All of lore.kernel.org
 help / color / mirror / Atom feed
From: Ingo Molnar <mingo@kernel.org>
To: Pavel Machek <pavel@ucw.cz>
Cc: Kees Cook <keescook@chromium.org>,
	Peter Zijlstra <peterz@infradead.org>,
	Arnaldo Carvalho de Melo <acme@redhat.com>,
	kernel list <linux-kernel@vger.kernel.org>,
	Ingo Molnar <mingo@redhat.com>,
	Alexander Shishkin <alexander.shishkin@linux.intel.com>,
	"kernel-hardening@lists.openwall.com"
	<kernel-hardening@lists.openwall.com>
Subject: [kernel-hardening] Re: rowhammer protection [was Re: Getting interrupt every million cache misses]
Date: Fri, 28 Oct 2016 09:07:01 +0200	[thread overview]
Message-ID: <20161028070701.GA11376@gmail.com> (raw)
In-Reply-To: <20161027212747.GA18147@amd>


* Pavel Machek <pavel@ucw.cz> wrote:

> +static void rh_overflow(struct perf_event *event, struct perf_sample_data *data, struct pt_regs *regs)
> +{
> +	u64 *ts = this_cpu_ptr(&rh_timestamp); /* this is NMI context */
> +	u64 now = ktime_get_mono_fast_ns();
> +	s64 delta = now - *ts;
> +
> +	*ts = now;
> +
> +	/* FIXME msec per usec, reverse logic? */
> +	if (delta < 64 * NSEC_PER_MSEC)
> +		mdelay(56);
> +}

I'd suggest making the absolute delay sysctl tunable, because 'wait 56 msecs' is 
very magic, and do we know it 100% that 56 msecs is what is needed everywhere?

Plus I'd also suggest exposing an 'NMI rowhammer delay count' in /proc/interrupts, 
to make it easier to debug this. (Perhaps only show the line if the count is 
nonzero.)

Finally, could we please also add a sysctl and Kconfig that allows this feature to 
be turned on/off, with the default bootup value determined by the Kconfig value 
(i.e. by the distribution)? Similar to CONFIG_SECURITY_SELINUX_BOOTPARAM_VALUE.

Thanks,

	Ingo

WARNING: multiple messages have this Message-ID (diff)
From: Ingo Molnar <mingo@kernel.org>
To: Pavel Machek <pavel@ucw.cz>
Cc: Kees Cook <keescook@chromium.org>,
	Peter Zijlstra <peterz@infradead.org>,
	Arnaldo Carvalho de Melo <acme@redhat.com>,
	kernel list <linux-kernel@vger.kernel.org>,
	Ingo Molnar <mingo@redhat.com>,
	Alexander Shishkin <alexander.shishkin@linux.intel.com>,
	"kernel-hardening@lists.openwall.com" 
	<kernel-hardening@lists.openwall.com>
Subject: Re: rowhammer protection [was Re: Getting interrupt every million cache misses]
Date: Fri, 28 Oct 2016 09:07:01 +0200	[thread overview]
Message-ID: <20161028070701.GA11376@gmail.com> (raw)
In-Reply-To: <20161027212747.GA18147@amd>


* Pavel Machek <pavel@ucw.cz> wrote:

> +static void rh_overflow(struct perf_event *event, struct perf_sample_data *data, struct pt_regs *regs)
> +{
> +	u64 *ts = this_cpu_ptr(&rh_timestamp); /* this is NMI context */
> +	u64 now = ktime_get_mono_fast_ns();
> +	s64 delta = now - *ts;
> +
> +	*ts = now;
> +
> +	/* FIXME msec per usec, reverse logic? */
> +	if (delta < 64 * NSEC_PER_MSEC)
> +		mdelay(56);
> +}

I'd suggest making the absolute delay sysctl tunable, because 'wait 56 msecs' is 
very magic, and do we know it 100% that 56 msecs is what is needed everywhere?

Plus I'd also suggest exposing an 'NMI rowhammer delay count' in /proc/interrupts, 
to make it easier to debug this. (Perhaps only show the line if the count is 
nonzero.)

Finally, could we please also add a sysctl and Kconfig that allows this feature to 
be turned on/off, with the default bootup value determined by the Kconfig value 
(i.e. by the distribution)? Similar to CONFIG_SECURITY_SELINUX_BOOTPARAM_VALUE.

Thanks,

	Ingo

  reply	other threads:[~2016-10-28  7:07 UTC|newest]

Thread overview: 56+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-10-26 20:54 Getting interrupt every million cache misses Pavel Machek
2016-10-27  8:28 ` Peter Zijlstra
2016-10-27  8:46   ` Pavel Machek
2016-10-27  9:15     ` Peter Zijlstra
2016-10-27  9:11   ` Pavel Machek
2016-10-27  9:33     ` Peter Zijlstra
2016-10-27 20:40       ` [kernel-hardening] " Kees Cook
2016-10-27 20:40         ` Kees Cook
2016-10-27 21:27         ` [kernel-hardening] rowhammer protection [was Re: Getting interrupt every million cache misses] Pavel Machek
2016-10-27 21:27           ` Pavel Machek
2016-10-28  7:07           ` Ingo Molnar [this message]
2016-10-28  7:07             ` Ingo Molnar
2016-10-28  8:50             ` [kernel-hardening] " Pavel Machek
2016-10-28  8:50               ` Pavel Machek
2016-10-28  8:59               ` [kernel-hardening] " Ingo Molnar
2016-10-28  8:59                 ` Ingo Molnar
2016-10-28 11:55                 ` [kernel-hardening] " Pavel Machek
2016-10-28 11:55                   ` Pavel Machek
2016-10-28  9:04               ` [kernel-hardening] " Peter Zijlstra
2016-10-28  9:04                 ` Peter Zijlstra
2016-10-28  9:27                 ` [kernel-hardening] " Vegard Nossum
2016-10-28  9:27                   ` Vegard Nossum
2016-10-28  9:35                   ` [kernel-hardening] " Ingo Molnar
2016-10-28  9:35                     ` Ingo Molnar
2016-10-28  9:47                     ` [kernel-hardening] " Vegard Nossum
2016-10-28  9:47                       ` Vegard Nossum
2016-10-28  9:53                     ` [kernel-hardening] " Mark Rutland
2016-10-28 11:27                 ` Pavel Machek
2016-10-28 11:27                   ` Pavel Machek
2016-10-28  9:51           ` [kernel-hardening] " Mark Rutland
2016-10-28 11:21             ` Pavel Machek
2016-10-28 14:05               ` Mark Rutland
2016-10-28 14:18                 ` Peter Zijlstra
2016-10-28 18:30                   ` Pavel Machek
2016-10-28 18:48                     ` Peter Zijlstra
2016-11-02 18:13                   ` Pavel Machek
2016-10-28 17:27                 ` Pavel Machek
2016-10-29 13:06                   ` Daniel Gruss
2016-10-29 19:42                     ` Pavel Machek
2016-10-29 20:05                       ` Daniel Gruss
2016-10-29 20:14                         ` Daniel Gruss
2016-10-29 21:05                         ` Pavel Machek
2016-10-29 21:07                           ` Daniel Gruss
2016-10-29 21:45                             ` Pavel Machek
2016-10-29 21:49                               ` Daniel Gruss
2016-10-29 22:01                                 ` Pavel Machek
2016-10-29 22:02                                   ` Daniel Gruss
2016-10-31  8:27                 ` Pavel Machek
2016-10-31 14:47                   ` Mark Rutland
2016-10-31 21:13                     ` Pavel Machek
2016-10-31 22:09                       ` Mark Rutland
2016-11-01  6:33                   ` Ingo Molnar
2016-11-01  7:20                     ` Daniel Micay
2016-11-01  7:53                     ` Daniel Gruss
2016-11-01  8:10                     ` Pavel Machek
2016-11-01  8:13                       ` Daniel Gruss

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20161028070701.GA11376@gmail.com \
    --to=mingo@kernel.org \
    --cc=acme@redhat.com \
    --cc=alexander.shishkin@linux.intel.com \
    --cc=keescook@chromium.org \
    --cc=kernel-hardening@lists.openwall.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@redhat.com \
    --cc=pavel@ucw.cz \
    --cc=peterz@infradead.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body. 
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.