From: Stas Sergeev <stsp@aknet.ru>
To: "Maciej W. Rozycki" <macro@linux-mips.org>
Cc: Linux kernel <linux-kernel@vger.kernel.org>,
Andrew Morton <akpm@osdl.org>
Subject: Re: 2.6.10-rc1-mm5
Date: Thu, 18 Nov 2004 20:27:11 +0300 [thread overview]
Message-ID: <419CDB6F.2090406@aknet.ru> (raw)
In-Reply-To: <Pine.LNX.4.58L.0411181557430.30376@blysk.ds.pg.gda.pl>
[-- Attachment #1: Type: text/plain, Size: 478 bytes --]
Hi,
Maciej W. Rozycki wrote:
> something CPU-intensive, like:
> and observe the NMI counter ticking every second.
Ah, thanks for explanations!
Indeed it works as you say.
Very usefull info, perhaps worth
being added to documentation?
So if you ACK the attached patch,
I ask Andrew to please apply it.
Andrew, the attached patch adds
the information about the NMI
watchdog frequency caveats to the
Documentation/nmi_watchdog.txt
Unless there are any objections,
please apply.
[-- Attachment #2: nmi_watchdog.txt.diff --]
[-- Type: text/x-patch, Size: 1406 bytes --]
--- linux/Documentation/nmi_watchdog.txt.old 2004-10-21 21:21:56.000000000 +0400
+++ linux/Documentation/nmi_watchdog.txt 2004-11-18 20:16:37.972370552 +0300
@@ -54,6 +54,20 @@
cannot even accept NMI interrupts, or the crash has made the kernel
unable to print messages.
+Be aware that when using local APIC, the frequency of NMI interrupts
+it generates, depends on the system load. The local APIC NMI watchdog,
+lacking a better source, uses the "cycles unhalted" event. As you may
+guess it doesn't tick when the CPU is in the halted state (which happens
+when the system is idle), but if your system locks up on anything but the
+"hlt" processor instruction, the watchdog will trigger very soon as the
+"cycles unhalted" event will happen every clock tick. If it locks up on
+"hlt", then you are out of luck -- the event will not happen at all and the
+watchdog won't trigger. This is a shortcoming of the local APIC watchdog
+-- unfortunately there is no "clock ticks" event that would work all the
+time. The I/O APIC watchdog is driven externally and has no such shortcoming.
+But its NMI frequency is much higher, resulting in a more significant hit
+to the overall system performance.
+
NOTE: starting with 2.4.2-ac18 the NMI-oopser is disabled by default,
you have to enable it with a boot time parameter. Prior to 2.4.2-ac18
the NMI-oopser is enabled unconditionally on x86 SMP boxes.
next prev parent reply other threads:[~2004-11-18 17:38 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
2004-11-13 21:02 2.6.10-rc1-mm5 Stas Sergeev
2004-11-13 21:38 ` 2.6.10-rc1-mm5 Andrew Morton
2004-11-15 18:21 ` 2.6.10-rc1-mm5 Stas Sergeev
2004-11-13 22:13 ` 2.6.10-rc1-mm5 Zwane Mwaikambo
2004-11-13 22:47 ` 2.6.10-rc1-mm5 Stas Sergeev
2004-11-15 1:16 ` 2.6.10-rc1-mm5 Zwane Mwaikambo
2004-11-15 1:49 ` 2.6.10-rc1-mm5 Maciej W. Rozycki
2004-11-15 2:14 ` 2.6.10-rc1-mm5 Zwane Mwaikambo
2004-11-14 9:24 ` 2.6.10-rc1-mm5 Dave Airlie
2004-11-14 13:48 ` 2.6.10-rc1-mm5 Stas Sergeev
2004-11-15 1:19 ` 2.6.10-rc1-mm5 Maciej W. Rozycki
2004-11-15 18:05 ` 2.6.10-rc1-mm5 Stas Sergeev
2004-11-15 23:33 ` 2.6.10-rc1-mm5 Maciej W. Rozycki
2004-11-15 23:53 ` 2.6.10-rc1-mm5 Andrew Morton
2004-11-16 1:47 ` 2.6.10-rc1-mm5 Maciej W. Rozycki
2004-11-16 17:29 ` 2.6.10-rc1-mm5 Stas Sergeev
2004-11-16 23:35 ` 2.6.10-rc1-mm5 Maciej W. Rozycki
2004-11-18 14:53 ` 2.6.10-rc1-mm5 Stas Sergeev
2004-11-18 16:17 ` 2.6.10-rc1-mm5 Maciej W. Rozycki
2004-11-18 17:27 ` Stas Sergeev [this message]
2004-11-18 18:37 ` 2.6.10-rc1-mm5 Maciej W. Rozycki
-- strict thread matches above, loose matches on Subject: below --
2004-11-14 12:11 2.6.10-rc1-mm5 Mikael Pettersson
2004-11-14 13:27 ` 2.6.10-rc1-mm5 Stas Sergeev
2004-11-13 0:30 2.6.10-rc1-mm5 Keshavamurthy, Anil S
2004-11-11 9:23 2.6.10-rc1-mm5 Andrew Morton
2004-11-11 10:09 ` 2.6.10-rc1-mm5 Magnus Damm
2004-11-11 10:11 ` 2.6.10-rc1-mm5 Andrew Morton
2004-11-11 10:24 ` 2.6.10-rc1-mm5 Magnus Damm
2004-11-11 10:45 ` 2.6.10-rc1-mm5 Brice Goglin
2004-11-11 11:00 ` 2.6.10-rc1-mm5 Stefano Rivoir
2004-11-11 11:13 ` 2.6.10-rc1-mm5 Andrew Morton
2004-11-11 11:04 ` 2.6.10-rc1-mm5 Markus Trippelsdorf
2004-11-11 11:08 ` 2.6.10-rc1-mm5 Andrew Morton
2004-11-11 21:55 ` 2.6.10-rc1-mm5 Martin J. Bligh
2004-11-11 22:14 ` 2.6.10-rc1-mm5 Andrew Morton
2004-11-12 9:27 ` 2.6.10-rc1-mm5 Jens Axboe
2004-11-11 22:41 ` 2.6.10-rc1-mm5 Andrew Morton
2004-11-12 19:00 ` 2.6.10-rc1-mm5 Alexander Nyberg
2004-11-13 0:11 ` 2.6.10-rc1-mm5 Greg KH
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=419CDB6F.2090406@aknet.ru \
--to=stsp@aknet.ru \
--cc=akpm@osdl.org \
--cc=linux-kernel@vger.kernel.org \
--cc=macro@linux-mips.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.