mirror of
https://git.kernel.org/pub/scm/linux/kernel/git/chenhuacai/linux-loongson
synced 2025-09-02 08:32:55 +00:00
loongarch-next
235 Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
Andrii Nakryiko
|
e26080d0da |
bpf: prepare btf_prepare_func_args() for handling static subprogs
Generalize btf_prepare_func_args() to support both global and static subprogs. We are going to utilize this property in the next patch, reusing btf_prepare_func_args() for subprog call logic instead of reparsing BTF information in a completely separate implementation. btf_prepare_func_args() now detects whether subprog is global or static makes slight logic adjustments for static func cases, like not failing fatally (-EFAULT) for conditions that are allowable for static subprogs. Somewhat subtle (but major!) difference is the handling of pointer arguments. Both global and static functions need to handle special context arguments (which are pointers to predefined type names), but static subprogs give up on any other pointers, falling back to marking subprog as "unreliable", disabling the use of BTF type information altogether. For global functions, though, we are assuming that such pointers to unrecognized types are just pointers to fixed-sized memory region (or error out if size cannot be established, like for `void *` pointers). This patch accommodates these small differences and sets up a stage for refactoring in the next patch, eliminating a separate BTF-based parsing logic in btf_check_func_arg_match(). Acked-by: Eduard Zingerman <eddyz87@gmail.com> Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/r/20231215011334.2307144-4-andrii@kernel.org Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Andrii Nakryiko
|
4ba1d0f234 |
bpf: abstract away global subprog arg preparation logic from reg state setup
btf_prepare_func_args() is used to understand expectations and restrictions on global subprog arguments. But current implementation is hard to extend, as it intermixes BTF-based func prototype parsing and interpretation logic with setting up register state at subprog entry. Worse still, those registers are not completely set up inside btf_prepare_func_args(), requiring some more logic later in do_check_common(). Like calling mark_reg_unknown() and similar initialization operations. This intermixing of BTF interpretation and register state setup is problematic. First, it causes duplication of BTF parsing logic for global subprog verification (to set up initial state of global subprog) and global subprog call sites analysis (when we need to check that whatever is being passed into global subprog matches expectations), performed in btf_check_subprog_call(). Given we want to extend global func argument with tags later, this duplication is problematic. So refactor btf_prepare_func_args() to do only BTF-based func proto and args parsing, returning high-level argument "expectations" only, with no regard to specifics of register state. I.e., if it's a context argument, instead of setting register state to PTR_TO_CTX, we return ARG_PTR_TO_CTX enum for that argument as "an argument specification" for further processing inside do_check_common(). Similarly for SCALAR arguments, PTR_TO_MEM, etc. This allows to reuse btf_prepare_func_args() in following patches at global subprog call site analysis time. It also keeps register setup code consistently in one place, do_check_common(). Besides all this, we cache this argument specs information inside env->subprog_info, eliminating the need to redo these potentially expensive BTF traversals, especially if BPF program's BTF is big and/or there are lots of global subprog calls. Acked-by: Eduard Zingerman <eddyz87@gmail.com> Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/r/20231215011334.2307144-2-andrii@kernel.org Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Andrii Nakryiko
|
406a6fa44b |
bpf: use bitfields for simple per-subprog bool flags
We have a bunch of bool flags for each subprog. Instead of wasting bytes for them, use bitfields instead. Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Acked-by: Eduard Zingerman <eddyz87@gmail.com> Link: https://lore.kernel.org/r/20231204233931.49758-5-andrii@kernel.org Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Andrei Matei
|
92e1567ee3 |
bpf: Add some comments to stack representation
Add comments to the datastructure tracking the stack state, as the mapping between each stack slot and where its state is stored is not entirely obvious. Signed-off-by: Andrei Matei <andreimatei1@gmail.com> Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Acked-by: Eduard Zingerman <eddyz87@gmail.com> Link: https://lore.kernel.org/bpf/20231208032519.260451-2-andreimatei1@gmail.com |
||
|
Andrii Nakryiko
|
41f6f64e69 |
bpf: support non-r10 register spill/fill to/from stack in precision tracking
Use instruction (jump) history to record instructions that performed register spill/fill to/from stack, regardless if this was done through read-only r10 register, or any other register after copying r10 into it *and* potentially adjusting offset. To make this work reliably, we push extra per-instruction flags into instruction history, encoding stack slot index (spi) and stack frame number in extra 10 bit flags we take away from prev_idx in instruction history. We don't touch idx field for maximum performance, as it's checked most frequently during backtracking. This change removes basically the last remaining practical limitation of precision backtracking logic in BPF verifier. It fixes known deficiencies, but also opens up new opportunities to reduce number of verified states, explored in the subsequent patches. There are only three differences in selftests' BPF object files according to veristat, all in the positive direction (less states). File Program Insns (A) Insns (B) Insns (DIFF) States (A) States (B) States (DIFF) -------------------------------------- ------------- --------- --------- ------------- ---------- ---------- ------------- test_cls_redirect_dynptr.bpf.linked3.o cls_redirect 2987 2864 -123 (-4.12%) 240 231 -9 (-3.75%) xdp_synproxy_kern.bpf.linked3.o syncookie_tc 82848 82661 -187 (-0.23%) 5107 5073 -34 (-0.67%) xdp_synproxy_kern.bpf.linked3.o syncookie_xdp 85116 84964 -152 (-0.18%) 5162 5130 -32 (-0.62%) Note, I avoided renaming jmp_history to more generic insn_hist to minimize number of lines changed and potential merge conflicts between bpf and bpf-next trees. Notice also cur_hist_entry pointer reset to NULL at the beginning of instruction verification loop. This pointer avoids the problem of relying on last jump history entry's insn_idx to determine whether we already have entry for current instruction or not. It can happen that we added jump history entry because current instruction is_jmp_point(), but also we need to add instruction flags for stack access. In this case, we don't want to entries, so we need to reuse last added entry, if it is present. Relying on insn_idx comparison has the same ambiguity problem as the one that was fixed recently in [0], so we avoid that. [0] https://patchwork.kernel.org/project/netdevbpf/patch/20231110002638.4168352-3-andrii@kernel.org/ Acked-by: Eduard Zingerman <eddyz87@gmail.com> Reported-by: Tao Lyu <tao.lyu@epfl.ch> Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/r/20231205184248.1502704-2-andrii@kernel.org Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Andrii Nakryiko
|
8fa4ecd49b |
bpf: enforce exact retval range on subprog/callback exit
Instead of relying on potentially imprecise tnum representation of expected return value range for callbacks and subprogs, validate that smin/smax range satisfy exact expected range of return values. E.g., if callback would need to return [0, 2] range, tnum can't represent this precisely and instead will allow [0, 3] range. By checking smin/smax range, we can make sure that subprog/callback indeed returns only valid [0, 2] range. Acked-by: Eduard Zingerman <eddyz87@gmail.com> Acked-by: Shung-Hsi Yu <shung-hsi.yu@suse.com> Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/r/20231202175705.885270-5-andrii@kernel.org Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Andrii Nakryiko
|
45b5623f2d |
bpf: rearrange bpf_func_state fields to save a bit of memory
It's a trivial rearrangement saving 8 bytes. We have 4 bytes of padding
at the end which can be filled with another field without increasing
struct bpf_func_state.
copy_func_state() logic remains correct without any further changes.
BEFORE
======
struct bpf_func_state {
struct bpf_reg_state regs[11]; /* 0 1320 */
/* --- cacheline 20 boundary (1280 bytes) was 40 bytes ago --- */
int callsite; /* 1320 4 */
u32 frameno; /* 1324 4 */
u32 subprogno; /* 1328 4 */
u32 async_entry_cnt; /* 1332 4 */
bool in_callback_fn; /* 1336 1 */
/* XXX 7 bytes hole, try to pack */
/* --- cacheline 21 boundary (1344 bytes) --- */
struct tnum callback_ret_range; /* 1344 16 */
bool in_async_callback_fn; /* 1360 1 */
bool in_exception_callback_fn; /* 1361 1 */
/* XXX 2 bytes hole, try to pack */
int acquired_refs; /* 1364 4 */
struct bpf_reference_state * refs; /* 1368 8 */
int allocated_stack; /* 1376 4 */
/* XXX 4 bytes hole, try to pack */
struct bpf_stack_state * stack; /* 1384 8 */
/* size: 1392, cachelines: 22, members: 13 */
/* sum members: 1379, holes: 3, sum holes: 13 */
/* last cacheline: 48 bytes */
};
AFTER
=====
struct bpf_func_state {
struct bpf_reg_state regs[11]; /* 0 1320 */
/* --- cacheline 20 boundary (1280 bytes) was 40 bytes ago --- */
int callsite; /* 1320 4 */
u32 frameno; /* 1324 4 */
u32 subprogno; /* 1328 4 */
u32 async_entry_cnt; /* 1332 4 */
struct tnum callback_ret_range; /* 1336 16 */
/* --- cacheline 21 boundary (1344 bytes) was 8 bytes ago --- */
bool in_callback_fn; /* 1352 1 */
bool in_async_callback_fn; /* 1353 1 */
bool in_exception_callback_fn; /* 1354 1 */
/* XXX 1 byte hole, try to pack */
int acquired_refs; /* 1356 4 */
struct bpf_reference_state * refs; /* 1360 8 */
struct bpf_stack_state * stack; /* 1368 8 */
int allocated_stack; /* 1376 4 */
/* size: 1384, cachelines: 22, members: 13 */
/* sum members: 1379, holes: 1, sum holes: 1 */
/* padding: 4 */
/* last cacheline: 40 bytes */
};
Acked-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/r/20231202175705.885270-2-andrii@kernel.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
||
|
Jakub Kicinski
|
45c226dde7 |
Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Cross-merge networking fixes after downstream PR. Conflicts: drivers/net/ethernet/intel/ice/ice_main.c |
||
|
Eduard Zingerman
|
bb124da69c |
bpf: keep track of max number of bpf_loop callback iterations
In some cases verifier can't infer convergence of the bpf_loop()
iteration. E.g. for the following program:
static int cb(__u32 idx, struct num_context* ctx)
{
ctx->i++;
return 0;
}
SEC("?raw_tp")
int prog(void *_)
{
struct num_context ctx = { .i = 0 };
__u8 choice_arr[2] = { 0, 1 };
bpf_loop(2, cb, &ctx, 0);
return choice_arr[ctx.i];
}
Each 'cb' simulation would eventually return to 'prog' and reach
'return choice_arr[ctx.i]' statement. At which point ctx.i would be
marked precise, thus forcing verifier to track multitude of separate
states with {.i=0}, {.i=1}, ... at bpf_loop() callback entry.
This commit allows "brute force" handling for such cases by limiting
number of callback body simulations using 'umax' value of the first
bpf_loop() parameter.
For this, extend bpf_func_state with 'callback_depth' field.
Increment this field when callback visiting state is pushed to states
traversal stack. For frame #N it's 'callback_depth' field counts how
many times callback with frame depth N+1 had been executed.
Use bpf_func_state specifically to allow independent tracking of
callback depths when multiple nested bpf_loop() calls are present.
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/r/20231121020701.26440-11-eddyz87@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
||
|
Eduard Zingerman
|
ab5cfac139 |
bpf: verify callbacks as if they are called unknown number of times
Prior to this patch callbacks were handled as regular function calls,
execution of callback body was modeled exactly once.
This patch updates callbacks handling logic as follows:
- introduces a function push_callback_call() that schedules callback
body verification in env->head stack;
- updates prepare_func_exit() to reschedule callback body verification
upon BPF_EXIT;
- as calls to bpf_*_iter_next(), calls to callback invoking functions
are marked as checkpoints;
- is_state_visited() is updated to stop callback based iteration when
some identical parent state is found.
Paths with callback function invoked zero times are now verified first,
which leads to necessity to modify some selftests:
- the following negative tests required adding release/unlock/drop
calls to avoid previously masked unrelated error reports:
- cb_refs.c:underflow_prog
- exceptions_fail.c:reject_rbtree_add_throw
- exceptions_fail.c:reject_with_cp_reference
- the following precision tracking selftests needed change in expected
log trace:
- verifier_subprog_precision.c:callback_result_precise
(note: r0 precision is no longer propagated inside callback and
I think this is a correct behavior)
- verifier_subprog_precision.c:parent_callee_saved_reg_precise_with_callback
- verifier_subprog_precision.c:parent_stack_slot_precise_with_callback
Reported-by: Andrew Werner <awerner32@gmail.com>
Closes: https://lore.kernel.org/bpf/CA+vRuzPChFNXmouzGG+wsy=6eMcfr1mFG0F3g7rbg-sedGKW3w@mail.gmail.com/
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/r/20231121020701.26440-7-eddyz87@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
||
|
Andrii Nakryiko
|
42feb6620a |
bpf: move verifier state printing code to kernel/bpf/log.c
Move a good chunk of code from verifier.c to log.c: verifier state verbose printing logic. This is an important and very much logging/debugging oriented code. It fits the overlall log.c's focus on verifier logging, and moving it allows to keep growing it without unnecessarily adding to verifier.c code that otherwise contains a core verification logic. There are not many shared dependencies between this code and the rest of verifier.c code, except a few single-line helpers for various register type checks and a bit of state "scratching" helpers. We move all such trivial helpers into include/bpf/bpf_verifier.h as static inlines. No functional changes in this patch. Acked-by: Eduard Zingerman <eddyz87@gmail.com> Acked-by: Stanislav Fomichev <sdf@google.com> Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/r/20231118034623.3320920-3-andrii@kernel.org Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Andrii Nakryiko
|
db840d389b |
bpf: move verbose_linfo() into kernel/bpf/log.c
verifier.c is huge. Let's try to move out parts that are logging-related into log.c, as we previously did with bpf_log() and other related stuff. This patch moves line info verbose output routines: it's pretty self-contained and isolated code, so there is no problem with this. Acked-by: Eduard Zingerman <eddyz87@gmail.com> Acked-by: Stanislav Fomichev <sdf@google.com> Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/r/20231118034623.3320920-2-andrii@kernel.org Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Andrii Nakryiko
|
ff8867af01 |
bpf: rename BPF_F_TEST_SANITY_STRICT to BPF_F_TEST_REG_INVARIANTS
Rename verifier internal flag BPF_F_TEST_SANITY_STRICT to more neutral BPF_F_TEST_REG_INVARIANTS. This is a follow up to [0]. A few selftests and veristat need to be adjusted in the same patch as well. [0] https://patchwork.kernel.org/project/netdevbpf/patch/20231112010609.848406-5-andrii@kernel.org/ Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/r/20231117171404.225508-1-andrii@kernel.org Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Andrii Nakryiko
|
5f99f312bd |
bpf: add register bounds sanity checks and sanitization
Add simple sanity checks that validate well-formed ranges (min <= max) across u64, s64, u32, and s32 ranges. Also for cases when the value is constant (either 64-bit or 32-bit), we validate that ranges and tnums are in agreement. These bounds checks are performed at the end of BPF_ALU/BPF_ALU64 operations, on conditional jumps, and for LDX instructions (where subreg zero/sign extension is probably the most important to check). This covers most of the interesting cases. Also, we validate the sanity of the return register when manually adjusting it for some special helpers. By default, sanity violation will trigger a warning in verifier log and resetting register bounds to "unbounded" ones. But to aid development and debugging, BPF_F_TEST_SANITY_STRICT flag is added, which will trigger hard failure of verification with -EFAULT on register bounds violations. This allows selftests to catch such issues. veristat will also gain a CLI option to enable this behavior. Acked-by: Eduard Zingerman <eddyz87@gmail.com> Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Acked-by: Shung-Hsi Yu <shung-hsi.yu@suse.com> Link: https://lore.kernel.org/r/20231112010609.848406-5-andrii@kernel.org Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Eduard Zingerman
|
2a0992829e |
bpf: correct loop detection for iterators convergence
It turns out that .branches > 0 in is_state_visited() is not a
sufficient condition to identify if two verifier states form a loop
when iterators convergence is computed. This commit adds logic to
distinguish situations like below:
(I) initial (II) initial
| |
V V
.---------> hdr ..
| | |
| V V
| .------... .------..
| | | | |
| V V V V
| ... ... .-> hdr ..
| | | | | |
| V V | V V
| succ <- cur | succ <- cur
| | | |
| V | V
| ... | ...
| | | |
'----' '----'
For both (I) and (II) successor 'succ' of the current state 'cur' was
previously explored and has branches count at 0. However, loop entry
'hdr' corresponding to 'succ' might be a part of current DFS path.
If that is the case 'succ' and 'cur' are members of the same loop
and have to be compared exactly.
Co-developed-by: Andrii Nakryiko <andrii.nakryiko@gmail.com>
Co-developed-by: Alexei Starovoitov <alexei.starovoitov@gmail.com>
Reviewed-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/r/20231024000917.12153-6-eddyz87@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
||
|
Eduard Zingerman
|
2793a8b015 |
bpf: exact states comparison for iterator convergence checks
Convergence for open coded iterators is computed in is_state_visited()
by examining states with branches count > 1 and using states_equal().
states_equal() computes sub-state relation using read and precision marks.
Read and precision marks are propagated from children states,
thus are not guaranteed to be complete inside a loop when branches
count > 1. This could be demonstrated using the following unsafe program:
1. r7 = -16
2. r6 = bpf_get_prandom_u32()
3. while (bpf_iter_num_next(&fp[-8])) {
4. if (r6 != 42) {
5. r7 = -32
6. r6 = bpf_get_prandom_u32()
7. continue
8. }
9. r0 = r10
10. r0 += r7
11. r8 = *(u64 *)(r0 + 0)
12. r6 = bpf_get_prandom_u32()
13. }
Here verifier would first visit path 1-3, create a checkpoint at 3
with r7=-16, continue to 4-7,3 with r7=-32.
Because instructions at 9-12 had not been visitied yet existing
checkpoint at 3 does not have read or precision mark for r7.
Thus states_equal() would return true and verifier would discard
current state, thus unsafe memory access at 11 would not be caught.
This commit fixes this loophole by introducing exact state comparisons
for iterator convergence logic:
- registers are compared using regs_exact() regardless of read or
precision marks;
- stack slots have to have identical type.
Unfortunately, this is too strict even for simple programs like below:
i = 0;
while(iter_next(&it))
i++;
At each iteration step i++ would produce a new distinct state and
eventually instruction processing limit would be reached.
To avoid such behavior speculatively forget (widen) range for
imprecise scalar registers, if those registers were not precise at the
end of the previous iteration and do not match exactly.
This a conservative heuristic that allows to verify wide range of
programs, however it precludes verification of programs that conjure
an imprecise value on the first loop iteration and use it as precise
on the second.
Test case iter_task_vma_for_each() presents one of such cases:
unsigned int seen = 0;
...
bpf_for_each(task_vma, vma, task, 0) {
if (seen >= 1000)
break;
...
seen++;
}
Here clang generates the following code:
<LBB0_4>:
24: r8 = r6 ; stash current value of
... body ... 'seen'
29: r1 = r10
30: r1 += -0x8
31: call bpf_iter_task_vma_next
32: r6 += 0x1 ; seen++;
33: if r0 == 0x0 goto +0x2 <LBB0_6> ; exit on next() == NULL
34: r7 += 0x10
35: if r8 < 0x3e7 goto -0xc <LBB0_4> ; loop on seen < 1000
<LBB0_6>:
... exit ...
Note that counter in r6 is copied to r8 and then incremented,
conditional jump is done using r8. Because of this precision mark for
r6 lags one state behind of precision mark on r8 and widening logic
kicks in.
Adding barrier_var(seen) after conditional is sufficient to force
clang use the same register for both counting and conditional jump.
This issue was discussed in the thread [1] which was started by
Andrew Werner <awerner32@gmail.com> demonstrating a similar bug
in callback functions handling. The callbacks would be addressed
in a followup patch.
[1] https://lore.kernel.org/bpf/97a90da09404c65c8e810cf83c94ac703705dc0e.camel@gmail.com/
Co-developed-by: Andrii Nakryiko <andrii.nakryiko@gmail.com>
Co-developed-by: Alexei Starovoitov <alexei.starovoitov@gmail.com>
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/r/20231024000917.12153-4-eddyz87@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
||
|
Chuyi Zhou
|
dfab99df14 |
bpf: teach the verifier to enforce css_iter and task_iter in RCU CS
css_iter and task_iter should be used in rcu section. Specifically, in sleepable progs explicit bpf_rcu_read_lock() is needed before use these iters. In normal bpf progs that have implicit rcu_read_lock(), it's OK to use them directly. This patch adds a new a KF flag KF_RCU_PROTECTED for bpf_iter_task_new and bpf_iter_css_new. It means the kfunc should be used in RCU CS. We check whether we are in rcu cs before we want to invoke this kfunc. If the rcu protection is guaranteed, we would let st->type = PTR_TO_STACK | MEM_RCU. Once user do rcu_unlock during the iteration, state MEM_RCU of regs would be cleared. is_iter_reg_valid_init() will reject if reg->type is UNTRUSTED. It is worth noting that currently, bpf_rcu_read_unlock does not clear the state of the STACK_ITER reg, since bpf_for_each_spilled_reg only considers STACK_SPILL. This patch also let bpf_for_each_spilled_reg search STACK_ITER. Signed-off-by: Chuyi Zhou <zhouchuyi@bytedance.com> Acked-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/r/20231018061746.111364-6-zhouchuyi@bytedance.com Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Kumar Kartikeya Dwivedi
|
b9ae0c9dd0 |
bpf: Add support for custom exception callbacks
By default, the subprog generated by the verifier to handle a thrown exception hardcodes a return value of 0. To allow user-defined logic and modification of the return value when an exception is thrown, introduce the 'exception_callback:' declaration tag, which marks a callback as the default exception handler for the program. The format of the declaration tag is 'exception_callback:<value>', where <value> is the name of the exception callback. Each main program can be tagged using this BTF declaratiion tag to associate it with an exception callback. In case the tag is absent, the default callback is used. As such, the exception callback cannot be modified at runtime, only set during verification. Allowing modification of the callback for the current program execution at runtime leads to issues when the programs begin to nest, as any per-CPU state maintaing this information will have to be saved and restored. We don't want it to stay in bpf_prog_aux as this takes a global effect for all programs. An alternative solution is spilling the callback pointer at a known location on the program stack on entry, and then passing this location to bpf_throw as a parameter. However, since exceptions are geared more towards a use case where they are ideally never invoked, optimizing for this use case and adding to the complexity has diminishing returns. Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com> Link: https://lore.kernel.org/r/20230912233214.1518551-7-memxor@gmail.com Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Kumar Kartikeya Dwivedi
|
f18b03faba |
bpf: Implement BPF exceptions
This patch implements BPF exceptions, and introduces a bpf_throw kfunc to allow programs to throw exceptions during their execution at runtime. A bpf_throw invocation is treated as an immediate termination of the program, returning back to its caller within the kernel, unwinding all stack frames. This allows the program to simplify its implementation, by testing for runtime conditions which the verifier has no visibility into, and assert that they are true. In case they are not, the program can simply throw an exception from the other branch. BPF exceptions are explicitly *NOT* an unlikely slowpath error handling primitive, and this objective has guided design choices of the implementation of the them within the kernel (with the bulk of the cost for unwinding the stack offloaded to the bpf_throw kfunc). The implementation of this mechanism requires use of add_hidden_subprog mechanism introduced in the previous patch, which generates a couple of instructions to move R1 to R0 and exit. The JIT then rewrites the prologue of this subprog to take the stack pointer and frame pointer as inputs and reset the stack frame, popping all callee-saved registers saved by the main subprog. The bpf_throw function then walks the stack at runtime, and invokes this exception subprog with the stack and frame pointers as parameters. Reviewers must take note that currently the main program is made to save all callee-saved registers on x86_64 during entry into the program. This is because we must do an equivalent of a lightweight context switch when unwinding the stack, therefore we need the callee-saved registers of the caller of the BPF program to be able to return with a sane state. Note that we have to additionally handle r12, even though it is not used by the program, because when throwing the exception the program makes an entry into the kernel which could clobber r12 after saving it on the stack. To be able to preserve the value we received on program entry, we push r12 and restore it from the generated subprogram when unwinding the stack. For now, bpf_throw invocation fails when lingering resources or locks exist in that path of the program. In a future followup, bpf_throw will be extended to perform frame-by-frame unwinding to release lingering resources for each stack frame, removing this limitation. Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com> Link: https://lore.kernel.org/r/20230912233214.1518551-5-memxor@gmail.com Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Kumar Kartikeya Dwivedi
|
335d1c5b54 |
bpf: Implement support for adding hidden subprogs
Introduce support in the verifier for generating a subprogram and include it as part of a BPF program dynamically after the do_check phase is complete. The first user will be the next patch which generates default exception callbacks if none are set for the program. The phase of invocation will be do_misc_fixups. Note that this is an internal verifier function, and should be used with instruction blocks which uphold the invariants stated in check_subprogs. Since these subprogs are always appended to the end of the instruction sequence of the program, it becomes relatively inexpensive to do the related adjustments to the subprog_info of the program. Only the fake exit subprogram is shifted forward, making room for our new subprog. This is useful to insert a new subprogram, get it JITed, and obtain its function pointer. The next patch will use this functionality to insert a default exception callback which will be invoked after unwinding the stack. Note that these added subprograms are invisible to userspace, and never reported in BPF_OBJ_GET_INFO_BY_ID etc. For now, only a single subprogram is supported, but more can be easily supported in the future. To this end, two function counts are introduced now, the existing func_cnt, and real_func_cnt, the latter including hidden programs. This allows us to conver the JIT code to use the real_func_cnt for management of resources while syscall path continues working with existing func_cnt. Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com> Link: https://lore.kernel.org/r/20230912233214.1518551-4-memxor@gmail.com Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Yonghong Song
|
01cc55af93 |
bpf: Add bpf_this_cpu_ptr/bpf_per_cpu_ptr support for allocated percpu obj
The bpf helpers bpf_this_cpu_ptr() and bpf_per_cpu_ptr() are re-purposed for allocated percpu objects. For an allocated percpu obj, the reg type is 'PTR_TO_BTF_ID | MEM_PERCPU | MEM_RCU'. The return type for these two re-purposed helpera is 'PTR_TO_MEM | MEM_RCU | MEM_ALLOC'. The MEM_ALLOC allows that the per-cpu data can be read and written. Since the memory allocator bpf_mem_alloc() returns a ptr to a percpu ptr for percpu data, the first argument of bpf_this_cpu_ptr() and bpf_per_cpu_ptr() is patched with a dereference before passing to the helper func. Signed-off-by: Yonghong Song <yonghong.song@linux.dev> Link: https://lore.kernel.org/r/20230827152749.1997202-1-yonghong.song@linux.dev Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Dave Marchevsky
|
2a6d50b50d |
bpf: Consider non-owning refs trusted
Recent discussions around default kptr "trustedness" led to changes such as commit |
||
|
Eduard Zingerman
|
1ffc85d929 |
bpf: Verify scalar ids mapping in regsafe() using check_ids()
Make sure that the following unsafe example is rejected by verifier:
1: r9 = ... some pointer with range X ...
2: r6 = ... unbound scalar ID=a ...
3: r7 = ... unbound scalar ID=b ...
4: if (r6 > r7) goto +1
5: r6 = r7
6: if (r6 > X) goto ...
--- checkpoint ---
7: r9 += r7
8: *(u64 *)r9 = Y
This example is unsafe because not all execution paths verify r7 range.
Because of the jump at (4) the verifier would arrive at (6) in two states:
I. r6{.id=b}, r7{.id=b} via path 1-6;
II. r6{.id=a}, r7{.id=b} via path 1-4, 6.
Currently regsafe() does not call check_ids() for scalar registers,
thus from POV of regsafe() states (I) and (II) are identical. If the
path 1-6 is taken by verifier first, and checkpoint is created at (6)
the path [1-4, 6] would be considered safe.
Changes in this commit:
- check_ids() is modified to disallow mapping multiple old_id to the
same cur_id.
- check_scalar_ids() is added, unlike check_ids() it treats ID zero as
a unique scalar ID.
- check_scalar_ids() needs to generate temporary unique IDs, field
'tmp_id_gen' is added to bpf_verifier_env::idmap_scratch to
facilitate this.
- regsafe() is updated to:
- use check_scalar_ids() for precise scalar registers.
- compare scalar registers using memcmp only for explore_alu_limits
branch. This simplifies control flow for scalar case, and has no
measurable performance impact.
- check_alu_op() is updated to avoid generating bpf_reg_state::id for
constant scalar values when processing BPF_MOV. ID is needed to
propagate range information for identical values, but there is
nothing to propagate for constants.
Fixes:
|
||
|
Eduard Zingerman
|
904e6ddf41 |
bpf: Use scalar ids in mark_chain_precision()
Change mark_chain_precision() to track precision in situations
like below:
r2 = unknown value
...
--- state #0 ---
...
r1 = r2 // r1 and r2 now share the same ID
...
--- state #1 {r1.id = A, r2.id = A} ---
...
if (r2 > 10) goto exit; // find_equal_scalars() assigns range to r1
...
--- state #2 {r1.id = A, r2.id = A} ---
r3 = r10
r3 += r1 // need to mark both r1 and r2
At the beginning of the processing of each state, ensure that if a
register with a scalar ID is marked as precise, all registers sharing
this ID are also marked as precise.
This property would be used by a follow-up change in regsafe().
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20230613153824.3324830-2-eddyz87@gmail.com
|
||
|
Andrii Nakryiko
|
d9439c21a9 |
bpf: improve precision backtrack logging
Add helper to format register and stack masks in more human-readable format. Adjust logging a bit during backtrack propagation and especially during forcing precision fallback logic to make it clearer what's going on (with log_level=2, of course), and also start reporting affected frame depth. This is in preparation for having more than one active frame later when precision propagation between subprog calls is added. Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/r/20230505043317.3629845-5-andrii@kernel.org Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Andrii Nakryiko
|
407958a0e9 |
bpf: encapsulate precision backtracking bookkeeping
Add struct backtrack_state and straightforward API around it to keep track of register and stack masks used and maintained during precision backtracking process. Having this logic separately allow to keep high-level backtracking algorithm cleaner, but also it sets us up to cleanly keep track of register and stack masks per frame, allowing (with some further logic adjustments) to perform precision backpropagation across multiple frames (i.e., subprog calls). Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/r/20230505043317.3629845-4-andrii@kernel.org Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Dave Marchevsky
|
d2dcc67df9 |
bpf: Migrate bpf_rbtree_add and bpf_list_push_{front,back} to possibly fail
Consider this code snippet:
struct node {
long key;
bpf_list_node l;
bpf_rb_node r;
bpf_refcount ref;
}
int some_bpf_prog(void *ctx)
{
struct node *n = bpf_obj_new(/*...*/), *m;
bpf_spin_lock(&glock);
bpf_rbtree_add(&some_tree, &n->r, /* ... */);
m = bpf_refcount_acquire(n);
bpf_rbtree_add(&other_tree, &m->r, /* ... */);
bpf_spin_unlock(&glock);
/* ... */
}
After bpf_refcount_acquire, n and m point to the same underlying memory,
and that node's bpf_rb_node field is being used by the some_tree insert,
so overwriting it as a result of the second insert is an error. In order
to properly support refcounted nodes, the rbtree and list insert
functions must be allowed to fail. This patch adds such support.
The kfuncs bpf_rbtree_add, bpf_list_push_{front,back} are modified to
return an int indicating success/failure, with 0 -> success, nonzero ->
failure.
bpf_obj_drop on failure
=======================
Currently the only reason an insert can fail is the example above: the
bpf_{list,rb}_node is already in use. When such a failure occurs, the
insert kfuncs will bpf_obj_drop the input node. This allows the insert
operations to logically fail without changing their verifier owning ref
behavior, namely the unconditional release_reference of the input
owning ref.
With insert that always succeeds, ownership of the node is always passed
to the collection, since the node always ends up in the collection.
With a possibly-failed insert w/ bpf_obj_drop, ownership of the node
is always passed either to the collection (success), or to bpf_obj_drop
(failure). Regardless, it's correct to continue unconditionally
releasing the input owning ref, as something is always taking ownership
from the calling program on insert.
Keeping owning ref behavior unchanged results in a nice default UX for
insert functions that can fail. If the program's reaction to a failed
insert is "fine, just get rid of this owning ref for me and let me go
on with my business", then there's no reason to check for failure since
that's default behavior. e.g.:
long important_failures = 0;
int some_bpf_prog(void *ctx)
{
struct node *n, *m, *o; /* all bpf_obj_new'd */
bpf_spin_lock(&glock);
bpf_rbtree_add(&some_tree, &n->node, /* ... */);
bpf_rbtree_add(&some_tree, &m->node, /* ... */);
if (bpf_rbtree_add(&some_tree, &o->node, /* ... */)) {
important_failures++;
}
bpf_spin_unlock(&glock);
}
If we instead chose to pass ownership back to the program on failed
insert - by returning NULL on success or an owning ref on failure -
programs would always have to do something with the returned ref on
failure. The most likely action is probably "I'll just get rid of this
owning ref and go about my business", which ideally would look like:
if (n = bpf_rbtree_add(&some_tree, &n->node, /* ... */))
bpf_obj_drop(n);
But bpf_obj_drop isn't allowed in a critical section and inserts must
occur within one, so in reality error handling would become a
hard-to-parse mess.
For refcounted nodes, we can replicate the "pass ownership back to
program on failure" logic with this patch's semantics, albeit in an ugly
way:
struct node *n = bpf_obj_new(/* ... */), *m;
bpf_spin_lock(&glock);
m = bpf_refcount_acquire(n);
if (bpf_rbtree_add(&some_tree, &n->node, /* ... */)) {
/* Do something with m */
}
bpf_spin_unlock(&glock);
bpf_obj_drop(m);
bpf_refcount_acquire is used to simulate "return owning ref on failure".
This should be an uncommon occurrence, though.
Addition of two verifier-fixup'd args to collection inserts
===========================================================
The actual bpf_obj_drop kfunc is
bpf_obj_drop_impl(void *, struct btf_struct_meta *), with bpf_obj_drop
macro populating the second arg with 0 and the verifier later filling in
the arg during insn fixup.
Because bpf_rbtree_add and bpf_list_push_{front,back} now might do
bpf_obj_drop, these kfuncs need a btf_struct_meta parameter that can be
passed to bpf_obj_drop_impl.
Similarly, because the 'node' param to those insert functions is the
bpf_{list,rb}_node within the node type, and bpf_obj_drop expects a
pointer to the beginning of the node, the insert functions need to be
able to find the beginning of the node struct. A second
verifier-populated param is necessary: the offset of {list,rb}_node within the
node type.
These two new params allow the insert kfuncs to correctly call
__bpf_obj_drop_impl:
beginning_of_node = bpf_rb_node_ptr - offset
if (already_inserted)
__bpf_obj_drop_impl(beginning_of_node, btf_struct_meta->record);
Similarly to other kfuncs with "hidden" verifier-populated params, the
insert functions are renamed with _impl prefix and a macro is provided
for common usage. For example, bpf_rbtree_add kfunc is now
bpf_rbtree_add_impl and bpf_rbtree_add is now a macro which sets
"hidden" args to 0.
Due to the two new args BPF progs will need to be recompiled to work
with the new _impl kfuncs.
This patch also rewrites the "hidden argument" explanation to more
directly say why the BPF program writer doesn't need to populate the
arguments with anything meaningful.
How does this new logic affect non-owning references?
=====================================================
Currently, non-owning refs are valid until the end of the critical
section in which they're created. We can make this guarantee because, if
a non-owning ref exists, the referent was added to some collection. The
collection will drop() its nodes when it goes away, but it can't go away
while our program is accessing it, so that's not a problem. If the
referent is removed from the collection in the same CS that it was added
in, it can't be bpf_obj_drop'd until after CS end. Those are the only
two ways to free the referent's memory and neither can happen until
after the non-owning ref's lifetime ends.
On first glance, having these collection insert functions potentially
bpf_obj_drop their input seems like it breaks the "can't be
bpf_obj_drop'd until after CS end" line of reasoning. But we care about
the memory not being _freed_ until end of CS end, and a previous patch
in the series modified bpf_obj_drop such that it doesn't free refcounted
nodes until refcount == 0. So the statement can be more accurately
rewritten as "can't be free'd until after CS end".
We can prove that this rewritten statement holds for any non-owning
reference produced by collection insert functions:
* If the input to the insert function is _not_ refcounted
* We have an owning reference to the input, and can conclude it isn't
in any collection
* Inserting a node in a collection turns owning refs into
non-owning, and since our input type isn't refcounted, there's no
way to obtain additional owning refs to the same underlying
memory
* Because our node isn't in any collection, the insert operation
cannot fail, so bpf_obj_drop will not execute
* If bpf_obj_drop is guaranteed not to execute, there's no risk of
memory being free'd
* Otherwise, the input to the insert function is refcounted
* If the insert operation fails due to the node's list_head or rb_root
already being in some collection, there was some previous successful
insert which passed refcount to the collection
* We have an owning reference to the input, it must have been
acquired via bpf_refcount_acquire, which bumped the refcount
* refcount must be >= 2 since there's a valid owning reference and the
node is already in a collection
* Insert triggering bpf_obj_drop will decr refcount to >= 1, never
resulting in a free
So although we may do bpf_obj_drop during the critical section, this
will never result in memory being free'd, and no changes to non-owning
ref logic are needed in this patch.
Signed-off-by: Dave Marchevsky <davemarchevsky@fb.com>
Link: https://lore.kernel.org/r/20230415201811.343116-6-davemarchevsky@fb.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
||
|
Andrii Nakryiko
|
bdcab4144f |
bpf: Simplify internal verifier log interface
Simplify internal verifier log API down to bpf_vlog_init() and bpf_vlog_finalize(). The former handles input arguments validation in one place and makes it easier to change it. The latter subsumes -ENOSPC (truncation) and -EFAULT handling and simplifies both caller's code (bpf_check() and btf_parse()). For btf_parse(), this patch also makes sure that verifier log finalization happens even if there is some error condition during BTF verification process prior to normal finalization step. Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Lorenz Bauer <lmb@isovalent.com> Link: https://lore.kernel.org/bpf/20230406234205.323208-14-andrii@kernel.org |
||
|
Andrii Nakryiko
|
fa1c7d5cc4 |
bpf: Keep track of total log content size in both fixed and rolling modes
Change how we do accounting in BPF_LOG_FIXED mode and adopt log->end_pos as *logical* log position. This means that we can go beyond physical log buffer size now and be able to tell what log buffer size should be to fit entire log contents without -ENOSPC. To do this for BPF_LOG_FIXED mode, we need to remove a short-circuiting logic of not vsnprintf()'ing further log content once we filled up user-provided buffer, which is done by bpf_verifier_log_needed() checks. We modify these checks to always keep going if log->level is non-zero (i.e., log is requested), even if log->ubuf was NULL'ed out due to copying data to user-space, or if entire log buffer is physically full. We adopt bpf_verifier_vlog() routine to work correctly with log->ubuf == NULL condition, performing log formatting into temporary kernel buffer, doing all the necessary accounting, but just avoiding copying data out if buffer is full or NULL'ed out. With these changes, it's now possible to do this sort of determination of log contents size in both BPF_LOG_FIXED and default rolling log mode. We need to keep in mind bpf_vlog_reset(), though, which shrinks log contents after successful verification of a particular code path. This log reset means that log->end_pos isn't always increasing, so to return back to users what should be the log buffer size to fit all log content without causing -ENOSPC even in the presence of log resetting, we need to keep maximum over "lifetime" of logging. We do this accounting in bpf_vlog_update_len_max() helper. A related and subtle aspect is that with this logical log->end_pos even in BPF_LOG_FIXED mode we could temporary "overflow" buffer, but then reset it back with bpf_vlog_reset() to a position inside user-supplied log_buf. In such situation we still want to properly maintain terminating zero. We will eventually return -ENOSPC even if final log buffer is small (we detect this through log->len_max check). This behavior is simpler to reason about and is consistent with current behavior of verifier log. Handling of this required a small addition to bpf_vlog_reset() logic to avoid doing put_user() beyond physical log buffer dimensions. Another issue to keep in mind is that we limit log buffer size to 32-bit value and keep such log length as u32, but theoretically verifier could produce huge log stretching beyond 4GB. Instead of keeping (and later returning) 64-bit log length, we cap it at UINT_MAX. Current UAPI makes it impossible to specify log buffer size bigger than 4GB anyways, so we don't really loose anything here and keep everything consistently 32-bit in UAPI. This property will be utilized in next patch. Doing the same determination of maximum log buffer for rolling mode is trivial, as log->end_pos and log->start_pos are already logical positions, so there is nothing new there. These changes do incidentally fix one small issue with previous logging logic. Previously, if use provided log buffer of size N, and actual log output was exactly N-1 bytes + terminating \0, kernel logic coun't distinguish this condition from log truncation scenario which would end up with truncated log contents of N-1 bytes + terminating \0 as well. But now with log->end_pos being logical position that could go beyond actual log buffer size, we can distinguish these two conditions, which we do in this patch. This plays nicely with returning log_size_actual (implemented in UAPI in the next patch), as we can now guarantee that if user takes such log_size_actual and provides log buffer of that exact size, they will not get -ENOSPC in return. All in all, all these changes do conceptually unify fixed and rolling log modes much better, and allow a nice feature requested by users: knowing what should be the size of the buffer to avoid -ENOSPC. We'll plumb this through the UAPI and the code in the next patch. Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Lorenz Bauer <lmb@isovalent.com> Link: https://lore.kernel.org/bpf/20230406234205.323208-12-andrii@kernel.org |
||
|
Andrii Nakryiko
|
1216640938 |
bpf: Switch BPF verifier log to be a rotating log by default
Currently, if user-supplied log buffer to collect BPF verifier log turns out to be too small to contain full log, bpf() syscall returns -ENOSPC, fails BPF program verification/load, and preserves first N-1 bytes of the verifier log (where N is the size of user-supplied buffer). This is problematic in a bunch of common scenarios, especially when working with real-world BPF programs that tend to be pretty complex as far as verification goes and require big log buffers. Typically, it's when debugging tricky cases at log level 2 (verbose). Also, when BPF program is successfully validated, log level 2 is the only way to actually see verifier state progression and all the important details. Even with log level 1, it's possible to get -ENOSPC even if the final verifier log fits in log buffer, if there is a code path that's deep enough to fill up entire log, even if normally it would be reset later on (there is a logic to chop off successfully validated portions of BPF verifier log). In short, it's not always possible to pre-size log buffer. Also, what's worse, in practice, the end of the log most often is way more important than the beginning, but verifier stops emitting log as soon as initial log buffer is filled up. This patch switches BPF verifier log behavior to effectively behave as rotating log. That is, if user-supplied log buffer turns out to be too short, verifier will keep overwriting previously written log, effectively treating user's log buffer as a ring buffer. -ENOSPC is still going to be returned at the end, to notify user that log contents was truncated, but the important last N bytes of the log would be returned, which might be all that user really needs. This consistent -ENOSPC behavior, regardless of rotating or fixed log behavior, allows to prevent backwards compatibility breakage. The only user-visible change is which portion of verifier log user ends up seeing *if buffer is too small*. Given contents of verifier log itself is not an ABI, there is no breakage due to this behavior change. Specialized tools that rely on specific contents of verifier log in -ENOSPC scenario are expected to be easily adapted to accommodate old and new behaviors. Importantly, though, to preserve good user experience and not require every user-space application to adopt to this new behavior, before exiting to user-space verifier will rotate log (in place) to make it start at the very beginning of user buffer as a continuous zero-terminated string. The contents will be a chopped off N-1 last bytes of full verifier log, of course. Given beginning of log is sometimes important as well, we add BPF_LOG_FIXED (which equals 8) flag to force old behavior, which allows tools like veristat to request first part of verifier log, if necessary. BPF_LOG_FIXED flag is also a simple and straightforward way to check if BPF verifier supports rotating behavior. On the implementation side, conceptually, it's all simple. We maintain 64-bit logical start and end positions. If we need to truncate the log, start position will be adjusted accordingly to lag end position by N bytes. We then use those logical positions to calculate their matching actual positions in user buffer and handle wrap around the end of the buffer properly. Finally, right before returning from bpf_check(), we rotate user log buffer contents in-place as necessary, to make log contents contiguous. See comments in relevant functions for details. Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Reviewed-by: Lorenz Bauer <lmb@isovalent.com> Link: https://lore.kernel.org/bpf/20230406234205.323208-4-andrii@kernel.org |
||
|
Andrii Nakryiko
|
4294a0a7ab |
bpf: Split off basic BPF verifier log into separate file
kernel/bpf/verifier.c file is large and growing larger all the time. So it's good to start splitting off more or less self-contained parts into separate files to keep source code size (somewhat) somewhat under control. This patch is a one step in this direction, moving some of BPF verifier log routines into a separate kernel/bpf/log.c. Right now it's most low-level and isolated routines to append data to log, reset log to previous position, etc. Eventually we could probably move verifier state printing logic here as well, but this patch doesn't attempt to do that yet. Subsequent patches will add more logic to verifier log management, so having basics in a separate file will make sure verifier.c doesn't grow more with new changes. Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Lorenz Bauer <lmb@isovalent.com> Link: https://lore.kernel.org/bpf/20230406234205.323208-2-andrii@kernel.org |
||
|
Andrii Nakryiko
|
4b5ce570db |
bpf: ensure state checkpointing at iter_next() call sites
State equivalence check and checkpointing performed in is_state_visited() employs certain heuristics to try to save memory by avoiding state checkpoints if not enough jumps and instructions happened since last checkpoint. This leads to unpredictability of whether a particular instruction will be checkpointed and how regularly. While normally this is not causing much problems (except inconveniences for predictable verifier tests, which we overcome with BPF_F_TEST_STATE_FREQ flag), turns out it's not the case for open-coded iterators. Checking and saving state checkpoints at iter_next() call is crucial for fast convergence of open-coded iterator loop logic, so we need to force it. If we don't do that, is_state_visited() might skip saving a checkpoint, causing unnecessarily long sequence of not checkpointed instructions and jumps, leading to exhaustion of jump history buffer, and potentially other undesired outcomes. It is expected that with correct open-coded iterators convergence will happen quickly, so we don't run a risk of exhausting memory. This patch adds, in addition to prune and jump instruction marks, also a "forced checkpoint" mark, and makes sure that any iter_next() call instruction is marked as such. Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/r/20230310060149.625887-1-andrii@kernel.org Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Andrii Nakryiko
|
06accc8779 |
bpf: add support for open-coded iterator loops
Teach verifier about the concept of the open-coded (or inline) iterators.
This patch adds generic iterator loop verification logic, new STACK_ITER
stack slot type to contain iterator state, and necessary kfunc plumbing
for iterator's constructor, destructor and next methods. Next patch
implements first specific iterator (numbers iterator for implementing
for() loop logic). Such split allows to have more focused commits for
verifier logic and separate commit that we could point later to
demonstrating what does it take to add a new kind of iterator.
Each kind of iterator has its own associated struct bpf_iter_<type>,
where <type> denotes a specific type of iterator. struct bpf_iter_<type>
state is supposed to live on BPF program stack, so there will be no way
to change its size later on without breaking backwards compatibility, so
choose wisely! But given this struct is specific to a given <type> of
iterator, this allows a lot of flexibility: simple iterators could be
fine with just one stack slot (8 bytes), like numbers iterator in the
next patch, while some other more complicated iterators might need way
more to keep their iterator state. Either way, such design allows to
avoid runtime memory allocations, which otherwise would be necessary if
we fixed on-the-stack size and it turned out to be too small for a given
iterator implementation.
The way BPF verifier logic is implemented, there are no artificial
restrictions on a number of active iterators, it should work correctly
using multiple active iterators at the same time. This also means you
can have multiple nested iteration loops. struct bpf_iter_<type>
reference can be safely passed to subprograms as well.
General flow is easiest to demonstrate with a simple example using
number iterator implemented in next patch. Here's the simplest possible
loop:
struct bpf_iter_num it;
int *v;
bpf_iter_num_new(&it, 2, 5);
while ((v = bpf_iter_num_next(&it))) {
bpf_printk("X = %d", *v);
}
bpf_iter_num_destroy(&it);
Above snippet should output "X = 2", "X = 3", "X = 4". Note that 5 is
exclusive and is not returned. This matches similar APIs (e.g., slices
in Go or Rust) that implement a range of elements, where end index is
non-inclusive.
In the above example, we see a trio of function:
- constructor, bpf_iter_num_new(), which initializes iterator state
(struct bpf_iter_num it) on the stack. If any of the input arguments
are invalid, constructor should make sure to still initialize it such
that subsequent bpf_iter_num_next() calls will return NULL. I.e., on
error, return error and construct empty iterator.
- next method, bpf_iter_num_next(), which accepts pointer to iterator
state and produces an element. Next method should always return
a pointer. The contract between BPF verifier is that next method will
always eventually return NULL when elements are exhausted. Once NULL is
returned, subsequent next calls should keep returning NULL. In the
case of numbers iterator, bpf_iter_num_next() returns a pointer to an int
(storage for this integer is inside the iterator state itself),
which can be dereferenced after corresponding NULL check.
- once done with the iterator, it's mandated that user cleans up its
state with the call to destructor, bpf_iter_num_destroy() in this
case. Destructor frees up any resources and marks stack space used by
struct bpf_iter_num as usable for something else.
Any other iterator implementation will have to implement at least these
three methods. It is enforced that for any given type of iterator only
applicable constructor/destructor/next are callable. I.e., verifier
ensures you can't pass number iterator state into, say, cgroup
iterator's next method.
It is important to keep the naming pattern consistent to be able to
create generic macros to help with BPF iter usability. E.g., one
of the follow up patches adds generic bpf_for_each() macro to bpf_misc.h
in selftests, which allows to utilize iterator "trio" nicely without
having to code the above somewhat tedious loop explicitly every time.
This is enforced at kfunc registration point by one of the previous
patches in this series.
At the implementation level, iterator state tracking for verification
purposes is very similar to dynptr. We add STACK_ITER stack slot type,
reserve necessary number of slots, depending on
sizeof(struct bpf_iter_<type>), and keep track of necessary extra state
in the "main" slot, which is marked with non-zero ref_obj_id. Other
slots are also marked as STACK_ITER, but have zero ref_obj_id. This is
simpler than having a separate "is_first_slot" flag.
Another big distinction is that STACK_ITER is *always refcounted*, which
simplifies implementation without sacrificing usability. So no need for
extra "iter_id", no need to anticipate reuse of STACK_ITER slots for new
constructors, etc. Keeping it simple here.
As far as the verification logic goes, there are two extensive comments:
in process_iter_next_call() and iter_active_depths_differ() explaining
some important and sometimes subtle aspects. Please refer to them for
details.
But from 10,000-foot point of view, next methods are the points of
forking a verification state, which are conceptually similar to what
verifier is doing when validating conditional jump. We branch out at
a `call bpf_iter_<type>_next` instruction and simulate two outcomes:
NULL (iteration is done) and non-NULL (new element is returned). NULL is
simulated first and is supposed to reach exit without looping. After
that non-NULL case is validated and it either reaches exit (for trivial
examples with no real loop), or reaches another `call bpf_iter_<type>_next`
instruction with the state equivalent to already (partially) validated
one. State equivalency at that point means we technically are going to
be looping forever without "breaking out" out of established "state
envelope" (i.e., subsequent iterations don't add any new knowledge or
constraints to the verifier state, so running 1, 2, 10, or a million of
them doesn't matter). But taking into account the contract stating that
iterator next method *has to* return NULL eventually, we can conclude
that loop body is safe and will eventually terminate. Given we validated
logic outside of the loop (NULL case), and concluded that loop body is
safe (though potentially looping many times), verifier can claim safety
of the overall program logic.
The rest of the patch is necessary plumbing for state tracking, marking,
validation, and necessary further kfunc plumbing to allow implementing
iterator constructor, destructor, and next methods.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/r/20230308184121.1165081-4-andrii@kernel.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
||
|
Andrii Nakryiko
|
215bf4962f |
bpf: add iterator kfuncs registration and validation logic
Add ability to register kfuncs that implement BPF open-coded iterator
contract and enforce naming and function proto convention. Enforcement
happens at the time of kfunc registration and significantly simplifies
the rest of iterators logic in the verifier.
More details follow in subsequent patches, but we enforce the following
conditions.
All kfuncs (constructor, next, destructor) have to be named consistenly
as bpf_iter_<type>_{new,next,destroy}(), respectively. <type> represents
iterator type, and iterator state should be represented as a matching
`struct bpf_iter_<type>` state type. Also, all iter kfuncs should have
a pointer to this `struct bpf_iter_<type>` as the very first argument.
Additionally:
- Constructor, i.e., bpf_iter_<type>_new(), can have arbitrary extra
number of arguments. Return type is not enforced either.
- Next method, i.e., bpf_iter_<type>_next(), has to return a pointer
type and should have exactly one argument: `struct bpf_iter_<type> *`
(const/volatile/restrict and typedefs are ignored).
- Destructor, i.e., bpf_iter_<type>_destroy(), should return void and
should have exactly one argument, similar to the next method.
- struct bpf_iter_<type> size is enforced to be positive and
a multiple of 8 bytes (to fit stack slots correctly).
Such strictness and consistency allows to build generic helpers
abstracting important, but boilerplate, details to be able to use
open-coded iterators effectively and ergonomically (see bpf_for_each()
in subsequent patches). It also simplifies the verifier logic in some
places. At the same time, this doesn't hurt generality of possible
iterator implementations. Win-win.
Constructor kfunc is marked with a new KF_ITER_NEW flags, next method is
marked with KF_ITER_NEXT (and should also have KF_RET_NULL, of course),
while destructor kfunc is marked as KF_ITER_DESTROY.
Additionally, we add a trivial kfunc name validation: it should be
a valid non-NULL and non-empty string.
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/r/20230308184121.1165081-3-andrii@kernel.org
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
||
|
Alexei Starovoitov
|
6fcd486b3a |
bpf: Refactor RCU enforcement in the verifier.
bpf_rcu_read_lock/unlock() are only available in clang compiled kernels. Lack of such key mechanism makes it impossible for sleepable bpf programs to use RCU pointers. Allow bpf_rcu_read_lock/unlock() in GCC compiled kernels (though GCC doesn't support btf_type_tag yet) and allowlist certain field dereferences in important data structures like tast_struct, cgroup, socket that are used by sleepable programs either as RCU pointer or full trusted pointer (which is valid outside of RCU CS). Use BTF_TYPE_SAFE_RCU and BTF_TYPE_SAFE_TRUSTED macros for such tagging. They will be removed once GCC supports btf_type_tag. With that refactor check_ptr_to_btf_access(). Make it strict in enforcing PTR_TRUSTED and PTR_UNTRUSTED while deprecating old PTR_TO_BTF_ID without modifier flags. There is a chance that this strict enforcement might break existing programs (especially on GCC compiled kernels), but this cleanup has to start sooner than later. Note PTR_TO_CTX access still yields old deprecated PTR_TO_BTF_ID. Once it's converted to strict PTR_TRUSTED or PTR_UNTRUSTED the kfuncs and helpers will be able to default to KF_TRUSTED_ARGS. KF_RCU will remain as a weaker version of KF_TRUSTED_ARGS where obj refcnt could be 0. Adjust rcu_read_lock selftest to run on gcc and clang compiled kernels. Signed-off-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: David Vernet <void@manifault.com> Link: https://lore.kernel.org/bpf/20230303041446.3630-7-alexei.starovoitov@gmail.com |
||
|
Joanne Koong
|
7e0dac2807 |
bpf: Refactor process_dynptr_func
This change cleans up process_dynptr_func's flow to be more intuitive and updates some comments with more context. Signed-off-by: Joanne Koong <joannelkoong@gmail.com> Link: https://lore.kernel.org/r/20230301154953.641654-3-joannelkoong@gmail.com Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Dave Marchevsky
|
6a3cd3318f |
bpf: Migrate release_on_unlock logic to non-owning ref semantics
This patch introduces non-owning reference semantics to the verifier,
specifically linked_list API kfunc handling. release_on_unlock logic for
refs is refactored - with small functional changes - to implement these
semantics, and bpf_list_push_{front,back} are migrated to use them.
When a list node is pushed to a list, the program still has a pointer to
the node:
n = bpf_obj_new(typeof(*n));
bpf_spin_lock(&l);
bpf_list_push_back(&l, n);
/* n still points to the just-added node */
bpf_spin_unlock(&l);
What the verifier considers n to be after the push, and thus what can be
done with n, are changed by this patch.
Common properties both before/after this patch:
* After push, n is only a valid reference to the node until end of
critical section
* After push, n cannot be pushed to any list
* After push, the program can read the node's fields using n
Before:
* After push, n retains the ref_obj_id which it received on
bpf_obj_new, but the associated bpf_reference_state's
release_on_unlock field is set to true
* release_on_unlock field and associated logic is used to implement
"n is only a valid ref until end of critical section"
* After push, n cannot be written to, the node must be removed from
the list before writing to its fields
* After push, n is marked PTR_UNTRUSTED
After:
* After push, n's ref is released and ref_obj_id set to 0. NON_OWN_REF
type flag is added to reg's type, indicating that it's a non-owning
reference.
* NON_OWN_REF flag and logic is used to implement "n is only a
valid ref until end of critical section"
* n can be written to (except for special fields e.g. bpf_list_node,
timer, ...)
Summary of specific implementation changes to achieve the above:
* release_on_unlock field, ref_set_release_on_unlock helper, and logic
to "release on unlock" based on that field are removed
* The anonymous active_lock struct used by bpf_verifier_state is
pulled out into a named struct bpf_active_lock.
* NON_OWN_REF type flag is introduced along with verifier logic
changes to handle non-owning refs
* Helpers are added to use NON_OWN_REF flag to implement non-owning
ref semantics as described above
* invalidate_non_owning_refs - helper to clobber all non-owning refs
matching a particular bpf_active_lock identity. Replaces
release_on_unlock logic in process_spin_lock.
* ref_set_non_owning - set NON_OWN_REF type flag after doing some
sanity checking
* ref_convert_owning_non_owning - convert owning reference w/
specified ref_obj_id to non-owning references. Set NON_OWN_REF
flag for each reg with that ref_obj_id and 0-out its ref_obj_id
* Update linked_list selftests to account for minor semantic
differences introduced by this patch
* Writes to a release_on_unlock node ref are not allowed, while
writes to non-owning reference pointees are. As a result the
linked_list "write after push" failure tests are no longer scenarios
that should fail.
* The test##missing_lock##op and test##incorrect_lock##op
macro-generated failure tests need to have a valid node argument in
order to have the same error output as before. Otherwise
verification will fail early and the expected error output won't be seen.
Signed-off-by: Dave Marchevsky <davemarchevsky@fb.com>
Link: https://lore.kernel.org/r/20230212092715.1422619-2-davemarchevsky@fb.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
||
|
Kumar Kartikeya Dwivedi
|
f8064ab90d |
bpf: Invalidate slices on destruction of dynptrs on stack
The previous commit implemented destroy_if_dynptr_stack_slot. It destroys the dynptr which given spi belongs to, but still doesn't invalidate the slices that belong to such a dynptr. While for the case of referenced dynptr, we don't allow their overwrite and return an error early, we still allow it and destroy the dynptr for unreferenced dynptr. To be able to enable precise and scoped invalidation of dynptr slices in this case, we must be able to associate the source dynptr of slices that have been obtained using bpf_dynptr_data. When doing destruction, only slices belonging to the dynptr being destructed should be invalidated, and nothing else. Currently, dynptr slices belonging to different dynptrs are indistinguishible. Hence, allocate a unique id to each dynptr (CONST_PTR_TO_DYNPTR and those on stack). This will be stored as part of reg->id. Whenever using bpf_dynptr_data, transfer this unique dynptr id to the returned PTR_TO_MEM_OR_NULL slice pointer, and store it in a new per-PTR_TO_MEM dynptr_id register state member. Finally, after establishing such a relationship between dynptrs and their slices, implement precise invalidation logic that only invalidates slices belong to the destroyed dynptr in destroy_if_dynptr_stack_slot. Acked-by: Joanne Koong <joannelkoong@gmail.com> Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com> Link: https://lore.kernel.org/r/20230121002241.2113993-5-memxor@gmail.com Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Andrii Nakryiko
|
a73bf9f2d9 |
bpf: reorganize struct bpf_reg_state fields
Move id and ref_obj_id fields after scalar data section (var_off and ranges). This is necessary to simplify next patch which will change regsafe()'s logic to be safer, as it makes the contents that has to be an exact match (type-specific parts, off, type, and var_off+ranges) a single sequential block of memory, while id and ref_obj_id should always be remapped and thus can't be memcp()'ed. There are few places that assume that var_off is after id/ref_obj_id to clear out id/ref_obj_id with the single memset(0). These are changed to explicitly zero-out id/ref_obj_id fields. Other places are adjusted to preserve exact byte-by-byte comparison behavior. No functional changes. Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/r/20221223054921.958283-3-andrii@kernel.org Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Eduard Zingerman
|
5dd9cdbc9d |
bpf: states_equal() must build idmap for all function frames
verifier.c:states_equal() must maintain register ID mapping across all
function frames. Otherwise the following example might be erroneously
marked as safe:
main:
fp[-24] = map_lookup_elem(...) ; frame[0].fp[-24].id == 1
fp[-32] = map_lookup_elem(...) ; frame[0].fp[-32].id == 2
r1 = &fp[-24]
r2 = &fp[-32]
call foo()
r0 = 0
exit
foo:
0: r9 = r1
1: r8 = r2
2: r7 = ktime_get_ns()
3: r6 = ktime_get_ns()
4: if (r6 > r7) goto skip_assign
5: r9 = r8
skip_assign: ; <--- checkpoint
6: r9 = *r9 ; (a) frame[1].r9.id == 2
; (b) frame[1].r9.id == 1
7: if r9 == 0 goto exit: ; mark_ptr_or_null_regs() transfers != 0 info
; for all regs sharing ID:
; (a) r9 != 0 => &frame[0].fp[-32] != 0
; (b) r9 != 0 => &frame[0].fp[-24] != 0
8: r8 = *r8 ; (a) r8 == &frame[0].fp[-32]
; (b) r8 == &frame[0].fp[-32]
9: r0 = *r8 ; (a) safe
; (b) unsafe
exit:
10: exit
While processing call to foo() verifier considers the following
execution paths:
(a) 0-10
(b) 0-4,6-10
(There is also path 0-7,10 but it is not interesting for the issue at
hand. (a) is verified first.)
Suppose that checkpoint is created at (6) when path (a) is verified,
next path (b) is verified and (6) is reached.
If states_equal() maintains separate 'idmap' for each frame the
mapping at (6) for frame[1] would be empty and
regsafe(r9)::check_ids() would add a pair 2->1 and return true,
which is an error.
If states_equal() maintains single 'idmap' for all frames the mapping
at (6) would be { 1->1, 2->2 } and regsafe(r9)::check_ids() would
return false when trying to add a pair 2->1.
This issue was suggested in the following discussion:
https://lore.kernel.org/bpf/CAEf4BzbFB5g4oUfyxk9rHy-PJSLQ3h8q9mV=rVoXfr_JVm8+1Q@mail.gmail.com/
Suggested-by: Andrii Nakryiko <andrii.nakryiko@gmail.com>
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Link: https://lore.kernel.org/r/20221209135733.28851-4-eddyz87@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
||
|
Kumar Kartikeya Dwivedi
|
6b75bd3d03 |
bpf: Refactor ARG_PTR_TO_DYNPTR checks into process_dynptr_func
ARG_PTR_TO_DYNPTR is akin to ARG_PTR_TO_TIMER, ARG_PTR_TO_KPTR, where the underlying register type is subjected to more special checks to determine the type of object represented by the pointer and its state consistency. Move dynptr checks to their own 'process_dynptr_func' function so that is consistent and in-line with existing code. This also makes it easier to reuse this code for kfunc handling. Then, reuse this consolidated function in kfunc dynptr handling too. Note that for kfuncs, the arg_type constraint of DYNPTR_TYPE_LOCAL has been lifted. Acked-by: David Vernet <void@manifault.com> Acked-by: Joanne Koong <joannelkoong@gmail.com> Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com> Link: https://lore.kernel.org/r/20221207204141.308952-2-memxor@gmail.com Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Andrii Nakryiko
|
bffdeaa8a5 |
bpf: decouple prune and jump points
BPF verifier marks some instructions as prune points. Currently these prune points serve two purposes. It's a point where verifier tries to find previously verified state and check current state's equivalence to short circuit verification for current code path. But also currently it's a point where jump history, used for precision backtracking, is updated. This is done so that non-linear flow of execution could be properly backtracked. Such coupling is coincidental and unnecessary. Some prune points are not part of some non-linear jump path, so don't need update of jump history. On the other hand, not all instructions which have to be recorded in jump history necessarily are good prune points. This patch splits prune and jump points into independent flags. Currently all prune points are marked as jump points to minimize amount of changes in this patch, but next patch will perform some optimization of prune vs jmp point placement. No functional changes are intended. Acked-by: John Fastabend <john.fastabend@gmail.com> Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/r/20221206233345.438540-2-andrii@kernel.org Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Yonghong Song
|
fca1aa7551 |
bpf: Handle MEM_RCU type properly
Commit |
||
|
Alexei Starovoitov
|
c67cae551f |
bpf: Tighten ptr_to_btf_id checks.
The networking programs typically don't require CAP_PERFMON, but through kfuncs like bpf_cast_to_kern_ctx() they can access memory through PTR_TO_BTF_ID. In such case enforce CAP_PERFMON. Also make sure that only GPL programs can access kernel data structures. All kfuncs require GPL already. Also remove allow_ptr_to_map_access. It's the same as allow_ptr_leaks and different name for the same check only causes confusion. Fixes: |
||
|
Yonghong Song
|
9bb00b2895 |
bpf: Add kfunc bpf_rcu_read_lock/unlock()
Add two kfunc's bpf_rcu_read_lock() and bpf_rcu_read_unlock(). These two kfunc's
can be used for all program types. The following is an example about how
rcu pointer are used w.r.t. bpf_rcu_read_lock()/bpf_rcu_read_unlock().
struct task_struct {
...
struct task_struct *last_wakee;
struct task_struct __rcu *real_parent;
...
};
Let us say prog does 'task = bpf_get_current_task_btf()' to get a
'task' pointer. The basic rules are:
- 'real_parent = task->real_parent' should be inside bpf_rcu_read_lock
region. This is to simulate rcu_dereference() operation. The
'real_parent' is marked as MEM_RCU only if (1). task->real_parent is
inside bpf_rcu_read_lock region, and (2). task is a trusted ptr. So
MEM_RCU marked ptr can be 'trusted' inside the bpf_rcu_read_lock region.
- 'last_wakee = real_parent->last_wakee' should be inside bpf_rcu_read_lock
region since it tries to access rcu protected memory.
- the ptr 'last_wakee' will be marked as PTR_UNTRUSTED since in general
it is not clear whether the object pointed by 'last_wakee' is valid or
not even inside bpf_rcu_read_lock region.
The verifier will reset all rcu pointer register states to untrusted
at bpf_rcu_read_unlock() kfunc call site, so any such rcu pointer
won't be trusted any more outside the bpf_rcu_read_lock() region.
The current implementation does not support nested rcu read lock
region in the prog.
Acked-by: Martin KaFai Lau <martin.lau@kernel.org>
Signed-off-by: Yonghong Song <yhs@fb.com>
Link: https://lore.kernel.org/r/20221124053217.2373910-1-yhs@fb.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
||
|
David Vernet
|
3f00c52393 |
bpf: Allow trusted pointers to be passed to KF_TRUSTED_ARGS kfuncs
Kfuncs currently support specifying the KF_TRUSTED_ARGS flag to signal
to the verifier that it should enforce that a BPF program passes it a
"safe", trusted pointer. Currently, "safe" means that the pointer is
either PTR_TO_CTX, or is refcounted. There may be cases, however, where
the kernel passes a BPF program a safe / trusted pointer to an object
that the BPF program wishes to use as a kptr, but because the object
does not yet have a ref_obj_id from the perspective of the verifier, the
program would be unable to pass it to a KF_ACQUIRE | KF_TRUSTED_ARGS
kfunc.
The solution is to expand the set of pointers that are considered
trusted according to KF_TRUSTED_ARGS, so that programs can invoke kfuncs
with these pointers without getting rejected by the verifier.
There is already a PTR_UNTRUSTED flag that is set in some scenarios,
such as when a BPF program reads a kptr directly from a map
without performing a bpf_kptr_xchg() call. These pointers of course can
and should be rejected by the verifier. Unfortunately, however,
PTR_UNTRUSTED does not cover all the cases for safety that need to
be addressed to adequately protect kfuncs. Specifically, pointers
obtained by a BPF program "walking" a struct are _not_ considered
PTR_UNTRUSTED according to BPF. For example, say that we were to add a
kfunc called bpf_task_acquire(), with KF_ACQUIRE | KF_TRUSTED_ARGS, to
acquire a struct task_struct *. If we only used PTR_UNTRUSTED to signal
that a task was unsafe to pass to a kfunc, the verifier would mistakenly
allow the following unsafe BPF program to be loaded:
SEC("tp_btf/task_newtask")
int BPF_PROG(unsafe_acquire_task,
struct task_struct *task,
u64 clone_flags)
{
struct task_struct *acquired, *nested;
nested = task->last_wakee;
/* Would not be rejected by the verifier. */
acquired = bpf_task_acquire(nested);
if (!acquired)
return 0;
bpf_task_release(acquired);
return 0;
}
To address this, this patch defines a new type flag called PTR_TRUSTED
which tracks whether a PTR_TO_BTF_ID pointer is safe to pass to a
KF_TRUSTED_ARGS kfunc or a BPF helper function. PTR_TRUSTED pointers are
passed directly from the kernel as a tracepoint or struct_ops callback
argument. Any nested pointer that is obtained from walking a PTR_TRUSTED
pointer is no longer PTR_TRUSTED. From the example above, the struct
task_struct *task argument is PTR_TRUSTED, but the 'nested' pointer
obtained from 'task->last_wakee' is not PTR_TRUSTED.
A subsequent patch will add kfuncs for storing a task kfunc as a kptr,
and then another patch will add selftests to validate.
Signed-off-by: David Vernet <void@manifault.com>
Link: https://lore.kernel.org/r/20221120051004.3605026-3-void@manifault.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
||
|
David Vernet
|
ef66c5475d |
bpf: Allow multiple modifiers in reg_type_str() prefix
reg_type_str() in the verifier currently only allows a single register type modifier to be present in the 'prefix' string which is eventually stored in the env type_str_buf. This currently works fine because there are no overlapping type modifiers, but once PTR_TRUSTED is added, that will no longer be the case. This patch updates reg_type_str() to support having multiple modifiers in the prefix string, and updates the size of type_str_buf to be 128 bytes. Signed-off-by: David Vernet <void@manifault.com> Link: https://lore.kernel.org/r/20221120051004.3605026-2-void@manifault.com Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Kumar Kartikeya Dwivedi
|
534e86bc6c |
bpf: Add 'release on unlock' logic for bpf_list_push_{front,back}
This commit implements the delayed release logic for bpf_list_push_front and bpf_list_push_back. Once a node has been added to the list, it's pointer changes to PTR_UNTRUSTED. However, it is only released once the lock protecting the list is unlocked. For such PTR_TO_BTF_ID | MEM_ALLOC with PTR_UNTRUSTED set but an active ref_obj_id, it is still permitted to read them as long as the lock is held. Writing to them is not allowed. This allows having read access to push items we no longer own until we release the lock guarding the list, allowing a little more flexibility when working with these APIs. Note that enabling write support has fairly tricky interactions with what happens inside the critical section. Just as an example, currently, bpf_obj_drop is not permitted, but if it were, being able to write to the PTR_UNTRUSTED pointer while the object gets released back to the memory allocator would violate safety properties we wish to guarantee (i.e. not crashing the kernel). The memory could be reused for a different type in the BPF program or even in the kernel as it gets eventually kfree'd. Not enabling bpf_obj_drop inside the critical section would appear to prevent all of the above, but that is more of an artifical limitation right now. Since the write support is tangled with how we handle potential aliasing of nodes inside the critical section that may or may not be part of the list anymore, it has been deferred to a future patch. Acked-by: Dave Marchevsky <davemarchevsky@fb.com> Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com> Link: https://lore.kernel.org/r/20221118015614.2013203-18-memxor@gmail.com Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Kumar Kartikeya Dwivedi
|
958cf2e273 |
bpf: Introduce bpf_obj_new
Introduce type safe memory allocator bpf_obj_new for BPF programs. The
kernel side kfunc is named bpf_obj_new_impl, as passing hidden arguments
to kfuncs still requires having them in prototype, unlike BPF helpers
which always take 5 arguments and have them checked using bpf_func_proto
in verifier, ignoring unset argument types.
Introduce __ign suffix to ignore a specific kfunc argument during type
checks, then use this to introduce support for passing type metadata to
the bpf_obj_new_impl kfunc.
The user passes BTF ID of the type it wants to allocates in program BTF,
the verifier then rewrites the first argument as the size of this type,
after performing some sanity checks (to ensure it exists and it is a
struct type).
The second argument is also fixed up and passed by the verifier. This is
the btf_struct_meta for the type being allocated. It would be needed
mostly for the offset array which is required for zero initializing
special fields while leaving the rest of storage in unitialized state.
It would also be needed in the next patch to perform proper destruction
of the object's special fields.
Under the hood, bpf_obj_new will call bpf_mem_alloc and bpf_mem_free,
using the any context BPF memory allocator introduced recently. To this
end, a global instance of the BPF memory allocator is initialized on
boot to be used for this purpose. This 'bpf_global_ma' serves all
allocations for bpf_obj_new. In the future, bpf_obj_new variants will
allow specifying a custom allocator.
Note that now that bpf_obj_new can be used to allocate objects that can
be linked to BPF linked list (when future linked list helpers are
available), we need to also free the elements using bpf_mem_free.
However, since the draining of elements is done outside the
bpf_spin_lock, we need to do migrate_disable around the call since
bpf_list_head_free can be called from map free path where migration is
enabled. Otherwise, when called from BPF programs migration is already
disabled.
A convenience macro is included in the bpf_experimental.h header to hide
over the ugly details of the implementation, leading to user code
looking similar to a language level extension which allocates and
constructs fields of a user type.
struct bar {
struct bpf_list_node node;
};
struct foo {
struct bpf_spin_lock lock;
struct bpf_list_head head __contains(bar, node);
};
void prog(void) {
struct foo *f;
f = bpf_obj_new(typeof(*f));
if (!f)
return;
...
}
A key piece of this story is still missing, i.e. the free function,
which will come in the next patch.
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Link: https://lore.kernel.org/r/20221118015614.2013203-14-memxor@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
||
|
Kumar Kartikeya Dwivedi
|
00b85860fe |
bpf: Rewrite kfunc argument handling
As we continue to add more features, argument types, kfunc flags, and different extensions to kfuncs, the code to verify the correctness of the kfunc prototype wrt the passed in registers has become ad-hoc and ugly to read. To make life easier, and make a very clear split between different stages of argument processing, move all the code into verifier.c and refactor into easier to read helpers and functions. This also makes sharing code within the verifier easier with kfunc argument processing. This will be more and more useful in later patches as we are now moving to implement very core BPF helpers as kfuncs, to keep them experimental before baking into UAPI. Remove all kfunc related bits now from btf_check_func_arg_match, as users have been converted away to refactored kfunc argument handling. Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com> Link: https://lore.kernel.org/r/20221118015614.2013203-12-memxor@gmail.com Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Kumar Kartikeya Dwivedi
|
d0d78c1df9 |
bpf: Allow locking bpf_spin_lock global variables
Global variables reside in maps accessible using direct_value_addr
callbacks, so giving each load instruction's rewrite a unique reg->id
disallows us from holding locks which are global.
The reason for preserving reg->id as a unique value for registers that
may point to spin lock is that two separate lookups are treated as two
separate memory regions, and any possible aliasing is ignored for the
purposes of spin lock correctness.
This is not great especially for the global variable case, which are
served from maps that have max_entries == 1, i.e. they always lead to
map values pointing into the same map value.
So refactor the active_spin_lock into a 'active_lock' structure which
represents the lock identity, and instead of the reg->id, remember two
fields, a pointer and the reg->id. The pointer will store reg->map_ptr
or reg->btf. It's only necessary to distinguish for the id == 0 case of
global variables, but always setting the pointer to a non-NULL value and
using the pointer to check whether the lock is held simplifies code in
the verifier.
This is generic enough to allow it for global variables, map lookups,
and allocated objects at the same time.
Note that while whether a lock is held can be answered by just comparing
active_lock.ptr to NULL, to determine whether the register is pointing
to the same held lock requires comparing _both_ ptr and id.
Finally, as a result of this refactoring, pseudo load instructions are
not given a unique reg->id, as they are doing lookup for the same map
value (max_entries is never greater than 1).
Essentially, we consider that the tuple of (ptr, id) will always be
unique for any kind of argument to bpf_spin_{lock,unlock}.
Note that this can be extended in the future to also remember offset
used for locking, so that we can introduce multiple bpf_spin_lock fields
in the same allocation.
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Link: https://lore.kernel.org/r/20221118015614.2013203-10-memxor@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
||
|
Martin KaFai Lau
|
271de525e1 |
bpf: Remove prog->active check for bpf_lsm and bpf_iter
The commit |
||
|
Roberto Sassu
|
b8d31762a0 |
btf: Allow dynamic pointer parameters in kfuncs
Allow dynamic pointers (struct bpf_dynptr_kern *) to be specified as parameters in kfuncs. Also, ensure that dynamic pointers passed as argument are valid and initialized, are a pointer to the stack, and of the type local. More dynamic pointer types can be supported in the future. To properly detect whether a parameter is of the desired type, introduce the stringify_struct() macro to compare the returned structure name with the desired name. In addition, protect against structure renames, by halting the build with BUILD_BUG_ON(), so that developers have to revisit the code. To check if a dynamic pointer passed to the kfunc is valid and initialized, and if its type is local, export the existing functions is_dynptr_reg_valid_init() and is_dynptr_type_expected(). Cc: Joanne Koong <joannelkoong@gmail.com> Cc: Kumar Kartikeya Dwivedi <memxor@gmail.com> Signed-off-by: Roberto Sassu <roberto.sassu@huawei.com> Acked-by: Kumar Kartikeya Dwivedi <memxor@gmail.com> Link: https://lore.kernel.org/r/20220920075951.929132-5-roberto.sassu@huaweicloud.com Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Dave Marchevsky
|
1bfe26fb08 |
bpf: Add verifier support for custom callback return range
Verifier logic to confirm that a callback function returns 0 or 1 was
added in commit
|
||
|
Kumar Kartikeya Dwivedi
|
b239da3420 |
bpf: Add helper macro bpf_for_each_reg_in_vstate
For a lot of use cases in future patches, we will want to modify the state of registers part of some same 'group' (e.g. same ref_obj_id). It won't just be limited to releasing reference state, but setting a type flag dynamically based on certain actions, etc. Hence, we need a way to easily pass a callback to the function that iterates over all registers in current bpf_verifier_state in all frames upto (and including) the curframe. While in C++ we would be able to easily use a lambda to pass state and the callback together, sadly we aren't using C++ in the kernel. The next best thing to avoid defining a function for each case seems like statement expressions in GNU C. The kernel already uses them heavily, hence they can passed to the macro in the style of a lambda. The statement expression will then be substituted in the for loop bodies. Variables __state and __reg are set to current bpf_func_state and reg for each invocation of the expression inside the passed in verifier state. Then, convert mark_ptr_or_null_regs, clear_all_pkt_pointers, release_reference, find_good_pkt_pointers, find_equal_scalars to use bpf_for_each_reg_in_vstate. Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com> Link: https://lore.kernel.org/r/20220904204145.3089-16-memxor@gmail.com Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Benjamin Tissoires
|
eb1f7f71c1 |
bpf/verifier: allow kfunc to return an allocated mem
For drivers (outside of network), the incoming data is not statically defined in a struct. Most of the time the data buffer is kzalloc-ed and thus we can not rely on eBPF and BTF to explore the data. This commit allows to return an arbitrary memory, previously allocated by the driver. An interesting extra point is that the kfunc can mark the exported memory region as read only or read/write. So, when a kfunc is not returning a pointer to a struct but to a plain type, we can consider it is a valid allocated memory assuming that: - one of the arguments is either called rdonly_buf_size or rdwr_buf_size - and this argument is a const from the caller point of view We can then use this parameter as the size of the allocated memory. The memory is either read-only or read-write based on the name of the size parameter. Acked-by: Kumar Kartikeya Dwivedi <memxor@gmail.com> Signed-off-by: Benjamin Tissoires <benjamin.tissoires@redhat.com> Link: https://lore.kernel.org/r/20220906151303.2780789-7-benjamin.tissoires@redhat.com Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Kumar Kartikeya Dwivedi
|
9d9d00ac29 |
bpf: Fix reference state management for synchronous callbacks
Currently, verifier verifies callback functions (sync and async) as if
they will be executed once, (i.e. it explores execution state as if the
function was being called once). The next insn to explore is set to
start of subprog and the exit from nested frame is handled using
curframe > 0 and prepare_func_exit. In case of async callback it uses a
customized variant of push_stack simulating a kind of branch to set up
custom state and execution context for the async callback.
While this approach is simple and works when callback really will be
executed only once, it is unsafe for all of our current helpers which
are for_each style, i.e. they execute the callback multiple times.
A callback releasing acquired references of the caller may do so
multiple times, but currently verifier sees it as one call inside the
frame, which then returns to caller. Hence, it thinks it released some
reference that the cb e.g. got access through callback_ctx (register
filled inside cb from spilled typed register on stack).
Similarly, it may see that an acquire call is unpaired inside the
callback, so the caller will copy the reference state of callback and
then will have to release the register with new ref_obj_ids. But again,
the callback may execute multiple times, but the verifier will only
account for acquired references for a single symbolic execution of the
callback, which will cause leaks.
Note that for async callback case, things are different. While currently
we have bpf_timer_set_callback which only executes it once, even for
multiple executions it would be safe, as reference state is NULL and
check_reference_leak would force program to release state before
BPF_EXIT. The state is also unaffected by analysis for the caller frame.
Hence async callback is safe.
Since we want the reference state to be accessible, e.g. for pointers
loaded from stack through callback_ctx's PTR_TO_STACK, we still have to
copy caller's reference_state to callback's bpf_func_state, but we
enforce that whatever references it adds to that reference_state has
been released before it hits BPF_EXIT. This requires introducing a new
callback_ref member in the reference state to distinguish between caller
vs callee references. Hence, check_reference_leak now errors out if it
sees we are in callback_fn and we have not released callback_ref refs.
Since there can be multiple nested callbacks, like frame 0 -> cb1 -> cb2
etc. we need to also distinguish between whether this particular ref
belongs to this callback frame or parent, and only error for our own, so
we store state->frameno (which is always non-zero for callbacks).
In short, callbacks can read parent reference_state, but cannot mutate
it, to be able to use pointers acquired by the caller. They must only
undo their changes (by releasing their own acquired_refs before
BPF_EXIT) on top of caller reference_state before returning (at which
point the caller and callback state will match anyway, so no need to
copy it back to caller).
Fixes:
|
||
|
Matthieu Baerts
|
f16214c102 |
bpf: Fix 'dubious one-bit signed bitfield' warnings
Our CI[1] reported these warnings when using Sparse:
$ touch net/mptcp/bpf.c
$ make C=1 net/mptcp/bpf.o
net/mptcp/bpf.c: note: in included file:
include/linux/bpf_verifier.h:348:26: error: dubious one-bit signed bitfield
include/linux/bpf_verifier.h:349:29: error: dubious one-bit signed bitfield
Set them as 'unsigned' to avoid warnings.
[1] https://github.com/multipath-tcp/mptcp_net-next/actions/runs/2643588487
Fixes:
|
||
|
Eduard Zingerman
|
1ade237119 |
bpf: Inline calls to bpf_loop when callback is known
Calls to `bpf_loop` are replaced with direct loops to avoid
indirection. E.g. the following:
bpf_loop(10, foo, NULL, 0);
Is replaced by equivalent of the following:
for (int i = 0; i < 10; ++i)
foo(i, NULL);
This transformation could be applied when:
- callback is known and does not change during program execution;
- flags passed to `bpf_loop` are always zero.
Inlining logic works as follows:
- During execution simulation function `update_loop_inline_state`
tracks the following information for each `bpf_loop` call
instruction:
- is callback known and constant?
- are flags constant and zero?
- Function `optimize_bpf_loop` increases stack depth for functions
where `bpf_loop` calls can be inlined and invokes `inline_bpf_loop`
to apply the inlining. The additional stack space is used to spill
registers R6, R7 and R8. These registers are used as loop counter,
loop maximal bound and callback context parameter;
Measurements using `benchs/run_bench_bpf_loop.sh` inside QEMU / KVM on
i7-4710HQ CPU show a drop in latency from 14 ns/op to 2 ns/op.
Signed-off-by: Eduard Zingerman <eddyz87@gmail.com>
Acked-by: Song Liu <songliubraving@fb.com>
Link: https://lore.kernel.org/r/20220620235344.569325-4-eddyz87@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
||
|
Hongyi Lu
|
6dbdc9f353 |
bpf: Fix spelling in bpf_verifier.h
Minor spelling fix spotted in bpf_verifier.h. Spelling is no big deal, but it is still an improvement when reading through the code. Signed-off-by: Hongyi Lu <jwnhy0@gmail.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Link: https://lore.kernel.org/bpf/20220613211633.58647-1-jwnhy0@gmail.com |
||
|
Joanne Koong
|
bc34dee65a |
bpf: Dynptr support for ring buffers
Currently, our only way of writing dynamically-sized data into a ring buffer is through bpf_ringbuf_output but this incurs an extra memcpy cost. bpf_ringbuf_reserve + bpf_ringbuf_commit avoids this extra memcpy, but it can only safely support reservation sizes that are statically known since the verifier cannot guarantee that the bpf program won’t access memory outside the reserved space. The bpf_dynptr abstraction allows for dynamically-sized ring buffer reservations without the extra memcpy. There are 3 new APIs: long bpf_ringbuf_reserve_dynptr(void *ringbuf, u32 size, u64 flags, struct bpf_dynptr *ptr); void bpf_ringbuf_submit_dynptr(struct bpf_dynptr *ptr, u64 flags); void bpf_ringbuf_discard_dynptr(struct bpf_dynptr *ptr, u64 flags); These closely follow the functionalities of the original ringbuf APIs. For example, all ringbuffer dynptrs that have been reserved must be either submitted or discarded before the program exits. Signed-off-by: Joanne Koong <joannelkoong@gmail.com> Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Acked-by: Andrii Nakryiko <andrii@kernel.org> Acked-by: David Vernet <void@manifault.com> Link: https://lore.kernel.org/bpf/20220523210712.3641569-4-joannelkoong@gmail.com |
||
|
Joanne Koong
|
97e03f5210 |
bpf: Add verifier support for dynptrs
This patch adds the bulk of the verifier work for supporting dynamic
pointers (dynptrs) in bpf.
A bpf_dynptr is opaque to the bpf program. It is a 16-byte structure
defined internally as:
struct bpf_dynptr_kern {
void *data;
u32 size;
u32 offset;
} __aligned(8);
The upper 8 bits of *size* is reserved (it contains extra metadata about
read-only status and dynptr type). Consequently, a dynptr only supports
memory less than 16 MB.
There are different types of dynptrs (eg malloc, ringbuf, ...). In this
patchset, the most basic one, dynptrs to a bpf program's local memory,
is added. For now only local memory that is of reg type PTR_TO_MAP_VALUE
is supported.
In the verifier, dynptr state information will be tracked in stack
slots. When the program passes in an uninitialized dynptr
(ARG_PTR_TO_DYNPTR | MEM_UNINIT), the stack slots corresponding
to the frame pointer where the dynptr resides at are marked
STACK_DYNPTR. For helper functions that take in initialized dynptrs (eg
bpf_dynptr_read + bpf_dynptr_write which are added later in this
patchset), the verifier enforces that the dynptr has been initialized
properly by checking that their corresponding stack slots have been
marked as STACK_DYNPTR.
The 6th patch in this patchset adds test cases that the verifier should
successfully reject, such as for example attempting to use a dynptr
after doing a direct write into it inside the bpf program.
Signed-off-by: Joanne Koong <joannelkoong@gmail.com>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: David Vernet <void@manifault.com>
Link: https://lore.kernel.org/bpf/20220523210712.3641569-2-joannelkoong@gmail.com
|
||
|
Kumar Kartikeya Dwivedi
|
8f14852e89 |
bpf: Tag argument to be released in bpf_func_proto
Add a new type flag for bpf_arg_type that when set tells verifier that for a release function, that argument's register will be the one for which meta.ref_obj_id will be set, and which will then be released using release_reference. To capture the regno, introduce a new field release_regno in bpf_call_arg_meta. This would be required in the next patch, where we may either pass NULL or a refcounted pointer as an argument to the release function bpf_kptr_xchg. Just releasing only when meta.ref_obj_id is set is not enough, as there is a case where the type of argument needed matches, but the ref_obj_id is set to 0. Hence, we must enforce that whenever meta.ref_obj_id is zero, the register that is to be released can only be NULL for a release function. Since we now indicate whether an argument is to be released in bpf_func_proto itself, is_release_function helper has lost its utitlity, hence refactor code to work without it, and just rely on meta.release_regno to know when to release state for a ref_obj_id. Still, the restriction of one release argument and only one ref_obj_id passed to BPF helper or kfunc remains. This may be lifted in the future. Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/bpf/20220424214901.2743946-3-memxor@gmail.com |
||
|
Martin KaFai Lau
|
4a9c7bbe2e |
bpf: Resolve to prog->aux->dst_prog->type only for BPF_PROG_TYPE_EXT
The commit |
||
|
Kumar Kartikeya Dwivedi
|
24d5bb806c |
bpf: Harden register offset checks for release helpers and kfuncs
Let's ensure that the PTR_TO_BTF_ID reg being passed in to release BPF
helpers and kfuncs always has its offset set to 0. While not a real
problem now, there's a very real possibility this will become a problem
when more and more kfuncs are exposed, and more BPF helpers are added
which can release PTR_TO_BTF_ID.
Previous commits already protected against non-zero var_off. One of the
case we are concerned about now is when we have a type that can be
returned by e.g. an acquire kfunc:
struct foo {
int a;
int b;
struct bar b;
};
... and struct bar is also a type that can be returned by another
acquire kfunc.
Then, doing the following sequence:
struct foo *f = bpf_get_foo(); // acquire kfunc
if (!f)
return 0;
bpf_put_bar(&f->b); // release kfunc
... would work with the current code, since the btf_struct_ids_match
takes reg->off into account for matching pointer type with release kfunc
argument type, but would obviously be incorrect, and most likely lead to
a kernel crash. A test has been included later to prevent regressions in
this area.
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20220304224645.3677453-5-memxor@gmail.com
|
||
|
Kumar Kartikeya Dwivedi
|
25b35dd281 |
bpf: Add check_func_arg_reg_off function
Lift the list of register types allowed for having fixed and variable offsets when passed as helper function arguments into a common helper, so that they can be reused for kfunc checks in later commits. Keeping a common helper aids maintainability and allows us to follow the same consistent rules across helpers and kfuncs. Also, convert check_func_arg to use this function. Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/bpf/20220304224645.3677453-2-memxor@gmail.com |
||
|
Jakub Kicinski
|
caaba96131 |
Merge https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next
Daniel Borkmann says:
====================
pull-request: bpf-next 2022-01-24
We've added 80 non-merge commits during the last 14 day(s) which contain
a total of 128 files changed, 4990 insertions(+), 895 deletions(-).
The main changes are:
1) Add XDP multi-buffer support and implement it for the mvneta driver,
from Lorenzo Bianconi, Eelco Chaudron and Toke Høiland-Jørgensen.
2) Add unstable conntrack lookup helpers for BPF by using the BPF kfunc
infra, from Kumar Kartikeya Dwivedi.
3) Extend BPF cgroup programs to export custom ret value to userspace via
two helpers bpf_get_retval() and bpf_set_retval(), from YiFei Zhu.
4) Add support for AF_UNIX iterator batching, from Kuniyuki Iwashima.
5) Complete missing UAPI BPF helper description and change bpf_doc.py script
to enforce consistent & complete helper documentation, from Usama Arif.
6) Deprecate libbpf's legacy BPF map definitions and streamline XDP APIs to
follow tc-based APIs, from Andrii Nakryiko.
7) Support BPF_PROG_QUERY for BPF programs attached to sockmap, from Di Zhu.
8) Deprecate libbpf's bpf_map__def() API and replace users with proper getters
and setters, from Christy Lee.
9) Extend libbpf's btf__add_btf() with an additional hashmap for strings to
reduce overhead, from Kui-Feng Lee.
10) Fix bpftool and libbpf error handling related to libbpf's hashmap__new()
utility function, from Mauricio Vásquez.
11) Add support to BTF program names in bpftool's program dump, from Raman Shukhau.
12) Fix resolve_btfids build to pick up host flags, from Connor O'Brien.
* https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next: (80 commits)
selftests, bpf: Do not yet switch to new libbpf XDP APIs
selftests, xsk: Fix rx_full stats test
bpf: Fix flexible_array.cocci warnings
xdp: disable XDP_REDIRECT for xdp frags
bpf: selftests: add CPUMAP/DEVMAP selftests for xdp frags
bpf: selftests: introduce bpf_xdp_{load,store}_bytes selftest
net: xdp: introduce bpf_xdp_pointer utility routine
bpf: generalise tail call map compatibility check
libbpf: Add SEC name for xdp frags programs
bpf: selftests: update xdp_adjust_tail selftest to include xdp frags
bpf: test_run: add xdp_shared_info pointer in bpf_test_finish signature
bpf: introduce frags support to bpf_prog_test_run_xdp()
bpf: move user_size out of bpf_test_init
bpf: add frags support to xdp copy helpers
bpf: add frags support to the bpf_xdp_adjust_tail() API
bpf: introduce bpf_xdp_get_buff_len helper
net: mvneta: enable jumbo frames if the loaded XDP program support frags
bpf: introduce BPF_F_XDP_HAS_FRAGS flag in prog_flags loading the ebpf program
net: mvneta: add frags support to XDP_TX
xdp: add frags support to xdp_return_{buff/frame}
...
====================
Link: https://lore.kernel.org/r/20220124221235.18993-1-daniel@iogearbox.net
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
|
||
|
Daniel Borkmann
|
be80a1d3f9 |
bpf: Generalize check_ctx_reg for reuse with other types
Generalize the check_ctx_reg() helper function into a more generic named one so that it can be reused for other register types as well to check whether their offset is non-zero. No functional change. Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: John Fastabend <john.fastabend@gmail.com> Acked-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Kumar Kartikeya Dwivedi
|
5c073f26f9 |
bpf: Add reference tracking support to kfunc
This patch adds verifier support for PTR_TO_BTF_ID return type of kfunc
to be a reference, by reusing acquire_reference_state/release_reference
support for existing in-kernel bpf helpers.
We make use of the three kfunc types:
- BTF_KFUNC_TYPE_ACQUIRE
Return true if kfunc_btf_id is an acquire kfunc. This will
acquire_reference_state for the returned PTR_TO_BTF_ID (this is the
only allow return value). Note that acquire kfunc must always return a
PTR_TO_BTF_ID{_OR_NULL}, otherwise the program is rejected.
- BTF_KFUNC_TYPE_RELEASE
Return true if kfunc_btf_id is a release kfunc. This will release the
reference to the passed in PTR_TO_BTF_ID which has a reference state
(from earlier acquire kfunc).
The btf_check_func_arg_match returns the regno (of argument register,
hence > 0) if the kfunc is a release kfunc, and a proper referenced
PTR_TO_BTF_ID is being passed to it.
This is similar to how helper call check uses bpf_call_arg_meta to
store the ref_obj_id that is later used to release the reference.
Similar to in-kernel helper, we only allow passing one referenced
PTR_TO_BTF_ID as an argument. It can also be passed in to normal
kfunc, but in case of release kfunc there must always be one
PTR_TO_BTF_ID argument that is referenced.
- BTF_KFUNC_TYPE_RET_NULL
For kfunc returning PTR_TO_BTF_ID, tells if it can be NULL, hence
force caller to mark the pointer not null (using check) before
accessing it. Note that taking into account the case fixed by commit
|
||
|
Kumar Kartikeya Dwivedi
|
d583691c47 |
bpf: Introduce mem, size argument pair support for kfunc
BPF helpers can associate two adjacent arguments together to pass memory of certain size, using ARG_PTR_TO_MEM and ARG_CONST_SIZE arguments. Since we don't use bpf_func_proto for kfunc, we need to leverage BTF to implement similar support. The ARG_CONST_SIZE processing for helpers is refactored into a common check_mem_size_reg helper that is shared with kfunc as well. kfunc ptr_to_mem support follows logic similar to global functions, where verification is done as if pointer is not null, even when it may be null. This leads to a simple to follow rule for writing kfunc: always check the argument pointer for NULL, except when it is PTR_TO_CTX. Also, the PTR_TO_CTX case is also only safe when the helper expecting pointer to program ctx is not exposed to other programs where same struct is not ctx type. In that case, the type check will fall through to other cases and would permit passing other types of pointers, possibly NULL at runtime. Currently, we require the size argument to be suffixed with "__sz" in the parameter name. This information is then recorded in kernel BTF and verified during function argument checking. In the future we can use BTF tagging instead, and modify the kernel function definitions. This will be a purely kernel-side change. This allows us to have some form of backwards compatibility for structures that are passed in to the kernel function with their size, and allow variable length structures to be passed in if they are accompanied by a size parameter. Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com> Link: https://lore.kernel.org/r/20220114163953.1455836-5-memxor@gmail.com Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Hao Luo
|
c25b2ae136 |
bpf: Replace PTR_TO_XXX_OR_NULL with PTR_TO_XXX | PTR_MAYBE_NULL
We have introduced a new type to make bpf_reg composable, by allocating bits in the type to represent flags. One of the flags is PTR_MAYBE_NULL which indicates a pointer may be NULL. This patch switches the qualified reg_types to use this flag. The reg_types changed in this patch include: 1. PTR_TO_MAP_VALUE_OR_NULL 2. PTR_TO_SOCKET_OR_NULL 3. PTR_TO_SOCK_COMMON_OR_NULL 4. PTR_TO_TCP_SOCK_OR_NULL 5. PTR_TO_BTF_ID_OR_NULL 6. PTR_TO_MEM_OR_NULL 7. PTR_TO_RDONLY_BUF_OR_NULL 8. PTR_TO_RDWR_BUF_OR_NULL Signed-off-by: Hao Luo <haoluo@google.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/r/20211217003152.48334-5-haoluo@google.com |
||
|
Hao Luo
|
d639b9d13a |
bpf: Introduce composable reg, ret and arg types.
There are some common properties shared between bpf reg, ret and arg values. For instance, a value may be a NULL pointer, or a pointer to a read-only memory. Previously, to express these properties, enumeration was used. For example, in order to test whether a reg value can be NULL, reg_type_may_be_null() simply enumerates all types that are possibly NULL. The problem of this approach is that it's not scalable and causes a lot of duplication. These properties can be combined, for example, a type could be either MAYBE_NULL or RDONLY, or both. This patch series rewrites the layout of reg_type, arg_type and ret_type, so that common properties can be extracted and represented as composable flag. For example, one can write ARG_PTR_TO_MEM | PTR_MAYBE_NULL which is equivalent to the previous ARG_PTR_TO_MEM_OR_NULL The type ARG_PTR_TO_MEM are called "base type" in this patch. Base types can be extended with flags. A flag occupies the higher bits while base types sits in the lower bits. This patch in particular sets up a set of macro for this purpose. The following patches will rewrite arg_types, ret_types and reg_types respectively. Signed-off-by: Hao Luo <haoluo@google.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/bpf/20211217003152.48334-2-haoluo@google.com |
||
|
Christy Lee
|
2e5766483c |
bpf: Right align verifier states in verifier logs.
Make the verifier logs more readable, print the verifier states on the corresponding instruction line. If the previous line was not a bpf instruction, then print the verifier states on its own line. Before: Validating test_pkt_access_subprog3() func#3... 86: R1=invP(id=0) R2=ctx(id=0,off=0,imm=0) R10=fp0 ; int test_pkt_access_subprog3(int val, struct __sk_buff *skb) 86: (bf) r6 = r2 87: R2=ctx(id=0,off=0,imm=0) R6_w=ctx(id=0,off=0,imm=0) 87: (bc) w7 = w1 88: R1=invP(id=0) R7_w=invP(id=0,umax_value=4294967295,var_off=(0x0; 0xffffffff)) ; return get_skb_len(skb) * get_skb_ifindex(val, skb, get_constant(123)); 88: (bf) r1 = r6 89: R1_w=ctx(id=0,off=0,imm=0) R6_w=ctx(id=0,off=0,imm=0) 89: (85) call pc+9 Func#4 is global and valid. Skipping. 90: R0_w=invP(id=0) 90: (bc) w8 = w0 91: R0_w=invP(id=0) R8_w=invP(id=0,umax_value=4294967295,var_off=(0x0; 0xffffffff)) ; return get_skb_len(skb) * get_skb_ifindex(val, skb, get_constant(123)); 91: (b7) r1 = 123 92: R1_w=invP123 92: (85) call pc+65 Func#5 is global and valid. Skipping. 93: R0=invP(id=0) After: 86: R1=invP(id=0) R2=ctx(id=0,off=0,imm=0) R10=fp0 ; int test_pkt_access_subprog3(int val, struct __sk_buff *skb) 86: (bf) r6 = r2 ; R2=ctx(id=0,off=0,imm=0) R6_w=ctx(id=0,off=0,imm=0) 87: (bc) w7 = w1 ; R1=invP(id=0) R7_w=invP(id=0,umax_value=4294967295,var_off=(0x0; 0xffffffff)) ; return get_skb_len(skb) * get_skb_ifindex(val, skb, get_constant(123)); 88: (bf) r1 = r6 ; R1_w=ctx(id=0,off=0,imm=0) R6_w=ctx(id=0,off=0,imm=0) 89: (85) call pc+9 Func#4 is global and valid. Skipping. 90: R0_w=invP(id=0) 90: (bc) w8 = w0 ; R0_w=invP(id=0) R8_w=invP(id=0,umax_value=4294967295,var_off=(0x0; 0xffffffff)) ; return get_skb_len(skb) * get_skb_ifindex(val, skb, get_constant(123)); 91: (b7) r1 = 123 ; R1_w=invP123 92: (85) call pc+65 Func#5 is global and valid. Skipping. 93: R0=invP(id=0) Signed-off-by: Christy Lee <christylee@fb.com> Acked-by: Andrii Nakryiko <andrii@kernel.org> Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Christy Lee
|
0f55f9ed21 |
bpf: Only print scratched registers and stack slots to verifier logs.
When printing verifier state for any log level, print full verifier state only on function calls or on errors. Otherwise, only print the registers and stack slots that were accessed. Log size differences: verif_scale_loop6 before: 234566564 verif_scale_loop6 after: 72143943 69% size reduction kfree_skb before: 166406 kfree_skb after: 55386 69% size reduction Before: 156: (61) r0 = *(u32 *)(r1 +0) 157: R0_w=invP(id=0,umax_value=4294967295,var_off=(0x0; 0xffffffff)) R1=ctx(id=0,off=0,imm=0) R2_w=invP0 R10=fp0 fp-8_w=00000000 fp-16_w=00\ 000000 fp-24_w=00000000 fp-32_w=00000000 fp-40_w=00000000 fp-48_w=00000000 fp-56_w=00000000 fp-64_w=00000000 fp-72_w=00000000 fp-80_w=00000\ 000 fp-88_w=00000000 fp-96_w=00000000 fp-104_w=00000000 fp-112_w=00000000 fp-120_w=00000000 fp-128_w=00000000 fp-136_w=00000000 fp-144_w=00\ 000000 fp-152_w=00000000 fp-160_w=00000000 fp-168_w=00000000 fp-176_w=00000000 fp-184_w=00000000 fp-192_w=00000000 fp-200_w=00000000 fp-208\ _w=00000000 fp-216_w=00000000 fp-224_w=00000000 fp-232_w=00000000 fp-240_w=00000000 fp-248_w=00000000 fp-256_w=00000000 fp-264_w=00000000 f\ p-272_w=00000000 fp-280_w=00000000 fp-288_w=00000000 fp-296_w=00000000 fp-304_w=00000000 fp-312_w=00000000 fp-320_w=00000000 fp-328_w=00000\ 000 fp-336_w=00000000 fp-344_w=00000000 fp-352_w=00000000 fp-360_w=00000000 fp-368_w=00000000 fp-376_w=00000000 fp-384_w=00000000 fp-392_w=\ 00000000 fp-400_w=00000000 fp-408_w=00000000 fp-416_w=00000000 fp-424_w=00000000 fp-432_w=00000000 fp-440_w=00000000 fp-448_w=00000000 ; return skb->len; 157: (95) exit Func#4 is safe for any args that match its prototype Validating get_constant() func#5... 158: R1=invP(id=0) R10=fp0 ; int get_constant(long val) 158: (bf) r0 = r1 159: R0_w=invP(id=1) R1=invP(id=1) R10=fp0 ; return val - 122; 159: (04) w0 += -122 160: R0_w=invP(id=0,umax_value=4294967295,var_off=(0x0; 0xffffffff)) R1=invP(id=1) R10=fp0 ; return val - 122; 160: (95) exit Func#5 is safe for any args that match its prototype Validating get_skb_ifindex() func#6... 161: R1=invP(id=0) R2=ctx(id=0,off=0,imm=0) R3=invP(id=0) R10=fp0 ; int get_skb_ifindex(int val, struct __sk_buff *skb, int var) 161: (bc) w0 = w3 162: R0_w=invP(id=0,umax_value=4294967295,var_off=(0x0; 0xffffffff)) R1=invP(id=0) R2=ctx(id=0,off=0,imm=0) R3=invP(id=0) R10=fp0 After: 156: (61) r0 = *(u32 *)(r1 +0) 157: R0_w=invP(id=0,umax_value=4294967295,var_off=(0x0; 0xffffffff)) R1=ctx(id=0,off=0,imm=0) ; return skb->len; 157: (95) exit Func#4 is safe for any args that match its prototype Validating get_constant() func#5... 158: R1=invP(id=0) R10=fp0 ; int get_constant(long val) 158: (bf) r0 = r1 159: R0_w=invP(id=1) R1=invP(id=1) ; return val - 122; 159: (04) w0 += -122 160: R0_w=invP(id=0,umax_value=4294967295,var_off=(0x0; 0xffffffff)) ; return val - 122; 160: (95) exit Func#5 is safe for any args that match its prototype Validating get_skb_ifindex() func#6... 161: R1=invP(id=0) R2=ctx(id=0,off=0,imm=0) R3=invP(id=0) R10=fp0 ; int get_skb_ifindex(int val, struct __sk_buff *skb, int var) 161: (bc) w0 = w3 162: R0_w=invP(id=0,umax_value=4294967295,var_off=(0x0; 0xffffffff)) R3=invP(id=0) Signed-off-by: Christy Lee <christylee@fb.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/bpf/20211216213358.3374427-2-christylee@fb.com |
||
|
Hou Tao
|
866de40744 |
bpf: Disallow BPF_LOG_KERNEL log level for bpf(BPF_BTF_LOAD)
BPF_LOG_KERNEL is only used internally, so disallow bpf_btf_load()
to set log level as BPF_LOG_KERNEL. The same checking has already
been done in bpf_check(), so factor out a helper to check the
validity of log attributes and use it in both places.
Fixes:
|
||
|
Kumar Kartikeya Dwivedi
|
2357672c54 |
bpf: Introduce BPF support for kernel module function calls
This change adds support on the kernel side to allow for BPF programs to call kernel module functions. Userspace will prepare an array of module BTF fds that is passed in during BPF_PROG_LOAD using fd_array parameter. In the kernel, the module BTFs are placed in the auxilliary struct for bpf_prog, and loaded as needed. The verifier then uses insn->off to index into the fd_array. insn->off 0 is reserved for vmlinux BTF (for backwards compat), so userspace must use an fd_array index > 0 for module kfunc support. kfunc_btf_tab is sorted based on offset in an array, and each offset corresponds to one descriptor, with a max limit up to 256 such module BTFs. We also change existing kfunc_tab to distinguish each element based on imm, off pair as each such call will now be distinct. Another change is to check_kfunc_call callback, which now include a struct module * pointer, this is to be used in later patch such that the kfunc_id and module pointer are matched for dynamically registered BTF sets from loadable modules, so that same kfunc_id in two modules doesn't lead to check_kfunc_call succeeding. For the duration of the check_kfunc_call, the reference to struct module exists, as it returns the pointer stored in kfunc_btf_tab. Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/bpf/20211002011757.311265-2-memxor@gmail.com |
||
|
Jakub Kicinski
|
d2e11fd2b7 |
Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Conflicting commits, all resolutions pretty trivial: drivers/bus/mhi/pci_generic.c |
||
|
Daniel Borkmann
|
2039f26f3a |
bpf: Fix leakage due to insufficient speculative store bypass mitigation
Spectre v4 gadgets make use of memory disambiguation, which is a set of techniques that execute memory access instructions, that is, loads and stores, out of program order; Intel's optimization manual, section 2.4.4.5: A load instruction micro-op may depend on a preceding store. Many microarchitectures block loads until all preceding store addresses are known. The memory disambiguator predicts which loads will not depend on any previous stores. When the disambiguator predicts that a load does not have such a dependency, the load takes its data from the L1 data cache. Eventually, the prediction is verified. If an actual conflict is detected, the load and all succeeding instructions are re-executed. |
||
|
Daniel Borkmann
|
e042aa532c |
bpf: Fix pointer arithmetic mask tightening under state pruning
In |
||
|
Alexei Starovoitov
|
7ddc80a476 |
bpf: Teach stack depth check about async callbacks.
Teach max stack depth checking algorithm about async callbacks that don't increase bpf program stack size. Also add sanity check that bpf_tail_call didn't sneak into async cb. It's impossible, since PTR_TO_CTX is not available in async cb, hence the program cannot contain bpf_tail_call(ctx,...); Signed-off-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Andrii Nakryiko <andrii@kernel.org> Acked-by: Toke Høiland-Jørgensen <toke@redhat.com> Link: https://lore.kernel.org/bpf/20210715005417.78572-10-alexei.starovoitov@gmail.com |
||
|
Alexei Starovoitov
|
bfc6bb74e4 |
bpf: Implement verifier support for validation of async callbacks.
bpf_for_each_map_elem() and bpf_timer_set_callback() helpers are relying on PTR_TO_FUNC infra in the verifier to validate addresses to subprograms and pass them into the helpers as function callbacks. In case of bpf_for_each_map_elem() the callback is invoked synchronously and the verifier treats it as a normal subprogram call by adding another bpf_func_state and new frame in __check_func_call(). bpf_timer_set_callback() doesn't invoke the callback directly. The subprogram will be called asynchronously from bpf_timer_cb(). Teach the verifier to validate such async callbacks as special kind of jump by pushing verifier state into stack and let pop_stack() process it. Special care needs to be taken during state pruning. The call insn doing bpf_timer_set_callback has to be a prune_point. Otherwise short timer callbacks might not have prune points in front of bpf_timer_set_callback() which means is_state_visited() will be called after this call insn is processed in __check_func_call(). Which means that another async_cb state will be pushed to be walked later and the verifier will eventually hit BPF_COMPLEXITY_LIMIT_JMP_SEQ limit. Since push_async_cb() looks like another push_stack() branch the infinite loop detection will trigger false positive. To recognize this case mark such states as in_async_callback_fn. To distinguish infinite loop in async callback vs the same callback called with different arguments for different map and timer add async_entry_cnt to bpf_func_state. Enforce return zero from async callbacks. Signed-off-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Andrii Nakryiko <andrii@kernel.org> Acked-by: Toke Høiland-Jørgensen <toke@redhat.com> Link: https://lore.kernel.org/bpf/20210715005417.78572-9-alexei.starovoitov@gmail.com |
||
|
Alexei Starovoitov
|
3e8ce29850 |
bpf: Prevent pointer mismatch in bpf_timer_init.
bpf_timer_init() arguments are:
1. pointer to a timer (which is embedded in map element).
2. pointer to a map.
Make sure that pointer to a timer actually belongs to that map.
Use map_uid (which is unique id of inner map) to reject:
inner_map1 = bpf_map_lookup_elem(outer_map, key1)
inner_map2 = bpf_map_lookup_elem(outer_map, key2)
if (inner_map1 && inner_map2) {
timer = bpf_map_lookup_elem(inner_map1);
if (timer)
// mismatch would have been allowed
bpf_timer_init(timer, inner_map2);
}
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Toke Høiland-Jørgensen <toke@redhat.com>
Link: https://lore.kernel.org/bpf/20210715005417.78572-6-alexei.starovoitov@gmail.com
|
||
|
Alexei Starovoitov
|
387544bfa2 |
bpf: Introduce fd_idx
Typical program loading sequence involves creating bpf maps and applying map FDs into bpf instructions in various places in the bpf program. This job is done by libbpf that is using compiler generated ELF relocations to patch certain instruction after maps are created and BTFs are loaded. The goal of fd_idx is to allow bpf instructions to stay immutable after compilation. At load time the libbpf would still create maps as usual, but it wouldn't need to patch instructions. It would store map_fds into __u32 fd_array[] and would pass that pointer to sys_bpf(BPF_PROG_LOAD). Signed-off-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/bpf/20210514003623.28033-9-alexei.starovoitov@gmail.com |
||
|
Lorenz Bauer
|
c9e73e3d2b |
bpf: verifier: Allocate idmap scratch in verifier env
func_states_equal makes a very short lived allocation for idmap, probably because it's too large to fit on the stack. However the function is called quite often, leading to a lot of alloc / free churn. Replace the temporary allocation with dedicated scratch space in struct bpf_verifier_env. Signed-off-by: Lorenz Bauer <lmb@cloudflare.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Edward Cree <ecree.xilinx@gmail.com> Link: https://lore.kernel.org/bpf/20210429134656.122225-4-lmb@cloudflare.com |
||
|
Daniel Borkmann
|
801c6058d1 |
bpf: Fix leakage of uninitialized bpf stack under speculation
The current implemented mechanisms to mitigate data disclosure under
speculation mainly address stack and map value oob access from the
speculative domain. However, Piotr discovered that uninitialized BPF
stack is not protected yet, and thus old data from the kernel stack,
potentially including addresses of kernel structures, could still be
extracted from that 512 bytes large window. The BPF stack is special
compared to map values since it's not zero initialized for every
program invocation, whereas map values /are/ zero initialized upon
their initial allocation and thus cannot leak any prior data in either
domain. In the non-speculative domain, the verifier ensures that every
stack slot read must have a prior stack slot write by the BPF program
to avoid such data leaking issue.
However, this is not enough: for example, when the pointer arithmetic
operation moves the stack pointer from the last valid stack offset to
the first valid offset, the sanitation logic allows for any intermediate
offsets during speculative execution, which could then be used to
extract any restricted stack content via side-channel.
Given for unprivileged stack pointer arithmetic the use of unknown
but bounded scalars is generally forbidden, we can simply turn the
register-based arithmetic operation into an immediate-based arithmetic
operation without the need for masking. This also gives the benefit
of reducing the needed instructions for the operation. Given after
the work in
|
||
|
Toke Høiland-Jørgensen
|
441e8c66b2 |
bpf: Return target info when a tracing bpf_link is queried
There is currently no way to discover the target of a tracing program attachment after the fact. Add this information to bpf_link_info and return it when querying the bpf_link fd. Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/bpf/20210413091607.58945-1-toke@redhat.com |
||
|
Yonghong Song
|
69c087ba62 |
bpf: Add bpf_for_each_map_elem() helper
The bpf_for_each_map_elem() helper is introduced which iterates all map elements with a callback function. The helper signature looks like long bpf_for_each_map_elem(map, callback_fn, callback_ctx, flags) and for each map element, the callback_fn will be called. For example, like hashmap, the callback signature may look like long callback_fn(map, key, val, callback_ctx) There are two known use cases for this. One is from upstream ([1]) where a for_each_map_elem helper may help implement a timeout mechanism in a more generic way. Another is from our internal discussion for a firewall use case where a map contains all the rules. The packet data can be compared to all these rules to decide allow or deny the packet. For array maps, users can already use a bounded loop to traverse elements. Using this helper can avoid using bounded loop. For other type of maps (e.g., hash maps) where bounded loop is hard or impossible to use, this helper provides a convenient way to operate on all elements. For callback_fn, besides map and map element, a callback_ctx, allocated on caller stack, is also passed to the callback function. This callback_ctx argument can provide additional input and allow to write to caller stack for output. If the callback_fn returns 0, the helper will iterate through next element if available. If the callback_fn returns 1, the helper will stop iterating and returns to the bpf program. Other return values are not used for now. Currently, this helper is only available with jit. It is possible to make it work with interpreter with so effort but I leave it as the future work. [1]: https://lore.kernel.org/bpf/20210122205415.113822-1-xiyou.wangcong@gmail.com/ 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/20210226204925.3884923-1-yhs@fb.com |
||
|
Dmitrii Banshchikov
|
e5069b9c23 |
bpf: Support pointers in global func args
Add an ability to pass a pointer to a type with known size in arguments of a global function. Such pointers may be used to overcome the limit on the maximum number of arguments, avoid expensive and tricky workarounds and to have multiple output arguments. A referenced type may contain pointers but indirect access through them isn't supported. The implementation consists of two parts. If a global function has an argument that is a pointer to a type with known size then: 1) In btf_check_func_arg_match(): check that the corresponding register points to NULL or to a valid memory region that is large enough to contain the expected argument's type. 2) In btf_prepare_func_args(): set the corresponding register type to PTR_TO_MEM_OR_NULL and its size to the size of the expected type. Only global functions are supported because allowance of pointers for static functions might break validation. Consider the following scenario. A static function has a pointer argument. A caller passes pointer to its stack memory. Because the callee can change referenced memory verifier cannot longer assume any particular slot type of the caller's stack memory hence the slot type is changed to SLOT_MISC. If there is an operation that relies on slot type other than SLOT_MISC then verifier won't be able to infer safety of the operation. When verifier sees a static function that has a pointer argument different from PTR_TO_CTX then it skips arguments check and continues with "inline" validation with more information available. The operation that relies on the particular slot type now succeeds. Because global functions were not allowed to have pointer arguments different from PTR_TO_CTX it's not possible to break existing and valid code. Signed-off-by: Dmitrii Banshchikov <me@ubique.spb.ru> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Andrii Nakryiko <andrii@kernel.org> Link: https://lore.kernel.org/bpf/20210212205642.620788-4-me@ubique.spb.ru |
||
|
Andrei Matei
|
01f810ace9 |
bpf: Allow variable-offset stack access
Before this patch, variable offset access to the stack was dissalowed for regular instructions, but was allowed for "indirect" accesses (i.e. helpers). This patch removes the restriction, allowing reading and writing to the stack through stack pointers with variable offsets. This makes stack-allocated buffers more usable in programs, and brings stack pointers closer to other types of pointers. The motivation is being able to use stack-allocated buffers for data manipulation. When the stack size limit is sufficient, allocating buffers on the stack is simpler than per-cpu arrays, or other alternatives. In unpriviledged programs, variable-offset reads and writes are disallowed (they were already disallowed for the indirect access case) because the speculative execution checking code doesn't support them. Additionally, when writing through a variable-offset stack pointer, if any pointers are in the accessible range, there's possilibities of later leaking pointers because the write cannot be tracked precisely. Writes with variable offset mark the whole range as initialized, even though we don't know which stack slots are actually written. This is in order to not reject future reads to these slots. Note that this doesn't affect writes done through helpers; like before, helpers need the whole stack range to be initialized to begin with. All the stack slots are in range are considered scalars after the write; variable-offset register spills are not tracked. For reads, all the stack slots in the variable range needs to be initialized (but see above about what writes do), otherwise the read is rejected. All register spilled in stack slots that might be read are marked as having been read, however reads through such pointers don't do register filling; the target register will always be either a scalar or a constant zero. Signed-off-by: Andrei Matei <andreimatei1@gmail.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/bpf/20210207011027.676572-2-andreimatei1@gmail.com |
||
|
Andrii Nakryiko
|
541c3bad8d |
bpf: Support BPF ksym variables in kernel modules
Add support for directly accessing kernel module variables from BPF programs using special ldimm64 instructions. This functionality builds upon vmlinux ksym support, but extends ldimm64 with src_reg=BPF_PSEUDO_BTF_ID to allow specifying kernel module BTF's FD in insn[1].imm field. During BPF program load time, verifier will resolve FD to BTF object and will take reference on BTF object itself and, for module BTFs, corresponding module as well, to make sure it won't be unloaded from under running BPF program. The mechanism used is similar to how bpf_prog keeps track of used bpf_maps. One interesting change is also in how per-CPU variable is determined. The logic is to find .data..percpu data section in provided BTF, but both vmlinux and module each have their own .data..percpu entries in BTF. So for module's case, the search for DATASEC record needs to look at only module's added BTF types. This is implemented with custom search function. Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Yonghong Song <yhs@fb.com> Acked-by: Hao Luo <haoluo@google.com> Link: https://lore.kernel.org/bpf/20210112075520.4103414-6-andrii@kernel.org |
||
|
Andrii Nakryiko
|
22dc4a0f5e |
bpf: Remove hard-coded btf_vmlinux assumption from BPF verifier
Remove a permeating assumption thoughout BPF verifier of vmlinux BTF. Instead, wherever BTF type IDs are involved, also track the instance of struct btf that goes along with the type ID. This allows to gradually add support for kernel module BTFs and using/tracking module types across BPF helper calls and registers. This patch also renames btf_id() function to btf_obj_id() to minimize naming clash with using btf_id to denote BTF *type* ID, rather than BTF *object*'s ID. Also, altough btf_vmlinux can't get destructed and thus doesn't need refcounting, module BTFs need that, so apply BTF refcounting universally when BPF program is using BTF-powered attachment (tp_btf, fentry/fexit, etc). This makes for simpler clean up code. Now that BTF type ID is not enough to uniquely identify a BTF type, extend BPF trampoline key to include BTF object ID. To differentiate that from target program BPF ID, set 31st bit of type ID. BTF type IDs (at least currently) are not allowed to take full 32 bits, so there is no danger of confusing that bit with a valid BTF type ID. Signed-off-by: Andrii Nakryiko <andrii@kernel.org> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Link: https://lore.kernel.org/bpf/20201203204634.1325171-10-andrii@kernel.org |
||
|
Alexei Starovoitov
|
6d94e741a8 |
bpf: Support for pointers beyond pkt_end.
This patch adds the verifier support to recognize inlined branch conditions. The LLVM knows that the branch evaluates to the same value, but the verifier couldn't track it. Hence causing valid programs to be rejected. The potential LLVM workaround: https://reviews.llvm.org/D87428 can have undesired side effects, since LLVM doesn't know that skb->data/data_end are being compared. LLVM has to introduce extra boolean variable and use inline_asm trick to force easier for the verifier assembly. Instead teach the verifier to recognize that r1 = skb->data; r1 += 10; r2 = skb->data_end; if (r1 > r2) { here r1 points beyond packet_end and subsequent if (r1 > r2) // always evaluates to "true". } Signed-off-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Tested-by: Jiri Olsa <jolsa@redhat.com> Acked-by: John Fastabend <john.fastabend@gmail.com> Link: https://lore.kernel.org/bpf/20201111031213.25109-2-alexei.starovoitov@gmail.com |
||
|
Hao Luo
|
4976b718c3 |
bpf: Introduce pseudo_btf_id
Pseudo_btf_id is a type of ld_imm insn that associates a btf_id to a ksym so that further dereferences on the ksym can use the BTF info to validate accesses. Internally, when seeing a pseudo_btf_id ld insn, the verifier reads the btf_id stored in the insn[0]'s imm field and marks the dst_reg as PTR_TO_BTF_ID. The btf_id points to a VAR_KIND, which is encoded in btf_vminux by pahole. If the VAR is not of a struct type, the dst reg will be marked as PTR_TO_MEM instead of PTR_TO_BTF_ID and the mem_size is resolved to the size of the VAR's type. >From the VAR btf_id, the verifier can also read the address of the ksym's corresponding kernel var from kallsyms and use that to fill dst_reg. Therefore, the proper functionality of pseudo_btf_id depends on (1) kallsyms and (2) the encoding of kernel global VARs in pahole, which should be available since pahole v1.18. Signed-off-by: Hao Luo <haoluo@google.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> Acked-by: Andrii Nakryiko <andriin@fb.com> Link: https://lore.kernel.org/bpf/20200929235049.2533242-2-haoluo@google.com |
||
|
Toke Høiland-Jørgensen
|
f7b12b6fea |
bpf: verifier: refactor check_attach_btf_id()
The check_attach_btf_id() function really does three things: 1. It performs a bunch of checks on the program to ensure that the attachment is valid. 2. It stores a bunch of state about the attachment being requested in the verifier environment and struct bpf_prog objects. 3. It allocates a trampoline for the attachment. This patch splits out (1.) and (3.) into separate functions which will perform the checks, but return the computed values instead of directly modifying the environment. This is done in preparation for reusing the checks when the actual attachment is happening, which will allow tracing programs to have multiple (compatible) attachments. This also fixes a bug where a bunch of checks were skipped if a trampoline already existed for the tracing target. Fixes: |
||
|
Toke Høiland-Jørgensen
|
efc68158c4 |
bpf: change logging calls from verbose() to bpf_log() and use log pointer
In preparation for moving code around, change a bunch of references to env->log (and the verbose() logging helper) to use bpf_log() and a direct pointer to struct bpf_verifier_log. While we're touching the function signature, mark the 'prog' argument to bpf_check_type_match() as const. Also enhance the bpf_verifier_log_needed() check to handle NULL pointers for the log struct so we can re-use the code with logging disabled. Acked-by: Andrii Nakryiko <andriin@fb.com> Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Alexei Starovoitov
|
09b28d76ea |
bpf: Add abnormal return checks.
LD_[ABS|IND] instructions may return from the function early. bpf_tail_call pseudo instruction is either fallthrough or return. Allow them in the subprograms only when subprograms are BTF annotated and have scalar return types. Allow ld_abs and tail_call in the main program even if it calls into subprograms. In the past that was not ok to do for ld_abs, since it was JITed with special exit sequence. Since bpf_gen_ld_abs() was introduced the ld_abs looks like normal exit insn from JIT point of view, so it's safe to allow them in the main program. Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Maciej Fijalkowski
|
ebf7d1f508 |
bpf, x64: rework pro/epilogue and tailcall handling in JIT
This commit serves two things:
1) it optimizes BPF prologue/epilogue generation
2) it makes possible to have tailcalls within BPF subprogram
Both points are related to each other since without 1), 2) could not be
achieved.
In [1], Alexei says:
"The prologue will look like:
nop5
xor eax,eax // two new bytes if bpf_tail_call() is used in this
// function
push rbp
mov rbp, rsp
sub rsp, rounded_stack_depth
push rax // zero init tail_call counter
variable number of push rbx,r13,r14,r15
Then bpf_tail_call will pop variable number rbx,..
and final 'pop rax'
Then 'add rsp, size_of_current_stack_frame'
jmp to next function and skip over 'nop5; xor eax,eax; push rpb; mov
rbp, rsp'
This way new function will set its own stack size and will init tail
call
counter with whatever value the parent had.
If next function doesn't use bpf_tail_call it won't have 'xor eax,eax'.
Instead it would need to have 'nop2' in there."
Implement that suggestion.
Since the layout of stack is changed, tail call counter handling can not
rely anymore on popping it to rbx just like it have been handled for
constant prologue case and later overwrite of rbx with actual value of
rbx pushed to stack. Therefore, let's use one of the register (%rcx) that
is considered to be volatile/caller-saved and pop the value of tail call
counter in there in the epilogue.
Drop the BUILD_BUG_ON in emit_prologue and in
emit_bpf_tail_call_indirect where instruction layout is not constant
anymore.
Introduce new poke target, 'tailcall_bypass' to poke descriptor that is
dedicated for skipping the register pops and stack unwind that are
generated right before the actual jump to target program.
For case when the target program is not present, BPF program will skip
the pop instructions and nop5 dedicated for jmpq $target. An example of
such state when only R6 of callee saved registers is used by program:
ffffffffc0513aa1: e9 0e 00 00 00 jmpq 0xffffffffc0513ab4
ffffffffc0513aa6: 5b pop %rbx
ffffffffc0513aa7: 58 pop %rax
ffffffffc0513aa8: 48 81 c4 00 00 00 00 add $0x0,%rsp
ffffffffc0513aaf: 0f 1f 44 00 00 nopl 0x0(%rax,%rax,1)
ffffffffc0513ab4: 48 89 df mov %rbx,%rdi
When target program is inserted, the jump that was there to skip
pops/nop5 will become the nop5, so CPU will go over pops and do the
actual tailcall.
One might ask why there simply can not be pushes after the nop5?
In the following example snippet:
ffffffffc037030c: 48 89 fb mov %rdi,%rbx
(...)
ffffffffc0370332: 5b pop %rbx
ffffffffc0370333: 58 pop %rax
ffffffffc0370334: 48 81 c4 00 00 00 00 add $0x0,%rsp
ffffffffc037033b: 0f 1f 44 00 00 nopl 0x0(%rax,%rax,1)
ffffffffc0370340: 48 81 ec 00 00 00 00 sub $0x0,%rsp
ffffffffc0370347: 50 push %rax
ffffffffc0370348: 53 push %rbx
ffffffffc0370349: 48 89 df mov %rbx,%rdi
ffffffffc037034c: e8 f7 21 00 00 callq 0xffffffffc0372548
There is the bpf2bpf call (at ffffffffc037034c) right after the tailcall
and jump target is not present. ctx is in %rbx register and BPF
subprogram that we will call into on ffffffffc037034c is relying on it,
e.g. it will pick ctx from there. Such code layout is therefore broken
as we would overwrite the content of %rbx with the value that was pushed
on the prologue. That is the reason for the 'bypass' approach.
Special care needs to be taken during the install/update/remove of
tailcall target. In case when target program is not present, the CPU
must not execute the pop instructions that precede the tailcall.
To address that, the following states can be defined:
A nop, unwind, nop
B nop, unwind, tail
C skip, unwind, nop
D skip, unwind, tail
A is forbidden (lead to incorrectness). The state transitions between
tailcall install/update/remove will work as follows:
First install tail call f: C->D->B(f)
* poke the tailcall, after that get rid of the skip
Update tail call f to f': B(f)->B(f')
* poke the tailcall (poke->tailcall_target) and do NOT touch the
poke->tailcall_bypass
Remove tail call: B(f')->C(f')
* poke->tailcall_bypass is poked back to jump, then we wait the RCU
grace period so that other programs will finish its execution and
after that we are safe to remove the poke->tailcall_target
Install new tail call (f''): C(f')->D(f'')->B(f'').
* same as first step
This way CPU can never be exposed to "unwind, tail" state.
Last but not least, when tailcalls get mixed with bpf2bpf calls, it
would be possible to encounter the endless loop due to clearing the
tailcall counter if for example we would use the tailcall3-like from BPF
selftests program that would be subprogram-based, meaning the tailcall
would be present within the BPF subprogram.
This test, broken down to particular steps, would do:
entry -> set tailcall counter to 0, bump it by 1, tailcall to func0
func0 -> call subprog_tail
(we are NOT skipping the first 11 bytes of prologue and this subprogram
has a tailcall, therefore we clear the counter...)
subprog -> do the same thing as entry
and then loop forever.
To address this, the idea is to go through the call chain of bpf2bpf progs
and look for a tailcall presence throughout whole chain. If we saw a single
tail call then each node in this call chain needs to be marked as a subprog
that can reach the tailcall. We would later feed the JIT with this info
and:
- set eax to 0 only when tailcall is reachable and this is the entry prog
- if tailcall is reachable but there's no tailcall in insns of currently
JITed prog then push rax anyway, so that it will be possible to
propagate further down the call chain
- finally if tailcall is reachable, then we need to precede the 'call'
insn with mov rax, [rbp - (stack_depth + 8)]
Tail call related cases from test_verifier kselftest are also working
fine. Sample BPF programs that utilize tail calls (sockex3, tracex5)
work properly as well.
[1]: https://lore.kernel.org/bpf/20200517043227.2gpq22ifoq37ogst@ast-mbp.dhcp.thefacebook.com/
Suggested-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|
||
|
Maciej Fijalkowski
|
7f6e4312e1 |
bpf: Limit caller's stack depth 256 for subprogs with tailcalls
Protect against potential stack overflow that might happen when bpf2bpf calls get combined with tailcalls. Limit the caller's stack depth for such case down to 256 so that the worst case scenario would result in 8k stack size (32 which is tailcall limit * 256 = 8k). Suggested-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com> Signed-off-by: Alexei Starovoitov <ast@kernel.org> |
||
|
Andrey Ignatov
|
41c48f3a98 |
bpf: Support access to bpf map fields
There are multiple use-cases when it's convenient to have access to bpf
map fields, both `struct bpf_map` and map type specific struct-s such as
`struct bpf_array`, `struct bpf_htab`, etc.
For example while working with sock arrays it can be necessary to
calculate the key based on map->max_entries (some_hash % max_entries).
Currently this is solved by communicating max_entries via "out-of-band"
channel, e.g. via additional map with known key to get info about target
map. That works, but is not very convenient and error-prone while
working with many maps.
In other cases necessary data is dynamic (i.e. unknown at loading time)
and it's impossible to get it at all. For example while working with a
hash table it can be convenient to know how much capacity is already
used (bpf_htab.count.counter for BPF_F_NO_PREALLOC case).
At the same time kernel knows this info and can provide it to bpf
program.
Fill this gap by adding support to access bpf map fields from bpf
program for both `struct bpf_map` and map type specific fields.
Support is implemented via btf_struct_access() so that a user can define
their own `struct bpf_map` or map type specific struct in their program
with only necessary fields and preserve_access_index attribute, cast a
map to this struct and use a field.
For example:
struct bpf_map {
__u32 max_entries;
} __attribute__((preserve_access_index));
struct bpf_array {
struct bpf_map map;
__u32 elem_size;
} __attribute__((preserve_access_index));
struct {
__uint(type, BPF_MAP_TYPE_ARRAY);
__uint(max_entries, 4);
__type(key, __u32);
__type(value, __u32);
} m_array SEC(".maps");
SEC("cgroup_skb/egress")
int cg_skb(void *ctx)
{
struct bpf_array *array = (struct bpf_array *)&m_array;
struct bpf_map *map = (struct bpf_map *)&m_array;
/* .. use map->max_entries or array->map.max_entries .. */
}
Similarly to other btf_struct_access() use-cases (e.g. struct tcp_sock
in net/ipv4/bpf_tcp_ca.c) the patch allows access to any fields of
corresponding struct. Only reading from map fields is supported.
For btf_struct_access() to work there should be a way to know btf id of
a struct that corresponds to a map type. To get btf id there should be a
way to get a stringified name of map-specific struct, such as
"bpf_array", "bpf_htab", etc for a map type. Two new fields are added to
`struct bpf_map_ops` to handle it:
* .map_btf_name keeps a btf name of a struct returned by map_alloc();
* .map_btf_id is used to cache btf id of that struct.
To make btf ids calculation cheaper they're calculated once while
preparing btf_vmlinux and cached same way as it's done for btf_id field
of `struct bpf_func_proto`
While calculating btf ids, struct names are NOT checked for collision.
Collisions will be checked as a part of the work to prepare btf ids used
in verifier in compile time that should land soon. The only known
collision for `struct bpf_htab` (kernel/bpf/hashtab.c vs
net/core/sock_map.c) was fixed earlier.
Both new fields .map_btf_name and .map_btf_id must be set for a map type
for the feature to work. If neither is set for a map type, verifier will
return ENOTSUPP on a try to access map_ptr of corresponding type. If
just one of them set, it's verifier misconfiguration.
Only `struct bpf_array` for BPF_MAP_TYPE_ARRAY and `struct bpf_htab` for
BPF_MAP_TYPE_HASH are supported by this patch. Other map types will be
supported separately.
The feature is available only for CONFIG_DEBUG_INFO_BTF=y and gated by
perfmon_capable() so that unpriv programs won't have access to bpf map
fields.
Signed-off-by: Andrey Ignatov <rdna@fb.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Link: https://lore.kernel.org/bpf/6479686a0cd1e9067993df57b4c3eef0e276fec9.1592600985.git.rdna@fb.com
|
||
|
Andrii Nakryiko
|
457f44363a |
bpf: Implement BPF ring buffer and verifier support for it
This commit adds a new MPSC ring buffer implementation into BPF ecosystem,
which allows multiple CPUs to submit data to a single shared ring buffer. On
the consumption side, only single consumer is assumed.
Motivation
----------
There are two distinctive motivators for this work, which are not satisfied by
existing perf buffer, which prompted creation of a new ring buffer
implementation.
- more efficient memory utilization by sharing ring buffer across CPUs;
- preserving ordering of events that happen sequentially in time, even
across multiple CPUs (e.g., fork/exec/exit events for a task).
These two problems are independent, but perf buffer fails to satisfy both.
Both are a result of a choice to have per-CPU perf ring buffer. Both can be
also solved by having an MPSC implementation of ring buffer. The ordering
problem could technically be solved for perf buffer with some in-kernel
counting, but given the first one requires an MPSC buffer, the same solution
would solve the second problem automatically.
Semantics and APIs
------------------
Single ring buffer is presented to BPF programs as an instance of BPF map of
type BPF_MAP_TYPE_RINGBUF. Two other alternatives considered, but ultimately
rejected.
One way would be to, similar to BPF_MAP_TYPE_PERF_EVENT_ARRAY, make
BPF_MAP_TYPE_RINGBUF could represent an array of ring buffers, but not enforce
"same CPU only" rule. This would be more familiar interface compatible with
existing perf buffer use in BPF, but would fail if application needed more
advanced logic to lookup ring buffer by arbitrary key. HASH_OF_MAPS addresses
this with current approach. Additionally, given the performance of BPF
ringbuf, many use cases would just opt into a simple single ring buffer shared
among all CPUs, for which current approach would be an overkill.
Another approach could introduce a new concept, alongside BPF map, to
represent generic "container" object, which doesn't necessarily have key/value
interface with lookup/update/delete operations. This approach would add a lot
of extra infrastructure that has to be built for observability and verifier
support. It would also add another concept that BPF developers would have to
familiarize themselves with, new syntax in libbpf, etc. But then would really
provide no additional benefits over the approach of using a map.
BPF_MAP_TYPE_RINGBUF doesn't support lookup/update/delete operations, but so
doesn't few other map types (e.g., queue and stack; array doesn't support
delete, etc).
The approach chosen has an advantage of re-using existing BPF map
infrastructure (introspection APIs in kernel, libbpf support, etc), being
familiar concept (no need to teach users a new type of object in BPF program),
and utilizing existing tooling (bpftool). For common scenario of using
a single ring buffer for all CPUs, it's as simple and straightforward, as
would be with a dedicated "container" object. On the other hand, by being
a map, it can be combined with ARRAY_OF_MAPS and HASH_OF_MAPS map-in-maps to
implement a wide variety of topologies, from one ring buffer for each CPU
(e.g., as a replacement for perf buffer use cases), to a complicated
application hashing/sharding of ring buffers (e.g., having a small pool of
ring buffers with hashed task's tgid being a look up key to preserve order,
but reduce contention).
Key and value sizes are enforced to be zero. max_entries is used to specify
the size of ring buffer and has to be a power of 2 value.
There are a bunch of similarities between perf buffer
(BPF_MAP_TYPE_PERF_EVENT_ARRAY) and new BPF ring buffer semantics:
- variable-length records;
- if there is no more space left in ring buffer, reservation fails, no
blocking;
- memory-mappable data area for user-space applications for ease of
consumption and high performance;
- epoll notifications for new incoming data;
- but still the ability to do busy polling for new data to achieve the
lowest latency, if necessary.
BPF ringbuf provides two sets of APIs to BPF programs:
- bpf_ringbuf_output() allows to *copy* data from one place to a ring
buffer, similarly to bpf_perf_event_output();
- bpf_ringbuf_reserve()/bpf_ringbuf_commit()/bpf_ringbuf_discard() APIs
split the whole process into two steps. First, a fixed amount of space is
reserved. If successful, a pointer to a data inside ring buffer data area
is returned, which BPF programs can use similarly to a data inside
array/hash maps. Once ready, this piece of memory is either committed or
discarded. Discard is similar to commit, but makes consumer ignore the
record.
bpf_ringbuf_output() has disadvantage of incurring extra memory copy, because
record has to be prepared in some other place first. But it allows to submit
records of the length that's not known to verifier beforehand. It also closely
matches bpf_perf_event_output(), so will simplify migration significantly.
bpf_ringbuf_reserve() avoids the extra copy of memory by providing a memory
pointer directly to ring buffer memory. In a lot of cases records are larger
than BPF stack space allows, so many programs have use extra per-CPU array as
a temporary heap for preparing sample. bpf_ringbuf_reserve() avoid this needs
completely. But in exchange, it only allows a known constant size of memory to
be reserved, such that verifier can verify that BPF program can't access
memory outside its reserved record space. bpf_ringbuf_output(), while slightly
slower due to extra memory copy, covers some use cases that are not suitable
for bpf_ringbuf_reserve().
The difference between commit and discard is very small. Discard just marks
a record as discarded, and such records are supposed to be ignored by consumer
code. Discard is useful for some advanced use-cases, such as ensuring
all-or-nothing multi-record submission, or emulating temporary malloc()/free()
within single BPF program invocation.
Each reserved record is tracked by verifier through existing
reference-tracking logic, similar to socket ref-tracking. It is thus
impossible to reserve a record, but forget to submit (or discard) it.
bpf_ringbuf_query() helper allows to query various properties of ring buffer.
Currently 4 are supported:
- BPF_RB_AVAIL_DATA returns amount of unconsumed data in ring buffer;
- BPF_RB_RING_SIZE returns the size of ring buffer;
- BPF_RB_CONS_POS/BPF_RB_PROD_POS returns current logical possition of
consumer/producer, respectively.
Returned values are momentarily snapshots of ring buffer state and could be
off by the time helper returns, so this should be used only for
debugging/reporting reasons or for implementing various heuristics, that take
into account highly-changeable nature of some of those characteristics.
One such heuristic might involve more fine-grained control over poll/epoll
notifications about new data availability in ring buffer. Together with
BPF_RB_NO_WAKEUP/BPF_RB_FORCE_WAKEUP flags for output/commit/discard helpers,
it allows BPF program a high degree of control and, e.g., more efficient
batched notifications. Default self-balancing strategy, though, should be
adequate for most applications and will work reliable and efficiently already.
Design and implementation
-------------------------
This reserve/commit schema allows a natural way for multiple producers, either
on different CPUs or even on the same CPU/in the same BPF program, to reserve
independent records and work with them without blocking other producers. This
means that if BPF program was interruped by another BPF program sharing the
same ring buffer, they will both get a record reserved (provided there is
enough space left) and can work with it and submit it independently. This
applies to NMI context as well, except that due to using a spinlock during
reservation, in NMI context, bpf_ringbuf_reserve() might fail to get a lock,
in which case reservation will fail even if ring buffer is not full.
The ring buffer itself internally is implemented as a power-of-2 sized
circular buffer, with two logical and ever-increasing counters (which might
wrap around on 32-bit architectures, that's not a problem):
- consumer counter shows up to which logical position consumer consumed the
data;
- producer counter denotes amount of data reserved by all producers.
Each time a record is reserved, producer that "owns" the record will
successfully advance producer counter. At that point, data is still not yet
ready to be consumed, though. Each record has 8 byte header, which contains
the length of reserved record, as well as two extra bits: busy bit to denote
that record is still being worked on, and discard bit, which might be set at
commit time if record is discarded. In the latter case, consumer is supposed
to skip the record and move on to the next one. Record header also encodes
record's relative offset from the beginning of ring buffer data area (in
pages). This allows bpf_ringbuf_commit()/bpf_ringbuf_discard() to accept only
the pointer to the record itself, without requiring also the pointer to ring
buffer itself. Ring buffer memory location will be restored from record
metadata header. This significantly simplifies verifier, as well as improving
API usability.
Producer counter increments are serialized under spinlock, so there is
a strict ordering between reservations. Commits, on the other hand, are
completely lockless and independent. All records become available to consumer
in the order of reservations, but only after all previous records where
already committed. It is thus possible for slow producers to temporarily hold
off submitted records, that were reserved later.
Reservation/commit/consumer protocol is verified by litmus tests in
Documentation/litmus-test/bpf-rb.
One interesting implementation bit, that significantly simplifies (and thus
speeds up as well) implementation of both producers and consumers is how data
area is mapped twice contiguously back-to-back in the virtual memory. This
allows to not take any special measures for samples that have to wrap around
at the end of the circular buffer data area, because the next page after the
last data page would be first data page again, and thus the sample will still
appear completely contiguous in virtual memory. See comment and a simple ASCII
diagram showing this visually in bpf_ringbuf_area_alloc().
Another feature that distinguishes BPF ringbuf from perf ring buffer is
a self-pacing notifications of new data being availability.
bpf_ringbuf_commit() implementation will send a notification of new record
being available after commit only if consumer has already caught up right up
to the record being committed. If not, consumer still has to catch up and thus
will see new data anyways without needing an extra poll notification.
Benchmarks (see tools/testing/selftests/bpf/benchs/bench_ringbuf.c) show that
this allows to achieve a very high throughput without having to resort to
tricks like "notify only every Nth sample", which are necessary with perf
buffer. For extreme cases, when BPF program wants more manual control of
notifications, commit/discard/output helpers accept BPF_RB_NO_WAKEUP and
BPF_RB_FORCE_WAKEUP flags, which give full control over notifications of data
availability, but require extra caution and diligence in using this API.
Comparison to alternatives
--------------------------
Before considering implementing BPF ring buffer from scratch existing
alternatives in kernel were evaluated, but didn't seem to meet the needs. They
largely fell into few categores:
- per-CPU buffers (perf, ftrace, etc), which don't satisfy two motivations
outlined above (ordering and memory consumption);
- linked list-based implementations; while some were multi-producer designs,
consuming these from user-space would be very complicated and most
probably not performant; memory-mapping contiguous piece of memory is
simpler and more performant for user-space consumers;
- io_uring is SPSC, but also requires fixed-sized elements. Naively turning
SPSC queue into MPSC w/ lock would have subpar performance compared to
locked reserve + lockless commit, as with BPF ring buffer. Fixed sized
elements would be too limiting for BPF programs, given existing BPF
programs heavily rely on variable-sized perf buffer already;
- specialized implementations (like a new printk ring buffer, [0]) with lots
of printk-specific limitations and implications, that didn't seem to fit
well for intended use with BPF programs.
[0] https://lwn.net/Articles/779550/
Signed-off-by: Andrii Nakryiko <andriin@fb.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Link: https://lore.kernel.org/bpf/20200529075424.3139988-2-andriin@fb.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
|