T3 and T4 systems may panic due to a NULL iv

Asset ID:	1-72-1537665.1
Update Date:	2017-09-23
Keywords:

Solution Type Problem Resolution Sure

Solution 1537665.1 : T3 and T4 systems may panic due to a NULL iv_handler

Applies to:

SPARC T4-4 - Version All Versions and later
SPARC T3-4 - Version All Versions and later
SPARC T3-1 - Version All Versions and later
SPARC T3-1B - Version All Versions and later
SPARC T3-2 - Version All Versions and later
Information in this document applies to any platform.
***Checked for relevance on 23-Sep-2017***

Symptoms

T4-4 systems may panic due to a NULL iv_handler with a stack looking similar to:

-- trap data  type: 0x9 (instruction access MMU miss)  rp: 0x2a1012079f0  --
  addr: 0x0
pc:  0x0 (invalid text address)
npc: 0x4 (invalid text address)
  global:                       %g1                  0
        %g2      0x300028a9a40  %g3               0x30
        %g4                  0  %g5      0x30007a42000
        %g6               0x1c  %g7      0x2a101207ca0
  out:  %o0                  0  %o1                  0
        %o2      0x30007a42000  %o3      0x2a101207d78
        %o4                0x1  %o5                  0
        %sp      0x2a101207291  %o7          0x100df44
  loc:  %l0          0x1009904  %l1                  0
        %l2                  0  %l3         0x70010cc0
        %l4                  0  %l5                  0
        %l6                  0  %l7      0x2a101197890
  in:   %i0               0x16  %i1                0x1
        %i2                  0  %i3             0x1000
        %i4  0x101010100000000  %i5               0x12
        %fp      0x2a101197131  %i7          0x1042f2c
<trap>unix:intr_thread+0x198(0x16, 0x1, 0x0, 0x1000, 0x101010100000000, 0x12)
unix:ktl0+0x64()
-- interrupt data  rp: 0x2a101197890
pc:  0x1042f34 unix:cpu_halt+0x10c:   call      unix:disable_vec_intr
npc: 0x1042f38 unix:cpu_halt+0x110:   nop
  global:                       %g1                0x1
        %g2      0x300028a9a40  %g3               0x30
        %g4                  0  %g5      0x30007a42000
        %g6               0x1c  %g7      0x2a101197ca0
  out:  %o0               0x16  %o1                0x1
        %o2                  0  %o3             0x1000
        %o4  0x101010100000000  %o5               0x12
        %sp      0x2a101197131  %o7          0x1042f2c
  loc:  %l0      0x6001466051c  %l1                0x1
        %l2               0x16  %l3                  0
        %l4                  0  %l5                0x2
        %l6      0x30007a42178  %l7                0x1
  in:   %i0      0x30007a42000  %i1                0x6
        %i2          0x1913a30  %i3          0x1913900
        %i4      0x30007a42000  %i5                  0
        %fp      0x2a1011971e1  %i7          0x106bd5c
<intr trap>unix:cpu_halt+0x10c()
unix:idle+0x128(0x0, 0x0)
unix:thread_start+0x4()
-- end of CPU6 idle thread stack --

Cause

The explanation of this issue is beyond the scope of this document.

If a panic is encountered with a similar stack frame and confirmation is desired then please open a service request and provide the output of the explorer script and the system crash dump (vmcore.x or vmdump.x).

Please see the references at the bottom of this document for documents providing additional details and guidance on what data to provide and how.

Solution

A fix for this issue has been integrated and released starting with system firmware "Sun System Firmware 8.2.1.b"

References

<NOTE:1312847.1> - Oracle Explorer Data Collector Resource Center
<NOTE:1020199.1> - How to Upload Support Files to Oracle Such as Explorer and Crash Dumps
<NOTE:1010911.1> - What Should I Send to Oracle After a Solaris Panic or Unexpected Reboot?
<BUG:15795199> - SUNBT7173177-1.11.X PANIC ON A NULL IV_HANDLER, THE SPECIFIC INTERRUPT BEING HAN

Attachments

This solution has no attachment