From: Borislav Petkov <borislav.petkov@amd.com>
To: Thomas Gleixner <tglx@linutronix.de>
Cc: LKML <linux-kernel@vger.kernel.org>,
Tony Luck <tony.luck@intel.com>,
Borislav Petkov <borislav.petkov@amd.com>,
Chen Gong <gong.chen@linux.intel.com>, <x86@kernel.org>,
Peter Zijlstra <peterz@infradead.org>
Subject: Re: [patch 5/5] x86: mce: Add cmci poll mode
Date: Thu, 7 Jun 2012 20:14:14 +0200 [thread overview]
Message-ID: <20120607181414.GH16941@aftab.osrc.amd.com> (raw)
In-Reply-To: <20120606215116.757323156@linutronix.de>
On Wed, Jun 06, 2012 at 09:53:24PM +0000, Thomas Gleixner wrote:
[ … ]
> Index: tip/arch/x86/kernel/cpu/mcheck/mce_intel.c
> ===================================================================
> --- tip.orig/arch/x86/kernel/cpu/mcheck/mce_intel.c
> +++ tip/arch/x86/kernel/cpu/mcheck/mce_intel.c
> @@ -15,6 +15,8 @@
> #include <asm/msr.h>
> #include <asm/mce.h>
>
> +#include "mce-internal.h"
> +
> /*
> * Support for Intel Correct Machine Check Interrupts. This allows
> * the CPU to raise an interrupt when a corrected machine check happened.
> @@ -30,7 +32,22 @@ static DEFINE_PER_CPU(mce_banks_t, mce_b
> */
> static DEFINE_RAW_SPINLOCK(cmci_discover_lock);
>
> -#define CMCI_THRESHOLD 1
> +#define CMCI_THRESHOLD 1
> +#define CMCI_POLL_INTERVAL (30 * HZ)
> +#define CMCI_STORM_INTERVAL (1 * HZ)
> +#define CMCI_STORM_TRESHOLD 5
Just a spelling correction:
CMCI_STORM_THRESHOLD
> +
> +static DEFINE_PER_CPU(unsigned long, cmci_time_stamp);
> +static DEFINE_PER_CPU(unsigned int, cmci_storm_cnt);
> +static DEFINE_PER_CPU(unsigned int, cmci_storm_state);
> +
> +enum {
> + CMCI_STORM_NONE,
> + CMCI_STORM_ACTIVE,
> + CMCI_STORM_SUBSIDED,
> +};
> +
> +static atomic_t cmci_storm_on_cpus;
>
> static int cmci_supported(int *banks)
> {
> @@ -53,6 +70,73 @@ static int cmci_supported(int *banks)
> return !!(cap & MCG_CMCI_P);
> }
>
> +void mce_intel_cmci_poll(void)
> +{
> + if (__this_cpu_read(cmci_storm_state) == CMCI_STORM_NONE)
> + return;
> + machine_check_poll(MCP_TIMESTAMP, &__get_cpu_var(mce_banks_owned));
> +}
> +
> +unsigned long mce_intel_adjust_timer(unsigned long interval)
> +{
> + if (interval < CMCI_POLL_INTERVAL)
> + return interval;
> +
> + switch (__this_cpu_read(cmci_storm_state)) {
> + case CMCI_STORM_ACTIVE:
> + /*
> + * We switch back to interrupt mode once the poll timer has
> + * silenced itself. That means no events recorded and the
> + * timer interval is back to our poll interval.
> + */
> + __this_cpu_write(cmci_storm_state, CMCI_STORM_SUBSIDED);
> + atomic_dec(&cmci_storm_on_cpus);
> +
> + case CMCI_STORM_SUBSIDED:
> + /*
> + * We wait for all cpus to go back to SUBSIDED
> + * state. When that happens we switch back to
> + * interrupt mode.
> + */
> + if (!atomic_read(&cmci_storm_on_cpus)) {
> + __this_cpu_write(cmci_storm_state, CMCI_STORM_NONE);
> + cmci_reenable();
> + cmci_recheck();
> + }
> + return CMCI_POLL_INTERVAL;
> + default:
> + /*
> + * We have shiny wheather, let the poll do whatever it
> + * thinks.
> + */
> + return interval;
> + }
> +}
> +
> +static bool cmci_storm_detect(void)
> +{
> + unsigned int cnt = __this_cpu_read(cmci_storm_cnt);
> + unsigned long ts = __this_cpu_read(cmci_time_stamp);
> + unsigned long now = jiffies;
> +
> + if (time_before_eq(now, ts + CMCI_STORM_INTERVAL)) {
> + cnt++;
> + } else {
> + cnt = 1;
> + __this_cpu_write(cmci_time_stamp, now);
> + }
> + __this_cpu_write(cmci_storm_cnt, cnt);
> +
> + if (cnt <= CMCI_STORM_TRESHOLD)
and here too.
> + return false;
> +
> + cmci_clear();
> + __this_cpu_write(cmci_storm_state, CMCI_STORM_ACTIVE);
> + atomic_inc(&cmci_storm_on_cpus);
> + mce_timer_kick(CMCI_POLL_INTERVAL);
> + return true;
> +}
> +
> /*
> * The interrupt handler. This is called on every event.
> * Just call the poller directly to log any events.
> @@ -61,6 +145,8 @@ static int cmci_supported(int *banks)
> */
> static void intel_threshold_interrupt(void)
> {
> + if (cmci_storm_detect())
> + return;
> machine_check_poll(MCP_TIMESTAMP, &__get_cpu_var(mce_banks_owned));
> mce_notify_irq();
> }
>
>
>
--
Regards/Gruss,
Boris.
Advanced Micro Devices GmbH
Einsteinring 24, 85609 Dornach
GM: Alberto Bozzo
Reg: Dornach, Landkreis Muenchen
HRB Nr. 43632 WEEE Registernr: 129 19551
next prev parent reply other threads:[~2012-06-07 18:14 UTC|newest]
Thread overview: 33+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-06-06 21:53 [patch 0/5] x86: mce: Bugfixes, cleanups and a new CMCI poll version Thomas Gleixner
2012-06-06 21:53 ` [patch 1/5] x86: mce: Disable preemption when calling raise_local() Thomas Gleixner
2012-06-06 21:53 ` [patch 2/5] x86: mce: Serialize mce injection Thomas Gleixner
2012-06-06 21:53 ` [patch 3/5] x86: mce: Split timer init Thomas Gleixner
2012-06-07 15:18 ` Borislav Petkov
2012-06-20 3:35 ` Hidetoshi Seto
2012-06-06 21:53 ` [patch 5/5] x86: mce: Add cmci poll mode Thomas Gleixner
2012-06-07 18:14 ` Borislav Petkov [this message]
2012-06-06 21:53 ` [patch 4/5] x86: mce: Remove the frozen cases in the hotplug code Thomas Gleixner
2012-06-07 17:49 ` Borislav Petkov
2012-06-07 10:08 ` [patch 0/5] x86: mce: Bugfixes, cleanups and a new CMCI poll version Chen Gong
2012-06-07 13:35 ` Borislav Petkov
2012-06-07 16:22 ` Luck, Tony
2012-06-08 7:49 ` Thomas Gleixner
2012-06-11 5:46 ` Chen Gong
2012-06-11 6:09 ` Chen Gong
2012-06-14 13:49 ` [PATCH] tmp patch to fix hotplug issue in CMCI storm Chen Gong
2012-06-14 14:07 ` Thomas Gleixner
2012-06-15 6:51 ` Chen Gong
2012-06-15 9:55 ` Thomas Gleixner
2012-06-18 6:42 ` Chen Gong
2012-06-18 6:45 ` [PATCH V2] " Chen Gong
2012-06-18 8:00 ` Thomas Gleixner
2012-06-18 10:13 ` Chen Gong
2012-06-18 12:23 ` Thomas Gleixner
2012-06-19 6:05 ` Chen Gong
2012-06-19 6:09 ` [PATCH V3] " Chen Gong
2012-07-04 8:12 ` Chen Gong
2012-07-16 3:16 ` Chen Gong
2012-07-16 8:22 ` Thomas Gleixner
2012-07-17 21:47 ` Chen Gong
2012-07-18 19:59 [V2] x86: mce: Bugfixes, cleanups and a new CMCI poll version Chen Gong
2012-07-18 19:59 ` [PATCH 5/5] x86: mce: Add cmci poll mode Chen Gong
2012-07-19 17:59 [RESEND PATCH 0/5 V2] x86: mce: Bugfixes, cleanups and a new CMCI poll version Chen Gong
2012-07-19 17:59 ` [PATCH 5/5] x86: mce: Add cmci poll mode Chen Gong
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=20120607181414.GH16941@aftab.osrc.amd.com \
--to=borislav.petkov@amd.com \
--cc=gong.chen@linux.intel.com \
--cc=linux-kernel@vger.kernel.org \
--cc=peterz@infradead.org \
--cc=tglx@linutronix.de \
--cc=tony.luck@intel.com \
--cc=x86@kernel.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
Powered by JetHome