|
Bug ID
|
6363985
|
|
Synopsis
|
acpica: SMB Alerts for some BIOSes result in high background system load
|
|
State
|
10-Fix Delivered (Fix available in build)
|
|
Category:Subcategory
|
driver:acpi
|
|
Keywords
|
BIOS | acpi | cpu | looping | metropolis | spin | system | time | vmstat | w2100z
|
|
Responsible Engineer
|
Alan Perry
|
|
Reported Against
|
s10u2_fcs
|
|
Duplicate Of
|
|
|
Introduced In
|
|
|
Commit to Fix
|
snv_40
|
|
Fixed In
|
snv_40
|
|
Release Fixed
|
solaris_nevada(snv_40)
,
solaris_10u4(s10u4_04) (Bug ID:2143384)
|
|
Related Bugs
|
6301776
,
6362141
,
6559975
|
|
Submit Date
|
15-December-2005
|
|
Last Update Date
|
9-February-2009
|
|
Description
|
My machine was sluggish, and debugging eventually determined that thermal events were causing an ACPI interrupt storm; see Comments for details.
[Additional description and search keywords added at customer request to make this issue easier to find]
This issue is commonly seen on W2100z Java Workstations after upgrading to Solaris 10 Update 2 or above. (S10 FCS and update 1 do not see this issue).
Symptoms :
High system time seen in vmstat :
# vmstat 5 5
kthr memory page disk faults cpu
r b w swap free re mf pi po fr de sr s0 s1 -- -- in sy cs us sy id
0 0 0 3180808 2882228 0 1 0 0 0 0 0 0 0 0 0 21511 983 21257 0 52 48
0 0 0 3151052 2844368 0 11 0 0 0 0 0 0 0 0 0 21442 996 21193 0 52 48
0 0 0 3151052 2844368 0 0 0 0 0 0 0 0 0 0 0 21481 1040 21239 0 52 48
0 0 0 3151052 2844368 0 0 0 0 0 0 0 0 0 0 0 21489 956 21240 0 52 48
0 0 0 3151052 2844368 0 0 0 0 0 0 0 0 0 0 0 21486 737 21242 0 52 48
System time on the W2100z sits around 50% as one of the two cpus is busy in ACPI code due to the ACPI interrupt storm as mentioned above.
|
|
Work Around
|
boot with acpi-user-options=0x8
|
|
Comments
|
N/A
|