Hi Pedro,

I make a new patch according to your mail.
Please help me review it.

Thanks,
Hui

On Tue, Mar 10, 2009 at 04:34, Pedro Alves <pedro@codesourcery.com> wrote:
> On Monday 23 February 2009 09:20:13, teawater wrote:
>> This is the new version patches that follow cvs-head.
>
> Thanks.
>
>>
>>  gdb/Makefile.in |    4
>>  gdb/record.c    | 1283 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++
>>  gdb/record.h    |   87 +++
>>  3 files changed, 1372 insertions(+), 2 deletions(-)
>>
>> Index: src/gdb/Makefile.in
>> ===================================================================
>> --- src.orig/gdb/Makefile.in  2009-02-21 15:53:27.000000000 +0000
>> +++ src/gdb/Makefile.in       2009-02-28 20:23:20.000000000 +0000
>> @@ -662,7 +662,7 @@ SFILES = ada-exp.y ada-lang.c ada-typepr
>>       valarith.c valops.c valprint.c value.c varobj.c vec.c \
>>       wrapper.c \
>>       xml-tdesc.c xml-support.c \
>> -     inferior.c
>> +     inferior.c record.c
>>
>>  LINTFILES = $(SFILES) $(YYFILES) $(CONFIG_SRCS) init.c
>>
>> @@ -813,7 +813,7 @@ COMMON_OBS = $(DEPFILES) $(CONFIG_OBS) $
>>       solib.o solib-null.o \
>>       prologue-value.o memory-map.o xml-support.o \
>>       target-descriptions.o target-memory.o xml-tdesc.o xml-builtin.o \
>> -     inferior.o osdata.o
>> +     inferior.o osdata.o record.o
>>
>>  TSOBS = inflow.o
>>
>> Index: src/gdb/record.c
>> ===================================================================
>> --- /dev/null 1970-01-01 00:00:00.000000000 +0000
>> +++ src/gdb/record.c  2009-02-28 20:23:20.000000000 +0000
>> @@ -0,0 +1,1283 @@
>> +/* Process record and replay target for GDB, the GNU debugger.
>> +
>> +   Copyright (C) 2008 Free Software Foundation, Inc.
>
> Oops, you'll have to update this.
>
>> +
>> +   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 3 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, see <http://www.gnu.org/licenses/>.  */
>> +
>> +#include "defs.h"
>> +#include "target.h"
>> +#include "gdbcmd.h"
>> +#include "regcache.h"
>> +#include "inferior.h"
>> +#include "gdbthread.h"
>> +#include "event-top.h"
>> +#include "annotate.h"
>> +#include "observer.h"
>> +#include "record.h"
>
> Why did you need annotate.h?  Can you check which headers are
> really needed here?
>
>> +
>> +#include <signal.h>
>> +
>> +#define DEFAULT_RECORD_INSN_MAX_NUM  200000
>> +
>> +int record_debug = 0;
>> +
>> +record_t record_first;
>> +record_t *record_list = &record_first;
>> +record_t *record_arch_list_head = NULL;
>> +record_t *record_arch_list_tail = NULL;
>> +struct regcache *record_regcache = NULL;
>> +
>> +/* 1 ask user. 0 auto delete the last record_t.  */
>                 ^ double-space please.
>
>> +static int record_stop_at_limit = 1;
>> +static int record_insn_max_num = DEFAULT_RECORD_INSN_MAX_NUM;
>> +static int record_insn_num = 0;
>> +
>> +struct target_ops record_ops;
>> +static int record_resume_step = 0;
>> +static enum target_signal record_resume_siggnal;
>> +static int record_get_sig = 0;
>> +static sigset_t record_maskall;
>> +static int record_gdb_operation_disable = 0;
>> +int record_will_store_registers = 0;
>
> You've got a bunch of magic variables here.  Could you
> add some comments explaining what they're for?
>
>> +
>> +extern struct bp_location *bp_location_chain;
>
> This is not right.  You're accessing a breakpoints.c
> internal implementation detail.  We need to come up with interfaces
> to replace this hack.
>
>> +
>> +/* The beneath function pointers.  */
>> +static struct target_ops *record_beneath_to_resume_ops = NULL;
>> +static void (*record_beneath_to_resume) (struct target_ops *, ptid_t, int,
>> +                                         enum target_signal) = NULL;
>> +static struct target_ops *record_beneath_to_wait_ops = NULL;
>> +static ptid_t (*record_beneath_to_wait) (struct target_ops *, ptid_t,
>> +                                      struct target_waitstatus *) = NULL;
>> +static struct target_ops *record_beneath_to_store_registers_ops = NULL;
>> +static void (*record_beneath_to_store_registers) (struct target_ops *,
>> +                                                  struct regcache *,
>> +                                               int regno) = NULL;
>> +static struct target_ops *record_beneath_to_xfer_partial_ops = NULL;
>> +static LONGEST (*record_beneath_to_xfer_partial) (struct target_ops * ops,
>> +                                               enum target_object object,
>> +                                               const char *annex,
>> +                                               gdb_byte * readbuf,
>> +                                               const gdb_byte * writebuf,
>> +                                               ULONGEST offset,
>> +                                               LONGEST len) = NULL;
>> +static int (*record_beneath_to_insert_breakpoint) (struct bp_target_info *) =
>> +  NULL;
>> +static int (*record_beneath_to_remove_breakpoint) (struct bp_target_info *) =
>> +  NULL;
>
> I can't say I'm happy with this mechanism yet, but it is much
> better than the previous version of making target.c aware of record.c's callbacks.
>
>
>> +
>> +static void
>> +record_list_release (record_t * rec)
>
>                                  ^ no space after *, please.
>
>> +{
>> +  record_t *tmp;
>> +
>> +  if (!rec)
>> +    return;
>> +
>> +  while (rec->next)
>> +    {
>> +      rec = rec->next;
>> +    }
>> +
>> +  while (rec->prev)
>> +    {
>> +      tmp = rec;
>> +      rec = rec->prev;
>> +      if (tmp->type == record_reg)
>> +     xfree (tmp->u.reg.val);
>> +      else if (tmp->type == record_mem)
>> +     xfree (tmp->u.mem.val);
>> +      xfree (tmp);
>> +    }
>> +
>> +  if (rec != &record_first)
>> +    xfree (rec);
>> +}
>> +
>> +static void
>> +record_list_release_next (void)
>> +{
>> +  record_t *rec = record_list;
>> +  record_t *tmp = rec->next;
>> +  rec->next = NULL;
>> +  while (tmp)
>> +    {
>> +      rec = tmp->next;
>> +      if (tmp->type == record_reg)
>> +     record_insn_num--;
>> +      else if (tmp->type == record_reg)
>> +     xfree (tmp->u.reg.val);
>> +      else if (tmp->type == record_mem)
>> +     xfree (tmp->u.mem.val);
>> +      xfree (tmp);
>> +      tmp = rec;
>> +    }
>> +}
>> +
>> +static void
>> +record_list_release_first (void)
>> +{
>> +  record_t *tmp = NULL;
>> +  enum record_type type;
>> +
>> +  if (!record_first.next)
>> +    return;
>> +
>> +  while (1)
>> +    {
>> +      type = record_first.next->type;
>> +
>> +      if (type == record_reg)
>> +     xfree (record_first.next->u.reg.val);
>> +      else if (type == record_mem)
>> +     xfree (record_first.next->u.mem.val);
>> +      tmp = record_first.next;
>> +      record_first.next = tmp->next;
>> +      xfree (tmp);
>> +
>> +      if (!record_first.next)
>> +     {
>> +       gdb_assert (record_insn_num == 1);
>> +       break;
>> +     }
>> +
>> +      record_first.next->prev = &record_first;
>> +
>> +      if (type == record_end)
>> +     break;
>> +    }
>> +
>> +  record_insn_num--;
>> +}
>> +
>> +/* Add a record_t to record_arch_list.  */
>> +static void
>> +record_arch_list_add (record_t * rec)
>> +{
>> +  if (record_debug > 1)
>> +    fprintf_unfiltered (gdb_stdlog,
>> +                     "Process record: record_arch_list_add %s.\n",
>> +                     host_address_to_string (rec));
>> +
>> +  if (record_arch_list_tail)
>> +    {
>> +      record_arch_list_tail->next = rec;
>> +      rec->prev = record_arch_list_tail;
>> +      record_arch_list_tail = rec;
>> +    }
>> +  else
>> +    {
>> +      record_arch_list_head = rec;
>> +      record_arch_list_tail = rec;
>> +    }
>> +}
>> +
>> +/* Record the value of a register ("num") to record_arch_list.  */
>
> When we want to mention the value of a parameter, we mentions it in
> uppercase, like so:
>
> /* Record the value of register NUM to record_arch_list.  */
>
> Also, there should be an empty line between the comment and
> the function itself.  Here and elsewhere.
>
>> +int
>> +record_arch_list_add_reg (int num)
>> +{
>> +  record_t *rec;
>> +
>> +  if (record_debug > 1)
>> +    fprintf_unfiltered (gdb_stdlog,
>> +                     "Process record: add register num = %d to "
>> +                     "record list.\n",
>> +                     num);
>> +
>> +  rec = (record_t *) xmalloc (sizeof (record_t));
>> +  rec->u.reg.val = (gdb_byte *) xmalloc (MAX_REGISTER_SIZE);
>> +  rec->prev = NULL;
>> +  rec->next = NULL;
>> +  rec->type = record_reg;
>> +  rec->u.reg.num = num;
>> +
>> +  regcache_raw_read (record_regcache, num, rec->u.reg.val);
>> +
>> +  record_arch_list_add (rec);
>> +
>> +  return 0;
>> +}
>> +
>> +/* Record the value of a region of memory whose address is "addr" and
>> +   length is "len" to record_arch_list.  */
>
> /* Record the value of a region of memory whose address is ADDR and
>   length is LEN to record_arch_list.  */
>
>> +
>> +int
>> +record_arch_list_add_mem (CORE_ADDR addr, int len)
>> +{
>> +  record_t *rec;
>> +
>> +  if (record_debug > 1)
>> +    fprintf_unfiltered (gdb_stdlog,
>> +                     "Process record: add mem addr = 0x%s len = %d to "
>> +                     "record list.\n",
>> +                     paddr_nz (addr), len);
>> +
>
>> +  if (!addr)
>> +    return 0;
>
> Why is this check for addr == 0 necessary here?
>
>> +
>> +  rec = (record_t *) xmalloc (sizeof (record_t));
>> +  rec->u.mem.val = (gdb_byte *) xmalloc (len);
>> +  rec->prev = NULL;
>> +  rec->next = NULL;
>> +  rec->type = record_mem;
>> +  rec->u.mem.addr = addr;
>> +  rec->u.mem.len = len;
>> +
>> +  if (target_read_memory (addr, rec->u.mem.val, len))
>> +    {
>> +      if (record_debug)
>> +     fprintf_unfiltered (gdb_stdlog,
>> +                         "Process record: error reading memory at "
>> +                         "addr = 0x%s len = %d.\n",
>> +                         paddr_nz (addr), len);
>> +      xfree (rec->u.mem.val);
>> +      xfree (rec);
>> +      return -1;
>> +    }
>> +
>> +  record_arch_list_add (rec);
>> +
>> +  return 0;
>> +}
>> +
>> +/* Add a record_end type record_t to record_arch_list.  */
>> +int
>> +record_arch_list_add_end (int need_dasm)
>> +{
>> +  record_t *rec;
>> +
>> +  if (record_debug > 1)
>> +    fprintf_unfiltered (gdb_stdlog,
>> +                     "Process record: add end need_dasm = %d to "
>> +                     "arch list.\n",
>> +                     need_dasm);
>> +
>> +  rec = (record_t *) xmalloc (sizeof (record_t));
>> +  rec->prev = NULL;
>> +  rec->next = NULL;
>> +  rec->type = record_end;
>> +
>> +  rec->u.need_dasm = need_dasm;
>> +
>> +  record_arch_list_add (rec);
>> +
>> +  return 0;
>> +}
>> +
>> +static void
>> +record_check_insn_num (int set_terminal)
>> +{
>> +  if (record_insn_max_num)
>> +    {
>> +      gdb_assert (record_insn_num <= record_insn_max_num);
>> +      if (record_insn_num == record_insn_max_num)
>> +     {
>> +       /* Ask user what to do.  */
>> +       if (record_stop_at_limit)
>> +         {
>> +           int q;
>> +           if (set_terminal)
>> +             target_terminal_ours ();
>> +           q = yquery (_("Do you want to auto delete previous execution "
>> +                         "log entries when record/replay buffer becomes "
>> +                         "full (record-stop-at-limit)?"));
>> +           if (set_terminal)
>> +             target_terminal_inferior ();
>> +           if (q)
>> +             record_stop_at_limit = 0;
>> +           else
>> +             error (_("Process record: inferior program stopped."));
>> +         }
>> +     }
>> +    }
>> +}
>> +
>> +static void
>> +record_normal_stop (void)
>> +{
>> +  finish_thread_state (minus_one_ptid);
>> +
>> +  if (!breakpoints_always_inserted_mode () && target_has_execution)
>> +    remove_breakpoints ();
>> +
>> +  target_terminal_ours ();
>> +
>> +  if (target_has_stack && !stop_stack_dummy)
>> +    set_current_sal_from_frame (get_current_frame (), 1);
>> +
>> +  select_frame (get_current_frame ());
>> +
>> +  annotate_stopped ();
>> +  if (!suppress_stop_observer)
>> +    {
>> +      if (!ptid_equal (inferior_ptid, null_ptid))
>> +     observer_notify_normal_stop (inferior_thread ()->stop_bpstat, 0);
>> +      else
>> +     observer_notify_normal_stop (NULL, 0);
>> +    }
>> +}
>
> Nope sorry, this is worse than before.  When I asked to not call
> normal_stop, I didn't mean that you should inline chunks of
> it here.  This is papering over a different problem.  Why
> is it that record.c needs to do this, while other targets do not?
> If we need to do something like this, it should be needed by all
> targets that can throw from target_wait (which, is documented
> as something you shouldn't do anyway).  Actually, do you still need
> any of this in current head?
>
>> +
>> +/* Before inferior step (when GDB record the running message, inferior
>> +   only can step), GDB will call this function to record the values to
>> +   record_list.  This function will call gdbarch_process_record to
>> +   record the running message of inferior and set them to
>> +   record_arch_list, and add it to record_list.  */
>> +
>> +static void
>> +record_message_cleanups (void *ignore)
>> +{
>> +  record_list_release (record_arch_list_tail);
>> +  set_executing (inferior_ptid, 0);
>> +  record_normal_stop ();
>> +}
>
> Same as above.  This isn't a record.c problem.
>
>> +
>> +void
>> +record_message (struct gdbarch *gdbarch)
>> +{
>> +  int ret;
>> +  struct cleanup *old_cleanups = make_cleanup (record_message_cleanups, 0);
>> +
>> +  record_arch_list_head = NULL;
>> +  record_arch_list_tail = NULL;
>> +
>> +  /* Check record_insn_num.  */
>> +  record_check_insn_num (1);
>> +
>> +  record_regcache = get_current_regcache ();
>> +
>> +  ret = gdbarch_process_record (gdbarch,
>> +                             regcache_read_pc (record_regcache));
>> +  if (ret > 0)
>> +    error (_("Process record: inferior program stopped."));
>> +  if (ret < 0)
>> +    error (_("Process record: failed to record execution log."));
>> +
>> +  discard_cleanups (old_cleanups);
>> +
>> +  record_list->next = record_arch_list_head;
>> +  record_arch_list_head->prev = record_list;
>> +  record_list = record_arch_list_tail;
>> +
>> +  if (record_insn_num == record_insn_max_num && record_insn_max_num)
>> +    record_list_release_first ();
>> +  else
>> +    record_insn_num++;
>> +}
>> +
>
>
>> +/* Things to clean up if we error or QUIT out of function that set
>> +   record_gdb_operation_disable (ie. command that caused target_wait to
>> +   be called).  */
>> +static void
>> +record_gdb_operation_disable_cleanups (void *ignore)
>> +{
>> +  record_gdb_operation_disable = 0;
>> +}
>> +
>> +struct cleanup *
>> +record_gdb_operation_disable_set (void)
>> +{
>> +  struct cleanup *old_cleanups =
>> +    make_cleanup (record_gdb_operation_disable_cleanups, 0);
>> +  record_gdb_operation_disable = 1;
>> +
>> +  return old_cleanups;
>> +}
>
> This is make_cleanup_restore_integer.
>
>> +
>> +static void
>> +record_open (char *name, int from_tty)
>> +{
>> +  struct target_ops *t;
>> +
>> +  if (record_debug)
>> +    fprintf_unfiltered (gdb_stdlog, "Process record: record_open\n");
>> +
>> +  /* check exec */
>
> and core.
>
>> +  if (!target_has_execution)
>> +    error (_("Process record: the program is not being run."));
>> +  if (non_stop)
>> +    error (_("Process record target can't debug inferior in non-stop mode "
>> +          "(non-stop)."));
>> +  if (target_async_permitted)
>> +    error (_("Process record target can't debug inferior in asynchronous "
>> +          "mode (target-async)."));
>> +
>> +  if (!gdbarch_process_record_p (current_gdbarch))
>> +    error (_("Process record: the current architecture doesn't support "
>> +          "record function."));
>> +
>> +  /* Check if record target is already running.  */
>> +  if (TARGET_IS_PROCESS_RECORD)
>> +    {
>> +      if (!nquery
>> +       (_("Process record target already running, do you want to delete "
>> +          "the old record log?")))
>> +     return;
>> +    }
>> +
>> +  /* Set the beneath function pointers.  */
>> +  for (t = current_target.beneath; t != NULL; t = t->beneath)
>> +    {
>> +      if (!record_beneath_to_resume)
>> +        {
>> +       record_beneath_to_resume = t->to_resume;
>> +       record_beneath_to_resume_ops = t;
>> +        }
>> +      if (!record_beneath_to_wait)
>> +        {
>> +       record_beneath_to_wait = t->to_wait;
>> +       record_beneath_to_wait_ops = t;
>> +        }
>> +      if (!record_beneath_to_store_registers)
>> +        {
>> +       record_beneath_to_store_registers = t->to_store_registers;
>> +       record_beneath_to_store_registers_ops = t;
>> +        }
>> +      if (!record_beneath_to_xfer_partial)
>> +        {
>> +       record_beneath_to_xfer_partial = t->to_xfer_partial;
>> +       record_beneath_to_xfer_partial_ops = t;
>> +        }
>> +      if (!record_beneath_to_insert_breakpoint)
>> +     record_beneath_to_insert_breakpoint = t->to_insert_breakpoint;
>> +      if (!record_beneath_to_remove_breakpoint)
>> +     record_beneath_to_remove_breakpoint = t->to_remove_breakpoint;
>> +    }
>> +  if (!record_beneath_to_resume)
>> +    error (_("Process record can't get to_resume."));
>> +  if (!record_beneath_to_wait)
>> +    error (_("Process record can't get to_wait."));
>> +  if (!record_beneath_to_store_registers)
>> +    error (_("Process record can't get to_store_registers."));
>> +  if (!record_beneath_to_xfer_partial)
>> +    error (_("Process record can't get to_xfer_partial."));
>> +  if (!record_beneath_to_insert_breakpoint)
>> +    error (_("Process record can't get to_insert_breakpoint."));
>> +  if (!record_beneath_to_remove_breakpoint)
>> +    error (_("Process record can't get to_remove_breakpoint."));
>
> As I said above, this is better than making target.c aware of
> these pointers, but, it still isn't ideal.  For one thing, if
> some other layer/target is pushed after the record target is opened,
> these will be wrong.  I would prefer that this beneath lookup
> would be done at each callback implementation (record_resume, etc.),
> but, I'm happy enough with this for now.  It is now contained, so we can
> clean this up afterwards...
>
>> +
>> +  push_target (&record_ops);
>> +
>> +  /* Reset */
>> +  record_insn_num = 0;
>> +  record_list = &record_first;
>> +  record_list->next = NULL;
>> +}
>> +
>> +static void
>> +record_close (int quitting)
>> +{
>> +  if (record_debug)
>> +    fprintf_unfiltered (gdb_stdlog, "Process record: record_close\n");
>> +
>> +  record_list_release (record_list);
>> +}
>
> Shouldn't this clear the record_beneath_* pointers as well?
>
>> +
>> +static void
>> +record_resume (struct target_ops *ops, ptid_t ptid, int step,
>> +               enum target_signal siggnal)
>> +{
>> +  record_resume_step = step;
>> +  record_resume_siggnal = siggnal;
>> +
>> +  if (!RECORD_IS_REPLAY)
>> +    {
>> +      record_message (current_gdbarch);
>> +      record_beneath_to_resume (record_beneath_to_resume_ops, ptid, 1,
>> +                                siggnal);
>> +    }
>> +}
>> +
>> +static void
>> +record_sig_handler (int signo)
>> +{
>> +  if (record_debug)
>> +    fprintf_unfiltered (gdb_stdlog, "Process record: get a signal\n");
>> +
>> +  record_resume_step = 1;
>> +  record_get_sig = 1;
>> +}
>
> This handler is magical.  Why is it setting resume_step, for instance?
> It would definitelly benefic from some comments.  In fact, most of the
> file is undercommented.
>
>> +
>> +static void
>> +record_wait_cleanups (void *ignore)
>> +{
>> +  if (execution_direction == EXEC_REVERSE)
>> +    {
>> +      if (record_list->next)
>> +     record_list = record_list->next;
>> +    }
>> +  else
>> +    record_list = record_list->prev;
>> +
>> +  set_executing (inferior_ptid, 0);
>> +  record_normal_stop ();
>> +}
>
> See comments about record_normal_stop above.
>
>> +
>> +/* record_wait
>
> Please remove the function name from the comment.  It's redundant.
>
>> +   In replay mode, this function examines the recorded log and
>> +   determines where to stop.  */
>> +
>> +static ptid_t
>> +record_wait (struct target_ops *ops,
>> +              ptid_t ptid, struct target_waitstatus *status)
>> +{
>> +  struct cleanup *set_cleanups = record_gdb_operation_disable_set ();
>> +
>> +  if (record_debug)
>> +    fprintf_unfiltered (gdb_stdlog,
>> +                     "Process record: record_wait "
>> +                     "record_resume_step = %d\n",
>> +                     record_resume_step);
>> +
>> +  if (!RECORD_IS_REPLAY)
>> +    {
>> +      if (record_resume_step)
>> +     {
>> +       /* This is a single step.  */
>> +       return record_beneath_to_wait (record_beneath_to_wait_ops,
>> +                                                      ptid, status);
>> +     }
>> +      else
>> +     {
>> +       if (record_resume_step)
>> +         {
>> +           /* This is a single step.  */
>> +           return record_beneath_to_wait (record_beneath_to_wait_ops,
>> +                                                          ptid, status);
>> +         }
>> +       else
>> +         {
>> +           /* This is not a single step.  */
>> +           ptid_t ret;
>> +           int is_breakpoint = 1;
>> +           CORE_ADDR pc = 0;
>> +           int pc_is_read = 0;
>> +           struct bp_location *bl;
>> +           struct breakpoint *b;
>> +
>> +           do
>> +             {
>> +               ret = record_beneath_to_wait (record_beneath_to_wait_ops,
>> +                                                            ptid, status);
>> +
>> +               if (status->kind == TARGET_WAITKIND_STOPPED
>> +                   && status->value.sig == TARGET_SIGNAL_TRAP)
>> +                 {
>> +                   /* Check if there is a breakpoint.  */
>> +                   pc_is_read = 0;
>> +                   registers_changed ();
>> +                   for (bl = bp_location_chain; bl; bl = bl->global_next)
>
> This will need to be fixed.  Can you use the breakpoint_here-like functions
> exported by breakpoint.h instead of referencing bp_location_chain directly?
>
>> +                     {
>> +                       b = bl->owner;
>> +                       gdb_assert (b);
>> +                       if (b->enable_state != bp_enabled
>> +                           && b->enable_state != bp_permanent)
>> +                         continue;
>> +                       if (!pc_is_read)
>> +                         {
>> +                           pc =
>> +                             regcache_read_pc (get_thread_regcache (ret));
>> +                           pc_is_read = 1;
>> +                         }
>> +                       switch (b->type)
>> +                         {
>> +                         default:
>> +                           if (bl->address == pc)
>> +                             goto breakpoint;
>> +                           break;
>> +
>> +                         case bp_watchpoint:
>> +                           /* XXX teawater: I still not very clear how to
>> +                              deal with it.  */
>> +                           goto breakpoint;
>> +                           break;
>> +
>> +                         case bp_catchpoint:
>> +                           gdb_assert (b->ops != NULL
>> +                                       && b->ops->breakpoint_hit != NULL);
>> +                           if (b->ops->breakpoint_hit (b))
>> +                             goto breakpoint;
>> +                           break;
>> +
>> +                         case bp_hardware_watchpoint:
>> +                         case bp_read_watchpoint:
>> +                         case bp_access_watchpoint:
>> +                           if (STOPPED_BY_WATCHPOINT (0))
>> +                             goto breakpoint;
>> +                           break;
>> +                         }
>> +                     }
>> +
>> +                   /* There is not a breakpoint.  */
>> +                   record_message (current_gdbarch);
>> +                   record_beneath_to_resume (record_beneath_to_resume_ops,
>> +                                                ptid, 1,
>> +                                             record_resume_siggnal);
>> +                   continue;
>> +                 }
>> +
>> +               is_breakpoint = 0;
>> +
>> +             breakpoint:
>> +               /* Add gdbarch_decr_pc_after_break to pc because gdb will
>> +                  expect the pc to be at address plus decr_pc_after_break
>> +                  when the inferior stops at a breakpoint.  */
>> +               if (is_breakpoint)
>> +                 {
>> +                   CORE_ADDR decr_pc_after_break =
>> +                     gdbarch_decr_pc_after_break (current_gdbarch);
>> +                   if (decr_pc_after_break)
>> +                     {
>> +                       if (!pc_is_read)
>> +                         pc =
>> +                           regcache_read_pc (get_thread_regcache (ret));
>> +                       regcache_write_pc (get_thread_regcache (ret),
>> +                                          pc + decr_pc_after_break);
>> +                     }
>> +                 }
>> +
>> +               break;
>> +             }
>> +           while (1);
>> +
>> +           return ret;
>> +         }
>> +     }
>> +    }
>> +  else
>> +    {
>> +      int need_dasm = 0;
>> +      struct regcache *regcache = get_current_regcache ();
>> +      int continue_flag = 1;
>> +      int first_record_end = 1;
>> +      struct cleanup *old_cleanups = make_cleanup (record_wait_cleanups, 0);
>> +      CORE_ADDR tmp_pc;
>> +
>> +      status->kind = TARGET_WAITKIND_STOPPED;
>> +
>> +      /* Check breakpoint when forward execute.  */
>> +      if (execution_direction == EXEC_FORWARD)
>> +     {
>> +       tmp_pc = regcache_read_pc (regcache);
>> +       if (breakpoint_inserted_here_p (tmp_pc))
>> +         {
>> +           if (record_debug)
>> +             fprintf_unfiltered (gdb_stdlog,
>> +                                 "Process record: break at 0x%s.\n",
>> +                                 paddr_nz (tmp_pc));
>> +           if (gdbarch_decr_pc_after_break (get_regcache_arch (regcache))
>> +               && !record_resume_step)
>> +             regcache_write_pc (regcache,
>> +                                tmp_pc +
>> +                                gdbarch_decr_pc_after_break
>> +                                (get_regcache_arch (regcache)));
>> +           goto replay_out;
>> +         }
>> +     }
>> +
>> +      record_get_sig = 0;
>> +      signal (SIGINT, record_sig_handler);
>> +      /* If GDB is in terminal_inferior mode, it will not get the signal.
>> +         And in GDB replay mode, GDB doesn't need to be in terminal_inferior
>> +         mode, because inferior will not executed.
>> +         Then set it to terminal_ours to make GDB get the signal.  */
>> +      target_terminal_ours ();
>> +
>> +      /* In EXEC_FORWARD mode, record_list points to the tail of prev
>> +         instruction.  */
>> +      if (execution_direction == EXEC_FORWARD && record_list->next)
>> +     record_list = record_list->next;
>> +
>> +      /* Loop over the record_list, looking for the next place to
>> +      stop.  */
>> +      do
>> +     {
>> +       /* Check for beginning and end of log.  */
>> +       if (execution_direction == EXEC_REVERSE
>> +           && record_list == &record_first)
>> +         {
>> +           /* Hit beginning of record log in reverse.  */
>> +           status->kind = TARGET_WAITKIND_NO_HISTORY;
>> +           break;
>> +         }
>> +       if (execution_direction != EXEC_REVERSE && !record_list->next)
>> +         {
>> +           /* Hit end of record log going forward.  */
>> +           status->kind = TARGET_WAITKIND_NO_HISTORY;
>> +           break;
>> +         }
>> +
>> +       /* Set ptid, register and memory according to record_list.  */
>> +       if (record_list->type == record_reg)
>> +         {
>> +           /* reg */
>> +           gdb_byte reg[MAX_REGISTER_SIZE];
>> +           if (record_debug > 1)
>> +             fprintf_unfiltered (gdb_stdlog,
>> +                                 "Process record: record_reg %s to "
>> +                                 "inferior num = %d.\n",
>> +                                 host_address_to_string (record_list),
>> +                                 record_list->u.reg.num);
>> +           regcache_cooked_read (regcache, record_list->u.reg.num, reg);
>> +           regcache_cooked_write (regcache, record_list->u.reg.num,
>> +                                  record_list->u.reg.val);
>> +           memcpy (record_list->u.reg.val, reg, MAX_REGISTER_SIZE);
>> +         }
>> +       else if (record_list->type == record_mem)
>> +         {
>> +           /* mem */
>> +           gdb_byte *mem = alloca (record_list->u.mem.len);
>> +           if (record_debug > 1)
>> +             fprintf_unfiltered (gdb_stdlog,
>> +                                 "Process record: record_mem %s to "
>> +                                 "inferior addr = 0x%s len = %d.\n",
>> +                                 host_address_to_string (record_list),
>> +                                 paddr_nz (record_list->u.mem.addr),
>> +                                 record_list->u.mem.len);
>> +
>> +           if (target_read_memory
>> +               (record_list->u.mem.addr, mem, record_list->u.mem.len))
>> +             error (_("Process record: error reading memory at "
>> +                      "addr = 0x%s len = %d."),
>> +                    paddr_nz (record_list->u.mem.addr),
>> +                    record_list->u.mem.len);
>> +
>> +           if (target_write_memory
>> +               (record_list->u.mem.addr, record_list->u.mem.val,
>> +                record_list->u.mem.len))
>> +             error (_
>> +                    ("Process record: error writing memory at "
>> +                     "addr = 0x%s len = %d."),
>> +                    paddr_nz (record_list->u.mem.addr),
>> +                    record_list->u.mem.len);
>> +
>> +           memcpy (record_list->u.mem.val, mem, record_list->u.mem.len);
>> +         }
>> +       else
>> +         {
>> +           if (record_debug > 1)
>> +             fprintf_unfiltered (gdb_stdlog,
>> +                                 "Process record: record_end %s to "
>> +                                 "inferior need_dasm = %d.\n",
>> +                                 host_address_to_string (record_list),
>> +                                 record_list->u.need_dasm);
>> +
>> +           if (execution_direction == EXEC_FORWARD)
>> +             need_dasm = record_list->u.need_dasm;
>> +           if (need_dasm)
>> +             gdbarch_process_record_dasm (current_gdbarch);
>> +
>> +           if (first_record_end && execution_direction == EXEC_REVERSE)
>> +             {
>> +               /* When reverse excute, the first record_end is the part of
>> +                  current instruction.  */
>> +               first_record_end = 0;
>> +             }
>> +           else
>> +             {
>> +               /* In EXEC_REVERSE mode, this is the record_end of prev
>> +                  instruction.
>> +                  In EXEC_FORWARD mode, this is the record_end of current
>> +                  instruction.  */
>> +               /* step */
>> +               if (record_resume_step)
>> +                 {
>> +                   if (record_debug > 1)
>> +                     fprintf_unfiltered (gdb_stdlog,
>> +                                         "Process record: step.\n");
>> +                   continue_flag = 0;
>> +                 }
>> +
>> +               /* check breakpoint */
>> +               tmp_pc = regcache_read_pc (regcache);
>> +               if (breakpoint_inserted_here_p (tmp_pc))
>> +                 {
>> +                   if (record_debug)
>> +                     fprintf_unfiltered (gdb_stdlog,
>> +                                         "Process record: break "
>> +                                         "at 0x%s.\n",
>> +                                         paddr_nz (tmp_pc));
>> +                   if (gdbarch_decr_pc_after_break (get_regcache_arch (regcache))
>> +                       && execution_direction == EXEC_FORWARD
>> +                       && !record_resume_step)
>> +                     regcache_write_pc (regcache,
>> +                                        tmp_pc +
>> +                                        gdbarch_decr_pc_after_break
>> +                                        (get_regcache_arch (regcache)));
>> +                   continue_flag = 0;
>> +                 }
>> +             }
>> +           if (execution_direction == EXEC_REVERSE)
>> +             need_dasm = record_list->u.need_dasm;
>> +         }
>> +
>> +next:
>> +       if (continue_flag)
>> +         {
>> +           if (execution_direction == EXEC_REVERSE)
>> +             {
>> +               if (record_list->prev)
>> +                 record_list = record_list->prev;
>> +             }
>> +           else
>> +             {
>> +               if (record_list->next)
>> +                 record_list = record_list->next;
>> +             }
>> +         }
>> +     }
>> +      while (continue_flag);
>> +
>> +      signal (SIGINT, handle_sigint);
>> +
>> +replay_out:
>> +      if (record_get_sig)
>> +     status->value.sig = TARGET_SIGNAL_INT;
>> +      else
>> +     status->value.sig = TARGET_SIGNAL_TRAP;
>> +
>> +      discard_cleanups (old_cleanups);
>> +    }
>> +
>> +  do_cleanups (set_cleanups);
>> +  return inferior_ptid;
>> +}
>
> I have to say that I find that function confusing, due to
> the use of gotos, and excessive nesting.  Personally, I much prefer
> code that does:
>
>  if (foo)
>    {
>      /* something */
>      return;
>    }
>
>  if (bar)
>    {
>      /* something */
>      return;
>    }
>
>  if (lala)
>    {
>      /* something */
>      return;
>    }
>
> Over:
>
>  if (foo)
>    {
>      /* something */
>      return;
>    }
>  else
>   {
>     if (bar)
>       {
>         /* something */
>         return;
>       }
>     else
>      {
>         if (lala)
>           {
>             /* something */
>             return;
>           }
>      }
>   }
>
>
>> +
>> +static void
>> +record_disconnect (struct target_ops *target, char *args, int from_tty)
>> +{
>> +  if (record_debug)
>> +    fprintf_unfiltered (gdb_stdlog, "Process record: record_disconnect\n");
>> +
>> +  unpush_target (&record_ops);
>> +  target_disconnect (args, from_tty);
>> +}
>> +
>> +static void
>> +record_detach (struct target_ops *ops, char *args, int from_tty)
>> +{
>> +  if (record_debug)
>> +    fprintf_unfiltered (gdb_stdlog, "Process record: record_detach\n");
>> +
>> +  unpush_target (&record_ops);
>> +  target_detach (args, from_tty);
>> +}
>
> This trick you're using happens to work, but, could you try
> this instead?  Here and elsewhere similarly.
>
>  struct target_ops *beneath = find_target_beaneath (ops);
>  unpush_target (ops);
>  beneath->to_detach (args, from_tty);
>
>> +
>> +static void
>> +record_mourn_inferior (struct target_ops *ops)
>> +{
>> +  if (record_debug)
>> +    fprintf_unfiltered (gdb_stdlog, "Process record: "
>> +                                 "record_mourn_inferior\n");
>> +
>> +  unpush_target (&record_ops);
>> +  target_mourn_inferior ();
>> +}
>> +
>> +/* Close process record target before killing the inferior process.  */
>> +static void
>> +record_kill (void)
>> +{
>> +  if (record_debug)
>> +    fprintf_unfiltered (gdb_stdlog, "Process record: record_kill\n");
>> +
>> +  unpush_target (&record_ops);
>> +  target_kill ();
>> +}
>> +
>> +/* Record registers change (by user or by GDB) to list as an instruction.  */
>> +static void
>> +record_registers_change (struct regcache *regcache, int regnum)
>> +{
>> +  /* Check record_insn_num.  */
>> +  record_check_insn_num (0);
>> +
>> +  record_arch_list_head = NULL;
>> +  record_arch_list_tail = NULL;
>> +
>> +  record_regcache = get_current_regcache ();
>> +
>> +  if (regnum < 0)
>> +    {
>> +      int i;
>> +      for (i = 0; i < gdbarch_num_regs (get_regcache_arch (regcache)); i++)
>> +     {
>> +       if (record_arch_list_add_reg (i))
>> +         {
>> +           record_list_release (record_arch_list_tail);
>> +           error (_("Process record: failed to record execution log."));
>> +         }
>> +     }
>> +    }
>> +  else
>> +    {
>> +      if (record_arch_list_add_reg (regnum))
>> +     {
>> +       record_list_release (record_arch_list_tail);
>> +       error (_("Process record: failed to record execution log."));
>> +     }
>> +    }
>> +  if (record_arch_list_add_end (0))
>> +    {
>> +      record_list_release (record_arch_list_tail);
>> +      error (_("Process record: failed to record execution log."));
>> +    }
>> +  record_list->next = record_arch_list_head;
>> +  record_arch_list_head->prev = record_list;
>> +  record_list = record_arch_list_tail;
>> +
>> +  if (record_insn_num == record_insn_max_num && record_insn_max_num)
>> +    record_list_release_first ();
>> +  else
>> +    record_insn_num++;
>> +}
>> +
>> +static void
>> +record_store_registers (struct target_ops *ops, struct regcache *regcache,
>> +                        int regno)
>> +{
>> +  if (!record_gdb_operation_disable)
>> +    {
>> +      if (RECORD_IS_REPLAY)
>> +     {
>> +       int n;
>> +       struct cleanup *old_cleanups;
>> +
>> +       /* Let user choose if he wants to write register or not.  */
>> +       if (regno < 0)
>> +         n =
>> +           nquery (_("Because GDB is in replay mode, changing the "
>> +                     "value of a register will make the execution "
>> +                     "log unusable from this point onward.  "
>> +                     "Change all registers?"));
>> +       else
>> +         n =
>> +           nquery (_("Because GDB is in replay mode, changing the value "
>> +                     "of a register will make the execution log unusable "
>> +                     "from this point onward.  Change register %s?"),
>> +                   gdbarch_register_name (get_regcache_arch (regcache),
>> +                                            regno));
>> +
>> +       if (!n)
>> +         {
>> +           /* Invalidate the value of regcache that was set in function
>> +              "regcache_raw_write".  */
>> +           if (regno < 0)
>> +             {
>> +               int i;
>> +               for (i = 0;
>> +                    i < gdbarch_num_regs (get_regcache_arch (regcache));
>> +                    i++)
>> +                 regcache_invalidate (regcache, i);
>> +             }
>> +           else
>> +             regcache_invalidate (regcache, regno);
>> +
>> +           error (_("Process record canceled the operation."));
>> +         }
>> +
>> +       /* Destroy the record from here forward.  */
>> +       record_list_release_next ();
>> +     }
>> +
>> +      record_registers_change (regcache, regno);
>> +    }
>> +  record_beneath_to_store_registers (record_beneath_to_store_registers_ops,
>> +                                     regcache, regno);
>> +}
>> +
>> +/* record_xfer_partial -- behavior is conditional on RECORD_IS_REPLAY.
>> +   In replay mode, we cannot write memory unles we are willing to
>> +   invalidate the record/replay log from this point forward.  */
>> +
>> +static LONGEST
>> +record_xfer_partial (struct target_ops *ops, enum target_object object,
>> +                  const char *annex, gdb_byte * readbuf,
>> +                  const gdb_byte * writebuf, ULONGEST offset, LONGEST len)
>> +{
>> +  if (!record_gdb_operation_disable
>> +      && (object == TARGET_OBJECT_MEMORY
>> +       || object == TARGET_OBJECT_RAW_MEMORY) && writebuf)
>> +    {
>> +      if (RECORD_IS_REPLAY)
>> +     {
>> +       /* Let user choose if he wants to write memory or not.  */
>> +       if (!nquery (_("Because GDB is in replay mode, writing to memory "
>> +                      "will make the execution log unusable from this "
>> +                      "point onward.  Write memory at address 0x%s?"),
>> +                    paddr_nz (offset)))
>> +         return -1;
>> +
>> +       /* Destroy the record from here forward.  */
>> +       record_list_release_next ();
>> +     }
>> +
>> +      /* Check record_insn_num */
>> +      record_check_insn_num (0);
>> +
>> +      /* Record registers change to list as an instruction.  */
>> +      record_arch_list_head = NULL;
>> +      record_arch_list_tail = NULL;
>> +      if (record_arch_list_add_mem (offset, len))
>> +     {
>> +       record_list_release (record_arch_list_tail);
>> +       if (record_debug)
>> +         fprintf_unfiltered (gdb_stdlog,
>> +                             _("Process record: failed to record "
>> +                               "execution log."));
>> +       return -1;
>> +     }
>> +      if (record_arch_list_add_end (0))
>> +     {
>> +       record_list_release (record_arch_list_tail);
>> +       if (record_debug)
>> +         fprintf_unfiltered (gdb_stdlog,
>> +                             _("Process record: failed to record "
>> +                               "execution log."));
>> +       return -1;
>> +     }
>> +      record_list->next = record_arch_list_head;
>> +      record_arch_list_head->prev = record_list;
>> +      record_list = record_arch_list_tail;
>> +
>> +      if (record_insn_num == record_insn_max_num && record_insn_max_num)
>> +     record_list_release_first ();
>> +      else
>> +     record_insn_num++;
>> +    }
>> +
>> +  return record_beneath_to_xfer_partial (record_beneath_to_xfer_partial_ops,
>> +                                         object, annex, readbuf, writebuf,
>> +                                         offset, len);
>> +}
>> +
>> +/* record_insert_breakpoint
>> +   record_remove_breakpoint
>> +   Behavior is conditional on RECORD_IS_REPLAY.
>> +   We will not actually insert or remove breakpoints when replaying,
>> +   nor when recording.  */
>> +
>> +static int
>> +record_insert_breakpoint (struct bp_target_info *bp_tgt)
>> +{
>> +  if (!RECORD_IS_REPLAY)
>> +    {
>> +      struct cleanup *old_cleanups = record_gdb_operation_disable_set ();
>> +      int ret = record_beneath_to_insert_breakpoint (bp_tgt);
>> +
>> +      do_cleanups (old_cleanups);
>> +
>> +      return ret;
>> +    }
>> +
>> +  return 0;
>> +}
>> +
>> +static int
>> +record_remove_breakpoint (struct bp_target_info *bp_tgt)
>> +{
>> +  if (!RECORD_IS_REPLAY)
>> +    {
>> +      struct cleanup *old_cleanups = record_gdb_operation_disable_set ();
>> +      int ret = record_beneath_to_remove_breakpoint (bp_tgt);
>> +
>> +      do_cleanups (old_cleanups);
>> +
>> +      return ret;
>> +    }
>> +
>> +  return 0;
>> +}
>> +
>> +static int
>> +record_can_execute_reverse (void)
>> +{
>> +  return 1;
>> +}
>> +
>> +static void
>> +init_record_ops (void)
>> +{
>> +  record_ops.to_shortname = "record";
>> +  record_ops.to_longname = "Process record and replay target";
>> +  record_ops.to_doc =
>> +    "Log program while executing and replay execution from log.";
>> +  record_ops.to_open = record_open;
>> +  record_ops.to_close = record_close;
>> +  record_ops.to_resume = record_resume;
>> +  record_ops.to_wait = record_wait;
>> +  record_ops.to_disconnect = record_disconnect;
>> +  record_ops.to_detach = record_detach;
>> +  record_ops.to_mourn_inferior = record_mourn_inferior;
>> +  record_ops.to_kill = record_kill;
>> +  record_ops.to_create_inferior = find_default_create_inferior;
>> +  record_ops.to_store_registers = record_store_registers;
>> +  record_ops.to_xfer_partial = record_xfer_partial;
>> +  record_ops.to_insert_breakpoint = record_insert_breakpoint;
>> +  record_ops.to_remove_breakpoint = record_remove_breakpoint;
>> +  record_ops.to_can_execute_reverse = record_can_execute_reverse;
>> +  record_ops.to_stratum = record_stratum;
>> +  record_ops.to_magic = OPS_MAGIC;
>> +}
>> +
>> +static void
>> +show_record_debug (struct ui_file *file, int from_tty,
>> +                struct cmd_list_element *c, const char *value)
>> +{
>> +  fprintf_filtered (file, _("Debugging of process record target is %s.\n"),
>> +                 value);
>> +}
>> +
>> +/* cmd_record_start -- alias for "target record".  */
>> +
>> +static void
>> +cmd_record_start (char *args, int from_tty)
>> +{
>> +  execute_command ("target record", from_tty);
>> +}
>> +
>> +/* cmd_record_delete -- truncate the record log from the present point
>> +   of replay until the end.  */
>> +
>> +static void
>> +cmd_record_delete (char *args, int from_tty)
>> +{
>> +  if (TARGET_IS_PROCESS_RECORD)
>> +    {
>> +      if (RECORD_IS_REPLAY)
>> +     {
>> +       if (!from_tty || query (_("Delete the log from this point forward "
>> +                                 "and begin to record the running message "
>> +                                 "at current PC?")))
>> +         record_list_release_next ();
>> +     }
>> +      else
>> +       printf_unfiltered (_("Already at end of record list.\n"));
>> +
>> +    }
>> +  else
>> +    printf_unfiltered (_("Process record is not started.\n"));
>> +}
>> +
>> +/* cmd_record_stop -- implement the "stoprecord" command.  */
>> +
>> +static void
>> +cmd_record_stop (char *args, int from_tty)
>> +{
>> +  if (TARGET_IS_PROCESS_RECORD)
>> +    {
>> +      if (!record_list || !from_tty || query (_("Delete recorded log and "
>> +                                             "stop recording?")))
>> +     unpush_target (&record_ops);
>> +    }
>> +  else
>> +    printf_unfiltered (_("Process record is not started.\n"));
>> +}
>> +
>> +/* set_record_insn_max_num -- set upper limit of record log size.  */
>> +
>> +static void
>> +set_record_insn_max_num (char *args, int from_tty, struct cmd_list_element *c)
>> +{
>> +  if (record_insn_num > record_insn_max_num && record_insn_max_num)
>> +    {
>> +      printf_unfiltered (_("Record instructions number is bigger than "
>> +                        "record instructions max number.  Auto delete "
>> +                        "the first ones?\n"));
>> +
>> +      while (record_insn_num > record_insn_max_num)
>> +     record_list_release_first ();
>> +    }
>> +}
>> +
>> +/* show_record_insn_number -- print the current index
>> +   into the record log (number of insns recorded so far).  */
>> +
>> +static void
>> +show_record_insn_number (char *ignore, int from_tty)
>> +{
>> +  printf_unfiltered (_("Record instruction number is %d.\n"),
>> +                  record_insn_num);
>> +}
>> +
>> +void
>> +_initialize_record (void)
>> +{
>> +  /* Init record_maskall.  */
>> +  if (sigfillset (&record_maskall) == -1)
>> +    perror_with_name (_("Process record: sigfillset failed"));
>
> This will not build on all hosts.  Is it still needed?  I can't
> find any other reference to this variable in this patch.
>
>> +
>> +  /* Init record_first.  */
>> +  record_first.prev = NULL;
>> +  record_first.next = NULL;
>> +  record_first.type = record_end;
>> +  record_first.u.need_dasm = 0;
>> +
>> +  init_record_ops ();
>> +  add_target (&record_ops);
>> +
>> +  add_setshow_zinteger_cmd ("record", no_class, &record_debug,
>> +                         _("Set debugging of record/replay feature."),
>> +                         _("Show debugging of record/replay feature."),
>> +                         _("When enabled, debugging output for "
>> +                           "record/replay feature is displayed."),
>> +                         NULL, show_record_debug, &setdebuglist,
>> +                         &showdebuglist);
>> +
>> +  add_com ("record", class_obscure, cmd_record_start,
>> +        _("Abbreviated form of \"target record\" command."));
>> +
>> +  add_com_alias ("rec", "record", class_obscure, 1);
>> +
>> +  /* XXX: I try to use some simple commands such as "disconnect" and
>> +     "detach" to support this functions.  But these commands all have
>> +     other affect to GDB such as call function "no_shared_libraries".
>> +     So I add special commands to GDB.  */
>> +  add_com ("delrecord", class_obscure, cmd_record_delete,
>> +        _("Delete the rest of execution log and start recording it anew."));
>> +  add_com_alias ("dr", "delrecord", class_obscure, 1);
>> +  add_com ("stoprecord", class_obscure, cmd_record_stop,
>> +        _("Stop the record/replay target."));
>> +  add_com_alias ("sr", "stoprecord", class_obscure, 1);
>> +
>> +  /* Record instructions number limit command.  */
>> +  add_setshow_boolean_cmd ("record-stop-at-limit", no_class,
>> +                         &record_stop_at_limit,
>> +                         _("Set whether record/replay stop when "
>> +                           "record/replay buffer becomes full."),
>> +                         _("Show whether record/replay stop when "
>> +                           "record/replay buffer becomes full."),
>> +                         _("Enable is default value.\n"
>> +                           "When enabled, if the record/replay buffer "
>> +                           "becomes full,\n"
>> +                              "ask user what to do.\n"
>> +                              "When disabled, if the record/replay buffer "
>> +                           "becomes full,\n"
>> +                              "delete it and start new recording."),
>> +                         NULL, NULL, &setlist, &showlist);
>> +  add_setshow_zinteger_cmd ("record-insn-number-max", no_class,
>> +                         &record_insn_max_num,
>> +                         _("Set record/replay buffer limit."),
>> +                         _("Show record/replay buffer limit."),
>> +                         _("Set the maximum number of instructions to be "
>> +                              "stored in the\n"
>> +                              "record/replay buffer.  "
>> +                              "Zero means unlimited (default 200000)."),
>> +                         set_record_insn_max_num,
>> +                         NULL, &setlist, &showlist);
>> +  add_info ("record-insn-number", show_record_insn_number,
>> +         _("Show the current number of instructions in the "
>> +           "record/replay buffer."));
>> +}
>> Index: src/gdb/record.h
>> ===================================================================
>> --- /dev/null 1970-01-01 00:00:00.000000000 +0000
>> +++ src/gdb/record.h  2009-02-28 20:23:20.000000000 +0000
>> @@ -0,0 +1,87 @@
>> +/* Process record and replay target for GDB, the GNU debugger.
>> +
>> +   Copyright (C) 2008 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 3 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, see <http://www.gnu.org/licenses/>.  */
>> +
>> +#ifndef _RECORD_H_
>> +#define _RECORD_H_
>> +
>> +#define TARGET_IS_PROCESS_RECORD   \
>> +     (current_target.beneath == &record_ops)
>
> Sorry, but I repeat the request I've made several times already.  This is
> not the right way to do this.  You need to add a new target_ops method or
> property that the core of GDB checks on.  It is not correct that make
> the core of GDB reference record_ops directly.  To come up with
> the target callback's name, at each call site of TARGET_IS_PROCESS_RECORD,
> consider what is the property of the current target that GDB needs to
> know about the current target.  Is it something like:
>
>  target_is_recording () ?
>  target_is_replaying () ?
>  target_is_read_only () ?
>
> etc.
>
>> +#define RECORD_IS_REPLAY \
>> +     (record_list->next || execution_direction == EXEC_REVERSE)
>
> AFAICS, this macro is not used outside of record.c.  It should move
> there, along with anything that isn't used outside of record.c.
>
>> +
>> +typedef struct record_reg_s
>> +{
>> +  int num;
>> +  gdb_byte *val;
>> +} record_reg_t;
>> +
>> +typedef struct record_mem_s
>> +{
>> +  CORE_ADDR addr;
>> +  int len;
>> +  gdb_byte *val;
>> +} record_mem_t;
>> +
>> +enum record_type
>> +{
>> +  record_end = 0,
>> +  record_reg,
>> +  record_mem
>> +};
>> +
>> +/* This is the core struct of record function.
>> +
>> +   An entity of record_t is a record of the value change of a register
>> +   ("record_reg") or a part of memory ("record_mem").  And each
>> +   instruction must has a record_t ("record_end") that points out this
>> +   is the last record_t of this instruction.
>> +
>> +   Each record_t is linked to "record_list" by "prev" and "next".
>> + */
>> +typedef struct record_s
>> +{
>> +  struct record_s *prev;
>> +  struct record_s *next;
>> +  enum record_type type;
>> +  union
>> +  {
>> +    /* reg */
>> +    record_reg_t reg;
>> +    /* mem */
>> +    record_mem_t mem;
>> +    /* end */
>> +    int need_dasm;
>> +  } u;
>> +} record_t;
>> +
>> +extern int record_debug;
>> +extern record_t *record_list;
>> +extern record_t *record_arch_list_head;
>> +extern record_t *record_arch_list_tail;
>> +extern struct regcache *record_regcache;
>
> Most of these things don't appear to be used anywhere else other
> than in record.c.  Can you remove these declarations from the
> public header, and make them static in record.c?
>
>> +
>> +extern struct target_ops record_ops;
>
> Once you get rid of TARGET_IS_PROCESS_RECORD this doesn't
> need to be public anymore.
>
>> +
>> +extern int record_arch_list_add_reg (int num);
>> +extern int record_arch_list_add_mem (CORE_ADDR addr, int len);
>> +extern int record_arch_list_add_end (int need_dasm);
>> +extern void record_message (struct gdbarch *gdbarch);
>> +extern struct cleanup * record_gdb_operation_disable_set (void);
>> +
>> +#endif /* _RECORD_H_ */
>>
>
> --
> Pedro Alves
>