korg/linux
0
0
Fork 0
mirror of https://mirrors.bfsu.edu.cn/git/linux.git synced 2024-11-19 18:24:14 +08:00
Code Issues Packages Projects Releases Wiki Activity
b03ebd992f
linux/arch/mn10300/kernel/kgdb.c
Linus Torvalds b235beea9e Clarify naming of thread info/stack allocators 
We've had the thread info allocated together with the thread stack for
most architectures for a long time (since the thread_info was split off
from the task struct), but that is about to change.

But the patches that move the thread info to be off-stack (and a part of
the task struct instead) made it clear how confused the allocator and
freeing functions are.

Because the common case was that we share an allocation with the thread
stack and the thread_info, the two pointers were identical.  That
identity then meant that we would have things like

	ti = alloc_thread_info_node(tsk, node);
	...
	tsk->stack = ti;

which certainly _worked_ (since stack and thread_info have the same
value), but is rather confusing: why are we assigning a thread_info to
the stack? And if we move the thread_info away, the "confusing" code
just gets to be entirely bogus.

So remove all this confusion, and make it clear that we are doing the
stack allocation by renaming and clarifying the function names to be
about the stack.  The fact that the thread_info then shares the
allocation is an implementation detail, and not really about the
allocation itself.

This is a pure renaming and type fix: we pass in the same pointer, it's
just that we clarify what the pointer means.

The ia64 code that actually only has one single allocation (for all of
task_struct, thread_info and kernel thread stack) now looks a bit odd,
but since "tsk->stack" is actually not even used there, that oddity
doesn't matter.  It would be a separate thing to clean that up, I
intentionally left the ia64 changes as a pure brute-force renaming and
type change.

Acked-by: Andy Lutomirski <luto@amacapital.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-06-24 15:09:37 -07:00

503 lines
12 KiB
C
Raw Blame History

/* kgdb support for MN10300
*
* Copyright (C) 2010 Red Hat, Inc. All Rights Reserved.
* Written by David Howells (dhowells@redhat.com)
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public Licence
* as published by the Free Software Foundation; either version
* 2 of the Licence, or (at your option) any later version.
*/
#include <linux/slab.h>
#include <linux/ptrace.h>
#include <linux/kgdb.h>
#include <linux/uaccess.h>
#include <unit/leds.h>
#include <unit/serial.h>
#include <asm/debugger.h>
#include <asm/serial-regs.h>
#include "internal.h"
/*
* Software single-stepping breakpoint save (used by __switch_to())
*/
static struct thread_info *kgdb_sstep_thread;
u8 *kgdb_sstep_bp_addr[2];
u8 kgdb_sstep_bp[2];
/*
* Copy kernel exception frame registers to the GDB register file
*/
void pt_regs_to_gdb_regs(unsigned long *gdb_regs, struct pt_regs *regs)
{
unsigned long ssp = (unsigned long) (regs + 1);
gdb_regs[GDB_FR_D0] = regs->d0;
gdb_regs[GDB_FR_D1] = regs->d1;
gdb_regs[GDB_FR_D2] = regs->d2;
gdb_regs[GDB_FR_D3] = regs->d3;
gdb_regs[GDB_FR_A0] = regs->a0;
gdb_regs[GDB_FR_A1] = regs->a1;
gdb_regs[GDB_FR_A2] = regs->a2;
gdb_regs[GDB_FR_A3] = regs->a3;
gdb_regs[GDB_FR_SP] = (regs->epsw & EPSW_nSL) ? regs->sp : ssp;
gdb_regs[GDB_FR_PC] = regs->pc;
gdb_regs[GDB_FR_MDR] = regs->mdr;
gdb_regs[GDB_FR_EPSW] = regs->epsw;
gdb_regs[GDB_FR_LIR] = regs->lir;
gdb_regs[GDB_FR_LAR] = regs->lar;
gdb_regs[GDB_FR_MDRQ] = regs->mdrq;
gdb_regs[GDB_FR_E0] = regs->e0;
gdb_regs[GDB_FR_E1] = regs->e1;
gdb_regs[GDB_FR_E2] = regs->e2;
gdb_regs[GDB_FR_E3] = regs->e3;
gdb_regs[GDB_FR_E4] = regs->e4;
gdb_regs[GDB_FR_E5] = regs->e5;
gdb_regs[GDB_FR_E6] = regs->e6;
gdb_regs[GDB_FR_E7] = regs->e7;
gdb_regs[GDB_FR_SSP] = ssp;
gdb_regs[GDB_FR_MSP] = 0;
gdb_regs[GDB_FR_USP] = regs->sp;
gdb_regs[GDB_FR_MCRH] = regs->mcrh;
gdb_regs[GDB_FR_MCRL] = regs->mcrl;
gdb_regs[GDB_FR_MCVF] = regs->mcvf;
gdb_regs[GDB_FR_DUMMY0] = 0;
gdb_regs[GDB_FR_DUMMY1] = 0;
gdb_regs[GDB_FR_FS0] = 0;
}
/*
* Extracts kernel SP/PC values understandable by gdb from the values
* saved by switch_to().
*/
void sleeping_thread_to_gdb_regs(unsigned long *gdb_regs, struct task_struct *p)
{
gdb_regs[GDB_FR_SSP] = p->thread.sp;
gdb_regs[GDB_FR_PC] = p->thread.pc;
gdb_regs[GDB_FR_A3] = p->thread.a3;
gdb_regs[GDB_FR_USP] = p->thread.usp;
gdb_regs[GDB_FR_FPCR] = p->thread.fpu_state.fpcr;
}
/*
* Fill kernel exception frame registers from the GDB register file
*/
void gdb_regs_to_pt_regs(unsigned long *gdb_regs, struct pt_regs *regs)
{
regs->d0 = gdb_regs[GDB_FR_D0];
regs->d1 = gdb_regs[GDB_FR_D1];
regs->d2 = gdb_regs[GDB_FR_D2];
regs->d3 = gdb_regs[GDB_FR_D3];
regs->a0 = gdb_regs[GDB_FR_A0];
regs->a1 = gdb_regs[GDB_FR_A1];
regs->a2 = gdb_regs[GDB_FR_A2];
regs->a3 = gdb_regs[GDB_FR_A3];
regs->sp = gdb_regs[GDB_FR_SP];
regs->pc = gdb_regs[GDB_FR_PC];
regs->mdr = gdb_regs[GDB_FR_MDR];
regs->epsw = gdb_regs[GDB_FR_EPSW];
regs->lir = gdb_regs[GDB_FR_LIR];
regs->lar = gdb_regs[GDB_FR_LAR];
regs->mdrq = gdb_regs[GDB_FR_MDRQ];
regs->e0 = gdb_regs[GDB_FR_E0];
regs->e1 = gdb_regs[GDB_FR_E1];
regs->e2 = gdb_regs[GDB_FR_E2];
regs->e3 = gdb_regs[GDB_FR_E3];
regs->e4 = gdb_regs[GDB_FR_E4];
regs->e5 = gdb_regs[GDB_FR_E5];
regs->e6 = gdb_regs[GDB_FR_E6];
regs->e7 = gdb_regs[GDB_FR_E7];
regs->sp = gdb_regs[GDB_FR_SSP];
/* gdb_regs[GDB_FR_MSP]; */
// regs->usp = gdb_regs[GDB_FR_USP];
regs->mcrh = gdb_regs[GDB_FR_MCRH];
regs->mcrl = gdb_regs[GDB_FR_MCRL];
regs->mcvf = gdb_regs[GDB_FR_MCVF];
/* gdb_regs[GDB_FR_DUMMY0]; */
/* gdb_regs[GDB_FR_DUMMY1]; */
// regs->fpcr = gdb_regs[GDB_FR_FPCR];
// regs->fs0 = gdb_regs[GDB_FR_FS0];
}
struct kgdb_arch arch_kgdb_ops = {
.gdb_bpt_instr = { 0xff },
.flags = KGDB_HW_BREAKPOINT,
};
static const unsigned char mn10300_kgdb_insn_sizes[256] =
{
/* 1 2 3 4 5 6 7 8 9 a b c d e f */
1, 3, 3, 3, 1, 3, 3, 3, 1, 3, 3, 3, 1, 3, 3, 3, /* 0 */
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* 1 */
2, 2, 2, 2, 3, 3, 3, 3, 2, 2, 2, 2, 3, 3, 3, 3, /* 2 */
3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 1, 1, 1, 1, /* 3 */
1, 1, 2, 2, 1, 1, 2, 2, 1, 1, 2, 2, 1, 1, 2, 2, /* 4 */
1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, /* 5 */
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* 6 */
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* 7 */
2, 1, 1, 1, 1, 2, 1, 1, 1, 1, 2, 1, 1, 1, 1, 2, /* 8 */
2, 1, 1, 1, 1, 2, 1, 1, 1, 1, 2, 1, 1, 1, 1, 2, /* 9 */
2, 1, 1, 1, 1, 2, 1, 1, 1, 1, 2, 1, 1, 1, 1, 2, /* a */
2, 1, 1, 1, 1, 2, 1, 1, 1, 1, 2, 1, 1, 1, 1, 2, /* b */
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 2, 2, /* c */
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, /* d */
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* e */
0, 2, 2, 2, 2, 2, 2, 4, 0, 3, 0, 4, 0, 6, 7, 1 /* f */
};
/*
* Attempt to emulate single stepping by means of breakpoint instructions.
* Although there is a single-step trace flag in EPSW, its use is not
* sufficiently documented and is only intended for use with the JTAG debugger.
*/
static int kgdb_arch_do_singlestep(struct pt_regs *regs)
{
unsigned long arg;
unsigned size;
u8 *pc = (u8 *)regs->pc, *sp = (u8 *)(regs + 1), cur;
u8 *x = NULL, *y = NULL;
int ret;
ret = probe_kernel_read(&cur, pc, 1);
if (ret < 0)
return ret;
size = mn10300_kgdb_insn_sizes[cur];
if (size > 0) {
x = pc + size;
goto set_x;
}
switch (cur) {
/* Bxx (d8,PC) */
case 0xc0 ... 0xca:
ret = probe_kernel_read(&arg, pc + 1, 1);
if (ret < 0)
return ret;
x = pc + 2;
if (arg >= 0 && arg <= 2)
goto set_x;
y = pc + (s8)arg;
goto set_x_and_y;
/* LXX (d8,PC) */
case 0xd0 ... 0xda:
x = pc + 1;
if (regs->pc == regs->lar)
goto set_x;
y = (u8 *)regs->lar;
goto set_x_and_y;
/* SETLB - loads the next four bytes into the LIR register
* (which mustn't include a breakpoint instruction) */
case 0xdb:
x = pc + 5;
goto set_x;
/* JMP (d16,PC) or CALL (d16,PC) */
case 0xcc:
case 0xcd:
ret = probe_kernel_read(&arg, pc + 1, 2);
if (ret < 0)
return ret;
x = pc + (s16)arg;
goto set_x;
/* JMP (d32,PC) or CALL (d32,PC) */
case 0xdc:
case 0xdd:
ret = probe_kernel_read(&arg, pc + 1, 4);
if (ret < 0)
return ret;
x = pc + (s32)arg;
goto set_x;
/* RETF */
case 0xde:
x = (u8 *)regs->mdr;
goto set_x;
/* RET */
case 0xdf:
ret = probe_kernel_read(&arg, pc + 2, 1);
if (ret < 0)
return ret;
ret = probe_kernel_read(&x, sp + (s8)arg, 4);
if (ret < 0)
return ret;
goto set_x;
case 0xf0:
ret = probe_kernel_read(&cur, pc + 1, 1);
if (ret < 0)
return ret;
if (cur >= 0xf0 && cur <= 0xf7) {
/* JMP (An) / CALLS (An) */
switch (cur & 3) {
case 0: x = (u8 *)regs->a0; break;
case 1: x = (u8 *)regs->a1; break;
case 2: x = (u8 *)regs->a2; break;
case 3: x = (u8 *)regs->a3; break;
}
goto set_x;
} else if (cur == 0xfc) {
/* RETS */
ret = probe_kernel_read(&x, sp, 4);
if (ret < 0)
return ret;
goto set_x;
} else if (cur == 0xfd) {
/* RTI */
ret = probe_kernel_read(&x, sp + 4, 4);
if (ret < 0)
return ret;
goto set_x;
} else {
x = pc + 2;
goto set_x;
}
break;
/* potential 3-byte conditional branches */
case 0xf8:
ret = probe_kernel_read(&cur, pc + 1, 1);
if (ret < 0)
return ret;
x = pc + 3;
if (cur >= 0xe8 && cur <= 0xeb) {
ret = probe_kernel_read(&arg, pc + 2, 1);
if (ret < 0)
return ret;
if (arg >= 0 && arg <= 3)
goto set_x;
y = pc + (s8)arg;
goto set_x_and_y;
}
goto set_x;
case 0xfa:
ret = probe_kernel_read(&cur, pc + 1, 1);
if (ret < 0)
return ret;
if (cur == 0xff) {
/* CALLS (d16,PC) */
ret = probe_kernel_read(&arg, pc + 2, 2);
if (ret < 0)
return ret;
x = pc + (s16)arg;
goto set_x;
}
x = pc + 4;
goto set_x;
case 0xfc:
ret = probe_kernel_read(&cur, pc + 1, 1);
if (ret < 0)
return ret;
if (cur == 0xff) {
/* CALLS (d32,PC) */
ret = probe_kernel_read(&arg, pc + 2, 4);
if (ret < 0)
return ret;
x = pc + (s32)arg;
goto set_x;
}
x = pc + 6;
goto set_x;
}
return 0;
set_x:
kgdb_sstep_bp_addr[0] = x;
kgdb_sstep_bp_addr[1] = NULL;
ret = probe_kernel_read(&kgdb_sstep_bp[0], x, 1);
if (ret < 0)
return ret;
ret = probe_kernel_write(x, &arch_kgdb_ops.gdb_bpt_instr, 1);
if (ret < 0)
return ret;
kgdb_sstep_thread = current_thread_info();
debugger_local_cache_flushinv_one(x);
return ret;
set_x_and_y:
kgdb_sstep_bp_addr[0] = x;
kgdb_sstep_bp_addr[1] = y;
ret = probe_kernel_read(&kgdb_sstep_bp[0], x, 1);
if (ret < 0)
return ret;
ret = probe_kernel_read(&kgdb_sstep_bp[1], y, 1);
if (ret < 0)
return ret;
ret = probe_kernel_write(x, &arch_kgdb_ops.gdb_bpt_instr, 1);
if (ret < 0)
return ret;
ret = probe_kernel_write(y, &arch_kgdb_ops.gdb_bpt_instr, 1);
if (ret < 0) {
probe_kernel_write(kgdb_sstep_bp_addr[0],
&kgdb_sstep_bp[0], 1);
} else {
kgdb_sstep_thread = current_thread_info();
}
debugger_local_cache_flushinv_one(x);
debugger_local_cache_flushinv_one(y);
return ret;
}
/*
* Remove emplaced single-step breakpoints, returning true if we hit one of
* them.
*/
static bool kgdb_arch_undo_singlestep(struct pt_regs *regs)
{
bool hit = false;
u8 *x = kgdb_sstep_bp_addr[0], *y = kgdb_sstep_bp_addr[1];
u8 opcode;
if (kgdb_sstep_thread == current_thread_info()) {
if (x) {
if (x == (u8 *)regs->pc)
hit = true;
if (probe_kernel_read(&opcode, x,
1) < 0 ||
opcode != 0xff)
BUG();
probe_kernel_write(x, &kgdb_sstep_bp[0], 1);
debugger_local_cache_flushinv_one(x);
}
if (y) {
if (y == (u8 *)regs->pc)
hit = true;
if (probe_kernel_read(&opcode, y,
1) < 0 ||
opcode != 0xff)
BUG();
probe_kernel_write(y, &kgdb_sstep_bp[1], 1);
debugger_local_cache_flushinv_one(y);
}
}
kgdb_sstep_bp_addr[0] = NULL;
kgdb_sstep_bp_addr[1] = NULL;
kgdb_sstep_thread = NULL;
return hit;
}
/*
* Catch a single-step-pending thread being deleted and make sure the global
* single-step state is cleared. At this point the breakpoints should have
* been removed by __switch_to().
*/
void arch_release_thread_stack(unsigned long *stack)
{
struct thread_info *ti = (void *)stack;
if (kgdb_sstep_thread == ti) {
kgdb_sstep_thread = NULL;
/* However, we may now be running in degraded mode, with most
* of the CPUs disabled until such a time as KGDB is reentered,
* so force immediate reentry */
kgdb_breakpoint();
}
}
/*
* Handle unknown packets and [CcsDk] packets
* - at this point breakpoints have been installed
*/
int kgdb_arch_handle_exception(int vector, int signo, int err_code,
char *remcom_in_buffer, char *remcom_out_buffer,
struct pt_regs *regs)
{
long addr;
char *ptr;
switch (remcom_in_buffer[0]) {
case 'c':
case 's':
/* try to read optional parameter, pc unchanged if no parm */
ptr = &remcom_in_buffer[1];
if (kgdb_hex2long(&ptr, &addr))
regs->pc = addr;
case 'D':
case 'k':
atomic_set(&kgdb_cpu_doing_single_step, -1);
if (remcom_in_buffer[0] == 's') {
kgdb_arch_do_singlestep(regs);
kgdb_single_step = 1;
atomic_set(&kgdb_cpu_doing_single_step,
raw_smp_processor_id());
}
return 0;
}
return -1; /* this means that we do not want to exit from the handler */
}
/*
* Handle event interception
* - returns 0 if the exception should be skipped, -ERROR otherwise.
*/
int debugger_intercept(enum exception_code excep, int signo, int si_code,
struct pt_regs *regs)
{
int ret;
if (kgdb_arch_undo_singlestep(regs)) {
excep = EXCEP_TRAP;
signo = SIGTRAP;
si_code = TRAP_TRACE;
}
ret = kgdb_handle_exception(excep, signo, si_code, regs);
debugger_local_cache_flushinv();
return ret;
}
/*
* Determine if we've hit a debugger special breakpoint
*/
int at_debugger_breakpoint(struct pt_regs *regs)
{
return regs->pc == (unsigned long)&__arch_kgdb_breakpoint;
}
/*
* Initialise kgdb
*/
int kgdb_arch_init(void)
{
return 0;
}
/*
* Do something, perhaps, but don't know what.
*/
void kgdb_arch_exit(void)
{
}
#ifdef CONFIG_SMP
void debugger_nmi_interrupt(struct pt_regs *regs, enum exception_code code)
{
kgdb_nmicallback(arch_smp_processor_id(), regs);
debugger_local_cache_flushinv();
}
void kgdb_roundup_cpus(unsigned long flags)
{
smp_jump_to_debugger();
}
#endif
Reference in New Issue View Git Blame Copy Permalink