Re: [RFC][PATCH 00/16] objtool: vmlinux.o and noinstr validation

From: Masami Hiramatsu
Date: Tue Mar 17 2020 - 10:20:10 EST


On Tue, 17 Mar 2020 10:26:49 +0100
Thomas Gleixner <tglx@xxxxxxxxxxxxx> wrote:

> Masami Hiramatsu <mhiramat@xxxxxxxxxx> writes:
> > BTW, out of curiously, if BUG*() or WARN*() cases happens in the noinstr
> > section, do we also need to move them (register dump, stack unwinding,
> > printk, console output, etc.) all into noinstr section?
>
> The current plan is to declare BUG()/WARN() "safe". On x86 it is kinda
> safe as it uses UD2. That raises an exception which handles the bug/warn
> after establishing the correct state.

OK, so at least the entry of BUG()/WARN() is in noinstr, but the
register/stack dump routines are out of noinstr.

>
> Of course it's debatable, but moving all of this into the noinstr
> section might be just a too wide scope.

Agreed.

Thank you,

--
Masami Hiramatsu <mhiramat@xxxxxxxxxx>