mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Andi Kleen <ak@suse.de>
To: adurbin@google.com, patches@x86-64.org, linux-kernel@vger.kernel.org
Subject: [PATCH] [9/12] x86_64: Avoid NMI Watchdog and/or long wait in setup_APIC_timer
Date: Thu,  9 Aug 2007 14:41:37 +0200 (CEST)	[thread overview]
Message-ID: <20070809124137.0FE4D14F3B@wotan.suse.de> (raw)
In-Reply-To: <20070809241.425881000@suse.de>


From: Aaron Durbin <adurbin@google.com>
In setup_APIC_timer with the HPET in use, a condition can arise while
waiting for the next irq slice to expire on the HPET which will either
cause an NMI watchdog to fire or a 3 minute busy loop if the NMI
watchdog is disabled.

The HPET comparator and the counter keep incrementing during its normal
operation. When a comparison event fires the comparator will increment
by the designated period. If the HPET trigger occurs right after
the 'int trigger = hpet_readl(HPET_T0_CMP);' line, we will will spin
for up to 3 minutes (with a clock of 25MHz) waiting for the HPET
counter to wrap around. However, when the NMI watchdog is enabled the
NMI watchdog will detect a lockup and reboot the machine. This
scenario can be exasperated by the presence of an SMI which will
increase the window of opportunity for the condition to occur.

The fix is to wait for the compartor to change which signals the
end of the tick slice.

Signed-off-by: Andi Kleen <ak@suse.de>

---
 arch/x86_64/kernel/apic.c |    8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

Index: linux/arch/x86_64/kernel/apic.c
===================================================================
--- linux.orig/arch/x86_64/kernel/apic.c
+++ linux/arch/x86_64/kernel/apic.c
@@ -791,10 +791,12 @@ static void setup_APIC_timer(unsigned in
 
 	/* wait for irq slice */
 	if (hpet_address && hpet_use_timer) {
+		/*
+		 * Wait for the comparator value to change which signals that
+		 * the tick slice has expired.
+		 */
 		int trigger = hpet_readl(HPET_T0_CMP);
-		while (hpet_readl(HPET_COUNTER) >= trigger)
-			/* do nothing */ ;
-		while (hpet_readl(HPET_COUNTER) <  trigger)
+		while (trigger == hpet_readl(HPET_T0_CMP))
 			/* do nothing */ ;
 	} else {
 		int c1, c2;

  parent reply	other threads:[~2007-08-09 12:46 UTC|newest]

Thread overview: 46+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-08-09 12:41 [PATCH] [0/12] x86 Late merge bug fixes for 2.6.23 Andi Kleen
2007-08-09 12:41 ` [PATCH] [1/12] x86: Work around mmio config space quirk on AMD Fam10h Andi Kleen
2007-08-10 10:19   ` [patches] " Joachim Deguara
2007-08-12  9:25   ` Linus Torvalds
2007-08-12 10:03     ` Andi Kleen
2007-08-12 19:15       ` Linus Torvalds
2007-08-13  3:06         ` Dave Jones
2007-08-13  3:26           ` Linus Torvalds
2007-08-13  3:38             ` dean gaudet
2007-08-09 12:41 ` [PATCH] [2/12] x86_64: Calgary - Fix mis-handled PCI topology Andi Kleen
2007-08-09 12:41 ` [PATCH] [3/12] x86_64: Don't mark __exitcall as __cold Andi Kleen
2007-08-09 12:41 ` [PATCH] [4/12] x86_64: Disable CLFLUSH support again Andi Kleen
2007-08-09 12:54   ` [patches] " Jan Beulich
2007-08-09 13:11   ` Muli Ben-Yehuda
2007-08-09 13:12     ` Andi Kleen
2007-08-09 13:29     ` [patches] " Jan Beulich
2007-08-09 16:24       ` Muli Ben-Yehuda
2007-08-09 12:41 ` [PATCH] [5/12] x86_64: Make patching more robust, fix paravirt issue Andi Kleen
2007-08-18  0:05   ` Jeremy Fitzhardinge
2007-08-18  0:11     ` Chris Wright
2007-08-18  9:49     ` Andi Kleen
2007-08-18 20:32       ` Chris Wright
2007-08-18 20:55         ` Linus Torvalds
2007-08-18 21:31           ` Chris Wright
2007-08-18 22:28             ` [PATCH] x86: skip paravirt patching when appropriate Chris Wright
2007-08-19  3:08               ` Zachary Amsden
2007-08-19  5:55               ` Zachary Amsden
2007-08-18 21:58       ` [PATCH] [5/12] x86_64: Make patching more robust, fix paravirt issue Jeremy Fitzhardinge
2007-08-20  1:50     ` Rusty Russell
2007-08-20  2:07       ` Chris Wright
2007-08-20  6:15         ` Jeremy Fitzhardinge
2007-08-21  7:30           ` Glauber de Oliveira Costa
2007-08-21 11:23             ` Andi Kleen
2007-08-21 12:26               ` Glauber de Oliveira Costa
2007-08-09 12:41 ` [PATCH] [6/12] x86_64: Early segment setup for VT Andi Kleen
2007-10-11 21:30   ` H. Peter Anvin
2007-08-09 12:41 ` [PATCH] [7/12] x86_64: really stop MCEs during code patching Andi Kleen
2007-08-09 12:41 ` [PATCH] [8/12] x86_64: Use global flag to disable broken local apic timer on AMD CPUs Andi Kleen
2007-08-09 16:46   ` Cal Peake
2007-08-09 16:50     ` Andi Kleen
2007-08-09 12:41 ` Andi Kleen [this message]
2007-08-09 12:41 ` [PATCH] [10/12] x86_64: Add warning in Documentation that zero-page is not a stable ABI Andi Kleen
2007-08-09 12:41 ` [PATCH] [11/12] x86_64: vdso.lds in arch/x86_64/vdso/.gitignore Andi Kleen
2007-08-09 12:41 ` [PATCH] [12/12] x86_64: Fix start_kernel warning Andi Kleen
2007-08-09 13:06 ` [PATCH] [0/12] x86 Late merge bug fixes for 2.6.23 Muli Ben-Yehuda
2007-08-09 13:12   ` Andi Kleen

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=20070809124137.0FE4D14F3B@wotan.suse.de \
    --to=ak@suse.de \
    --cc=adurbin@google.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=patches@x86-64.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

all inboxes | Powered by JetHome®