linux-kernel - Re: [PATCH 1/1] i386: Geode's TSC is not neccessary to mark tu unstable

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] [day] [month] [year] [list]

Message-ID: <a781481a0707220917m710f2df3x1a27bdbb523099d7@mail.gmail.com>
Date:	Sun, 22 Jul 2007 21:47:06 +0530
From:	"Satyam Sharma" <satyam.sharma@...il.com>
To:	"Andi Kleen" <ak@...e.de>
Cc:	"Andres Salomon" <dilinger@...ued.net>,
	"Juergen Beisert" <juergen127@...uzholzen.de>,
	"Andrew Morton" <akpm@...ux-foundation.org>,
	"TAKADA Yoshihito" <takada@....nifty.com>,
	linux-kernel@...r.kernel.org,
	"Jordan Crouse" <jordan.crouse@....com>,
	"Andres Salomon" <dilinger@...ian.org>,
	"Alan Cox" <alan@...rguk.ukuu.org.uk>
Subject: Re: [PATCH 1/1] i386: Geode's TSC is not neccessary to mark tu unstable

> > > On Sun, 15 Jul 2007 21:06:27 +0200
> > > Juergen Beisert <juergen127@...uzholzen.de> wrote:
> > > > Replace NSC/Cyrix specific chipset access macros by inlined functions.
> > > > With the macros a line like this fails (and does nothing):
> > > >   setCx86(CX86_CCR2, getCx86(CX86_CCR2) | 0x88);
> > > > With inlined functions this line will work as expected.


> > On Thursday 19 July 2007 03:02, Andrew Morton wrote:
> > >
> > > I don't get it.  Why would the macros behave differently from inlined
> > > functions?


> On Thu, 19 Jul 2007 08:49:05 +0200
> Juergen Beisert <juergen127@...uzholzen.de> wrote:
> >
> > X86 magic. The access order is important. The first access must always be the
> > offset at 0x22. This access enables the next access to 0x23 (data). If you do
> > it in wrong order, it fails. With the macros you get something like 0x22,
> > 0x22, 0x23, 0x23. With the inline functions 0x22,0x23,0x22,0x23.


On 7/19/07, Andres Salomon <dilinger@...ued.net> wrote:
>
> Wow, that's a really cool bug; nice work!  Don't forget to update
> arch/i386/kernel/cpu/mtrr/state.c, though; it uses setCx86() as well.  It needs
> to include processor-cyrix.h.


On 7/19/07, Andi Kleen <ak@...e.de> wrote:
>
> It also needs some big fat comments


Ok, I was discussing macros-in-C on some other thread and got
reminded about this one. Anyway, I don't really think there was
anything weird / surprising about this case at all -- it's just another
manifestation of the same age-old time-tested advise all our respective
grandmothers have always given us:

        Never pass arguments that have side-effects to macros.

Of course, ideally the user shouldn't even know that the API call he's
using is a macro or a function -- which puts the onus upon the person
who *wrote* that API to ensure that he doesn't write macros for what
could, and should, easily be functions. Macros are generally evil and
always horrible, all IMHO, of course.

Satyam
-
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/