mirror of
https://sourceware.org/git/binutils-gdb.git
synced 2024-12-09 04:21:49 +08:00
16726dd15b
and stack pointer during stepping, to speed things up. A. Changes to not select a frame until we need a selected frame: * blockframe.c (flush_cached_frames): Call select_frame (NULL, -1). * infrun.c (wait_for_inferior): Move call to select_frame back to normal_stop. This reverts a change of 13 Apr 94 (it says Jeff Law, but the change was my idea); the only reason for that change was so we could save and restore the selected frame in wait_for_inferior, and now that flush_cached frames clears the selected frame, that should work OK now. B. Changes to not create a current_frame until we need one: * blockframe.c (get_current_frame): If current_frame is NULL, try to create an innermost frame. * sparc-tdep.c (sparc_pop_frame), infcmd.c (run-stack_dummy), infrun.c (wait_for_inferior), thread.c (thread_switch), convex-tdep.c (set_thread_command), a29k-tdep.c (pop_frame), alpha-tdep.c (alpha_pop_frame), convex-xdep.c (core_file_command), h8300-tdep.c (h8300_pop_frame), h8500-tdep.c (h8300_pop_frame), hppa-tdep.c (hppa_pop_frame), i386-tdep.c (i386_pop_frame), i960-tdep.c (pop_frame), m68k-tdep.c (m68k_pop_frame), mips-tdep.c (mips_pop_frame), rs6000-tdep.c (push_dummy_frame, pop_dummy_frame, pop_frame), sh-tdep.c (pop_frame), config/arm/tm-arm.h (POP_FRAME), config/convex/tm-convex.h (POP_FRAME), config/gould/tm-pn.h (POP_FRAME), config/ns32k/tm-merlin.h (POP_FRAME), config/ns32k/tm-umax.h (POP_FRAME), config/tahoe/tm-tahoe.h (POP_FRAME), config/vax/tm-vax.h (POP_FRAME): Don't call create_new_frame. * corelow.c (core_open), altos-xdep.c (core_file_command), arm-xdep.c (core_file_command), gould-xdep.c (core_file_command), m3-nat.c (select_thread), sun386-nat.c (core_file_command), umax-xdep.c (core_file_command): Don't call create_new_frame; do call flush_cached_frames. * blockframe.c (reinit_frame_cache): Don't call create_new_frame or select_frame. C. Changes to get rid of stop_frame_address and instead only fetch the frame pointer when we need it. * breakpoint.c (bpstat_stop_status): Remove argument frame_address; use FRAME_FP (get_current_frame ()). * infrun.c (wait_for_inferior): Don't pass frame pointer to bpstat_stop_status. * infrun.c (wait_for_inferior): Use FRAME_FP (get_current_frame ()) instead of stop_frame_address. * infrun.c (save_inferior_status, restore_inferior_status), inferior.h (struct inferior_status): Don't save and restore stop_frame_address. * inferior.h, infcmd.c, thread.c (thread_switch), m3-nat.c (select_thread): Remove stop_frame_address and uses thereof. D. Same thing for the stack pointer. * infrun.c (wait_for_inferior): Remove stop_sp and replace uses thereof with read_sp (). E. Change to eliminate one nasty little spot where we were wanting to know the frame pointer from before the current step (idea from GDB 3.5, which saved my ass, because my other ideas of how to fix it were very baroque). * infrun.c: Remove prev_frame_address. * infrun.c (wait_for_inferior, step_over_function): Use step_frame_address instead of prev_frame_address. F. Same basic idea for the stack pointer. * inferior.h, infcmd.c: New variable step_sp. * infcmd.c (step_1, until_next_command): Set it. * infrun.c: Remove prev_sp and replace uses by step_sp. * infrun.c (wait_for_inferior): If we get out of the step range, then set step_sp to the current stack pointer before we start going again.
975 lines
28 KiB
C
975 lines
28 KiB
C
/* Target-machine dependent code for the AMD 29000
|
||
Copyright 1990, 1991, 1992, 1993, 1994 Free Software Foundation, Inc.
|
||
Contributed by Cygnus Support. Written by Jim Kingdon.
|
||
|
||
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., 675 Mass Ave, Cambridge, MA 02139, USA. */
|
||
|
||
#include "defs.h"
|
||
#include "gdbcore.h"
|
||
#include "frame.h"
|
||
#include "value.h"
|
||
#include "symtab.h"
|
||
#include "inferior.h"
|
||
#include "gdbcmd.h"
|
||
|
||
/* If all these bits in an instruction word are zero, it is a "tag word"
|
||
which precedes a function entry point and gives stack traceback info.
|
||
This used to be defined as 0xff000000, but that treated 0x00000deb as
|
||
a tag word, while it is really used as a breakpoint. */
|
||
#define TAGWORD_ZERO_MASK 0xff00f800
|
||
|
||
extern CORE_ADDR text_start; /* FIXME, kludge... */
|
||
|
||
/* The user-settable top of the register stack in virtual memory. We
|
||
won't attempt to access any stored registers above this address, if set
|
||
nonzero. */
|
||
|
||
static CORE_ADDR rstack_high_address = UINT_MAX;
|
||
|
||
/* Structure to hold cached info about function prologues. */
|
||
struct prologue_info
|
||
{
|
||
CORE_ADDR pc; /* First addr after fn prologue */
|
||
unsigned rsize, msize; /* register stack frame size, mem stack ditto */
|
||
unsigned mfp_used : 1; /* memory frame pointer used */
|
||
unsigned rsize_valid : 1; /* Validity bits for the above */
|
||
unsigned msize_valid : 1;
|
||
unsigned mfp_valid : 1;
|
||
};
|
||
|
||
/* Examine the prologue of a function which starts at PC. Return
|
||
the first addess past the prologue. If MSIZE is non-NULL, then
|
||
set *MSIZE to the memory stack frame size. If RSIZE is non-NULL,
|
||
then set *RSIZE to the register stack frame size (not including
|
||
incoming arguments and the return address & frame pointer stored
|
||
with them). If no prologue is found, *RSIZE is set to zero.
|
||
If no prologue is found, or a prologue which doesn't involve
|
||
allocating a memory stack frame, then set *MSIZE to zero.
|
||
|
||
Note that both msize and rsize are in bytes. This is not consistent
|
||
with the _User's Manual_ with respect to rsize, but it is much more
|
||
convenient.
|
||
|
||
If MFP_USED is non-NULL, *MFP_USED is set to nonzero if a memory
|
||
frame pointer is being used. */
|
||
CORE_ADDR
|
||
examine_prologue (pc, rsize, msize, mfp_used)
|
||
CORE_ADDR pc;
|
||
unsigned *msize;
|
||
unsigned *rsize;
|
||
int *mfp_used;
|
||
{
|
||
long insn;
|
||
CORE_ADDR p = pc;
|
||
struct minimal_symbol *msymbol = lookup_minimal_symbol_by_pc (pc);
|
||
struct prologue_info *mi = 0;
|
||
|
||
if (msymbol != NULL)
|
||
mi = (struct prologue_info *) msymbol -> info;
|
||
|
||
if (mi != 0)
|
||
{
|
||
int valid = 1;
|
||
if (rsize != NULL)
|
||
{
|
||
*rsize = mi->rsize;
|
||
valid &= mi->rsize_valid;
|
||
}
|
||
if (msize != NULL)
|
||
{
|
||
*msize = mi->msize;
|
||
valid &= mi->msize_valid;
|
||
}
|
||
if (mfp_used != NULL)
|
||
{
|
||
*mfp_used = mi->mfp_used;
|
||
valid &= mi->mfp_valid;
|
||
}
|
||
if (valid)
|
||
return mi->pc;
|
||
}
|
||
|
||
if (rsize != NULL)
|
||
*rsize = 0;
|
||
if (msize != NULL)
|
||
*msize = 0;
|
||
if (mfp_used != NULL)
|
||
*mfp_used = 0;
|
||
|
||
/* Prologue must start with subtracting a constant from gr1.
|
||
Normally this is sub gr1,gr1,<rsize * 4>. */
|
||
insn = read_memory_integer (p, 4);
|
||
if ((insn & 0xffffff00) != 0x25010100)
|
||
{
|
||
/* If the frame is large, instead of a single instruction it
|
||
might be a pair of instructions:
|
||
const <reg>, <rsize * 4>
|
||
sub gr1,gr1,<reg>
|
||
*/
|
||
int reg;
|
||
/* Possible value for rsize. */
|
||
unsigned int rsize0;
|
||
|
||
if ((insn & 0xff000000) != 0x03000000)
|
||
{
|
||
p = pc;
|
||
goto done;
|
||
}
|
||
reg = (insn >> 8) & 0xff;
|
||
rsize0 = (((insn >> 8) & 0xff00) | (insn & 0xff));
|
||
p += 4;
|
||
insn = read_memory_integer (p, 4);
|
||
if ((insn & 0xffffff00) != 0x24010100
|
||
|| (insn & 0xff) != reg)
|
||
{
|
||
p = pc;
|
||
goto done;
|
||
}
|
||
if (rsize != NULL)
|
||
*rsize = rsize0;
|
||
}
|
||
else
|
||
{
|
||
if (rsize != NULL)
|
||
*rsize = (insn & 0xff);
|
||
}
|
||
p += 4;
|
||
|
||
/* Next instruction ought to be asgeu V_SPILL,gr1,rab.
|
||
* We don't check the vector number to allow for kernel debugging. The
|
||
* kernel will use a different trap number.
|
||
* If this insn is missing, we just keep going; Metaware R2.3u compiler
|
||
* generates prologue that intermixes initializations and puts the asgeu
|
||
* way down.
|
||
*/
|
||
insn = read_memory_integer (p, 4);
|
||
if ((insn & 0xff00ffff) == (0x5e000100|RAB_HW_REGNUM))
|
||
{
|
||
p += 4;
|
||
}
|
||
|
||
/* Next instruction usually sets the frame pointer (lr1) by adding
|
||
<size * 4> from gr1. However, this can (and high C does) be
|
||
deferred until anytime before the first function call. So it is
|
||
OK if we don't see anything which sets lr1.
|
||
To allow for alternate register sets (gcc -mkernel-registers) the msp
|
||
register number is a compile time constant. */
|
||
|
||
/* Normally this is just add lr1,gr1,<size * 4>. */
|
||
insn = read_memory_integer (p, 4);
|
||
if ((insn & 0xffffff00) == 0x15810100)
|
||
p += 4;
|
||
else
|
||
{
|
||
/* However, for large frames it can be
|
||
const <reg>, <size *4>
|
||
add lr1,gr1,<reg>
|
||
*/
|
||
int reg;
|
||
CORE_ADDR q;
|
||
|
||
if ((insn & 0xff000000) == 0x03000000)
|
||
{
|
||
reg = (insn >> 8) & 0xff;
|
||
q = p + 4;
|
||
insn = read_memory_integer (q, 4);
|
||
if ((insn & 0xffffff00) == 0x14810100
|
||
&& (insn & 0xff) == reg)
|
||
p = q;
|
||
}
|
||
}
|
||
|
||
/* Next comes "add lr{<rsize-1>},msp,0", but only if a memory
|
||
frame pointer is in use. We just check for add lr<anything>,msp,0;
|
||
we don't check this rsize against the first instruction, and
|
||
we don't check that the trace-back tag indicates a memory frame pointer
|
||
is in use.
|
||
To allow for alternate register sets (gcc -mkernel-registers) the msp
|
||
register number is a compile time constant.
|
||
|
||
The recommended instruction is actually "sll lr<whatever>,msp,0".
|
||
We check for that, too. Originally Jim Kingdon's code seemed
|
||
to be looking for a "sub" instruction here, but the mask was set
|
||
up to lose all the time. */
|
||
insn = read_memory_integer (p, 4);
|
||
if (((insn & 0xff80ffff) == (0x15800000|(MSP_HW_REGNUM<<8))) /* add */
|
||
|| ((insn & 0xff80ffff) == (0x81800000|(MSP_HW_REGNUM<<8)))) /* sll */
|
||
{
|
||
p += 4;
|
||
if (mfp_used != NULL)
|
||
*mfp_used = 1;
|
||
}
|
||
|
||
/* Next comes a subtraction from msp to allocate a memory frame,
|
||
but only if a memory frame is
|
||
being used. We don't check msize against the trace-back tag.
|
||
|
||
To allow for alternate register sets (gcc -mkernel-registers) the msp
|
||
register number is a compile time constant.
|
||
|
||
Normally this is just
|
||
sub msp,msp,<msize>
|
||
*/
|
||
insn = read_memory_integer (p, 4);
|
||
if ((insn & 0xffffff00) ==
|
||
(0x25000000|(MSP_HW_REGNUM<<16)|(MSP_HW_REGNUM<<8)))
|
||
{
|
||
p += 4;
|
||
if (msize != NULL)
|
||
*msize = insn & 0xff;
|
||
}
|
||
else
|
||
{
|
||
/* For large frames, instead of a single instruction it might
|
||
be
|
||
|
||
const <reg>, <msize>
|
||
consth <reg>, <msize> ; optional
|
||
sub msp,msp,<reg>
|
||
*/
|
||
int reg;
|
||
unsigned msize0;
|
||
CORE_ADDR q = p;
|
||
|
||
if ((insn & 0xff000000) == 0x03000000)
|
||
{
|
||
reg = (insn >> 8) & 0xff;
|
||
msize0 = ((insn >> 8) & 0xff00) | (insn & 0xff);
|
||
q += 4;
|
||
insn = read_memory_integer (q, 4);
|
||
/* Check for consth. */
|
||
if ((insn & 0xff000000) == 0x02000000
|
||
&& (insn & 0x0000ff00) == reg)
|
||
{
|
||
msize0 |= (insn << 8) & 0xff000000;
|
||
msize0 |= (insn << 16) & 0x00ff0000;
|
||
q += 4;
|
||
insn = read_memory_integer (q, 4);
|
||
}
|
||
/* Check for sub msp,msp,<reg>. */
|
||
if ((insn & 0xffffff00) ==
|
||
(0x24000000|(MSP_HW_REGNUM<<16)|(MSP_HW_REGNUM<<8))
|
||
&& (insn & 0xff) == reg)
|
||
{
|
||
p = q + 4;
|
||
if (msize != NULL)
|
||
*msize = msize0;
|
||
}
|
||
}
|
||
}
|
||
|
||
/* Next instruction might be asgeu V_SPILL,gr1,rab.
|
||
* We don't check the vector number to allow for kernel debugging. The
|
||
* kernel will use a different trap number.
|
||
* Metaware R2.3u compiler
|
||
* generates prologue that intermixes initializations and puts the asgeu
|
||
* way down after everything else.
|
||
*/
|
||
insn = read_memory_integer (p, 4);
|
||
if ((insn & 0xff00ffff) == (0x5e000100|RAB_HW_REGNUM))
|
||
{
|
||
p += 4;
|
||
}
|
||
|
||
done:
|
||
if (msymbol != NULL)
|
||
{
|
||
if (mi == 0)
|
||
{
|
||
/* Add a new cache entry. */
|
||
mi = (struct prologue_info *)xmalloc (sizeof (struct prologue_info));
|
||
msymbol -> info = (char *)mi;
|
||
mi->rsize_valid = 0;
|
||
mi->msize_valid = 0;
|
||
mi->mfp_valid = 0;
|
||
}
|
||
/* else, cache entry exists, but info is incomplete. */
|
||
mi->pc = p;
|
||
if (rsize != NULL)
|
||
{
|
||
mi->rsize = *rsize;
|
||
mi->rsize_valid = 1;
|
||
}
|
||
if (msize != NULL)
|
||
{
|
||
mi->msize = *msize;
|
||
mi->msize_valid = 1;
|
||
}
|
||
if (mfp_used != NULL)
|
||
{
|
||
mi->mfp_used = *mfp_used;
|
||
mi->mfp_valid = 1;
|
||
}
|
||
}
|
||
return p;
|
||
}
|
||
|
||
/* Advance PC across any function entry prologue instructions
|
||
to reach some "real" code. */
|
||
|
||
CORE_ADDR
|
||
skip_prologue (pc)
|
||
CORE_ADDR pc;
|
||
{
|
||
return examine_prologue (pc, (unsigned *)NULL, (unsigned *)NULL,
|
||
(int *)NULL);
|
||
}
|
||
/*
|
||
* Examine the one or two word tag at the beginning of a function.
|
||
* The tag word is expect to be at 'p', if it is not there, we fail
|
||
* by returning 0. The documentation for the tag word was taken from
|
||
* page 7-15 of the 29050 User's Manual. We are assuming that the
|
||
* m bit is in bit 22 of the tag word, which seems to be the agreed upon
|
||
* convention today (1/15/92).
|
||
* msize is return in bytes.
|
||
*/
|
||
static int /* 0/1 - failure/success of finding the tag word */
|
||
examine_tag(p, is_trans, argcount, msize, mfp_used)
|
||
CORE_ADDR p;
|
||
int *is_trans;
|
||
int *argcount;
|
||
unsigned *msize;
|
||
int *mfp_used;
|
||
{
|
||
unsigned int tag1, tag2;
|
||
|
||
tag1 = read_memory_integer (p, 4);
|
||
if ((tag1 & TAGWORD_ZERO_MASK) != 0) /* Not a tag word */
|
||
return 0;
|
||
if (tag1 & (1<<23)) /* A two word tag */
|
||
{
|
||
tag2 = read_memory_integer (p+4, 4);
|
||
if (msize)
|
||
*msize = tag2;
|
||
}
|
||
else /* A one word tag */
|
||
{
|
||
if (msize)
|
||
*msize = tag1 & 0x7ff;
|
||
}
|
||
if (is_trans)
|
||
*is_trans = ((tag1 & (1<<21)) ? 1 : 0);
|
||
/* Note that this includes the frame pointer and the return address
|
||
register, so the actual number of registers of arguments is two less.
|
||
argcount can be zero, however, sometimes, for strange assembler
|
||
routines. */
|
||
if (argcount)
|
||
*argcount = (tag1 >> 16) & 0x1f;
|
||
if (mfp_used)
|
||
*mfp_used = ((tag1 & (1<<22)) ? 1 : 0);
|
||
return(1);
|
||
}
|
||
|
||
/* Initialize the frame. In addition to setting "extra" frame info,
|
||
we also set ->frame because we use it in a nonstandard way, and ->pc
|
||
because we need to know it to get the other stuff. See the diagram
|
||
of stacks and the frame cache in tm-a29k.h for more detail. */
|
||
static void
|
||
init_frame_info (innermost_frame, fci)
|
||
int innermost_frame;
|
||
struct frame_info *fci;
|
||
{
|
||
CORE_ADDR p;
|
||
long insn;
|
||
unsigned rsize;
|
||
unsigned msize;
|
||
int mfp_used, trans;
|
||
struct symbol *func;
|
||
|
||
p = fci->pc;
|
||
|
||
if (innermost_frame)
|
||
fci->frame = read_register (GR1_REGNUM);
|
||
else
|
||
fci->frame = fci->next->frame + fci->next->rsize;
|
||
|
||
#if CALL_DUMMY_LOCATION == ON_STACK
|
||
This wont work;
|
||
#else
|
||
if (PC_IN_CALL_DUMMY (p, 0, 0))
|
||
#endif
|
||
{
|
||
fci->rsize = DUMMY_FRAME_RSIZE;
|
||
/* This doesn't matter since we never try to get locals or args
|
||
from a dummy frame. */
|
||
fci->msize = 0;
|
||
/* Dummy frames always use a memory frame pointer. */
|
||
fci->saved_msp =
|
||
read_register_stack_integer (fci->frame + DUMMY_FRAME_RSIZE - 4, 4);
|
||
fci->flags |= (TRANSPARENT|MFP_USED);
|
||
return;
|
||
}
|
||
|
||
func = find_pc_function (p);
|
||
if (func != NULL)
|
||
p = BLOCK_START (SYMBOL_BLOCK_VALUE (func));
|
||
else
|
||
{
|
||
/* Search backward to find the trace-back tag. However,
|
||
do not trace back beyond the start of the text segment
|
||
(just as a sanity check to avoid going into never-never land). */
|
||
#if 1
|
||
while (p >= text_start
|
||
&& ((insn = read_memory_integer (p, 4)) & TAGWORD_ZERO_MASK) != 0)
|
||
p -= 4;
|
||
#else /* 0 */
|
||
char pat[4] = {0, 0, 0, 0};
|
||
char mask[4];
|
||
char insn_raw[4];
|
||
store_unsigned_integer (mask, 4, TAGWORD_ZERO_MASK);
|
||
/* Enable this once target_search is enabled and tested. */
|
||
target_search (4, pat, mask, p, -4, text_start, p+1, &p, &insn_raw);
|
||
insn = extract_unsigned_integer (insn_raw, 4);
|
||
#endif /* 0 */
|
||
|
||
if (p < text_start)
|
||
{
|
||
/* Couldn't find the trace-back tag.
|
||
Something strange is going on. */
|
||
fci->saved_msp = 0;
|
||
fci->rsize = 0;
|
||
fci->msize = 0;
|
||
fci->flags = TRANSPARENT;
|
||
return;
|
||
}
|
||
else
|
||
/* Advance to the first word of the function, i.e. the word
|
||
after the trace-back tag. */
|
||
p += 4;
|
||
}
|
||
|
||
/* We've found the start of the function.
|
||
Try looking for a tag word that indicates whether there is a
|
||
memory frame pointer and what the memory stack allocation is.
|
||
If one doesn't exist, try using a more exhaustive search of
|
||
the prologue. */
|
||
|
||
if (examine_tag(p-4,&trans,(int *)NULL,&msize,&mfp_used)) /* Found good tag */
|
||
examine_prologue (p, &rsize, 0, 0);
|
||
else /* No tag try prologue */
|
||
examine_prologue (p, &rsize, &msize, &mfp_used);
|
||
|
||
fci->rsize = rsize;
|
||
fci->msize = msize;
|
||
fci->flags = 0;
|
||
if (mfp_used)
|
||
fci->flags |= MFP_USED;
|
||
if (trans)
|
||
fci->flags |= TRANSPARENT;
|
||
if (innermost_frame)
|
||
{
|
||
fci->saved_msp = read_register (MSP_REGNUM) + msize;
|
||
}
|
||
else
|
||
{
|
||
if (mfp_used)
|
||
fci->saved_msp =
|
||
read_register_stack_integer (fci->frame + rsize - 4, 4);
|
||
else
|
||
fci->saved_msp = fci->next->saved_msp + msize;
|
||
}
|
||
}
|
||
|
||
void
|
||
init_extra_frame_info (fci)
|
||
struct frame_info *fci;
|
||
{
|
||
if (fci->next == 0)
|
||
/* Assume innermost frame. May produce strange results for "info frame"
|
||
but there isn't any way to tell the difference. */
|
||
init_frame_info (1, fci);
|
||
else {
|
||
/* We're in get_prev_frame_info.
|
||
Take care of everything in init_frame_pc. */
|
||
;
|
||
}
|
||
}
|
||
|
||
void
|
||
init_frame_pc (fromleaf, fci)
|
||
int fromleaf;
|
||
struct frame_info *fci;
|
||
{
|
||
fci->pc = (fromleaf ? SAVED_PC_AFTER_CALL (fci->next) :
|
||
fci->next ? FRAME_SAVED_PC (fci->next) : read_pc ());
|
||
init_frame_info (fromleaf, fci);
|
||
}
|
||
|
||
/* Local variables (i.e. LOC_LOCAL) are on the memory stack, with their
|
||
offsets being relative to the memory stack pointer (high C) or
|
||
saved_msp (gcc). */
|
||
|
||
CORE_ADDR
|
||
frame_locals_address (fi)
|
||
struct frame_info *fi;
|
||
{
|
||
if (fi->flags & MFP_USED)
|
||
return fi->saved_msp;
|
||
else
|
||
return fi->saved_msp - fi->msize;
|
||
}
|
||
|
||
/* Routines for reading the register stack. The caller gets to treat
|
||
the register stack as a uniform stack in memory, from address $gr1
|
||
straight through $rfb and beyond. */
|
||
|
||
/* Analogous to read_memory except the length is understood to be 4.
|
||
Also, myaddr can be NULL (meaning don't bother to read), and
|
||
if actual_mem_addr is non-NULL, store there the address that it
|
||
was fetched from (or if from a register the offset within
|
||
registers). Set *LVAL to lval_memory or lval_register, depending
|
||
on where it came from. The contents written into MYADDR are in
|
||
target format. */
|
||
void
|
||
read_register_stack (memaddr, myaddr, actual_mem_addr, lval)
|
||
CORE_ADDR memaddr;
|
||
char *myaddr;
|
||
CORE_ADDR *actual_mem_addr;
|
||
enum lval_type *lval;
|
||
{
|
||
long rfb = read_register (RFB_REGNUM);
|
||
long rsp = read_register (RSP_REGNUM);
|
||
|
||
/* If we don't do this 'info register' stops in the middle. */
|
||
if (memaddr >= rstack_high_address)
|
||
{
|
||
/* a bogus value */
|
||
static char val[] = {~0, ~0, ~0, ~0};
|
||
/* It's in a local register, but off the end of the stack. */
|
||
int regnum = (memaddr - rsp) / 4 + LR0_REGNUM;
|
||
if (myaddr != NULL)
|
||
{
|
||
/* Provide bogusness */
|
||
memcpy (myaddr, val, 4);
|
||
}
|
||
supply_register(regnum, val); /* More bogusness */
|
||
if (lval != NULL)
|
||
*lval = lval_register;
|
||
if (actual_mem_addr != NULL)
|
||
*actual_mem_addr = REGISTER_BYTE (regnum);
|
||
}
|
||
/* If it's in the part of the register stack that's in real registers,
|
||
get the value from the registers. If it's anywhere else in memory
|
||
(e.g. in another thread's saved stack), skip this part and get
|
||
it from real live memory. */
|
||
else if (memaddr < rfb && memaddr >= rsp)
|
||
{
|
||
/* It's in a register. */
|
||
int regnum = (memaddr - rsp) / 4 + LR0_REGNUM;
|
||
if (regnum > LR0_REGNUM + 127)
|
||
error ("Attempt to read register stack out of range.");
|
||
if (myaddr != NULL)
|
||
read_register_gen (regnum, myaddr);
|
||
if (lval != NULL)
|
||
*lval = lval_register;
|
||
if (actual_mem_addr != NULL)
|
||
*actual_mem_addr = REGISTER_BYTE (regnum);
|
||
}
|
||
else
|
||
{
|
||
/* It's in the memory portion of the register stack. */
|
||
if (myaddr != NULL)
|
||
read_memory (memaddr, myaddr, 4);
|
||
if (lval != NULL)
|
||
*lval = lval_memory;
|
||
if (actual_mem_addr != NULL)
|
||
*actual_mem_addr = memaddr;
|
||
}
|
||
}
|
||
|
||
/* Analogous to read_memory_integer
|
||
except the length is understood to be 4. */
|
||
long
|
||
read_register_stack_integer (memaddr, len)
|
||
CORE_ADDR memaddr;
|
||
int len;
|
||
{
|
||
char buf[4];
|
||
read_register_stack (memaddr, buf, NULL, NULL);
|
||
return extract_signed_integer (buf, 4);
|
||
}
|
||
|
||
/* Copy 4 bytes from GDB memory at MYADDR into inferior memory
|
||
at MEMADDR and put the actual address written into in
|
||
*ACTUAL_MEM_ADDR. */
|
||
static void
|
||
write_register_stack (memaddr, myaddr, actual_mem_addr)
|
||
CORE_ADDR memaddr;
|
||
char *myaddr;
|
||
CORE_ADDR *actual_mem_addr;
|
||
{
|
||
long rfb = read_register (RFB_REGNUM);
|
||
long rsp = read_register (RSP_REGNUM);
|
||
/* If we don't do this 'info register' stops in the middle. */
|
||
if (memaddr >= rstack_high_address)
|
||
{
|
||
/* It's in a register, but off the end of the stack. */
|
||
if (actual_mem_addr != NULL)
|
||
*actual_mem_addr = 0;
|
||
}
|
||
else if (memaddr < rfb)
|
||
{
|
||
/* It's in a register. */
|
||
int regnum = (memaddr - rsp) / 4 + LR0_REGNUM;
|
||
if (regnum < LR0_REGNUM || regnum > LR0_REGNUM + 127)
|
||
error ("Attempt to read register stack out of range.");
|
||
if (myaddr != NULL)
|
||
write_register (regnum, *(long *)myaddr);
|
||
if (actual_mem_addr != NULL)
|
||
*actual_mem_addr = 0;
|
||
}
|
||
else
|
||
{
|
||
/* It's in the memory portion of the register stack. */
|
||
if (myaddr != NULL)
|
||
write_memory (memaddr, myaddr, 4);
|
||
if (actual_mem_addr != NULL)
|
||
*actual_mem_addr = memaddr;
|
||
}
|
||
}
|
||
|
||
/* Find register number REGNUM relative to FRAME and put its
|
||
(raw) contents in *RAW_BUFFER. Set *OPTIMIZED if the variable
|
||
was optimized out (and thus can't be fetched). If the variable
|
||
was fetched from memory, set *ADDRP to where it was fetched from,
|
||
otherwise it was fetched from a register.
|
||
|
||
The argument RAW_BUFFER must point to aligned memory. */
|
||
void
|
||
get_saved_register (raw_buffer, optimized, addrp, frame, regnum, lvalp)
|
||
char *raw_buffer;
|
||
int *optimized;
|
||
CORE_ADDR *addrp;
|
||
FRAME frame;
|
||
int regnum;
|
||
enum lval_type *lvalp;
|
||
{
|
||
struct frame_info *fi;
|
||
CORE_ADDR addr;
|
||
enum lval_type lval;
|
||
|
||
if (frame == 0)
|
||
return;
|
||
|
||
fi = get_frame_info (frame);
|
||
|
||
/* Once something has a register number, it doesn't get optimized out. */
|
||
if (optimized != NULL)
|
||
*optimized = 0;
|
||
if (regnum == RSP_REGNUM)
|
||
{
|
||
if (raw_buffer != NULL)
|
||
{
|
||
store_address (raw_buffer, REGISTER_RAW_SIZE (regnum), fi->frame);
|
||
}
|
||
if (lvalp != NULL)
|
||
*lvalp = not_lval;
|
||
return;
|
||
}
|
||
else if (regnum == PC_REGNUM)
|
||
{
|
||
if (raw_buffer != NULL)
|
||
{
|
||
store_address (raw_buffer, REGISTER_RAW_SIZE (regnum), fi->pc);
|
||
}
|
||
|
||
/* Not sure we have to do this. */
|
||
if (lvalp != NULL)
|
||
*lvalp = not_lval;
|
||
|
||
return;
|
||
}
|
||
else if (regnum == MSP_REGNUM)
|
||
{
|
||
if (raw_buffer != NULL)
|
||
{
|
||
if (fi->next != NULL)
|
||
{
|
||
store_address (raw_buffer, REGISTER_RAW_SIZE (regnum),
|
||
fi->next->saved_msp);
|
||
}
|
||
else
|
||
read_register_gen (MSP_REGNUM, raw_buffer);
|
||
}
|
||
/* The value may have been computed, not fetched. */
|
||
if (lvalp != NULL)
|
||
*lvalp = not_lval;
|
||
return;
|
||
}
|
||
else if (regnum < LR0_REGNUM || regnum >= LR0_REGNUM + 128)
|
||
{
|
||
/* These registers are not saved over procedure calls,
|
||
so just print out the current values. */
|
||
if (raw_buffer != NULL)
|
||
read_register_gen (regnum, raw_buffer);
|
||
if (lvalp != NULL)
|
||
*lvalp = lval_register;
|
||
if (addrp != NULL)
|
||
*addrp = REGISTER_BYTE (regnum);
|
||
return;
|
||
}
|
||
|
||
addr = fi->frame + (regnum - LR0_REGNUM) * 4;
|
||
if (raw_buffer != NULL)
|
||
read_register_stack (addr, raw_buffer, &addr, &lval);
|
||
if (lvalp != NULL)
|
||
*lvalp = lval;
|
||
if (addrp != NULL)
|
||
*addrp = addr;
|
||
}
|
||
|
||
|
||
/* Discard from the stack the innermost frame,
|
||
restoring all saved registers. */
|
||
|
||
void
|
||
pop_frame ()
|
||
{
|
||
FRAME frame = get_current_frame ();
|
||
struct frame_info *fi = get_frame_info (frame);
|
||
CORE_ADDR rfb = read_register (RFB_REGNUM);
|
||
CORE_ADDR gr1 = fi->frame + fi->rsize;
|
||
CORE_ADDR lr1;
|
||
CORE_ADDR original_lr0;
|
||
int must_fix_lr0 = 0;
|
||
int i;
|
||
|
||
/* If popping a dummy frame, need to restore registers. */
|
||
if (PC_IN_CALL_DUMMY (read_register (PC_REGNUM),
|
||
read_register (SP_REGNUM),
|
||
FRAME_FP (fi)))
|
||
{
|
||
int lrnum = LR0_REGNUM + DUMMY_ARG/4;
|
||
for (i = 0; i < DUMMY_SAVE_SR128; ++i)
|
||
write_register (SR_REGNUM (i + 128),read_register (lrnum++));
|
||
for (i = 0; i < DUMMY_SAVE_SR160; ++i)
|
||
write_register (SR_REGNUM(i+160), read_register (lrnum++));
|
||
for (i = 0; i < DUMMY_SAVE_GREGS; ++i)
|
||
write_register (RETURN_REGNUM + i, read_register (lrnum++));
|
||
/* Restore the PCs and prepare to restore LR0. */
|
||
write_register(PC_REGNUM, read_register (lrnum++));
|
||
write_register(NPC_REGNUM, read_register (lrnum++));
|
||
write_register(PC2_REGNUM, read_register (lrnum++));
|
||
original_lr0 = read_register (lrnum++);
|
||
must_fix_lr0 = 1;
|
||
}
|
||
|
||
/* Restore the memory stack pointer. */
|
||
write_register (MSP_REGNUM, fi->saved_msp);
|
||
/* Restore the register stack pointer. */
|
||
write_register (GR1_REGNUM, gr1);
|
||
|
||
/* If we popped a dummy frame, restore lr0 now that gr1 has been restored. */
|
||
if (must_fix_lr0)
|
||
write_register (LR0_REGNUM, original_lr0);
|
||
|
||
/* Check whether we need to fill registers. */
|
||
lr1 = read_register (LR0_REGNUM + 1);
|
||
if (lr1 > rfb)
|
||
{
|
||
/* Fill. */
|
||
int num_bytes = lr1 - rfb;
|
||
int i;
|
||
long word;
|
||
write_register (RAB_REGNUM, read_register (RAB_REGNUM) + num_bytes);
|
||
write_register (RFB_REGNUM, lr1);
|
||
for (i = 0; i < num_bytes; i += 4)
|
||
{
|
||
/* Note: word is in host byte order. */
|
||
word = read_memory_integer (rfb + i, 4);
|
||
write_register (LR0_REGNUM + ((rfb - gr1) % 0x80) + i / 4, word);
|
||
}
|
||
}
|
||
flush_cached_frames ();
|
||
}
|
||
|
||
/* Push an empty stack frame, to record the current PC, etc. */
|
||
|
||
void
|
||
push_dummy_frame ()
|
||
{
|
||
long w;
|
||
CORE_ADDR rab, gr1;
|
||
CORE_ADDR msp = read_register (MSP_REGNUM);
|
||
int lrnum, i;
|
||
CORE_ADDR original_lr0;
|
||
|
||
/* Read original lr0 before changing gr1. This order isn't really needed
|
||
since GDB happens to have a snapshot of all the regs and doesn't toss
|
||
it when gr1 is changed. But it's The Right Thing To Do. */
|
||
original_lr0 = read_register (LR0_REGNUM);
|
||
|
||
/* Allocate the new frame. */
|
||
gr1 = read_register (GR1_REGNUM) - DUMMY_FRAME_RSIZE;
|
||
write_register (GR1_REGNUM, gr1);
|
||
|
||
rab = read_register (RAB_REGNUM);
|
||
if (gr1 < rab)
|
||
{
|
||
/* We need to spill registers. */
|
||
int num_bytes = rab - gr1;
|
||
CORE_ADDR rfb = read_register (RFB_REGNUM);
|
||
int i;
|
||
long word;
|
||
|
||
write_register (RFB_REGNUM, rfb - num_bytes);
|
||
write_register (RAB_REGNUM, gr1);
|
||
for (i = 0; i < num_bytes; i += 4)
|
||
{
|
||
/* Note: word is in target byte order. */
|
||
read_register_gen (LR0_REGNUM + i / 4, (char *) &word);
|
||
write_memory (rfb - num_bytes + i, (char *) &word, 4);
|
||
}
|
||
}
|
||
|
||
/* There are no arguments in to the dummy frame, so we don't need
|
||
more than rsize plus the return address and lr1. */
|
||
write_register (LR0_REGNUM + 1, gr1 + DUMMY_FRAME_RSIZE + 2 * 4);
|
||
|
||
/* Set the memory frame pointer. */
|
||
write_register (LR0_REGNUM + DUMMY_FRAME_RSIZE / 4 - 1, msp);
|
||
|
||
/* Allocate arg_slop. */
|
||
write_register (MSP_REGNUM, msp - 16 * 4);
|
||
|
||
/* Save registers. */
|
||
lrnum = LR0_REGNUM + DUMMY_ARG/4;
|
||
for (i = 0; i < DUMMY_SAVE_SR128; ++i)
|
||
write_register (lrnum++, read_register (SR_REGNUM (i + 128)));
|
||
for (i = 0; i < DUMMY_SAVE_SR160; ++i)
|
||
write_register (lrnum++, read_register (SR_REGNUM (i + 160)));
|
||
for (i = 0; i < DUMMY_SAVE_GREGS; ++i)
|
||
write_register (lrnum++, read_register (RETURN_REGNUM + i));
|
||
/* Save the PCs and LR0. */
|
||
write_register (lrnum++, read_register (PC_REGNUM));
|
||
write_register (lrnum++, read_register (NPC_REGNUM));
|
||
write_register (lrnum++, read_register (PC2_REGNUM));
|
||
|
||
/* Why are we saving LR0? What would clobber it? (the dummy frame should
|
||
be below it on the register stack, no?). */
|
||
write_register (lrnum++, original_lr0);
|
||
}
|
||
|
||
|
||
|
||
/*
|
||
This routine takes three arguments and makes the cached frames look
|
||
as if these arguments defined a frame on the cache. This allows the
|
||
rest of `info frame' to extract the important arguments without much
|
||
difficulty. Since an individual frame on the 29K is determined by
|
||
three values (FP, PC, and MSP), we really need all three to do a
|
||
good job. */
|
||
|
||
FRAME
|
||
setup_arbitrary_frame (argc, argv)
|
||
int argc;
|
||
FRAME_ADDR *argv;
|
||
{
|
||
FRAME fid;
|
||
|
||
if (argc != 3)
|
||
error ("AMD 29k frame specifications require three arguments: rsp pc msp");
|
||
|
||
fid = create_new_frame (argv[0], argv[1]);
|
||
|
||
if (!fid)
|
||
fatal ("internal: create_new_frame returned invalid frame id");
|
||
|
||
/* Creating a new frame munges the `frame' value from the current
|
||
GR1, so we restore it again here. FIXME, untangle all this
|
||
29K frame stuff... */
|
||
fid->frame = argv[0];
|
||
|
||
/* Our MSP is in argv[2]. It'd be intelligent if we could just
|
||
save this value in the FRAME. But the way it's set up (FIXME),
|
||
we must save our caller's MSP. We compute that by adding our
|
||
memory stack frame size to our MSP. */
|
||
fid->saved_msp = argv[2] + fid->msize;
|
||
|
||
return fid;
|
||
}
|
||
|
||
|
||
|
||
enum a29k_processor_types processor_type = a29k_unknown;
|
||
|
||
void
|
||
a29k_get_processor_type ()
|
||
{
|
||
unsigned int cfg_reg = (unsigned int) read_register (CFG_REGNUM);
|
||
|
||
/* Most of these don't have freeze mode. */
|
||
processor_type = a29k_no_freeze_mode;
|
||
|
||
switch ((cfg_reg >> 28) & 0xf)
|
||
{
|
||
case 0:
|
||
fprintf_filtered (gdb_stderr, "Remote debugging an Am29000");
|
||
break;
|
||
case 1:
|
||
fprintf_filtered (gdb_stderr, "Remote debugging an Am29005");
|
||
break;
|
||
case 2:
|
||
fprintf_filtered (gdb_stderr, "Remote debugging an Am29050");
|
||
processor_type = a29k_freeze_mode;
|
||
break;
|
||
case 3:
|
||
fprintf_filtered (gdb_stderr, "Remote debugging an Am29035");
|
||
break;
|
||
case 4:
|
||
fprintf_filtered (gdb_stderr, "Remote debugging an Am29030");
|
||
break;
|
||
case 5:
|
||
fprintf_filtered (gdb_stderr, "Remote debugging an Am2920*");
|
||
break;
|
||
case 6:
|
||
fprintf_filtered (gdb_stderr, "Remote debugging an Am2924*");
|
||
break;
|
||
case 7:
|
||
fprintf_filtered (gdb_stderr, "Remote debugging an Am29040");
|
||
break;
|
||
default:
|
||
fprintf_filtered (gdb_stderr, "Remote debugging an unknown Am29k\n");
|
||
/* Don't bother to print the revision. */
|
||
return;
|
||
}
|
||
fprintf_filtered (gdb_stderr, " revision %c\n", 'A' + ((cfg_reg >> 24) & 0x0f));
|
||
}
|
||
|
||
void
|
||
_initialize_29k()
|
||
{
|
||
extern CORE_ADDR text_end;
|
||
|
||
/* FIXME, there should be a way to make a CORE_ADDR variable settable. */
|
||
add_show_from_set
|
||
(add_set_cmd ("rstack_high_address", class_support, var_uinteger,
|
||
(char *)&rstack_high_address,
|
||
"Set top address in memory of the register stack.\n\
|
||
Attempts to access registers saved above this address will be ignored\n\
|
||
or will produce the value -1.", &setlist),
|
||
&showlist);
|
||
|
||
/* FIXME, there should be a way to make a CORE_ADDR variable settable. */
|
||
add_show_from_set
|
||
(add_set_cmd ("call_scratch_address", class_support, var_uinteger,
|
||
(char *)&text_end,
|
||
"Set address in memory where small amounts of RAM can be used\n\
|
||
when making function calls into the inferior.", &setlist),
|
||
&showlist);
|
||
}
|