linux-kernel - Re: [RFC PATCH] perf: align raw sample data on 64-bit boundaries

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:	Wed, 19 May 2010 09:39:10 +0200
From:	Frederic Weisbecker <fweisbec@...il.com>
To:	Robert Richter <robert.richter@....com>
Cc:	Peter Zijlstra <a.p.zijlstra@...llo.nl>,
	Stephane Eranian <eranian@...gle.com>,
	Ingo Molnar <mingo@...e.hu>,
	Arnaldo Carvalho de Melo <acme@...radead.org>,
	Paul Mackerras <paulus@...ba.org>,
	LKML <linux-kernel@...r.kernel.org>
Subject: Re: [RFC PATCH] perf: align raw sample data on 64-bit boundaries

On Tue, May 18, 2010 at 05:12:27PM +0200, Robert Richter wrote:
> On 19.04.10 14:19:57, Stephane Eranian wrote:
> > > +       perf_sample_data_init(&data, 0);
> > > +       if (event->attr.sample_type & PERF_SAMPLE_RAW) {
> > > +               for (i = 1; i < size; i++)
> > > +                       rdmsrl(msr++, *buf++);
> > > +               raw.size = sizeof(u64) * size;
> > > +               raw.data = buffer;
> > > +               data.raw = &raw;
> > > +       }
> > > +
> > 
> > Need to add the padding: raw.size = sizeof(u64) * size + sizeof(u32);
> 
> When thinking about this I was wondering if it wouldn't make sense to
> better fix the alignment and move the data buffer to a 64 bit
> boundary. So take a look at the enclosed RFC patch. Though it breaks
> the ABI it would solve some problems. I think more than it introduces.
> Hopefully I found all effected code locations using it.
> 
> -Robert
> 
> --
> 
> From 2427dda67b072f27ecff678f8829b9e2fc537988 Mon Sep 17 00:00:00 2001
> From: Robert Richter <robert.richter@....com>
> Date: Fri, 7 May 2010 15:32:45 +0200
> Subject: [PATCH] perf: align raw sample data on 64-bit boundaries
> 
> In the current implementation 64 bit raw sample data values are not
> aligned due to the 32 bit size header. The size header is located
> before the data buffer on a 64 bit boundary. This leads to unaligned
> memory access to the data buffer for arrays or structures containing
> 64 bit values. To avoid this, the patch adds a 32 bit reserved value
> to the raw sample data header. The data buffer starts then at a 64 bit
> boundary.
> 
> This changes the ABI and requires changes in the userland tools. For
> tools/perf this is at a single location in event.c only. This could
> also introduce some overhead on smaller architectures, but currently
> this is only used on x86 or for transferring raw tracepoint
> data.


No this is used on any architectures that event have a minimal support for
perf events.

I use tracepoint raw samples in sparc 64 for example (which has much more
than the minimal support).



> Though an ABI change should be avoided, it is worth to align raw
> sample data on 64-bit boundaries as the change fixes unaligned memory
> access, eases the implementation for raw sample data and reduces the
> risk of data corruption due to different pad structures inserted by
> the compiler.
> 
> Signed-off-by: Robert Richter <robert.richter@....com>
> ---


I don't think we should do this. Ok it's true we've screwed up
something there but breaking the ABI is going to make the things
even worst I think.

I would feel better with a new PERF_SAMPLE_RAW_ALIGNED sample_type
and schedule the deprecation of PERF_SAMPLE_RAW for later but keep
it for some releases.

What do you think?

--
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/