linux-kernel - Re: [PATCH v7 2/3] arm64: implement ftrace with regs

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]

Date:   Mon, 4 Feb 2019 13:03:09 +0100
From:   Torsten Duwe <duwe@....de>
To:     Ard Biesheuvel <ard.biesheuvel@...aro.org>
Cc:     Julien Thierry <julien.thierry@....com>,
        Mark Rutland <mark.rutland@....com>,
        Will Deacon <will.deacon@....com>,
        Catalin Marinas <catalin.marinas@....com>,
        Steven Rostedt <rostedt@...dmis.org>,
        Josh Poimboeuf <jpoimboe@...hat.com>,
        Ingo Molnar <mingo@...hat.com>, Arnd Bergmann <arnd@...db.de>,
        AKASHI Takahiro <takahiro.akashi@...aro.org>,
        Amit Daniel Kachhap <amit.kachhap@....com>,
        "Singh, Balbir" <bsingharora@...il.com>,
        linux-arm-kernel <linux-arm-kernel@...ts.infradead.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        live-patching@...r.kernel.org
Subject: Re: [PATCH v7 2/3] arm64: implement ftrace with regs

On Tue, Jan 22, 2019 at 02:55:12PM +0100, Ard Biesheuvel wrote:
> On Tue, 22 Jan 2019 at 14:28, Torsten Duwe <duwe@....de> wrote:
> >
> > On Tue, Jan 22, 2019 at 10:18:17AM +0000, Julien Thierry wrote:
> > > Hi Torsten,
> > >
> > > A few suggestions below.
> > >
> > > > +#ifdef CONFIG_DYNAMIC_FTRACE_WITH_REGS
> > > > +#define ARCH_SUPPORTS_FTRACE_OPS 1
> > > > +#define REC_IP_BRANCH_OFFSET AARCH64_INSN_SIZE
> > > > +/* All we need is some magic value. Simply use "_mCount:" */
> > >
> > > Nit: Should the casing be "_mcount" ?
> >
> > No! The string makes it clear what it's supposed to be and the peculiar
> > casing makes it unique and leaves no doubt were it came from. So whenever
> > you see this register value in a crash dump you don't have to wonder about
> > weird linkage errors, as it surely did not originate from a symtab.
> >
> 
> In that case, do you need to deal with endianness here?
> 
> > > > +#define MCOUNT_ADDR                (0x5f6d436f756e743a)

Strictly speaking, yes. OTOH "wrong-andian" machines always show a difference
when memory is dumped in bigger units than bytes, so when you see the register
value in hex...

Since all that's needed is a somewhat unique constant, let's not over-engineer
this ok?

If there are no other comments I'll send out v8 with the discussed changes.

	Torsten