On Sat, 2025-10-25 at 13:30 +0800, KaFai Wan wrote:
> When conditional jumps are performed on the same register (e.g., r0 <= r0,
> r0 > r0, r0 < r0) where the register holds a scalar with range, the verifier
> incorrectly attempts to adjust the register's min/max bounds. This leads to
> invalid range bounds and triggers a BUG warning:
>
> verifier bug: REG INVARIANTS VIOLATION (true_reg1): range bounds violation
> u64=[0x1, 0x0] s64=[0x1, 0x0] u32=[0x1, 0x0] s32=[0x1, 0x0] var_off=(0x0, 0x0)
> WARNING: CPU: 0 PID: 92 at kernel/bpf/verifier.c:2731
> reg_bounds_sanity_check+0x163/0x220
> Hardware name: QEMU Ubuntu 24.04 PC (i440FX + PIIX, 1996), BIOS
> 1.16.3-debian-1.16.3-2 04/01/2014
> RIP: 0010:reg_bounds_sanity_check+0x163/0x220
> Call Trace:
> <TASK>
> reg_set_min_max+0xf7/0x1d0
> check_cond_jmp_op+0x57b/0x1730
> ? print_bpf_insn+0x3d5/0xa50
> do_check_common+0x33ac/0x33c0
> ...
>
> The root cause is in regs_refine_cond_op() where BPF_JLT/BPF_JSLT operations
> adjust both min/max bounds on the same register, causing invalid bounds.
>
> Since comparing a register with itself should not change its bounds (the
> comparison result is always known: r0 == r0 is always true, r0 < r0 is
> always false), the bounds adjustment is unnecessary.
>
> Fix this by:
> 1. Enhance is_branch_taken() and is_scalar_branch_taken() to properly
> handle branch direction computation for same register comparisons
> across all BPF jump operations
> 2. For unknown branch directions (e.g., BPF_JSET), add early return in
> reg_set_min_max() to avoid bounds adjustment on the same register
>
> The fix ensures that unnecessary bounds adjustments are skipped, preventing
> the verifier bug while maintaining correct branch direction analysis.
>
> Reported-by: Kaiyan Mei <[email protected]>
> Reported-by: Yinhao Hu <[email protected]>
> Closes:
> https://lore.kernel.org/all/[email protected]/
> Fixes: 0df1a55afa83 ("bpf: Warn on internal verifier errors")
> Signed-off-by: KaFai Wan <[email protected]>
> ---
> kernel/bpf/verifier.c | 32 ++++++++++++++++++++++++++++++++
> 1 file changed, 32 insertions(+)
>
> diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
> index 6d175849e57a..653fa96ed0df 100644
> --- a/kernel/bpf/verifier.c
> +++ b/kernel/bpf/verifier.c
> @@ -16037,6 +16037,12 @@ static int is_scalar_branch_taken(struct
> bpf_reg_state *reg1, struct bpf_reg_sta
> }
> break;
> case BPF_JSET:
> + if (reg1 == reg2) {
> + if (tnum_is_const(t1))
> + return t1.value != 0;
> + else
> + return (smin1 <= 0 && smax1 >= 0) ? -1 : 1;
> + }
I think this logic is fine, but it needs tests for multiple cases.
> if (!is_reg_const(reg2, is_jmp32)) {
> swap(reg1, reg2);
> swap(t1, t2);
> @@ -16172,6 +16178,25 @@ static int is_pkt_ptr_branch_taken(struct
> bpf_reg_state *dst_reg,
> static int is_branch_taken(struct bpf_reg_state *reg1, struct bpf_reg_state
> *reg2,
> u8 opcode, bool is_jmp32)
> {
> + if (reg1 == reg2) {
> + switch (opcode) {
> + case BPF_JGE:
> + case BPF_JLE:
> + case BPF_JSGE:
> + case BPF_JSLE:
> + case BPF_JEQ:
> + return 1;
> + case BPF_JGT:
> + case BPF_JLT:
> + case BPF_JSGT:
> + case BPF_JSLT:
> + case BPF_JNE:
> + return 0;
> + default:
> + break;
> + }
> + }
> +
I think Alexei was against my suggestion to put it in
is_branch_taken() and preferred is_scalar_branch_taken() instead.
> if (reg_is_pkt_pointer_any(reg1) && reg_is_pkt_pointer_any(reg2) &&
> !is_jmp32)
> return is_pkt_ptr_branch_taken(reg1, reg2, opcode);
>
> @@ -16429,6 +16454,13 @@ static int reg_set_min_max(struct bpf_verifier_env
> *env,
> if (false_reg1->type != SCALAR_VALUE || false_reg2->type !=
> SCALAR_VALUE)
> return 0;
>
> + /* We compute branch direction for same registers in is_branch_taken()
> and
> + * is_scalar_branch_taken(). For unknown branch directions (e.g.,
> BPF_JSET)
> + * on the same registers, we don't need to adjusts the min/max values.
> + */
> + if (false_reg1 == false_reg2)
> + return 0;
> +
> /* fallthrough (FALSE) branch */
> regs_refine_cond_op(false_reg1, false_reg2, rev_opcode(opcode),
> is_jmp32);
> reg_bounds_sync(false_reg1);
