Re: [PATCH v5 23/26] perf annotate-arm64: Support 'add' instruction tracking

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] Failure to clear `lifetime_active` and `lifetime_end` in `update_add_insn_state` for the destination register.
--

commit 1ca68e83139c25b5687ab59598c804f32ea92afb
Author: Tengda Wu <wutengda@huaweicloud.com>

perf annotate-arm64: Support 'add' instruction tracking

This commit extends update_insn_state_arm64() to track 'add'
instructions for structure member address calculation. It enables the
data type tracker to follow member address calculations and propagate
types correctly instead of dropping type info on 'add'.

> diff --git a/tools/perf/util/annotate-arch/annotate-arm64.c b/tools/perf/util/annotate-arch/annotate-arm64.c
> index 2f8bedf583c1a..e237b91426849 100644
> --- a/tools/perf/util/annotate-arch/annotate-arm64.c
> +++ b/tools/perf/util/annotate-arch/annotate-arm64.c
> @@ -864,6 +864,72 @@ static void update_mov_insn_state(struct type_state *state,
>  	pr_debug_type_name(&tsr->type, tsr->kind);
>  }
>  
> +static void update_add_insn_state(struct type_state *state,
> +				  struct data_loc_info *dloc,
> +				  struct disasm_line *dl,
> +				  struct annotated_op_loc *src,
> +				  struct annotated_op_loc *dst)
> +{
> +	struct type_state_reg *tsr;
> +	struct type_state_reg src_tsr;
> +	u32 insn_offset = dl->al.offset;
> +	int sreg = src->reg1;
> +	int dreg = dst->reg1;
> +	int reg_offset;
> +
> +	if (!has_reg_type(state, dreg))
> +		return;
> +
> +	tsr = &state->regs[dreg];
> +	tsr->copied_from = -1;

[Severity: Medium]
Does this code need to clear lifetime_active and lifetime_end on the
destination register?

When the add instruction overwrites a destination register that previously
tracked a variable with an active DWARF lifetime, it looks like the stale
DWARF lifetime bounds are incorrectly retained on the destination register.
Could this cause caller-saved registers to incorrectly survive function call
invalidation and corrupt data type tracking during profiling?

[ ... ]

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/cover.1788872630.git.wutengda@huaweicloud.com?part=23