Document fcos/A0193-1
FCO #: A0193-1
SYNOPSIS: Sun Fire 15K systems with hsPCI
DATE: May/15/2002
KEYWORDS: Sun Fire 15K systems with hsPCI
----------------------------------------------------------------------------
- Sun Proprietary/Confidential: Internal Use Only -
----------------------------------------------------------------------------
FIELD CHANGE ORDER
(For Authorized Distribution by Enterprise Services)
FCO #: A0193-1 Date: May/15/2002
SYNOPSIS: Sun Fire 15K systems with hsPCI boards having Schizo
2.2 ASICs can suffer a panic due to a timing race
condition.
Sun Alert: Y
TOP FIN/FCO REPORT: Yes
PRODUCT_REFERENCE: F15K hsPCI boards with Schizo 2.2 ASICs
PRODUCT CATEGORY: Server / System Component
PRODUCT AFFECTED:
Mkt_ID Platform Model Description Serial Number
------ -------- ----- ----------- -------------
- F15K - Sun Fire 15K _
X-Options Affected
--------- -------
Mkt_ID Platform Model Description Serial Number
------ -------- ----- ----------- -------------
X4575A F15K - hsPCI Assembly -
AFFECTED PARTS:
Part Number Description Model
----------- ----------- -----
501-5397-08(Or Less) hsPCI I/O Board -
REFERENCES :
ESC: 535301
ESC: 535705
ESC: 536269
ESC: 536227
ESC: 535927
ESC: 536469
ECO: WO_23260
ECO: WO_23112
ECO: WO_23410
PatchID: 112665 Nexus Driver
Manual:806-3511-10 Sun Fire 15K HW Installation and De-Installation Guide
LEAP: 1963
FIN: I0820-1
SunAlert: 44582
PROBLEM DESCRIPTION:
Under certain conditions Sun Fire 15K servers, with hsPCI I/O boards that
contain Schizo 2.2 ASICs, may experience a domain panic. Only Sun Fire
15K servers shipped prior to April 1, 2002 are affected. Sun Fire 12K
servers are not impacted as none where shipped with affected hsPCI I/O boards.
No silent data corruption occurs has a result of this issue.
Domain panics may occur with the following Sun Fire 15K configurations:
Hardware Components:
hsPCI I/O Board revision 2.2 (501-5397-08 or lower)
JNI 32-bit PCI-to-Fibre Channel HBA (FCI-1063-x)
SunSwift PCI SCSI (Fresh Choice) adapters (X1032A)
Nexus Driver:
pcisch (PCI Bus nexus driver 1.199)
NOTE: Version 1.199 of the Nexus Driver is the default version shipped
with Solaris 8. This is the version installed unless Patch
112665 has been applied.
Domain Configuration:
20+ CPUs utilizing ISP (SCSI HBA Driver)
or JNI FCI-1063-x HBA are more susceptible.
This failure has only been observed with ISP and/or JNI drivers.
However, not all configurations with the ISP and/or the JNI driver are
affected.
The failure is configuration specific. When the failure manifests itself as a
panic, these drivers are in the panic string, which helps to identify the
failure.
When the panic has been seen with the ISP driver, the panic string is:
"isp_scsi_impl_pktfree: freeing free packet"
For the JNI driver used with the JNI FCI-1063-x HBA, the system will log
"INB_SCSI_COMPLETE interrupt with INVALID tag" errors before the domain
panics with a "panic assertion failed:" These error messages and panic
strings correlate with domain configurations utilizing the JNI FCI-1063-x
Host-Bus-Adapter.
Domain configurations utilizing an ISP (SCSI Host Bus Adapter Driver)
configuration generate the "isp_scsi_impl_pktfree: freeing free packet"
panic string.
In the JNI configuration the following errors were observed;
"INB_SCSI_COMPLETE interrupt with INVALID tag"
before the domain crashed with a "panic assertion failed:"
With the hsPCI configuration the domain will crash with an
"isp_scsi_impl_pktfree: freeing free packet" panic string.
Root cause of this problem is due to a transaction ordering issue within the
I/O controller. The I/O controller does not follow certain ordering rules
and may have data remaining from a previous read/write while the current
transaction is being processed.
Corrective action was made available in manufacturing via ECO# WO_23260
by dash rolling FRU part number 501-5397 to -09 on March 07, 2002.
Corrective Action was made available in Enterprise Services via
LEAP# 1963 on April 17, 2002.
Sun Legal approved Customer Letter can be located at;
http://sdpsweb.EBay/FIN_FCO/FCO/FCO_A0192-1_Dir/CustomerLetter.ps
CUSTOMER LIST: Reference the following URL for a list of affected customer
shipments;
http://sdpsweb.ebay/FIN_FCO/FCO/FCO_A0193-1_Dir/F15Kcust.sdc
Note: To view document click on the above URL, then save to your local
disk using your Netscape 'file' button and select 'save as', then
open file locally using StarOffice.
PLANNED IMPLEMENTION COMPLETION DATE: October 31, 2002
IMPLEMENTATION:
---
| X | MANDATORY (Fully Pro-Active)
---
---
| | CONTROLLED PRO-ACTIVE (per Sun Geo Plan)
---
---
| | UPON FAILURE
---
REPLACEMENT TIME ESTIMATE: 2 hours
SPECIAL CONSIDERATION:
The below link to a Sun Alert has a "Restricted" Distribution. Please
print
and use this Sun Alert in addition to, or in place of, the customer letter.
Communicate to your *affected* customers only. Typically, Sun Alerts have a
wider distribution on Contract and Free SunSolve, but the Sun Alert program
has been enhanced to include what is known as "Targeted Sun Alerts" for
affected customers only.
http://sdpsweb.EBay/FIN_FCO/FCO/FCO_A0193-1_Dir/44582_public.html
Note: If you have questions please contact the Sun Alert Program Office.
Ref; http://sunalert.ebay/progorg.html
CORRECTIVE ACTION :
All swap activities should be directed to the following Enterprise Services
timezone representatives to ensure proper prioritization with the Global
Prioritization Committee (GPC);
EMEA: Richard Porter
AMER: Gary Replogle
APAC: Kam-Weng Goh
The following is a list of GPC Global Sales Organization timezone
representatives;
AMER (US): Jeff Barteld
AMER (INTL): Kerry Roller
EMEA: Jon Ireland
APAC: Peter Chadford
Proactivly Replace Schizo 2.2 ASIC based hsPCI assemblies, 501-5397-08 (or
below)
with Schizo 2.3 ASIC based hsPCI assemblies, 501-5397-09 (or above).
Additionally, upgrade of Nexus Driver with point patch 112665 is required.
Until this point patch becomes available on SunSolve it can temporary be
located at the below URL;
http://sdpsweb.ebay/FIN_FCO/FCO/FCO_A0193-1_Dir/112665-01.tar
Tag all returned boards with "FCO A0193-1" and return via Overnight
Freight.
COMMENTS:
BILLING TYPE:
Warranty: Sun will provide parts at no charge under Warranty
Service. On-Site Labor Rates are based on how the
system was initially installed.
Contract: Sun will provide parts at no charge. On-Site Labor Rates
are based on the type of service contract.
Non Contract: Sun will provide parts at no charge. Installation by
Sun is available based on the On-Site Labor Rates
defined in the Price List.
--------------------------------------------------------------------------
Implementation Footnote:
________________________
i) In case of Mandatory FCOs, Enterprise Services will attempt to contact
all known customers to recommend the part upgrade.
ii) For controlled proactive swap FCOs, Enterprise Services mission critical
support teams will initiate proactive swap efforts for their respective
accounts, as required.
iii) For Replace upon Failure FCOs, Enterprise Services partners will implement
the necessary corrective actions as and when they are required.
--------------------------------------------------------------------------
All released FINs and FCOs can be accessed using your favorite network
browser as follows:
SunWeb Access:
______________
* Access the top level URL of http://sdpsweb.EBay/FIN_FCO/
* From there, select the appropriate link to query or browse the FIN and
FCO Homepage collections.
SunSolve Online Access:
_______________________
* Access the SunSolve Online URL at http://sunsolve.Central/
* From there, select the appropriate link to browse the FIN or FCO index.
Internet Access:
_______________
* Access the top level URL of https://infoserver.Sun.COM
--------------------------------------------------------------------------
General:
________
Send questions or comments to finfco-manager@sdpsweb.EBay
---------------------------------------------------------------------------
Copyright (c) 1997-2003 Sun Microsystems, Inc.