[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference kernel::csguk_systems

Title:CSGUK_SYSTEMS
Notice:No restrictions on keyword creation
Moderator:KERNEL::ADAMS
Created:Wed Mar 01 1989
Last Modified:Thu Nov 28 1996
Last Successful Update:Fri Jun 06 1997
Number of topics:242
Total number of notes:1855

199.0. "watch induced crash @ Rocal - see .1" by COMICS::GLEDHILL () Wed Dec 28 1994 13:39

T.RTitleUserPersonal
Name
DateLines
199.1rocol watch induced crash.COMICS::GLEDHILLWed Jan 04 1995 01:31104
From:	NICES::GLEDHILL      4-JAN-1995 01:14:49.52
To:	GLEDHILL
CC:	GLEDHILL
Subj:	 

*****  THIS IS A READ ONLY COPY FROM NICE   -   please handle accordingly  *****
********************************************************************************

Log No            24056.00-4C6-1UVO          Queue      GLEDHILL       
Log D/T           20-DEC-1994 17:04          Owner      GLEDHILL                 
LSDT D/T           3-JAN-1995 09:00          Loc/Phone  UVO  3245
Status as at       4-JAN-1995 01:15 is OPEN                                    
EXT REQ Stat Code             Escalation Indicator  Y
Hold Indicator    N           Planned Indicator     Y

---------------------------------Customer---------------------------------------
Company          ROCOL LTD                                                      
Department       ROCOL HOUSE                             
Street           SWILLINGTON                             
City             LEEDS                         
Postal Code      LS26 8BS                    PO No                          

Caller           JOHN ELLIOT                 Title      MR             
Phone            0532 866511                 Extension  ASK   
Service Wish     ** dump on ryrgrs::$4$DIA1:[24056_ROCOL]SYSDUMP.DMP;1 **       

--------------------------------------------------------------------------------
******************************INITIAL CALL DETAILS******************************
UPDATED BY / DATE-TIME        :: Theo Bartley  Wed 21-Dec-1994::09:56
PROBLEM DESCRIPTION / CONFIG  :: System bugchecked
ACTION PLAN                   :: Look at dump.
NEXT UPDATE TIME & DATE       :: Cust's modem is in use.  He will call back
                                 later when it becomes free
   ******************************   Update   ******************************
UPDATED BY /DATE-TIME         :: Theo Bartley / Wed 21-Dec-1994::11:13
CURRENT STATUS                :: Customer is unable to give us access to the
                                 system to look at the dump online.  He will
                                 send the dump in on tape.  No errors in the
                                 errorlog, so customer won't include errorlog
                                 on the tape.  Customer ran WATCH just before
                                 system crashed.
ACTION PLAN                   :: Analyse dump.
NEXT UPDATE TIME & DATE       :: Call cust if tape hasn't arrived by Fri 12:00

   ******************************   Update   ******************************
UPDATED BY /DATE-TIME         :: MIKE AYLING  23-DEC-1994  10:10
CURRENT STATUS                :: TK50 TAPE RECEIVED
ACTION PLAN                   :: TO BE ANALYSED BY SPECIALIST WHEN AVAILABLE
NEXT UPDATE TIME & DATE       :: 

   ******************************   Update   ******************************
UPDATED BY /DATE-TIME         :: Norm Pettet 23-Dec-1994:16:58
CURRENT STATUS                :: dump copied to ryrgrs::$4$DIA1:[24056_ROCOL]
                                 Directory $4$DIA1:[24056_ROCOL]

                                 SYSDUMP.DMP;1          90892

                                 Total of 1 file, 90892 blocks.

ACTION PLAN                   :: 
NEXT UPDATE TIME & DATE       :: 


   ******************************   Update   ******************************
UPDATED BY /DATE-TIME         :: Norm Pettet
CURRENT STATUS                :: process crash - lloks like a process run-down
                                 problem - further re-work required
ACTION PLAN                   :: 
NEXT UPDATE TIME & DATE       :: 

   ******************************   Update   ******************************
UPDATED BY /DATE-TIME         :: Norm Pettet 31-Dec-1994
CURRENT STATUS                :: Reading the supplied information from the
                                 customer, it appears he was using DEBUG and
                                 was exiting when user IAN had a problem. User
                                 IAN logged in again then customer, using WATCH
                                 from DECUS, monitored his process. It was then
                                 that the system crashed. I can't find however 
				 the initial login from user IAN.
ACTION PLAN                   :: Suggest discussing with customer - if this is
                                 the first time this crash has occured then
                                 wait for it to happen again (if ever)

--------------------------------------------------------------------------------
--------------------------------------------------------------------------------
canasta description.


 Status:          UNIDENTIFIED                        
 VMS Version:     6.1                          More Admin Reference     
 CPU Type:        3100-95                             
 Bugcheck:        SSRVEXCEPT                                                  
 Process Name:    IAN                                                         
 Image Name:      SSU                                                         
 SID Register:    13000202                                                    
 Argument Count:  00000005                                                    
 Condition Code:  0000000C                                                    
 Reason Mask:     00000001                                                    
 Virtual Address: 00C00008                                                    
 Exception PC:    80000010                                                    
 Exception PSL:   00000009                                                    
 Module:          EXE$QIOW_3                                                  
 Offset:          00000000                                                    
 Instruction:     CALLG                                                       
199.2watch induced crash @ Rocal - see .1COMICS::GLEDHILLWed Jan 04 1995 01:33363
--------------------------------------------------------------------------------
Log No             24056.00-4C6-1UVO           Desc type      TS
Sequence no        01                          Authr badge no 064297
                                               Creation D/T   23-DEC-1994 18:56
--------------------------------------------------------------------------------
VAX/VMS System dump analyzer
Dump taken on 20-DEC-1994 16:48:05.38
SSRVEXCEPT, Unexpected system service exception
System crash information
------------------------
Time of system crash: 20-DEC-1994 16:48:05.38
Version of system: VAX/VMS VERSION V6.1
System Version Major ID/Minor ID: 1/0
VAXcluster node: WIZZ, a MicroVAX 3100-95
Crash CPU ID/Primary CPU ID:  00/00
Bitmask of CPUs active/available:  00000001/00000001
CPU bugcheck codes:
        CPU 00 -- SSRVEXCEPT, Unexpected system service exception
CPU 00 reason for Bugcheck: SSRVEXCEPT, Unexpected system service exception
Process currently executing on this CPU: IAN
Current image file: $2$DKA100:[SYS0.SYSCOMMON.][SYSEXE]SSU.EXE;1
Current IPL: 0  (decimal)
CPU database address:  8505A000
General registers:
        R0  = 00000009   R1  = 7FFE7790   R2  = 81839004   R3  = 00C00001
        R4  = 84403980   R5  = 842FBAC0   R6  = 81839004   R7  = 00000001
        R8  = 00005B30   R9  = 00000000   R10 = 00005690   R11 = 00000000
        AP  = 7FFE776C   FP  = 7FFE7754   SP  = 7FFE7754   PC  = 838AE5E3
        PSL = 00000000
Processor registers:
        P0BR   = 88ED4000     SBR    = 04DC4C00     ASTLVL = 00000004
        P0LR   = 00000462     SLR    = 0008A900     SISR   = 00000000
        P1BR   = 8877BE00     PCBB   = 0440A620     ICCS   = 00000041
        P1LR   = 001FF3A2     SCBB   = 04DBDE00     SID    = 13000202
        TODR   = C62660A9     BPCR   = ECC80024     PAMODE = 00000000
        MMEPTE = 88EEC000     MMESTS = 1C004001     PCSCR  = 01000200
        ICSR   = 00000001     ECR    = 0000008A     TBSTS  = 800001D0
        PCCTL  = FFFFFC13     PCSTS  = FFFFF800     CEFSTS = 00019200
        CCTL   = 00000021     BCEDSTS= 00000000     BCETSTS= 00000000
        NESTS  = 00000000     NEOCMD = 0000F004     NEICMD = 00000000
        MESR   = 00006000     MMCDSR = 09CBC000
        DSER   = 00000000     CBTCR  = 00000000
        ISP    = 8505B600
        KSP    = 7FFE7754
        ESP    = 7FFE9800
        SSP    = 7FFECA44
        USP    = 7FE76908
                No spinlocks currently owned by CPU 00
Current operating stack (KERNEL):

                7FFE7734  00000000
                7FFE7738  00005690      DAP$K_CRC_TBL1+002AD
                7FFE773C  00000000
                7FFE7740  7FFE776C      CTL$GL_KSTKBAS+0056C
                7FFE7744  7FFE7754      CTL$GL_KSTKBAS+00554
                7FFE7748  7FFE774C      CTL$GL_KSTKBAS+0054C
                7FFE774C  838AE5E3      EXE$EXCPTNE
                7FFE7750  00000000

         SP =>  7FFE7754  00000000
                7FFE7758  00000000
                7FFE775C  7FE76944
                7FFE7760  7FFE77E4      CTL$GL_KSTKBAS+005E4
                7FFE7764  80000014      EXE$QIOW_3+00004
                7FFE7768  838B0AD1      EXE$CONTSIGNAL+0007C
                7FFE776C  00000002
                7FFE7770  7FFE7790      CTL$GL_KSTKBAS+00590
                7FFE7774  7FFE7778      CTL$GL_KSTKBAS+00578
                7FFE7778  00000004
                7FFE777C  7FFE77E4      CTL$GL_KSTKBAS+005E4
                7FFE7780  FFFFFFFD      LKB$K_SCSWAIT
                7FFE7784  00000000
                7FFE7788  00C00008      PSL$M_PRVMOD+00008
                7FFE778C  0000000A
                7FFE7790  00000005
                7FFE7794  0000000C
                7FFE7798  00000001
                7FFE779C  00C00008      PSL$M_PRVMOD+00008
                7FFE77A0  80000010      EXE$QIOW_3
                7FFE77A4  00000009
                7FFE77A8  838B0A4D      EXE$SRCHANDLER+00076
                7FFE77AC  00000002
                7FFE77B0  7FFE77D0      CTL$GL_KSTKBAS+005D0
                7FFE77B4  7FFE77B8      CTL$GL_KSTKBAS+005B8
                7FFE77B8  00000004
                7FFE77BC  7FFE77E4      CTL$GL_KSTKBAS+005E4
                7FFE77C0  00000000
                7FFE77C4  84423618
                7FFE77C8  00000001
                7FFE77CC  05000001      NFB$C_PLI_LCK
                7FFE77D0  00000005
                7FFE77D4  0000000C
                7FFE77D8  00000000
                7FFE77DC  00000000
                7FFE77E0  00000000
                7FFE77E4  00C00008      PSL$M_PRVMOD+00008
                7FFE77E8  00000000
                7FFE77EC  7FE76944
                7FFE77F0  7FE76908
                7FFE77F4  838AE858      EXE$EXCEPTION+00227
                7FFE77F8  0000782E      DAP$K_CRC_TBLC+003AC
                7FFE77FC  03C00000



SDA> ex/in 838B0A4D-20;20	; From stack
%SDA-W-INSKIPPED, unreasonable instruction stream - 1 bytes skipped
EXE$SRCHANDLER+00057:  BRW     EXE$CONTSIGNAL+0003D
EXE$SRCHANDLER+0005A:  BRW     EXE$CONTSIGNAL
EXE$SRCHANDLER+0005D:  CALLG   (SP),EXE$CONTSIGNAL+000FD
EXE$SRCHANDLER+00062:  BLBC    R0,EXE$CONTSIGNAL+00054
EXE$SRCHANDLER+00065:  BBC     #01,20(SP),EXE$SRCHANDLER+00070
EXE$SRCHANDLER+0006A:  INSV    #04,#00,#03,28(SP)
EXE$SRCHANDLER+00070:  JSB     EXE$CALL_HANDL_JACKET	; -------
EXE$SRCHANDLER+00076:  BLBC    R0,EXE$SRCHANDLER+0005D		|
								|
SDA> ex/in EXE$CALL_HANDL_JACKET;20			<--------
EXE$CALL_HANDL_JACKET:  MOVPSL  R0
EXE$CALL_HANDL_JACKET+00002:  CMPZV   #18,#02,R0,#03
EXE$CALL_HANDL_JACKET+00007:  BNEQ    EXE$CALL_HANDL_JACKET+00016
EXE$CALL_HANDL_JACKET+00009:  BLBC    @#CTL$GL_VP_FLAGS,EXE$CALL_HANDL_JACKET+00
016
EXE$CALL_HANDL_JACKET+00010:  JSB     @#V_EXE$SET_VP_JACKET
EXE$CALL_HANDL_JACKET+00016:  JMP     @#EXE$QIOW_3	;-------
								|
								|
EXE$QIOW_2+00007:  RET						|
EXE$QIOW_3:  CALLG   04(SP),(R1) ; crash here R1=C00008  <-------


844039B0   PCB$L_ONQTIME                   0105C864
844039B4   PCB$L_WAITIME                   0105C695

SDA> ex/in 00006C41-20;20
%SDA-W-INSKIPPED, unreasonable instruction stream - 1 bytes skipped
DAP$K_CRC_TBL7+00280:  CALLS   #03,@DAP$K_CRC_TBLC+00006
DAP$K_CRC_TBL7+00287:  CLRQ    -(SP)
DAP$K_CRC_TBL7+00289:  PUSHL   R5
DAP$K_CRC_TBL7+0028B:  MOVL    2739(R2),R0
DAP$K_CRC_TBL7+00290:  PUSHAB  04(R0)
DAP$K_CRC_TBL7+00293:  CLRQ    -(SP)
DAP$K_CRC_TBL7+00295:  PUSHL   04(R2)
DAP$K_CRC_TBL7+00298:  CALLS   #07,@DAP$K_CRC_TBLC+00026
DAP$K_CRC_TBL7+0029F:  CMPL    R0,#01
SDA>


SDA> ex/in 80000010-20;20
%SDA-W-INSKIPPED, unreasonable instruction stream - 18 bytes skipped
SYS$S0_VECTOR_BASE+00002:  CHMK    #002D
SYS$S0_VECTOR_BASE+00006:  BLBC    R0,EXE$QIOW_2+00007
EXE$QIOW_2+00001:  PUSHL   10(AP)
EXE$QIOW_2+00004:  BRW     EXE$SYNCH+00005
EXE$QIOW_2+00007:  RET
EXE$QIOW_3:  CALLG   04(SP),(R1)
Process index: 0095   Name: IAN   Extended PID: 21600A95
--------------------------------------------------------
Status : 02040001 res,phdres,inter
Status2: 00000001 quantum_resched
PCB address              84403980    JIB address              84418EC0
PHD address              88EA7400    Swapfile disk address    00000000
Master internal PID      00050095    Subprocess count                0
Internal PID             00050095    Creator internal PID     00000000
Extended PID             21600A95    Creator extended PID     00000000
State                       CUR      Termination mailbox          0000
Current priority                9    AST's enabled                KESU
Base priority                   4    AST's active                 NONE
UIC                [00300,000007]    AST's remaining                94
Mutex count                     0    Buffered I/O count/limit      100/100
Waiting EF cluster              1    Direct I/O count/limit        100/100
Starting wait time       1B001B1B    BUFIO byte count/limit      31872/31936
Event flag wait mask     03FFFFFF    # open files allowed left      98
Local EF cluster 0       E0000001    Timer entries allowed left     10
Local EF cluster 1       02000000    Active page table count         0
Global cluster 2 pointer 00000000    Process WS page count         410
Global cluster 3 pointer 00000000    Global WS page count          143
SDA>
SDA> sho call
Call Frame Information
----------------------
        Call Frame Generated by CALLG Instruction
Condition Handler       7FFE7754  00000000
SP Align Bits = 00      7FFE7758  00000000
   Saved  AP            7FFE775C  7FE76944
   Saved  FP            7FFE7760  7FFE77E4      CTL$GL_KSTKBAS+005E4
   Return PC            7FFE7764  80000014      EXE$QIOW_3+00004
Align Stack by 0 Bytes =>
SDA> sho call/n
Call Frame Information
----------------------
        Call Frame Generated by CALLG Instruction
Condition Handler       7FFE77E4  00C00008      PSL$M_PRVMOD+00008
SP Align Bits = 00      7FFE77E8  00000000
   Saved  AP            7FFE77EC  7FE76944
   Saved  FP            7FFE77F0  7FE76908
   Return PC            7FFE77F4  838AE858      EXE$EXCEPTION+00227
Align Stack by 0 Bytes =>
SDA>
Call Frame Information
----------------------
        Call Frame Generated by CALLS Instruction

Condition Handler       7FE76908  00000000
SP Align Bits = 00      7FE7690C  2FFC0000
   Saved  AP            7FE76910  7FE7699C
   Saved  FP            7FE76914  7FE76968
   Return PC            7FE76918  00006C41      DAP$K_CRC_TBL7+0029F
        R2              7FE7691C  00076210
        R3              7FE76920  00000001
        R4              7FE76924  00000000
        R5              7FE76928  00000001
        R6              7FE7692C  0000127C      VCRP$K_EC_DLL_LAST+00071
        R7              7FE76930  7FE769A0
        R8              7FE76934  00005B30      DAP$K_CRC_TBL1+0074D
        R9              7FE76938  00000000
        R10             7FE7693C  00005690      DAP$K_CRC_TBL1+002AD
        R11             7FE76940  00005590      DAP$K_CRC_TBL1+001AD
Align Stack by 0 Bytes =>
Argument List           7FE76944  00000007
Call Frame Information
----------------------
                        7FE76948  000000F0
                        7FE7694C  00000000
                        7FE76950  00000000
                        7FE76954  0008C204
                        7FE76958  00000001
                        7FE7695C  00000000
                        7FE76960  00000000

Process index: 0095   Name: IAN   Extended PID: 21600A95
--------------------------------------------------------
                            Process activated images
                            ------------------------
ICB       Start     End       Type            Image Name  Major ID,Minor ID
--------  --------  --------  --------------  -----------------------------
7FF966B8  00000200  000077FF  MAIN            SSU  0,0
7FF96798  00052800  0006A7FF  GLOBAL     SHR  VAXCRTL  4,3
7FF979C0  0006A800  0006AFFF  GLOBAL          CMA$TIS_SHR  1,1
7FF978E0  00008200  000325FF  GLOBAL     SHR  MTHRTL  129,32780
7FF96F00  00032600  000527FF  GLOBAL     SHR  LIBRTL  1,14
7FF97950  00007800  000081FF  GLOBAL PRT SHR  PTD$SERVICES_SHR  1,0

Total images = 6                Pages allocated = 855
SDA>
--------------------------------------------------------------------------------
Log No             24056.00-4C6-1UVO           Desc type      TS
Sequence no        02                          Authr badge no 231847
                                               Creation D/T    2-JAN-1995 21:42
--------------------------------------------------------------------------------
--------------------------------------------------------------------------------
Log No             24056.00-4C6-1UVO           Desc type      TS
Sequence no        03                          Authr badge no 231847
                                               Creation D/T    3-JAN-1995 10:54
--------------------------------------------------------------

                     <<< Note 199.1 by COMICS::GLEDHILL >>>
                        -< rocol watch induced crash. >-

From:	NICES::GLEDHILL      4-JAN-1995 01:14:49.52
To:	GLEDHILL
CC:	GLEDHILL
Subj:	 

*****  THIS IS A READ ONLY COPY FROM NICE   -   please handle accordingly  *****
********************************************************************************

Log No            24056.00-4C6-1UVO          Queue      GLEDHILL       
Log D/T           20-DEC-1994 17:04          Owner      GLEDHILL                 
LSDT D/T           3-JAN-1995 09:00          Loc/Phone  UVO  3245
Status as at       4-JAN-1995 01:15 is OPEN                                    
EXT REQ Stat Code             Escalation Indicator  Y
Hold Indicator    N           Planned Indicator     Y

---------------------------------Customer---------------------------------------
Company          ROCOL LTD                                                      
Department       ROCOL HOUSE                             
Street           SWILLINGTON                             
City             LEEDS                         
Postal Code      LS26 8BS                    PO No                          

Caller           JOHN ELLIOT                 Title      MR             
Phone            0532 866511                 Extension  ASK   
Service Wish     ** dump on ryrgrs::$4$DIA1:[24056_ROCOL]SYSDUMP.DMP;1 **       

--------------------------------------------------------------------------------
******************************INITIAL CALL DETAILS******************************
UPDATED BY / DATE-TIME        :: Theo Bartley  Wed 21-Dec-1994::09:56
PROBLEM DESCRIPTION / CONFIG  :: System bugchecked
ACTION PLAN                   :: Look at dump.
NEXT UPDATE TIME & DATE       :: Cust's modem is in use.  He will call back
                                 later when it becomes free
   ******************************   Update   ******************************
UPDATED BY /DATE-TIME         :: Theo Bartley / Wed 21-Dec-1994::11:13
CURRENT STATUS                :: Customer is unable to give us access to the
                                 system to look at the dump online.  He will
                                 send the dump in on tape.  No errors in the
                                 errorlog, so customer won't include errorlog
                                 on the tape.  Customer ran WATCH just before
                                 system crashed.
ACTION PLAN                   :: Analyse dump.
NEXT UPDATE TIME & DATE       :: Call cust if tape hasn't arrived by Fri 12:00

   ******************************   Update   ******************************
UPDATED BY /DATE-TIME         :: MIKE AYLING  23-DEC-1994  10:10
CURRENT STATUS                :: TK50 TAPE RECEIVED
ACTION PLAN                   :: TO BE ANALYSED BY SPECIALIST WHEN AVAILABLE
NEXT UPDATE TIME & DATE       :: 

   ******************************   Update   ******************************
UPDATED BY /DATE-TIME         :: Norm Pettet 23-Dec-1994:16:58
CURRENT STATUS                :: dump copied to ryrgrs::$4$DIA1:[24056_ROCOL]
                                 Directory $4$DIA1:[24056_ROCOL]

                                 SYSDUMP.DMP;1          90892

                                 Total of 1 file, 90892 blocks.

ACTION PLAN                   :: 
NEXT UPDATE TIME & DATE       :: 


   ******************************   Update   ******************************
UPDATED BY /DATE-TIME         :: Norm Pettet
CURRENT STATUS                :: process crash - lloks like a process run-down
                                 problem - further re-work required
ACTION PLAN                   :: 
NEXT UPDATE TIME & DATE       :: 

   ******************************   Update   ******************************
UPDATED BY /DATE-TIME         :: Norm Pettet 31-Dec-1994
CURRENT STATUS                :: Reading the supplied information from the
                                 customer, it appears he was using DEBUG and
                                 was exiting when user IAN had a problem. User
                                 IAN logged in again then customer, using WATCH
                                 from DECUS, monitored his process. It was then
                                 that the system crashed. I can't find however 
				 the initial login from user IAN.
ACTION PLAN                   :: Suggest discussing with customer - if this is
                                 the first time this crash has occured then
                                 wait for it to happen again (if ever)

--------------------------------------------------------------------------------
--------------------------------------------------------------------------------
canasta description.


 Status:          UNIDENTIFIED                        
 VMS Version:     6.1                          More Admin Reference     
 CPU Type:        3100-95                             
 Bugcheck:        SSRVEXCEPT                                                  
 Process Name:    IAN                                                         
 Image Name:      SSU                                                         
 SID Register:    13000202                                                    
 Argument Count:  00000005                                                    
 Condition Code:  0000000C                                                    
 Reason Mask:     00000001                                                    
 Virtual Address: 00C00008                                                    
 Exception PC:    80000010                                                    
 Exception PSL:   00000009                                                    
 Module:          EXE$QIOW_3                                                  
 Offset:          00000000                                                    
 Instruction:     CALLG                                                       

199.3Look in the vector tablesCOMICS::GLEDHILLWed Jan 04 1995 01:36307
------------------
This is an interesting one (thanks roland!) I think the clue is in the info
the customer sent in saying he was using watch from decus, but thought I would
go thru it make sure anyway!.

here is from the process. 

SDA> SHOW CALL
Call Frame Information
----------------------
	Call Frame Generated by CALLG Instruction

Condition Handler	7FFE7754  00000000	
SP Align Bits = 00	7FFE7758  00000000
   Saved  AP		7FFE775C  7FE76944	
   Saved  FP		7FFE7760  7FFE77E4	CTL$GL_KSTKBAS+005E4
   Return PC		7FFE7764  80000014	EXE$QIOW_3+00004
Align Stack by 0 Bytes =>

SDA> SHOW CALL/NEXT
Call Frame Information
----------------------
	Call Frame Generated by CALLG Instruction

Condition Handler	7FFE77E4  00C00008	
SP Align Bits = 00	7FFE77E8  00000000
   Saved  AP		7FFE77EC  7FE76944	
   Saved  FP		7FFE77F0  7FE76908	
   Return PC		7FFE77F4  838AE858	EXCEPTION+00458
Align Stack by 0 Bytes =>
** the above is on the kernel stack, the below on the user stack, this is
the call frame that took us into K mode.

SDA> SHOW CALL/NEXT
Call Frame Information
----------------------
	Call Frame Generated by CALLS Instruction

Condition Handler	7FE76908  00000000	
SP Align Bits = 00	7FE7690C  2FFC0000
   Saved  AP		7FE76910  7FE7699C	
   Saved  FP		7FE76914  7FE76968	
   Return PC		7FE76918  00006C41	**
etc 
    Press RETURN for more.
SDA> EX/INST 6C41
00006C41:  CMPL    R0,#01
SDA> EX/INST 6C41-20;20
....
00006C37:  PUSHL   04(R2)
00006C41:  CMPL    R0,#01
00006C3A:  CALLS   #07,@000074A8 <-
SDA> EX 74A8
000074A8:  00007828   "(x.."


SDA> ex/inst 7828
00007828:  XFC     
SDA> ex/inst
SDA> ex/inst
0000782A:  CHMK    #4249
** this wont be a normal system service, most likely a user written or vms
supplied one in a privileged image (checking black book vms codes are 0 - 255, 
the vms supplied privileged higher than this, the -ve ones reserved for
customer use) 

 ** show proc/image show this is in ptd$services, ie the psuedo terminal driver
thing. 
                            Process activated images
                            ------------------------

ICB       Start     End       Type            Image Name  Major ID,Minor ID
--------  --------  --------  --------------  -----------------------------
7FF96798  00000200  000077FF  MAIN            SSU  0,0
7FF96F00  00052800  0006A7FF  GLOBAL     SHR  VAXCRTL  4,3
7FF97840  0006A800  0006AFFF  GLOBAL          CMA$TIS_SHR  1,1
7FF978B0  00008200  000325FF  GLOBAL     SHR  MTHRTL  129,32780
7FF97920  00032600  000527FF  GLOBAL     SHR  LIBRTL  1,14
7FF966B8  00007800  000081FF  GLOBAL PRT SHR  PTD$SERVICES_SHR  1,0

(actually checking the listings shows that this is the transfer vector for
ptd$write, see below for listings).

To check that this is a valid  privileged image we need to check the P1 space
stuff. I did this and all seems in order, comparing this with another process
running the same images, they are the same.

IE 
SDA> SHOW SYM/ALL CTL$GL_USR
...
CTL$GL_USRCHMK                  = 7FFEFF18 : 7FFE5804    

This is 
CTL$A_DISPVEC+00004:  JSB     @#00007863 - this is in ptd$services as we would 
CTL$A_DISPVEC+0000A:  RSB                  have expected.

SDA> SH PROC/CHAN
Process index: 0095   Name: IAN   Extended PID: 21600A95
--------------------------------------------------------

* ok so now have established that this is meant to be and we are calling
a valid (dec supplied) system service.  Looking at the listings of ptd$write

(this corrresponds to our 7828)
0028   209 	.ALIGN	QUAD
0028   210 	.TRANSFER	PTD$WRITE		; Define routine public name
0028   211 	.MASK		PTD$WRITE_K		; Define entry mask
002A   212 	CHMK	#PTD$WRITE_NUMBER		; Change to Kernel mode and execute
and here is the code that this calls.

01E3   729 PTD$WRITE::
01E3   730 	.ENTRY	PTD$WRITE_K,		-	; Build read request
01E5   731 		^M<R2,R3,R4,R5,R6,R7,R8,R9,R10,R11>
01E5   732 	MOVZWL	WRITE_CHAN(AP),R0		; Get channel number
01E9   733 	BSBW	PTD$VALIDATE_CHAN		; Validate channel number
01EC   734 	MOVL	WRITE_AST(AP),R0		; Get write AST address
01F0   735 	BEQL	10$				; EQL no AST
01F2   736 	MOVL	WRITE_ASTPRM(AP),R1		; Get AST parameter
01F6   737 	MNEGL	#1,R6				; No EFN
01F9   738 	BSBW	PTD$BUILD_ACB			; Build an ACB
01FC   739 						; (raises IPL to IPL$_ASTDEL)
01FC   740 	MOVL	R3,R11				; Copy ACB address
01FF   741 	BRB	20$				; Continue
0201   742 10$:
0201   743 	CLRL	R11				; No AST to deliver
0203   744 
0203   745 20$:
0203   746 	MOVL	WRITE_BUF(AP),R0		; Get write buffer 
0207   747 	MOVZWL	WRITE_BUF_LEN(AP),R1		; Get write buffer size
020B   748 	BSBW	PTD$VALIDATE_BUFF		; Validate buffer
020E   749 	BLBC	R0,40$				; LBC error handle it
0211   750 	MOVL	R1,R7				; Copy write buffer size
0214   751 	CLRL	(R2)		  		; Clear out I/O status longword
0216   752 	MOVL	R2,R6				; Copy system address of first character to write
0219   753 	CLRL	R9				; Assume no echo buffer
021B   754 	MOVL	WRITE_ECHOBUF(AP),R0		; Get echo buffer address
021F   755 	BEQL	30$				; EQL no echo buffer 
0221   756 	MOVZWL	WRITE_ECHOBUF_LEN(AP),R1	; Get Echo buffer size
0225   757 	BSBW	PTD$VALIDATE_BUFF		; Validate Echo buffer
0228   758 	BLBC	R0,40$				; LBC error exit
022B   759 	CLRL	(R2)				; Clear I/O status longword in echo buffer
022D   760 	MOVL	R2,R9				; Copy system address of first
0230   761 						; character position in echo buffer
0230   762 	MOVL	R1,R10				; Copy echo buffer size
0233   763 
0233   764 30$:
0233   765 	MOVL	UCB$L_TT_PORT(R5),R0		; Get port vector address
0238   766 	JMP	@PORT_FT_WRITE(R0)		; Jump to driver to finish request

**I think we crashed after here, if we check the stack we see that the saved r0 
contains  the ucb$l_tt_port address. Also note tht there are two signal
arrays on the stack (the lower one is the relevant one) 
--------------------------------------------------------------------------------
for more details ....
Now this process is running ssu, the session support utility, this runs a
process that presents a split screen interace to the user, so that they can
run 2 programs on one screen. The ssu process communicates to the programs via
these peuedo terminal devices.

                            Process active channels
                            -----------------------

Channel	 Window		  Status	Device/file accessed
-------	 ------		  ------	--------------------
  0010	00000000	          	DKA100:
  0020	8422CE40	          	DKA100:(9074,1,0)
  0030	840C6A40	          	DKA100:(3664,4,0)
  0040	00000000	          	VTA230:
  0050	840C35C0	          	DKA100:(193,1,0) (section file)
  0060	00000000	          	VTA230:
  0070	840BE700	          	DKA100:(344,1,0) (section file)
  0080	840D6640	          	DKA100:(171,1,0) (section file)
  0090	840C1280	          	DKA100:(282,1,0) (section file)
  00A0	840C6E80	          	DKA100:(187,1,0) (section file)
  00B0	840C38C0	          	DKA100:(3705,2,0) (section file)
  00C0	840C7600	          	DKA100:(138,1,0) (section file)
  00D0	00000000	          	VTA230:
  00E0	00000000	          	VTA230:
  00F0	00000000	          	FTA295:
ie the vta devices are used to talke to the physical terminal - the fta
devices to the application(s). 

However this process is being watched by watch and this watches other processes 
by revectoring the read + write etc routines into their own routines first so
that they can log what the users are doing. These routines are stored in
the ucb of the device and in class and vector tables. 

What it actually does is clone new port and class vector tables, putting most 
of the contents the same as in the original, modifying the ones it wants to 
revector, plus modify the get/put next routines in the ucb.

Looking at fta295s UCB
842FBBE8   UCB$L_TT_GETNXT                 84424415	
842FBBEC   UCB$L_TT_PUTNXT                 8442442A	
842FBBF0   UCB$L_TT_CLASS                  84423664	
842FBBF4   UCB$L_TT_PORT                   84423618	


this is code in nonpaged pool, comparing with other ftdevices which 
have these addresses in this (ft) or ttdriver 

84392368   UCB$L_TT_GETNXT                 843E9170	TTDRIVER+02BB0
8439236C   UCB$L_TT_PUTNXT                 843E8DD8	TTDRIVER+02818
84392370   UCB$L_TT_CLASS                  843E6E50	DDT+0010F
84392374   UCB$L_TT_PORT                   84355FAC	FTDRIVER+0016C

So in ptd$write we jump to 54 hex of the port vect, looking at this we see

SDA> eval  @(84423618 + 54)
Hex = 843EB315   Decimal = -2076265707		TTDRIVER+04D55
SDA> ex/inst ttdriver + 4d55;50
TTDRIVER+04D55:  PUSHL   R0
TTDRIVER+04D57:  MOVL    00D4(R5),R0
TTDRIVER+04D5C:  BEQL    TTDRIVER+04D67 (ucb$l_tt_logucb <> 0 so don't branch)
TTDRIVER+04D5E:  TSTW    70(R0)  (ucb$w_refc = 3 so don't branch)
TTDRIVER+04D61:  BEQL    TTDRIVER+04D67
TTDRIVER+04D63:  MOVL    (SP)+,R0
TTDRIVER+04D66:  RSB     

So we would have gone thru this and rsbed, plus r0 will end up with what we
started with. So we will rsb, but to what, remember we went into here with a
jmp and left with a rsb. 

Now on entry to a system service (see chaper 6 in internal book i think)
we have a standard 5 argument call frame(may have extra stuff on stack for a 
non-standard system service, but checking the ptd dispatcher we see that it
clears the stack (movl fp,sp) so we will rsb to the top thing on the stack 
which wil be the condition handler = 0. Checking the stack. (didn't bother to
see why the 2nd exception occurred)


	 SP =>  7FFE7754  00000000	
		7FFE7758  00000000	
		7FFE775C  7FE76944	
		7FFE7760  7FFE77E4	CTL$GL_KSTKBAS+005E4
		7FFE7764  80000014	EXE$QIOW_3+00004
		7FFE7768  838B0AD1	EXCEPTION+026D1
		7FFE776C  00000002	
		7FFE7770  7FFE7790	CTL$GL_KSTKBAS+00590
		7FFE7774  7FFE7778	CTL$GL_KSTKBAS+00578
		7FFE7778  00000004	
		7FFE777C  7FFE77E4	CTL$GL_KSTKBAS+005E4
		7FFE7780  FFFFFFFD	LKB$K_SCSWAIT
		7FFE7784  00000000	
		7FFE7788  00C00008	PSL$M_PRVMOD+00008
		7FFE778C  0000000A	
		7FFE7790  00000005	
		7FFE7794  0000000C	
		7FFE7798  00000001	
		7FFE779C  00C00008	PSL$M_PRVMOD+00008
		7FFE77A0  80000010	EXE$QIOW_3
		7FFE77A4  00000009	
		7FFE77A8  838B0A4D	EXCEPTION+0264D

		7FFE77AC  00000002	-> call frame of original exception
		7FFE77B0  7FFE77D0	CTL$GL_KSTKBAS+005D0
		7FFE77B4  7FFE77B8	CTL$GL_KSTKBAS+005B8

		7FFE77B8  00000004	- > mech array
		7FFE77BC  7FFE77E4	CTL$GL_KSTKBAS+005E4
		7FFE77C0  00000000	
		7FFE77C4  84423618	- > ro = ucb$l_tt_port
		7FFE77C8  00000001	- > r1
		7FFE77CC  05000001	
	
 		7FFE77D0  00000005	-> sig array
		7FFE77D4  0000000C	   accvio
		7FFE77D8  00000000	   rm
		7FFE77DC  00000000	   va
		7FFE77E0  00000000	   PC
top of ss-->    7FFE77E4  00C00008	PSL$M_PRVMOD+00008 PSL
call frame     
should be here
 		7FFE77E8  00000000	register save mask
		7FFE77EC  7FE76944	saved ap
		7FFE77F0  7FE76908	saved fp
		7FFE77F4  838AE858	EXCEPTION+00458 (return pc - ie 
					sys_service exit)
							 

exception saved PC/PSL from cmkl
		7FFE77F8  0000782E	saved PC
		7FFE77FC  03C00000	saved PSL

So we crash due to rsbing to 0. Only question that remains is what was wrong
in the prot and class vector table. I think the reasons for this is that the
port/class  vector of the ptd devices are different to the standard tt ones.
In particular ptd is larger than that for the standard terminal device. 
Watch doesn't realise this, and allocates new port class vectors of the
standard size, ofset 54 in the port table goes past the end of the standard one
so we get jumping to an irelevent address. The actual adress we dump to is
the class_setup ucb address within the class table which is adjacent (higher) 
in memory to the port table. 

ie port+ 54 is 8442367c which is class + 08 which corresponds to the 
class_setup_ucb routine
 
842FBBF0   UCB$L_TT_CLASS                  84423664	
842FBBF4   UCB$L_TT_PORT                   84423618	

Must assume from this that it is not safe to use watch on any psuedo-terminal
device. This would include any process communicating via ssu but also perhaps
workstations processes (as these use ptd devices).PS don't think watch  was used
on this processes, but on the client process (index 96), ie on the other end of 
the psuedo terminal.