All of lore.kernel.org
 help / color / mirror / Atom feed
From: Tony Luck <tony.luck@intel.com>
To: Borislav Petkov <bp@alien8.de>
Cc: Yazen Ghannam <yazen.ghannam@amd.com>,
	Smita.KoralahalliChannabasappa@amd.com,
	dave.hansen@linux.intel.com, x86@kernel.org,
	linux-edac@vger.kernel.org, linux-kernel@vger.kernel.org,
	patches@lists.linux.dev, Tony Luck <tony.luck@intel.com>
Subject: [PATCH v8 2/3] x86/mce: Add per-bank CMCI storm mitigation
Date: Fri, 29 Sep 2023 11:16:25 -0700	[thread overview]
Message-ID: <20230929181626.210782-3-tony.luck@intel.com> (raw)
In-Reply-To: <20230929181626.210782-1-tony.luck@intel.com>

This is the core functionality to track CMCI storms at the
machine check bank granularity. Subsequent patches will add
the vendor specific hooks to supply input to the storm
detection and take actions on the start/end of a storm.

Maintain a bitmap history for each bank showing whether the bank
logged an corrected error or not each time it is polled.

In normal operation the interval between polls of this banks
determines how far to shift the history. The 64 bit width corresponds
to about one second.

When a storm is observed a CPU vendor specific action is taken to reduce
or stop CMCI from the bank that is the source of the storm.  The bank
is added to the bitmap of banks for this CPU to poll. The polling rate
is increased to once per second.  During a storm each bit in the history
indicates the status of the bank each time it is polled. Thus the history
covers just over a minute.

Declare a storm for that bank if the number of corrected interrupts
seen in that history is above some threshold (defined as 5 in this
series, could be tuned later if there is data to suggest a better
value).

A storm on a bank ends if enough consecutive polls of the bank show
no corrected errors (defined as 30, may also change). That calls the
CPU vendor specific function to revert to normal operational mode,
and changes the polling rate back to the default.

Updated with review comments from Yazen.
Link: https://lore.kernel.org/r/c76723df-f2f1-4888-9e05-61917145503c@amd.com

Signed-off-by: Tony Luck <tony.luck@intel.com>
---

Note that the storm begin/end warnings are now printed using
printk_deferred(KERN_NOTICE ...) instead of pr_notice. This
is my solution for a locking issue when in poll mode because
of the "cmci_poll_lock" introduced by:

commit c3629dd7e67d ("x86/mce: Prevent duplicate error records")

Is this the right way to handle this?


 arch/x86/kernel/cpu/mce/internal.h  | 39 +++++++++++++-
 arch/x86/kernel/cpu/mce/core.c      | 25 ++++++---
 arch/x86/kernel/cpu/mce/threshold.c | 83 +++++++++++++++++++++++++++++
 3 files changed, 138 insertions(+), 9 deletions(-)

diff --git a/arch/x86/kernel/cpu/mce/internal.h b/arch/x86/kernel/cpu/mce/internal.h
index 616732ec16f7..5e75f5f81464 100644
--- a/arch/x86/kernel/cpu/mce/internal.h
+++ b/arch/x86/kernel/cpu/mce/internal.h
@@ -54,7 +54,44 @@ static inline void intel_clear_lmce(void) { }
 static inline bool intel_filter_mce(struct mce *m) { return false; }
 #endif
 
-void mce_timer_kick(unsigned long interval);
+void mce_timer_kick(bool storm);
+void cmci_storm_begin(unsigned int bank);
+void cmci_storm_end(unsigned int bank);
+void mce_track_storm(struct mce *mce);
+
+/*
+ * history:	bitmask tracking whether errors were seen or not seen in
+ *		the most recent polls of a bank.
+ * timestamp:	last time (in jiffies) that the bank was polled
+ * storm:	Is this bank in storm mode?
+ */
+struct storm_bank {
+	u64 history;
+	u64 timestamp;
+	bool storm;
+};
+
+/*
+ * banks:		per-cpu, per-bank details
+ * stormy_bank_count:	count of MC banks in storm state
+ * poll_mode:		CPU is in poll mode
+ */
+struct mca_storm_desc {
+	struct storm_bank	banks[MAX_NR_BANKS];
+	u8			stormy_bank_count;
+	bool			poll_mode;
+};
+DECLARE_PER_CPU(struct mca_storm_desc, storm_desc);
+
+/* How many errors within the history buffer mark the start of a storm. */
+#define STORM_BEGIN_THRESHOLD	5
+
+/*
+ * How many polls of machine check bank without an error before declaring
+ * the storm is over. Since it is tracked by the bitmaks in the history
+ * field of struct storm_bank the mask is 30 bits [0 ... 29].
+ */
+#define STORM_END_POLL_THRESHOLD	29
 
 #ifdef CONFIG_ACPI_APEI
 int apei_write_mce(struct mce *m);
diff --git a/arch/x86/kernel/cpu/mce/core.c b/arch/x86/kernel/cpu/mce/core.c
index f6e87443b37a..fccbf4a7e783 100644
--- a/arch/x86/kernel/cpu/mce/core.c
+++ b/arch/x86/kernel/cpu/mce/core.c
@@ -680,6 +680,8 @@ bool machine_check_poll(enum mcp_flags flags, mce_banks_t *b)
 		barrier();
 		m.status = mce_rdmsrl(mca_msr_reg(i, MCA_STATUS));
 
+		mce_track_storm(&m);
+
 		/* If this entry is not valid, ignore it */
 		if (!(m.status & MCI_STATUS_VAL))
 			continue;
@@ -1652,22 +1654,29 @@ static void mce_timer_fn(struct timer_list *t)
 	else
 		iv = min(iv * 2, round_jiffies_relative(check_interval * HZ));
 
-	__this_cpu_write(mce_next_interval, iv);
-	__start_timer(t, iv);
+	if (__this_cpu_read(storm_desc.poll_mode)) {
+		__start_timer(t, HZ);
+	} else {
+		__this_cpu_write(mce_next_interval, iv);
+		__start_timer(t, iv);
+	}
 }
 
 /*
- * Ensure that the timer is firing in @interval from now.
+ * When a storm starts on any bank on this CPU, switch to polling
+ * once per second. When the storm ends, revert to the default
+ * polling interval.
  */
-void mce_timer_kick(unsigned long interval)
+void mce_timer_kick(bool storm)
 {
 	struct timer_list *t = this_cpu_ptr(&mce_timer);
-	unsigned long iv = __this_cpu_read(mce_next_interval);
 
-	__start_timer(t, interval);
+	__this_cpu_write(storm_desc.poll_mode, storm);
 
-	if (interval < iv)
-		__this_cpu_write(mce_next_interval, interval);
+	if (storm)
+		__start_timer(t, HZ);
+	else
+		__this_cpu_write(mce_next_interval, check_interval * HZ);
 }
 
 /* Must not be called in IRQ context where del_timer_sync() can deadlock */
diff --git a/arch/x86/kernel/cpu/mce/threshold.c b/arch/x86/kernel/cpu/mce/threshold.c
index ef4e7bb5fd88..c9e32d92cf8f 100644
--- a/arch/x86/kernel/cpu/mce/threshold.c
+++ b/arch/x86/kernel/cpu/mce/threshold.c
@@ -29,3 +29,86 @@ DEFINE_IDTENTRY_SYSVEC(sysvec_threshold)
 	trace_threshold_apic_exit(THRESHOLD_APIC_VECTOR);
 	apic_eoi();
 }
+
+DEFINE_PER_CPU(struct mca_storm_desc, storm_desc);
+
+static void mce_handle_storm(unsigned int bank, bool on)
+{
+	switch (boot_cpu_data.x86_vendor) {
+	}
+}
+
+void cmci_storm_begin(unsigned int bank)
+{
+	struct mca_storm_desc *storm = this_cpu_ptr(&storm_desc);
+
+	__set_bit(bank, this_cpu_ptr(mce_poll_banks));
+	storm->banks[bank].storm = true;
+
+	/*
+	 * If this is the first bank on this CPU to enter storm mode
+	 * start polling.
+	 */
+	if (++storm->stormy_bank_count == 1)
+		mce_timer_kick(true);
+}
+
+void cmci_storm_end(unsigned int bank)
+{
+	struct mca_storm_desc *storm = this_cpu_ptr(&storm_desc);
+
+	__clear_bit(bank, this_cpu_ptr(mce_poll_banks));
+	storm->banks[bank].history = 0;
+	storm->banks[bank].storm = false;
+
+	/* If no banks left in storm mode, stop polling. */
+	if (!this_cpu_dec_return(storm_desc.stormy_bank_count))
+		mce_timer_kick(false);
+}
+
+void mce_track_storm(struct mce *mce)
+{
+	struct mca_storm_desc *storm = this_cpu_ptr(&storm_desc);
+	unsigned long now = jiffies, delta;
+	unsigned int shift = 1;
+	u64 history = 0;
+
+	/*
+	 * When a bank is in storm mode it is polled once per second and
+	 * the history mask will record about the last minute of poll results.
+	 * If it is not in storm mode, then the bank is only checked when
+	 * there is a CMCI interrupt. Check how long it has been since
+	 * this bank was last checked, and adjust the amount of "shift"
+	 * to apply to history.
+	 */
+	if (!storm->banks[mce->bank].storm) {
+		delta = now - storm->banks[mce->bank].timestamp;
+		shift = (delta + HZ) / HZ;
+	}
+
+	/* If it has been a long time since the last poll, clear history. */
+	if (shift < 64)
+		history = storm->banks[mce->bank].history << shift;
+
+	storm->banks[mce->bank].timestamp = now;
+
+	/* History keeps track of corrected errors. VAL=1 && UC=0 */
+	if ((mce->status & MCI_STATUS_VAL) && mce_is_correctable(mce))
+		history |= 1;
+
+	storm->banks[mce->bank].history = history;
+
+	if (storm->banks[mce->bank].storm) {
+		if (history & GENMASK_ULL(STORM_END_POLL_THRESHOLD, 0))
+			return;
+		printk_deferred(KERN_NOTICE "CPU%d BANK%d CMCI storm subsided\n", smp_processor_id(), mce->bank);
+		mce_handle_storm(mce->bank, false);
+		cmci_storm_end(mce->bank);
+	} else {
+		if (hweight64(history) < STORM_BEGIN_THRESHOLD)
+			return;
+		printk_deferred(KERN_NOTICE "CPU%d BANK%d CMCI storm detected\n", smp_processor_id(), mce->bank);
+		mce_handle_storm(mce->bank, true);
+		cmci_storm_begin(mce->bank);
+	}
+}
-- 
2.41.0


  parent reply	other threads:[~2023-09-29 20:13 UTC|newest]

Thread overview: 105+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-04-06  6:35 [RFC PATCH 0/5] Handle corrected machine check interrupt storms Smita Koralahalli
2022-04-06  6:35 ` [PATCH 1/5] x86/mce: Remove old CMCI storm mitigation code Smita Koralahalli
2022-04-06  6:35 ` [PATCH 2/5] x86/mce: Add per-bank CMCI storm mitigation Smita Koralahalli
2022-04-06  6:35 ` [RFC PATCH 3/5] x86/mce: Introduce a function pointer mce_handle_storm Smita Koralahalli
2022-04-06 22:38   ` Luck, Tony
2022-04-06  6:35 ` [RFC PATCH 4/5] x86/mce: Move storm handling to core Smita Koralahalli
2022-06-21  5:08   ` Luck, Tony
2022-06-27 17:36     ` [PATCH v2 0/5] Handle corrected machine check interrupt storms Tony Luck
2022-06-27 17:36       ` [PATCH v2 1/5] x86/mce: Remove old CMCI storm mitigation code Tony Luck
2022-06-27 17:36       ` [PATCH v2 2/5] x86/mce: Add per-bank CMCI storm mitigation Tony Luck
2022-06-27 17:36       ` [PATCH v2 3/5] x86/mce: Introduce mce_handle_storm() to deal with begin/end of storms Tony Luck
2022-06-27 17:36       ` [PATCH v2 4/5] x86/mce: Move storm handling to core Tony Luck
2022-06-27 17:36       ` [PATCH v2 5/5] x86/mce: Handle AMD threshold interrupt storms Tony Luck
2023-03-17 14:50       ` [PATCH v2 0/5] Handle corrected machine check " Yazen Ghannam
2023-03-17 17:20         ` [PATCH v3 " Tony Luck
2023-03-17 17:20           ` [PATCH v3 1/5] x86/mce: Remove old CMCI storm mitigation code Tony Luck
2023-03-17 17:20           ` [PATCH v3 2/5] x86/mce: Add per-bank CMCI storm mitigation Tony Luck
2023-03-17 17:20           ` [PATCH v3 3/5] x86/mce: Introduce mce_handle_storm() to deal with begin/end of storms Tony Luck
2023-03-23 15:22             ` Yazen Ghannam
2023-03-23 18:00               ` Tony Luck
2023-03-17 17:20           ` [PATCH v3 4/5] x86/mce: Move storm handling to core Tony Luck
2023-03-23 15:27             ` Yazen Ghannam
2023-03-23 18:10               ` Luck, Tony
2023-03-23 20:26                 ` Luck, Tony
2023-03-24 20:44                   ` Yazen Ghannam
2023-03-29 15:26                   ` Yazen Ghannam
2023-04-03 19:03                     ` Luck, Tony
2023-04-03 21:07                     ` [PATCH v4 0/5] Handle corrected machine check interrupt storms Tony Luck
2023-04-03 21:07                       ` [PATCH v4 1/5] x86/mce: Remove old CMCI storm mitigation code Tony Luck
2023-04-03 21:07                       ` [PATCH v4 2/5] x86/mce: Add per-bank CMCI storm mitigation Tony Luck
2023-04-11 12:32                         ` Borislav Petkov
2023-04-11 14:06                           ` Yazen Ghannam
2023-04-11 16:06                             ` Luck, Tony
2023-04-11 17:17                               ` Borislav Petkov
2023-04-03 21:07                       ` [PATCH v4 3/5] x86/mce: Introduce mce_handle_storm() to deal with begin/end of storms Tony Luck
2023-04-03 21:07                       ` [PATCH v4 4/5] x86/mce: Move storm handling to core Tony Luck
2023-04-03 21:07                       ` [PATCH v4 5/5] x86/mce: Handle AMD threshold interrupt storms Tony Luck
2023-04-11 17:38                       ` [PATCH v5 0/5] Handle corrected machine check " Tony Luck
2023-04-11 17:38                         ` [PATCH v5 1/5] x86/mce: Remove old CMCI storm mitigation code Tony Luck
2023-04-11 17:38                         ` [PATCH v5 2/5] x86/mce: Add per-bank CMCI storm mitigation Tony Luck
2023-06-13 17:45                           ` Borislav Petkov
2023-06-16 18:15                             ` Tony Luck
2023-04-11 17:38                         ` [PATCH v5 3/5] x86/mce: Introduce mce_handle_storm() to deal with begin/end of storms Tony Luck
2023-04-11 17:38                         ` [PATCH v5 4/5] x86/mce: Move storm handling to core Tony Luck
2023-04-11 17:38                         ` [PATCH v5 5/5] x86/mce: Handle AMD threshold interrupt storms Tony Luck
2023-06-16 18:27                         ` [PATCH v6 0/4] Handle corrected machine check " Tony Luck
2023-06-16 18:27                           ` [PATCH v6 1/4] x86/mce: Remove old CMCI storm mitigation code Tony Luck
2023-06-16 18:27                           ` [PATCH v6 2/4] x86/mce: Add per-bank CMCI storm mitigation Tony Luck
2023-06-23 12:09                             ` Borislav Petkov
2023-06-23 15:40                               ` Luck, Tony
2023-07-17  8:58                                 ` Borislav Petkov
2023-06-16 18:27                           ` [PATCH v6 3/4] x86/mce: Handle AMD threshold interrupt storms Tony Luck
2023-06-23 14:45                             ` Borislav Petkov
2023-06-23 15:54                               ` Yazen Ghannam
2023-06-16 18:27                           ` [PATCH v6 4/4] x86/mce: Handle Intel " Tony Luck
2023-07-18 21:08                           ` [PATCH v7 0/3] Handle corrected machine check " Tony Luck
2023-07-18 21:08                             ` [PATCH v7 1/3] x86/mce: Remove old CMCI storm mitigation code Tony Luck
2023-07-18 21:08                             ` [PATCH v7 2/3] x86/mce: Add per-bank CMCI storm mitigation Tony Luck
2023-09-19 17:44                               ` Yazen Ghannam
2023-09-20 15:56                               ` Yazen Ghannam
2023-09-20 16:09                                 ` Luck, Tony
2023-07-18 21:08                             ` [PATCH v7 3/3] x86/mce: Handle Intel threshold interrupt storms Tony Luck
2023-09-19 17:59                               ` Yazen Ghannam
2023-09-29 18:16                             ` [PATCH v8 0/3] Handle corrected machine check " Tony Luck
2023-09-29 18:16                               ` [PATCH v8 1/3] x86/mce: Remove old CMCI storm mitigation code Tony Luck
2023-09-29 18:16                               ` Tony Luck [this message]
2023-09-30  6:02                                 ` [PATCH v8 2/3] x86/mce: Add per-bank CMCI storm mitigation kernel test robot
2023-09-29 18:16                               ` [PATCH v8 3/3] x86/mce: Handle Intel threshold interrupt storms Tony Luck
2023-10-02 17:57                               ` [PATCH v8 0/3] Handle corrected machine check " Luck, Tony
2023-10-04 18:36                               ` [PATCH v9 " Tony Luck
2023-10-04 18:36                                 ` [PATCH v9 1/3] x86/mce: Remove old CMCI storm mitigation code Tony Luck
2023-10-04 18:36                                 ` [PATCH v9 2/3] x86/mce: Add per-bank CMCI storm mitigation Tony Luck
2023-10-11  9:11                                   ` kernel test robot
2023-10-11 15:16                                     ` Luck, Tony
2023-10-11 15:42                                       ` Feng Tang
2023-10-11 17:23                                         ` Luck, Tony
2023-10-12  5:36                                           ` Feng Tang
2023-10-12  5:56                                             ` Feng Tang
2023-10-12  2:35                                         ` Philip Li
2023-10-19 15:12                                   ` Borislav Petkov
2023-10-23 18:14                                     ` Tony Luck
2023-11-14 19:23                                       ` Borislav Petkov
2023-11-14 22:04                                         ` Tony Luck
2023-11-21 11:54                                           ` Borislav Petkov
2023-11-27 19:50                                             ` Tony Luck
2023-11-27 20:14                                               ` Tony Luck
2023-11-28  0:42                                                 ` Tony Luck
2023-11-28 15:32                                                   ` Yazen Ghannam
2023-12-14 16:58                                                   ` Borislav Petkov
2023-12-14 18:03                                                     ` Luck, Tony
2023-10-04 18:36                                 ` [PATCH v9 3/3] x86/mce: Handle Intel threshold interrupt storms Tony Luck
2023-11-15 19:54                                 ` [PATCH v10 0/3] Handle corrected machine check " Tony Luck
2023-11-15 19:54                                   ` [PATCH v10 1/3] x86/mce: Remove old CMCI storm mitigation code Tony Luck
2023-12-15 16:10                                     ` [tip: ras/core] " tip-bot2 for Tony Luck
2023-11-15 19:54                                   ` [PATCH v10 2/3] x86/mce: Add per-bank CMCI storm mitigation Tony Luck
2023-12-15 16:10                                     ` [tip: ras/core] " tip-bot2 for Tony Luck
2023-11-15 19:54                                   ` [PATCH v10 3/3] x86/mce: Handle Intel threshold interrupt storms Tony Luck
2023-12-15 16:10                                     ` [tip: ras/core] " tip-bot2 for Tony Luck
2023-12-15 17:21                                       ` Luck, Tony
2023-12-15 17:37                                         ` Borislav Petkov
2023-03-17 17:20           ` [PATCH v3 5/5] x86/mce: Handle AMD " Tony Luck
2022-04-06  6:35 ` [RFC PATCH " Smita Koralahalli
2022-04-06 22:44   ` Luck, Tony
2022-04-08  7:48     ` Koralahalli Channabasappa, Smita
2022-04-08 19:29       ` Luck, Tony

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=20230929181626.210782-3-tony.luck@intel.com \
    --to=tony.luck@intel.com \
    --cc=Smita.KoralahalliChannabasappa@amd.com \
    --cc=bp@alien8.de \
    --cc=dave.hansen@linux.intel.com \
    --cc=linux-edac@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=patches@lists.linux.dev \
    --cc=x86@kernel.org \
    --cc=yazen.ghannam@amd.com \
    /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.