From 9d02a4283e9ce4e9ca11ff00615bdacdb0515a1a Mon Sep 17 00:00:00 2001 From: Jonas Bonn Date: Sat, 4 Jun 2011 11:05:39 +0300 Subject: [PATCH] OpenRISC: Boot code Architecture code and early setup routines for booting Linux. Signed-off-by: Jonas Bonn Reviewed-by: Arnd Bergmann --- arch/openrisc/kernel/entry.S | 1128 +++++++++++++++++++++ arch/openrisc/kernel/head.S | 1607 ++++++++++++++++++++++++++++++ arch/openrisc/kernel/init_task.c | 41 + arch/openrisc/kernel/setup.c | 381 +++++++ 4 files changed, 3157 insertions(+) create mode 100644 arch/openrisc/kernel/entry.S create mode 100644 arch/openrisc/kernel/head.S create mode 100644 arch/openrisc/kernel/init_task.c create mode 100644 arch/openrisc/kernel/setup.c diff --git a/arch/openrisc/kernel/entry.S b/arch/openrisc/kernel/entry.S new file mode 100644 index 000000000000..d5f9c35a583f --- /dev/null +++ b/arch/openrisc/kernel/entry.S @@ -0,0 +1,1128 @@ +/* + * OpenRISC entry.S + * + * Linux architectural port borrowing liberally from similar works of + * others. All original copyrights apply as per the original source + * declaration. + * + * Modifications for the OpenRISC architecture: + * Copyright (C) 2003 Matjaz Breskvar + * Copyright (C) 2005 Gyorgy Jeney + * Copyright (C) 2010-2011 Jonas Bonn + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ + +#include + +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#define DISABLE_INTERRUPTS(t1,t2) \ + l.mfspr t2,r0,SPR_SR ;\ + l.movhi t1,hi(~(SPR_SR_IEE|SPR_SR_TEE)) ;\ + l.ori t1,t1,lo(~(SPR_SR_IEE|SPR_SR_TEE)) ;\ + l.and t2,t2,t1 ;\ + l.mtspr r0,t2,SPR_SR + +#define ENABLE_INTERRUPTS(t1) \ + l.mfspr t1,r0,SPR_SR ;\ + l.ori t1,t1,lo(SPR_SR_IEE|SPR_SR_TEE) ;\ + l.mtspr r0,t1,SPR_SR + +/* =========================================================[ macros ]=== */ + +/* + * We need to disable interrupts at beginning of RESTORE_ALL + * since interrupt might come in after we've loaded EPC return address + * and overwrite EPC with address somewhere in RESTORE_ALL + * which is of course wrong! + */ + +#define RESTORE_ALL \ + DISABLE_INTERRUPTS(r3,r4) ;\ + l.lwz r3,PT_PC(r1) ;\ + l.mtspr r0,r3,SPR_EPCR_BASE ;\ + l.lwz r3,PT_SR(r1) ;\ + l.mtspr r0,r3,SPR_ESR_BASE ;\ + l.lwz r2,PT_GPR2(r1) ;\ + l.lwz r3,PT_GPR3(r1) ;\ + l.lwz r4,PT_GPR4(r1) ;\ + l.lwz r5,PT_GPR5(r1) ;\ + l.lwz r6,PT_GPR6(r1) ;\ + l.lwz r7,PT_GPR7(r1) ;\ + l.lwz r8,PT_GPR8(r1) ;\ + l.lwz r9,PT_GPR9(r1) ;\ + l.lwz r10,PT_GPR10(r1) ;\ + l.lwz r11,PT_GPR11(r1) ;\ + l.lwz r12,PT_GPR12(r1) ;\ + l.lwz r13,PT_GPR13(r1) ;\ + l.lwz r14,PT_GPR14(r1) ;\ + l.lwz r15,PT_GPR15(r1) ;\ + l.lwz r16,PT_GPR16(r1) ;\ + l.lwz r17,PT_GPR17(r1) ;\ + l.lwz r18,PT_GPR18(r1) ;\ + l.lwz r19,PT_GPR19(r1) ;\ + l.lwz r20,PT_GPR20(r1) ;\ + l.lwz r21,PT_GPR21(r1) ;\ + l.lwz r22,PT_GPR22(r1) ;\ + l.lwz r23,PT_GPR23(r1) ;\ + l.lwz r24,PT_GPR24(r1) ;\ + l.lwz r25,PT_GPR25(r1) ;\ + l.lwz r26,PT_GPR26(r1) ;\ + l.lwz r27,PT_GPR27(r1) ;\ + l.lwz r28,PT_GPR28(r1) ;\ + l.lwz r29,PT_GPR29(r1) ;\ + l.lwz r30,PT_GPR30(r1) ;\ + l.lwz r31,PT_GPR31(r1) ;\ + l.lwz r1,PT_SP(r1) ;\ + l.rfe + + +#define EXCEPTION_ENTRY(handler) \ + .global handler ;\ +handler: ;\ + /* r1, EPCR, ESR a already saved */ ;\ + l.sw PT_GPR2(r1),r2 ;\ + l.sw PT_GPR3(r1),r3 ;\ + l.sw PT_ORIG_GPR11(r1),r11 ;\ + /* r4 already save */ ;\ + l.sw PT_GPR5(r1),r5 ;\ + l.sw PT_GPR6(r1),r6 ;\ + l.sw PT_GPR7(r1),r7 ;\ + l.sw PT_GPR8(r1),r8 ;\ + l.sw PT_GPR9(r1),r9 ;\ + /* r10 already saved */ ;\ + l.sw PT_GPR11(r1),r11 ;\ + /* r12 already saved */ ;\ + l.sw PT_GPR13(r1),r13 ;\ + l.sw PT_GPR14(r1),r14 ;\ + l.sw PT_GPR15(r1),r15 ;\ + l.sw PT_GPR16(r1),r16 ;\ + l.sw PT_GPR17(r1),r17 ;\ + l.sw PT_GPR18(r1),r18 ;\ + l.sw PT_GPR19(r1),r19 ;\ + l.sw PT_GPR20(r1),r20 ;\ + l.sw PT_GPR21(r1),r21 ;\ + l.sw PT_GPR22(r1),r22 ;\ + l.sw PT_GPR23(r1),r23 ;\ + l.sw PT_GPR24(r1),r24 ;\ + l.sw PT_GPR25(r1),r25 ;\ + l.sw PT_GPR26(r1),r26 ;\ + l.sw PT_GPR27(r1),r27 ;\ + l.sw PT_GPR28(r1),r28 ;\ + l.sw PT_GPR29(r1),r29 ;\ + /* r30 already save */ ;\ +/* l.sw PT_GPR30(r1),r30*/ ;\ + l.sw PT_GPR31(r1),r31 ;\ + l.sw PT_SYSCALLNO(r1),r0 + +#define UNHANDLED_EXCEPTION(handler,vector) \ + .global handler ;\ +handler: ;\ + /* r1, EPCR, ESR already saved */ ;\ + l.sw PT_GPR2(r1),r2 ;\ + l.sw PT_GPR3(r1),r3 ;\ + l.sw PT_ORIG_GPR11(r1),r11 ;\ + l.sw PT_GPR5(r1),r5 ;\ + l.sw PT_GPR6(r1),r6 ;\ + l.sw PT_GPR7(r1),r7 ;\ + l.sw PT_GPR8(r1),r8 ;\ + l.sw PT_GPR9(r1),r9 ;\ + /* r10 already saved */ ;\ + l.sw PT_GPR11(r1),r11 ;\ + /* r12 already saved */ ;\ + l.sw PT_GPR13(r1),r13 ;\ + l.sw PT_GPR14(r1),r14 ;\ + l.sw PT_GPR15(r1),r15 ;\ + l.sw PT_GPR16(r1),r16 ;\ + l.sw PT_GPR17(r1),r17 ;\ + l.sw PT_GPR18(r1),r18 ;\ + l.sw PT_GPR19(r1),r19 ;\ + l.sw PT_GPR20(r1),r20 ;\ + l.sw PT_GPR21(r1),r21 ;\ + l.sw PT_GPR22(r1),r22 ;\ + l.sw PT_GPR23(r1),r23 ;\ + l.sw PT_GPR24(r1),r24 ;\ + l.sw PT_GPR25(r1),r25 ;\ + l.sw PT_GPR26(r1),r26 ;\ + l.sw PT_GPR27(r1),r27 ;\ + l.sw PT_GPR28(r1),r28 ;\ + l.sw PT_GPR29(r1),r29 ;\ + /* r31 already saved */ ;\ + l.sw PT_GPR30(r1),r30 ;\ +/* l.sw PT_GPR31(r1),r31 */ ;\ + l.sw PT_SYSCALLNO(r1),r0 ;\ + l.addi r3,r1,0 ;\ + /* r4 is exception EA */ ;\ + l.addi r5,r0,vector ;\ + l.jal unhandled_exception ;\ + l.nop ;\ + l.j _ret_from_exception ;\ + l.nop + +/* + * NOTE: one should never assume that SPR_EPC, SPR_ESR, SPR_EEAR + * contain the same values as when exception we're handling + * occured. in fact they never do. if you need them use + * values saved on stack (for SPR_EPC, SPR_ESR) or content + * of r4 (for SPR_EEAR). for details look at EXCEPTION_HANDLE() + * in 'arch/or32/kernel/head.S' + */ + +/* =====================================================[ exceptions] === */ + +/* ---[ 0x100: RESET exception ]----------------------------------------- */ + +EXCEPTION_ENTRY(_tng_kernel_start) + l.jal _start + l.andi r0,r0,0 + +/* ---[ 0x200: BUS exception ]------------------------------------------- */ + +EXCEPTION_ENTRY(_bus_fault_handler) + /* r4: EA of fault (set by EXCEPTION_HANDLE) */ + l.jal do_bus_fault + l.addi r3,r1,0 /* pt_regs */ + + l.j _ret_from_exception + l.nop + +/* ---[ 0x300: Data Page Fault exception ]------------------------------- */ + +EXCEPTION_ENTRY(_data_page_fault_handler) + /* set up parameters for do_page_fault */ + l.addi r3,r1,0 // pt_regs + /* r4 set be EXCEPTION_HANDLE */ // effective address of fault + l.ori r5,r0,0x300 // exception vector + + /* + * __PHX__: TODO + * + * all this can be written much simpler. look at + * DTLB miss handler in the CONFIG_GUARD_PROTECTED_CORE part + */ +#ifdef CONFIG_OPENRISC_NO_SPR_SR_DSX + l.lwz r6,PT_PC(r3) // address of an offending insn + l.lwz r6,0(r6) // instruction that caused pf + + l.srli r6,r6,26 // check opcode for jump insn + l.sfeqi r6,0 // l.j + l.bf 8f + l.sfeqi r6,1 // l.jal + l.bf 8f + l.sfeqi r6,3 // l.bnf + l.bf 8f + l.sfeqi r6,4 // l.bf + l.bf 8f + l.sfeqi r6,0x11 // l.jr + l.bf 8f + l.sfeqi r6,0x12 // l.jalr + l.bf 8f + + l.nop + + l.j 9f + l.nop +8: + + l.lwz r6,PT_PC(r3) // address of an offending insn + l.addi r6,r6,4 + l.lwz r6,0(r6) // instruction that caused pf + l.srli r6,r6,26 // get opcode +9: + +#else + + l.mfspr r6,r0,SPR_SR // SR +// l.lwz r6,PT_SR(r3) // ESR + l.andi r6,r6,SPR_SR_DSX // check for delay slot exception + l.sfeqi r6,0x1 // exception happened in delay slot + l.bnf 7f + l.lwz r6,PT_PC(r3) // address of an offending insn + + l.addi r6,r6,4 // offending insn is in delay slot +7: + l.lwz r6,0(r6) // instruction that caused pf + l.srli r6,r6,26 // check opcode for write access +#endif + + l.sfgeui r6,0x34 // check opcode for write access + l.bnf 1f + l.sfleui r6,0x37 + l.bnf 1f + l.ori r6,r0,0x1 // write access + l.j 2f + l.nop +1: l.ori r6,r0,0x0 // !write access +2: + + /* call fault.c handler in or32/mm/fault.c */ + l.jal do_page_fault + l.nop + l.j _ret_from_exception + l.nop + +/* ---[ 0x400: Insn Page Fault exception ]------------------------------- */ + +EXCEPTION_ENTRY(_insn_page_fault_handler) + /* set up parameters for do_page_fault */ + l.addi r3,r1,0 // pt_regs + /* r4 set be EXCEPTION_HANDLE */ // effective address of fault + l.ori r5,r0,0x400 // exception vector + l.ori r6,r0,0x0 // !write access + + /* call fault.c handler in or32/mm/fault.c */ + l.jal do_page_fault + l.nop + l.j _ret_from_exception + l.nop + + +/* ---[ 0x500: Timer exception ]----------------------------------------- */ + +EXCEPTION_ENTRY(_timer_handler) + l.jal timer_interrupt + l.addi r3,r1,0 /* pt_regs */ + + l.j _ret_from_intr + l.nop + +/* ---[ 0x600: Aligment exception ]-------------------------------------- */ + +EXCEPTION_ENTRY(_alignment_handler) + /* r4: EA of fault (set by EXCEPTION_HANDLE) */ + l.jal do_unaligned_access + l.addi r3,r1,0 /* pt_regs */ + + l.j _ret_from_exception + l.nop + +#if 0 +EXCEPTION_ENTRY(_aligment_handler) +// l.mfspr r2,r0,SPR_EEAR_BASE /* Load the efective addres */ + l.addi r2,r4,0 +// l.mfspr r5,r0,SPR_EPCR_BASE /* Load the insn address */ + l.lwz r5,PT_PC(r1) + + l.lwz r3,0(r5) /* Load insn */ + l.srli r4,r3,26 /* Shift left to get the insn opcode */ + + l.sfeqi r4,0x00 /* Check if the load/store insn is in delay slot */ + l.bf jmp + l.sfeqi r4,0x01 + l.bf jmp + l.sfeqi r4,0x03 + l.bf jmp + l.sfeqi r4,0x04 + l.bf jmp + l.sfeqi r4,0x11 + l.bf jr + l.sfeqi r4,0x12 + l.bf jr + l.nop + l.j 1f + l.addi r5,r5,4 /* Increment PC to get return insn address */ + +jmp: + l.slli r4,r3,6 /* Get the signed extended jump length */ + l.srai r4,r4,4 + + l.lwz r3,4(r5) /* Load the real load/store insn */ + + l.add r5,r5,r4 /* Calculate jump target address */ + + l.j 1f + l.srli r4,r3,26 /* Shift left to get the insn opcode */ + +jr: + l.slli r4,r3,9 /* Shift to get the reg nb */ + l.andi r4,r4,0x7c + + l.lwz r3,4(r5) /* Load the real load/store insn */ + + l.add r4,r4,r1 /* Load the jump register value from the stack */ + l.lwz r5,0(r4) + + l.srli r4,r3,26 /* Shift left to get the insn opcode */ + + +1: +// l.mtspr r0,r5,SPR_EPCR_BASE + l.sw PT_PC(r1),r5 + + l.sfeqi r4,0x26 + l.bf lhs + l.sfeqi r4,0x25 + l.bf lhz + l.sfeqi r4,0x22 + l.bf lws + l.sfeqi r4,0x21 + l.bf lwz + l.sfeqi r4,0x37 + l.bf sh + l.sfeqi r4,0x35 + l.bf sw + l.nop + +1: l.j 1b /* I don't know what to do */ + l.nop + +lhs: l.lbs r5,0(r2) + l.slli r5,r5,8 + l.lbz r6,1(r2) + l.or r5,r5,r6 + l.srli r4,r3,19 + l.andi r4,r4,0x7c + l.add r4,r4,r1 + l.j align_end + l.sw 0(r4),r5 + +lhz: l.lbz r5,0(r2) + l.slli r5,r5,8 + l.lbz r6,1(r2) + l.or r5,r5,r6 + l.srli r4,r3,19 + l.andi r4,r4,0x7c + l.add r4,r4,r1 + l.j align_end + l.sw 0(r4),r5 + +lws: l.lbs r5,0(r2) + l.slli r5,r5,24 + l.lbz r6,1(r2) + l.slli r6,r6,16 + l.or r5,r5,r6 + l.lbz r6,2(r2) + l.slli r6,r6,8 + l.or r5,r5,r6 + l.lbz r6,3(r2) + l.or r5,r5,r6 + l.srli r4,r3,19 + l.andi r4,r4,0x7c + l.add r4,r4,r1 + l.j align_end + l.sw 0(r4),r5 + +lwz: l.lbz r5,0(r2) + l.slli r5,r5,24 + l.lbz r6,1(r2) + l.slli r6,r6,16 + l.or r5,r5,r6 + l.lbz r6,2(r2) + l.slli r6,r6,8 + l.or r5,r5,r6 + l.lbz r6,3(r2) + l.or r5,r5,r6 + l.srli r4,r3,19 + l.andi r4,r4,0x7c + l.add r4,r4,r1 + l.j align_end + l.sw 0(r4),r5 + +sh: + l.srli r4,r3,9 + l.andi r4,r4,0x7c + l.add r4,r4,r1 + l.lwz r5,0(r4) + l.sb 1(r2),r5 + l.srli r5,r5,8 + l.j align_end + l.sb 0(r2),r5 + +sw: + l.srli r4,r3,9 + l.andi r4,r4,0x7c + l.add r4,r4,r1 + l.lwz r5,0(r4) + l.sb 3(r2),r5 + l.srli r5,r5,8 + l.sb 2(r2),r5 + l.srli r5,r5,8 + l.sb 1(r2),r5 + l.srli r5,r5,8 + l.j align_end + l.sb 0(r2),r5 + +align_end: + l.j _ret_from_intr + l.nop +#endif + +/* ---[ 0x700: Illegal insn exception ]---------------------------------- */ + +EXCEPTION_ENTRY(_illegal_instruction_handler) + /* r4: EA of fault (set by EXCEPTION_HANDLE) */ + l.jal do_illegal_instruction + l.addi r3,r1,0 /* pt_regs */ + + l.j _ret_from_exception + l.nop + +/* ---[ 0x800: External interrupt exception ]---------------------------- */ + +EXCEPTION_ENTRY(_external_irq_handler) +#ifdef CONFIG_OPENRISC_ESR_EXCEPTION_BUG_CHECK + l.lwz r4,PT_SR(r1) // were interrupts enabled ? + l.andi r4,r4,SPR_SR_IEE + l.sfeqi r4,0 + l.bnf 1f // ext irq enabled, all ok. + l.nop + + l.addi r1,r1,-0x8 + l.movhi r3,hi(42f) + l.ori r3,r3,lo(42f) + l.sw 0x0(r1),r3 + l.jal printk + l.sw 0x4(r1),r4 + l.addi r1,r1,0x8 + + .section .rodata, "a" +42: + .string "\n\rESR interrupt bug: in _external_irq_handler (ESR %x)\n\r" + .align 4 + .previous + + l.ori r4,r4,SPR_SR_IEE // fix the bug +// l.sw PT_SR(r1),r4 +1: +#endif + l.addi r3,r1,0 + l.movhi r8,hi(do_IRQ) + l.ori r8,r8,lo(do_IRQ) + l.jalr r8 + l.nop + l.j _ret_from_intr + l.nop + +/* ---[ 0x900: DTLB miss exception ]------------------------------------- */ + + +/* ---[ 0xa00: ITLB miss exception ]------------------------------------- */ + + +/* ---[ 0xb00: Range exception ]----------------------------------------- */ + +UNHANDLED_EXCEPTION(_vector_0xb00,0xb00) + +/* ---[ 0xc00: Syscall exception ]--------------------------------------- */ + +/* + * Syscalls are a special type of exception in that they are + * _explicitly_ invoked by userspace and can therefore be + * held to conform to the same ABI as normal functions with + * respect to whether registers are preserved across the call + * or not. + */ + +/* Upon syscall entry we just save the callee-saved registers + * and not the call-clobbered ones. + */ + +_string_syscall_return: + .string "syscall return %ld \n\r\0" + .align 4 + +ENTRY(_sys_call_handler) + /* syscalls run with interrupts enabled */ + ENABLE_INTERRUPTS(r29) // enable interrupts, r29 is temp + + /* r1, EPCR, ESR a already saved */ + l.sw PT_GPR2(r1),r2 + /* r3-r8 must be saved because syscall restart relies + * on us being able to restart the syscall args... technically + * they should be clobbered, otherwise + */ + l.sw PT_GPR3(r1),r3 + /* r4 already saved */ + /* r4 holds the EEAR address of the fault, load the original r4 */ + l.lwz r4,PT_GPR4(r1) + l.sw PT_GPR5(r1),r5 + l.sw PT_GPR6(r1),r6 + l.sw PT_GPR7(r1),r7 + l.sw PT_GPR8(r1),r8 + l.sw PT_GPR9(r1),r9 + /* r10 already saved */ + l.sw PT_GPR11(r1),r11 + l.sw PT_ORIG_GPR11(r1),r11 + /* r12,r13 already saved */ + + /* r14-r28 (even) aren't touched by the syscall fast path below + * so we don't need to save them. However, the functions that return + * to userspace via a call to switch() DO need to save these because + * switch() effectively clobbers them... saving these registers for + * such functions is handled in their syscall wrappers (see fork, vfork, + * and clone, below). + + /* r30 is the only register we clobber in the fast path */ + /* r30 already saved */ +/* l.sw PT_GPR30(r1),r30 */ + /* This is used by do_signal to determine whether to check for + * syscall restart or not */ + l.sw PT_SYSCALLNO(r1),r11 + +_syscall_check_trace_enter: + /* If TIF_SYSCALL_TRACE is set, then we want to do syscall tracing */ + l.lwz r30,TI_FLAGS(r10) + l.andi r30,r30,_TIF_SYSCALL_TRACE + l.sfne r30,r0 + l.bf _syscall_trace_enter + l.nop + +_syscall_check: + /* Ensure that the syscall number is reasonable */ + l.sfgeui r11,__NR_syscalls + l.bf _syscall_badsys + l.nop + +_syscall_call: + l.movhi r29,hi(sys_call_table) + l.ori r29,r29,lo(sys_call_table) + l.slli r11,r11,2 + l.add r29,r29,r11 + l.lwz r29,0(r29) + + l.jalr r29 + l.nop + +_syscall_return: + /* All syscalls return here... just pay attention to ret_from_fork + * which does it in a round-about way. + */ + l.sw PT_GPR11(r1),r11 // save return value + +#if 0 +_syscall_debug: + l.movhi r3,hi(_string_syscall_return) + l.ori r3,r3,lo(_string_syscall_return) + l.ori r27,r0,1 + l.sw -4(r1),r27 + l.sw -8(r1),r11 + l.addi r1,r1,-8 + l.movhi r27,hi(printk) + l.ori r27,r27,lo(printk) + l.jalr r27 + l.nop + l.addi r1,r1,8 +#endif + +_syscall_check_trace_leave: + /* r30 is a callee-saved register so this should still hold the + * _TIF_SYSCALL_TRACE flag from _syscall_check_trace_enter above... + * _syscall_trace_leave expects syscall result to be in pt_regs->r11. + */ + l.sfne r30,r0 + l.bf _syscall_trace_leave + l.nop + +/* This is where the exception-return code begins... interrupts need to be + * disabled the rest of the way here because we can't afford to miss any + * interrupts that set NEED_RESCHED or SIGNALPENDING... really true? */ + +_syscall_check_work: + /* Here we need to disable interrupts */ + DISABLE_INTERRUPTS(r27,r29) + l.lwz r30,TI_FLAGS(r10) + l.andi r30,r30,_TIF_WORK_MASK + l.sfne r30,r0 + + l.bnf _syscall_resume_userspace + l.nop + + /* Work pending follows a different return path, so we need to + * make sure that all the call-saved registers get into pt_regs + * before branching... + */ + l.sw PT_GPR14(r1),r14 + l.sw PT_GPR16(r1),r16 + l.sw PT_GPR18(r1),r18 + l.sw PT_GPR20(r1),r20 + l.sw PT_GPR22(r1),r22 + l.sw PT_GPR24(r1),r24 + l.sw PT_GPR26(r1),r26 + l.sw PT_GPR28(r1),r28 + + /* _work_pending needs to be called with interrupts disabled */ + l.j _work_pending + l.nop + +_syscall_resume_userspace: +// ENABLE_INTERRUPTS(r29) + + +/* This is the hot path for returning to userspace from a syscall. If there's + * work to be done and the branch to _work_pending was taken above, then the + * return to userspace will be done via the normal exception return path... + * that path restores _all_ registers and will overwrite the "clobbered" + * registers with whatever garbage is in pt_regs -- that's OK because those + * registers are clobbered anyway and because the extra work is insignificant + * in the context of the extra work that _work_pending is doing. + +/* Once again, syscalls are special and only guarantee to preserve the + * same registers as a normal function call */ + +/* The assumption here is that the registers r14-r28 (even) are untouched and + * don't need to be restored... be sure that that's really the case! + */ + +/* This is still too much... we should only be restoring what we actually + * clobbered... we should even be using 'scratch' (odd) regs above so that + * we don't need to restore anything, hardly... + */ + + l.lwz r2,PT_GPR2(r1) + + /* Restore args */ + /* r3-r8 are technically clobbered, but syscall restart needs these + * to be restored... + */ + l.lwz r3,PT_GPR3(r1) + l.lwz r4,PT_GPR4(r1) + l.lwz r5,PT_GPR5(r1) + l.lwz r6,PT_GPR6(r1) + l.lwz r7,PT_GPR7(r1) + l.lwz r8,PT_GPR8(r1) + + l.lwz r9,PT_GPR9(r1) + l.lwz r10,PT_GPR10(r1) + l.lwz r11,PT_GPR11(r1) + + /* r30 is the only register we clobber in the fast path */ + l.lwz r30,PT_GPR30(r1) + + /* Here we use r13-r19 (odd) as scratch regs */ + l.lwz r13,PT_PC(r1) + l.lwz r15,PT_SR(r1) + l.lwz r1,PT_SP(r1) + /* Interrupts need to be disabled for setting EPCR and ESR + * so that another interrupt doesn't come in here and clobber + * them before we can use them for our l.rfe */ + DISABLE_INTERRUPTS(r17,r19) + l.mtspr r0,r13,SPR_EPCR_BASE + l.mtspr r0,r15,SPR_ESR_BASE + l.rfe + +/* End of hot path! + * Keep the below tracing and error handling out of the hot path... +*/ + +_syscall_trace_enter: + /* Here we pass pt_regs to do_syscall_trace_enter. Make sure + * that function is really getting all the info it needs as + * pt_regs isn't a complete set of userspace regs, just the + * ones relevant to the syscall... + * + * Note use of delay slot for setting argument. + */ + l.jal do_syscall_trace_enter + l.addi r3,r1,0 + + /* Restore arguments (not preserved across do_syscall_trace_enter) + * so that we can do the syscall for real and return to the syscall + * hot path. + */ + l.lwz r11,PT_SYSCALLNO(r1) + l.lwz r3,PT_GPR3(r1) + l.lwz r4,PT_GPR4(r1) + l.lwz r5,PT_GPR5(r1) + l.lwz r6,PT_GPR6(r1) + l.lwz r7,PT_GPR7(r1) + + l.j _syscall_check + l.lwz r8,PT_GPR8(r1) + +_syscall_trace_leave: + l.jal do_syscall_trace_leave + l.addi r3,r1,0 + + l.j _syscall_check_work + l.nop + +_syscall_badsys: + /* Here we effectively pretend to have executed an imaginary + * syscall that returns -ENOSYS and then return to the regular + * syscall hot path. + * Note that "return value" is set in the delay slot... + */ + l.j _syscall_return + l.addi r11,r0,-ENOSYS + +/******* END SYSCALL HANDLING *******/ + +/* ---[ 0xd00: Trap exception ]------------------------------------------ */ + +UNHANDLED_EXCEPTION(_vector_0xd00,0xd00) + +/* ---[ 0xe00: Trap exception ]------------------------------------------ */ + +EXCEPTION_ENTRY(_trap_handler) + /* r4: EA of fault (set by EXCEPTION_HANDLE) */ + l.jal do_trap + l.addi r3,r1,0 /* pt_regs */ + + l.j _ret_from_exception + l.nop + +/* ---[ 0xf00: Reserved exception ]-------------------------------------- */ + +UNHANDLED_EXCEPTION(_vector_0xf00,0xf00) + +/* ---[ 0x1000: Reserved exception ]------------------------------------- */ + +UNHANDLED_EXCEPTION(_vector_0x1000,0x1000) + +/* ---[ 0x1100: Reserved exception ]------------------------------------- */ + +UNHANDLED_EXCEPTION(_vector_0x1100,0x1100) + +/* ---[ 0x1200: Reserved exception ]------------------------------------- */ + +UNHANDLED_EXCEPTION(_vector_0x1200,0x1200) + +/* ---[ 0x1300: Reserved exception ]------------------------------------- */ + +UNHANDLED_EXCEPTION(_vector_0x1300,0x1300) + +/* ---[ 0x1400: Reserved exception ]------------------------------------- */ + +UNHANDLED_EXCEPTION(_vector_0x1400,0x1400) + +/* ---[ 0x1500: Reserved exception ]------------------------------------- */ + +UNHANDLED_EXCEPTION(_vector_0x1500,0x1500) + +/* ---[ 0x1600: Reserved exception ]------------------------------------- */ + +UNHANDLED_EXCEPTION(_vector_0x1600,0x1600) + +/* ---[ 0x1700: Reserved exception ]------------------------------------- */ + +UNHANDLED_EXCEPTION(_vector_0x1700,0x1700) + +/* ---[ 0x1800: Reserved exception ]------------------------------------- */ + +UNHANDLED_EXCEPTION(_vector_0x1800,0x1800) + +/* ---[ 0x1900: Reserved exception ]------------------------------------- */ + +UNHANDLED_EXCEPTION(_vector_0x1900,0x1900) + +/* ---[ 0x1a00: Reserved exception ]------------------------------------- */ + +UNHANDLED_EXCEPTION(_vector_0x1a00,0x1a00) + +/* ---[ 0x1b00: Reserved exception ]------------------------------------- */ + +UNHANDLED_EXCEPTION(_vector_0x1b00,0x1b00) + +/* ---[ 0x1c00: Reserved exception ]------------------------------------- */ + +UNHANDLED_EXCEPTION(_vector_0x1c00,0x1c00) + +/* ---[ 0x1d00: Reserved exception ]------------------------------------- */ + +UNHANDLED_EXCEPTION(_vector_0x1d00,0x1d00) + +/* ---[ 0x1e00: Reserved exception ]------------------------------------- */ + +UNHANDLED_EXCEPTION(_vector_0x1e00,0x1e00) + +/* ---[ 0x1f00: Reserved exception ]------------------------------------- */ + +UNHANDLED_EXCEPTION(_vector_0x1f00,0x1f00) + +/* ========================================================[ return ] === */ + +_work_pending: + /* + * if (current_thread_info->flags & _TIF_NEED_RESCHED) + * schedule(); + */ + l.lwz r5,TI_FLAGS(r10) + l.andi r3,r5,_TIF_NEED_RESCHED + l.sfnei r3,0 + l.bnf _work_notifysig + l.nop + l.jal schedule + l.nop + l.j _resume_userspace + l.nop + +/* Handle pending signals and notify-resume requests. + * do_notify_resume must be passed the latest pushed pt_regs, not + * necessarily the "userspace" ones. Also, pt_regs->syscallno + * must be set so that the syscall restart functionality works. + */ +_work_notifysig: + l.jal do_notify_resume + l.ori r3,r1,0 /* pt_regs */ + +_resume_userspace: + DISABLE_INTERRUPTS(r3,r4) + l.lwz r3,TI_FLAGS(r10) + l.andi r3,r3,_TIF_WORK_MASK + l.sfnei r3,0 + l.bf _work_pending + l.nop + +_restore_all: + RESTORE_ALL + /* This returns to userspace code */ + + +ENTRY(_ret_from_intr) +ENTRY(_ret_from_exception) + l.lwz r4,PT_SR(r1) + l.andi r3,r4,SPR_SR_SM + l.sfeqi r3,0 + l.bnf _restore_all + l.nop + l.j _resume_userspace + l.nop + +ENTRY(ret_from_fork) + l.jal schedule_tail + l.nop + + /* _syscall_returns expect r11 to contain return value */ + l.lwz r11,PT_GPR11(r1) + + /* The syscall fast path return expects call-saved registers + * r12-r28 to be untouched, so we restore them here as they + * will have been effectively clobbered when arriving here + * via the call to switch() + */ + l.lwz r12,PT_GPR12(r1) + l.lwz r14,PT_GPR14(r1) + l.lwz r16,PT_GPR16(r1) + l.lwz r18,PT_GPR18(r1) + l.lwz r20,PT_GPR20(r1) + l.lwz r22,PT_GPR22(r1) + l.lwz r24,PT_GPR24(r1) + l.lwz r26,PT_GPR26(r1) + l.lwz r28,PT_GPR28(r1) + + l.j _syscall_return + l.nop + +/* Since syscalls don't save call-clobbered registers, the args to + * kernel_thread_helper will need to be passed through callee-saved + * registers and copied to the parameter registers when the thread + * begins running. + * + * See arch/openrisc/kernel/process.c: + * The args are passed as follows: + * arg1 (r3) : passed in r20 + * arg2 (r4) : passed in r22 + */ + +ENTRY(_kernel_thread_helper) + l.or r3,r20,r0 + l.or r4,r22,r0 + l.movhi r31,hi(kernel_thread_helper) + l.ori r31,r31,lo(kernel_thread_helper) + l.jr r31 + l.nop + + +/* ========================================================[ switch ] === */ + +/* + * This routine switches between two different tasks. The process + * state of one is saved on its kernel stack. Then the state + * of the other is restored from its kernel stack. The memory + * management hardware is updated to the second process's state. + * Finally, we can return to the second process, via the 'return'. + * + * Note: there are two ways to get to the "going out" portion + * of this code; either by coming in via the entry (_switch) + * or via "fork" which must set up an environment equivalent + * to the "_switch" path. If you change this (or in particular, the + * SAVE_REGS macro), you'll have to change the fork code also. + */ + + +/* _switch MUST never lay on page boundry, cause it runs from + * effective addresses and beeing interrupted by iTLB miss would kill it. + * dTLB miss seams to never accour in the bad place since data accesses + * are from task structures which are always page aligned. + * + * The problem happens in RESTORE_ALL_NO_R11 where we first set the EPCR + * register, then load the previous register values and only at the end call + * the l.rfe instruction. If get TLB miss in beetwen the EPCR register gets + * garbled and we end up calling l.rfe with the wrong EPCR. (same probably + * holds for ESR) + * + * To avoid this problems it is sufficient to align _switch to + * some nice round number smaller than it's size... + */ + +/* ABI rules apply here... we either enter _switch via schedule() or via + * an imaginary call to which we shall return at return_from_fork. Either + * way, we are a function call and only need to preserve the callee-saved + * registers when we return. As such, we don't need to save the registers + * on the stack that we won't be returning as they were... + */ + + .align 0x400 +ENTRY(_switch) + /* We don't store SR as _switch only gets called in a context where + * the SR will be the same going in and coming out... */ + + /* Set up new pt_regs struct for saving task state */ + l.addi r1,r1,-(INT_FRAME_SIZE) + + /* No need to store r1/PT_SP as it goes into KSP below */ + l.sw PT_GPR2(r1),r2 + l.sw PT_GPR9(r1),r9 + /* This is wrong, r12 shouldn't be here... but GCC is broken for the time being + * and expects r12 to be callee-saved... */ + l.sw PT_GPR12(r1),r12 + l.sw PT_GPR14(r1),r14 + l.sw PT_GPR16(r1),r16 + l.sw PT_GPR18(r1),r18 + l.sw PT_GPR20(r1),r20 + l.sw PT_GPR22(r1),r22 + l.sw PT_GPR24(r1),r24 + l.sw PT_GPR26(r1),r26 + l.sw PT_GPR28(r1),r28 + l.sw PT_GPR30(r1),r30 + + l.addi r11,r10,0 /* Save old 'current' to 'last' return value*/ + + /* We use thread_info->ksp for storing the address of the above + * structure so that we can get back to it later... we don't want + * to lose the value of thread_info->ksp, though, so store it as + * pt_regs->sp so that we can easily restore it when we are made + * live again... + */ + + /* Save the old value of thread_info->ksp as pt_regs->sp */ + l.lwz r29,TI_KSP(r10) + l.sw PT_SP(r1),r29 + + /* Swap kernel stack pointers */ + l.sw TI_KSP(r10),r1 /* Save old stack pointer */ + l.or r10,r4,r0 /* Set up new current_thread_info */ + l.lwz r1,TI_KSP(r10) /* Load new stack pointer */ + + /* Restore the old value of thread_info->ksp */ + l.lwz r29,PT_SP(r1) + l.sw TI_KSP(r10),r29 + + /* ...and restore the registers, except r11 because the return value + * has already been set above. + */ + l.lwz r2,PT_GPR2(r1) + l.lwz r9,PT_GPR9(r1) + /* No need to restore r10 */ + /* ...and do not restore r11 */ + + /* This is wrong, r12 shouldn't be here... but GCC is broken for the time being + * and expects r12 to be callee-saved... */ + l.lwz r12,PT_GPR12(r1) + l.lwz r14,PT_GPR14(r1) + l.lwz r16,PT_GPR16(r1) + l.lwz r18,PT_GPR18(r1) + l.lwz r20,PT_GPR20(r1) + l.lwz r22,PT_GPR22(r1) + l.lwz r24,PT_GPR24(r1) + l.lwz r26,PT_GPR26(r1) + l.lwz r28,PT_GPR28(r1) + l.lwz r30,PT_GPR30(r1) + + /* Unwind stack to pre-switch state */ + l.addi r1,r1,(INT_FRAME_SIZE) + + /* Return via the link-register back to where we 'came from', where that can be + * either schedule() or return_from_fork()... */ + l.jr r9 + l.nop + +/* ==================================================================== */ + +/* These all use the delay slot for setting the argument register, so the + * jump is always happening after the l.addi instruction. + * + * These are all just wrappers that don't touch the link-register r9, so the + * return from the "real" syscall function will return back to the syscall + * code that did the l.jal that brought us here. + */ + +/* fork requires that we save all the callee-saved registers because they + * are all effectively clobbered by the call to _switch. Here we store + * all the registers that aren't touched by the syscall fast path and thus + * weren't saved there. + */ + +_fork_save_extra_regs_and_call: + l.sw PT_GPR14(r1),r14 + l.sw PT_GPR16(r1),r16 + l.sw PT_GPR18(r1),r18 + l.sw PT_GPR20(r1),r20 + l.sw PT_GPR22(r1),r22 + l.sw PT_GPR24(r1),r24 + l.sw PT_GPR26(r1),r26 + l.jr r29 + l.sw PT_GPR28(r1),r28 + +ENTRY(sys_clone) + l.movhi r29,hi(_sys_clone) + l.ori r29,r29,lo(_sys_clone) + l.j _fork_save_extra_regs_and_call + l.addi r7,r1,0 + +ENTRY(sys_fork) + l.movhi r29,hi(_sys_fork) + l.ori r29,r29,lo(_sys_fork) + l.j _fork_save_extra_regs_and_call + l.addi r3,r1,0 + +ENTRY(sys_execve) + l.j _sys_execve + l.addi r6,r1,0 + +ENTRY(sys_sigaltstack) + l.j _sys_sigaltstack + l.addi r5,r1,0 + +ENTRY(sys_rt_sigreturn) + l.j _sys_rt_sigreturn + l.addi r3,r1,0 + +/* This is a catch-all syscall for atomic instructions for the OpenRISC 1000. + * The functions takes a variable number of parameters depending on which + * particular flavour of atomic you want... parameter 1 is a flag identifying + * the atomic in question. Currently, this function implements the + * following variants: + * + * XCHG: + * @flag: 1 + * @ptr1: + * @ptr2: + * Atomically exchange the values in pointers 1 and 2. + * + */ + +ENTRY(sys_or1k_atomic) + /* FIXME: This ignores r3 and always does an XCHG */ + DISABLE_INTERRUPTS(r17,r19) + l.lwz r30,0(r4) + l.lwz r28,0(r5) + l.sw 0(r4),r28 + l.sw 0(r5),r30 + ENABLE_INTERRUPTS(r17) + l.jr r9 + l.or r11,r0,r0 + +/* ============================================================[ EOF ]=== */ diff --git a/arch/openrisc/kernel/head.S b/arch/openrisc/kernel/head.S new file mode 100644 index 000000000000..c75018d22644 --- /dev/null +++ b/arch/openrisc/kernel/head.S @@ -0,0 +1,1607 @@ +/* + * OpenRISC head.S + * + * Linux architectural port borrowing liberally from similar works of + * others. All original copyrights apply as per the original source + * declaration. + * + * Modifications for the OpenRISC architecture: + * Copyright (C) 2003 Matjaz Breskvar + * Copyright (C) 2010-2011 Jonas Bonn + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#define tophys(rd,rs) \ + l.movhi rd,hi(-KERNELBASE) ;\ + l.add rd,rd,rs + +#define CLEAR_GPR(gpr) \ + l.or gpr,r0,r0 + +#define LOAD_SYMBOL_2_GPR(gpr,symbol) \ + l.movhi gpr,hi(symbol) ;\ + l.ori gpr,gpr,lo(symbol) + + +#define UART_BASE_ADD 0x90000000 + +#define EXCEPTION_SR (SPR_SR_DME | SPR_SR_IME | SPR_SR_DCE | SPR_SR_ICE | SPR_SR_SM) +#define SYSCALL_SR (SPR_SR_DME | SPR_SR_IME | SPR_SR_DCE | SPR_SR_ICE | SPR_SR_IEE | SPR_SR_TEE | SPR_SR_SM) + +/* ============================================[ tmp store locations ]=== */ + +/* + * emergency_print temporary stores + */ +#define EMERGENCY_PRINT_STORE_GPR4 l.sw 0x20(r0),r4 +#define EMERGENCY_PRINT_LOAD_GPR4 l.lwz r4,0x20(r0) + +#define EMERGENCY_PRINT_STORE_GPR5 l.sw 0x24(r0),r5 +#define EMERGENCY_PRINT_LOAD_GPR5 l.lwz r5,0x24(r0) + +#define EMERGENCY_PRINT_STORE_GPR6 l.sw 0x28(r0),r6 +#define EMERGENCY_PRINT_LOAD_GPR6 l.lwz r6,0x28(r0) + +#define EMERGENCY_PRINT_STORE_GPR7 l.sw 0x2c(r0),r7 +#define EMERGENCY_PRINT_LOAD_GPR7 l.lwz r7,0x2c(r0) + +#define EMERGENCY_PRINT_STORE_GPR8 l.sw 0x30(r0),r8 +#define EMERGENCY_PRINT_LOAD_GPR8 l.lwz r8,0x30(r0) + +#define EMERGENCY_PRINT_STORE_GPR9 l.sw 0x34(r0),r9 +#define EMERGENCY_PRINT_LOAD_GPR9 l.lwz r9,0x34(r0) + + +/* + * TLB miss handlers temorary stores + */ +#define EXCEPTION_STORE_GPR9 l.sw 0x10(r0),r9 +#define EXCEPTION_LOAD_GPR9 l.lwz r9,0x10(r0) + +#define EXCEPTION_STORE_GPR2 l.sw 0x64(r0),r2 +#define EXCEPTION_LOAD_GPR2 l.lwz r2,0x64(r0) + +#define EXCEPTION_STORE_GPR3 l.sw 0x68(r0),r3 +#define EXCEPTION_LOAD_GPR3 l.lwz r3,0x68(r0) + +#define EXCEPTION_STORE_GPR4 l.sw 0x6c(r0),r4 +#define EXCEPTION_LOAD_GPR4 l.lwz r4,0x6c(r0) + +#define EXCEPTION_STORE_GPR5 l.sw 0x70(r0),r5 +#define EXCEPTION_LOAD_GPR5 l.lwz r5,0x70(r0) + +#define EXCEPTION_STORE_GPR6 l.sw 0x74(r0),r6 +#define EXCEPTION_LOAD_GPR6 l.lwz r6,0x74(r0) + + +/* + * EXCEPTION_HANDLE temporary stores + */ + +#define EXCEPTION_T_STORE_GPR30 l.sw 0x78(r0),r30 +#define EXCEPTION_T_LOAD_GPR30(reg) l.lwz reg,0x78(r0) + +#define EXCEPTION_T_STORE_GPR10 l.sw 0x7c(r0),r10 +#define EXCEPTION_T_LOAD_GPR10(reg) l.lwz reg,0x7c(r0) + +#define EXCEPTION_T_STORE_SP l.sw 0x80(r0),r1 +#define EXCEPTION_T_LOAD_SP(reg) l.lwz reg,0x80(r0) + +/* + * For UNHANLDED_EXCEPTION + */ + +#define EXCEPTION_T_STORE_GPR31 l.sw 0x84(r0),r31 +#define EXCEPTION_T_LOAD_GPR31(reg) l.lwz reg,0x84(r0) + +/* =========================================================[ macros ]=== */ + + +#define GET_CURRENT_PGD(reg,t1) \ + LOAD_SYMBOL_2_GPR(reg,current_pgd) ;\ + tophys (t1,reg) ;\ + l.lwz reg,0(t1) + + +/* + * DSCR: this is a common hook for handling exceptions. it will save + * the needed registers, set up stack and pointer to current + * then jump to the handler while enabling MMU + * + * PRMS: handler - a function to jump to. it has to save the + * remaining registers to kernel stack, call + * appropriate arch-independant exception handler + * and finaly jump to ret_from_except + * + * PREQ: unchanged state from the time exception happened + * + * POST: SAVED the following registers original value + * to the new created exception frame pointed to by r1 + * + * r1 - ksp pointing to the new (exception) frame + * r4 - EEAR exception EA + * r10 - current pointing to current_thread_info struct + * r12 - syscall 0, since we didn't come from syscall + * r13 - temp it actually contains new SR, not needed anymore + * r31 - handler address of the handler we'll jump to + * + * handler has to save remaining registers to the exception + * ksp frame *before* tainting them! + * + * NOTE: this function is not reentrant per se. reentrancy is guaranteed + * by processor disabling all exceptions/interrupts when exception + * accours. + * + * OPTM: no need to make it so wasteful to extract ksp when in user mode + */ + +#define EXCEPTION_HANDLE(handler) \ + EXCEPTION_T_STORE_GPR30 ;\ + l.mfspr r30,r0,SPR_ESR_BASE ;\ + l.andi r30,r30,SPR_SR_SM ;\ + l.sfeqi r30,0 ;\ + EXCEPTION_T_STORE_GPR10 ;\ + l.bnf 2f /* kernel_mode */ ;\ + EXCEPTION_T_STORE_SP /* delay slot */ ;\ +1: /* user_mode: */ ;\ + LOAD_SYMBOL_2_GPR(r1,current_thread_info_set) ;\ + tophys (r30,r1) ;\ + /* r10: current_thread_info */ ;\ + l.lwz r10,0(r30) ;\ + tophys (r30,r10) ;\ + l.lwz r1,(TI_KSP)(r30) ;\ + /* fall through */ ;\ +2: /* kernel_mode: */ ;\ + /* create new stack frame, save only needed gprs */ ;\ + /* r1: KSP, r10: current, r4: EEAR, r31: __pa(KSP) */ ;\ + /* r12: temp, syscall indicator */ ;\ + l.addi r1,r1,-(INT_FRAME_SIZE) ;\ + /* r1 is KSP, r30 is __pa(KSP) */ ;\ + tophys (r30,r1) ;\ + l.sw PT_GPR12(r30),r12 ;\ + l.mfspr r12,r0,SPR_EPCR_BASE ;\ + l.sw PT_PC(r30),r12 ;\ + l.mfspr r12,r0,SPR_ESR_BASE ;\ + l.sw PT_SR(r30),r12 ;\ + /* save r30 */ ;\ + EXCEPTION_T_LOAD_GPR30(r12) ;\ + l.sw PT_GPR30(r30),r12 ;\ + /* save r10 as was prior to exception */ ;\ + EXCEPTION_T_LOAD_GPR10(r12) ;\ + l.sw PT_GPR10(r30),r12 ;\ + /* save PT_SP as was prior to exception */ ;\ + EXCEPTION_T_LOAD_SP(r12) ;\ + l.sw PT_SP(r30),r12 ;\ + /* save exception r4, set r4 = EA */ ;\ + l.sw PT_GPR4(r30),r4 ;\ + l.mfspr r4,r0,SPR_EEAR_BASE ;\ + /* r12 == 1 if we come from syscall */ ;\ + CLEAR_GPR(r12) ;\ + /* ----- turn on MMU ----- */ ;\ + l.ori r30,r0,(EXCEPTION_SR) ;\ + l.mtspr r0,r30,SPR_ESR_BASE ;\ + /* r30: EA address of handler */ ;\ + LOAD_SYMBOL_2_GPR(r30,handler) ;\ + l.mtspr r0,r30,SPR_EPCR_BASE ;\ + l.rfe + +/* + * this doesn't work + * + * + * #ifdef CONFIG_JUMP_UPON_UNHANDLED_EXCEPTION + * #define UNHANDLED_EXCEPTION(handler) \ + * l.ori r3,r0,0x1 ;\ + * l.mtspr r0,r3,SPR_SR ;\ + * l.movhi r3,hi(0xf0000100) ;\ + * l.ori r3,r3,lo(0xf0000100) ;\ + * l.jr r3 ;\ + * l.nop 1 + * + * #endif + */ + +/* DSCR: this is the same as EXCEPTION_HANDLE(), we are just + * a bit more carefull (if we have a PT_SP or current pointer + * corruption) and set them up from 'current_set' + * + */ +#define UNHANDLED_EXCEPTION(handler) \ + EXCEPTION_T_STORE_GPR31 ;\ + EXCEPTION_T_STORE_GPR10 ;\ + EXCEPTION_T_STORE_SP ;\ + /* temporary store r3, r9 into r1, r10 */ ;\ + l.addi r1,r3,0x0 ;\ + l.addi r10,r9,0x0 ;\ + /* the string referenced by r3 must be low enough */ ;\ + l.jal _emergency_print ;\ + l.ori r3,r0,lo(_string_unhandled_exception) ;\ + l.mfspr r3,r0,SPR_NPC ;\ + l.jal _emergency_print_nr ;\ + l.andi r3,r3,0x1f00 ;\ + /* the string referenced by r3 must be low enough */ ;\ + l.jal _emergency_print ;\ + l.ori r3,r0,lo(_string_epc_prefix) ;\ + l.jal _emergency_print_nr ;\ + l.mfspr r3,r0,SPR_EPCR_BASE ;\ + l.jal _emergency_print ;\ + l.ori r3,r0,lo(_string_nl) ;\ + /* end of printing */ ;\ + l.addi r3,r1,0x0 ;\ + l.addi r9,r10,0x0 ;\ + /* extract current, ksp from current_set */ ;\ + LOAD_SYMBOL_2_GPR(r1,_unhandled_stack_top) ;\ + LOAD_SYMBOL_2_GPR(r10,init_thread_union) ;\ + /* create new stack frame, save only needed gprs */ ;\ + /* r1: KSP, r10: current, r31: __pa(KSP) */ ;\ + /* r12: temp, syscall indicator, r13 temp */ ;\ + l.addi r1,r1,-(INT_FRAME_SIZE) ;\ + /* r1 is KSP, r31 is __pa(KSP) */ ;\ + tophys (r31,r1) ;\ + l.sw PT_GPR12(r31),r12 ;\ + l.mfspr r12,r0,SPR_EPCR_BASE ;\ + l.sw PT_PC(r31),r12 ;\ + l.mfspr r12,r0,SPR_ESR_BASE ;\ + l.sw PT_SR(r31),r12 ;\ + /* save r31 */ ;\ + EXCEPTION_T_LOAD_GPR31(r12) ;\ + l.sw PT_GPR31(r31),r12 ;\ + /* save r10 as was prior to exception */ ;\ + EXCEPTION_T_LOAD_GPR10(r12) ;\ + l.sw PT_GPR10(r31),r12 ;\ + /* save PT_SP as was prior to exception */ ;\ + EXCEPTION_T_LOAD_SP(r12) ;\ + l.sw PT_SP(r31),r12 ;\ + l.sw PT_GPR13(r31),r13 ;\ + /* --> */ ;\ + /* save exception r4, set r4 = EA */ ;\ + l.sw PT_GPR4(r31),r4 ;\ + l.mfspr r4,r0,SPR_EEAR_BASE ;\ + /* r12 == 1 if we come from syscall */ ;\ + CLEAR_GPR(r12) ;\ + /* ----- play a MMU trick ----- */ ;\ + l.ori r31,r0,(EXCEPTION_SR) ;\ + l.mtspr r0,r31,SPR_ESR_BASE ;\ + /* r31: EA address of handler */ ;\ + LOAD_SYMBOL_2_GPR(r31,handler) ;\ + l.mtspr r0,r31,SPR_EPCR_BASE ;\ + l.rfe + +/* =====================================================[ exceptions] === */ + +/* ---[ 0x100: RESET exception ]----------------------------------------- */ + .org 0x100 + /* Jump to .init code at _start which lives in the .head section + * and will be discarded after boot. + */ + LOAD_SYMBOL_2_GPR(r4, _start) + tophys (r3,r4) /* MMU disabled */ + l.jr r3 + l.nop + +/* ---[ 0x200: BUS exception ]------------------------------------------- */ + .org 0x200 +_dispatch_bus_fault: + EXCEPTION_HANDLE(_bus_fault_handler) + +/* ---[ 0x300: Data Page Fault exception ]------------------------------- */ + .org 0x300 +_dispatch_do_dpage_fault: +// totaly disable timer interrupt +// l.mtspr r0,r0,SPR_TTMR +// DEBUG_TLB_PROBE(0x300) +// EXCEPTION_DEBUG_VALUE_ER_ENABLED(0x300) + EXCEPTION_HANDLE(_data_page_fault_handler) + +/* ---[ 0x400: Insn Page Fault exception ]------------------------------- */ + .org 0x400 +_dispatch_do_ipage_fault: +// totaly disable timer interrupt +// l.mtspr r0,r0,SPR_TTMR +// DEBUG_TLB_PROBE(0x400) +// EXCEPTION_DEBUG_VALUE_ER_ENABLED(0x400) + EXCEPTION_HANDLE(_insn_page_fault_handler) + +/* ---[ 0x500: Timer exception ]----------------------------------------- */ + .org 0x500 + EXCEPTION_HANDLE(_timer_handler) + +/* ---[ 0x600: Aligment exception ]-------------------------------------- */ + .org 0x600 + EXCEPTION_HANDLE(_alignment_handler) + +/* ---[ 0x700: Illegal insn exception ]---------------------------------- */ + .org 0x700 + EXCEPTION_HANDLE(_illegal_instruction_handler) + +/* ---[ 0x800: External interrupt exception ]---------------------------- */ + .org 0x800 + EXCEPTION_HANDLE(_external_irq_handler) + +/* ---[ 0x900: DTLB miss exception ]------------------------------------- */ + .org 0x900 + l.j boot_dtlb_miss_handler + l.nop + +/* ---[ 0xa00: ITLB miss exception ]------------------------------------- */ + .org 0xa00 + l.j boot_itlb_miss_handler + l.nop + +/* ---[ 0xb00: Range exception ]----------------------------------------- */ + .org 0xb00 + UNHANDLED_EXCEPTION(_vector_0xb00) + +/* ---[ 0xc00: Syscall exception ]--------------------------------------- */ + .org 0xc00 + EXCEPTION_HANDLE(_sys_call_handler) + +/* ---[ 0xd00: Trap exception ]------------------------------------------ */ + .org 0xd00 + UNHANDLED_EXCEPTION(_vector_0xd00) + +/* ---[ 0xe00: Trap exception ]------------------------------------------ */ + .org 0xe00 +// UNHANDLED_EXCEPTION(_vector_0xe00) + EXCEPTION_HANDLE(_trap_handler) + +/* ---[ 0xf00: Reserved exception ]-------------------------------------- */ + .org 0xf00 + UNHANDLED_EXCEPTION(_vector_0xf00) + +/* ---[ 0x1000: Reserved exception ]------------------------------------- */ + .org 0x1000 + UNHANDLED_EXCEPTION(_vector_0x1000) + +/* ---[ 0x1100: Reserved exception ]------------------------------------- */ + .org 0x1100 + UNHANDLED_EXCEPTION(_vector_0x1100) + +/* ---[ 0x1200: Reserved exception ]------------------------------------- */ + .org 0x1200 + UNHANDLED_EXCEPTION(_vector_0x1200) + +/* ---[ 0x1300: Reserved exception ]------------------------------------- */ + .org 0x1300 + UNHANDLED_EXCEPTION(_vector_0x1300) + +/* ---[ 0x1400: Reserved exception ]------------------------------------- */ + .org 0x1400 + UNHANDLED_EXCEPTION(_vector_0x1400) + +/* ---[ 0x1500: Reserved exception ]------------------------------------- */ + .org 0x1500 + UNHANDLED_EXCEPTION(_vector_0x1500) + +/* ---[ 0x1600: Reserved exception ]------------------------------------- */ + .org 0x1600 + UNHANDLED_EXCEPTION(_vector_0x1600) + +/* ---[ 0x1700: Reserved exception ]------------------------------------- */ + .org 0x1700 + UNHANDLED_EXCEPTION(_vector_0x1700) + +/* ---[ 0x1800: Reserved exception ]------------------------------------- */ + .org 0x1800 + UNHANDLED_EXCEPTION(_vector_0x1800) + +/* ---[ 0x1900: Reserved exception ]------------------------------------- */ + .org 0x1900 + UNHANDLED_EXCEPTION(_vector_0x1900) + +/* ---[ 0x1a00: Reserved exception ]------------------------------------- */ + .org 0x1a00 + UNHANDLED_EXCEPTION(_vector_0x1a00) + +/* ---[ 0x1b00: Reserved exception ]------------------------------------- */ + .org 0x1b00 + UNHANDLED_EXCEPTION(_vector_0x1b00) + +/* ---[ 0x1c00: Reserved exception ]------------------------------------- */ + .org 0x1c00 + UNHANDLED_EXCEPTION(_vector_0x1c00) + +/* ---[ 0x1d00: Reserved exception ]------------------------------------- */ + .org 0x1d00 + UNHANDLED_EXCEPTION(_vector_0x1d00) + +/* ---[ 0x1e00: Reserved exception ]------------------------------------- */ + .org 0x1e00 + UNHANDLED_EXCEPTION(_vector_0x1e00) + +/* ---[ 0x1f00: Reserved exception ]------------------------------------- */ + .org 0x1f00 + UNHANDLED_EXCEPTION(_vector_0x1f00) + + .org 0x2000 +/* ===================================================[ kernel start ]=== */ + +/* .text*/ + +/* This early stuff belongs in HEAD, but some of the functions below definitely + * don't... */ + + __HEAD + .global _start +_start: + /* + * ensure a deterministic start + */ + + l.ori r3,r0,0x1 + l.mtspr r0,r3,SPR_SR + + CLEAR_GPR(r1) + CLEAR_GPR(r2) + CLEAR_GPR(r3) + CLEAR_GPR(r4) + CLEAR_GPR(r5) + CLEAR_GPR(r6) + CLEAR_GPR(r7) + CLEAR_GPR(r8) + CLEAR_GPR(r9) + CLEAR_GPR(r10) + CLEAR_GPR(r11) + CLEAR_GPR(r12) + CLEAR_GPR(r13) + CLEAR_GPR(r14) + CLEAR_GPR(r15) + CLEAR_GPR(r16) + CLEAR_GPR(r17) + CLEAR_GPR(r18) + CLEAR_GPR(r19) + CLEAR_GPR(r20) + CLEAR_GPR(r21) + CLEAR_GPR(r22) + CLEAR_GPR(r23) + CLEAR_GPR(r24) + CLEAR_GPR(r25) + CLEAR_GPR(r26) + CLEAR_GPR(r27) + CLEAR_GPR(r28) + CLEAR_GPR(r29) + CLEAR_GPR(r30) + CLEAR_GPR(r31) + + /* + * set up initial ksp and current + */ + LOAD_SYMBOL_2_GPR(r1,init_thread_union+0x2000) // setup kernel stack + LOAD_SYMBOL_2_GPR(r10,init_thread_union) // setup current + tophys (r31,r10) + l.sw TI_KSP(r31), r1 + + l.ori r4,r0,0x0 + + + /* + * .data contains initialized data, + * .bss contains uninitialized data - clear it up + */ +clear_bss: + LOAD_SYMBOL_2_GPR(r24, __bss_start) + LOAD_SYMBOL_2_GPR(r26, _end) + tophys(r28,r24) + tophys(r30,r26) + CLEAR_GPR(r24) + CLEAR_GPR(r26) +1: + l.sw (0)(r28),r0 + l.sfltu r28,r30 + l.bf 1b + l.addi r28,r28,4 + +enable_ic: + l.jal _ic_enable + l.nop + +enable_dc: + l.jal _dc_enable + l.nop + +flush_tlb: + /* + * I N V A L I D A T E T L B e n t r i e s + */ + LOAD_SYMBOL_2_GPR(r5,SPR_DTLBMR_BASE(0)) + LOAD_SYMBOL_2_GPR(r6,SPR_ITLBMR_BASE(0)) + l.addi r7,r0,128 /* Maximum number of sets */ +1: + l.mtspr r5,r0,0x0 + l.mtspr r6,r0,0x0 + + l.addi r5,r5,1 + l.addi r6,r6,1 + l.sfeq r7,r0 + l.bnf 1b + l.addi r7,r7,-1 + + +/* The MMU needs to be enabled before or32_early_setup is called */ + +enable_mmu: + /* + * enable dmmu & immu + * SR[5] = 0, SR[6] = 0, 6th and 7th bit of SR set to 0 + */ + l.mfspr r30,r0,SPR_SR + l.movhi r28,hi(SPR_SR_DME | SPR_SR_IME) + l.ori r28,r28,lo(SPR_SR_DME | SPR_SR_IME) + l.or r30,r30,r28 + l.mtspr r0,r30,SPR_SR + l.nop + l.nop + l.nop + l.nop + l.nop + l.nop + l.nop + l.nop + l.nop + l.nop + l.nop + l.nop + l.nop + l.nop + l.nop + l.nop + + // reset the simulation counters + l.nop 5 + + LOAD_SYMBOL_2_GPR(r24, or32_early_setup) + l.jalr r24 + l.nop + +clear_regs: + /* + * clear all GPRS to increase determinism + */ + CLEAR_GPR(r2) + CLEAR_GPR(r3) + CLEAR_GPR(r4) + CLEAR_GPR(r5) + CLEAR_GPR(r6) + CLEAR_GPR(r7) + CLEAR_GPR(r8) + CLEAR_GPR(r9) + CLEAR_GPR(r11) + CLEAR_GPR(r12) + CLEAR_GPR(r13) + CLEAR_GPR(r14) + CLEAR_GPR(r15) + CLEAR_GPR(r16) + CLEAR_GPR(r17) + CLEAR_GPR(r18) + CLEAR_GPR(r19) + CLEAR_GPR(r20) + CLEAR_GPR(r21) + CLEAR_GPR(r22) + CLEAR_GPR(r23) + CLEAR_GPR(r24) + CLEAR_GPR(r25) + CLEAR_GPR(r26) + CLEAR_GPR(r27) + CLEAR_GPR(r28) + CLEAR_GPR(r29) + CLEAR_GPR(r30) + CLEAR_GPR(r31) + +jump_start_kernel: + /* + * jump to kernel entry (start_kernel) + */ + LOAD_SYMBOL_2_GPR(r30, start_kernel) + l.jr r30 + l.nop + +/* ========================================[ cache ]=== */ + + /* aligment here so we don't change memory offsets with + * memory controler defined + */ + .align 0x2000 + +_ic_enable: + /* Check if IC present and skip enabling otherwise */ + l.mfspr r24,r0,SPR_UPR + l.andi r26,r24,SPR_UPR_ICP + l.sfeq r26,r0 + l.bf 9f + l.nop + + /* Disable IC */ + l.mfspr r6,r0,SPR_SR + l.addi r5,r0,-1 + l.xori r5,r5,SPR_SR_ICE + l.and r5,r6,r5 + l.mtspr r0,r5,SPR_SR + + /* Establish cache block size + If BS=0, 16; + If BS=1, 32; + r14 contain block size + */ + l.mfspr r24,r0,SPR_ICCFGR + l.andi r26,r24,SPR_ICCFGR_CBS + l.srli r28,r26,7 + l.ori r30,r0,16 + l.sll r14,r30,r28 + + /* Establish number of cache sets + r16 contains number of cache sets + r28 contains log(# of cache sets) + */ + l.andi r26,r24,SPR_ICCFGR_NCS + l.srli r28,r26,3 + l.ori r30,r0,1 + l.sll r16,r30,r28 + + /* Invalidate IC */ + l.addi r6,r0,0 + l.sll r5,r14,r28 +// l.mul r5,r14,r16 +// l.trap 1 +// l.addi r5,r0,IC_SIZE +1: + l.mtspr r0,r6,SPR_ICBIR + l.sfne r6,r5 + l.bf 1b + l.add r6,r6,r14 + // l.addi r6,r6,IC_LINE + + /* Enable IC */ + l.mfspr r6,r0,SPR_SR + l.ori r6,r6,SPR_SR_ICE + l.mtspr r0,r6,SPR_SR + l.nop + l.nop + l.nop + l.nop + l.nop + l.nop + l.nop + l.nop + l.nop + l.nop +9: + l.jr r9 + l.nop + +_dc_enable: + /* Check if DC present and skip enabling otherwise */ + l.mfspr r24,r0,SPR_UPR + l.andi r26,r24,SPR_UPR_DCP + l.sfeq r26,r0 + l.bf 9f + l.nop + + /* Disable DC */ + l.mfspr r6,r0,SPR_SR + l.addi r5,r0,-1 + l.xori r5,r5,SPR_SR_DCE + l.and r5,r6,r5 + l.mtspr r0,r5,SPR_SR + + /* Establish cache block size + If BS=0, 16; + If BS=1, 32; + r14 contain block size + */ + l.mfspr r24,r0,SPR_DCCFGR + l.andi r26,r24,SPR_DCCFGR_CBS + l.srli r28,r26,7 + l.ori r30,r0,16 + l.sll r14,r30,r28 + + /* Establish number of cache sets + r16 contains number of cache sets + r28 contains log(# of cache sets) + */ + l.andi r26,r24,SPR_DCCFGR_NCS + l.srli r28,r26,3 + l.ori r30,r0,1 + l.sll r16,r30,r28 + + /* Invalidate DC */ + l.addi r6,r0,0 + l.sll r5,r14,r28 +1: + l.mtspr r0,r6,SPR_DCBIR + l.sfne r6,r5 + l.bf 1b + l.add r6,r6,r14 + + /* Enable DC */ + l.mfspr r6,r0,SPR_SR + l.ori r6,r6,SPR_SR_DCE + l.mtspr r0,r6,SPR_SR +9: + l.jr r9 + l.nop + +/* ===============================================[ page table masks ]=== */ + +/* bit 4 is used in hardware as write back cache bit. we never use this bit + * explicitly, so we can reuse it as _PAGE_FILE bit and mask it out when + * writing into hardware pte's + */ + +#define DTLB_UP_CONVERT_MASK 0x3fa +#define ITLB_UP_CONVERT_MASK 0x3a + +/* for SMP we'd have (this is a bit subtle, CC must be always set + * for SMP, but since we have _PAGE_PRESENT bit always defined + * we can just modify the mask) + */ +#define DTLB_SMP_CONVERT_MASK 0x3fb +#define ITLB_SMP_CONVERT_MASK 0x3b + +/* ---[ boot dtlb miss handler ]----------------------------------------- */ + +boot_dtlb_miss_handler: + +/* mask for DTLB_MR register: - (0) sets V (valid) bit, + * - (31-12) sets bits belonging to VPN (31-12) + */ +#define DTLB_MR_MASK 0xfffff001 + +/* mask for DTLB_TR register: - (2) sets CI (cache inhibit) bit, + * - (4) sets A (access) bit, + * - (5) sets D (dirty) bit, + * - (8) sets SRE (superuser read) bit + * - (9) sets SWE (superuser write) bit + * - (31-12) sets bits belonging to VPN (31-12) + */ +#define DTLB_TR_MASK 0xfffff332 + +/* These are for masking out the VPN/PPN value from the MR/TR registers... + * it's not the same as the PFN */ +#define VPN_MASK 0xfffff000 +#define PPN_MASK 0xfffff000 + + + EXCEPTION_STORE_GPR6 + +#if 0 + l.mfspr r6,r0,SPR_ESR_BASE // + l.andi r6,r6,SPR_SR_SM // are we in kernel mode ? + l.sfeqi r6,0 // r6 == 0x1 --> SM + l.bf exit_with_no_dtranslation // + l.nop +#endif + + /* this could be optimized by moving storing of + * non r6 registers here, and jumping r6 restore + * if not in supervisor mode + */ + + EXCEPTION_STORE_GPR2 + EXCEPTION_STORE_GPR3 + EXCEPTION_STORE_GPR4 + EXCEPTION_STORE_GPR5 + + l.mfspr r4,r0,SPR_EEAR_BASE // get the offending EA + +immediate_translation: + CLEAR_GPR(r6) + + l.srli r3,r4,0xd // r3 <- r4 / 8192 (sets are relative to page size (8Kb) NOT VPN size (4Kb) + + l.mfspr r6, r0, SPR_DMMUCFGR + l.andi r6, r6, SPR_DMMUCFGR_NTS + l.srli r6, r6, SPR_DMMUCFGR_NTS_OFF + l.ori r5, r0, 0x1 + l.sll r5, r5, r6 // r5 = number DMMU sets + l.addi r6, r5, -1 // r6 = nsets mask + l.and r2, r3, r6 // r2 <- r3 % NSETS_MASK + + l.or r6,r6,r4 // r6 <- r4 + l.ori r6,r6,~(VPN_MASK) // r6 <- VPN :VPN .xfff - clear up lo(r6) to 0x**** *fff + l.movhi r5,hi(DTLB_MR_MASK) // r5 <- ffff:0000.x000 + l.ori r5,r5,lo(DTLB_MR_MASK) // r5 <- ffff:1111.x001 - apply DTLB_MR_MASK + l.and r5,r5,r6 // r5 <- VPN :VPN .x001 - we have DTLBMR entry + l.mtspr r2,r5,SPR_DTLBMR_BASE(0) // set DTLBMR + + /* set up DTLB with no translation for EA <= 0xbfffffff */ + LOAD_SYMBOL_2_GPR(r6,0xbfffffff) + l.sfgeu r6,r4 // flag if r6 >= r4 (if 0xbfffffff >= EA) + l.bf 1f // goto out + l.and r3,r4,r4 // delay slot :: 24 <- r4 (if flag==1) + + tophys(r3,r4) // r3 <- PA +1: + l.ori r3,r3,~(PPN_MASK) // r3 <- PPN :PPN .xfff - clear up lo(r6) to 0x**** *fff + l.movhi r5,hi(DTLB_TR_MASK) // r5 <- ffff:0000.x000 + l.ori r5,r5,lo(DTLB_TR_MASK) // r5 <- ffff:1111.x330 - apply DTLB_MR_MASK + l.and r5,r5,r3 // r5 <- PPN :PPN .x330 - we have DTLBTR entry + l.mtspr r2,r5,SPR_DTLBTR_BASE(0) // set DTLBTR + + EXCEPTION_LOAD_GPR6 + EXCEPTION_LOAD_GPR5 + EXCEPTION_LOAD_GPR4 + EXCEPTION_LOAD_GPR3 + EXCEPTION_LOAD_GPR2 + + l.rfe // SR <- ESR, PC <- EPC + +exit_with_no_dtranslation: + /* EA out of memory or not in supervisor mode */ + EXCEPTION_LOAD_GPR6 + EXCEPTION_LOAD_GPR4 + l.j _dispatch_bus_fault + +/* ---[ boot itlb miss handler ]----------------------------------------- */ + +boot_itlb_miss_handler: + +/* mask for ITLB_MR register: - sets V (valid) bit, + * - sets bits belonging to VPN (15-12) + */ +#define ITLB_MR_MASK 0xfffff001 + +/* mask for ITLB_TR register: - sets A (access) bit, + * - sets SXE (superuser execute) bit + * - sets bits belonging to VPN (15-12) + */ +#define ITLB_TR_MASK 0xfffff050 + +/* +#define VPN_MASK 0xffffe000 +#define PPN_MASK 0xffffe000 +*/ + + + + EXCEPTION_STORE_GPR2 + EXCEPTION_STORE_GPR3 + EXCEPTION_STORE_GPR4 + EXCEPTION_STORE_GPR5 + EXCEPTION_STORE_GPR6 + +#if 0 + l.mfspr r6,r0,SPR_ESR_BASE // + l.andi r6,r6,SPR_SR_SM // are we in kernel mode ? + l.sfeqi r6,0 // r6 == 0x1 --> SM + l.bf exit_with_no_itranslation + l.nop +#endif + + + l.mfspr r4,r0,SPR_EEAR_BASE // get the offending EA + +earlyearly: + CLEAR_GPR(r6) + + l.srli r3,r4,0xd // r3 <- r4 / 8192 (sets are relative to page size (8Kb) NOT VPN size (4Kb) + + l.mfspr r6, r0, SPR_IMMUCFGR + l.andi r6, r6, SPR_IMMUCFGR_NTS + l.srli r6, r6, SPR_IMMUCFGR_NTS_OFF + l.ori r5, r0, 0x1 + l.sll r5, r5, r6 // r5 = number IMMU sets from IMMUCFGR + l.addi r6, r5, -1 // r6 = nsets mask + l.and r2, r3, r6 // r2 <- r3 % NSETS_MASK + + l.or r6,r6,r4 // r6 <- r4 + l.ori r6,r6,~(VPN_MASK) // r6 <- VPN :VPN .xfff - clear up lo(r6) to 0x**** *fff + l.movhi r5,hi(ITLB_MR_MASK) // r5 <- ffff:0000.x000 + l.ori r5,r5,lo(ITLB_MR_MASK) // r5 <- ffff:1111.x001 - apply ITLB_MR_MASK + l.and r5,r5,r6 // r5 <- VPN :VPN .x001 - we have ITLBMR entry + l.mtspr r2,r5,SPR_ITLBMR_BASE(0) // set ITLBMR + + /* + * set up ITLB with no translation for EA <= 0x0fffffff + * + * we need this for head.S mapping (EA = PA). if we move all functions + * which run with mmu enabled into entry.S, we might be able to eliminate this. + * + */ + LOAD_SYMBOL_2_GPR(r6,0x0fffffff) + l.sfgeu r6,r4 // flag if r6 >= r4 (if 0xb0ffffff >= EA) + l.bf 1f // goto out + l.and r3,r4,r4 // delay slot :: 24 <- r4 (if flag==1) + + tophys(r3,r4) // r3 <- PA +1: + l.ori r3,r3,~(PPN_MASK) // r3 <- PPN :PPN .xfff - clear up lo(r6) to 0x**** *fff + l.movhi r5,hi(ITLB_TR_MASK) // r5 <- ffff:0000.x000 + l.ori r5,r5,lo(ITLB_TR_MASK) // r5 <- ffff:1111.x050 - apply ITLB_MR_MASK + l.and r5,r5,r3 // r5 <- PPN :PPN .x050 - we have ITLBTR entry + l.mtspr r2,r5,SPR_ITLBTR_BASE(0) // set ITLBTR + + EXCEPTION_LOAD_GPR6 + EXCEPTION_LOAD_GPR5 + EXCEPTION_LOAD_GPR4 + EXCEPTION_LOAD_GPR3 + EXCEPTION_LOAD_GPR2 + + l.rfe // SR <- ESR, PC <- EPC + +exit_with_no_itranslation: + EXCEPTION_LOAD_GPR4 + EXCEPTION_LOAD_GPR6 + l.j _dispatch_bus_fault + l.nop + +/* ====================================================================== */ +/* + * Stuff below here shouldn't go into .head section... maybe this stuff + * can be moved to entry.S ??? + */ + +/* ==============================================[ DTLB miss handler ]=== */ + +/* + * Comments: + * Exception handlers are entered with MMU off so the following handler + * needs to use physical addressing + * + */ + + .text +ENTRY(dtlb_miss_handler) + EXCEPTION_STORE_GPR2 + EXCEPTION_STORE_GPR3 + EXCEPTION_STORE_GPR4 + EXCEPTION_STORE_GPR5 + EXCEPTION_STORE_GPR6 + /* + * get EA of the miss + */ + l.mfspr r2,r0,SPR_EEAR_BASE + /* + * pmd = (pmd_t *)(current_pgd + pgd_index(daddr)); + */ + GET_CURRENT_PGD(r3,r5) // r3 is current_pgd, r5 is temp + l.srli r4,r2,0x18 // >> PAGE_SHIFT + (PAGE_SHIFT - 2) + l.slli r4,r4,0x2 // to get address << 2 + l.add r5,r4,r3 // r4 is pgd_index(daddr) + /* + * if (pmd_none(*pmd)) + * goto pmd_none: + */ + tophys (r4,r5) + l.lwz r3,0x0(r4) // get *pmd value + l.sfne r3,r0 + l.bnf d_pmd_none + l.andi r3,r3,~PAGE_MASK //0x1fff // ~PAGE_MASK + /* + * if (pmd_bad(*pmd)) + * pmd_clear(pmd) + * goto pmd_bad: + */ +// l.sfeq r3,r0 // check *pmd value +// l.bf d_pmd_good + l.addi r3,r0,0xffffe000 // PAGE_MASK +// l.j d_pmd_bad +// l.sw 0x0(r4),r0 // clear pmd +d_pmd_good: + /* + * pte = *pte_offset(pmd, daddr); + */ + l.lwz r4,0x0(r4) // get **pmd value + l.and r4,r4,r3 // & PAGE_MASK + l.srli r5,r2,0xd // >> PAGE_SHIFT, r2 == EEAR + l.andi r3,r5,0x7ff // (1UL << PAGE_SHIFT - 2) - 1 + l.slli r3,r3,0x2 // to get address << 2 + l.add r3,r3,r4 + l.lwz r2,0x0(r3) // this is pte at last + /* + * if (!pte_present(pte)) + */ + l.andi r4,r2,0x1 + l.sfne r4,r0 // is pte present + l.bnf d_pte_not_present + l.addi r3,r0,0xffffe3fa // PAGE_MASK | DTLB_UP_CONVERT_MASK + /* + * fill DTLB TR register + */ + l.and r4,r2,r3 // apply the mask + // Determine number of DMMU sets + l.mfspr r6, r0, SPR_DMMUCFGR + l.andi r6, r6, SPR_DMMUCFGR_NTS + l.srli r6, r6, SPR_DMMUCFGR_NTS_OFF + l.ori r3, r0, 0x1 + l.sll r3, r3, r6 // r3 = number DMMU sets DMMUCFGR + l.addi r6, r3, -1 // r6 = nsets mask + l.and r5, r5, r6 // calc offset: & (NUM_TLB_ENTRIES-1) + //NUM_TLB_ENTRIES + l.mtspr r5,r4,SPR_DTLBTR_BASE(0) + /* + * fill DTLB MR register + */ + l.mfspr r2,r0,SPR_EEAR_BASE + l.addi r3,r0,0xffffe000 // PAGE_MASK + l.and r4,r2,r3 // apply PAGE_MASK to EA (__PHX__ do we really need this?) + l.ori r4,r4,0x1 // set hardware valid bit: DTBL_MR entry + l.mtspr r5,r4,SPR_DTLBMR_BASE(0) + + EXCEPTION_LOAD_GPR2 + EXCEPTION_LOAD_GPR3 + EXCEPTION_LOAD_GPR4 + EXCEPTION_LOAD_GPR5 + EXCEPTION_LOAD_GPR6 + l.rfe +d_pmd_bad: + l.nop 1 + EXCEPTION_LOAD_GPR2 + EXCEPTION_LOAD_GPR3 + EXCEPTION_LOAD_GPR4 + EXCEPTION_LOAD_GPR5 + EXCEPTION_LOAD_GPR6 + l.rfe +d_pmd_none: +d_pte_not_present: + EXCEPTION_LOAD_GPR2 + EXCEPTION_LOAD_GPR3 + EXCEPTION_LOAD_GPR4 + EXCEPTION_LOAD_GPR5 + EXCEPTION_LOAD_GPR6 + l.j _dispatch_do_dpage_fault + l.nop + +/* ==============================================[ ITLB miss handler ]=== */ +ENTRY(itlb_miss_handler) + EXCEPTION_STORE_GPR2 + EXCEPTION_STORE_GPR3 + EXCEPTION_STORE_GPR4 + EXCEPTION_STORE_GPR5 + EXCEPTION_STORE_GPR6 + /* + * get EA of the miss + */ + l.mfspr r2,r0,SPR_EEAR_BASE + + /* + * pmd = (pmd_t *)(current_pgd + pgd_index(daddr)); + * + */ + GET_CURRENT_PGD(r3,r5) // r3 is current_pgd, r5 is temp + l.srli r4,r2,0x18 // >> PAGE_SHIFT + (PAGE_SHIFT - 2) + l.slli r4,r4,0x2 // to get address << 2 + l.add r5,r4,r3 // r4 is pgd_index(daddr) + /* + * if (pmd_none(*pmd)) + * goto pmd_none: + */ + tophys (r4,r5) + l.lwz r3,0x0(r4) // get *pmd value + l.sfne r3,r0 + l.bnf i_pmd_none + l.andi r3,r3,0x1fff // ~PAGE_MASK + /* + * if (pmd_bad(*pmd)) + * pmd_clear(pmd) + * goto pmd_bad: + */ + +// l.sfeq r3,r0 // check *pmd value +// l.bf i_pmd_good + l.addi r3,r0,0xffffe000 // PAGE_MASK +// l.j i_pmd_bad +// l.sw 0x0(r4),r0 // clear pmd + +i_pmd_good: + /* + * pte = *pte_offset(pmd, iaddr); + * + */ + l.lwz r4,0x0(r4) // get **pmd value + l.and r4,r4,r3 // & PAGE_MASK + l.srli r5,r2,0xd // >> PAGE_SHIFT, r2 == EEAR + l.andi r3,r5,0x7ff // (1UL << PAGE_SHIFT - 2) - 1 + l.slli r3,r3,0x2 // to get address << 2 + l.add r3,r3,r4 + l.lwz r2,0x0(r3) // this is pte at last + /* + * if (!pte_present(pte)) + * + */ + l.andi r4,r2,0x1 + l.sfne r4,r0 // is pte present + l.bnf i_pte_not_present + l.addi r3,r0,0xffffe03a // PAGE_MASK | ITLB_UP_CONVERT_MASK + /* + * fill ITLB TR register + */ + l.and r4,r2,r3 // apply the mask + l.andi r3,r2,0x7c0 // _PAGE_EXEC | _PAGE_SRE | _PAGE_SWE | _PAGE_URE | _PAGE_UWE +// l.andi r3,r2,0x400 // _PAGE_EXEC + l.sfeq r3,r0 + l.bf itlb_tr_fill //_workaround + // Determine number of IMMU sets + l.mfspr r6, r0, SPR_IMMUCFGR + l.andi r6, r6, SPR_IMMUCFGR_NTS + l.srli r6, r6, SPR_IMMUCFGR_NTS_OFF + l.ori r3, r0, 0x1 + l.sll r3, r3, r6 // r3 = number IMMU sets IMMUCFGR + l.addi r6, r3, -1 // r6 = nsets mask + l.and r5, r5, r6 // calc offset: & (NUM_TLB_ENTRIES-1) + +/* + * __PHX__ :: fixme + * we should not just blindly set executable flags, + * but it does help with ping. the clean way would be to find out + * (and fix it) why stack doesn't have execution permissions + */ + +itlb_tr_fill_workaround: + l.ori r4,r4,0xc0 // | (SPR_ITLBTR_UXE | ITLBTR_SXE) +itlb_tr_fill: + l.mtspr r5,r4,SPR_ITLBTR_BASE(0) + /* + * fill DTLB MR register + */ + l.mfspr r2,r0,SPR_EEAR_BASE + l.addi r3,r0,0xffffe000 // PAGE_MASK + l.and r4,r2,r3 // apply PAGE_MASK to EA (__PHX__ do we really need this?) + l.ori r4,r4,0x1 // set hardware valid bit: DTBL_MR entry + l.mtspr r5,r4,SPR_ITLBMR_BASE(0) + + EXCEPTION_LOAD_GPR2 + EXCEPTION_LOAD_GPR3 + EXCEPTION_LOAD_GPR4 + EXCEPTION_LOAD_GPR5 + EXCEPTION_LOAD_GPR6 + l.rfe + +i_pmd_bad: + l.nop 1 + EXCEPTION_LOAD_GPR2 + EXCEPTION_LOAD_GPR3 + EXCEPTION_LOAD_GPR4 + EXCEPTION_LOAD_GPR5 + EXCEPTION_LOAD_GPR6 + l.rfe +i_pmd_none: +i_pte_not_present: + EXCEPTION_LOAD_GPR2 + EXCEPTION_LOAD_GPR3 + EXCEPTION_LOAD_GPR4 + EXCEPTION_LOAD_GPR5 + EXCEPTION_LOAD_GPR6 + l.j _dispatch_do_ipage_fault + l.nop + +/* ==============================================[ boot tlb handlers ]=== */ + + +/* =================================================[ debugging aids ]=== */ + + .align 64 +_immu_trampoline: + .space 64 +_immu_trampoline_top: + +#define TRAMP_SLOT_0 (0x0) +#define TRAMP_SLOT_1 (0x4) +#define TRAMP_SLOT_2 (0x8) +#define TRAMP_SLOT_3 (0xc) +#define TRAMP_SLOT_4 (0x10) +#define TRAMP_SLOT_5 (0x14) +#define TRAMP_FRAME_SIZE (0x18) + +ENTRY(_immu_trampoline_workaround) + // r2 EEA + // r6 is physical EEA + tophys(r6,r2) + + LOAD_SYMBOL_2_GPR(r5,_immu_trampoline) + tophys (r3,r5) // r3 is trampoline (physical) + + LOAD_SYMBOL_2_GPR(r4,0x15000000) + l.sw TRAMP_SLOT_0(r3),r4 + l.sw TRAMP_SLOT_1(r3),r4 + l.sw TRAMP_SLOT_4(r3),r4 + l.sw TRAMP_SLOT_5(r3),r4 + + // EPC = EEA - 0x4 + l.lwz r4,0x0(r6) // load op @ EEA + 0x0 (fc address) + l.sw TRAMP_SLOT_3(r3),r4 // store it to _immu_trampoline_data + l.lwz r4,-0x4(r6) // load op @ EEA - 0x4 (f8 address) + l.sw TRAMP_SLOT_2(r3),r4 // store it to _immu_trampoline_data + + l.srli r5,r4,26 // check opcode for write access + l.sfeqi r5,0 // l.j + l.bf 0f + l.sfeqi r5,0x11 // l.jr + l.bf 1f + l.sfeqi r5,1 // l.jal + l.bf 2f + l.sfeqi r5,0x12 // l.jalr + l.bf 3f + l.sfeqi r5,3 // l.bnf + l.bf 4f + l.sfeqi r5,4 // l.bf + l.bf 5f +99: + l.nop + l.j 99b // should never happen + l.nop 1 + + // r2 is EEA + // r3 is trampoline address (physical) + // r4 is instruction + // r6 is physical(EEA) + // + // r5 + +2: // l.jal + + /* 19 20 aa aa l.movhi r9,0xaaaa + * a9 29 bb bb l.ori r9,0xbbbb + * + * where 0xaaaabbbb is EEA + 0x4 shifted right 2 + */ + + l.addi r6,r2,0x4 // this is 0xaaaabbbb + + // l.movhi r9,0xaaaa + l.ori r5,r0,0x1920 // 0x1920 == l.movhi r9 + l.sh (TRAMP_SLOT_0+0x0)(r3),r5 + l.srli r5,r6,16 + l.sh (TRAMP_SLOT_0+0x2)(r3),r5 + + // l.ori r9,0xbbbb + l.ori r5,r0,0xa929 // 0xa929 == l.ori r9 + l.sh (TRAMP_SLOT_1+0x0)(r3),r5 + l.andi r5,r6,0xffff + l.sh (TRAMP_SLOT_1+0x2)(r3),r5 + + /* falthrough, need to set up new jump offset */ + + +0: // l.j + l.slli r6,r4,6 // original offset shifted left 6 - 2 +// l.srli r6,r6,6 // original offset shifted right 2 + + l.slli r4,r2,4 // old jump position: EEA shifted left 4 +// l.srli r4,r4,6 // old jump position: shifted right 2 + + l.addi r5,r3,0xc // new jump position (physical) + l.slli r5,r5,4 // new jump position: shifted left 4 + + // calculate new jump offset + // new_off = old_off + (old_jump - new_jump) + + l.sub r5,r4,r5 // old_jump - new_jump + l.add r5,r6,r5 // orig_off + (old_jump - new_jump) + l.srli r5,r5,6 // new offset shifted right 2 + + // r5 is new jump offset + // l.j has opcode 0x0... + l.sw TRAMP_SLOT_2(r3),r5 // write it back + + l.j trampoline_out + l.nop + +/* ----------------------------- */ + +3: // l.jalr + + /* 19 20 aa aa l.movhi r9,0xaaaa + * a9 29 bb bb l.ori r9,0xbbbb + * + * where 0xaaaabbbb is EEA + 0x4 shifted right 2 + */ + + l.addi r6,r2,0x4 // this is 0xaaaabbbb + + // l.movhi r9,0xaaaa + l.ori r5,r0,0x1920 // 0x1920 == l.movhi r9 + l.sh (TRAMP_SLOT_0+0x0)(r3),r5 + l.srli r5,r6,16 + l.sh (TRAMP_SLOT_0+0x2)(r3),r5 + + // l.ori r9,0xbbbb + l.ori r5,r0,0xa929 // 0xa929 == l.ori r9 + l.sh (TRAMP_SLOT_1+0x0)(r3),r5 + l.andi r5,r6,0xffff + l.sh (TRAMP_SLOT_1+0x2)(r3),r5 + + l.lhz r5,(TRAMP_SLOT_2+0x0)(r3) // load hi part of jump instruction + l.andi r5,r5,0x3ff // clear out opcode part + l.ori r5,r5,0x4400 // opcode changed from l.jalr -> l.jr + l.sh (TRAMP_SLOT_2+0x0)(r3),r5 // write it back + + /* falthrough */ + +1: // l.jr + l.j trampoline_out + l.nop + +/* ----------------------------- */ + +4: // l.bnf +5: // l.bf + l.slli r6,r4,6 // original offset shifted left 6 - 2 +// l.srli r6,r6,6 // original offset shifted right 2 + + l.slli r4,r2,4 // old jump position: EEA shifted left 4 +// l.srli r4,r4,6 // old jump position: shifted right 2 + + l.addi r5,r3,0xc // new jump position (physical) + l.slli r5,r5,4 // new jump position: shifted left 4 + + // calculate new jump offset + // new_off = old_off + (old_jump - new_jump) + + l.add r6,r6,r4 // (orig_off + old_jump) + l.sub r6,r6,r5 // (orig_off + old_jump) - new_jump + l.srli r6,r6,6 // new offset shifted right 2 + + // r6 is new jump offset + l.lwz r4,(TRAMP_SLOT_2+0x0)(r3) // load jump instruction + l.srli r4,r4,16 + l.andi r4,r4,0xfc00 // get opcode part + l.slli r4,r4,16 + l.or r6,r4,r6 // l.b(n)f new offset + l.sw TRAMP_SLOT_2(r3),r6 // write it back + + /* we need to add l.j to EEA + 0x8 */ + tophys (r4,r2) // may not be needed (due to shifts down_ + l.addi r4,r4,(0x8 - 0x8) // jump target = r2 + 0x8 (compensate for 0x8) + // jump position = r5 + 0x8 (0x8 compensated) + l.sub r4,r4,r5 // jump offset = target - new_position + 0x8 + + l.slli r4,r4,4 // the amount of info in imediate of jump + l.srli r4,r4,6 // jump instruction with offset + l.sw TRAMP_SLOT_4(r3),r4 // write it to 4th slot + + /* fallthrough */ + +trampoline_out: + // set up new EPC to point to our trampoline code + LOAD_SYMBOL_2_GPR(r5,_immu_trampoline) + l.mtspr r0,r5,SPR_EPCR_BASE + + // immu_trampoline is (4x) CACHE_LINE aligned + // and only 6 instructions long, + // so we need to invalidate only 2 lines + + /* Establish cache block size + If BS=0, 16; + If BS=1, 32; + r14 contain block size + */ + l.mfspr r21,r0,SPR_ICCFGR + l.andi r21,r21,SPR_ICCFGR_CBS + l.srli r21,r21,7 + l.ori r23,r0,16 + l.sll r14,r23,r21 + + l.mtspr r0,r5,SPR_ICBIR + l.add r5,r5,r14 + l.mtspr r0,r5,SPR_ICBIR + + l.jr r9 + l.nop + + +/* + * DSCR: prints a string referenced by r3. + * + * PRMS: r3 - address of the first character of null + * terminated string to be printed + * + * PREQ: UART at UART_BASE_ADD has to be initialized + * + * POST: caller should be aware that r3, r9 are changed + */ +ENTRY(_emergency_print) + EMERGENCY_PRINT_STORE_GPR4 + EMERGENCY_PRINT_STORE_GPR5 + EMERGENCY_PRINT_STORE_GPR6 + EMERGENCY_PRINT_STORE_GPR7 +2: + l.lbz r7,0(r3) + l.sfeq r7,r0 + l.bf 9f + l.nop + +// putc: + l.movhi r4,hi(UART_BASE_ADD) + + l.addi r6,r0,0x20 +1: l.lbz r5,5(r4) + l.andi r5,r5,0x20 + l.sfeq r5,r6 + l.bnf 1b + l.nop + + l.sb 0(r4),r7 + + l.addi r6,r0,0x60 +1: l.lbz r5,5(r4) + l.andi r5,r5,0x60 + l.sfeq r5,r6 + l.bnf 1b + l.nop + + /* next character */ + l.j 2b + l.addi r3,r3,0x1 + +9: + EMERGENCY_PRINT_LOAD_GPR7 + EMERGENCY_PRINT_LOAD_GPR6 + EMERGENCY_PRINT_LOAD_GPR5 + EMERGENCY_PRINT_LOAD_GPR4 + l.jr r9 + l.nop + +ENTRY(_emergency_print_nr) + EMERGENCY_PRINT_STORE_GPR4 + EMERGENCY_PRINT_STORE_GPR5 + EMERGENCY_PRINT_STORE_GPR6 + EMERGENCY_PRINT_STORE_GPR7 + EMERGENCY_PRINT_STORE_GPR8 + + l.addi r8,r0,32 // shift register + +1: /* remove leading zeros */ + l.addi r8,r8,-0x4 + l.srl r7,r3,r8 + l.andi r7,r7,0xf + + /* don't skip the last zero if number == 0x0 */ + l.sfeqi r8,0x4 + l.bf 2f + l.nop + + l.sfeq r7,r0 + l.bf 1b + l.nop + +2: + l.srl r7,r3,r8 + + l.andi r7,r7,0xf + l.sflts r8,r0 + l.bf 9f + + l.sfgtui r7,0x9 + l.bnf 8f + l.nop + l.addi r7,r7,0x27 + +8: + l.addi r7,r7,0x30 +// putc: + l.movhi r4,hi(UART_BASE_ADD) + + l.addi r6,r0,0x20 +1: l.lbz r5,5(r4) + l.andi r5,r5,0x20 + l.sfeq r5,r6 + l.bnf 1b + l.nop + + l.sb 0(r4),r7 + + l.addi r6,r0,0x60 +1: l.lbz r5,5(r4) + l.andi r5,r5,0x60 + l.sfeq r5,r6 + l.bnf 1b + l.nop + + /* next character */ + l.j 2b + l.addi r8,r8,-0x4 + +9: + EMERGENCY_PRINT_LOAD_GPR8 + EMERGENCY_PRINT_LOAD_GPR7 + EMERGENCY_PRINT_LOAD_GPR6 + EMERGENCY_PRINT_LOAD_GPR5 + EMERGENCY_PRINT_LOAD_GPR4 + l.jr r9 + l.nop + + +/* + * This should be used for debugging only. + * It messes up the Linux early serial output + * somehow, so use it sparingly and essentially + * only if you need to debug something that goes wrong + * before Linux gets the early serial going. + * + * Furthermore, you'll have to make sure you set the + * UART_DEVISOR correctly according to the system + * clock rate. + * + * + */ + + + +#define SYS_CLK 20000000 +//#define SYS_CLK 1843200 +#define OR32_CONSOLE_BAUD 115200 +#define UART_DIVISOR SYS_CLK/(16*OR32_CONSOLE_BAUD) + +ENTRY(_early_uart_init) + l.movhi r3,hi(UART_BASE_ADD) + + l.addi r4,r0,0x7 + l.sb 0x2(r3),r4 + + l.addi r4,r0,0x0 + l.sb 0x1(r3),r4 + + l.addi r4,r0,0x3 + l.sb 0x3(r3),r4 + + l.lbz r5,3(r3) + l.ori r4,r5,0x80 + l.sb 0x3(r3),r4 + l.addi r4,r0,((UART_DIVISOR>>8) & 0x000000ff) + l.sb UART_DLM(r3),r4 + l.addi r4,r0,((UART_DIVISOR) & 0x000000ff) + l.sb UART_DLL(r3),r4 + l.sb 0x3(r3),r5 + + l.jr r9 + l.nop + +_string_copying_linux: + .string "\n\n\n\n\n\rCopying Linux... \0" + +_string_ok_booting: + .string "Ok, booting the kernel.\n\r\0" + +_string_unhandled_exception: + .string "\n\rRunarunaround: Unhandled exception 0x\0" + +_string_epc_prefix: + .string ": EPC=0x\0" + +_string_nl: + .string "\n\r\0" + + .global _string_esr_irq_bug +_string_esr_irq_bug: + .string "\n\rESR external interrupt bug, for details look into entry.S\n\r\0" + + + +/* ========================================[ page aligned structures ]=== */ + +/* + * .data section should be page aligned + * (look into arch/or32/kernel/vmlinux.lds) + */ + .section .data,"aw" + .align 8192 + .global empty_zero_page +empty_zero_page: + .space 8192 + + .global swapper_pg_dir +swapper_pg_dir: + .space 8192 + + .global _unhandled_stack +_unhandled_stack: + .space 8192 +_unhandled_stack_top: + +/* ============================================================[ EOF ]=== */ diff --git a/arch/openrisc/kernel/init_task.c b/arch/openrisc/kernel/init_task.c new file mode 100644 index 000000000000..45744a384927 --- /dev/null +++ b/arch/openrisc/kernel/init_task.c @@ -0,0 +1,41 @@ +/* + * OpenRISC init_task.c + * + * Linux architectural port borrowing liberally from similar works of + * others. All original copyrights apply as per the original source + * declaration. + * + * Modifications for the OpenRISC architecture: + * Copyright (C) 2003 Matjaz Breskvar + * Copyright (C) 2010-2011 Jonas Bonn + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ + +#include +#include + +static struct signal_struct init_signals = INIT_SIGNALS(init_signals); +static struct sighand_struct init_sighand = INIT_SIGHAND(init_sighand); + +/* + * Initial thread structure. + * + * We need to make sure that this is THREAD_SIZE aligned due to the + * way process stacks are handled. This is done by having a special + * "init_task" linker map entry.. + */ +union thread_union init_thread_union __init_task_data = { + INIT_THREAD_INFO(init_task) +}; + +/* + * Initial task structure. + * + * All other task structs will be allocated on slabs in fork.c + */ +struct task_struct init_task = INIT_TASK(init_task); +EXPORT_SYMBOL(init_task); diff --git a/arch/openrisc/kernel/setup.c b/arch/openrisc/kernel/setup.c new file mode 100644 index 000000000000..1422f747f52b --- /dev/null +++ b/arch/openrisc/kernel/setup.c @@ -0,0 +1,381 @@ +/* + * OpenRISC setup.c + * + * Linux architectural port borrowing liberally from similar works of + * others. All original copyrights apply as per the original source + * declaration. + * + * Modifications for the OpenRISC architecture: + * Copyright (C) 2003 Matjaz Breskvar + * Copyright (C) 2010-2011 Jonas Bonn + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + * + * This file handles the architecture-dependent parts of initialization + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include +#include +#include +#include + +#include "vmlinux.h" + +char __initdata cmd_line[COMMAND_LINE_SIZE] = CONFIG_CMDLINE; + +static unsigned long __init setup_memory(void) +{ + unsigned long bootmap_size; + unsigned long ram_start_pfn; + unsigned long free_ram_start_pfn; + unsigned long ram_end_pfn; + phys_addr_t memory_start, memory_end; + struct memblock_region *region; + + memory_end = memory_start = 0; + + /* Find main memory where is the kernel */ + for_each_memblock(memory, region) { + memory_start = region->base; + memory_end = region->base + region->size; + printk(KERN_INFO "%s: Memory: 0x%x-0x%x\n", __func__, + memory_start, memory_end); + } + + if (!memory_end) { + panic("No memory!"); + } + + ram_start_pfn = PFN_UP(memory_start); + /* free_ram_start_pfn is first page after kernel */ + free_ram_start_pfn = PFN_UP(__pa(&_end)); + ram_end_pfn = PFN_DOWN(memblock_end_of_DRAM()); + + max_pfn = ram_end_pfn; + + /* + * initialize the boot-time allocator (with low memory only). + * + * This makes the memory from the end of the kernel to the end of + * RAM usable. + * init_bootmem sets the global values min_low_pfn, max_low_pfn. + */ + bootmap_size = init_bootmem(free_ram_start_pfn, + ram_end_pfn - ram_start_pfn); + free_bootmem(PFN_PHYS(free_ram_start_pfn), + (ram_end_pfn - free_ram_start_pfn) << PAGE_SHIFT); + reserve_bootmem(PFN_PHYS(free_ram_start_pfn), bootmap_size, + BOOTMEM_DEFAULT); + + for_each_memblock(reserved, region) { + printk(KERN_INFO "Reserved - 0x%08x-0x%08x\n", + (u32) region->base, (u32) region->size); + reserve_bootmem(region->base, region->size, BOOTMEM_DEFAULT); + } + + return ram_end_pfn; +} + +struct cpuinfo cpuinfo; + +static void print_cpuinfo(void) +{ + unsigned long upr = mfspr(SPR_UPR); + unsigned long vr = mfspr(SPR_VR); + unsigned int version; + unsigned int revision; + + version = (vr & SPR_VR_VER) >> 24; + revision = (vr & SPR_VR_REV); + + printk(KERN_INFO "CPU: OpenRISC-%x (revision %d) @%d MHz\n", + version, revision, cpuinfo.clock_frequency / 1000000); + + if (!(upr & SPR_UPR_UP)) { + printk(KERN_INFO + "-- no UPR register... unable to detect configuration\n"); + return; + } + + if (upr & SPR_UPR_DCP) + printk(KERN_INFO + "-- dcache: %4d bytes total, %2d bytes/line, %d way(s)\n", + cpuinfo.dcache_size, cpuinfo.dcache_block_size, 1); + else + printk(KERN_INFO "-- dcache disabled\n"); + if (upr & SPR_UPR_ICP) + printk(KERN_INFO + "-- icache: %4d bytes total, %2d bytes/line, %d way(s)\n", + cpuinfo.icache_size, cpuinfo.icache_block_size, 1); + else + printk(KERN_INFO "-- icache disabled\n"); + + if (upr & SPR_UPR_DMP) + printk(KERN_INFO "-- dmmu: %4d entries, %lu way(s)\n", + 1 << ((mfspr(SPR_DMMUCFGR) & SPR_DMMUCFGR_NTS) >> 2), + 1 + (mfspr(SPR_DMMUCFGR) & SPR_DMMUCFGR_NTW)); + if (upr & SPR_UPR_IMP) + printk(KERN_INFO "-- immu: %4d entries, %lu way(s)\n", + 1 << ((mfspr(SPR_IMMUCFGR) & SPR_IMMUCFGR_NTS) >> 2), + 1 + (mfspr(SPR_IMMUCFGR) & SPR_IMMUCFGR_NTW)); + + printk(KERN_INFO "-- additional features:\n"); + if (upr & SPR_UPR_DUP) + printk(KERN_INFO "-- debug unit\n"); + if (upr & SPR_UPR_PCUP) + printk(KERN_INFO "-- performance counters\n"); + if (upr & SPR_UPR_PMP) + printk(KERN_INFO "-- power management\n"); + if (upr & SPR_UPR_PICP) + printk(KERN_INFO "-- PIC\n"); + if (upr & SPR_UPR_TTP) + printk(KERN_INFO "-- timer\n"); + if (upr & SPR_UPR_CUP) + printk(KERN_INFO "-- custom unit(s)\n"); +} + +void __init setup_cpuinfo(void) +{ + struct device_node *cpu; + unsigned long iccfgr, dccfgr; + unsigned long cache_set_size, cache_ways; + + cpu = of_find_compatible_node(NULL, NULL, "opencores,or1200-rtlsvn481"); + if (!cpu) + panic("No compatible CPU found in device tree...\n"); + + iccfgr = mfspr(SPR_ICCFGR); + cache_ways = 1 << (iccfgr & SPR_ICCFGR_NCW); + cache_set_size = 1 << ((iccfgr & SPR_ICCFGR_NCS) >> 3); + cpuinfo.icache_block_size = 16 << ((iccfgr & SPR_ICCFGR_CBS) >> 7); + cpuinfo.icache_size = + cache_set_size * cache_ways * cpuinfo.icache_block_size; + + dccfgr = mfspr(SPR_DCCFGR); + cache_ways = 1 << (dccfgr & SPR_DCCFGR_NCW); + cache_set_size = 1 << ((dccfgr & SPR_DCCFGR_NCS) >> 3); + cpuinfo.dcache_block_size = 16 << ((dccfgr & SPR_DCCFGR_CBS) >> 7); + cpuinfo.dcache_size = + cache_set_size * cache_ways * cpuinfo.dcache_block_size; + + if (of_property_read_u32(cpu, "clock-frequency", + &cpuinfo.clock_frequency)) { + printk(KERN_WARNING + "Device tree missing CPU 'clock-frequency' parameter." + "Assuming frequency 25MHZ" + "This is probably not what you want."); + } + + of_node_put(cpu); + + print_cpuinfo(); +} + +/** + * or32_early_setup + * + * Handles the pointer to the device tree that this kernel is to use + * for establishing the available platform devices. + * + * For now, this is limited to using the built-in device tree. In the future, + * it is intended that this function will take a pointer to the device tree + * that is potentially built-in, but potentially also passed in by the + * bootloader, or discovered by some equally clever means... + */ + +void __init or32_early_setup(void) +{ + + early_init_devtree(__dtb_start); + + printk(KERN_INFO "Compiled-in FDT at 0x%p\n", __dtb_start); +} + +static int __init openrisc_device_probe(void) +{ + of_platform_populate(NULL, NULL, NULL, NULL); + + return 0; +} + +device_initcall(openrisc_device_probe); + +static inline unsigned long extract_value_bits(unsigned long reg, + short bit_nr, short width) +{ + return (reg >> bit_nr) & (0 << width); +} + +static inline unsigned long extract_value(unsigned long reg, unsigned long mask) +{ + while (!(mask & 0x1)) { + reg = reg >> 1; + mask = mask >> 1; + } + return mask & reg; +} + +void __init detect_unit_config(unsigned long upr, unsigned long mask, + char *text, void (*func) (void)) +{ + if (text != NULL) + printk("%s", text); + + if (upr & mask) { + if (func != NULL) + func(); + else + printk("present\n"); + } else + printk("not present\n"); +} + +/* + * calibrate_delay + * + * Lightweight calibrate_delay implementation that calculates loops_per_jiffy + * from the clock frequency passed in via the device tree + * + */ + +void __cpuinit calibrate_delay(void) +{ + const int *val; + struct device_node *cpu = NULL; + cpu = of_find_compatible_node(NULL, NULL, "opencores,or1200-rtlsvn481"); + val = of_get_property(cpu, "clock-frequency", NULL); + if (!val) + panic("no cpu 'clock-frequency' parameter in device tree"); + loops_per_jiffy = *val / HZ; + pr_cont("%lu.%02lu BogoMIPS (lpj=%lu)\n", + loops_per_jiffy / (500000 / HZ), + (loops_per_jiffy / (5000 / HZ)) % 100, loops_per_jiffy); +} + +void __init setup_arch(char **cmdline_p) +{ + unsigned long max_low_pfn; + + unflatten_device_tree(); + + setup_cpuinfo(); + + /* process 1's initial memory region is the kernel code/data */ + init_mm.start_code = (unsigned long)&_stext; + init_mm.end_code = (unsigned long)&_etext; + init_mm.end_data = (unsigned long)&_edata; + init_mm.brk = (unsigned long)&_end; + +#ifdef CONFIG_BLK_DEV_INITRD + initrd_start = (unsigned long)&__initrd_start; + initrd_end = (unsigned long)&__initrd_end; + if (initrd_start == initrd_end) { + initrd_start = 0; + initrd_end = 0; + } + initrd_below_start_ok = 1; +#endif + + /* setup bootmem allocator */ + max_low_pfn = setup_memory(); + + /* paging_init() sets up the MMU and marks all pages as reserved */ + paging_init(); + +#if defined(CONFIG_VT) && defined(CONFIG_DUMMY_CONSOLE) + if (!conswitchp) + conswitchp = &dummy_con; +#endif + + *cmdline_p = cmd_line; + + printk(KERN_INFO "OpenRISC Linux -- http://openrisc.net\n"); +} + +static int show_cpuinfo(struct seq_file *m, void *v) +{ + unsigned long vr; + int version, revision; + + vr = mfspr(SPR_VR); + version = (vr & SPR_VR_VER) >> 24; + revision = vr & SPR_VR_REV; + + return seq_printf(m, + "cpu\t\t: OpenRISC-%x\n" + "revision\t: %d\n" + "frequency\t: %ld\n" + "dcache size\t: %d bytes\n" + "dcache block size\t: %d bytes\n" + "icache size\t: %d bytes\n" + "icache block size\t: %d bytes\n" + "immu\t\t: %d entries, %lu ways\n" + "dmmu\t\t: %d entries, %lu ways\n" + "bogomips\t: %lu.%02lu\n", + version, + revision, + loops_per_jiffy * HZ, + cpuinfo.dcache_size, + cpuinfo.dcache_block_size, + cpuinfo.icache_size, + cpuinfo.icache_block_size, + 1 << ((mfspr(SPR_DMMUCFGR) & SPR_DMMUCFGR_NTS) >> 2), + 1 + (mfspr(SPR_DMMUCFGR) & SPR_DMMUCFGR_NTW), + 1 << ((mfspr(SPR_IMMUCFGR) & SPR_IMMUCFGR_NTS) >> 2), + 1 + (mfspr(SPR_IMMUCFGR) & SPR_IMMUCFGR_NTW), + (loops_per_jiffy * HZ) / 500000, + ((loops_per_jiffy * HZ) / 5000) % 100); +} + +static void *c_start(struct seq_file *m, loff_t * pos) +{ + /* We only have one CPU... */ + return *pos < 1 ? (void *)1 : NULL; +} + +static void *c_next(struct seq_file *m, void *v, loff_t * pos) +{ + ++*pos; + return NULL; +} + +static void c_stop(struct seq_file *m, void *v) +{ +} + +const struct seq_operations cpuinfo_op = { + .start = c_start, + .next = c_next, + .stop = c_stop, + .show = show_cpuinfo, +};