[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20241213062131.gk75bqcm3eclhb4d@jpoimboe>
Date: Thu, 12 Dec 2024 22:21:31 -0800
From: Josh Poimboeuf <jpoimboe@...hat.com>
To: Valentin Schneider <vschneid@...hat.com>
Cc: linux-kernel@...r.kernel.org, Peter Zijlstra <peterz@...radead.org>
Subject: Re: [PATCH v2 3/3] objtool: Add noinstr validation for static
branches/calls
On Tue, Dec 03, 2024 at 01:02:21PM +0100, Valentin Schneider wrote:
> On 26/11/24 20:47, Josh Poimboeuf wrote:
> > @@ -3532,14 +3577,29 @@ static inline bool noinstr_call_dest(struct objtool_file *file,
> > return false;
> > }
> >
> > +static char *static_call_name(struct symbol *func)
> > +{
> > + return func->name + strlen("__SCT__");
> > +}
> > +
> > static int validate_call(struct objtool_file *file,
> > struct instruction *insn,
> > struct insn_state *state)
> > {
> > - if (state->noinstr && state->instr <= 0 &&
> > - !noinstr_call_dest(file, insn, insn_call_dest(insn))) {
> > - WARN_INSN(insn, "call to %s() leaves .noinstr.text section", call_dest_name(insn));
> > - return 1;
> > + if (state->noinstr && state->instr <= 0) {
> > + struct symbol *dest = insn_call_dest(insn);
> > +
>
> Interestingly only after I tried making __flush_tlb_all() noinstr did this
> start causing a segfault due to a NULL dest. I added a NULL check here but
> I wonder if I'm not just plastering over some other issue?
>
> > + if (dest->static_call_tramp) {
Yeah, this line can probably just be:
if (dest && dest->static_call_tramp)
--
Josh
Powered by blists - more mailing lists