linux_dsm_epyc7002/arch/nios2/kernel/entry.S

561 lines
14 KiB
ArmAsm
Raw Normal View History

/*
* linux/arch/nios2/kernel/entry.S
*
* Copyright (C) 2013-2014 Altera Corporation
* Copyright (C) 2009, Wind River Systems Inc
*
* Implemented by fredrik.markstrom@gmail.com and ivarholmqvist@gmail.com
*
* Copyright (C) 1999-2002, Greg Ungerer (gerg@snapgear.com)
* Copyright (C) 1998 D. Jeff Dionne <jeff@lineo.ca>,
* Kenneth Albanowski <kjahds@kjahds.com>,
* Copyright (C) 2000 Lineo Inc. (www.lineo.com)
* Copyright (C) 2004 Microtronix Datacom Ltd.
*
* This file is subject to the terms and conditions of the GNU General Public
* License. See the file "COPYING" in the main directory of this archive
* for more details.
*
* Linux/m68k support by Hamish Macdonald
*
* 68060 fixes by Jesper Skov
* ColdFire support by Greg Ungerer (gerg@snapgear.com)
* 5307 fixes by David W. Miller
* linux 2.4 support David McCullough <davidm@snapgear.com>
*/
#include <linux/sys.h>
#include <linux/linkage.h>
#include <asm/asm-offsets.h>
#include <asm/asm-macros.h>
#include <asm/thread_info.h>
#include <asm/errno.h>
#include <asm/setup.h>
#include <asm/entry.h>
#include <asm/unistd.h>
#include <asm/processor.h>
.macro GET_THREAD_INFO reg
.if THREAD_SIZE & 0xffff0000
andhi \reg, sp, %hi(~(THREAD_SIZE-1))
.else
addi \reg, r0, %lo(~(THREAD_SIZE-1))
and \reg, \reg, sp
.endif
.endm
.macro kuser_cmpxchg_check
/*
* Make sure our user space atomic helper is restarted if it was
* interrupted in a critical region.
* ea-4 = address of interrupted insn (ea must be preserved).
* sp = saved regs.
* cmpxchg_ldw = first critical insn, cmpxchg_stw = last critical insn.
* If ea <= cmpxchg_stw and ea > cmpxchg_ldw then saved EA is set to
* cmpxchg_ldw + 4.
*/
/* et = cmpxchg_stw + 4 */
movui et, (KUSER_BASE + 4 + (cmpxchg_stw - __kuser_helper_start))
bgtu ea, et, 1f
subi et, et, (cmpxchg_stw - cmpxchg_ldw) /* et = cmpxchg_ldw + 4 */
bltu ea, et, 1f
stw et, PT_EA(sp) /* fix up EA */
mov ea, et
1:
.endm
.section .rodata
.align 4
exception_table:
.word unhandled_exception /* 0 - Reset */
.word unhandled_exception /* 1 - Processor-only Reset */
.word external_interrupt /* 2 - Interrupt */
.word handle_trap /* 3 - Trap Instruction */
.word instruction_trap /* 4 - Unimplemented instruction */
.word handle_illegal /* 5 - Illegal instruction */
.word handle_unaligned /* 6 - Misaligned data access */
.word handle_unaligned /* 7 - Misaligned destination address */
.word handle_diverror /* 8 - Division error */
.word protection_exception_ba /* 9 - Supervisor-only instr. address */
.word protection_exception_instr /* 10 - Supervisor only instruction */
.word protection_exception_ba /* 11 - Supervisor only data address */
.word unhandled_exception /* 12 - Double TLB miss (data) */
.word protection_exception_pte /* 13 - TLB permission violation (x) */
.word protection_exception_pte /* 14 - TLB permission violation (r) */
.word protection_exception_pte /* 15 - TLB permission violation (w) */
.word unhandled_exception /* 16 - MPU region violation */
trap_table:
.word handle_system_call /* 0 */
.word handle_trap_1 /* 1 */
.word handle_trap_2 /* 2 */
.word handle_trap_3 /* 3 */
.word handle_trap_reserved /* 4 */
.word handle_trap_reserved /* 5 */
.word handle_trap_reserved /* 6 */
.word handle_trap_reserved /* 7 */
.word handle_trap_reserved /* 8 */
.word handle_trap_reserved /* 9 */
.word handle_trap_reserved /* 10 */
.word handle_trap_reserved /* 11 */
.word handle_trap_reserved /* 12 */
.word handle_trap_reserved /* 13 */
.word handle_trap_reserved /* 14 */
.word handle_trap_reserved /* 15 */
.word handle_trap_reserved /* 16 */
.word handle_trap_reserved /* 17 */
.word handle_trap_reserved /* 18 */
.word handle_trap_reserved /* 19 */
.word handle_trap_reserved /* 20 */
.word handle_trap_reserved /* 21 */
.word handle_trap_reserved /* 22 */
.word handle_trap_reserved /* 23 */
.word handle_trap_reserved /* 24 */
.word handle_trap_reserved /* 25 */
.word handle_trap_reserved /* 26 */
.word handle_trap_reserved /* 27 */
.word handle_trap_reserved /* 28 */
.word handle_trap_reserved /* 29 */
#ifdef CONFIG_KGDB
.word handle_kgdb_breakpoint /* 30 KGDB breakpoint */
#else
.word instruction_trap /* 30 */
#endif
.word handle_breakpoint /* 31 */
.text
.set noat
.set nobreak
ENTRY(inthandler)
SAVE_ALL
kuser_cmpxchg_check
/* Clear EH bit before we get a new excpetion in the kernel
* and after we have saved it to the exception frame. This is done
* whether it's trap, tlb-miss or interrupt. If we don't do this
* estatus is not updated the next exception.
*/
rdctl r24, status
movi r9, %lo(~STATUS_EH)
and r24, r24, r9
wrctl status, r24
/* Read cause and vector and branch to the associated handler */
mov r4, sp
rdctl r5, exception
movia r9, exception_table
add r24, r9, r5
ldw r24, 0(r24)
jmp r24
/***********************************************************************
* Handle traps
***********************************************************************
*/
ENTRY(handle_trap)
ldwio r24, -4(ea) /* instruction that caused the exception */
srli r24, r24, 4
andi r24, r24, 0x7c
movia r9,trap_table
add r24, r24, r9
ldw r24, 0(r24)
jmp r24
/***********************************************************************
* Handle system calls
***********************************************************************
*/
ENTRY(handle_system_call)
/* Enable interrupts */
rdctl r10, status
ori r10, r10, STATUS_PIE
wrctl status, r10
/* Reload registers destroyed by common code. */
ldw r4, PT_R4(sp)
ldw r5, PT_R5(sp)
local_restart:
/* Check that the requested system call is within limits */
movui r1, __NR_syscalls
bgeu r2, r1, ret_invsyscall
slli r1, r2, 2
movhi r11, %hiadj(sys_call_table)
add r1, r1, r11
ldw r1, %lo(sys_call_table)(r1)
beq r1, r0, ret_invsyscall
/* Check if we are being traced */
GET_THREAD_INFO r11
ldw r11,TI_FLAGS(r11)
BTBNZ r11,r11,TIF_SYSCALL_TRACE,traced_system_call
/* Execute the system call */
callr r1
/* If the syscall returns a negative result:
* Set r7 to 1 to indicate error,
* Negate r2 to get a positive error code
* If the syscall returns zero or a positive value:
* Set r7 to 0.
* The sigreturn system calls will skip the code below by
* adding to register ra. To avoid destroying registers
*/
translate_rc_and_ret:
movi r1, 0
bge r2, zero, 3f
sub r2, zero, r2
movi r1, 1
3:
stw r2, PT_R2(sp)
stw r1, PT_R7(sp)
end_translate_rc_and_ret:
ret_from_exception:
ldw r1, PT_ESTATUS(sp)
/* if so, skip resched, signals */
TSTBNZ r1, r1, ESTATUS_EU, Luser_return
restore_all:
rdctl r10, status /* disable intrs */
andi r10, r10, %lo(~STATUS_PIE)
wrctl status, r10
RESTORE_ALL
eret
/* If the syscall number was invalid return ENOSYS */
ret_invsyscall:
movi r2, -ENOSYS
br translate_rc_and_ret
/* This implements the same as above, except it calls
* do_syscall_trace_enter and do_syscall_trace_exit before and after the
* syscall in order for utilities like strace and gdb to work.
*/
traced_system_call:
SAVE_SWITCH_STACK
call do_syscall_trace_enter
RESTORE_SWITCH_STACK
/* Create system call register arguments. The 5th and 6th
arguments on stack are already in place at the beginning
of pt_regs. */
ldw r2, PT_R2(sp)
ldw r4, PT_R4(sp)
ldw r5, PT_R5(sp)
ldw r6, PT_R6(sp)
ldw r7, PT_R7(sp)
/* Fetch the syscall function, we don't need to check the boundaries
* since this is already done.
*/
slli r1, r2, 2
movhi r11,%hiadj(sys_call_table)
add r1, r1, r11
ldw r1, %lo(sys_call_table)(r1)
callr r1
/* If the syscall returns a negative result:
* Set r7 to 1 to indicate error,
* Negate r2 to get a positive error code
* If the syscall returns zero or a positive value:
* Set r7 to 0.
* The sigreturn system calls will skip the code below by
* adding to register ra. To avoid destroying registers
*/
translate_rc_and_ret2:
movi r1, 0
bge r2, zero, 4f
sub r2, zero, r2
movi r1, 1
4:
stw r2, PT_R2(sp)
stw r1, PT_R7(sp)
end_translate_rc_and_ret2:
SAVE_SWITCH_STACK
call do_syscall_trace_exit
RESTORE_SWITCH_STACK
br ret_from_exception
Luser_return:
GET_THREAD_INFO r11 /* get thread_info pointer */
ldw r10, TI_FLAGS(r11) /* get thread_info->flags */
ANDI32 r11, r10, _TIF_WORK_MASK
beq r11, r0, restore_all /* Nothing to do */
BTBZ r1, r10, TIF_NEED_RESCHED, Lsignal_return
/* Reschedule work */
call schedule
br ret_from_exception
Lsignal_return:
ANDI32 r1, r10, _TIF_SIGPENDING | _TIF_NOTIFY_RESUME
beq r1, r0, restore_all
mov r4, sp /* pt_regs */
SAVE_SWITCH_STACK
call do_notify_resume
beq r2, r0, no_work_pending
RESTORE_SWITCH_STACK
/* prepare restart syscall here without leaving kernel */
ldw r2, PT_R2(sp) /* reload syscall number in r2 */
ldw r4, PT_R4(sp) /* reload syscall arguments r4-r9 */
ldw r5, PT_R5(sp)
ldw r6, PT_R6(sp)
ldw r7, PT_R7(sp)
ldw r8, PT_R8(sp)
ldw r9, PT_R9(sp)
br local_restart /* restart syscall */
no_work_pending:
RESTORE_SWITCH_STACK
br ret_from_exception
/***********************************************************************
* Handle external interrupts.
***********************************************************************
*/
/*
* This is the generic interrupt handler (for all hardware interrupt
* sources). It figures out the vector number and calls the appropriate
* interrupt service routine directly.
*/
external_interrupt:
rdctl r12, ipending
rdctl r9, ienable
and r12, r12, r9
/* skip if no interrupt is pending */
beq r12, r0, ret_from_interrupt
movi r24, -1
stw r24, PT_ORIG_R2(sp)
/*
* Process an external hardware interrupt.
*/
addi ea, ea, -4 /* re-issue the interrupted instruction */
stw ea, PT_EA(sp)
2: movi r4, %lo(-1) /* Start from bit position 0,
highest priority */
/* This is the IRQ # for handler call */
1: andi r10, r12, 1 /* Isolate bit we are interested in */
srli r12, r12, 1 /* shift count is costly without hardware
multiplier */
addi r4, r4, 1
beq r10, r0, 1b
mov r5, sp /* Setup pt_regs pointer for handler call */
call do_IRQ
rdctl r12, ipending /* check again if irq still pending */
rdctl r9, ienable /* Isolate possible interrupts */
and r12, r12, r9
bne r12, r0, 2b
/* br ret_from_interrupt */ /* fall through to ret_from_interrupt */
ENTRY(ret_from_interrupt)
ldw r1, PT_ESTATUS(sp) /* check if returning to kernel */
TSTBNZ r1, r1, ESTATUS_EU, Luser_return
#ifdef CONFIG_PREEMPTION
GET_THREAD_INFO r1
ldw r4, TI_PREEMPT_COUNT(r1)
bne r4, r0, restore_all
ldw r4, TI_FLAGS(r1) /* ? Need resched set */
BTBZ r10, r4, TIF_NEED_RESCHED, restore_all
ldw r4, PT_ESTATUS(sp) /* ? Interrupts off */
andi r10, r4, ESTATUS_EPIE
beq r10, r0, restore_all
call preempt_schedule_irq
#endif
br restore_all
/***********************************************************************
* A few syscall wrappers
***********************************************************************
*/
/*
* int clone(unsigned long clone_flags, unsigned long newsp,
* int __user * parent_tidptr, int __user * child_tidptr,
* int tls_val)
*/
ENTRY(sys_clone)
SAVE_SWITCH_STACK
subi sp, sp, 4 /* make space for tls pointer */
stw r8, 0(sp) /* pass tls pointer (r8) via stack (5th argument) */
nios2: enable HAVE_COPY_THREAD_TLS, switch to kernel_clone_args This is part of a larger series that aims at getting rid of the copy_thread()/copy_thread_tls() split that makes the process creation codepaths in the kernel more convoluted and error-prone than they need to be. I'm converting all the remaining arches that haven't yet switched and am collecting individual acks. Once I have them, I'll send the whole series removing the copy_thread()/copy_thread_tls() split, the HAVE_COPY_THREAD_TLS define and the legacy do_fork() helper. The only kernel-wide process creation entry point for anything not going directly through the syscall path will then be based on struct kernel_clone_args. No more danger of weird process creation abi quirks between architectures hopefully, and easier to maintain overall. It also unblocks implementing clone3() on architectures not support copy_thread_tls(). Any architecture that wants to implement clone3() will need to select HAVE_COPY_THREAD_TLS and thus need to implement copy_thread_tls(). So both goals are connected but independently beneficial. HAVE_COPY_THREAD_TLS means that a given architecture supports CLONE_SETTLS and not setting it should usually mean that the architectures doesn't implement it but that's not how things are. In fact all architectures support CLONE_TLS it's just that they don't follow the calling convention that HAVE_COPY_THREAD_TLS implies. That means all architectures can be switched over to select HAVE_COPY_THREAD_TLS. Once that is done we can remove that macro (yay, less code), the unnecessary do_fork() export in kernel/fork.c, and also rename copy_thread_tls() back to copy_thread(). At this point copy_thread() becomes the main architecture specific part of process creation but it will be the same layout and calling convention for all architectures. (Once that is done we can probably cleanup each copy_thread() function even more but that's for the future.) Since nios2 does support CLONE_SETTLS there's no reason to not select HAVE_COPY_THREAD_TLS. This brings us one step closer to getting rid of the copy_thread()/copy_thread_tls() split we still have and ultimately the HAVE_COPY_THREAD_TLS define in general. A lot of architectures have already converted and nios2 is one of the few hat haven't yet. This also unblocks implementing the clone3() syscall on nios2. Once that is done we can get of another ARCH_WANTS_* macro. Once Any architecture that supports HAVE_COPY_THREAD_TLS cannot call the do_fork() helper anymore. This is fine and intended since it should be removed in favor of the new, cleaner _do_fork() calling convention based on struct kernel_clone_args. In fact, most architectures have already switched. With this patch, nios2 joins the other arches which can't use the fork(), vfork(), clone(), clone3() syscalls directly and who follow the new process creation calling convention that is based on struct kernel_clone_args which we introduced a while back. This means less custom assembly in the architectures entry path to set up the registers before calling into the process creation helper and it is easier to to support new features without having to adapt calling conventions. It also unifies all process creation paths between fork(), vfork(), clone(), and clone3(). (We can't fix the ABI nightmare that legacy clone() is but we can prevent stuff like this happening in the future.) For some more context, please see: commit 606e9ad20094f6d500166881d301f31a51bc8aa7 Merge: ac61145a725a 457677c70c76 Author: Linus Torvalds <torvalds@linux-foundation.org> Date: Sat Jan 11 15:33:48 2020 -0800 Merge tag 'clone3-tls-v5.5-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux Pull thread fixes from Christian Brauner: "This contains a series of patches to fix CLONE_SETTLS when used with clone3(). The clone3() syscall passes the tls argument through struct clone_args instead of a register. This means, all architectures that do not implement copy_thread_tls() but still support CLONE_SETTLS via copy_thread() expecting the tls to be located in a register argument based on clone() are currently unfortunately broken. Their tls value will be garbage. The patch series fixes this on all architectures that currently define __ARCH_WANT_SYS_CLONE3. It also adds a compile-time check to ensure that any architecture that enables clone3() in the future is forced to also implement copy_thread_tls(). My ultimate goal is to get rid of the copy_thread()/copy_thread_tls() split and just have copy_thread_tls() at some point in the not too distant future (Maybe even renaming copy_thread_tls() back to simply copy_thread() once the old function is ripped from all arches). This is dependent now on all arches supporting clone3(). While all relevant arches do that now there are still four missing: ia64, m68k, sh and sparc. They have the system call reserved, but not implemented. Once they all implement clone3() we can get rid of ARCH_WANT_SYS_CLONE3 and HAVE_COPY_THREAD_TLS. Note that in the meantime, m68k has already switched to the new calling convention. And I've got sparc patches acked by Dave and ia64 is already done too. You can find a link to a booting qemu nios2 system with all the changes here at [1]. [1]: https://asciinema.org/a/333353 Cc: linux-kernel@vger.kernel.org Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de> Acked-by: Ley Foon Tan <ley.foon.tan@intel.com> Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
2020-05-07 20:14:27 +07:00
call nios2_clone
addi sp, sp, 4
RESTORE_SWITCH_STACK
ret
ENTRY(sys_rt_sigreturn)
SAVE_SWITCH_STACK
mov r4, sp
call do_rt_sigreturn
RESTORE_SWITCH_STACK
addi ra, ra, (end_translate_rc_and_ret - translate_rc_and_ret)
ret
/***********************************************************************
* A few other wrappers and stubs
***********************************************************************
*/
protection_exception_pte:
rdctl r6, pteaddr
slli r6, r6, 10
call do_page_fault
br ret_from_exception
protection_exception_ba:
rdctl r6, badaddr
call do_page_fault
br ret_from_exception
protection_exception_instr:
call handle_supervisor_instr
br ret_from_exception
handle_breakpoint:
call breakpoint_c
br ret_from_exception
#ifdef CONFIG_NIOS2_ALIGNMENT_TRAP
handle_unaligned:
SAVE_SWITCH_STACK
call handle_unaligned_c
RESTORE_SWITCH_STACK
br ret_from_exception
#else
handle_unaligned:
call handle_unaligned_c
br ret_from_exception
#endif
handle_illegal:
call handle_illegal_c
br ret_from_exception
handle_diverror:
call handle_diverror_c
br ret_from_exception
#ifdef CONFIG_KGDB
handle_kgdb_breakpoint:
call kgdb_breakpoint_c
br ret_from_exception
#endif
handle_trap_1:
call handle_trap_1_c
br ret_from_exception
handle_trap_2:
call handle_trap_2_c
br ret_from_exception
handle_trap_3:
handle_trap_reserved:
call handle_trap_3_c
br ret_from_exception
/*
* Beware - when entering resume, prev (the current task) is
* in r4, next (the new task) is in r5, don't change these
* registers.
*/
ENTRY(resume)
rdctl r7, status /* save thread status reg */
stw r7, TASK_THREAD + THREAD_KPSR(r4)
andi r7, r7, %lo(~STATUS_PIE) /* disable interrupts */
wrctl status, r7
SAVE_SWITCH_STACK
stw sp, TASK_THREAD + THREAD_KSP(r4)/* save kernel stack pointer */
ldw sp, TASK_THREAD + THREAD_KSP(r5)/* restore new thread stack */
movia r24, _current_thread /* save thread */
GET_THREAD_INFO r1
stw r1, 0(r24)
RESTORE_SWITCH_STACK
ldw r7, TASK_THREAD + THREAD_KPSR(r5)/* restore thread status reg */
wrctl status, r7
ret
ENTRY(ret_from_fork)
call schedule_tail
br ret_from_exception
ENTRY(ret_from_kernel_thread)
call schedule_tail
mov r4,r17 /* arg */
callr r16 /* function */
br ret_from_exception
/*
* Kernel user helpers.
*
* Each segment is 64-byte aligned and will be mapped to the <User space>.
* New segments (if ever needed) must be added after the existing ones.
* This mechanism should be used only for things that are really small and
* justified, and not be abused freely.
*
*/
/* Filling pads with undefined instructions. */
.macro kuser_pad sym size
.if ((. - \sym) & 3)
.rept (4 - (. - \sym) & 3)
.byte 0
.endr
.endif
.rept ((\size - (. - \sym)) / 4)
.word 0xdeadbeef
.endr
.endm
.align 6
.globl __kuser_helper_start
__kuser_helper_start:
__kuser_helper_version: /* @ 0x1000 */
.word ((__kuser_helper_end - __kuser_helper_start) >> 6)
__kuser_cmpxchg: /* @ 0x1004 */
/*
* r4 pointer to exchange variable
* r5 old value
* r6 new value
*/
cmpxchg_ldw:
ldw r2, 0(r4) /* load current value */
sub r2, r2, r5 /* compare with old value */
bne r2, zero, cmpxchg_ret
/* We had a match, store the new value */
cmpxchg_stw:
stw r6, 0(r4)
cmpxchg_ret:
ret
kuser_pad __kuser_cmpxchg, 64
.globl __kuser_sigtramp
__kuser_sigtramp:
movi r2, __NR_rt_sigreturn
trap
kuser_pad __kuser_sigtramp, 64
.globl __kuser_helper_end
__kuser_helper_end: