Sun Fire [TM] V1280/E2900, Netra 1280/1290 system: unable to boot due to "ERROR: Fast Data Access MMU Miss" and/or "Fatal SCSI error at script address 148 Illegal instruction"

Asset ID:	1-72-1583559.1
Update Date:	2017-10-11
Keywords:

Solution Type Problem Resolution Sure

Solution 1583559.1 : Sun Fire [TM] V1280/E2900, Netra 1280/1290 system: unable to boot due to "ERROR: Fast Data Access MMU Miss" and/or "Fatal SCSI error at script address 148 Illegal instruction"

Applies to:

Sun Fire V1280 Server - Version All Versions to All Versions [Release All Releases]
Sun Fire E2900 Server - Version All Versions to All Versions [Release All Releases]
Sun Netra 1290 Server - Version All Versions to All Versions [Release All Releases]
Sun Netra 1280 Server - Version All Versions to All Versions [Release All Releases]
Information in this document applies to any platform.

Symptoms

Failure to boot with SC console log of the following "ERROR: Fast Data Access MMU Miss" or "Fatal SCSI error at script address 148 Illegal instruction"

Example1:

.....

{/N0/SB0/P0} DCB_ENTER_OBP command succeeded
{/N0/SB0/P1} DCB_ENTER_OBP command succeeded
{/N0/SB0/P2} DCB_ENTER_OBP command succeeded
{/N0/SB0/P3} DCB_ENTER_OBP command succeeded

pci bootbus-controller pci
Probing /ssm@0,0/pci@18,700000 Device 1 Nothing there
Probing /ssm@0,0/pci@18,700000 Device 2 Nothing there
Probing /ssm@0,0/pci@18,700000 Device 3 ide disk cdrom
Probing /ssm@0,0/pci@18,600000 Device 1 network
Probing /ssm@0,0/pci@18,600000 Device 2 scsi disk tape scsi disk tape
pci pci
Probing /ssm@0,0/pci@19,700000 Device 1 Nothing there
Probing /ssm@0,0/pci@19,700000 Device 2 fibre-channel
Probing /ssm@0,0/pci@19,700000 Device 3 fibre-channel
Probing /ssm@0,0/pci@19,600000 Device 1 network
Probing /ssm@0,0/pci@19,600000 Device 2 network
Authorized uses only. All activity will be monitored and reported.
Copyright 2001-2004 Sun Microsystems, Inc. All rights reserved.
Use is subject to license terms.
SmartFirmware, Copyright (C) 1996-2001. All rights reserved.

Fatal SCSI error at script address 148 Illegal instruction

Example 2:

....

pci bootbus-controller pci
Probing /ssm@0,0/pci@18,700000 Device 1 Nothing there
Probing /ssm@0,0/pci@18,700000 Device 2 Nothing there
Probing /ssm@0,0/pci@18,700000 Device 3 ide disk cdrom
Probing /ssm@0,0/pci@18,600000 Device 1 network
Probing /ssm@0,0/pci@18,600000 Device 2 scsi disk tape scsi disk tape
pci pci
Probing /ssm@0,0/pci@19,700000 Device 1 Nothing there
Probing /ssm@0,0/pci@19,700000 Device 2 fibre-channel
Probing /ssm@0,0/pci@19,700000 Device 3 fibre-channel
Probing /ssm@0,0/pci@19,600000 Device 1 network
Probing /ssm@0,0/pci@19,600000 Device 2 network
Authorized uses only. All activity will be monitored and reported.
Copyright 2001-2004 Sun Microsystems, Inc. All rights reserved.
Use is subject to license terms.
SmartFirmware, Copyright (C) 1996-2001. All rights reserved.

Script interrupt: Reserved phase
ERROR: Fast Data Access MMU Miss

Example 3:

{0} ok boot
TL = 1, TT = 68. ERROR: Fast Data Access MMU Miss
TSTATE= 0x1407 [ccr = 0x0, asi = 0x0, pstate = 0x14, cwp = 0x7]
TPC= 00000000f0036f4c
TNPC= 00000000f0036f50
SFSR= 000000000080800b, TAGACCESS = 0000000000000000
D-SFAR = 0000000000000006
TICK= 800000d60f332d88, TICKCMP = ffffffffffffffff

{0} ok boot -F failsafe
TL = 1, TT = 68. ERROR: Fast Data Access MMU Miss
TSTATE= 0x1407 [ccr = 0x0, asi = 0x0, pstate = 0x14, cwp = 0x7]
TPC= 00000000f0036f4c
TNPC= 00000000f0036f50
SFSR= 000000000080800b, TAGACCESS = 0000000000000000
D-SFAR = 0000000000000006
TICK= 8000005d1e5fd050, TICKCMP = ffffffffffffffff

=======================================================================

reset-all was run followed by probe-scsi-all

Below is the output of probe-scsi-all where it shows that it gives error while trying to sense the internal boot disks.

ok probe-scsi-all

/ssm@0,0/pci@18,600000/scsi@2,1
Target 8
Unit 0 Disk HITACHI HUS15143BSUN146GPA02 286739329 Blocks, 140009 MB
Target 9
Unit 0 Disk HITACHI HUS15143BSUN146GPA02 286739329 Blocks, 140009 MB
Target a
Unit 0 Disk HITACHI HUS15143BSUN146GPA02 286739329 Blocks, 140009 MB
Target b
Unit 0 Disk HITACHI HUS15143BSUN146GPA02 286739329 Blocks, 140009 MB
Target c
Unit 0 Disk HITACHI HUS15143BSUN146GPA02 286739329 Blocks, 140009 MB
Target d
Unit 0 Disk SEAGATE ST314655LSUN146G0491 286739329 Blocks, 140009 MB
Target f
Unit 0 Processor SUN StorEdge 3320 D1180

/ssm@0,0/pci@18,600000/scsi@2
Target 1
Unit 0 Disk HITACHI HUS15143BSUN146GPA02 286739329 Blocks, 140009 MB.
Target 8
Unit 0 Disk HITACHI HUS15143BSUN146GPA02 286739329 Blocks, 140009 MB
Target a
Unit 0 Disk HITACHI HUS15143BSUN146GPA02 286739329 Blocks, 140009 MB
Target b
Unit 0 Disk HITACHI HUS15143BSUN146GPA02 286739329 Blocks, 140009 MB
Target c
Unit 0 Disk HITACHI HUS15143BSUN146GPA02 286739329 Blocks, 140009 MB
Target d
Unit 0 Disk HITACHI HUS15143BSUN146GPA02 286739329 Blocks, 140009 MB

/ssm@0,0/pci@18,700000/scsi@2,1

/ssm@0,0/pci@18,700000/scsi@2

TL = 1, TT = 68. ERROR: Fast Data Access MMU Miss
TSTATE= 0x1405 [ccr = 0x0, asi = 0x0, pstate = 0x14, cwp = 0x5]
TPC= 00000000f000fb6c
TNPC= 00000000f000fb70
SFSR= 000000000080800f, TAGACCESS = 0000000100000000
D-SFAR = 00000001000000fd
TICK= 8000004ccde8d4d8, TICKCMP = ffffffffffffffff

{0} ok

Changes

There are some scenarios that can lead to this behavior.

Some of them are listed here:

Customer shutdown the platform and install new SAN storage and the platform fails to boot
internal disk replacement
internal devices replacement

Cause

A unexpected SCSI Fatal event occurs due to possible timeout from the target or HBA or SCSI chip

Basically OBP call to jump to a selected Target device (i.e. SCSI ID)

1172   dup 32f73 and if    \ Fatal Error        ( status )
1173      ." Fatal SCSI error " .script-address
1174      show-status
1175      error-reset
1176      fatal-error true exit
1177   then                                      ( status )
1178
1179   \ Timeout (no interrupt status)
1180   ?dup 0= if timed-out true exit then    ( status )
1181
1182   \ The only cases left are "arbitration complete" (04000), which is
1183   \ not interesting, and "reselected" (01000)
1184   ." Unexpected SCSI Interrupt:" cr show-status fatal-error true
1185;

Solution

Three possible resolution to the issue:

Platform has just added SAN storage and/or an HBA

Revert the platform to previously known good working configuration
Remove the SCSI/QLC/Emulex HBA that was recently installed

Platform has made changes prior to the last reboot

Re-seat the HW involved in the previous change
check if probe-scsi-all now see everything after having issues a reset-all

Platform has not made any changes prior to the last reboot

Ensure SC and board firmware are current and latest
Boot the platform with an alternate boot device such as another disk, net, or cdrom
If an alternate boot fails replace the SCSI HBA/cable and/or the onboard SCSI (IB_SSC) and media bay
probe-scsi-all should be failing and show no disk in this case

To discuss this information further with Oracle experts and industry peers, we encourage you to review, join or start a discussion in an appropriate
My Oracle Support Community - Oracle Sun Technologies Community.

Attachments

This solution has no attachment