2                          Debugging on Linux for s/390 & z/Architecture
   3                                       by
   4                Denis Joseph Barrow (,
   5                Copyright (C) 2000-2001 IBM Deutschland Entwicklung GmbH, IBM Corporation
   6                              Best viewed with fixed width fonts 
   8Overview of Document:
  10This document is intended to give a good overview of how to debug
  11Linux for s/390 & z/Architecture. It isn't intended as a complete reference & not a
  12tutorial on the fundamentals of C & assembly. It doesn't go into
  13390 IO in any detail. It is intended to complement the documents in the
  14reference section below & any other worthwhile references you get.
  16It is intended like the Enterprise Systems Architecture/390 Reference Summary
  17to be printed out & used as a quick cheat sheet self help style reference when
  18problems occur.
  22Register Set
  23Address Spaces on Intel Linux
  24Address Spaces on Linux for s/390 & z/Architecture
  25The Linux for s/390 & z/Architecture Kernel Task Structure
  26Register Usage & Stackframes on Linux for s/390 & z/Architecture
  27A sample program with comments
  28Compiling programs for debugging on Linux for s/390 & z/Architecture
  29Figuring out gcc compile errors
  30Debugging Tools
  33Performance Debugging 
  34Debugging under VM
  35s/390 & z/Architecture IO Overview
  36Debugging IO on s/390 & z/Architecture under VM
  37GDB on s/390 & z/Architecture
  38Stack chaining in gdb by hand
  39Examining core dumps
  41Debugging modules
  42The proc file system
  43Starting points for debugging scripting languages etc.
  44Dumptool & Lcrash
  47Special Thanks
  49Register Set
  51The current architectures have the following registers.
  5316  General propose registers, 32 bit on s/390 64 bit on z/Architecture, r0-r15 or gpr0-gpr15 used for arithmetic & addressing. 
  5516 Control registers, 32 bit on s/390 64 bit on z/Architecture, ( cr0-cr15 kernel usage only ) used for memory management,
  56interrupt control,debugging control etc.
  5816 Access registers ( ar0-ar15 ) 32 bit on s/390 & z/Architecture
  59not used by normal programs but potentially could 
  60be used as temporary storage. Their main purpose is their 1 to 1
  61association with general purpose registers and are used in
  62the kernel for copying data between kernel & user address spaces.
  63Access register 0 ( & access register 1 on z/Architecture ( needs 64 bit 
  64pointer ) ) is currently used by the pthread library as a pointer to
  65the current running threads private area.
  6716 64 bit floating point registers (fp0-fp15 ) IEEE & HFP floating 
  68point format compliant on G5 upwards & a Floating point control reg (FPC) 
  694  64 bit registers (fp0,fp2,fp4 & fp6) HFP only on older machines.
  71Linux (currently) always uses IEEE & emulates G5 IEEE format on older machines,
  72( provided the kernel is configured for this ).
  75The PSW is the most important register on the machine it
  76is 64 bit on s/390 & 128 bit on z/Architecture & serves the roles of 
  77a program counter (pc), condition code register,memory space designator.
  78In IBM standard notation I am counting bit 0 as the MSB.
  79It has several advantages over a normal program counter
  80in that you can change address translation & program counter 
  81in a single instruction. To change address translation,
  82e.g. switching address translation off requires that you
  83have a logical=physical mapping for the address you are
  84currently running at.
  86      Bit           Value
  87s/390 z/Architecture
  880       0     Reserved ( must be 0 ) otherwise specification exception occurs.
  901       1     Program Event Recording 1 PER enabled, 
  91              PER is used to facilitate debugging e.g. single stepping.
  932-4    2-4    Reserved ( must be 0 ). 
  955       5     Dynamic address translation 1=DAT on.
  976       6     Input/Output interrupt Mask
  997       7     External interrupt Mask used primarily for interprocessor signalling & 
 100              clock interrupts.
 1028-11  8-11    PSW Key used for complex memory protection mechanism not used under linux
 10412      12    1 on s/390 0 on z/Architecture
 10613      13    Machine Check Mask 1=enable machine check interrupts
 10814      14    Wait State set this to 1 to stop the processor except for interrupts & give 
 109              time to other LPARS used in CPU idle in the kernel to increase overall 
 110              usage of processor resources.
 11215      15    Problem state ( if set to 1 certain instructions are disabled )
 113              all linux user programs run with this bit 1 
 114              ( useful info for debugging under VM ).
 11616-17 16-17   Address Space Control
 118              00 Primary Space Mode when DAT on
 119              The linux kernel currently runs in this mode, CR1 is affiliated with 
 120              this mode & points to the primary segment table origin etc.
 122              01 Access register mode this mode is used in functions to 
 123              copy data between kernel & user space.
 125              10 Secondary space mode not used in linux however CR7 the
 126              register affiliated with this mode is & this & normally
 127              CR13=CR7 to allow us to copy data between kernel & user space.
 128              We do this as follows:
 129              We set ar2 to 0 to designate its
 130              affiliated gpr ( gpr2 )to point to primary=kernel space.
 131              We set ar4 to 1 to designate its
 132              affiliated gpr ( gpr4 ) to point to secondary=home=user space
 133              & then essentially do a memcopy(gpr2,gpr4,size) to
 134              copy data between the address spaces, the reason we use home space for the
 135              kernel & don't keep secondary space free is that code will not run in 
 136              secondary space.
 138              11 Home Space Mode all user programs run in this mode.
 139              it is affiliated with CR13.
 14118-19 18-19   Condition codes (CC)
 14320    20      Fixed point overflow mask if 1=FPU exceptions for this event 
 144              occur ( normally 0 ) 
 14621    21      Decimal overflow mask if 1=FPU exceptions for this event occur 
 147              ( normally 0 )
 14922    22      Exponent underflow mask if 1=FPU exceptions for this event occur 
 150              ( normally 0 )
 15223    23      Significance Mask if 1=FPU exceptions for this event occur 
 153              ( normally 0 )
 15524-31 24-30   Reserved Must be 0.
 157      31      Extended Addressing Mode
 158      32      Basic Addressing Mode
 159              Used to set addressing mode
 160              PSW 31   PSW 32
 161                0         0        24 bit
 162                0         1        31 bit
 163                1         1        64 bit
 16532             1=31 bit addressing mode 0=24 bit addressing mode (for backward 
 166               compatibility), linux always runs with this bit set to 1
 16833-64          Instruction address.
 169      33-63    Reserved must be 0
 170      64-127   Address
 171               In 24 bits mode bits 64-103=0 bits 104-127 Address 
 172               In 31 bits mode bits 64-96=0 bits 97-127 Address
 173               Note: unlike 31 bit mode on s/390 bit 96 must be zero
 174               when loading the address with LPSWE otherwise a 
 175               specification exception occurs, LPSW is fully backward
 176               compatible.
 179Prefix Page(s)
 181This per cpu memory area is too intimately tied to the processor not to mention.
 182It exists between the real addresses 0-4096 on s/390 & 0-8192 z/Architecture & is exchanged 
 183with a 1 page on s/390 or 2 pages on z/Architecture in absolute storage by the set 
 184prefix instruction in linux'es startup. 
 185This page is mapped to a different prefix for each processor in an SMP configuration
 186( assuming the os designer is sane of course :-) ).
 187Bytes 0-512 ( 200 hex ) on s/390 & 0-512,4096-4544,4604-5119 currently on z/Architecture 
 188are used by the processor itself for holding such information as exception indications & 
 189entry points for exceptions.
 190Bytes after 0xc00 hex are used by linux for per processor globals on s/390 & z/Architecture 
 191( there is a gap on z/Architecture too currently between 0xc00 & 1000 which linux uses ).
 192The closest thing to this on traditional architectures is the interrupt
 193vector table. This is a good thing & does simplify some of the kernel coding
 194however it means that we now cannot catch stray NULL pointers in the
 195kernel without hard coded checks.
 199Address Spaces on Intel Linux
 202The traditional Intel Linux is approximately mapped as follows forgive
 203the ascii art.
 2040xFFFFFFFF 4GB Himem                        *****************
 205                                            *               *
 206                                            * Kernel Space  *
 207                                            *               *
 208                                            *****************          ****************
 209User Space Himem (typically 0xC0000000 3GB )*  User Stack   *          *              *
 210                                            *****************          *              *
 211                                            *  Shared Libs  *          * Next Process *          
 212                                            *****************          *     to       *  
 213                                            *               *    <==   *     Run      *  <==
 214                                            *  User Program *          *              *
 215                                            *   Data BSS    *          *              *
 216                                            *    Text       *          *              *
 217                                            *   Sections    *          *              *
 2180x00000000                                  *****************          ****************
 220Now it is easy to see that on Intel it is quite easy to recognise a kernel address 
 221as being one greater than user space himem ( in this case 0xC0000000).
 222& addresses of less than this are the ones in the current running program on this
 223processor ( if an smp box ).
 224If using the virtual machine ( VM ) as a debugger it is quite difficult to
 225know which user process is running as the address space you are looking at
 226could be from any process in the run queue.
 228The limitation of Intels addressing technique is that the linux
 229kernel uses a very simple real address to virtual addressing technique
 230of Real Address=Virtual Address-User Space Himem.
 231This means that on Intel the kernel linux can typically only address
 232Himem=0xFFFFFFFF-0xC0000000=1GB & this is all the RAM these machines
 233can typically use.
 234They can lower User Himem to 2GB or lower & thus be
 235able to use 2GB of RAM however this shrinks the maximum size
 236of User Space from 3GB to 2GB they have a no win limit of 4GB unless
 237they go to 64 Bit.
 240On 390 our limitations & strengths make us slightly different.
 241For backward compatibility we are only allowed use 31 bits (2GB)
 242of our 32 bit addresses, however, we use entirely separate address 
 243spaces for the user & kernel.
 245This means we can support 2GB of non Extended RAM on s/390, & more
 246with the Extended memory management swap device & 
 247currently 4TB of physical memory currently on z/Architecture.
 250Address Spaces on Linux for s/390 & z/Architecture
 253Our addressing scheme is as follows
 256Himem 0x7fffffff 2GB on s/390    *****************          ****************
 257currently 0x3ffffffffff (2^42)-1 *  User Stack   *          *              *
 258on z/Architecture.               *****************          *              *
 259                                 *  Shared Libs  *          *              *      
 260                                 *****************          *              *  
 261                                 *               *          *    Kernel    *  
 262                                 *  User Program *          *              *
 263                                 *   Data BSS    *          *              *
 264                                 *    Text       *          *              *
 265                                 *   Sections    *          *              *
 2660x00000000                       *****************          ****************
 268This also means that we need to look at the PSW problem state bit
 269or the addressing mode to decide whether we are looking at
 270user or kernel space.
 272Virtual Addresses on s/390 & z/Architecture
 275A virtual address on s/390 is made up of 3 parts
 276The SX ( segment index, roughly corresponding to the PGD & PMD in linux terminology ) 
 277being bits 1-11.
 278The PX ( page index, corresponding to the page table entry (pte) in linux terminology )
 279being bits 12-19. 
 280The remaining bits BX (the byte index are the offset in the page )
 281i.e. bits 20 to 31.
 283On z/Architecture in linux we currently make up an address from 4 parts.
 284The region index bits (RX) 0-32 we currently use bits 22-32
 285The segment index (SX) being bits 33-43
 286The page index (PX) being bits  44-51
 287The byte index (BX) being bits  52-63
 2901) s/390 has no PMD so the PMD is really the PGD also.
 291A lot of this stuff is defined in pgtable.h.
 2932) Also seeing as s/390's page indexes are only 1k  in size 
 294(bits 12-19 x 4 bytes per pte ) we use 1 ( page 4k )
 295to make the best use of memory by updating 4 segment indices 
 296entries each time we mess with a PMD & use offsets 
 2970,1024,2048 & 3072 in this page as for our segment indexes.
 298On z/Architecture our page indexes are now 2k in size
 299( bits 12-19 x 8 bytes per pte ) we do a similar trick
 300but only mess with 2 segment indices each time we mess with
 301a PMD.
 3033) As z/Architecture supports up to a massive 5-level page table lookup we
 304can only use 3 currently on Linux ( as this is all the generic kernel
 305currently supports ) however this may change in future
 306this allows us to access ( according to my sums )
 3074TB of virtual storage per process i.e.
 3084096*512(PTES)*1024(PMDS)*2048(PGD) = 4398046511104 bytes,
 309enough for another 2 or 3 of years I think :-).
 310to do this we use a region-third-table designation type in
 311our address space control registers.
 314The Linux for s/390 & z/Architecture Kernel Task Structure
 316Each process/thread under Linux for S390 has its own kernel task_struct
 317defined in linux/include/linux/sched.h
 318The S390 on initialisation & resuming of a process on a cpu sets
 319the __LC_KERNEL_STACK variable in the spare prefix area for this cpu
 320(which we use for per-processor globals).
 322The kernel stack pointer is intimately tied with the task structure for
 323each processor as follows.
 325                      s/390
 326            ************************
 327            *  1 page kernel stack *
 328            *        ( 4K )        *
 329            ************************
 330            *   1 page task_struct *        
 331            *        ( 4K )        *
 3328K aligned  ************************ 
 334                 z/Architecture
 335            ************************
 336            *  2 page kernel stack *
 337            *        ( 8K )        *
 338            ************************
 339            *  2 page task_struct  *        
 340            *        ( 8K )        *
 34116K aligned ************************ 
 343What this means is that we don't need to dedicate any register or global variable
 344to point to the current running process & can retrieve it with the following
 345very simple construct for s/390 & one very similar for z/Architecture.
 347static inline struct task_struct * get_current(void)
 349        struct task_struct *current;
 350        __asm__("lhi   %0,-8192\n\t"
 351                "nr    %0,15"
 352                : "=r" (current) );
 353        return current;
 356i.e. just anding the current kernel stack pointer with the mask -8192.
 357Thankfully because Linux doesn't have support for nested IO interrupts
 358& our devices have large buffers can survive interrupts being shut for 
 359short amounts of time we don't need a separate stack for interrupts.
 364Register Usage & Stackframes on Linux for s/390 & z/Architecture
 368This is the code that gcc produces at the top & the bottom of
 369each function. It usually is fairly consistent & similar from 
 370function to function & if you know its layout you can probably
 371make some headway in finding the ultimate cause of a problem
 372after a crash without a source level debugger.
 374Note: To follow stackframes requires a knowledge of C or Pascal &
 375limited knowledge of one assembly language.
 377It should be noted that there are some differences between the
 378s/390 & z/Architecture stack layouts as the z/Architecture stack layout didn't have
 379to maintain compatibility with older linkage formats.
 384This is a built in compiler function for runtime allocation
 385of extra space on the callers stack which is obviously freed
 386up on function exit ( e.g. the caller may choose to allocate nothing
 387of a buffer of 4k if required for temporary purposes ), it generates 
 388very efficient code ( a few cycles  ) when compared to alternatives 
 389like malloc.
 391automatics: These are local variables on the stack,
 392i.e they aren't in registers & they aren't static.
 395This is a pointer to the stack pointer before entering a
 396framed functions ( see frameless function ) prologue got by 
 397dereferencing the address of the current stack pointer,
 398 i.e. got by accessing the 32 bit value at the stack pointers
 399current location.
 402This is a pointer to the back of the literal pool which
 403is an area just behind each procedure used to store constants
 404in each function.
 406call-clobbered: The caller probably needs to save these registers if there 
 407is something of value in them, on the stack or elsewhere before making a 
 408call to another procedure so that it can restore it later.
 411The code generated by the compiler to return to the caller.
 414A frameless function in Linux for s390 & z/Architecture is one which doesn't 
 415need more than the register save area ( 96 bytes on s/390, 160 on z/Architecture )
 416given to it by the caller.
 417A frameless function never:
 4181) Sets up a back chain.
 4192) Calls alloca.
 4203) Calls other normal functions
 4214) Has automatics.
 424This is a pointer to the global-offset-table in ELF
 425( Executable Linkable Format, Linux'es most common executable format ),
 426all globals & shared library objects are found using this pointer.
 429ELF shared libraries are typically only loaded when routines in the shared
 430library are actually first called at runtime. This is lazy binding.
 433This is a table found from the GOT which contains pointers to routines
 434in other shared libraries which can't be called to by easier means.
 437The code generated by the compiler to set up the stack frame.
 440This is extra area allocated on the stack of the calling function if the
 441parameters for the callee's cannot all be put in registers, the same
 442area can be reused by each function the caller calls.
 445A COFF  executable format based concept of a procedure reference 
 446actually being 8 bytes or more as opposed to a simple pointer to the routine.
 447This is typically defined as follows
 448Routine Descriptor offset 0=Pointer to Function
 449Routine Descriptor offset 4=Pointer to Table of Contents
 450The table of contents/TOC is roughly equivalent to a GOT pointer.
 451& it means that shared libraries etc. can be shared between several
 452environments each with their own TOC.
 455static-chain: This is used in nested functions a concept adopted from pascal 
 456by gcc not used in ansi C or C++ ( although quite useful ), basically it
 457is a pointer used to reference local variables of enclosing functions.
 458You might come across this stuff once or twice in your lifetime.
 461The function below should return 11 though gcc may get upset & toss warnings 
 462about unused variables.
 463int FunctionA(int a)
 465        int b;
 466        FunctionC(int c)
 467        {
 468                b=c+1;
 469        }
 470        FunctionC(10);
 471        return(b);
 475s/390 & z/Architecture Register usage
 477r0       used by syscalls/assembly                  call-clobbered
 478r1       used by syscalls/assembly                  call-clobbered
 479r2       argument 0 / return value 0                call-clobbered
 480r3       argument 1 / return value 1 (if long long) call-clobbered
 481r4       argument 2                                 call-clobbered
 482r5       argument 3                                 call-clobbered
 483r6       argument 4                                 saved
 484r7       pointer-to arguments 5 to ...              saved      
 485r8       this & that                                saved
 486r9       this & that                                saved
 487r10      static-chain ( if nested function )        saved
 488r11      frame-pointer ( if function used alloca )  saved
 489r12      got-pointer                                saved
 490r13      base-pointer                               saved
 491r14      return-address                             saved
 492r15      stack-pointer                              saved
 494f0       argument 0 / return value ( float/double ) call-clobbered
 495f2       argument 1                                 call-clobbered
 496f4       z/Architecture argument 2                  saved
 497f6       z/Architecture argument 3                  saved
 498The remaining floating points
 499f1,f3,f5 f7-f15 are call-clobbered.
 5031) The only requirement is that registers which are used
 504by the callee are saved, e.g. the compiler is perfectly
 505capable of using r11 for purposes other than a frame a
 506frame pointer if a frame pointer is not needed.
 5072) In functions with variable arguments e.g. printf the calling procedure 
 508is identical to one without variable arguments & the same number of 
 509parameters. However, the prologue of this function is somewhat more
 510hairy owing to it having to move these parameters to the stack to
 511get va_start, va_arg & va_end to work.
 5123) Access registers are currently unused by gcc but are used in
 513the kernel. Possibilities exist to use them at the moment for
 514temporary storage but it isn't recommended.
 5154) Only 4 of the floating point registers are used for
 516parameter passing as older machines such as G3 only have only 4
 517& it keeps the stack frame compatible with other compilers.
 518However with IEEE floating point emulation under linux on the
 519older machines you are free to use the other 12.
 5205) A long long or double parameter cannot be have the 
 521first 4 bytes in a register & the second four bytes in the 
 522outgoing args area. It must be purely in the outgoing args
 523area if crossing this boundary.
 5246) Floating point parameters are mixed with outgoing args
 525on the outgoing args area in the order the are passed in as parameters.
 5267) Floating point arguments 2 & 3 are saved in the outgoing args area for 
 530Stack Frame Layout
 532s/390     z/Architecture
 5330         0             back chain ( a 0 here signifies end of back chain )
 5344         8             eos ( end of stack, not used on Linux for S390 used in other linkage formats )
 5358         16            glue used in other s/390 linkage formats for saved routine descriptors etc.
 53612        24            glue used in other s/390 linkage formats for saved routine descriptors etc.
 53716        32            scratch area
 53820        40            scratch area
 53924        48            saved r6 of caller function
 54028        56            saved r7 of caller function
 54132        64            saved r8 of caller function
 54236        72            saved r9 of caller function
 54340        80            saved r10 of caller function
 54444        88            saved r11 of caller function
 54548        96            saved r12 of caller function
 54652        104           saved r13 of caller function
 54756        112           saved r14 of caller function
 54860        120           saved r15 of caller function
 54964        128           saved f4 of caller function
 55072        132           saved f6 of caller function
 55180                      undefined
 55296        160           outgoing args passed from caller to callee
 55396+x      160+x         possible stack alignment ( 8 bytes desirable )
 55496+x+y    160+x+y       alloca space of caller ( if used )
 55596+x+y+z  160+x+y+z     automatics of caller ( if used )
 5560                       back-chain
 558A sample program with comments.
 561Comments on the function test
 5631) It didn't need to set up a pointer to the constant pool gpr13 as it isn't used
 564( :-( ).
 5652) This is a frameless function & no stack is bought.
 5663) The compiler was clever enough to recognise that it could return the
 567value in r2 as well as use it for the passed in parameter ( :-) ).
 5684) The basr ( branch relative & save ) trick works as follows the instruction 
 569has a special case with r0,r0 with some instruction operands is understood as 
 570the literal value 0, some risc architectures also do this ). So now
 571we are branching to the next address & the address new program counter is
 572in r13,so now we subtract the size of the function prologue we have executed
 573+ the size of the literal pool to get to the top of the literal pool
 5740040037c int test(int b)
 575{                                                          # Function prologue below
 576  40037c:       90 de f0 34     stm     %r13,%r14,52(%r15) # Save registers r13 & r14
 577  400380:       0d d0           basr    %r13,%r0           # Set up pointer to constant pool using
 578  400382:       a7 da ff fa     ahi     %r13,-6            # basr trick
 579        return(5+b);
 580                                                           # Huge main program
 581  400386:       a7 2a 00 05     ahi     %r2,5              # add 5 to r2
 583                                                           # Function epilogue below 
 584  40038a:       98 de f0 34     lm      %r13,%r14,52(%r15) # restore registers r13 & 14
 585  40038e:       07 fe           br      %r14               # return
 588Comments on the function main
 5901) The compiler did this function optimally ( 8-) )
 592Literal pool for main.
 593400390: ff ff ff ec     .long 0xffffffec
 594main(int argc,char *argv[])
 595{                                                          # Function prologue below
 596  400394:       90 bf f0 2c     stm     %r11,%r15,44(%r15) # Save necessary registers
 597  400398:       18 0f           lr      %r0,%r15           # copy stack pointer to r0
 598  40039a:       a7 fa ff a0     ahi     %r15,-96           # Make area for callee saving 
 599  40039e:       0d d0           basr    %r13,%r0           # Set up r13 to point to
 600  4003a0:       a7 da ff f0     ahi     %r13,-16           # literal pool
 601  4003a4:       50 00 f0 00     st      %r0,0(%r15)        # Save backchain
 603        return(test(5));                                   # Main Program Below
 604  4003a8:       58 e0 d0 00     l       %r14,0(%r13)       # load relative address of test from
 605                                                           # literal pool
 606  4003ac:       a7 28 00 05     lhi     %r2,5              # Set first parameter to 5
 607  4003b0:       4d ee d0 00     bas     %r14,0(%r14,%r13)  # jump to test setting r14 as return
 608                                                           # address using branch & save instruction.
 610                                                           # Function Epilogue below
 611  4003b4:       98 bf f0 8c     lm      %r11,%r15,140(%r15)# Restore necessary registers.
 612  4003b8:       07 fe           br      %r14               # return to do program exit 
 616Compiler updates
 619main(int argc,char *argv[])
 621  4004fc:       90 7f f0 1c             stm     %r7,%r15,28(%r15)
 622  400500:       a7 d5 00 04             bras    %r13,400508 <main+0xc>
 623  400504:       00 40 04 f4             .long   0x004004f4 
 624  # compiler now puts constant pool in code to so it saves an instruction 
 625  400508:       18 0f                   lr      %r0,%r15
 626  40050a:       a7 fa ff a0             ahi     %r15,-96
 627  40050e:       50 00 f0 00             st      %r0,0(%r15)
 628        return(test(5));
 629  400512:       58 10 d0 00             l       %r1,0(%r13)
 630  400516:       a7 28 00 05             lhi     %r2,5
 631  40051a:       0d e1                   basr    %r14,%r1
 632  # compiler adds 1 extra instruction to epilogue this is done to
 633  # avoid processor pipeline stalls owing to data dependencies on g5 &
 634  # above as register 14 in the old code was needed directly after being loaded 
 635  # by the lm   %r11,%r15,140(%r15) for the br %14.
 636  40051c:       58 40 f0 98             l       %r4,152(%r15)
 637  400520:       98 7f f0 7c             lm      %r7,%r15,124(%r15)
 638  400524:       07 f4                   br      %r4
 642Hartmut ( our compiler developer ) also has been threatening to take out the
 643stack backchain in optimised code as this also causes pipeline stalls, you
 644have been warned.
 64664 bit z/Architecture code disassembly
 649If you understand the stuff above you'll understand the stuff
 650below too so I'll avoid repeating myself & just say that 
 651some of the instructions have g's on the end of them to indicate
 652they are 64 bit & the stack offsets are a bigger, 
 653the only other difference you'll find between 32 & 64 bit is that
 654we now use f4 & f6 for floating point arguments on 64 bit.
 65500000000800005b0 <test>:
 656int test(int b)
 658        return(5+b);
 659    800005b0:   a7 2a 00 05             ahi     %r2,5
 660    800005b4:   b9 14 00 22             lgfr    %r2,%r2 # downcast to integer
 661    800005b8:   07 fe                   br      %r14
 662    800005ba:   07 07                   bcr     0,%r7
 66700000000800005bc <main>:
 668main(int argc,char *argv[])
 670    800005bc:   eb bf f0 58 00 24       stmg    %r11,%r15,88(%r15)
 671    800005c2:   b9 04 00 1f             lgr     %r1,%r15
 672    800005c6:   a7 fb ff 60             aghi    %r15,-160
 673    800005ca:   e3 10 f0 00 00 24       stg     %r1,0(%r15)
 674        return(test(5));
 675    800005d0:   a7 29 00 05             lghi    %r2,5
 676    # brasl allows jumps > 64k & is overkill here bras would do fune
 677    800005d4:   c0 e5 ff ff ff ee       brasl   %r14,800005b0 <test> 
 678    800005da:   e3 40 f1 10 00 04       lg      %r4,272(%r15)
 679    800005e0:   eb bf f0 f8 00 04       lmg     %r11,%r15,248(%r15)
 680    800005e6:   07 f4                   br      %r4
 685Compiling programs for debugging on Linux for s/390 & z/Architecture
 687-gdwarf-2 now works it should be considered the default debugging
 688format for s/390 & z/Architecture as it is more reliable for debugging
 689shared libraries,  normal -g debugging works much better now
 690Thanks to the IBM java compiler developers bug reports. 
 692This is typically done adding/appending the flags -g or -gdwarf-2 to the 
 693CFLAGS & LDFLAGS variables Makefile of the program concerned.
 695If using gdb & you would like accurate displays of registers &
 696 stack traces compile without optimisation i.e make sure
 697that there is no -O2 or similar on the CFLAGS line of the Makefile &
 698the emitted gcc commands, obviously this will produce worse code 
 699( not advisable for shipment ) but it is an  aid to the debugging process.
 701This aids debugging because the compiler will copy parameters passed in
 702in registers onto the stack so backtracing & looking at passed in
 703parameters will work, however some larger programs which use inline functions
 704will not compile without optimisation.
 706Debugging with optimisation has since much improved after fixing
 707some bugs, please make sure you are using gdb-5.0 or later developed 
 708after Nov'2000.
 710Figuring out gcc compile errors
 712If you are getting a lot of syntax errors compiling a program & the problem
 713isn't blatantly obvious from the source.
 714It often helps to just preprocess the file, this is done with the -E
 715option in gcc.
 716What this does is that it runs through the very first phase of compilation
 717( compilation in gcc is done in several stages & gcc calls many programs to
 718achieve its end result ) with the -E option gcc just calls the gcc preprocessor (cpp).
 719The c preprocessor does the following, it joins all the files #included together
 720recursively ( #include files can #include other files ) & also the c file you wish to compile.
 721It puts a fully qualified path of the #included files in a comment & it
 722does macro expansion.
 723This is useful for debugging because
 7241) You can double check whether the files you expect to be included are the ones
 725that are being included ( e.g. double check that you aren't going to the i386 asm directory ).
 7262) Check that macro definitions aren't clashing with typedefs,
 7273) Check that definitions aren't being used before they are being included.
 7284) Helps put the line emitting the error under the microscope if it contains macros.
 730For convenience the Linux kernel's makefile will do preprocessing automatically for you
 731by suffixing the file you want built with .i ( instead of .o )
 734from the linux directory type
 735make arch/s390/kernel/signal.i
 736this will build
 738s390-gcc -D__KERNEL__ -I/home1/barrow/linux/include -Wall -Wstrict-prototypes -O2 -fomit-frame-pointer
 739-fno-strict-aliasing -D__SMP__ -pipe -fno-strength-reduce   -E arch/s390/kernel/signal.c
 740> arch/s390/kernel/signal.i  
 742Now look at signal.i you should see something like.
 745# 1 "/home1/barrow/linux/include/asm/types.h" 1
 746typedef unsigned short umode_t;
 747typedef __signed__ char __s8;
 748typedef unsigned char __u8;
 749typedef __signed__ short __s16;
 750typedef unsigned short __u16;
 752If instead you are getting errors further down e.g.
 753unknown instruction:2515 "move.l" or better still unknown instruction:2515 
 754"Fixme not implemented yet, call Martin" you are probably are attempting to compile some code 
 755meant for another architecture or code that is simply not implemented, with a fixme statement
 756stuck into the inline assembly code so that the author of the file now knows he has work to do.
 757To look at the assembly emitted by gcc just before it is about to call gas ( the gnu assembler )
 758use the -S option.
 759Again for your convenience the Linux kernel's Makefile will hold your hand &
 760do all this donkey work for you also by building the file with the .s suffix.
 762from the Linux directory type 
 763make arch/s390/kernel/signal.s 
 765s390-gcc -D__KERNEL__ -I/home1/barrow/linux/include -Wall -Wstrict-prototypes -O2 -fomit-frame-pointer
 766-fno-strict-aliasing -D__SMP__ -pipe -fno-strength-reduce  -S arch/s390/kernel/signal.c 
 767-o arch/s390/kernel/signal.s  
 770This will output something like, ( please note the constant pool & the useful comments
 771in the prologue to give you a hand at interpreting it ).
 774        .string "misaligned (__u16 *) in __xchg\n"
 776        .string "misaligned (__u32 *) in __xchg\n"
 777.L$PG1: # Pool sys_sigsuspend
 779        .long   -262401
 781        .long   -1
 783        .long   schedule-.L$PG1
 785        .long   do_signal-.L$PG1
 786        .align 4
 787.globl sys_sigsuspend
 788        .type    sys_sigsuspend,@function
 790#       leaf function           0
 791#       automatics              16
 792#       outgoing args           0
 793#       need frame pointer      0
 794#       call alloca             0
 795#       has varargs             0
 796#       incoming args (stack)   0
 797#       function length         168
 798        STM     8,15,32(15)
 799        LR      0,15
 800        AHI     15,-112
 801        BASR    13,0
 802.L$CO1: AHI     13,.L$PG1-.L$CO1
 803        ST      0,0(15)
 804        LR    8,2
 805        N     5,.LC192-.L$PG1(13) 
 807Adding -g to the above output makes the output even more useful
 808e.g. typing
 809make CC:="s390-gcc -g" kernel/sched.s
 811which compiles.
 812s390-gcc -g -D__KERNEL__ -I/home/barrow/linux-2.3/include -Wall -Wstrict-prototypes -O2 -fomit-frame-pointer -fno-strict-aliasing -pipe -fno-strength-reduce   -S kernel/sched.c -o kernel/sched.s 
 814also outputs stabs ( debugger ) info, from this info you can find out the
 815offsets & sizes of various elements in structures.
 816e.g. the stab for the structure
 817struct rlimit {
 818        unsigned long   rlim_cur;
 819        unsigned long   rlim_max;
 822.stabs "rlimit:T(151,2)=s8rlim_cur:(0,5),0,32;rlim_max:(0,5),32,32;;",128,0,0,0
 823from this stab you can see that 
 824rlimit_cur starts at bit offset 0 & is 32 bits in size
 825rlimit_max starts at bit offset 32 & is 32 bits in size.
 828Debugging Tools:
 833This is a tool with many options the most useful being ( if compiled with -g).
 834objdump --source <victim program or object file> > <victims debug listing >
 837The whole kernel can be compiled like this ( Doing this will make a 17MB kernel
 838& a 200 MB listing ) however you have to strip it before building the image
 839using the strip command to make it a more reasonable size to boot it.
 841A source/assembly mixed dump of the kernel can be done with the line
 842objdump --source vmlinux > vmlinux.lst
 843Also, if the file isn't compiled -g, this will output as much debugging information
 844as it can (e.g. function names). This is very slow as it spends lots
 845of time searching for debugging info. The following self explanatory line should be used 
 846instead if the code isn't compiled -g, as it is much faster:
 847objdump --disassemble-all --syms vmlinux > vmlinux.lst  
 849As hard drive space is valuable most of us use the following approach.
 8501) Look at the emitted psw on the console to find the crash address in the kernel.
 8512) Look at the file ( in the linux directory ) produced when building 
 852the kernel to find the closest address less than the current PSW to find the
 853offending function.
 8543) use grep or similar to search the source tree looking for the source file
 855 with this function if you don't know where it is.
 8564) rebuild this object file with -g on, as an example suppose the file was
 857( /arch/s390/kernel/signal.o ) 
 8585) Assuming the file with the erroneous function is signal.c Move to the base of the 
 859Linux source tree.
 8606) rm /arch/s390/kernel/signal.o
 8617) make /arch/s390/kernel/signal.o
 8628) watch the gcc command line emitted
 8639) type it in again or alternatively cut & paste it on the console adding the -g option.
 86410) objdump --source arch/s390/kernel/signal.o > signal.lst
 865This will output the source & the assembly intermixed, as the snippet below shows
 866This will unfortunately output addresses which aren't the same
 867as the kernel ones you should be able to get around the mental arithmetic
 868by playing with the --adjust-vma parameter to objdump.
 873static inline void spin_lock(spinlock_t *lp)
 875      a0:       18 34           lr      %r3,%r4
 876      a2:       a7 3a 03 bc     ahi     %r3,956
 877        __asm__ __volatile("    lhi   1,-1\n"
 878      a6:       a7 18 ff ff     lhi     %r1,-1
 879      aa:       1f 00           slr     %r0,%r0
 880      ac:       ba 01 30 00     cs      %r0,%r1,0(%r3)
 881      b0:       a7 44 ff fd     jm      aa <sys_sigsuspend+0x2e>
 882        saveset = current->blocked;
 883      b4:       d2 07 f0 68     mvc     104(8,%r15),972(%r4)
 884      b8:       43 cc
 885        return (set->sig[0] & mask) != 0;
 8886) If debugging under VM go down to that section in the document for more info.
 891I now have a tool which takes the pain out of --adjust-vma
 892& you are able to do something like
 893make /arch/s390/kernel/traps.lst
 894& it automatically generates the correctly relocated entries for
 895the text segment in traps.lst.
 896This tool is now standard in linux distro's in scripts/makelst
 900Q. What is it ?
 901A. It is a tool for intercepting calls to the kernel & logging them
 902to a file & on the screen.
 904Q. What use is it ?
 905A. You can use it to find out what files a particular program opens.
 909Example 1
 911If you wanted to know does ping work but didn't have the source 
 912strace ping -c 1  
 913& then look at the man pages for each of the syscalls below,
 914( In fact this is sometimes easier than looking at some spaghetti
 915source which conditionally compiles for several architectures ).
 916Not everything that it throws out needs to make sense immediately.
 918Just looking quickly you can see that it is making up a RAW socket
 919for the ICMP protocol.
 920Doing an alarm(10) for a 10 second timeout
 921& doing a gettimeofday call before & after each read to see 
 922how long the replies took, & writing some text to stdout so the user
 923has an idea what is going on.
 926getuid()                                = 0
 927setuid(0)                               = 0
 928stat("/usr/share/locale/C/", 0xbffff134) = -1 ENOENT (No such file or directory)
 929stat("/usr/share/locale/libc/C", 0xbffff134) = -1 ENOENT (No such file or directory)
 930stat("/usr/local/share/locale/C/", 0xbffff134) = -1 ENOENT (No such file or directory)
 931getpid()                                = 353
 932setsockopt(3, SOL_SOCKET, SO_BROADCAST, [1], 4) = 0
 933setsockopt(3, SOL_SOCKET, SO_RCVBUF, [49152], 4) = 0
 934fstat(1, {st_mode=S_IFCHR|0620, st_rdev=makedev(3, 1), ...}) = 0
 935mmap(0, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x40008000
 936ioctl(1, TCGETS, {B9600 opost isig icanon echo ...}) = 0
 937write(1, "PING ( 56 d"..., 42PING ( 56 data bytes
 938) = 42
 939sigaction(SIGINT, {0x8049ba0, [], SA_RESTART}, {SIG_DFL}) = 0 
 940sigaction(SIGALRM, {0x8049600, [], SA_RESTART}, {SIG_DFL}) = 0
 941gettimeofday({948904719, 138951}, NULL) = 0
 942sendto(3, "\10\0D\201a\1\0\0\17#\2178\307\36"..., 64, 0, {sin_family=AF_INET,
 943sin_port=htons(0), sin_addr=inet_addr("")}, 16) = 64
 944sigaction(SIGALRM, {0x8049600, [], SA_RESTART}, {0x8049600, [], SA_RESTART}) = 0
 945sigaction(SIGALRM, {0x8049ba0, [], SA_RESTART}, {0x8049600, [], SA_RESTART}) = 0
 946alarm(10)                               = 0
 947recvfrom(3, "E\0\0T\0005\0\0@\1|r\177\0\0\1\177"..., 192, 0, 
 948{sin_family=AF_INET, sin_port=htons(50882), sin_addr=inet_addr("")}, [16]) = 84
 949gettimeofday({948904719, 160224}, NULL) = 0
 950recvfrom(3, "E\0\0T\0006\0\0\377\1\275p\177\0"..., 192, 0, 
 951{sin_family=AF_INET, sin_port=htons(50882), sin_addr=inet_addr("")}, [16]) = 84
 952gettimeofday({948904719, 166952}, NULL) = 0
 953write(1, "64 bytes from icmp_se"..., 
 9545764 bytes from icmp_seq=0 ttl=255 time=28.0 ms
 956Example 2
 958strace passwd 2>&1 | grep open
 959produces the following output
 960open("/etc/", O_RDONLY)      = 3
 961open("/opt/kde/lib/", O_RDONLY) = -1 ENOENT (No such file or directory)
 962open("/lib/", O_RDONLY)        = 3
 963open("/dev", O_RDONLY)                  = 3
 964open("/var/run/utmp", O_RDONLY)         = 3
 965open("/etc/passwd", O_RDONLY)           = 3
 966open("/etc/shadow", O_RDONLY)           = 3
 967open("/etc/login.defs", O_RDONLY)       = 4
 968open("/dev/tty", O_RDONLY)              = 4 
 970The 2>&1 is done to redirect stderr to stdout & grep is then filtering this input 
 971through the pipe for each line containing the string open.
 974Example 3
 976Getting sophisticated
 977telnetd crashes & I don't know why
 9811) Replace the following line in /etc/inetd.conf
 982telnet  stream  tcp     nowait  root    /usr/sbin/in.telnetd -h 
 984telnet  stream  tcp     nowait  root    /blah
 9862) Create the file /blah with the following contents to start tracing telnetd 
 988/usr/bin/strace -o/t1 -f /usr/sbin/in.telnetd -h 
 9893) chmod 700 /blah to make it executable only to root
 991killall -HUP inetd
 992or ps aux | grep inetd
 993get inetd's process id
 994& kill -HUP inetd to restart it.
 996Important options
 998-o is used to tell strace to output to a file in our case t1 in the root directory
 999-f is to follow children i.e.
1000e.g in our case above telnetd will start the login process & subsequently a shell like bash.
1001You will be able to tell which is which from the process ID's listed on the left hand side
1002of the strace output.
1003-p<pid> will tell strace to attach to a running process, yup this can be done provided
1004 it isn't being traced or debugged already & you have enough privileges,
1005the reason 2 processes cannot trace or debug the same program is that strace
1006becomes the parent process of the one being debugged & processes ( unlike people )
1007can have only one parent.
1010However the file /t1 will get big quite quickly
1011to test it telnet
1013now look at what files in.telnetd execve'd
1014413   execve("/usr/sbin/in.telnetd", ["/usr/sbin/in.telnetd", "-h"], [/* 17 vars */]) = 0
1015414   execve("/bin/login", ["/bin/login", "-h", "localhost", "-p"], [/* 2 vars */]) = 0 
1017Whey it worked!.
1020Other hints:
1022If the program is not very interactive ( i.e. not much keyboard input )
1023& is crashing in one architecture but not in another you can do 
1024an strace of both programs under as identical a scenario as you can
1025on both architectures outputting to a file then.
1026do a diff of the two traces using the diff program
1028diff output1 output2
1029& maybe you'll be able to see where the call paths differed, this
1030is possibly near the cause of the crash. 
1032More info
1034Look at man pages for strace & the various syscalls
1035e.g. man strace, man alarm, man socket.
1038Performance Debugging
1040gcc is capable of compiling in profiling code just add the -p option
1041to the CFLAGS, this obviously affects program size & performance.
1042This can be used by the gprof gnu profiling tool or the
1043gcov the gnu code coverage tool ( code coverage is a means of testing
1044code quality by checking if all the code in an executable in exercised by
1045a tester ).
1048Using top to find out where processes are sleeping in the kernel
1050To do this copy the from the root directory where
1051the linux kernel was built to the /boot directory on your 
1052linux machine.
1053Start top
1054Now type fU<return>
1055You should see a new field called WCHAN which
1056tells you where each process is sleeping here is a typical output.
1058 6:59pm  up 41 min,  1 user,  load average: 0.00, 0.00, 0.00
105928 processes: 27 sleeping, 1 running, 0 zombie, 0 stopped
1060CPU states:  0.0% user,  0.1% system,  0.0% nice, 99.8% idle
1061Mem:   254900K av,   45976K used,  208924K free,       0K shrd,   28636K buff
1062Swap:       0K av,       0K used,       0K free                    8620K cached
1065  750 root      12   0   848  848   700 do_select S       0  0.1  0.3   0:00 in.telnetd
1066  767 root      16   0  1140 1140   964           R       0  0.1  0.4   0:00 top
1067    1 root       8   0   212  212   180 do_select S       0  0.0  0.0   0:00 init
1068    2 root       9   0     0    0     0 down_inte SW      0  0.0  0.0   0:00 kmcheck
1070The time command
1072Another related command is the time command which gives you an indication
1073of where a process is spending the majority of its time.
1075time ping -c 5 nc
1077real    0m4.054s
1078user    0m0.010s
1079sys     0m0.010s
1081Debugging under VM
1086Addresses & values in the VM debugger are always hex never decimal
1087Address ranges are of the format <HexValue1>-<HexValue2> or <HexValue1>.<HexValue2> 
1088e.g. The address range  0x2000 to 0x3000 can be described as 2000-3000 or 2000.1000
1090The VM Debugger is case insensitive.
1092VM's strengths are usually other debuggers weaknesses you can get at any resource
1093no matter how sensitive e.g. memory management resources,change address translation
1094in the PSW. For kernel hacking you will reap dividends if you get good at it.
1096The VM Debugger displays operators but not operands, probably because some
1097of it was written when memory was expensive & the programmer was probably proud that
1098it fitted into 2k of memory & the programmers & didn't want to shock hardcore VM'ers by
1099changing the interface :-), also the debugger displays useful information on the same line & 
1100the author of the code probably felt that it was a good idea not to go over 
1101the 80 columns on the screen. 
1103As some of you are probably in a panic now this isn't as unintuitive as it may seem
1104as the 390 instructions are easy to decode mentally & you can make a good guess at a lot 
1105of them as all the operands are nibble ( half byte aligned ) & if you have an objdump listing
1106also it is quite easy to follow, if you don't have an objdump listing keep a copy of
1107the s/390 Reference Summary & look at between pages 2 & 7 or alternatively the
1108s/390 principles of operation.
1109e.g. even I can guess that 
11100001AFF8' LR    180F        CC 0
1111is a ( load register ) lr r0,r15 
1113Also it is very easy to tell the length of a 390 instruction from the 2 most significant
1114bits in the instruction ( not that this info is really useful except if you are trying to
1115make sense of a hexdump of code ).
1116Here is a table
1117Bits                    Instruction Length
111900                          2 Bytes
112001                          4 Bytes
112110                          4 Bytes
112211                          6 Bytes
1127The debugger also displays other useful info on the same line such as the
1128addresses being operated on destination addresses of branches & condition codes.
113000019736' AHI   A7DAFF0E    CC 1
1131000198BA' BRC   A7840004 -> 000198C2'   CC 0
1132000198CE' STM   900EF068 >> 0FA95E78    CC 2
1136Useful VM debugger commands
1139I suppose I'd better mention this before I start
1140to list the current active traces do 
1141Q TR
1142there can be a maximum of 255 of these per set
1143( more about trace sets later ).
1144To stop traces issue a
1145TR END.
1146To delete a particular breakpoint issue
1147TR DEL <breakpoint number>
1149The PA1 key drops to CP mode so you can issue debugger commands,
1150Doing alt c (on my 3270 console at least ) clears the screen. 
1151hitting b <enter> comes back to the running operating system
1152from cp mode ( in our case linux ).
1153It is typically useful to add shortcuts to your profile.exec file
1154if you have one ( this is roughly equivalent to autoexec.bat in DOS ).
1155file here are a few from mine.
1156/* this gives me command history on issuing f12 */
1157set pf12 retrieve 
1158/* this continues */
1159set pf8 imm b
1160/* goes to trace set a */
1161set pf1 imm tr goto a
1162/* goes to trace set b */
1163set pf2 imm tr goto b
1164/* goes to trace set c */
1165set pf3 imm tr goto c
1169Instruction Tracing
1171Setting a simple breakpoint
1172TR I PSWA <address>
1173To debug a particular function try
1174TR I R <function address range>
1175TR I on its own will single step.
1176TR I DATA <MNEMONIC> <OPTIONAL RANGE> will trace for particular mnemonics
1178TR I DATA 4D R 0197BC.4000
1179will trace for BAS'es ( opcode 4D ) in the range 0197BC.4000
1180if you were inclined you could add traces for all branch instructions &
1181suffix them with the run prefix so you would have a backtrace on screen 
1182when a program crashes.
1183TR BR <INTO OR FROM> will trace branches into or out of an address.
1185TR BR INTO 0 is often quite useful if a program is getting awkward & deciding
1186to branch to 0 & crashing as this will stop at the address before in jumps to 0.
1187TR I R <address range> RUN cmd d g
1188single steps a range of addresses but stays running &
1189displays the gprs on each step.
1193Displaying & modifying Registers
1195D G will display all the gprs
1196Adding a extra G to all the commands is necessary to access the full 64 bit 
1197content in VM on z/Architecture obviously this isn't required for access registers
1198as these are still 32 bit.
1199e.g. DGG instead of DG 
1200D X will display all the control registers
1201D AR will display all the access registers
1202D AR4-7 will display access registers 4 to 7
1203CPU ALL D G will display the GRPS of all CPUS in the configuration
1204D PSW will display the current PSW
1205st PSW 2000 will put the value 2000 into the PSW &
1206cause crash your machine.
1207D PREFIX displays the prefix offset
1210Displaying Memory
1212To display memory mapped using the current PSW's mapping try
1213D <range>
1214To make VM display a message each time it hits a particular address & continue try
1215D I<range> will disassemble/display a range of instructions.
1216ST addr 32 bit word will store a 32 bit aligned address
1217D T<range> will display the EBCDIC in an address ( if you are that way inclined )
1218D R<range> will display real addresses ( without DAT ) but with prefixing.
1219There are other complex options to display if you need to get at say home space
1220but are in primary space the easiest thing to do is to temporarily
1221modify the PSW to the other addressing mode, display the stuff & then
1222restore it.
1228If you want to issue a debugger command without halting your virtual machine with the
1229PA1 key try prefixing the command with #CP e.g.
1230#cp tr i pswa 2000
1231also suffixing most debugger commands with RUN will cause them not
1232to stop just display the mnemonic at the current instruction on the console.
1233If you have several breakpoints you want to put into your program &
1234you get fed up of cross referencing with
1235you can do the following trick for several symbols.
1236grep do_signal 
1237which emits the following among other things
12380001f4e0 T do_signal 
1239now you can do
1241TR I PSWA 0001f4e0 cmd msg * do_signal
1242This sends a message to your own console each time do_signal is entered.
1243( As an aside I wrote a perl script once which automatically generated a REXX
1244script with breakpoints on every kernel procedure, this isn't a good idea
1245because there are thousands of these routines & VM can only set 255 breakpoints
1246at a time so you nearly had to spend as long pruning the file down as you would 
1247entering the msg's by hand ),however, the trick might be useful for a single object file.
1248On linux'es 3270 emulator x3270 there is a very useful option under the file ment
1249Save Screens In File this is very good of keeping a copy of traces. 
1251From CMS help <command name> will give you online help on a particular command. 
1255Also CP has a file called profile.exec which automatically gets called
1256on startup of CMS ( like autoexec.bat ), keeping on a DOS analogy session
1257CP has a feature similar to doskey, it may be useful for you to
1258use profile.exec to define some keystrokes. 
1261This does a single step in VM on pressing F8. 
1262SET PF10  ^
1263This sets up the ^ key.
1264which can be used for ^c (ctrl-c),^z (ctrl-z) which can't be typed directly into some 3270 consoles.
1265SET PF11 ^-
1266This types the starting keystrokes for a sysrq see SysRq below.
1268This retrieves command history on pressing F12.
1271Sometimes in VM the display is set up to scroll automatically this
1272can be very annoying if there are messages you wish to look at
1273to stop this do
1274TERM MORE 255 255
1275This will nearly stop automatic screen updates, however it will
1276cause a denial of service if lots of messages go to the 3270 console,
1277so it would be foolish to use this as the default on a production machine.
1280Tracing particular processes
1282The kernel's text segment is intentionally at an address in memory that it will
1283very seldom collide with text segments of user programs ( thanks Martin ),
1284this simplifies debugging the kernel.
1285However it is quite common for user processes to have addresses which collide
1286this can make debugging a particular process under VM painful under normal
1287circumstances as the process may change when doing a 
1288TR I R <address range>.
1289Thankfully after reading VM's online help I figured out how to debug
1290I particular process.
1292Your first problem is to find the STD ( segment table designation )
1293of the program you wish to debug.
1294There are several ways you can do this here are a few
12951) objdump --syms <program to be debugged> | grep main
1296To get the address of main in the program.
1297tr i pswa <address of main>
1298Start the program, if VM drops to CP on what looks like the entry
1299point of the main function this is most likely the process you wish to debug.
1300Now do a D X13 or D XG13 on z/Architecture.
1301On 31 bit the STD is bits 1-19 ( the STO segment table origin ) 
1302& 25-31 ( the STL segment table length ) of CR13.
1303now type
1304TR I R STD <CR13's value> 0.7fffffff
1306TR I R STD 8F32E1FF 0.7fffffff
1307Another very useful variation is
1308TR STORE INTO STD <CR13's value> <address range>
1309for finding out when a particular variable changes.
1311An alternative way of finding the STD of a currently running process 
1312is to do the following, ( this method is more complex but
1313could be quite convenient if you aren't updating the kernel much &
1314so your kernel structures will stay constant for a reasonable period of
1315time ).
1317grep task /proc/<pid>/status
1318from this you should see something like
1319task: 0f160000 ksp: 0f161de8 pt_regs: 0f161f68
1320This now gives you a pointer to the task structure.
1321Now make CC:="s390-gcc -g" kernel/sched.s
1322To get the task_struct stabinfo.
1323( task_struct is defined in include/linux/sched.h ).
1324Now we want to look at
1326on my machine the active_mm in the task structure stab is
1328its offset is 672/8=84=0x54
1329the pgd member in the mm_struct stab is
1331so its offset is 96/8=12=0xc
1333so we'll
1334hexdump -s 0xf160054 /dev/mem | more
1335i.e. task_struct+active_mm offset
1336to look at the active_mm member
1337f160054 0fee cc60 0019 e334 0000 0000 0000 0011
1338hexdump -s 0x0feecc6c /dev/mem | more
1339i.e. active_mm+pgd offset
1340feecc6c 0f2c 0000 0000 0001 0000 0001 0000 0010
1341we get something like
1342now do 
1343TR I R STD <pgd|0x7f> 0.7fffffff
1344i.e. the 0x7f is added because the pgd only
1345gives the page table origin & we need to set the low bits
1346to the maximum possible segment table length.
1347TR I R STD 0f2c007f 0.7fffffff
1348on z/Architecture you'll probably need to do
1349TR I R STD <pgd|0x7> 0.ffffffffffffffff
1350to set the TableType to 0x1 & the Table length to 3.
1354Tracing Program Exceptions
1356If you get a crash which says something like
1357illegal operation or specification exception followed by a register dump
1358You can restart linux & trace these using the tr prog <range or value> trace option.
1362The most common ones you will normally be tracing for is
13631=operation exception
13642=privileged operation exception
13654=protection exception
13665=addressing exception
13676=specification exception
136810=segment translation exception
136911=page translation exception
1371The full list of these is on page 22 of the current s/390 Reference Summary.
1373tr prog 10 will trace segment translation exceptions.
1374tr prog on its own will trace all program interruption codes.
1376Trace Sets
1378On starting VM you are initially in the INITIAL trace set.
1379You can do a Q TR to verify this.
1380If you have a complex tracing situation where you wish to wait for instance 
1381till a driver is open before you start tracing IO, but know in your
1382heart that you are going to have to make several runs through the code till you
1383have a clue whats going on. 
1385What you can do is
1386TR I PSWA <Driver open address>
1387hit b to continue till breakpoint
1388reach the breakpoint
1389now do your
1390TR GOTO B 
1391TR IO 7c08-7c09 inst int run 
1392or whatever the IO channels you wish to trace are & hit b
1394To got back to the initial trace set do
1396& the TR I PSWA <Driver open address> will be the only active breakpoint again.
1399Tracing linux syscalls under VM
1401Syscalls are implemented on Linux for S390 by the Supervisor call instruction (SVC) there 256 
1402possibilities of these as the instruction is made up of a  0xA opcode & the second byte being
1403the syscall number. They are traced using the simple command.
1404TR SVC  <Optional value or range>
1405the syscalls are defined in linux/arch/s390/include/asm/unistd.h
1406e.g. to trace all file opens just do
1407TR SVC 5 ( as this is the syscall number of open )
1410SMP Specific commands
1412To find out how many cpus you have
1413Q CPUS displays all the CPU's available to your virtual machine
1414To find the cpu that the current cpu VM debugger commands are being directed at do
1415Q CPU to change the current cpu VM debugger commands are being directed at do
1416CPU <desired cpu no>
1418On a SMP guest issue a command to all CPUs try prefixing the command with cpu all.
1419To issue a command to a particular cpu try cpu <cpu number> e.g.
1420CPU 01 TR I R 2000.3000
1421If you are running on a guest with several cpus & you have a IO related problem
1422& cannot follow the flow of code but you know it isn't smp related.
1423from the bash prompt issue
1424shutdown -h now or halt.
1425do a Q CPUS to find out how many cpus you have
1426detach each one of them from cp except cpu 0 
1427by issuing a 
1428DETACH CPU 01-(number of cpus in configuration)
1429& boot linux again.
1430TR SIGP will trace inter processor signal processor instructions.
1431DEFINE CPU 01-(number in configuration) 
1432will get your guests cpus back.
1435Help for displaying ascii textstrings
1437On the very latest VM Nucleus'es VM can now display ascii
1438( thanks Neale for the hint ) by doing
1439D TX<lowaddr>.<len>
1441D TX0.100
1445Under older VM debuggers ( I love EBDIC too ) you can use this little program I wrote which
1446will convert a command line of hex digits to ascii text which can be compiled under linux & 
1447you can copy the hex digits from your x3270 terminal to your xterm if you are debugging
1448from a linuxbox.
1450This is quite useful when looking at a parameter passed in as a text string
1451under VM ( unless you are good at decoding ASCII in your head ).
1453e.g. consider tracing an open syscall
1454TR SVC 5
1455We have stopped at a breakpoint
1456000151B0' SVC   0A05     -> 0001909A'   CC 0
1458D 20.8 to check the SVC old psw in the prefix area & see was it from userspace
1459( for the layout of the prefix area consult P18 of the s/390 390 Reference Summary 
1460if you have it available ).
1461V00000020  070C2000 800151B2
1462The problem state bit wasn't set &  it's also too early in the boot sequence
1463for it to be a userspace SVC if it was we would have to temporarily switch the 
1464psw to user space addressing so we could get at the first parameter of the open in
1466Next do a 
1467D G2
1468GPR  2 =  00014CB4
1469Now display what gpr2 is pointing to
1470D 00014CB4.20
1471V00014CB4  2F646576 2F636F6E 736F6C65 00001BF5
1472V00014CC4  FC00014C B4001001 E0001000 B8070707
1473Now copy the text till the first 00 hex ( which is the end of the string
1474to an xterm & do hex2ascii on it.
1475hex2ascii 2F646576 2F636F6E 736F6C65 00 
1477Decoded Hex:=/ d e v / c o n s o l e 0x00 
1478We were opening the console device,
1480You can compile the code below yourself for practice :-),
1482 *    hex2ascii.c
1483 *    a useful little tool for converting a hexadecimal command line to ascii
1484 *
1485 *    Author(s): Denis Joseph Barrow (,
1486 *    (C) 2000 IBM Deutschland Entwicklung GmbH, IBM Corporation.
1487 */   
1488#include <stdio.h>
1490int main(int argc,char *argv[])
1492  int cnt1,cnt2,len,toggle=0;
1493  int startcnt=1;
1494  unsigned char c,hex;
1496  if(argc>1&&(strcmp(argv[1],"-a")==0))
1497     startcnt=2;
1498  printf("Decoded Hex:=");
1499  for(cnt1=startcnt;cnt1<argc;cnt1++)
1500  {
1501    len=strlen(argv[cnt1]);
1502    for(cnt2=0;cnt2<len;cnt2++)
1503    {
1504       c=argv[cnt1][cnt2];
1505       if(c>='0'&&c<='9')
1506          c=c-'0';
1507       if(c>='A'&&c<='F')
1508          c=c-'A'+10;
1509       if(c>='a'&&c<='f')
1510          c=c-'a'+10;
1511       switch(toggle)
1512       {
1513          case 0:
1514             hex=c<<4;
1515             toggle=1;
1516          break;
1517          case 1:
1518             hex+=c;
1519             if(hex<32||hex>127)
1520             {
1521                if(startcnt==1)
1522                   printf("0x%02X ",(int)hex);
1523                else
1524                   printf(".");
1525             }
1526             else
1527             {
1528               printf("%c",hex);
1529               if(startcnt==1)
1530                  printf(" ");
1531             }
1532             toggle=0;
1533          break;
1534       }
1535    }
1536  }
1537  printf("\n");
1543Stack tracing under VM
1545A basic backtrace
1548Here are the tricks I use 9 out of 10 times it works pretty well,
1550When your backchain reaches a dead end
1552This can happen when an exception happens in the kernel & the kernel is entered twice
1553if you reach the NULL pointer at the end of the back chain you should be
1554able to sniff further back if you follow the following tricks.
15551) A kernel address should be easy to recognise since it is in
1556primary space & the problem state bit isn't set & also
1557The Hi bit of the address is set.
15582) Another backchain should also be easy to recognise since it is an 
1559address pointing to another address approximately 100 bytes or 0x70 hex
1560behind the current stackpointer.
1563Here is some practice.
1564boot the kernel & hit PA1 at some random time
1565d g to display the gprs, this should display something like
1566GPR  0 =  00000001  00156018  0014359C  00000000
1567GPR  4 =  00000001  001B8888  000003E0  00000000
1568GPR  8 =  00100080  00100084  00000000  000FE000
1569GPR 12 =  00010400  8001B2DC  8001B36A  000FFED8
1570Note that GPR14 is a return address but as we are real men we are going to
1571trace the stack.
1572display 0x40 bytes after the stack pointer.
1574V000FFED8  000FFF38 8001B838 80014C8E 000FFF38
1575V000FFEE8  00000000 00000000 000003E0 00000000
1576V000FFEF8  00100080 00100084 00000000 000FE000
1577V000FFF08  00010400 8001B2DC 8001B36A 000FFED8
1580Ah now look at whats in sp+56 (sp+0x38) this is 8001B36A our saved r14 if
1581you look above at our stackframe & also agrees with GPR14.
1583now backchain 
1584d 000FFF38.40
1585we now are taking the contents of SP to get our first backchain.
1587V000FFF38  000FFFA0 00000000 00014995 00147094
1588V000FFF48  00147090 001470A0 000003E0 00000000
1589V000FFF58  00100080 00100084 00000000 001BF1D0
1590V000FFF68  00010400 800149BA 80014CA6 000FFF38
1592This displays a 2nd return address of 80014CA6
1594now do d 000FFFA0.40 for our 3rd backchain
1596V000FFFA0  04B52002 0001107F 00000000 00000000
1597V000FFFB0  00000000 00000000 FF000000 0001107F
1598V000FFFC0  00000000 00000000 00000000 00000000
1599V000FFFD0  00010400 80010802 8001085A 000FFFA0
1602our 3rd return address is 8001085A
1604as the 04B52002 looks suspiciously like rubbish it is fair to assume that the kernel entry routines
1605for the sake of optimisation don't set up a backchain.
1607now look at to see if the addresses make any sense.
1609grep -i 0001b3
1610outputs among other things
16110001b304 T cpu_idle 
1612so 8001B36A
1613is cpu_idle+0x66 ( quiet the cpu is asleep, don't wake it )
1616grep -i 00014 
1617produces among other things
161800014a78 T start_kernel  
1619so 0014CA6 is start_kernel+some hex number I can't add in my head.
1621grep -i 00108 
1622this produces
162300010800 T _stext
1624so   8001085A is _stext+0x5a
1626Congrats you've done your first backchain.
1630s/390 & z/Architecture IO Overview
1633I am not going to give a course in 390 IO architecture as this would take me quite a
1634while & I'm no expert. Instead I'll give a 390 IO architecture summary for Dummies if you have 
1635the s/390 principles of operation available read this instead. If nothing else you may find a few 
1636useful keywords in here & be able to use them on a web search engine like altavista to find 
1637more useful information.
1639Unlike other bus architectures modern 390 systems do their IO using mostly
1640fibre optics & devices such as tapes & disks can be shared between several mainframes,
1641also S390 can support up to 65536 devices while a high end PC based system might be choking
1642with around 64. Here is some of the common IO terminology
1645This is the logical number most IO commands use to talk to an IO device there can be up to
16460x10000 (65536) of these in a configuration typically there is a few hundred. Under VM
1647for simplicity they are allocated contiguously, however on the native hardware they are not
1648they typically stay consistent between boots provided no new hardware is inserted or removed.
1649Under Linux for 390 we use these as IRQ's & also when issuing an IO command (CLEAR SUBCHANNEL,
1651TEST SUBCHANNEL ) we use this as the ID of the device we wish to talk to, the most
1652important of these instructions are START SUBCHANNEL ( to start IO ), TEST SUBCHANNEL ( to check
1653whether the IO completed successfully ), & HALT SUBCHANNEL ( to kill IO ), a subchannel
1654can have up to 8 channel paths to a device this offers redundancy if one is not available.
1657Device Number:
1658This number remains static & Is closely tied to the hardware, there are 65536 of these
1659also they are made up of a CHPID ( Channel Path ID, the most significant 8 bits ) 
1660& another lsb 8 bits. These remain static even if more devices are inserted or removed
1661from the hardware, there is a 1 to 1 mapping between Subchannels & Device Numbers provided
1662devices aren't inserted or removed.
1664Channel Control Words:
1665CCWS are linked lists of instructions initially pointed to by an operation request block (ORB),
1666which is initially given to Start Subchannel (SSCH) command along with the subchannel number
1667for the IO subsystem to process while the CPU continues executing normal code.
1668These come in two flavours, Format 0 ( 24 bit for backward )
1669compatibility & Format 1 ( 31 bit ). These are typically used to issue read & write 
1670( & many other instructions ) they consist of a length field & an absolute address field.
1671For each IO typically get 1 or 2 interrupts one for channel end ( primary status ) when the
1672channel is idle & the second for device end ( secondary status ) sometimes you get both
1673concurrently, you check how the IO went on by issuing a TEST SUBCHANNEL at each interrupt,
1674from which you receive an Interruption response block (IRB). If you get channel & device end 
1675status in the IRB without channel checks etc. your IO probably went okay. If you didn't you
1676probably need a doctor to examine the IRB & extended status word etc.
1677If an error occurs, more sophisticated control units have a facility known as
1678concurrent sense this means that if an error occurs Extended sense information will
1679be presented in the Extended status word in the IRB if not you have to issue a
1680subsequent SENSE CCW command after the test subchannel. 
1683TPI( Test pending interrupt) can also be used for polled IO but in multitasking multiprocessor
1684systems it isn't recommended except for checking special cases ( i.e. non looping checks for
1685pending IO etc. ).
1687Store Subchannel & Modify Subchannel can be used to examine & modify operating characteristics
1688of a subchannel ( e.g. channel paths ).
1690Other IO related Terms:
1691Sysplex: S390's Clustering Technology
1692QDIO: S390's new high speed IO architecture to support devices such as gigabit ethernet,
1693this architecture is also designed to be forward compatible with up & coming 64 bit machines.
1696General Concepts 
1698Input Output Processors (IOP's) are responsible for communicating between
1699the mainframe CPU's & the channel & relieve the mainframe CPU's from the
1700burden of communicating with IO devices directly, this allows the CPU's to 
1701concentrate on data processing. 
1703IOP's can use one or more links ( known as channel paths ) to talk to each 
1704IO device. It first checks for path availability & chooses an available one,
1705then starts ( & sometimes terminates IO ).
1706There are two types of channel path: ESCON & the Parallel IO interface.
1708IO devices are attached to control units, control units provide the
1709logic to interface the channel paths & channel path IO protocols to 
1710the IO devices, they can be integrated with the devices or housed separately
1711& often talk to several similar devices ( typical examples would be raid 
1712controllers or a control unit which connects to 1000 3270 terminals ).
1715    +---------------------------------------------------------------+
1716    | +-----+ +-----+ +-----+ +-----+  +----------+  +----------+   |
1717    | | CPU | | CPU | | CPU | | CPU |  |  Main    |  | Expanded |   |
1718    | |     | |     | |     | |     |  |  Memory  |  |  Storage |   |
1719    | +-----+ +-----+ +-----+ +-----+  +----------+  +----------+   | 
1720    |---------------------------------------------------------------+
1721    |   IOP        |      IOP      |       IOP                      |
1722    |---------------------------------------------------------------
1723    | C | C | C | C | C | C | C | C | C | C | C | C | C | C | C | C | 
1724    ----------------------------------------------------------------
1725         ||                                              ||
1726         ||  Bus & Tag Channel Path                      || ESCON
1727         ||  ======================                      || Channel
1728         ||  ||                  ||                      || Path
1729    +----------+               +----------+         +----------+
1730    |          |               |          |         |          |
1731    |    CU    |               |    CU    |         |    CU    |
1732    |          |               |          |         |          |
1733    +----------+               +----------+         +----------+
1734       |      |                     |                |       |
1735+----------+ +----------+      +----------+   +----------+ +----------+
1736|I/O Device| |I/O Device|      |I/O Device|   |I/O Device| |I/O Device|
1737+----------+ +----------+      +----------+   +----------+ +----------+
1738  CPU = Central Processing Unit    
1739  C = Channel                      
1740  IOP = IP Processor               
1741  CU = Control Unit
1743The 390 IO systems come in 2 flavours the current 390 machines support both
1745The Older 360 & 370 Interface,sometimes called the Parallel I/O interface,
1746sometimes called Bus-and Tag & sometimes Original Equipment Manufacturers
1747Interface (OEMI).
1749This byte wide Parallel channel path/bus has parity & data on the "Bus" cable 
1750& control lines on the "Tag" cable. These can operate in byte multiplex mode for
1751sharing between several slow devices or burst mode & monopolize the channel for the
1752whole burst. Up to 256 devices can be addressed  on one of these cables. These cables are
1753about one inch in diameter. The maximum unextended length supported by these cables is
1754125 Meters but this can be extended up to 2km with a fibre optic channel extended 
1755such as a 3044. The maximum burst speed supported is 4.5 megabytes per second however
1756some really old processors support only transfer rates of 3.0, 2.0 & 1.0 MB/sec.
1757One of these paths can be daisy chained to up to 8 control units.
1760ESCON if fibre optic it is also called FICON 
1761Was introduced by IBM in 1990. Has 2 fibre optic cables & uses either leds or lasers
1762for communication at a signaling rate of up to 200 megabits/sec. As 10bits are transferred
1763for every 8 bits info this drops to 160 megabits/sec & to 18.6 Megabytes/sec once
1764control info & CRC are added. ESCON only operates in burst mode.
1766ESCONs typical max cable length is 3km for the led version & 20km for the laser version
1767known as XDF ( extended distance facility ). This can be further extended by using an
1768ESCON director which triples the above mentioned ranges. Unlike Bus & Tag as ESCON is
1769serial it uses a packet switching architecture the standard Bus & Tag control protocol
1770is however present within the packets. Up to 256 devices can be attached to each control
1771unit that uses one of these interfaces.
1773Common 390 Devices include:
1774Network adapters typically OSA2,3172's,2116's & OSA-E gigabit ethernet adapters,
1775Consoles 3270 & 3215 ( a teletype emulated under linux for a line mode console ).
1776DASD's direct access storage devices ( otherwise known as hard disks ).
1777Tape Drives.
1778CTC ( Channel to Channel Adapters ),
1779ESCON or Parallel Cables used as a very high speed serial link
1780between 2 machines. We use 2 cables under linux to do a bi-directional serial link.
1783Debugging IO on s/390 & z/Architecture under VM
1786Now we are ready to go on with IO tracing commands under VM
1788A few self explanatory queries:
1789Q OSA
1790Q CTC
1791Q DISK ( This command is CMS specific )
1792Q DASD
1799Q OSA on my machine returns
1800OSA  7C08 ON OSA   7C08 SUBCHANNEL = 0000
1801OSA  7C09 ON OSA   7C09 SUBCHANNEL = 0001
1802OSA  7C14 ON OSA   7C14 SUBCHANNEL = 0002
1803OSA  7C15 ON OSA   7C15 SUBCHANNEL = 0003
1805If you have a guest with certain privileges you may be able to see devices
1806which don't belong to you. To avoid this, add the option V.
1808Q V OSA
1810Now using the device numbers returned by this command we will
1811Trace the io starting up on the first device 7c08 & 7c09
1812In our simplest case we can trace the 
1813start subchannels
1814like TR SSCH 7C08-7C09
1815or the halt subchannels
1816or TR HSCH 7C08-7C09
1817MSCH's ,STSCH's I think you can guess the rest
1819Ingo's favourite trick is tracing all the IO's & CCWS & spooling them into the reader of another
1820VM guest so he can ftp the logfile back to his own machine.I'll do a small bit of this & give you
1821 a look at the output.
18231) Spool stdout to VM reader
1824SP PRT TO (another vm guest ) or * for the local vm guest
18252) Fill the reader with the trace
1826TR IO 7c08-7c09 INST INT CCW PRT RUN
18273) Start up linux 
1828i 00c  
18294) Finish the trace
1830TR END
18315) close the reader
1832C PRT
18336) list reader contents
18357) copy it to linux4's minidisk 
1836RECEIVE / LOG TXT A1 ( replace
1838filel & press F11 to look at it
1839You should see something like:
184100020942' SSCH  B2334000    0048813C    CC 0    SCH 0000    DEV 7C08
1842          CPA 000FFDF0   PARM 00E2C9C4    KEY 0  FPI C0  LPM 80
1843          CCW    000FFDF0  E4200100 00487FE8   0000  E4240100 ........
1844          IDAL                                      43D8AFE8
1845          IDAL                                      0FB76000
184600020B0A'   I/O DEV 7C08 -> 000197BC'   SCH 0000   PARM 00E2C9C4
184700021628' TSCH  B2354000 >> 00488164    CC 0    SCH 0000    DEV 7C08
1848          CCWA 000FFDF8   DEV STS 0C  SCH STS 00  CNT 00EC
1849           KEY 0   FPI C0  CC 0   CTLS 4007
185000022238' STSCH B2344000 >> 00488108    CC 0    SCH 0000    DEV 7C08
1852If you don't like messing up your readed ( because you possibly booted from it )
1853you can alternatively spool it to another readers guest.
1856Other common VM device related commands
1858These commands are listed only because they have
1859been of use to me in the past & may be of use to
1860you too. For more complete info on each of the commands
1861use type HELP <command> from CMS.
1862detaching devices
1863DET <devno range>
1864ATT <devno range> <guest> 
1865attach a device to guest * for your own guest
1866READY <devno> cause VM to issue a fake interrupt.
1868The VARY command is normally only available to VM administrators.
1869VARY ON PATH <path> TO <devno range>
1870VARY OFF PATH <PATH> FROM <devno range>
1871This is used to switch on or off channel paths to devices.
1873Q CHPID <channel path ID>
1874This displays state of devices using this channel path
1875D SCHIB <subchannel>
1876This displays the subchannel information SCHIB block for the device.
1877this I believe is also only available to administrators.
1878DEFINE CTC <devno>
1879defines a virtual CTC channel to channel connection
18802 need to be defined on each guest for the CTC driver to use.
1881COUPLE  devno userid remote devno
1882Joins a local virtual device to a remote virtual device
1883( commonly used for the CTC driver ).
1885Building a VM ramdisk under CMS which linux can use
1886def vfb-<blocksize> <subchannel> <number blocks>
1887blocksize is commonly 4096 for linux.
1888Formatting it
1889format <subchannel> <driver letter e.g. x> (blksize <blocksize>
1891Sharing a disk between multiple guests
1892LINK userid devno1 devno2 mode password
1896GDB on S390
1898N.B. if compiling for debugging gdb works better without optimisation 
1899( see Compiling programs for debugging )
1903gdb <victim program> <optional corefile>
1905Online help
1907help: gives help on commands
1910help display
1911Note gdb's online help is very good use it.
1916info registers: displays registers other than floating point.
1917info all-registers: displays floating points as well.
1918disassemble: disassembles
1920disassemble without parameters will disassemble the current function
1921disassemble $pc $pc+10 
1923Viewing & modifying variables
1925print or p: displays variable or register
1926e.g. p/x $sp will display the stack pointer
1928display: prints variable or register each time program stops
1930display/x $pc will display the program counter
1931display argc
1933undisplay : undo's display's
1935info breakpoints: shows all current breakpoints
1937info stack: shows stack back trace ( if this doesn't work too well, I'll show you the
1938stacktrace by hand below ).
1940info locals: displays local variables.
1942info args: display current procedure arguments.
1944set args: will set argc & argv each time the victim program is invoked.
1946set <variable>=value
1947set argc=100
1948set $pc=0
1952Modifying execution
1954step: steps n lines of sourcecode
1955step steps 1 line.
1956step 100 steps 100 lines of code.
1958next: like step except this will not step into subroutines
1960stepi: steps a single machine code instruction.
1961e.g. stepi 100
1963nexti: steps a single machine code instruction but will not step into subroutines.
1965finish: will run until exit of the current routine
1967run: (re)starts a program
1969cont: continues a program
1971quit: exits gdb.
1978sets a breakpoint
1981break main
1983break *$pc
1985break *0x400618
1987heres a really useful one for large programs
1989Set a breakpoint for all functions matching REGEXP
1991rbr 390
1992will set a breakpoint with all functions with 390 in their name.
1994info breakpoints
1995lists all breakpoints
1997delete: delete breakpoint by number or delete them all
1999delete 1 will delete the first breakpoint
2000delete will delete them all
2002watch: This will set a watchpoint ( usually hardware assisted ),
2003This will watch a variable till it changes
2005watch cnt, will watch the variable cnt till it changes.
2006As an aside unfortunately gdb's, architecture independent watchpoint code
2007is inconsistent & not very good, watchpoints usually work but not always.
2009info watchpoints: Display currently active watchpoints
2011condition: ( another useful one )
2012Specify breakpoint number N to break only if COND is true.
2013Usage is `condition N COND', where N is an integer and COND is an
2014expression to be evaluated whenever breakpoint N is reached.
2018User defined functions/macros
2020define: ( Note this is very very useful,simple & powerful )
2021usage define <name> <list of commands> end
2023examples which you should consider putting into .gdbinit in your home directory
2024define d
2026disassemble $pc $pc+10
2029define e
2031disassemble $pc $pc+10
2035Other hard to classify stuff
2037signal n:
2038sends the victim program a signal.
2039e.g. signal 3 will send a SIGQUIT.
2041info signals:
2042what gdb does when the victim receives certain signals.
2046list lists current function source
2047list 1,10 list first 10 lines of current file.
2048list test.c:1,10
2052Adds directories to be searched for source if gdb cannot find the source.
2053(note it is a bit sensitive about slashes)
2054e.g. To add the root of the filesystem to the searchpath do
2055directory //
2058call <function>
2059This calls a function in the victim program, this is pretty powerful
2061(gdb) call printf("hello world")
2063$1 = 11 
2065You might now be thinking that the line above didn't work, something extra had to be done.
2066(gdb) call fflush(stdout)
2067hello world$2 = 0
2068As an aside the debugger also calls malloc & free under the hood 
2069to make space for the "hello world" string.
20751) command completion works just like bash 
2076( if you are a bad typist like me this really helps )
2077e.g. hit br <TAB> & cursor up & down :-).
20792) if you have a debugging problem that takes a few steps to recreate
2080put the steps into a file called .gdbinit in your current working directory
2081if you have defined a few extra useful user defined commands put these in 
2082your home directory & they will be read each time gdb is launched.
2084A typical .gdbinit file might be.
2085break main
2087break runtime_exception
2091stack chaining in gdb by hand
2093This is done using a the same trick described for VM 
2094p/x (*($sp+56))&0x7fffffff get the first backchain.
2096For z/Architecture
2097Replace 56 with 112 & ignore the &0x7fffffff
2098in the macros below & do nasty casts to longs like the following
2099as gdb unfortunately deals with printed arguments as ints which
2100messes up everything.
2101i.e. here is a 3rd backchain dereference
2102p/x *(long *)(***(long ***)$sp+112)
2105this outputs 
2106$5 = 0x528f18 
2107on my machine.
2108Now you can use 
2109info symbol (*($sp+56))&0x7fffffff 
2110you might see something like.
2111rl_getc + 36 in section .text  telling you what is located at address 0x528f18
2112Now do.
2113p/x (*(*$sp+56))&0x7fffffff 
2114This outputs
2115$6 = 0x528ed0
2116Now do.
2117info symbol (*(*$sp+56))&0x7fffffff
2118rl_read_key + 180 in section .text
2119now do
2120p/x (*(**$sp+56))&0x7fffffff
2121& so on.
2123Disassembling instructions without debug info
2125gdb typically complains if there is a lack of debugging
2126symbols in the disassemble command with 
2127"No function contains specified address." To get around
2128this do 
2129x/<number lines to disassemble>xi <address>
2131x/20xi 0x400730
2135Note: Remember gdb has history just like bash you don't need to retype the
2136whole line just use the up & down arrows.
2140For more info
2142From your linuxbox do 
2143man gdb or info gdb.
2145core dumps
2147What a core dump ?,
2148A core dump is a file generated by the kernel ( if allowed ) which contains the registers,
2149& all active pages of the program which has crashed.
2150From this file gdb will allow you to look at the registers & stack trace & memory of the
2151program as if it just crashed on your system, it is usually called core & created in the
2152current working directory.
2153This is very useful in that a customer can mail a core dump to a technical support department
2154& the technical support department can reconstruct what happened.
2155Provided they have an identical copy of this program with debugging symbols compiled in &
2156the source base of this build is available.
2157In short it is far more useful than something like a crash log could ever hope to be.
2159In theory all that is missing to restart a core dumped program is a kernel patch which
2160will do the following.
21611) Make a new kernel task structure
21622) Reload all the dumped pages back into the kernel's memory management structures.
21633) Do the required clock fixups
21644) Get all files & network connections for the process back into an identical state ( really difficult ).
21655) A few more difficult things I haven't thought of.
2169Why have I never seen one ?.
2170Probably because you haven't used the command 
2171ulimit -c unlimited in bash
2172to allow core dumps, now do 
2173ulimit -a 
2174to verify that the limit was accepted.
2176A sample core dump
2177To create this I'm going to do
2178ulimit -c unlimited
2180to launch gdb (my victim app. ) now be bad & do the following from another 
2181telnet/xterm session to the same machine
2182ps -aux | grep gdb
2183kill -SIGSEGV <gdb's pid>
2184or alternatively use killall -SIGSEGV gdb if you have the killall command.
2185Now look at the core dump.
2186./gdb core
2187Displays the following
2188GNU gdb 4.18
2189Copyright 1998 Free Software Foundation, Inc.
2190GDB is free software, covered by the GNU General Public License, and you are
2191welcome to change it and/or distribute copies of it under certain conditions.
2192Type "show copying" to see the conditions.
2193There is absolutely no warranty for GDB.  Type "show warranty" for details.
2194This GDB was configured as "s390-ibm-linux"...
2195Core was generated by `./gdb'.
2196Program terminated with signal 11, Segmentation fault.
2197Reading symbols from /usr/lib/
2198Reading symbols from /lib/
2199Reading symbols from /lib/
2200Reading symbols from /lib/
2201#0  0x40126d1a in read () from /lib/
2202Setting up the environment for debugging gdb.
2203Breakpoint 1 at 0x4dc6f8: file utils.c, line 471.
2204Breakpoint 2 at 0x4d87a4: file top.c, line 2609.
2205(top-gdb) info stack
2206#0  0x40126d1a in read () from /lib/
2207#1  0x528f26 in rl_getc (stream=0x7ffffde8) at input.c:402
2208#2  0x528ed0 in rl_read_key () at input.c:381
2209#3  0x5167e6 in readline_internal_char () at readline.c:454
2210#4  0x5168ee in readline_internal_charloop () at readline.c:507
2211#5  0x51692c in readline_internal () at readline.c:521
2212#6  0x5164fe in readline (prompt=0x7ffff810 "\177\xC2\x81\xC3\xBF\xC2\x81\xC3\xB8x\177\xC2\x81\xC3\xBF\xC2\x81\xC3\xB7\xC2\x81\xC3\x98\177\xC2\x81\xC3\xBF\xC2\x81\xC3\xB8x\xC2\x81\xC3\x80")
2213    at readline.c:349
2214#7  0x4d7a8a in command_line_input (prrompt=0x564420 "(gdb) ", repeat=1,
2215    annotation_suffix=0x4d6b44 "prompt") at top.c:2091
2216#8  0x4d6cf0 in command_loop () at top.c:1345
2217#9  0x4e25bc in main (argc=1, argv=0x7ffffdf4) at main.c:635
2222This is a program which lists the shared libraries which a library needs,
2223Note you also get the relocations of the shared library text segments which
2224help when using objdump --source.
2226 ldd ./gdb
2227outputs => /usr/lib/ (0x40018000) => /lib/ (0x4005e000) => /lib/ (0x40084000)
2231/lib/ => /lib/ (0x40000000)
2234Debugging shared libraries
2236Most programs use shared libraries, however it can be very painful
2237when you single step instruction into a function like printf for the 
2238first time & you end up in functions like _dl_runtime_resolve this is
2239the doing lazy binding, lazy binding is a concept in ELF where 
2240shared library functions are not loaded into memory unless they are 
2241actually used, great for saving memory but a pain to debug.
2242To get around this either relink the program -static or exit gdb type 
2243export LD_BIND_NOW=true this will stop lazy binding & restart the gdb'ing 
2244the program in question.
2248Debugging modules
2250As modules are dynamically loaded into the kernel their address can be
2251anywhere to get around this use the -m option with insmod to emit a load
2252map which can be piped into a file if required.
2254The proc file system
2256What is it ?.
2257It is a filesystem created by the kernel with files which are created on demand
2258by the kernel if read, or can be used to modify kernel parameters,
2259it is a powerful concept.
2263cat /proc/sys/net/ipv4/ip_forward 
2264On my machine outputs 
2266telling me ip_forwarding is not on to switch it on I can do
2267echo 1 >  /proc/sys/net/ipv4/ip_forward
2268cat it again
2269cat /proc/sys/net/ipv4/ip_forward 
2270On my machine now outputs
2272IP forwarding is on.
2273There is a lot of useful info in here best found by going in & having a look around,
2274so I'll take you through some entries I consider important.
2276All the processes running on the machine have there own entry defined by
2278So lets have a look at the init process
2279cd /proc/1
2281cat cmdline
2283init [2]
2285cd /proc/1/fd
2286This contains numerical entries of all the open files,
2287some of these you can cat e.g. stdout (2)
2289cat /proc/29/maps
2290on my machine emits
229200400000-00478000 r-xp 00000000 5f:00 4103       /bin/bash
229300478000-0047e000 rw-p 00077000 5f:00 4103       /bin/bash
22940047e000-00492000 rwxp 00000000 00:00 0
229540000000-40015000 r-xp 00000000 5f:00 14382      /lib/
229640015000-40016000 rw-p 00014000 5f:00 14382      /lib/
229740016000-40017000 rwxp 00000000 00:00 0
229840017000-40018000 rw-p 00000000 00:00 0
229940018000-4001b000 r-xp 00000000 5f:00 14435      /lib/
23004001b000-4001c000 rw-p 00002000 5f:00 14435      /lib/
23014001c000-4010d000 r-xp 00000000 5f:00 14387      /lib/
23024010d000-40111000 rw-p 000f0000 5f:00 14387      /lib/
230340111000-40114000 rw-p 00000000 00:00 0
230440114000-4011e000 r-xp 00000000 5f:00 14408      /lib/
23054011e000-4011f000 rw-p 00009000 5f:00 14408      /lib/
23067fffd000-80000000 rwxp ffffe000 00:00 0
2309Showing us the shared libraries init uses where they are in memory
2310& memory access permissions for each virtual memory area.
2312/proc/1/cwd is a softlink to the current working directory.
2313/proc/1/root is the root of the filesystem for this process. 
2315/proc/1/mem is the current running processes memory which you
2316can read & write to like a file.
2317strace uses this sometimes as it is a bit faster than the
2318rather inefficient ptrace interface for peeking at DATA.
2321cat status 
2323Name:   init
2324State:  S (sleeping)
2325Pid:    1
2326PPid:   0
2327Uid:    0       0       0       0
2328Gid:    0       0       0       0
2330VmSize:      408 kB
2331VmLck:         0 kB
2332VmRSS:       208 kB
2333VmData:       24 kB
2334VmStk:         8 kB
2335VmExe:       368 kB
2336VmLib:         0 kB
2337SigPnd: 0000000000000000
2338SigBlk: 0000000000000000
2339SigIgn: 7fffffffd7f0d8fc
2340SigCgt: 00000000280b2603
2341CapInh: 00000000fffffeff
2342CapPrm: 00000000ffffffff
2343CapEff: 00000000fffffeff
2345User PSW:    070de000 80414146
2346task: 004b6000 tss: 004b62d8 ksp: 004b7ca8 pt_regs: 004b7f68
2347User GPRS:
234800000400  00000000  0000000b  7ffffa90
234900000000  00000000  00000000  0045d9f4
23500045cafc  7ffffa90  7fffff18  0045cb08
235100010400  804039e8  80403af8  7ffff8b0
2352User ACRS:
235300000000  00000000  00000000  00000000
235400000001  00000000  00000000  00000000
235500000000  00000000  00000000  00000000
235600000000  00000000  00000000  00000000
2357Kernel BackChain  CallChain    BackChain  CallChain
2358       004b7ca8   8002bd0c     004b7d18   8002b92c
2359       004b7db8   8005cd50     004b7e38   8005d12a
2360       004b7f08   80019114                     
2361Showing among other things memory usage & status of some signals &
2362the processes'es registers from the kernel task_structure
2363as well as a backchain which may be useful if a process crashes
2364in the kernel for some unknown reason.
2366Some driver debugging techniques
2368debug feature
2370Some of our drivers now support a "debug feature" in
2371/proc/s390dbf see s390dbf.txt in the linux/Documentation directory
2372for more info.
2374to switch on the lcs "debug feature"
2375echo 5 > /proc/s390dbf/lcs/level
2376& then after the error occurred.
2377cat /proc/s390dbf/lcs/sprintf >/logfile
2378the logfile now contains some information which may help
2379tech support resolve a problem in the field.
2383high level debugging network drivers
2385ifconfig is a quite useful command
2386it gives the current state of network drivers.
2388If you suspect your network device driver is dead
2389one way to check is type 
2390ifconfig <network device> 
2391e.g. tr0
2392You should see something like
2393tr0       Link encap:16/4 Mbps Token Ring (New)  HWaddr 00:04:AC:20:8E:48
2394          inet addr:  Bcast:  Mask:
2395          UP BROADCAST RUNNING MULTICAST  MTU:2000  Metric:1
2396          RX packets:246134 errors:0 dropped:0 overruns:0 frame:0
2397          TX packets:5 errors:0 dropped:0 overruns:0 carrier:0
2398          collisions:0 txqueuelen:100
2400if the device doesn't say up
2402/etc/rc.d/init.d/network start 
2403( this starts the network stack & hopefully calls ifconfig tr0 up ).
2404ifconfig looks at the output of /proc/net/dev & presents it in a more presentable form
2405Now ping the device from a machine in the same subnet.
2406if the RX packets count & TX packets counts don't increment you probably
2407have problems.
2409cat /proc/net/arp
2410Do you see any hardware addresses in the cache if not you may have problems.
2411Next try
2412ping -c 5 <broadcast_addr> i.e. the Bcast field above in the output of
2413ifconfig. Do you see any replies from machines other than the local machine
2414if not you may have problems. also if the TX packets count in ifconfig
2415hasn't incremented either you have serious problems in your driver 
2416(e.g. the txbusy field of the network device being stuck on ) 
2417or you may have multiple network devices connected.
2422There is a new device layer for channel devices, some
2423drivers e.g. lcs are registered with this layer.
2424If the device uses the channel device layer you'll be
2425able to find what interrupts it uses & the current state 
2426of the device.
2427See the manpage chandev.8 &type cat /proc/chandev for more info.
2431Starting points for debugging scripting languages etc.
2436bash -x <scriptname>
2437e.g. bash -x /usr/bin/bashbug
2438displays the following lines as it executes them.
2439+ MACHINE=i586
2440+ OS=linux-gnu
2441+ CC=gcc
2442+ CFLAGS= -DPROGRAM='bash' -DHOSTTYPE='i586' -DOSTYPE='linux-gnu' -DMACHTYPE='i586-pc-linux-gnu' -DSHELL -DHAVE_CONFIG_H   -I. -I. -I./lib -O2 -pipe
2443+ RELEASE=2.01
2445+ RELSTATUS=release
2446+ MACHTYPE=i586-pc-linux-gnu   
2448perl -d <scriptname> runs the perlscript in a fully interactive debugger
2449<like gdb>.
2450Type 'h' in the debugger for help.
2452for debugging java type
2453jdb <filename> another fully interactive gdb style debugger.
2454& type ? in the debugger for help.
2458Dumptool & Lcrash ( lkcd )
2460Michael Holzheu & others here at IBM have a fairly mature port of 
2461SGI's lcrash tool which allows one to look at kernel structures in a
2462running kernel.
2464It also complements a tool called dumptool which dumps all the kernel's
2465memory pages & registers to either a tape or a disk.
2466This can be used by tech support or an ambitious end user do
2467post mortem debugging of a machine like gdb core dumps.
2469Going into how to use this tool in detail will be explained
2470in other documentation supplied by IBM with the patches & the 
2471lcrash homepage & the lcrash manpage.
2473How they work
2475Lcrash is a perfectly normal program,however, it requires 2 
2476additional files, Kerntypes which is built using a patch to the 
2477linux kernel sources in the linux root directory & the
2479Kerntypes is an objectfile whose sole purpose in life
2480is to provide stabs debug info to lcrash, to do this
2481Kerntypes is built from kerntypes.c which just includes the most commonly
2482referenced header files used when debugging, lcrash can then read the
2483.stabs section of this file.
2485Debugging a live system it uses /dev/mem
2486alternatively for post mortem debugging it uses the data 
2487collected by dumptool.
2493This is now supported by linux for s/390 & z/Architecture.
2494To enable it do compile the kernel with 
2495Kernel Hacking -> Magic SysRq Key Enabled
2496echo "1" > /proc/sys/kernel/sysrq
2497also type
2498echo "8" >/proc/sys/kernel/printk
2499To make printk output go to console.
2500On 390 all commands are prefixed with
2503^-t will show tasks.
2504^-? or some unknown command will display help.
2505The sysrq key reading is very picky ( I have to type the keys in an
2506 xterm session & paste them  into the x3270 console )
2507& it may be wise to predefine the keys as described in the VM hints above
2509This is particularly useful for syncing disks unmounting & rebooting
2510if the machine gets partially hung.
2512Read Documentation/sysrq.txt for more info
2516Enterprise Systems Architecture Reference Summary
2517Enterprise Systems Architecture Principles of Operation
2518Hartmut Penners s390 stack frame sheet.
2519IBM Mainframe Channel Attachment a technology brief from a CISCO webpage
2520Various bits of man & info pages of Linux.
2521Linux & GDB source.
2522Various info & man pages.
2523CMS Help on tracing commands.
2524Linux for s/390 Elf Application Binary Interface
2525Linux for z/Series Elf Application Binary Interface ( Both Highly Recommended )
2526z/Architecture Principles of Operation SA22-7832-00
2527Enterprise Systems Architecture/390 Reference Summary SA22-7209-01 & the
2528Enterprise Systems Architecture/390 Principles of Operation SA22-7201-05
2530Special Thanks
2532Special thanks to Neale Ferguson who maintains a much
2533prettier HTML version of this page at
2535Bob Grainger Stefan Bader & others for reporting bugs