linux-kernel - Re: [TRIVIAL PATCH 16/26] x86: Convert print

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 for Android: free password hash cracker in your pocket

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

Date:	Thu, 13 Dec 2012 20:05:44 +0100
From:	Borislav Petkov <bp@...en8.de>
To:	"H. Peter Anvin" <hpa@...or.com>
Cc:	Joe Perches <joe@...ches.com>, "Luck, Tony" <tony.luck@...el.com>,
	Jiri Kosina <trivial@...nel.org>,
	Jeff Dike <jdike@...toit.com>,
	Richard Weinberger <richard@....at>,
	Thomas Gleixner <tglx@...utronix.de>,
	Ingo Molnar <mingo@...hat.com>,
	"x86@...nel.org" <x86@...nel.org>,
	"linux-edac@...r.kernel.org" <linux-edac@...r.kernel.org>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
	"user-mode-linux-devel@...ts.sourceforge.net" 
	<user-mode-linux-devel@...ts.sourceforge.net>,
	"user-mode-linux-user@...ts.sourceforge.net" 
	<user-mode-linux-user@...ts.sourceforge.net>
Subject: Re: [TRIVIAL PATCH 16/26] x86: Convert print_symbol to %pSR

On Thu, Dec 13, 2012 at 10:42:28AM -0800, H. Peter Anvin wrote:
> On 12/13/2012 10:37 AM, Borislav Petkov wrote:
> >>
> >> If appropriate, the code could be changed to
> >>
> >> 	(void *)(unsigned long)m->ip
> > 
> > Can we explicitly cast it to what it is so that we can be explicit as to
> > what we're casting it? IOW:
> > 
> > 	(void *)(__u64)m->ip;
> > 
> > Does that even work on 32bit?
> > 
> > Also, does the compiler bitch about this useless cast when building with
> > W=123?
> > 
> 
> Uh... no.
> 
> The point is that (void *)(unsigned long) casts it to an integer of
> pointer size -- in userspace you would to (void *)(size_t) or (void
> *)(uintptr_t) -- so that the compiler knows "I meant to do that."

Ok, I grok it now. So, in most cases, mce.ip comes from pt_regs.ip which
is unsigned long so we're fine for both 32- and 64-bit. There's only the
case with precise RIP reporting where we get the rip from an MSR but
even then, it should be 4 bytes on 32-bit since our addresses there are
4 bytes.

So, actually struct mce should've had ->ip defined as an unsigned long
from the very beginning. But we can't change that now because this
struct is shared with userspace :(

So yes Joe, make it (void *)(unsigned long).

Thanks guys.

-- 
Regards/Gruss,
    Boris.

Sent from a fat crate under my desk. Formatting is fine.
--
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/