Bug 467927 - [RHEL5] patch enabling deep C states makes a RHTS machine hang on boot
[RHEL5] patch enabling deep C states makes a RHTS machine hang on boot
Product: Red Hat Enterprise Linux 5
Classification: Red Hat
Component: kernel (Show other bugs)
All Linux
medium Severity medium
: rc
: ---
Assigned To: John Feeney
Mike Gahagan
: Regression
: 465251 (view as bug list)
Depends On:
  Show dependency treegraph
Reported: 2008-10-21 14:18 EDT by Aristeu Rozanski
Modified: 2013-01-10 02:07 EST (History)
8 users (show)

See Also:
Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of:
Last Closed: 2009-01-20 15:17:04 EST
Type: ---
Regression: ---
Mount Type: ---
Documentation: ---
Verified Versions:
Category: ---
oVirt Team: ---
RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: ---

Attachments (Terms of Use)
Only use C3 states that have no bm_control flag if they're accessed via _CST (870 bytes, patch)
2008-10-30 09:23 EDT, Matthew Garrett
no flags Details | Diff

  None (edit)
Description Aristeu Rozanski 2008-10-21 14:18:44 EDT
Description of problem:
since kernel -99, athlon3.rhts.bos.redhat.com is hanging at boot time. after
some bisecting, I found that the patch linux-2.6-acpi-enable-deep-c-states-for-idle-efficiency.patch is the culprit. removing it on -99 or on -120 fixes
the problem and the machine boots without problems.

The problem is 100% reproducible.
Comment 2 Matthew Garrett 2008-10-24 17:48:31 EDT
At what point in the boot does the system hang?
Comment 3 Aristeu Rozanski 2008-10-27 09:48:31 EDT
Right before the serial initialization. I'll find a dmesg output and attach it
Comment 4 Aristeu Rozanski 2008-10-27 11:13:29 EDT
Here's the output:

the machine has no remote console (except serial console) and has LAPIC disabled
by BIOS (thus no NMI watchdog support), so I couldn't get more info.
Comment 5 Matthew Garrett 2008-10-28 17:19:43 EDT
Hmm. Does it boot if you force-enable the apic with the lapic boot option?
Comment 6 Aristeu Rozanski 2008-10-28 17:26:57 EDT
No. It crashes with nmi_watchdog enabled/disabled.
Comment 7 Matthew Garrett 2008-10-28 17:48:30 EDT
Ok. Does booting with the


parameter work? If so, what is the maximum number you can increase that to and still get a working boot?
Comment 8 Aristeu Rozanski 2008-10-29 11:36:19 EDT
this machine (on RHTS) is not reserved to me anymore. I'll reserve and test it,
unless you want to do it.
Comment 9 Matthew Garrett 2008-10-29 11:54:12 EDT
Yeah, rhts is giving me weird errors whenever I attempt to reserve it (Transaction aborted: Empty recipe_set at /usr/share/rhts/lib/Workflows.pm line 321), so might be worth you giving it a shot.
Comment 14 Matthew Garrett 2008-10-30 01:02:27 EDT
Looks like we need to backport ed3110efb538d7acbf635095c1382118f7414f75 from upstream - doing a test build now.
Comment 15 Matthew Garrett 2008-10-30 09:23:41 EDT
Created attachment 321932 [details]
Only use C3 states that have no bm_control flag if they're accessed via _CST

Confirmed to fix the system
Comment 19 Don Zickus 2008-11-04 11:51:08 EST
in kernel-2.6.18-122.el5
You can download this test kernel from http://people.redhat.com/dzickus/el5
Comment 21 Matthew Garrett 2008-11-18 08:41:35 EST
*** Bug 465251 has been marked as a duplicate of this bug. ***
Comment 24 errata-xmlrpc 2009-01-20 15:17:04 EST
An advisory has been issued which should help the problem
described in this bug report. This report is therefore being
closed with a resolution of ERRATA. For more information
on therefore solution and/or where to find the updated files,
please follow the link below. You may reopen this bug report
if the solution does not work for you.


Note You need to log in before you can comment on or make changes to this bug.