
5376: (79) r1 = *(u64 *)(r10 -40) ;
5377: (79) r1 = *(u64 *)(r1 +8) ;
; if (i >= map->cnt) @ strobemeta_probe.bpf.c:396
5378: (dd) if r1 s<= r9 goto pc-5 ;
; descr->key_lens[i] = 0; @ strobemeta_probe.bpf.c:398
5379: (b4) w1 = 0 ;
5380: (6b) *(u16 *)(r8 -30) = r1 ;
; task, data, off, STROBE_MAX_STR_LEN, map->entries[i].key); @ strobemeta_probe.bpf.c:400
5381: (79) r3 = *(u64 *)(r7 -8) ;
5382: (7b) *(u64 *)(r10 -24) = r6 ;
; task, data, off, STROBE_MAX_STR_LEN, map->entries[i].key); @ strobemeta_probe.bpf.c:400
5383: (bc) w6 = w6 ;
; barrier_var(payload_off); @ strobemeta_probe.bpf.c:280
5384: (bf) r2 = r6 ;
5385: (bf) r1 = r4 ;
As can be seen, line 394 is emitted thrice, 396 is emitted twice, and
line 400 is duplicated as well. Note that there are no intermingling
other lines of source code in between these duplicates, so the issue is
not compiler reordering assembly instruction such that multiple original
source code lines are in effect.
It becomes more obvious what's going on if we look at *full* original line info
information (using btfdump for this, [0]):
#2764: line: insn #5363 --> 394:3 @ ./././strobemeta_probe.bpf.c
for (int i = 0; i < STROBE_MAX_MAP_ENTRIES; ++i) {
#2765: line: insn #5373 --> 394:21 @ ./././strobemeta_probe.bpf.c
for (int i = 0; i < STROBE_MAX_MAP_ENTRIES; ++i) {
#2766: line: insn #5375 --> 394:47 @ ./././strobemeta_probe.bpf.c
for (int i = 0; i < STROBE_MAX_MAP_ENTRIES; ++i) {
#2767: line: insn #5377 --> 394:3 @ ./././strobemeta_probe.bpf.c
for (int i = 0; i < STROBE_MAX_MAP_ENTRIES; ++i) {
#2768: line: insn #5378 --> 414:10 @ ./././strobemeta_probe.bpf.c
return off;
We can see that there are four line info records covering
instructions #5363 through #5377 (instruction indices are shifted due to
subprog instruction being appended to main program), all of them are
pointing to the same C source code line #394. But each of them points to
a different part of that line, which is denoted by differing column
numbers (3, 21, 47, 3).
But verifier log doesn't distinguish between parts of the same source code line
and doesn't emit this column number information, so for end user it's just a
repetitive visual noise. So let's improve the detection of repeated source code
line and avoid this.
With the changes in this patch, we get this output for the same piece of BPF
program log:
AFTER
=====
; for (int i = 0; i < STROBE_MAX_MAP_ENTRIES; ++i) { @ strobemeta_probe.bpf.c:394
5369: (07) r8 += 2 ;
5370: (07) r7 += 16 ;
5371: (07) r9 += 1 ;
5372: (79) r4 = *(u64 *)(r10 -32) ;
5373: (55) if r9 != 0xf goto pc+2
; if (i >= map->cnt) @ strobemeta_probe.bpf.c:396
5376: (79) r1 = *(u64 *)(r10 -40) ;
5377: (79) r1 = *(u64 *)(r1 +8) ;
5378: (dd) if r1 s<= r9 goto pc-5 ;
; descr->key_lens[i] = 0; @ strobemeta_probe.bpf.c:398
5379: (b4) w1 = 0 ;
5380: (6b) *(u16 *)(r8 -30) = r1 ;
; task, data, off, STROBE_MAX_STR_LEN, map->entries[i].key); @ strobemeta_probe.bpf.c:400
5381: (79) r3 = *(u64 *)(r7 -8) ;
5382: (7b) *(u64 *)(r10 -24) = r6 ;
5383: (bc) w6 = w6 ;
; barrier_var(payload_off); @ strobemeta_probe.bpf.c:280
5384: (bf) r2 = r6 ;
5385: (bf) r1 = r4 ;
All the duplication is gone and the log is cleaner and less distracting.
[0] https://github.com/anakryiko/btfdump
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/r/20240214174100.2847419-1-andrii@kernel.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
Real-world BPF applications keep growing in size. Medium-sized production
application can easily have 50K+ verified instructions, and its line
info section in .BTF.ext has more than 3K entries.
When verifier emits log with log_level>=1, it annotates assembly code
with matched original C source code. Currently it uses linear search
over line info records to find a match. As complexity of BPF
applications grows, this O(K * N) approach scales poorly.
So, let's instead of linear O(N) search for line info record use faster
equivalent O(log(N)) binary search algorithm. It's not a plain binary
search, as we don't look for exact match. It's an upper bound search
variant, looking for rightmost line info record that starts at or before
given insn_off.
Some unscientific measurements were done before and after this change.
They were done in VM and fluctuate a bit, but overall the speed up is
undeniable.
BASELINE
========
File Program Duration (us) Insns
-------------------------------- ---------------- ------------- ------
katran.bpf.o balancer_ingress 2497130 343552
pyperf600.bpf.linked3.o on_event 12389611 627288
strobelight_pyperf_libbpf.o on_py_event 387399 52445
-------------------------------- ---------------- ------------- ------
BINARY SEARCH
=============
File Program Duration (us) Insns
-------------------------------- ---------------- ------------- ------
katran.bpf.o balancer_ingress 2339312 343552
pyperf600.bpf.linked3.o on_event 5602203 627288
strobelight_pyperf_libbpf.o on_py_event 294761 52445
-------------------------------- ---------------- ------------- ------
While Katran's speed up is pretty modest (about 105ms, or 6%), for
production pyperf BPF program (on_py_event) it's much greater already,
going from 387ms down to 295ms (23% improvement).
Looking at BPF selftests's biggest pyperf example, we can see even more
dramatic improvement, shaving more than 50% of time, going from 12.3s
down to 5.6s.
Different amount of improvement is the function of overall amount of BPF
assembly instructions in .bpf.o files (which contributes to how much
line info records there will be and thus, on average, how much time linear
search will take), among other things:
$ llvm-objdump -d katran.bpf.o | wc -l
3863
$ llvm-objdump -d strobelight_pyperf_libbpf.o | wc -l
6997
$ llvm-objdump -d pyperf600.bpf.linked3.o | wc -l
87854
Granted, this only applies to debugging cases (e.g., using veristat, or
failing verification in production), but seems worth doing to improve
overall developer experience anyways.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: https://lore.kernel.org/bpf/20240214002311.2197116-1-andrii@kernel.org
|
|
In some situations, if you fail to zero-initialize the
bpf_{prog,map,btf,link}_info structs supplied to the set of LIBBPF
helpers bpf_{prog,map,btf,link}_get_info_by_fd(), you can expect the
helper to return an error. This can possibly leave people in a
situation where they're scratching their heads for an unnnecessary
amount of time. Make an explicit remark about the requirement of
zero-initializing the supplied bpf_{prog,map,btf,link}_info structs
for the respective LIBBPF helpers.
Internally, LIBBPF helpers bpf_{prog,map,btf,link}_get_info_by_fd()
call into bpf_obj_get_info_by_fd() where the bpf(2)
BPF_OBJ_GET_INFO_BY_FD command is used. This specific command is
effectively backed by restrictions enforced by the
bpf_check_uarg_tail_zero() helper. This function ensures that if the
size of the supplied bpf_{prog,map,btf,link}_info structs are larger
than what the kernel can handle, trailing bits are zeroed. This can be
a problem when compiling against UAPI headers that don't necessarily
match the sizes of the same underlying types known to the kernel.
Signed-off-by: Matt Bobrowski <mattbobrowski@google.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: https://lore.kernel.org/bpf/ZcyEb8x4VbhieWsL@google.com
|
|
As BPF applications grow in size and complexity and are separated into
multiple .bpf.c files that are statically linked together, it becomes
harder and harder to match verifier's BPF assembly level output to
original C code. While often annotated C source code is unique enough to
be able to identify the file it belongs to, quite often this is actually
problematic as parts of source code can be quite generic.
Long story short, it is very useful to see source code file name and
line number information along with the original C code. Verifier already
knows this information, we just need to output it.
This patch extends verifier log with file name and line number
information, emitted next to original (presumably C) source code,
annotating BPF assembly output, like so:
; <original C code> @ <filename>.bpf.c:<line>
If file name has directory names in it, they are stripped away. This
should be fine in practice as file names tend to be pretty unique with
C code anyways, and keeping log size smaller is always good.
In practice this might look something like below, where some code is
coming from application files, while others are from libbpf's usdt.bpf.h
header file:
; if (STROBEMETA_READ( @ strobemeta_probe.bpf.c:534
5592: (79) r1 = *(u64 *)(r10 -56) ; R1_w=mem_or_null(id=1589,sz=7680) R10=fp0
5593: (7b) *(u64 *)(r10 -56) = r1 ; R1_w=mem_or_null(id=1589,sz=7680) R10=fp0
5594: (79) r3 = *(u64 *)(r10 -8) ; R3_w=scalar() R10=fp0 fp-8=mmmmmmmm
...
170: (71) r1 = *(u8 *)(r8 +15) ; frame1: R1_w=scalar(...) R8_w=map_value(map=__bpf_usdt_spec,ks=4,vs=208)
171: (67) r1 <<= 56 ; frame1: R1_w=scalar(...)
172: (c7) r1 s>>= 56 ; frame1: R1_w=scalar(smin=smin32=-128,smax=smax32=127)
; val <<= arg_spec->arg_bitshift; @ usdt.bpf.h:183
173: (67) r1 <<= 32 ; frame1: R1_w=scalar(...)
174: (77) r1 >>= 32 ; frame1: R1_w=scalar(smin=0,smax=umax=0xffffffff,var_off=(0x0; 0xffffffff))
175: (79) r2 = *(u64 *)(r10 -8) ; frame1: R2_w=scalar() R10=fp0 fp-8=mmmmmmmm
176: (6f) r2 <<= r1 ; frame1: R1_w=scalar(smin=0,smax=umax=0xffffffff,var_off=(0x0; 0xffffffff)) R2_w=scalar()
177: (7b) *(u64 *)(r10 -8) = r2 ; frame1: R2_w=scalar(id=61) R10=fp0 fp-8_w=scalar(id=61)
; if (arg_spec->arg_signed) @ usdt.bpf.h:184
178: (bf) r3 = r2 ; frame1: R2_w=scalar(id=61) R3_w=scalar(id=61)
179: (7f) r3 >>= r1 ; frame1: R1_w=scalar(smin=0,smax=umax=0xffffffff,var_off=(0x0; 0xffffffff)) R3_w=scalar()
; if (arg_spec->arg_signed) @ usdt.bpf.h:184
180: (71) r4 = *(u8 *)(r8 +14)
181: safe
log_fixup tests needed a minor adjustment as verifier log output
increased a bit and that test is quite sensitive to such changes.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/r/20240212235944.2816107-1-andrii@kernel.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
Andrii Nakryiko says:
====================
Fix global subprog PTR_TO_CTX arg handling
Fix confusing and incorrect inference of PTR_TO_CTX argument type in BPF
global subprogs. For some program types (iters, tracepoint, any program type
that doesn't have fixed named "canonical" context type) when user uses (in
a correct and valid way) a pointer argument to user-defined anonymous struct
type, verifier will incorrectly assume that it has to be PTR_TO_CTX argument.
While it should be just a PTR_TO_MEM argument with allowed size calculated
from user-provided (even if anonymous) struct.
This did come up in practice and was very confusing to users, so let's prevent
this going forward. We had to do a slight refactoring of
btf_get_prog_ctx_type() to make it easy to support a special s390x KPROBE use
cases. See details in respective patches.
v1->v2:
- special-case typedef bpf_user_pt_regs_t handling for KPROBE programs,
fixing s390x after changes in patch #2.
====================
Link: https://lore.kernel.org/r/20240212233221.2575350-1-andrii@kernel.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
Add tests validating that kernel handles pointer to anonymous struct
argument as PTR_TO_MEM case, not as PTR_TO_CTX case.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/r/20240212233221.2575350-5-andrii@kernel.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
For program types that don't have named context type name (e.g., BPF
iterator programs or tracepoint programs), ctx_tname will be a non-NULL
empty string. For such programs it shouldn't be possible to have
PTR_TO_CTX argument for global subprogs based on type name alone.
arg:ctx tag is the only way to have PTR_TO_CTX passed into global
subprog for such program types.
Fix this loophole, which currently would assume PTR_TO_CTX whenever
user uses a pointer to anonymous struct as an argument to their global
subprogs. This happens in practice with the following (quite common, in
practice) approach:
typedef struct { /* anonymous */
int x;
} my_type_t;
int my_subprog(my_type_t *arg) { ... }
User's intent is to have PTR_TO_MEM argument for `arg`, but verifier
will complain about expecting PTR_TO_CTX.
This fix also closes unintended s390x-specific KPROBE handling of
PTR_TO_CTX case. Selftest change is necessary to accommodate this.
Fixes: 91cc1a99740e ("bpf: Annotate context types")
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/r/20240212233221.2575350-4-andrii@kernel.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
Expected canonical argument type for global function arguments
representing PTR_TO_CTX is `bpf_user_pt_regs_t *ctx`. This currently
works on s390x by accident because kernel resolves such typedef to
underlying struct (which is anonymous on s390x), and erroneously
accepting it as expected context type. We are fixing this problem next,
which would break s390x arch, so we need to handle `bpf_user_pt_regs_t`
case explicitly for KPROBE programs.
Fixes: 91cc1a99740e ("bpf: Annotate context types")
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/r/20240212233221.2575350-3-andrii@kernel.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
Return result of btf_get_prog_ctx_type() is never used and callers only
check NULL vs non-NULL case to determine if given type matches expected
PTR_TO_CTX type. So rename function to `btf_is_prog_ctx_type()` and
return a simple true/false. We'll use this simpler interface to handle
kprobe program type's special typedef case in the next patch.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/r/20240212233221.2575350-2-andrii@kernel.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
Originally, this patch removed a redundant check in
BPF_CGROUP_RUN_PROG_INET_EGRESS, as the check was already being done in
the function it called, __cgroup_bpf_run_filter_skb. For v2, it was
reccomended that I remove the check from __cgroup_bpf_run_filter_skb,
and add the checks to the other macro that calls that function,
BPF_CGROUP_RUN_PROG_INET_INGRESS.
To sum it up, checking that the socket exists and that it is a full
socket is now part of both macros BPF_CGROUP_RUN_PROG_INET_EGRESS and
BPF_CGROUP_RUN_PROG_INET_INGRESS, and it is no longer part of the
function they call, __cgroup_bpf_run_filter_skb.
v3->v4: Fixed weird merge conflict.
v2->v3: Sent to bpf-next instead of generic patch
v1->v2: Addressed feedback about where check should be removed.
Signed-off-by: Oliver Crumrine <ozlinuxc@gmail.com>
Acked-by: Stanislav Fomichev <sdf@google.com>
Link: https://lore.kernel.org/r/7lv62yiyvmj5a7eozv2iznglpkydkdfancgmbhiptrgvgan5sy@3fl3onchgdz3
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
|
|
Kui-Feng Lee says:
====================
Allow passing null pointers to the operators provided by a struct_ops
object. This is an RFC to collect feedbacks/opinions.
The function pointers that are passed to struct_ops operators (the function
pointers) are always considered reliable until now. They cannot be
null. However, in certain scenarios, it should be possible to pass null
pointers to these operators. For instance, sched_ext may pass a null
pointer in the struct task type to an operator that is provided by its
struct_ops objects.
The proposed solution here is to add PTR_MAYBE_NULL annotations to
arguments and create instances of struct bpf_ctx_arg_aux (arg_info) for
these arguments. These arg_infos will be installed at
prog->aux->ctx_arg_info and will be checked by the BPF verifier when
loading the programs. When a struct_ops program accesses arguments in the
ctx, the verifier will call btf_ctx_access() (through
bpf_verifier_ops->is_valid_access) to verify the access. btf_ctx_access()
will check arg_info and use the information of the matched arg_info to
properly set reg_type.
For nullable arguments, this patch sets an arg_info to label them with
PTR_TO_BTF_ID | PTR_TRUSTED | PTR_MAYBE_NULL. This enforces the verifier to
check programs and ensure that they properly check the pointer. The
programs should check if the pointer is null before reading/writing the
pointed memory.
The implementer of a struct_ops should annotate the arguments that can
be null. The implementer should define a stub function (empty) as a
placeholder for each defined operator. The name of a stub function
should be in the pattern "<st_op_type>__<operator name>". For example,
for test_maybe_null of struct bpf_testmod_ops, it's stub function name
should be "bpf_testmod_ops__test_maybe_null". You mark an argument
nullable by suffixing the argument name with "__nullable" at the stub
function. Here is the example in bpf_testmod.c.
static int bpf_testmod_ops__test_maybe_null(int dummy,
struct task_struct *task__nullable)
{
return 0;
}
This means that the argument 1 (2nd) of bpf_testmod_ops->test_maybe_null,
which is a function pointer that can be null. With this annotation, the
verifier will understand how to check programs using this arguments. A BPF
program that implement test_maybe_null should check the pointer to make
sure it is not null before using it. For example,
if (task__nullable)
save_tgid = task__nullable->tgid
Without the check, the verifier will reject the program.
Since we already has stub functions for kCFI, we just reuse these stub
functions with the naming convention mentioned earlier. These stub
functions with the naming convention is only required if there are nullable
arguments to annotate. For functions without nullable arguments, stub
functions are not necessary for the purpose of this patch.
---
Major changes from v7:
- Update a comment that is out of date.
Major changes from v6:
- Remove "len" from bpf_struct_ops_desc_release().
- Rename arg_info(s) to info, and rename all_arg_info to arg_info in
prepare_arg_info().
- Rename arg_info to info in struct bpf_struct_ops_arg_info.
Major changes from v5:
- Rename all member_arg_info variables.
- Refactor to bpf_struct_ops_desc_release() to share code
between btf_free_struct_ops_tab() and bpf_struct_ops_desc_init().
- Refactor to btf_param_match_suffix(). (Add a new patch as the part 2.)
- Clean up the commit log and remaining code in the patch of test cases.
- Update a comment in struct_ops_maybe_null.c.
Major changes from v4:
- Remove the support of pointers to types other than struct
types. That would be a separate patchset.
- Remove the patch about extending PTR_TO_BTF_ID.
- Remove the test against various pointer types from selftests.
- Remove the patch "bpf: Remove an unnecessary check" and send that
patch separately.
- Remove member_arg_info_cnt from struct bpf_struct_ops_desc.
- Use btf_id from FUNC_PROTO of a function pointer instead of a stub
function.
Major changes from v3:
- Move the code collecting argument information to prepare_arg_info()
called in the loop in bpf_struct_ops_desc_init().
- Simplify the memory allocation by having separated arg_info for
each member of a struct_ops type.
- Extend PTR_TO_BTF_ID to pointers to scalar types and array types,
not only to struct types.
Major changes from v2:
- Remove dead code.
- Add comments to explain the code itself.
Major changes from v1:
- Annotate arguments by suffixing argument names with "__nullable" at
stub functions.
v7: https://lore.kernel.org/all/20240209020053.1132710-1-thinker.li@gmail.com/
v6: https://lore.kernel.org/all/20240208065103.2154768-1-thinker.li@gmail.com/
v5: https://lore.kernel.org/all/20240206063833.2520479-1-thinker.li@gmail.com/
v4: https://lore.kernel.org/all/20240202220516.1165466-1-thinker.li@gmail.com/
v3: https://lore.kernel.org/all/20240122212217.1391878-1-thinker.li@gmail.com/
v2: https://lore.kernel.org/all/20240118224922.336006-1-thinker.li@gmail.com/
====================
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
|
|
Test if the verifier verifies nullable pointer arguments correctly for BPF
struct_ops programs.
"test_maybe_null" in struct bpf_testmod_ops is the operator defined for the
test cases here.
A BPF program should check a pointer for NULL beforehand to access the
value pointed by the nullable pointer arguments, or the verifier should
reject the programs. The test here includes two parts; the programs
checking pointers properly and the programs not checking pointers
beforehand. The test checks if the verifier accepts the programs checking
properly and rejects the programs not checking at all.
Signed-off-by: Kui-Feng Lee <thinker.li@gmail.com>
Link: https://lore.kernel.org/r/20240209023750.1153905-5-thinker.li@gmail.com
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
|
|
Collect argument information from the type information of stub functions to
mark arguments of BPF struct_ops programs with PTR_MAYBE_NULL if they are
nullable. A nullable argument is annotated by suffixing "__nullable" at
the argument name of stub function.
For nullable arguments, this patch sets a struct bpf_ctx_arg_aux to label
their reg_type with PTR_TO_BTF_ID | PTR_TRUSTED | PTR_MAYBE_NULL. This
makes the verifier to check programs and ensure that they properly check
the pointer. The programs should check if the pointer is null before
accessing the pointed memory.
The implementer of a struct_ops type should annotate the arguments that can
be null. The implementer should define a stub function (empty) as a
placeholder for each defined operator. The name of a stub function should
be in the pattern "<st_op_type>__<operator name>". For example, for
test_maybe_null of struct bpf_testmod_ops, it's stub function name should
be "bpf_testmod_ops__test_maybe_null". You mark an argument nullable by
suffixing the argument name with "__nullable" at the stub function.
Since we already has stub functions for kCFI, we just reuse these stub
functions with the naming convention mentioned earlier. These stub
functions with the naming convention is only required if there are nullable
arguments to annotate. For functions having not nullable arguments, stub
functions are not necessary for the purpose of this patch.
This patch will prepare a list of struct bpf_ctx_arg_aux, aka arg_info, for
each member field of a struct_ops type. "arg_info" will be assigned to
"prog->aux->ctx_arg_info" of BPF struct_ops programs in
check_struct_ops_btf_id() so that it can be used by btf_ctx_access() later
to set reg_type properly for the verifier.
Signed-off-by: Kui-Feng Lee <thinker.li@gmail.com>
Link: https://lore.kernel.org/r/20240209023750.1153905-4-thinker.li@gmail.com
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
|
|
Move __kfunc_param_match_suffix() to btf.c and rename it as
btf_param_match_suffix(). It can be reused by bpf_struct_ops later.
Signed-off-by: Kui-Feng Lee <thinker.li@gmail.com>
Link: https://lore.kernel.org/r/20240209023750.1153905-3-thinker.li@gmail.com
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
|
|
Enable the providers to use types defined in a module instead of in the
kernel (btf_vmlinux).
Signed-off-by: Kui-Feng Lee <thinker.li@gmail.com>
Link: https://lore.kernel.org/r/20240209023750.1153905-2-thinker.li@gmail.com
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
|
|
* Use "Instruction Set Architecture (ISA)" instead of "Instruction Set
Specification"
* Remove version number
As previously discussed on the mailing list at
https://mailarchive.ietf.org/arch/msg/bpf/SEpn3OL9TabNRn-4rDX9A6XVbjM/
Signed-off-by: Dave Thaler <dthaler1968@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: David Vernet <void@manifault.com>
Link: https://lore.kernel.org/bpf/20240208221449.12274-1-dthaler1968@gmail.com
|
|
Due to internal differences between LLVM and GCC the current
implementation for the CO-RE macros does not fit GCC parser, as it will
optimize those expressions even before those would be accessible by the
BPF backend.
As examples, the following would be optimized out with the original
definitions:
- As enums are converted to their integer representation during
parsing, the IR would not know how to distinguish an integer
constant from an actual enum value.
- Types need to be kept as temporary variables, as the existing type
casts of the 0 address (as expanded for LLVM), are optimized away by
the GCC C parser, never really reaching GCCs IR.
Although, the macros appear to add extra complexity, the expanded code
is removed from the compilation flow very early in the compilation
process, not really affecting the quality of the generated assembly.
Signed-off-by: Cupertino Miranda <cupertino.miranda@oracle.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20240213173543.1397708-1-cupertino.miranda@oracle.com
|
|
[Changes from V1:
- Avoid conflict by rebasing with latest master.]
Some BPF tests use loop unrolling compiler pragmas that are clang
specific and not supported by GCC. These pragmas, along with their
GCC equivalences are:
#pragma clang loop unroll_count(N)
#pragma GCC unroll N
#pragma clang loop unroll(full)
#pragma GCC unroll 65534
#pragma clang loop unroll(disable)
#pragma GCC unroll 1
#pragma unroll [aka #pragma clang loop unroll(enable)]
There is no GCC equivalence to this pragma. It enables unrolling on
loops that the compiler would not ordinarily unroll even with
-O2|-funroll-loops, but it is not equivalent to full unrolling
either.
This patch adds a new header progs/bpf_compiler.h that defines the
following macros, which correspond to each pair of compiler-specific
pragmas above:
__pragma_loop_unroll_count(N)
__pragma_loop_unroll_full
__pragma_loop_no_unroll
__pragma_loop_unroll
The selftests using loop unrolling pragmas are then changed to include
the header and use these macros in place of the explicit pragmas.
Tested in bpf-next master.
No regressions.
Signed-off-by: Jose E. Marchesi <jose.marchesi@oracle.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Yonghong Song <yonghong.song@linux.dev>
Link: https://lore.kernel.org/bpf/20240208203612.29611-1-jose.marchesi@oracle.com
|
|
Add two tests to ensure fentry programs cannot attach to
bpf_spin_{lock,unlock}() helpers. The tracing_failure.c files
can be used in the future for other tracing failure cases.
Signed-off-by: Yonghong Song <yonghong.song@linux.dev>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20240207070107.335341-1-yonghong.song@linux.dev
|
|
Currently tracing is supposed not to allow for bpf_spin_{lock,unlock}()
helper calls. This is to prevent deadlock for the following cases:
- there is a prog (prog-A) calling bpf_spin_{lock,unlock}().
- there is a tracing program (prog-B), e.g., fentry, attached
to bpf_spin_lock() and/or bpf_spin_unlock().
- prog-B calls bpf_spin_{lock,unlock}().
For such a case, when prog-A calls bpf_spin_{lock,unlock}(),
a deadlock will happen.
The related source codes are below in kernel/bpf/helpers.c:
notrace BPF_CALL_1(bpf_spin_lock, struct bpf_spin_lock *, lock)
notrace BPF_CALL_1(bpf_spin_unlock, struct bpf_spin_lock *, lock)
notrace is supposed to prevent fentry prog from attaching to
bpf_spin_{lock,unlock}().
But actually this is not the case and fentry prog can successfully
attached to bpf_spin_lock(). Siddharth Chintamaneni reported
the issue in [1]. The following is the macro definition for
above BPF_CALL_1:
#define BPF_CALL_x(x, name, ...) \
static __always_inline \
u64 ____##name(__BPF_MAP(x, __BPF_DECL_ARGS, __BPF_V, __VA_ARGS__)); \
typedef u64 (*btf_##name)(__BPF_MAP(x, __BPF_DECL_ARGS, __BPF_V, __VA_ARGS__)); \
u64 name(__BPF_REG(x, __BPF_DECL_REGS, __BPF_N, __VA_ARGS__)); \
u64 name(__BPF_REG(x, __BPF_DECL_REGS, __BPF_N, __VA_ARGS__)) \
{ \
return ((btf_##name)____##name)(__BPF_MAP(x,__BPF_CAST,__BPF_N,__VA_ARGS__));\
} \
static __always_inline \
u64 ____##name(__BPF_MAP(x, __BPF_DECL_ARGS, __BPF_V, __VA_ARGS__))
#define BPF_CALL_1(name, ...) BPF_CALL_x(1, name, __VA_ARGS__)
The notrace attribute is actually applied to the static always_inline function
____bpf_spin_{lock,unlock}(). The actual callback function
bpf_spin_{lock,unlock}() is not marked with notrace, hence
allowing fentry prog to attach to two helpers, and this
may cause the above mentioned deadlock. Siddharth Chintamaneni
actually has a reproducer in [2].
To fix the issue, a new macro NOTRACE_BPF_CALL_1 is introduced which
will add notrace attribute to the original function instead of
the hidden always_inline function and this fixed the problem.
[1] https://lore.kernel.org/bpf/CAE5sdEigPnoGrzN8WU7Tx-h-iFuMZgW06qp0KHWtpvoXxf1OAQ@mail.gmail.com/
[2] https://lore.kernel.org/bpf/CAE5sdEg6yUc_Jz50AnUXEEUh6O73yQ1Z6NV2srJnef0ZrQkZew@mail.gmail.com/
Fixes: d83525ca62cf ("bpf: introduce bpf_spin_lock")
Signed-off-by: Yonghong Song <yonghong.song@linux.dev>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: https://lore.kernel.org/bpf/20240207070102.335167-1-yonghong.song@linux.dev
|
|
Since 20d59ee55172 ("libbpf: add bpf_core_cast() macro"), libbpf is now
exporting a const arg version of bpf_rdonly_cast(). This causes the
following conflicting type error when generating kfunc prototypes from
BTF:
In file included from skeleton/pid_iter.bpf.c:5:
/home/dxu/dev/linux/tools/bpf/bpftool/bootstrap/libbpf/include/bpf/bpf_core_read.h:297:14: error: conflicting types for 'bpf_rdonly_cast'
extern void *bpf_rdonly_cast(const void *obj__ign, __u32 btf_id__k) __ksym __weak;
^
./vmlinux.h:135625:14: note: previous declaration is here
extern void *bpf_rdonly_cast(void *obj__ign, u32 btf_id__k) __weak __ksym;
This is b/c the kernel defines bpf_rdonly_cast() with non-const arg.
Since const arg is more permissive and thus backwards compatible, we
change the kernel definition as well to avoid conflicting type errors.
Signed-off-by: Daniel Xu <dxu@dxuuu.xyz>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Yonghong Song <yonghong.song@linux.dev>
Link: https://lore.kernel.org/bpf/dfd3823f11ffd2d4c838e961d61ec9ae8a646773.1707080349.git.dxu@dxuuu.xyz
|
|
In various performance profiles of kernels with BPF programs attached,
bpf_local_storage_lookup() appears as a significant portion of CPU
cycles spent. To enable the compiler generate more optimal code, turn
bpf_local_storage_lookup() into a static inline function, where only the
cache insertion code path is outlined
Notably, outlining cache insertion helps avoid bloating callers by
duplicating setting up calls to raw_spin_{lock,unlock}_irqsave() (on
architectures which do not inline spin_lock/unlock, such as x86), which
would cause the compiler produce worse code by deciding to outline
otherwise inlinable functions. The call overhead is neutral, because we
make 2 calls either way: either calling raw_spin_lock_irqsave() and
raw_spin_unlock_irqsave(); or call __bpf_local_storage_insert_cache(),
which calls raw_spin_lock_irqsave(), followed by a tail-call to
raw_spin_unlock_irqsave() where the compiler can perform TCO and (in
optimized uninstrumented builds) turns it into a plain jump. The call to
__bpf_local_storage_insert_cache() can be elided entirely if
cacheit_lockit is a false constant expression.
Based on results from './benchs/run_bench_local_storage.sh' (21 trials,
reboot between each trial; x86 defconfig + BPF, clang 16) this produces
improvements in throughput and latency in the majority of cases, with an
average (geomean) improvement of 8%:
+---- Hashmap Control --------------------
|
| + num keys: 10
| : <before> | <after>
| +-+ hashmap (control) sequential get +----------------------+----------------------
| +- hits throughput | 14.789 M ops/s | 14.745 M ops/s ( ~ )
| +- hits latency | 67.679 ns/op | 67.879 ns/op ( ~ )
| +- important_hits throughput | 14.789 M ops/s | 14.745 M ops/s ( ~ )
|
| + num keys: 1000
| : <before> | <after>
| +-+ hashmap (control) sequential get +----------------------+----------------------
| +- hits throughput | 12.233 M ops/s | 12.170 M ops/s ( ~ )
| +- hits latency | 81.754 ns/op | 82.185 ns/op ( ~ )
| +- important_hits throughput | 12.233 M ops/s | 12.170 M ops/s ( ~ )
|
| + num keys: 10000
| : <before> | <after>
| +-+ hashmap (control) sequential get +----------------------+----------------------
| +- hits throughput | 7.220 M ops/s | 7.204 M ops/s ( ~ )
| +- hits latency | 138.522 ns/op | 138.842 ns/op ( ~ )
| +- important_hits throughput | 7.220 M ops/s | 7.204 M ops/s ( ~ )
|
| + num keys: 100000
| : <before> | <after>
| +-+ hashmap (control) sequential get +----------------------+----------------------
| +- hits throughput | 5.061 M ops/s | 5.165 M ops/s (+2.1%)
| +- hits latency | 198.483 ns/op | 194.270 ns/op (-2.1%)
| +- important_hits throughput | 5.061 M ops/s | 5.165 M ops/s (+2.1%)
|
| + num keys: 4194304
| : <before> | <after>
| +-+ hashmap (control) sequential get +----------------------+----------------------
| +- hits throughput | 2.864 M ops/s | 2.882 M ops/s ( ~ )
| +- hits latency | 365.220 ns/op | 361.418 ns/op (-1.0%)
| +- important_hits throughput | 2.864 M ops/s | 2.882 M ops/s ( ~ )
|
+---- Local Storage ----------------------
|
| + num_maps: 1
| : <before> | <after>
| +-+ local_storage cache sequential get +----------------------+----------------------
| +- hits throughput | 33.005 M ops/s | 39.068 M ops/s (+18.4%)
| +- hits latency | 30.300 ns/op | 25.598 ns/op (-15.5%)
| +- important_hits throughput | 33.005 M ops/s | 39.068 M ops/s (+18.4%)
| :
| : <before> | <after>
| +-+ local_storage cache interleaved get +----------------------+----------------------
| +- hits throughput | 37.151 M ops/s | 44.926 M ops/s (+20.9%)
| +- hits latency | 26.919 ns/op | 22.259 ns/op (-17.3%)
| +- important_hits throughput | 37.151 M ops/s | 44.926 M ops/s (+20.9%)
|
| + num_maps: 10
| : <before> | <after>
| +-+ local_storage cache sequential get +----------------------+----------------------
| +- hits throughput | 32.288 M ops/s | 38.099 M ops/s (+18.0%)
| +- hits latency | 30.972 ns/op | 26.248 ns/op (-15.3%)
| +- important_hits throughput | 3.229 M ops/s | 3.810 M ops/s (+18.0%)
| :
| : <before> | <after>
| +-+ local_storage cache interleaved get +----------------------+----------------------
| +- hits throughput | 34.473 M ops/s | 41.145 M ops/s (+19.4%)
| +- hits latency | 29.010 ns/op | 24.307 ns/op (-16.2%)
| +- important_hits throughput | 12.312 M ops/s | 14.695 M ops/s (+19.4%)
|
| + num_maps: 16
| : <before> | <after>
| +-+ local_storage cache sequential get +----------------------+----------------------
| +- hits throughput | 32.524 M ops/s | 38.341 M ops/s (+17.9%)
| +- hits latency | 30.748 ns/op | 26.083 ns/op (-15.2%)
| +- important_hits throughput | 2.033 M ops/s | 2.396 M ops/s (+17.9%)
| :
| : <before> | <after>
| +-+ local_storage cache interleaved get +----------------------+----------------------
| +- hits throughput | 34.575 M ops/s | 41.338 M ops/s (+19.6%)
| +- hits latency | 28.925 ns/op | 24.193 ns/op (-16.4%)
| +- important_hits throughput | 11.001 M ops/s | 13.153 M ops/s (+19.6%)
|
| + num_maps: 17
| : <before> | <after>
| +-+ local_storage cache sequential get +----------------------+----------------------
| +- hits throughput | 28.861 M ops/s | 32.756 M ops/s (+13.5%)
| +- hits latency | 34.649 ns/op | 30.530 ns/op (-11.9%)
| +- important_hits throughput | 1.700 M ops/s | 1.929 M ops/s (+13.5%)
| :
| : <before> | <after>
| +-+ local_storage cache interleaved get +----------------------+----------------------
| +- hits throughput | 31.529 M ops/s | 36.110 M ops/s (+14.5%)
| +- hits latency | 31.719 ns/op | 27.697 ns/op (-12.7%)
| +- important_hits throughput | 9.598 M ops/s | 10.993 M ops/s (+14.5%)
|
| + num_maps: 24
| : <before> | <after>
| +-+ local_storage cache sequential get +----------------------+----------------------
| +- hits throughput | 18.602 M ops/s | 19.937 M ops/s (+7.2%)
| +- hits latency | 53.767 ns/op | 50.166 ns/op (-6.7%)
| +- important_hits throughput | 0.776 M ops/s | 0.831 M ops/s (+7.2%)
| :
| : <before> | <after>
| +-+ local_storage cache interleaved get +----------------------+----------------------
| +- hits throughput | 21.718 M ops/s | 23.332 M ops/s (+7.4%)
| +- hits latency | 46.047 ns/op | 42.865 ns/op (-6.9%)
| +- important_hits throughput | 6.110 M ops/s | 6.564 M ops/s (+7.4%)
|
| + num_maps: 32
| : <before> | <after>
| +-+ local_storage cache sequential get +----------------------+----------------------
| +- hits throughput | 14.118 M ops/s | 14.626 M ops/s (+3.6%)
| +- hits latency | 70.856 ns/op | 68.381 ns/op (-3.5%)
| +- important_hits throughput | 0.442 M ops/s | 0.458 M ops/s (+3.6%)
| :
| : <before> | <after>
| +-+ local_storage cache interleaved get +----------------------+----------------------
| +- hits throughput | 17.111 M ops/s | 17.906 M ops/s (+4.6%)
| +- hits latency | 58.451 ns/op | 55.865 ns/op (-4.4%)
| +- important_hits throughput | 4.776 M ops/s | 4.998 M ops/s (+4.6%)
|
| + num_maps: 100
| : <before> | <after>
| +-+ local_storage cache sequential get +----------------------+----------------------
| +- hits throughput | 5.281 M ops/s | 5.528 M ops/s (+4.7%)
| +- hits latency | 192.398 ns/op | 183.059 ns/op (-4.9%)
| +- important_hits throughput | 0.053 M ops/s | 0.055 M ops/s (+4.9%)
| :
| : <before> | <after>
| +-+ local_storage cache interleaved get +----------------------+----------------------
| +- hits throughput | 6.265 M ops/s | 6.498 M ops/s (+3.7%)
| +- hits latency | 161.436 ns/op | 152.877 ns/op (-5.3%)
| +- important_hits throughput | 1.636 M ops/s | 1.697 M ops/s (+3.7%)
|
| + num_maps: 1000
| : <before> | <after>
| +-+ local_storage cache sequential get +----------------------+----------------------
| +- hits throughput | 0.355 M ops/s | 0.354 M ops/s ( ~ )
| +- hits latency | 2826.538 ns/op | 2827.139 ns/op ( ~ )
| +- important_hits throughput | 0.000 M ops/s | 0.000 M ops/s ( ~ )
| :
| : <before> | <after>
| +-+ local_storage cache interleaved get +----------------------+----------------------
| +- hits throughput | 0.404 M ops/s | 0.403 M ops/s ( ~ )
| +- hits latency | 2481.190 ns/op | 2487.555 ns/op ( ~ )
| +- important_hits throughput | 0.102 M ops/s | 0.101 M ops/s ( ~ )
The on_lookup test in {cgrp,task}_ls_recursion.c is removed
because the bpf_local_storage_lookup is no longer traceable
and adding tracepoint will make the compiler generate worse
code: https://lore.kernel.org/bpf/ZcJmok64Xqv6l4ZS@elver.google.com/
Signed-off-by: Marco Elver <elver@google.com>
Cc: Martin KaFai Lau <martin.lau@linux.dev>
Acked-by: Yonghong Song <yonghong.song@linux.dev>
Link: https://lore.kernel.org/r/20240207122626.3508658-1-elver@google.com
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
|
|
Geliang Tang says:
====================
bpf: Add DEBUG_INFO_BTF checks for __register_bpf_struct_ops
This patch set avoids module loading failure when the module
trying to register a struct_ops and the module has its btf section
stripped. This will then work similarly as module kfunc registration in
commit 3de4d22cc9ac ("bpf, btf: Warn but return no error for NULL btf from __register_btf_kfunc_id_set()")
v5:
- drop CONFIG_MODULE_ALLOW_BTF_MISMATCH check as Martin suggested.
v4:
- add a new patch to fix error checks for btf_get_module_btf.
- rename the helper to check_btf_kconfigs.
v3:
- fix this build error:
kernel/bpf/btf.c:7750:11: error: incomplete definition of type 'struct module'
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202402040934.Fph0XeEo-lkp@intel.com/
v2:
- add register_check_missing_btf helper as Jiri suggested.
====================
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
|
|
Similar to the handling in the functions __register_btf_kfunc_id_set()
and register_btf_id_dtor_kfuncs(), this patch uses the newly added
helper check_btf_kconfigs() to handle module with its btf section
stripped.
While at it, the patch also adds the missed IS_ERR() check to fix the
commit f6be98d19985 ("bpf, net: switch to dynamic registration")
Fixes: f6be98d19985 ("bpf, net: switch to dynamic registration")
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
Link: https://lore.kernel.org/r/69082b9835463fe36f9e354bddf2d0a97df39c2b.1707373307.git.tanggeliang@kylinos.cn
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
|
|
This patch extracts duplicate code on error path when btf_get_module_btf()
returns NULL from the functions __register_btf_kfunc_id_set() and
register_btf_id_dtor_kfuncs() into a new helper named check_btf_kconfigs()
to check CONFIG_DEBUG_INFO_BTF and CONFIG_DEBUG_INFO_BTF_MODULES in it.
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
Acked-by: Jiri Olsa <jolsa@kernel.org>
Link: https://lore.kernel.org/r/fa5537fc55f1e4d0bfd686598c81b7ab9dbd82b7.1707373307.git.tanggeliang@kylinos.cn
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
|
|
The same as __register_btf_kfunc_id_set(), to let the modules with
stripped btf section loaded, this patch changes the return value of
register_btf_id_dtor_kfuncs() too from -ENOENT to 0 when btf is NULL.
Signed-off-by: Geliang Tang <tanggeliang@kylinos.cn>
Link: https://lore.kernel.org/r/eab65586d7fb0e72f2707d3747c7d4a5d60c823f.1707373307.git.tanggeliang@kylinos.cn
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
|
|
'config BPF' exists in both init/Kconfig and kernel/bpf/Kconfig.
Commit b24abcff918a ("bpf, kconfig: Add consolidated menu entry for bpf
with core options") added the second one to kernel/bpf/Kconfig instead
of moving the existing one.
Merge them together.
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Yonghong Song <yonghong.song@linux.dev>
Link: https://lore.kernel.org/bpf/20240204075634.32969-1-masahiroy@kernel.org
|
|
After the series "Annotate kfuncs in .BTF_ids section"[0], kfuncs can be
generated from bpftool. Let's mark the existing cpumask kfunc declarations
__weak so they don't conflict with definitions that will eventually come
from vmlinux.h.
[0]. https://lore.kernel.org/all/cover.1706491398.git.dxu@dxuuu.xyz
Suggested-by: Andrii Nakryiko <andrii.nakryiko@gmail.com>
Signed-off-by: Yafang Shao <laoar.shao@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Daniel Xu <dxu@dxuuu.xyz>
Link: https://lore.kernel.org/bpf/20240206081416.26242-5-laoar.shao@gmail.com
|
|
We should verify the return value of cpumask_success__load().
Signed-off-by: Yafang Shao <laoar.shao@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20240206081416.26242-4-laoar.shao@gmail.com
|
|
Viktor Malik says:
====================
tools/resolve_btfids: fix cross-compilation to non-host endianness
The .BTF_ids section is pre-filled with zeroed BTF ID entries during the
build and afterwards patched by resolve_btfids with correct values.
Since resolve_btfids always writes in host-native endianness, it relies
on libelf to do the translation when the target ELF is cross-compiled to
a different endianness (this was introduced in commit 61e8aeda9398
("bpf: Fix libelf endian handling in resolv_btfids")).
Unfortunately, the translation will corrupt the flags fields of SET8
entries because these were written during vmlinux compilation and are in
the correct endianness already. This will lead to numerous selftests
failures such as:
$ sudo ./test_verifier 502 502
#502/p sleepable fentry accept FAIL
Failed to load prog 'Invalid argument'!
bpf_fentry_test1 is not sleepable
verification time 34 usec
stack depth 0
processed 0 insns (limit 1000000) max_states_per_insn 0 total_states 0 peak_states 0 mark_read 0
Summary: 0 PASSED, 0 SKIPPED, 1 FAILED
Since it's not possible to instruct libelf to translate just certain
values, let's manually bswap the flags (both global and entry flags) in
resolve_btfids when needed, so that libelf then translates everything
correctly.
The first patch of the series refactors resolve_btfids by using types
from btf_ids.h instead of accessing the BTF ID data using magic offsets.
Acked-by: Jiri Olsa <jolsa@kernel.org>
---
Changes in v4:
- remove unnecessary vars and pointer casts (suggested by Daniel Xu)
Changes in v3:
- add byte swap of global 'flags' field in btf_id_set8 (suggested by
Jiri Olsa)
- cleaner refactoring of sets_patch (suggested by Jiri Olsa)
- add compile-time assertion that IDs are at the beginning of pairs
struct in btf_id_set8 (suggested by Daniel Borkmann)
Changes in v2:
- use type defs from btf_ids.h (suggested by Andrii Nakryiko)
====================
Link: https://lore.kernel.org/r/cover.1707223196.git.vmalik@redhat.com
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
|
|
The .BTF_ids section is pre-filled with zeroed BTF ID entries during the
build and afterwards patched by resolve_btfids with correct values.
Since resolve_btfids always writes in host-native endianness, it relies
on libelf to do the translation when the target ELF is cross-compiled to
a different endianness (this was introduced in commit 61e8aeda9398
("bpf: Fix libelf endian handling in resolv_btfids")).
Unfortunately, the translation will corrupt the flags fields of SET8
entries because these were written during vmlinux compilation and are in
the correct endianness already. This will lead to numerous selftests
failures such as:
$ sudo ./test_verifier 502 502
#502/p sleepable fentry accept FAIL
Failed to load prog 'Invalid argument'!
bpf_fentry_test1 is not sleepable
verification time 34 usec
stack depth 0
processed 0 insns (limit 1000000) max_states_per_insn 0 total_states 0 peak_states 0 mark_read 0
Summary: 0 PASSED, 0 SKIPPED, 1 FAILED
Since it's not possible to instruct libelf to translate just certain
values, let's manually bswap the flags (both global and entry flags) in
resolve_btfids when needed, so that libelf then translates everything
correctly.
Fixes: ef2c6f370a63 ("tools/resolve_btfids: Add support for 8-byte BTF sets")
Signed-off-by: Viktor Malik <vmalik@redhat.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/7b6bff690919555574ce0f13d2a5996cacf7bf69.1707223196.git.vmalik@redhat.com
|
|
Instead of using magic offsets to access BTF ID set data, leverage types
from btf_ids.h (btf_id_set and btf_id_set8) which define the actual
layout of the data. Thanks to this change, set sorting should also
continue working if the layout changes.
This requires to sync the definition of 'struct btf_id_set8' from
include/linux/btf_ids.h to tools/include/linux/btf_ids.h. We don't sync
the rest of the file at the moment, b/c that would require to also sync
multiple dependent headers and we don't need any other defs from
btf_ids.h.
Signed-off-by: Viktor Malik <vmalik@redhat.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Daniel Xu <dxu@dxuuu.xyz>
Link: https://lore.kernel.org/bpf/ff7f062ddf6a00815fda3087957c4ce667f50532.1707223196.git.vmalik@redhat.com
|
|
When the feature_flags and xdp_zc_max_segs fields were added to the libbpf
bpf_xdp_query_opts, the code writing them did not use the OPTS_SET() macro.
This causes libbpf to write to those fields unconditionally, which means
that programs compiled against an older version of libbpf (with a smaller
size of the bpf_xdp_query_opts struct) will have its stack corrupted by
libbpf writing out of bounds.
The patch adding the feature_flags field has an early bail out if the
feature_flags field is not part of the opts struct (via the OPTS_HAS)
macro, but the patch adding xdp_zc_max_segs does not. For consistency, this
fix just changes the assignments to both fields to use the OPTS_SET()
macro.
Fixes: 13ce2daa259a ("xsk: add new netlink attribute dedicated for ZC max frags")
Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20240206125922.1992815-1-toke@redhat.com
|
|
[Differences from V2:
- Remove conditionals in the source files pragmas, as the
pragma is supported by both GCC and clang.]
Both GCC and clang implement the -Wno-address-of-packed-member
warning, which is enabled by -Wall, that warns about taking the
address of a packed struct field when it can lead to an "unaligned"
address.
This triggers the following errors (-Werror) when building three
particular BPF selftests with GCC:
progs/test_cls_redirect.c
986 | if (ipv4_is_fragment((void *)&encap->ip)) {
progs/test_cls_redirect_dynptr.c
410 | pkt_ipv4_checksum((void *)&encap_gre->ip);
progs/test_cls_redirect.c
521 | pkt_ipv4_checksum((void *)&encap_gre->ip);
progs/test_tc_tunnel.c
232 | set_ipv4_csum((void *)&h_outer.ip);
These warnings do not signal any real problem in the tests as far as I
can see.
This patch adds pragmas to these test files that inhibit the
-Waddress-of-packed-member warning.
Tested in bpf-next master.
No regressions.
Signed-off-by: Jose E. Marchesi <jose.marchesi@oracle.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Yonghong Song <yonghong.song@linux.dev>
Link: https://lore.kernel.org/bpf/20240206102330.7113-1-jose.marchesi@oracle.com
|
|
* "imm32" should just be "imm"
* Add blank line to fix formatting error reported by Stephen Rothwell [0]
[0]: https://lore.kernel.org/bpf/20240206153301.4ead0bad@canb.auug.org.au/T/#u
Signed-off-by: Dave Thaler <dthaler1968@gmail.com>
Acked-by: David Vernet <void@manifault.com>
Link: https://lore.kernel.org/r/20240206045146.4965-1-dthaler1968@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
Mark dynptr kfuncs as __weak to allow
verifier_global_subprogs/arg_ctx_{perf,kprobe,raw_tp} subtests to be
loadable on old kernels. Because bpf_dynptr_from_xdp() kfunc is used
from arg_tag_dynptr BPF program in progs/verifier_global_subprogs.c
*and* is not marked as __weak, loading any subtest from
verifier_global_subprogs fails on old kernels that don't have
bpf_dynptr_from_xdp() kfunc defined. Even if arg_tag_dynptr program
itself is not loaded, libbpf bails out on non-weak reference to
bpf_dynptr_from_xdp (that can't be resolved), which shared across all
programs in progs/verifier_global_subprogs.c.
So mark all dynptr-related kfuncs as __weak to unblock libbpf CI ([0]).
In the upcoming "kfunc in vmlinux.h" work we should make sure that
kfuncs are always declared __weak as well.
[0] https://github.com/libbpf/libbpf/actions/runs/7792673215/job/21251250831?pr=776#step:4:7961
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/r/20240206004008.1541513-1-andrii@kernel.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
If PERF_EVENT program has __arg_ctx argument with matching
architecture-specific pt_regs/user_pt_regs/user_regs_struct pointer
type, libbpf should still perform type rewrite for old kernels, but not
emit the warning. Fix copy/paste from kernel code where 0 is meant to
signify "no error" condition. For libbpf we need to return "true" to
proceed with type rewrite (which for PERF_EVENT program will be
a canonical `struct bpf_perf_event_data *` type).
Fixes: 9eea8fafe33e ("libbpf: fix __arg_ctx type enforcement for perf_event programs")
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/r/20240206002243.1439450-1-andrii@kernel.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
Magnus Karlsson says:
====================
xsk: support redirect to any socket bound to the same umem
This patch set adds support for directing a packet to any socket bound
to the same umem. This makes it possible to use the XDP program to
select what socket the packet should be received on. The user can
populate the XSKMAP with various sockets and as long as they share the
same umem, the XDP program can pick any one of them.
The implementation is straight-forward. Instead of testing that the
incoming packet is targeting the same device and queue id as the
socket is bound to, just check that the umem the packet was received
on is the same as the socket we want it to be received on. This
guarantees that the redirect is legal as it is already in the correct
umem.
Patch #1 implements the feature and patch #2 adds documentation.
Thanks: Magnus
====================
Link: https://lore.kernel.org/r/20240205123553.22180-1-magnus.karlsson@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
Document the ability to redirect to any socket bound to the same umem.
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Link: https://lore.kernel.org/r/20240205123553.22180-3-magnus.karlsson@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
Add support for directing a packet to any socket bound to the same
umem. This makes it possible to use the XDP program to select what
socket the packet should be received on. The user can populate the
XSKMAP with various sockets and as long as they share the same umem,
the XDP program can pick any one of them.
Suggested-by: Yuval El-Hanany <yuvale@radware.com>
Signed-off-by: Magnus Karlsson <magnus.karlsson@intel.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://lore.kernel.org/r/20240205123553.22180-2-magnus.karlsson@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
Kumar Kartikeya Dwivedi says:
====================
Transfer RCU lock state across subprog calls
David suggested during the discussion in [0] that we should handle RCU
locks in a similar fashion to spin locks where the verifier understands
when a lock held in a caller is released in callee, or lock taken in
callee is released in a caller, or the callee is called within a lock
critical section. This set extends the same semantics to RCU read locks
and adds a few selftests to verify correct behavior. This issue has also
come up for sched-ext programs.
This would now allow static subprog calls to be made without errors
within RCU read sections, for subprogs to release RCU locks of callers
and return to them, or for subprogs to take RCU lock which is later
released in the caller.
[0]: https://lore.kernel.org/bpf/20240204120206.796412-1-memxor@gmail.com
Changelog:
----------
v1 -> v2:
v1: https://lore.kernel.org/bpf/20240204230231.1013964-1-memxor@gmail.com
* Add tests for global subprog behaviour (Yafang)
* Add Acks, Tested-by (Yonghong, Yafang)
====================
Link: https://lore.kernel.org/r/20240205055646.1112186-1-memxor@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
Add selftests covering the following cases:
- A static or global subprog called from within a RCU read section works
- A static subprog taking an RCU read lock which is released in caller works
- A static subprog releasing the caller's RCU read lock works
Global subprogs that leave the lock in an imbalanced state will not
work, as they are verified separately, so ensure those cases fail as
well.
Acked-by: Yonghong Song <yonghong.song@linux.dev>
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Acked-by: David Vernet <void@manifault.com>
Link: https://lore.kernel.org/r/20240205055646.1112186-3-memxor@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
|
Allow transferring an imbalanced RCU lock state between subprog calls
during verification. This allows patterns where a subprog call returns
with an RCU lock held, or a subprog call releases an RCU lock held by
the caller. Currently, the verifier would end up complaining if the RCU
lock is not released when processing an exit from a subprog, which is
non-ideal if its execution is supposed to be enclosed in an RCU read
section of the caller.
Instead, simply only check whether we are processing exit for frame#0
and do not complain on an active RCU lock otherwise. We only need to
update the check when processing BPF_EXIT insn, as copy_verifier_state
is already set up to do the right thing.
Suggested-by: David Vernet <void@manifault.com>
Tested-by: Yafang Shao <laoar.shao@gmail.com>
Acked-by: Yonghong Song <yonghong.song@linux.dev>
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Acked-by: David Vernet <void@manifault.com>
Link: https://lore.kernel.org/r/20240205055646.1112186-2-memxor@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
Dave Thaler
Alexei Starovoitov
Eduard Zingerman
Marcos Paulo de Souza
Yonghong Song
Andrii Nakryiko
Matt Bobrowski
Oliver Crumrine
Martin KaFai Lau
Kui-Feng Lee
Cupertino Miranda
Jose E. Marchesi
Daniel Xu
Marco Elver
Geliang Tang
Masahiro Yamada
Yafang Shao
Viktor Malik
Toke Høiland-Jørgensen
Magnus Karlsson
Kumar Kartikeya Dwivedi
