linux-kernel - Re: [PATCH v2 06/10] perf annotate: Track arithmetic instructions on pointers

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CAJUgMyKBFkF9rKFnzP6S_c9ma3P7a-NsCubV1DiTwoGUbByk+w@mail.gmail.com>
Date: Wed, 3 Sep 2025 16:26:50 -0400
From: Zecheng Li <zecheng@...gle.com>
To: Namhyung Kim <namhyung@...nel.org>
Cc: Peter Zijlstra <peterz@...radead.org>, Ingo Molnar <mingo@...hat.com>, 
	Arnaldo Carvalho de Melo <acme@...nel.org>, Mark Rutland <mark.rutland@....com>, 
	Alexander Shishkin <alexander.shishkin@...ux.intel.com>, Jiri Olsa <jolsa@...nel.org>, 
	Ian Rogers <irogers@...gle.com>, Adrian Hunter <adrian.hunter@...el.com>, 
	"Liang, Kan" <kan.liang@...ux.intel.com>, Masami Hiramatsu <mhiramat@...nel.org>, 
	Xu Liu <xliuprof@...gle.com>, linux-perf-users@...r.kernel.org, 
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2 06/10] perf annotate: Track arithmetic instructions on pointers

> I was thinking we can use TSR_KIND_POINTER here.  Probably we need to
> distinguish it from the existing percpu base use case. Maybe you want
> to rename it.  Then you don't need to find a pointer type in the DWARF.

Hi Namhyung, thanks for the review.

Yes. I think that's a better way to represent a pointer to a type.
With that tag we can also support registers representing addresses
(is_reg_var_addr == true). And when moving an address register to a
memory location we probably can simply add the * to denote a pointer
without finding the exact pointer type.

>
> > +                     tsr->offset = 0;
>
> I think it can refer to a member in a struct.  How about this?
>
>                         tsr->offset = state->offset - offset;

True, it may load an address of a struct member. Do you mean
stack->offset - offset?

> > @@ -388,6 +493,7 @@ static void update_insn_state_x86(struct type_state *state,
> >               tsr->type = state->regs[src->reg1].type;
> >               tsr->kind = state->regs[src->reg1].kind;
> >               tsr->imm_value = state->regs[src->reg1].imm_value;
> > +             tsr->offset = 0;
> >               tsr->ok = true;

I also find for mov register to register this should be

tsr->offset = state->regs[src->reg1].offset;

Let me update the patch.

Zecheng