Lect02
0 likes•545 views
1. The document discusses system calls from the user's perspective, providing a high-level overview of common system call categories like process management, file I/O, and directory management. 2. It explains that system calls provide a controlled entry point for applications to make requests to the operating system kernel to perform privileged operations. 3. Examples of system calls on different platforms like Linux, Windows, and MIPS are provided to illustrate the concept.
Lect02
- 1. System Calls 1
- 2. Learning Outcomes • A high-level understanding of System Calls – Mostly from the user’s perspective • From textbook (section 1.6) • Exposure architectural details of the MIPS R3000 – Detailed understanding of the of exception handling mechanism • From “Hardware Guide” on class web site • Understanding of the existence of compiler function calling conventions – Including details of the MIPS ‘C’ compiler calling convention • Understanding of how the application kernel boundary is crossed with system calls in general • Including an appreciation of the relationship between a case study (OS/161 system call handling) and the general case. 2
- 3. Operating System System Calls Kernel Level Operating System Requests Applications (System Calls) User Level Applications Applications 3
- 4. System Calls • Can be viewed as special function calls – Provides for a controlled entry into the kernel – While in kernel, they perform a privileged operation – Returns to original caller with the result • The system call interface represents the abstract machine provided by the operating system. 4
- 5. A Brief Overview of Classes System Calls • From the user’s perspective – Process Management – File I/O – Directories management – Some other selected Calls – There are many more • On Linux, see man syscalls for a list 5
- 6. Some System Calls For Process Management 6
- 7. Some System Calls For File Management 7
- 8. Some System Calls For Directory Management 8
- 9. Some System Calls For Miscellaneous Tasks 9
- 10. System Calls • A stripped down shell: while (TRUE) { /* repeat forever */ type_prompt( ); /* display prompt */ read_command (command, parameters) /* input from terminal */ if (fork() != 0) { /* fork off child process */ /* Parent code */ waitpid( -1, &status, 0); /* wait for child to exit */ } else { /* Child code */ execve (command, parameters, 0); /* execute command */ } } 10
- 11. System Calls Some Win32 API calls 11
- 12. The MIPS R2000/R3000 • Before looking at system call mechanics in some detail, we need a basic understanding of the MIPS R3000 12
- 13. MIPS R3000 • Load/store architecture – No instructions that operate on memory except load and store – Simple load/stores to/from memory from/to registers • Store word: sw r4, (r5) – Store contents of r4 in memory using address contained in register r5 • Load word: lw r3,(r7) – Load contents of memory into r3 using address contained in r7 – Delay of one instruction after load before data available in destination register » Must always an instruction between a load from memory and the subsequent use of the register. – lw, sw, lb, sb, lh, sh,…. 13
- 14. MIPS R3000 • Arithmetic and logical operations are register to register operations • E.g., add r3, r2, r1 • No arithmetic operations on memory • Example – add r3, r2, r1 ⇒ r3 = r2 + r1 • Some other instructions – add, sub, and, or, xor, sll, srl 14
- 15. MIPS R3000 • All instructions are encoded in 32-bit • Some instructions have immediate operands – Immediate values are constants encoded in the instruction itself – Only 16-bit value – Examples • Add Immediate: addi r2, r1, 2048 ⇒ r2 = r1 + 2048 • Load Immediate : li r2, 1234 ⇒ r2 = 1234 15
- 16. MIPS Registers • User-mode accessible registers – 32 general purpose registers • r0 hardwired to zero • r31 the link register for jump- and-link (JAL) instruction – HI/LO • 2 * 32-bits for multiply and divide – PC • Not directly visible • Modified implicitly by jump and branch instructions 16
- 17. Branching and Jumping li r2, 1 • Branching and jumping have a sw r0,(r3) branch delay slot j 1f – The instruction li r2, 2 following a branch li r2, 3 or jump is always 1: sw r2, (r3) executed prior to destination 17
- 18. Jump and Link Instruction M • JAL is used to implement function 0x10 jal 1f calls 0x14 nop – r31 = PC+8 0x18 lw r4,(r6) • Return Address M register (RA) is 1: used to return from 0x2a sw r2, (r3) function call M 0x38 jr r31 18 0x3a nop
- 19. MIPS R3000 • RISC architecture – 5 stage pipeline 19
- 20. Coprocessor 0 • The processor control registers are located in CP0 – Exception/Interrupt management registers – Translation management registers CP0 • CP0 is manipulated using mtc0 CP1 (floating point) (move to) and mfc0 (move PC: 0x0300 from) instructions HI/LO – mtc0/mfc0 are only accessible in R1 kernel mode. Rn 20
- 21. CP0 Registers • Exception Management • Miscellaneous • c0_cause – Cause of the recent – c0_prid exception • Processor Identifier • c0_status – Current status of the • Memory Management CPU • c0_epc • c0_index – Address of the • c0_random instruction that caused the exception • c0_entryhi • c0_badvaddr • c0_entrylo – Address accessed that • c0_context caused the exception – More about these later in course 21
- 22. c0_status • For practical purposes, you can ignore most bits – Green background is the focus 22
- 23. c0_status • IM • KU – Individual interrupt mask – 0 = kernel bits – 1 = user mode – 6 external • IE – 2 software – 0 = all interrupts masked – 1 = interrupts enable • Mask determined via IM bits • c, p, o = current, previous, old 23
- 24. c0_cause • IP • BD – Interrupts pending – If set, the instruction • 8 bits indicating current that caused the state of interrupt lines exception was in a • CE branch delay slot – Coprocessor error • ExcCode • Attempt to access disabled Copro. – The code number of the exception taken 24
- 25. Exception Codes 25
- 26. Exception Codes 26
- 27. c0_epc • The Exception Program nop Counter sw r3 (r4) c0_epc – Points to address of where nop to restart execution after handling the exception or interrupt – Example • Assume sw r3,(r4) causes a fault exception Aside: We are ignore BD-bit in c0_cause which is also used in reality on rare occasions. 27
- 28. Exception Vectors 28
- 29. Simple Exception Walk-through Application User Mode Kernel Mode Interrupt Handler 29
- 30. Hardware exception handling PC EPC 0x12345678 ? Cause • Let’s now walk through an exception ? – Assume an interrupt occurred as the Status KUo IEo KUp IEp KUc IEc previous instruction ? ? ? ? 1 1 completed – Note: We are in user mode with interrupts enabled 30
- 31. Hardware exception handling PC EPC 0x12345678 0x12345678 Cause ? • Instruction address at Status KUo IEo KUp IEp KUc IEc which to restart after ? ? ? ? 1 1 the interrupt is transferred to EPC 31
- 32. Hardware exception handling Interrupts PC EPC disabled 0x12345678 and previous 0x12345678 state shifted Causealong ? Kernel Mode is Status KUo IEo KUp IEp KUc IEc set, and ? ? 1 1 0 0 previous mode shifted along 32
- 33. Hardware exception handling PC EPC 0x12345678 0x12345678 Cause 0 Status KUo IEo KUp IEp KUc IEc ? ? 1 1 0 0 Code for the exception placed in Cause. Note Interrupt code = 0 33
- 34. Hardware exception handling PC EPC 0x80000080 0x12345678 Cause 0 Status KUo IEo KUp IEp KUc IEc ? ? 1 1 0 0 Address of general exception vector placed in PC 34
- 35. Hardware exception handling PC EPC 0x80000080 0x12345678 • CPU is now running in Cause kernel mode at 0x80000080, with 0 interrupts disabled Status KUo IEo KUp IEp KUc IEc • All information required to – Find out what caused the ? ? 1 1 0 0 exception – Restart after exception Badvaddr handling is in coprocessor registers 35
- 36. Returning from an exception • For now, lets ignore – how the exception is actually handled – how user-level registers are preserved • Let’s simply look at how we return from the exception 36
- 37. Returning from an exception PC EPC 0x80001234 0x12345678 • This code to return is Cause 0 lw r27, saved_epc Status KUo IEo KUp IEp KUc IEc nop jr r27 Load the contents1of 0 0 ? ? 1 rfe EPC which is usually moved earlier to somewhere in memory by the exception handler 37
- 38. Returning from an exception PC EPC 0x12345678 0x12345678 • This code to return is Cause 0 lw r27, saved_epc Status KUo IEo KUp IEp KUc IEc nop jr r27 ? ? 1 1 0 0 Store the EPC back in rfe the PC 38
- 39. Returning from an exception PC EPC 0x12345678 In the branch delay slot, 0x12345678 execute a restore from • This code to return is Causeinstruction exception 0 lw r27, saved_epc Status KUo IEo KUp IEp KUc IEc nop jr r27 ? ? ? ? 1 1 rfe 39
- 40. Returning from an exception PC EPC 0x12345678 0x12345678 • We are now back in the Cause same state we were in 0 when the exception happened Status KUo IEo KUp IEp KUc IEc ? ? ? ? 1 1 40
- 41. Function Stack Frames • Each function call allocates Stack a new stack frame for local Frame variables, the return Pointer address, previous frame f1() stack pointer etc. frame – Frame pointer: start of Stack current stack frame Pointer – Stack pointer: end of current stack frame • Example: assume f1() calls f2(), which calls f3(). 41
- 42. Function Stack Frames • Each function call allocates Stack a new stack frame for local variables, the return address, previous frame f1() stack pointer etc. frame – Frame pointer: start of Frame current stack frame Pointer – Stack pointer: end of current f2() stack stack frame frame Stack • Example: assume f1() calls Pointer f2(), which calls f3(). 42
- 43. Function Stack Frames • Each function call allocates Stack a new stack frame for local variables, the return address, previous frame f1() stack pointer etc. frame – Frame pointer: start of current stack frame – Stack pointer: end of current f2() stack stack frame Frame frame • Example: assume f1() calls Pointer f2(), which calls f3(). f3() stack Stack frame Pointer 43
- 44. Compiler Register Conventions • Given 32 registers, which registers are used for – Local variables? – Argument passing? – Function call results? – Stack Pointer? 44
- 46. Stack Frame • MIPS calling convention for gcc – Args 1-4 have space reserved for them 46
- 47. Example Code main () int sixargs(int a, int b, { int c, int d, int e, int i; int f) { i = return a + b + c + d sixargs(1,2,3,4,5,6); + e + f; } } 47
- 48. 0040011c <main>: 40011c: 27bdffd8 addiu sp,sp,-40 400120: afbf0024 sw ra,36(sp) 400124: afbe0020 sw s8,32(sp) 400128: 03a0f021 move s8,sp 40012c: 24020005 li v0,5 400130: afa20010 sw v0,16(sp) 400134: 24020006 li v0,6 400138: afa20014 sw v0,20(sp) 40013c: 24040001 li a0,1 400140: 24050002 li a1,2 400144: 24060003 li a2,3 400148: 0c10002c jal 4000b0 <sixargs> 40014c: 24070004 li a3,4 400150: afc20018 sw v0,24(s8) 400154: 03c0e821 move sp,s8 400158: 8fbf0024 lw ra,36(sp) 40015c: 8fbe0020 lw s8,32(sp) 400160: 03e00008 jr ra 400164: 27bd0028 addiu sp,sp,40 ... 48
- 49. 004000b0 <sixargs>: 4000b0: 27bdfff8 addiu sp,sp,-8 4000b4: afbe0000 sw s8,0(sp) 4000b8: 03a0f021 move s8,sp 4000bc: afc40008 sw a0,8(s8) 4000c0: afc5000c sw a1,12(s8) 4000c4: afc60010 sw a2,16(s8) 4000c8: afc70014 sw a3,20(s8) 4000cc: 8fc30008 lw v1,8(s8) 4000d0: 8fc2000c lw v0,12(s8) 4000d4: 00000000 nop 4000d8: 00621021 addu v0,v1,v0 4000dc: 8fc30010 lw v1,16(s8) 4000e0: 00000000 nop 4000e4: 00431021 addu v0,v0,v1 4000e8: 8fc30014 lw v1,20(s8) 4000ec: 00000000 nop 4000f0: 00431021 addu v0,v0,v1 4000f4: 8fc30018 lw v1,24(s8) 4000f8: 00000000 nop 49
- 50. 4000fc: 00431021 addu v0,v0,v1 400100: 8fc3001c lw v1,28(s8) 400104: 00000000 nop 400108: 00431021 addu v0,v0,v1 40010c: 03c0e821 move sp,s8 400110: 8fbe0000 lw s8,0(sp) 400114: 03e00008 jr ra 400118: 27bd0008 addiu sp,sp,8 50
- 51. System Calls Continued 51
- 52. User and Kernel Execution • Simplistically, execution state consists of – Registers, processor mode, PC, SP • User applications and the kernel have their own execution state. • System call mechanism safely transfers from user execution to kernel execution and back. 52
- 53. System Call Mechanism in Principle • Processor mode – Switched from user-mode to kernel-mode • Switched back when returning to user mode • SP – User-level SP is saved and a kernel SP is initialised • User-level SP restored when returning to user-mode • PC – User-level PC is saved and PC set to kernel entry point • User-level PC restored when returning to user-level – Kernel entry via the designated entry point must be strictly enforced 53
- 54. System Call Mechanism in Principle • Registers – Set at user-level to indicate system call type and its arguments • A convention between applications and the kernel – Some registers are preserved at user-level or kernel- level in order to restart user-level execution • Depends on language calling convention etc. – Result of system call placed in registers when returning to user-level • Another convention 54
- 55. Why do we need system calls? • Why not simply jump into the kernel via a function call???? – Function calls do not • Change from user to kernel mode – and eventually back again • Restrict possible entry points to secure locations 55
- 56. Steps in Making a System Call There are 11 steps in making the system call read (fd, buffer, nbytes) 56
- 57. MIPS System Calls • System calls are invoked via a syscall instruction. – The syscall instruction causes an exception and transfers control to the general exception handler – A convention (an agreement between the kernel and applications) is required as to how user-level software indicates • Which system call is required • Where its arguments are • Where the result should go 57
- 58. OS/161 Systems Calls • OS/161 uses the following conventions – Arguments are passed and returned via the normal C function calling convention – Additionally • Reg v0 contains the system call number • On return, reg a3 contains – 0: if success, v0 contains successful result – not 0: if failure, v0 has the errno. » v0 stored in errno » -1 returned in v0 58
- 59. ra Preserved ra fp fp Convention Convention for sp for kernel kernel entry sp gp Preserved for gp exit k1 C calling convention k1 k0 k0 s7 Preserved s7 M M s0 s0 t9 t9 M M t0 Success? t0 a3 Args in a3 a2 a2 a1 a1 a0 a0 v1 Result v1 v0 v0 AT AT zero SysCall No. zero 59
- 60. • Seriously low- level code follows move a0,s3 • This code is not addiu a1,sp,16 for the faint jal 40068c <read> hearted li a2,1024 move s0,v0 blez s0,400194 <docat+0x94> 60
- 61. User-Level System Call Walk Through – Calling read() int read(int filehandle, void *buffer, size_t size) • Three arguments, one return value • Code fragment calling the read function 400124: 02602021 move a0,s3 400128: 27a50010 addiu a1,sp,16 40012c: 0c1001a3 jal 40068c <read> 400130: 24060400 li a2,1024 400134: 00408021 move s0,v0 400138: 1a000016 blez s0,400194 <docat+0x94> • Args are loaded, return value is tested 61
- 62. Inside the read() syscall function part 1 0040068c <read>: 40068c: 08100190 j 400640 <__syscall> 400690: 24020005 li v0,5 • Appropriate registers are preserved – Arguments (a0-a3), return address (ra), etc. • The syscall number (5) is loaded into v0 • Jump (not jump and link) to the common syscall routine 62
- 63. The read() syscall function part 2 Generate a syscall exception 00400640 <__syscall>: 400640: 0000000c syscall 400644: 10e00005 beqz a3,40065c <__syscall+0x1c> 400648: 00000000 nop 40064c: 3c011000 lui at,0x1000 400650: ac220000 sw v0,0(at) 400654: 2403ffff li v1,-1 400658: 2402ffff li v0,-1 40065c: 03e00008 jr ra 400660: 00000000 nop 63
- 64. The read() syscall function Test success, if yes, part 2 branch to return 00400640 <__syscall>: from function 400640: 0000000c syscall 400644: 10e00005 beqz a3,40065c <__syscall+0x1c> 400648: 00000000 nop 40064c: 3c011000 lui at,0x1000 400650: ac220000 sw v0,0(at) 400654: 2403ffff li v1,-1 400658: 2402ffff li v0,-1 40065c: 03e00008 jr ra 400660: 00000000 nop 64
- 65. The read() syscall function part 2 00400640 <__syscall>: If failure, store code 400640: 0000000c syscall in errno 400644: 10e00005 beqz a3,40065c <__syscall+0x1c> 400648: 00000000 nop 40064c: 3c011000 lui at,0x1000 400650: ac220000 sw v0,0(at) 400654: 2403ffff li v1,-1 400658: 2402ffff li v0,-1 40065c: 03e00008 jr ra 400660: 00000000 nop 65
- 66. The read() syscall function part 2 00400640 <__syscall>: Set read() result to 400640: 0000000c syscall -1 400644: 10e00005 beqz a3,40065c <__syscall+0x1c> 400648: 00000000 nop 40064c: 3c011000 lui at,0x1000 400650: ac220000 sw v0,0(at) 400654: 2403ffff li v1,-1 400658: 2402ffff li v0,-1 40065c: 03e00008 jr ra 400660: 00000000 nop 66
- 67. The read() syscall function part 2 00400640 <__syscall>: Return to location 400640: 0000000c syscall after where read() 400644: 10e00005 was called beqz a3,40065c <__syscall+0x1c> 400648: 00000000 nop 40064c: 3c011000 lui at,0x1000 400650: ac220000 sw v0,0(at) 400654: 2403ffff li v1,-1 400658: 2402ffff li v0,-1 40065c: 03e00008 jr ra 400660: 00000000 nop 67
- 68. Summary • From the caller’s perspective, the read() system call behaves like a normal function call – It preserves the calling convention of the language • However, the actual function implements its own convention by agreement with the kernel – Our OS/161 example assumes the kernel preserves appropriate registers(s0-s8, sp, gp, ra). • Most languages have similar support libraries that interface with the operating system. 68
- 69. System Calls - Kernel Side • Things left to do – Change to kernel stack – Preserve registers by saving to memory (the stack) – Leave saved registers somewhere accessible to • Read arguments • Store return values – Do the “read()” – Restore registers – Switch back to user stack – Return to application 69
- 70. exception: move k1, sp /* Save previous stack pointer in k1 */ mfc0 k0, c0_status /* Get status register */ andi k0, k0, CST_Kup /* Check the we-were-in-user-mode bit */ beq k0, $0, 1f /* If clear, from kernel, already have stack */ nop /* delay slot */ /* Coming from user mode - load kernel stack into sp */ la k0, curkstack /* get address of "curkstack" */ lw sp, 0(k0) /* get its value */ nop /* Note k0, k1 delay slot for the load */ registers 1: mfc0 k0, c0_cause /* available for Now, load the exception cause. */ j common_exception /* kernel use Skip to common code */ nop /* delay slot */ 70
- 71. exception: move k1, sp /* Save previous stack pointer in k1 */ mfc0 k0, c0_status /* Get status register */ andi k0, k0, CST_Kup /* Check the we-were-in-user-mode bit */ beq k0, $0, 1f /* If clear, from kernel, already have stack */ nop /* delay slot */ /* Coming from user mode - load kernel stack into sp */ la k0, curkstack /* get address of "curkstack" */ lw sp, 0(k0) /* get its value */ nop /* delay slot for the load */ 1: mfc0 k0, c0_cause /* Now, load the exception cause. */ j common_exception /* Skip to common code */ nop /* delay slot */ 71
- 72. common_exception: /* * At this point: * Interrupts are off. (The processor did this for us.) * k0 contains the exception cause value. * k1 contains the old stack pointer. * sp points into the kernel stack. * All other registers are untouched. */ /* * Allocate stack space for 37 words to hold the trap frame, * plus four more words for a minimal argument block. */ addi sp, sp, -164 72
- 73. /* The order here must match mips/include/trapframe.h. */ sw ra, 160(sp) /* dummy for gdb */ sw s8, 156(sp) /* save s8 */ sw sp, 152(sp) /* dummy for gdb */ sw gp, 148(sp) /* save gp */ sw k1, 144(sp) /* dummy for gdb */ sw k0, 140(sp) /* dummy for gdb */ These six stores are a “hack” to avoid sw k1, 152(sp) /* real saved sp */ nop /* delay slot for store */ confusing GDB You can ignore the mfc0 k1, c0_epc /* Copr.0 reg 13 == PC for exception of why and details */ sw k1, 160(sp) /* real saved PC */ how 73
- 74. /* The order here must match mips/include/trapframe.h. */ sw ra, 160(sp) /* dummy for gdb */ sw s8, 156(sp) /* save s8 */ The real work starts sw sp, 152(sp) /* dummy for gdb */ here sw gp, 148(sp) /* save gp */ sw k1, 144(sp) /* dummy for gdb */ sw k0, 140(sp) /* dummy for gdb */ sw k1, 152(sp) /* real saved sp */ nop /* delay slot for store */ mfc0 k1, c0_epc /* Copr.0 reg 13 == PC for exception */ sw k1, 160(sp) /* real saved PC */ 74
- 75. sw t9, 136(sp) sw t8, 132(sp) sw s7, 128(sp) sw s6, 124(sp) sw s5, 120(sp) Save all the registers sw s4, 116(sp) on the kernel stack sw s3, 112(sp) sw s2, 108(sp) sw s1, 104(sp) sw s0, 100(sp) sw t7, 96(sp) sw t6, 92(sp) sw t5, 88(sp) sw t4, 84(sp) sw t3, 80(sp) sw t2, 76(sp) sw t1, 72(sp) sw t0, 68(sp) sw a3, 64(sp) sw a2, 60(sp) sw a1, 56(sp) sw a0, 52(sp) sw v1, 48(sp) sw v0, 44(sp) sw AT, 40(sp) sw ra, 36(sp) 75
- 76. /* * Save special registers. */ mfhi t0 We can now use the mflo t1 other registers (t0, t1) sw t0, 32(sp) that we have sw t1, 28(sp) preserved on the stack /* * Save remaining exception context information. */ sw k0, 24(sp) /* k0 was loaded with cause earlier */ mfc0 t1, c0_status /* Copr.0 reg 11 == status */ sw t1, 20(sp) mfc0 t2, c0_vaddr /* Copr.0 reg 8 == faulting vaddr */ sw t2, 16(sp) /* * Pretend to save $0 for gdb's benefit. */ sw $0, 12(sp) 76
- 77. /* * Prepare to call mips_trap(struct trapframe *) */ addiu a0, sp, 16 /* set argument */ jal mips_trap /* call it */ nop /* delay slot */ Create a pointer to the base of the saved registers and state in the first argument register 77
- 78. struct trapframe { u_int32_t tf_vaddr; /* vaddr register */ Kernel Stack u_int32_t tf_status; /* status register */ u_int32_t tf_cause; /* cause register */ u_int32_t tf_lo; u_int32_t tf_hi; u_int32_t tf_ra;/* Saved register 31 */ u_int32_t tf_at;/* Saved register 1 (AT) */ epc u_int32_t tf_v0;/* Saved register 2 (v0) */ s8 u_int32_t tf_v1;/* etc. */ sp u_int32_t tf_a0; u_int32_t tf_a1; By creating a pointer to gp u_int32_t tf_a2; k1 u_int32_t tf_a3; here of type struct k0 u_int32_t tf_t0; M trapframe *, we can t9 u_int32_t tf_t7; access the user’s saved t8 u_int32_t tf_s0; M registers as normal M u_int32_t tf_s7; variables within ‘C’ at u_int32_t tf_t8; u_int32_t tf_t9; ra u_int32_t tf_k0;/* dummy (see exception.S comments) */ hi u_int32_t tf_k1;/* dummy */ lo u_int32_t tf_gp; u_int32_t tf_sp; cause u_int32_t tf_s8; status u_int32_t tf_epc; /* coprocessor 0 epc register */ vaddr }; 78
- 79. /* Now we arrive in the ‘C’ kernel * General trap (exception) handling function for mips. * This is called by the assembly-language exception handler once * the trapframe has been set up. */ void mips_trap(struct trapframe *tf) { u_int32_t code, isutlb, iskern; int savespl; /* The trap frame is supposed to be 37 registers long. */ assert(sizeof(struct trapframe)==(37*4)); /* Save the value of curspl, which belongs to the old context. */ savespl = curspl; /* Right now, interrupts should be off. */ curspl = SPL_HIGH; 79
- 80. What happens next? • The kernel deals with whatever caused the exception – Syscall – Interrupt – Page fault – It potentially modifies the trapframe, etc • E.g., Store return code in v0, zero in a3 • ‘mips_trap’ eventually returns 80
- 81. exception_return: /* 16(sp) no need to restore tf_vaddr */ lw t0, 20(sp) /* load status register value into t0 */ nop /* load delay slot */ mtc0 t0, c0_status /* store it back to coprocessor 0 */ /* 24(sp) no need to restore tf_cause */ /* restore special registers */ lw t1, 28(sp) lw t0, 32(sp) mtlo t1 mthi t0 /* load the general registers */ lw ra, 36(sp) lw AT, 40(sp) lw v0, 44(sp) lw v1, 48(sp) lw a0, 52(sp) lw a1, 56(sp) lw a2, 60(sp) lw a3, 64(sp) 81
- 82. lw t0, 68(sp) lw t1, 72(sp) lw t2, 76(sp) lw t3, 80(sp) lw t4, 84(sp) lw t5, 88(sp) lw t6, 92(sp) lw t7, 96(sp) lw s0, 100(sp) lw s1, 104(sp) lw s2, 108(sp) lw s3, 112(sp) lw s4, 116(sp) lw s5, 120(sp) lw s6, 124(sp) lw s7, 128(sp) lw t8, 132(sp) lw t9, 136(sp) /* 140(sp) "saved" k0 was dummy garbage anyway */ /* 144(sp) "saved" k1 was dummy garbage anyway */ 82
- 83. lw gp, 148(sp) /* restore gp */ /* 152(sp) stack pointer - below */ lw s8, 156(sp) /* restore s8 */ lw k0, 160(sp) /* fetch exception return PC into k0 */ lw sp, 152(sp) /* fetch saved sp (must be last) */ /* done */ jr k0 /* jump back */ rfe /* in delay slot */ .end common_exception Note again that only k0, k1 have been trashed 83