869 lines
26 KiB
C
869 lines
26 KiB
C
/* Target-dependent code for Hitachi Super-H, for GDB.
|
|
Copyright (C) 1993, 1994, 1995, 1996 Free Software Foundation, Inc.
|
|
|
|
This file is part of GDB.
|
|
|
|
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 program is distributed in the hope that it will be useful,
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
GNU General Public License for more details.
|
|
|
|
You should have received a copy of the GNU General Public License
|
|
along with this program; if not, write to the Free Software
|
|
Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */
|
|
|
|
/*
|
|
Contributed by Steve Chamberlain
|
|
sac@cygnus.com
|
|
*/
|
|
|
|
#include "defs.h"
|
|
#include "frame.h"
|
|
#include "obstack.h"
|
|
#include "symtab.h"
|
|
#include "symfile.h"
|
|
#include "gdbtypes.h"
|
|
#include "gdbcmd.h"
|
|
#include "gdbcore.h"
|
|
#include "value.h"
|
|
#include "dis-asm.h"
|
|
#include "inferior.h" /* for BEFORE_TEXT_END etc. */
|
|
#include "gdb_string.h"
|
|
|
|
extern int remote_write_size; /* in remote.c */
|
|
|
|
/* A set of original names, to be used when restoring back to generic
|
|
registers from a specific set. */
|
|
|
|
char *sh_generic_reg_names[] = REGISTER_NAMES;
|
|
|
|
char *sh_reg_names[] = {
|
|
"r0", "r1", "r2", "r3", "r4", "r5", "r6", "r7",
|
|
"r8", "r9", "r10", "r11", "r12", "r13", "r14", "r15",
|
|
"pc", "pr", "gbr", "vbr", "mach", "macl", "sr",
|
|
"", "",
|
|
"", "", "", "", "", "", "", "",
|
|
"", "", "", "", "", "", "", "",
|
|
"", "",
|
|
"", "", "", "", "", "", "", "",
|
|
"", "", "", "", "", "", "", "",
|
|
};
|
|
|
|
char *sh3_reg_names[] = {
|
|
"r0", "r1", "r2", "r3", "r4", "r5", "r6", "r7",
|
|
"r8", "r9", "r10", "r11", "r12", "r13", "r14", "r15",
|
|
"pc", "pr", "gbr", "vbr", "mach", "macl", "sr",
|
|
"", "",
|
|
"", "", "", "", "", "", "", "",
|
|
"", "", "", "", "", "", "", "",
|
|
"ssr", "spc",
|
|
"r0b0", "r1b0", "r2b0", "r3b0", "r4b0", "r5b0", "r6b0", "r7b0",
|
|
"r0b1", "r1b1", "r2b1", "r3b1", "r4b1", "r5b1", "r6b1", "r7b1"
|
|
};
|
|
|
|
char *sh3e_reg_names[] = {
|
|
"r0", "r1", "r2", "r3", "r4", "r5", "r6", "r7",
|
|
"r8", "r9", "r10", "r11", "r12", "r13", "r14", "r15",
|
|
"pc", "pr", "gbr", "vbr", "mach", "macl", "sr",
|
|
"fpul", "fpscr",
|
|
"fr0", "fr1", "fr2", "fr3", "fr4", "fr5", "fr6", "fr7",
|
|
"fr8", "fr9", "fr10", "fr11", "fr12", "fr13", "fr14", "fr15",
|
|
"ssr", "spc",
|
|
"r0b0", "r1b0", "r2b0", "r3b0", "r4b0", "r5b0", "r6b0", "r7b0",
|
|
"r0b1", "r1b1", "r2b1", "r3b1", "r4b1", "r5b1", "r6b1", "r7b1",
|
|
};
|
|
|
|
struct {
|
|
char **regnames;
|
|
int mach;
|
|
} sh_processor_type_table[] = {
|
|
{ sh_reg_names, bfd_mach_sh },
|
|
{ sh3_reg_names, bfd_mach_sh3 },
|
|
{ sh3e_reg_names, bfd_mach_sh3e },
|
|
{ NULL, 0 }
|
|
};
|
|
|
|
/* Prologue looks like
|
|
[mov.l <regs>,@-r15]...
|
|
[sts.l pr,@-r15]
|
|
[mov.l r14,@-r15]
|
|
[mov r15,r14]
|
|
*/
|
|
|
|
#define IS_STS(x) ((x) == 0x4f22)
|
|
#define IS_PUSH(x) (((x) & 0xff0f) == 0x2f06)
|
|
#define GET_PUSHED_REG(x) (((x) >> 4) & 0xf)
|
|
#define IS_MOV_SP_FP(x) ((x) == 0x6ef3)
|
|
#define IS_ADD_SP(x) (((x) & 0xff00) == 0x7f00)
|
|
#define IS_MOV_R3(x) (((x) & 0xff00) == 0x1a00)
|
|
#define IS_SHLL_R3(x) ((x) == 0x4300)
|
|
#define IS_ADD_R3SP(x) ((x) == 0x3f3c)
|
|
|
|
/* Skip any prologue before the guts of a function */
|
|
|
|
CORE_ADDR
|
|
sh_skip_prologue (start_pc)
|
|
CORE_ADDR start_pc;
|
|
{
|
|
int w;
|
|
|
|
w = read_memory_integer (start_pc, 2);
|
|
while (IS_STS (w)
|
|
|| IS_PUSH (w)
|
|
|| IS_MOV_SP_FP (w)
|
|
|| IS_MOV_R3 (w)
|
|
|| IS_ADD_R3SP (w)
|
|
|| IS_ADD_SP (w)
|
|
|| IS_SHLL_R3 (w))
|
|
{
|
|
start_pc += 2;
|
|
w = read_memory_integer (start_pc, 2);
|
|
}
|
|
|
|
return start_pc;
|
|
}
|
|
|
|
/* Disassemble an instruction. */
|
|
|
|
int
|
|
gdb_print_insn_sh (memaddr, info)
|
|
bfd_vma memaddr;
|
|
disassemble_info *info;
|
|
{
|
|
if (TARGET_BYTE_ORDER == BIG_ENDIAN)
|
|
return print_insn_sh (memaddr, info);
|
|
else
|
|
return print_insn_shl (memaddr, info);
|
|
}
|
|
|
|
/* Given a GDB frame, determine the address of the calling function's frame.
|
|
This will be used to create a new GDB frame struct, and then
|
|
INIT_EXTRA_FRAME_INFO and INIT_FRAME_PC will be called for the new frame.
|
|
|
|
For us, the frame address is its stack pointer value, so we look up
|
|
the function prologue to determine the caller's sp value, and return it. */
|
|
|
|
CORE_ADDR
|
|
sh_frame_chain (frame)
|
|
struct frame_info *frame;
|
|
{
|
|
if (PC_IN_CALL_DUMMY (frame->pc, frame->frame, frame->frame))
|
|
return frame->frame; /* dummy frame same as caller's frame */
|
|
if (!inside_entry_file (frame->pc))
|
|
return read_memory_integer (FRAME_FP (frame) + frame->f_offset, 4);
|
|
else
|
|
return 0;
|
|
}
|
|
|
|
/* Find REGNUM on the stack. Otherwise, it's in an active register. One thing
|
|
we might want to do here is to check REGNUM against the clobber mask, and
|
|
somehow flag it as invalid if it isn't saved on the stack somewhere. This
|
|
would provide a graceful failure mode when trying to get the value of
|
|
caller-saves registers for an inner frame. */
|
|
|
|
CORE_ADDR
|
|
sh_find_callers_reg (fi, regnum)
|
|
struct frame_info *fi;
|
|
int regnum;
|
|
{
|
|
struct frame_saved_regs fsr;
|
|
|
|
for (; fi; fi = fi->next)
|
|
if (PC_IN_CALL_DUMMY (fi->pc, fi->frame, fi->frame))
|
|
/* When the caller requests PR from the dummy frame, we return PC because
|
|
that's where the previous routine appears to have done a call from. */
|
|
return generic_read_register_dummy (fi->pc, fi->frame, regnum);
|
|
else
|
|
{
|
|
FRAME_FIND_SAVED_REGS(fi, fsr);
|
|
if (fsr.regs[regnum] != 0)
|
|
return read_memory_integer (fsr.regs[regnum],
|
|
REGISTER_RAW_SIZE(regnum));
|
|
}
|
|
return read_register (regnum);
|
|
}
|
|
|
|
/* Put here the code to store, into a struct frame_saved_regs, the
|
|
addresses of the saved registers of frame described by FRAME_INFO.
|
|
This includes special registers such as pc and fp saved in special
|
|
ways in the stack frame. sp is even more special: the address we
|
|
return for it IS the sp for the next frame. */
|
|
|
|
void
|
|
sh_frame_find_saved_regs (fi, fsr)
|
|
struct frame_info *fi;
|
|
struct frame_saved_regs *fsr;
|
|
{
|
|
int where[NUM_REGS];
|
|
int rn;
|
|
int have_fp = 0;
|
|
int depth;
|
|
int pc;
|
|
int opc;
|
|
int insn;
|
|
int r3_val = 0;
|
|
char * dummy_regs = generic_find_dummy_frame (fi->pc, fi->frame);
|
|
|
|
if (dummy_regs)
|
|
{
|
|
/* DANGER! This is ONLY going to work if the char buffer format of
|
|
the saved registers is byte-for-byte identical to the
|
|
CORE_ADDR regs[NUM_REGS] format used by struct frame_saved_regs! */
|
|
memcpy (&fsr->regs, dummy_regs, sizeof(fsr));
|
|
return;
|
|
}
|
|
|
|
opc = pc = get_pc_function_start (fi->pc);
|
|
|
|
insn = read_memory_integer (pc, 2);
|
|
|
|
fi->leaf_function = 1;
|
|
fi->f_offset = 0;
|
|
|
|
for (rn = 0; rn < NUM_REGS; rn++)
|
|
where[rn] = -1;
|
|
|
|
depth = 0;
|
|
|
|
/* Loop around examining the prologue insns until we find something
|
|
that does not appear to be part of the prologue. But give up
|
|
after 20 of them, since we're getting silly then. */
|
|
|
|
while (pc < opc + 20 * 2)
|
|
{
|
|
/* See where the registers will be saved to */
|
|
if (IS_PUSH (insn))
|
|
{
|
|
pc += 2;
|
|
rn = GET_PUSHED_REG (insn);
|
|
where[rn] = depth;
|
|
insn = read_memory_integer (pc, 2);
|
|
depth += 4;
|
|
}
|
|
else if (IS_STS (insn))
|
|
{
|
|
pc += 2;
|
|
where[PR_REGNUM] = depth;
|
|
insn = read_memory_integer (pc, 2);
|
|
/* If we're storing the pr then this isn't a leaf */
|
|
fi->leaf_function = 0;
|
|
depth += 4;
|
|
}
|
|
else if (IS_MOV_R3 (insn))
|
|
{
|
|
r3_val = ((insn & 0xff) ^ 0x80) - 0x80;
|
|
pc += 2;
|
|
insn = read_memory_integer (pc, 2);
|
|
}
|
|
else if (IS_SHLL_R3 (insn))
|
|
{
|
|
r3_val <<= 1;
|
|
pc += 2;
|
|
insn = read_memory_integer (pc, 2);
|
|
}
|
|
else if (IS_ADD_R3SP (insn))
|
|
{
|
|
depth += -r3_val;
|
|
pc += 2;
|
|
insn = read_memory_integer (pc, 2);
|
|
}
|
|
else if (IS_ADD_SP (insn))
|
|
{
|
|
pc += 2;
|
|
depth -= ((insn & 0xff) ^ 0x80) - 0x80;
|
|
insn = read_memory_integer (pc, 2);
|
|
}
|
|
else
|
|
break;
|
|
}
|
|
|
|
/* Now we know how deep things are, we can work out their addresses */
|
|
|
|
for (rn = 0; rn < NUM_REGS; rn++)
|
|
{
|
|
if (where[rn] >= 0)
|
|
{
|
|
if (rn == FP_REGNUM)
|
|
have_fp = 1;
|
|
|
|
fsr->regs[rn] = fi->frame - where[rn] + depth - 4;
|
|
}
|
|
else
|
|
{
|
|
fsr->regs[rn] = 0;
|
|
}
|
|
}
|
|
|
|
if (have_fp)
|
|
{
|
|
fsr->regs[SP_REGNUM] = read_memory_integer (fsr->regs[FP_REGNUM], 4);
|
|
}
|
|
else
|
|
{
|
|
fsr->regs[SP_REGNUM] = fi->frame - 4;
|
|
}
|
|
|
|
fi->f_offset = depth - where[FP_REGNUM] - 4;
|
|
/* Work out the return pc - either from the saved pr or the pr
|
|
value */
|
|
}
|
|
|
|
/* initialize the extra info saved in a FRAME */
|
|
|
|
void
|
|
sh_init_extra_frame_info (fromleaf, fi)
|
|
int fromleaf;
|
|
struct frame_info *fi;
|
|
{
|
|
struct frame_saved_regs fsr;
|
|
|
|
if (fi->next)
|
|
fi->pc = FRAME_SAVED_PC (fi->next);
|
|
|
|
if (PC_IN_CALL_DUMMY (fi->pc, fi->frame, fi->frame))
|
|
{
|
|
/* We need to setup fi->frame here because run_stack_dummy gets it wrong
|
|
by assuming it's always FP. */
|
|
fi->frame = generic_read_register_dummy (fi->pc, fi->frame,
|
|
SP_REGNUM);
|
|
fi->return_pc = generic_read_register_dummy (fi->pc, fi->frame,
|
|
PC_REGNUM);
|
|
fi->f_offset = -(CALL_DUMMY_LENGTH + 4);
|
|
fi->leaf_function = 0;
|
|
return;
|
|
}
|
|
else
|
|
{
|
|
FRAME_FIND_SAVED_REGS (fi, fsr);
|
|
fi->return_pc = sh_find_callers_reg (fi, PR_REGNUM);
|
|
}
|
|
}
|
|
|
|
/* Discard from the stack the innermost frame,
|
|
restoring all saved registers. */
|
|
|
|
void
|
|
sh_pop_frame ()
|
|
{
|
|
register struct frame_info *frame = get_current_frame ();
|
|
register CORE_ADDR fp;
|
|
register int regnum;
|
|
struct frame_saved_regs fsr;
|
|
|
|
if (PC_IN_CALL_DUMMY (frame->pc, frame->frame, frame->frame))
|
|
generic_pop_dummy_frame ();
|
|
else
|
|
{
|
|
fp = FRAME_FP (frame);
|
|
get_frame_saved_regs (frame, &fsr);
|
|
|
|
/* Copy regs from where they were saved in the frame */
|
|
for (regnum = 0; regnum < NUM_REGS; regnum++)
|
|
if (fsr.regs[regnum])
|
|
write_register (regnum, read_memory_integer (fsr.regs[regnum], 4));
|
|
|
|
write_register (PC_REGNUM, frame->return_pc);
|
|
write_register (SP_REGNUM, fp + 4);
|
|
}
|
|
flush_cached_frames ();
|
|
}
|
|
|
|
/* Function: push_arguments
|
|
Setup the function arguments for calling a function in the inferior.
|
|
|
|
On the Hitachi SH architecture, there are four registers (R4 to R7)
|
|
which are dedicated for passing function arguments. Up to the first
|
|
four arguments (depending on size) may go into these registers.
|
|
The rest go on the stack.
|
|
|
|
Arguments that are smaller than 4 bytes will still take up a whole
|
|
register or a whole 32-bit word on the stack, and will be
|
|
right-justified in the register or the stack word. This includes
|
|
chars, shorts, and small aggregate types.
|
|
|
|
Arguments that are larger than 4 bytes may be split between two or
|
|
more registers. If there are not enough registers free, an argument
|
|
may be passed partly in a register (or registers), and partly on the
|
|
stack. This includes doubles, long longs, and larger aggregates.
|
|
As far as I know, there is no upper limit to the size of aggregates
|
|
that will be passed in this way; in other words, the convention of
|
|
passing a pointer to a large aggregate instead of a copy is not used.
|
|
|
|
An exceptional case exists for struct arguments (and possibly other
|
|
aggregates such as arrays) if the size is larger than 4 bytes but
|
|
not a multiple of 4 bytes. In this case the argument is never split
|
|
between the registers and the stack, but instead is copied in its
|
|
entirety onto the stack, AND also copied into as many registers as
|
|
there is room for. In other words, space in registers permitting,
|
|
two copies of the same argument are passed in. As far as I can tell,
|
|
only the one on the stack is used, although that may be a function
|
|
of the level of compiler optimization. I suspect this is a compiler
|
|
bug. Arguments of these odd sizes are left-justified within the
|
|
word (as opposed to arguments smaller than 4 bytes, which are
|
|
right-justified).
|
|
|
|
|
|
If the function is to return an aggregate type such as a struct, it
|
|
is either returned in the normal return value register R0 (if its
|
|
size is no greater than one byte), or else the caller must allocate
|
|
space into which the callee will copy the return value (if the size
|
|
is greater than one byte). In this case, a pointer to the return
|
|
value location is passed into the callee in register R2, which does
|
|
not displace any of the other arguments passed in via registers R4
|
|
to R7. */
|
|
|
|
CORE_ADDR
|
|
sh_push_arguments (nargs, args, sp, struct_return, struct_addr)
|
|
int nargs;
|
|
value_ptr *args;
|
|
CORE_ADDR sp;
|
|
unsigned char struct_return;
|
|
CORE_ADDR struct_addr;
|
|
{
|
|
int stack_offset, stack_alloc;
|
|
int argreg;
|
|
int argnum;
|
|
struct type *type;
|
|
CORE_ADDR regval;
|
|
char *val;
|
|
char valbuf[4];
|
|
int len;
|
|
int odd_sized_struct;
|
|
|
|
/* first force sp to a 4-byte alignment */
|
|
sp = sp & ~3;
|
|
|
|
/* The "struct return pointer" pseudo-argument has its own dedicated
|
|
register */
|
|
if (struct_return)
|
|
write_register (STRUCT_RETURN_REGNUM, struct_addr);
|
|
|
|
/* Now make sure there's space on the stack */
|
|
for (argnum = 0, stack_alloc = 0;
|
|
argnum < nargs; argnum++)
|
|
stack_alloc += ((TYPE_LENGTH(VALUE_TYPE(args[argnum])) + 3) & ~3);
|
|
sp -= stack_alloc; /* make room on stack for args */
|
|
|
|
|
|
/* Now load as many as possible of the first arguments into
|
|
registers, and push the rest onto the stack. There are 16 bytes
|
|
in four registers available. Loop thru args from first to last. */
|
|
|
|
argreg = ARG0_REGNUM;
|
|
for (argnum = 0, stack_offset = 0; argnum < nargs; argnum++)
|
|
{
|
|
type = VALUE_TYPE (args[argnum]);
|
|
len = TYPE_LENGTH (type);
|
|
memset(valbuf, 0, sizeof(valbuf));
|
|
if (len < 4)
|
|
{ /* value gets right-justified in the register or stack word */
|
|
memcpy(valbuf + (4 - len),
|
|
(char *) VALUE_CONTENTS (args[argnum]), len);
|
|
val = valbuf;
|
|
}
|
|
else
|
|
val = (char *) VALUE_CONTENTS (args[argnum]);
|
|
|
|
if (len > 4 && (len & 3) != 0)
|
|
odd_sized_struct = 1; /* such structs go entirely on stack */
|
|
else
|
|
odd_sized_struct = 0;
|
|
while (len > 0)
|
|
{
|
|
if (argreg > ARGLAST_REGNUM || odd_sized_struct)
|
|
{ /* must go on the stack */
|
|
write_memory (sp + stack_offset, val, 4);
|
|
stack_offset += 4;
|
|
}
|
|
/* NOTE WELL!!!!! This is not an "else if" clause!!!
|
|
That's because some *&^%$ things get passed on the stack
|
|
AND in the registers! */
|
|
if (argreg <= ARGLAST_REGNUM)
|
|
{ /* there's room in a register */
|
|
regval = extract_address (val, REGISTER_RAW_SIZE(argreg));
|
|
write_register (argreg++, regval);
|
|
}
|
|
/* Store the value 4 bytes at a time. This means that things
|
|
larger than 4 bytes may go partly in registers and partly
|
|
on the stack. */
|
|
len -= REGISTER_RAW_SIZE(argreg);
|
|
val += REGISTER_RAW_SIZE(argreg);
|
|
}
|
|
}
|
|
return sp;
|
|
}
|
|
|
|
/* Function: push_return_address (pc)
|
|
Set up the return address for the inferior function call.
|
|
Needed for targets where we don't actually execute a JSR/BSR instruction */
|
|
|
|
CORE_ADDR
|
|
sh_push_return_address (pc, sp)
|
|
CORE_ADDR pc;
|
|
CORE_ADDR sp;
|
|
{
|
|
write_register (PR_REGNUM, CALL_DUMMY_ADDRESS ());
|
|
return sp;
|
|
}
|
|
|
|
/* Function: fix_call_dummy
|
|
Poke the callee function's address into the destination part of
|
|
the CALL_DUMMY. The address is actually stored in a data word
|
|
following the actualy CALL_DUMMY instructions, which will load
|
|
it into a register using PC-relative addressing. This function
|
|
expects the CALL_DUMMY to look like this:
|
|
|
|
mov.w @(2,PC), R8
|
|
jsr @R8
|
|
nop
|
|
trap
|
|
<destination>
|
|
*/
|
|
|
|
#if 0
|
|
void
|
|
sh_fix_call_dummy (dummy, pc, fun, nargs, args, type, gcc_p)
|
|
char *dummy;
|
|
CORE_ADDR pc;
|
|
CORE_ADDR fun;
|
|
int nargs;
|
|
value_ptr *args;
|
|
struct type *type;
|
|
int gcc_p;
|
|
{
|
|
*(unsigned long *) (dummy + 8) = fun;
|
|
}
|
|
#endif
|
|
|
|
/* Function: get_saved_register
|
|
Just call the generic_get_saved_register function. */
|
|
|
|
void
|
|
get_saved_register (raw_buffer, optimized, addrp, frame, regnum, lval)
|
|
char *raw_buffer;
|
|
int *optimized;
|
|
CORE_ADDR *addrp;
|
|
struct frame_info *frame;
|
|
int regnum;
|
|
enum lval_type *lval;
|
|
{
|
|
generic_get_saved_register (raw_buffer, optimized, addrp,
|
|
frame, regnum, lval);
|
|
}
|
|
|
|
|
|
/* Modify the actual processor type. */
|
|
|
|
int
|
|
sh_target_architecture_hook (ap)
|
|
const bfd_arch_info_type *ap;
|
|
{
|
|
int i, j;
|
|
|
|
if (ap->arch != bfd_arch_sh)
|
|
return 0;
|
|
|
|
for (i = 0; sh_processor_type_table[i].regnames != NULL; i++)
|
|
{
|
|
if (sh_processor_type_table[i].mach == ap->mach)
|
|
{
|
|
for (j = 0; j < NUM_REGS; ++j)
|
|
reg_names[j] = sh_processor_type_table[i].regnames[j];
|
|
return 1;
|
|
}
|
|
}
|
|
|
|
fatal ("Architecture `%s' unreconized", ap->printable_name);
|
|
}
|
|
|
|
/* Print the registers in a form similar to the E7000 */
|
|
|
|
static void
|
|
sh_show_regs (args, from_tty)
|
|
char *args;
|
|
int from_tty;
|
|
{
|
|
int cpu;
|
|
if (target_architecture->arch == bfd_arch_sh)
|
|
cpu = target_architecture->mach;
|
|
else
|
|
cpu = 0;
|
|
|
|
printf_filtered ("PC=%08x SR=%08x PR=%08x MACH=%08x MACHL=%08x\n",
|
|
read_register (PC_REGNUM),
|
|
read_register (SR_REGNUM),
|
|
read_register (PR_REGNUM),
|
|
read_register (MACH_REGNUM),
|
|
read_register (MACL_REGNUM));
|
|
|
|
printf_filtered ("GBR=%08x VBR=%08x",
|
|
read_register (GBR_REGNUM),
|
|
read_register (VBR_REGNUM));
|
|
if (cpu == bfd_mach_sh3 || cpu == bfd_mach_sh3e)
|
|
{
|
|
printf_filtered (" SSR=%08x SPC=%08x",
|
|
read_register (SSR_REGNUM),
|
|
read_register (SPC_REGNUM));
|
|
if (cpu == bfd_mach_sh3e)
|
|
{
|
|
printf_filtered (" FPUL=%08x FPSCR=%08x",
|
|
read_register (FPUL_REGNUM),
|
|
read_register (FPSCR_REGNUM));
|
|
}
|
|
}
|
|
|
|
printf_filtered ("\nR0-R7 %08x %08x %08x %08x %08x %08x %08x %08x\n",
|
|
read_register (0),
|
|
read_register (1),
|
|
read_register (2),
|
|
read_register (3),
|
|
read_register (4),
|
|
read_register (5),
|
|
read_register (6),
|
|
read_register (7));
|
|
printf_filtered ("R8-R15 %08x %08x %08x %08x %08x %08x %08x %08x\n",
|
|
read_register (8),
|
|
read_register (9),
|
|
read_register (10),
|
|
read_register (11),
|
|
read_register (12),
|
|
read_register (13),
|
|
read_register (14),
|
|
read_register (15));
|
|
if (cpu == bfd_mach_sh3e)
|
|
{
|
|
printf_filtered ("FP0-FP7 %08x %08x %08x %08x %08x %08x %08x %08x\n",
|
|
read_register (FP0_REGNUM + 0),
|
|
read_register (FP0_REGNUM + 1),
|
|
read_register (FP0_REGNUM + 2),
|
|
read_register (FP0_REGNUM + 3),
|
|
read_register (FP0_REGNUM + 4),
|
|
read_register (FP0_REGNUM + 5),
|
|
read_register (FP0_REGNUM + 6),
|
|
read_register (FP0_REGNUM + 7));
|
|
printf_filtered ("FP8-FP15 %08x %08x %08x %08x %08x %08x %08x %08x\n",
|
|
read_register (FP0_REGNUM + 8),
|
|
read_register (FP0_REGNUM + 9),
|
|
read_register (FP0_REGNUM + 10),
|
|
read_register (FP0_REGNUM + 11),
|
|
read_register (FP0_REGNUM + 12),
|
|
read_register (FP0_REGNUM + 13),
|
|
read_register (FP0_REGNUM + 14),
|
|
read_register (FP0_REGNUM + 15));
|
|
}
|
|
}
|
|
|
|
/* Function: extract_return_value
|
|
Find a function's return value in the appropriate registers (in regbuf),
|
|
and copy it into valbuf. */
|
|
|
|
void
|
|
sh_extract_return_value (type, regbuf, valbuf)
|
|
struct type *type;
|
|
void *regbuf;
|
|
void *valbuf;
|
|
{
|
|
int len = TYPE_LENGTH(type);
|
|
|
|
if (len <= 4)
|
|
memcpy (valbuf, ((char *) regbuf) + 4 - len, len);
|
|
else if (len <= 8)
|
|
memcpy (valbuf, ((char *) regbuf) + 8 - len, len);
|
|
else
|
|
error ("bad size for return value");
|
|
}
|
|
|
|
void
|
|
_initialize_sh_tdep ()
|
|
{
|
|
struct cmd_list_element *c;
|
|
|
|
tm_print_insn = gdb_print_insn_sh;
|
|
|
|
target_architecture_hook = sh_target_architecture_hook;
|
|
|
|
add_com ("regs", class_vars, sh_show_regs, "Print all registers");
|
|
}
|
|
|
|
#ifdef NO_SINGLE_STEP
|
|
/* Non-zero if we just simulated a single-step ptrace call. This is
|
|
needed because we cannot remove the breakpoints in the inferior
|
|
process until after the `wait' in `wait_for_inferior'. Used for
|
|
4.4bsd for mips, where the kernel does not emulate single-step. */
|
|
|
|
int one_stepped;
|
|
CORE_ADDR target_addr; /* Branch target offset, if we have a
|
|
breakpoint there... */
|
|
CORE_ADDR step_addr; /* Offset of instruction after instruction
|
|
to be stepped, if we have a breakpoint
|
|
there. */
|
|
long step_cache [3]; /* Cache for instructions wiped out by
|
|
step breakpoint(s)... */
|
|
|
|
/* single_step() is called just before we want to resume the inferior,
|
|
if we want to single-step it but there is no hardware or kernel single-step
|
|
support (as on NetBSD). We find all the possible targets of the
|
|
coming instruction and breakpoint them.
|
|
|
|
single_step is also called just after the inferior stops. If we had
|
|
set up a simulated single-step, we undo our damage. */
|
|
|
|
/* thoughts:
|
|
|
|
For the current instruction, check to see if we're in a delay slot.
|
|
If we are, the next instruction executed will either be the target of
|
|
the branch or jump instruction preceding the current instruction, or
|
|
it will be the instruction following the current instruction. If
|
|
we are not, then the next instruction executed will either be the
|
|
instruction following the current instruction, or the instruction
|
|
following that (if the current instruction is a branch likely instruction
|
|
and the branch is not taken).
|
|
|
|
So, if we are in a delay slot then we set a breakpoint for the target
|
|
of the preceding instruction. Unless the preceding instruction was
|
|
a jump instruction (only jumps are unconditional), we also set a break-
|
|
point at the instruction following the current one and the instruction
|
|
following that. Setting two breakpoints after the current instruction
|
|
is cheaper and easier than figuring out whether the current instruction
|
|
is a branch likely instruction. */
|
|
|
|
#define IS_JMP(x) (((x) & 0xf0ff) == 0x402b)
|
|
#define IS_JSR(x) (((x) & 0xf0ff) == 0x400b)
|
|
|
|
#define IS_RTS(x) ((x) == 0x000b)
|
|
|
|
#define IS_BRAF(x) (((x) & 0xf0ff) == 0x0023)
|
|
#define IS_BSRF(x) (((x) & 0xf0ff) == 0x0003)
|
|
|
|
#define IS_BSR(x) (((x) & 0xf000) == 0xb000)
|
|
#define IS_BRA(x) (((x) & 0xf000) == 0xa000)
|
|
#define IS_BTS(x) (((x) & 0xff00) == 0x8d00)
|
|
#define IS_BFS(x) (((x) & 0xff00) == 0x8f00)
|
|
|
|
#define IS_BT(x) (((x) & 0xff00) == 0x8900)
|
|
#define IS_BF(x) (((x) & 0xff00) == 0x8b00)
|
|
|
|
|
|
#define IS_PCLOADFROMREG(x) ((IS_JMP(x)) || (IS_JSR(x)))
|
|
#define IS_ADDPCBYREG(x) ((IS_BRAF(x)) || (IS_BSRF(x)))
|
|
#define IS_ADDPCBYIMM(x) ((IS_BSR(x)) || (IS_BRA(x)) || (IS_BFS(x)) \
|
|
|| (IS_BTS(x)))
|
|
#define IS_ADDPCBYIMM_ND(x) ((IS_BF(x)) || (IS_BT(x)))
|
|
|
|
#define IS_DELAYEDBRANCH(x) (IS_PCLOADFROMREG(x) || IS_RTS(x) \
|
|
|| IS_ADDPCBYREG(x) || IS_ADDPCBYIMM(x))
|
|
|
|
#define IS_BRANCH(x) (IS_DELAYEDBRANCH(x) || IS_ADDPCBYIMM_ND(x))
|
|
|
|
|
|
#define M2REG(x) (((x) & 0x0f00) >> 8)
|
|
#define GETIMM8(x) ((x) & 0x00ff)
|
|
#define IMM8SIGN(x) ((x) & 0x0080)
|
|
#define GETIMM12(x) ((x) & 0x0fff)
|
|
#define IMM12SIGN(x) ((x) & 0x0800)
|
|
|
|
void
|
|
single_step (ignore)
|
|
enum target_signal ignore; /* not used */
|
|
{
|
|
CORE_ADDR pc;
|
|
CORE_ADDR epc;
|
|
CORE_ADDR next;
|
|
unsigned short delay_instruction;
|
|
CORE_ADDR offset;
|
|
unsigned short insn;
|
|
|
|
if (!one_stepped)
|
|
{
|
|
pc = epc = read_register (PC_REGNUM);
|
|
insn = read_memory_integer(pc, sizeof(insn));
|
|
|
|
if (IS_DELAYEDBRANCH(insn))
|
|
pc += 2;
|
|
next = pc + 2;
|
|
target_addr = 0;
|
|
step_addr = next;
|
|
|
|
if (IS_BRANCH(insn))
|
|
{
|
|
if (IS_PCLOADFROMREG(insn)) {
|
|
target_addr = read_register(M2REG (insn));
|
|
step_addr = 0;
|
|
} else if (IS_RTS(insn)) {
|
|
target_addr = read_register(PR_REGNUM);
|
|
step_addr = 0;
|
|
} else if (IS_ADDPCBYREG(insn)) {
|
|
target_addr = next + read_register(M2REG (insn));
|
|
step_addr = 0;
|
|
} else if (IS_BT(insn) || IS_BF(insn) || IS_BTS(insn)
|
|
|| IS_BFS(insn)) {
|
|
target_addr = GETIMM8(insn);
|
|
if (IMM8SIGN(insn))
|
|
target_addr |= ~(CORE_ADDR)0x00ff;
|
|
target_addr = epc + 4 + (target_addr << 1);
|
|
} else if (IS_BSR(insn) || IS_BRA(insn)) {
|
|
target_addr = GETIMM12(insn);
|
|
if (IMM12SIGN(insn))
|
|
target_addr |= ~(CORE_ADDR)0x0fff;
|
|
target_addr = next + (target_addr << 1);
|
|
step_addr = 0;
|
|
}
|
|
}
|
|
|
|
/* Don't try to put down two breakpoints in the same spot... */
|
|
if (step_addr == target_addr)
|
|
target_addr = 0;
|
|
|
|
if (step_addr)
|
|
{
|
|
target_insert_breakpoint (step_addr, (char *)&step_cache [0]);
|
|
if (step_addr + 2 != target_addr)
|
|
target_insert_breakpoint (step_addr + 2, (char *)&step_cache [1]);
|
|
}
|
|
if (target_addr)
|
|
{
|
|
target_insert_breakpoint (target_addr, (char *)&step_cache [2]);
|
|
}
|
|
|
|
/* If the breakpoint occurred in a branch instruction,
|
|
re-run the branch (the breakpoint instruction should
|
|
be gone by now)... */
|
|
if (epc != pc)
|
|
{
|
|
write_register (PC_REGNUM, epc);
|
|
}
|
|
one_stepped = 1;
|
|
return;
|
|
}
|
|
else
|
|
{
|
|
pc = epc = read_register(PC_REGNUM);
|
|
|
|
write_pc(pc -= 2);
|
|
|
|
insn = read_memory_integer(pc, sizeof(insn));
|
|
|
|
/* Remove step breakpoints */
|
|
if (step_addr)
|
|
{
|
|
target_remove_breakpoint (step_addr, (char *)&step_cache [0]);
|
|
if (step_addr + 2 != target_addr)
|
|
target_remove_breakpoint (step_addr + 2, (char *)&step_cache [1]);
|
|
}
|
|
|
|
if (target_addr)
|
|
{
|
|
target_remove_breakpoint (target_addr, (char *)&step_cache [2]);
|
|
target_addr = 0;
|
|
}
|
|
|
|
one_stepped = 0;
|
|
}
|
|
}
|
|
#endif /* NO_SINGLE_STEP */
|