Alexei Starovoitov 98d7ca374b bpf: Track delta between "linked" registers. ... Compilers can generate the code r1 = r2 r1 += 0x1 if r2 < 1000 goto ... use knowledge of r2 range in subsequent r1 operations So remember constant delta between r2 and r1 and update r1 after 'if' condition. Unfortunately LLVM still uses this pattern for loops with 'can_loop' construct: for (i = 0; i < 1000 && can_loop; i++) The "undo" pass was introduced in LLVM https://reviews.llvm.org/D121937 to prevent this optimization, but it cannot cover all cases. Instead of fighting middle end optimizer in BPF backend teach the verifier about this pattern. Signed-off-by: Alexei Starovoitov <ast@kernel.org> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Acked-by: Eduard Zingerman <eddyz87@gmail.com> Link: https://lore.kernel.org/bpf/20240613013815.953-3-alexei.starovoitov@gmail.com		2024-06-14 21:52:39 +02:00
..
.gitignore
atomic_and.c	bpf, x86: Fix BPF_FETCH atomic and/or/xor with r0 as src	2021-02-22 18:03:11 +01:00
atomic_bounds.c	bpf: Propagate stack bounds to registers in atomics w/ BPF_FETCH	2021-02-02 18:23:29 -08:00
atomic_cmpxchg.c	bpf: Fix accesses to uninit stack slots	2023-12-08 14:19:00 -08:00
atomic_fetch_add.c	bpf: Add tests for new BPF atomic operations	2021-01-14 18:34:29 -08:00
atomic_fetch.c	bpf, selftests: Add test case for atomic fetch on spilled pointer	2021-12-14 19:33:06 -08:00
atomic_invalid.c	bpf: Small BPF verifier log improvements	2022-03-03 16:54:10 +01:00
atomic_or.c	bpf: Explicitly zero-extend R0 after 32-bit cmpxchg	2021-03-04 19:06:03 -08:00
atomic_xchg.c	bpf: Add tests for new BPF atomic operations	2021-01-14 18:34:29 -08:00
atomic_xor.c	selftests/bpf: Fix endianness issues in atomic tests	2021-02-10 11:55:22 -08:00
basic_call.c
basic_instr.c	selftests/bpf: Fix a test_verifier failure	2023-07-27 18:54:16 -07:00
basic_stx_ldx.c
basic.c
bpf_loop_inline.c	selftests/bpf: Skip callback tests if jit is disabled in test_verifier	2024-01-23 20:27:41 -08:00
bpf_st_mem.c	selftests/bpf: Add test for immediate spilled to stack	2023-11-01 22:30:27 -07:00
calls.c	bpf: Fix accesses to uninit stack slots	2023-12-08 14:19:00 -08:00
ctx_sk_lookup.c	selftests/bpf: Add tests for accessing ingress_ifindex in bpf_sk_lookup	2021-11-10 16:29:59 -08:00
ctx_skb.c	selftests/bpf: Add F_NEEDS_EFFICIENT_UNALIGNED_ACCESS to some tests	2023-07-05 14:34:23 +02:00
dead_code.c	selftests, bpf: Test that dead ldx_w insns are accepted	2021-08-13 17:46:26 +02:00
direct_value_access.c
event_output.c
jit.c	bpf: add selftests for lsh, rsh, arsh with reg operand	2022-10-19 16:53:51 -07:00
jmp32.c	selftests/bpf: Add F_NEEDS_EFFICIENT_UNALIGNED_ACCESS to some tests	2023-07-05 14:34:23 +02:00
jset.c	bpf, selftests: Adjust few selftest outcomes wrt unreachable code	2021-06-14 23:06:38 +02:00
jump.c	bpf, selftests: Add verifier test case for imm=0,umin=0,umax=1 scalar	2022-07-01 12:56:27 -07:00
junk_insn.c
ld_abs.c
ld_dw.c
ld_imm64.c	bpf: handle ldimm64 properly in check_cfg()	2023-11-09 20:11:20 -08:00
map_kptr.c	selftests/bpf: Add F_NEEDS_EFFICIENT_UNALIGNED_ACCESS to some tests	2023-07-05 14:34:23 +02:00
perf_event_sample_period.c	selftests/bpf: Use __BYTE_ORDER__	2021-10-25 20:39:42 -07:00
precise.c	bpf: Track delta between "linked" registers.	2024-06-14 21:52:39 +02:00
scale.c
sleepable.c	bpf: Allow BPF_PROG_TYPE_STRUCT_OPS programs to be sleepable	2023-01-25 10:25:57 -08:00
wide_access.c