Bug 6846 - DMESG error: check_tsc_sync_source failed
Summary: DMESG error: check_tsc_sync_source failed
Status: RESOLVED OBSOLETE
Alias: None
Product: BSPs
Classification: Build System, Metadata & Runtime
Component: bsps-runtime (show other bugs)
Version: 1.7
Hardware: Other x86
: Medium normal
Target Milestone: 1.8
Assignee: Lucian Musat
QA Contact: Cristina Agurida
URL:
Whiteboard:
Depends on:
Blocks:
 
Reported: 2014-10-16 07:55 UTC by Lucian Musat
Modified: 2015-04-29 09:33 UTC (History)
3 users (show)

See Also:
OS type for building Yocto: ---
Type of Regression: ---
Verified:
Documentation change: Don't know

Attachments
Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Lucian Musat 2014-10-16 07:55:37 UTC 
Central error: [    0.182214] tsc: Marking TSC unstable due to check_tsc_sync_source failed

***********************

[    0.159841] ... max period:             000000007fffffff

[    0.159843] ... fixed-purpose events:   3

[    0.159845] ... event mask:             0000000700000003

[    0.169044] CPU 1 irqstacks, hard=f5cf0000 soft=f5cf2000

[    0.169049] x86: Booting SMP configuration:

[    0.169052] .... node  #0, CPUs:      #1

[    0.179142] Initializing CPU#1

[    0.180100] Disabled fast string operations

[    0.182190] TSC synchronization [CPU#0 -> CPU#1]:

[    0.182196] Measured 4312 cycles TSC warp between CPUs, turning off TSC clock.

[    0.182214] tsc: Marking TSC unstable due to check_tsc_sync_source failed

[    0.064018] CPU 2 irqstacks, hard=f5d06000 soft=f5d08000

[    0.064018]  #2

[    0.002000] Initializing CPU#2

[    0.002000] Disabled fast string operations

[    0.078268] CPU 3 irqstacks, hard=f5d12000 soft=f5d14000

[    0.078274]  #3

[    0.002000] Initializing CPU#3

[    0.002000] Disabled fast string operations

[    0.089510] x86: Booted up 1 node, 4 CPUs

[    0.089518] smpboot: Total of 4 processors activated (12769.56 BogoMIPS)


This error appears on the genericx86 image
Comment 1 Darren Hart 2014-10-16 09:13:20 UTC 
Which image (exact URL), which hardware platform? Please update the Platform field.
Comment 2 Lucian Musat 2014-10-16 10:33:04 UTC 
Hello Darren,

The tests for genericx86 are done on a Jetway Atom N2600 board. 

Our test script builds the rpm and psplash locally and downloads from http://autobuilder.yoctoproject.org/pub/releases/yocto-1.7.rc4/machines/genericx86/core-image-sato-sdk-genericx86.tar.gz for deployment on the board.

test-remote-image script is being used to automatically download the required tar.gz file from the root AB link (it goes by itself in <AB Link>/machines/<machine name>/<image type>.tar.gz). Our image type for tests is always core-image-sato-sdk. 

Thanks.
Comment 3 Darren Hart 2014-10-16 11:48:23 UTC 
So, it's very possible the TSC is indeed bad on this platform. Also possible that it has some horrible firmware SMM hack that also yields the TSC to be useless. This, sadly, is not an uncommon problem.
Comment 4 Darren Hart 2014-10-16 11:49:28 UTC 
Have you tested this image on another platform? Does it show the same error?
Comment 5 Lucian Musat 2014-12-16 09:39:08 UTC 
Hello Darren,

I have tested the genericx86 image on jetway and qemu. Only place the error appears is on jetway and I think that there is a problem with the board, as we have all sorts of issues with it and we will replace is as soon as possible.
This error does not appear on any other image type or board but jetway.
Comment 6 Lucian Musat 2015-04-29 09:33:07 UTC 
We don't test on the jetway any more so marking this as obsolete.