From: Michael Snyder <msnyder@vmware.com>
To: "gdb-patches@sourceware.org" <gdb-patches@sourceware.org>
Cc: Daniel Jacobowitz <drow@false.org>,
Pedro Alves <pedro@codesourcery.com>,
teawater <teawater@gmail.com>
Subject: [RFA] Reverse Debugging, 3/5
Date: Wed, 01 Oct 2008 19:20:00 -0000 [thread overview]
Message-ID: <48E3CD0B.8020003@vmware.com> (raw)
[-- Attachment #1: Type: text/plain, Size: 0 bytes --]
[-- Attachment #2: infrun.txt --]
[-- Type: text/plain, Size: 7997 bytes --]
This patch (#3) adds to the infrun event loop so that it
knows what to do differently if execution-direction is
reverse. This effectively makes reverse step, next,
and continue work properly, except that there is no way
for them to be invoked (that comes in a later patch).
This patch can go in on top of #1 and #2 without any
user-visible changes. Tested under RHEL native with
no change in results.
2008-09-30 Michael Snyder <msnyder@vmware.com>
Event handling interface for reverse debugging.
* infrun.c (enum inferior_stop_reason): Add NO_HISTORY reason.
(handle_inferior_event): Handle TARGET_WAITKIND_NO_HISTORY.
Handle stepping over a function call in reverse.
Handle stepping thru a line range in reverse.
Handle setting a step-resume breakpoint in reverse.
Handle stepping into a function in reverse.
Handle stepping between line ranges in reverse.
(print_stop_reason): Print reason for NO_HISTORY.
Index: infrun.c
===================================================================
RCS file: /cvs/src/src/gdb/infrun.c,v
retrieving revision 1.322
retrieving revision 1.322.2.2
diff -u -p -r1.322 -r1.322.2.2
--- infrun.c 22 Sep 2008 15:26:53 -0000 1.322
+++ infrun.c 30 Sep 2008 23:50:51 -0000 1.322.2.2
@@ -1193,11 +1193,17 @@ proceed (CORE_ADDR addr, enum target_sig
if (addr == (CORE_ADDR) -1)
{
- if (pc == stop_pc && breakpoint_here_p (pc))
+ if (pc == stop_pc && breakpoint_here_p (pc)
+ && target_get_execution_direction () != EXEC_REVERSE)
/* There is a breakpoint at the address we will resume at,
step one instruction before inserting breakpoints so that
we do not stop right away (and report a second hit at this
- breakpoint). */
+ breakpoint).
+
+ Note, we don't do this in reverse, because we won't
+ actually be executing the breakpoint insn anyway.
+ We'll be (un-)executing the previous instruction. */
+
oneproc = 1;
else if (gdbarch_single_step_through_delay_p (gdbarch)
&& gdbarch_single_step_through_delay (gdbarch,
@@ -1426,7 +1432,9 @@ enum inferior_stop_reason
/* Inferior exited. */
EXITED,
/* Inferior received signal, and user asked to be notified. */
- SIGNAL_RECEIVED
+ SIGNAL_RECEIVED,
+ /* Reverse execution -- target ran out of history info. */
+ NO_HISTORY
};
/* The PTID we'll do a target_wait on.*/
@@ -2141,6 +2149,12 @@ handle_inferior_event (struct execution_
ecs->event_thread->stop_signal = ecs->ws.value.sig;
break;
+ case TARGET_WAITKIND_NO_HISTORY:
+ /* Reverse execution: target ran out of history info. */
+ print_stop_reason (NO_HISTORY, 0);
+ stop_stepping (ecs);
+ return;
+
/* We had an event in the inferior, but we are not interested
in handling it at this level. The lower layers have already
done what needs to be done, if anything.
@@ -2861,6 +2875,17 @@ infrun: BPSTAT_WHAT_SET_LONGJMP_RESUME (
keep_going (ecs);
return;
}
+ if (stop_pc == ecs->stop_func_start &&
+ target_get_execution_direction () == EXEC_REVERSE)
+ {
+ /* We are stepping over a function call in reverse, and
+ just hit the step-resume breakpoint at the start
+ address of the function. Go back to single-stepping,
+ which should take us back to the function call. */
+ ecs->event_thread->stepping_over_breakpoint = 1;
+ keep_going (ecs);
+ return;
+ }
break;
case BPSTAT_WHAT_CHECK_SHLIBS:
@@ -3026,10 +3051,25 @@ infrun: BPSTAT_WHAT_SET_LONGJMP_RESUME (
&& stop_pc < ecs->event_thread->step_range_end)
{
if (debug_infrun)
- fprintf_unfiltered (gdb_stdlog, "infrun: stepping inside range [0x%s-0x%s]\n",
+ fprintf_unfiltered (gdb_stdlog, "infrun: stepping inside range [0x%s-0x%s]\n",
paddr_nz (ecs->event_thread->step_range_start),
paddr_nz (ecs->event_thread->step_range_end));
- keep_going (ecs);
+
+ /* When stepping backward, stop at beginning of line range
+ (unles it's the function entry point, in which case
+ keep going back to the call point). */
+ if (stop_pc == ecs->event_thread->step_range_start &&
+ stop_pc != ecs->stop_func_start &&
+ target_get_execution_direction () == EXEC_REVERSE)
+ {
+ ecs->event_thread->stop_step = 1;
+ print_stop_reason (END_STEPPING_RANGE, 0);
+ stop_stepping (ecs);
+ }
+ else
+ {
+ keep_going (ecs);
+ }
return;
}
@@ -3116,10 +3156,28 @@ infrun: BPSTAT_WHAT_SET_LONGJMP_RESUME (
if (ecs->event_thread->step_over_calls == STEP_OVER_ALL)
{
- /* We're doing a "next", set a breakpoint at callee's return
- address (the address at which the caller will
- resume). */
- insert_step_resume_breakpoint_at_caller (get_current_frame ());
+ /* We're doing a "next".
+
+ Normal (forward) execution: set a breakpoint at the
+ callee's return address (the address at which the caller
+ will resume).
+
+ Reverse (backward) execution. set the step-resume
+ breakpoint at the start of the function that we just
+ stepped into (backwards), and continue to there. When we
+ get there, we'll need to single-step back to the
+ caller. */
+
+ if (target_get_execution_direction () == EXEC_REVERSE)
+ {
+ struct symtab_and_line sr_sal;
+ init_sal (&sr_sal);
+ sr_sal.pc = ecs->stop_func_start;
+ insert_step_resume_breakpoint_at_sal (sr_sal, null_frame_id);
+ }
+ else
+ insert_step_resume_breakpoint_at_caller (get_current_frame ());
+
keep_going (ecs);
return;
}
@@ -3176,9 +3234,21 @@ infrun: BPSTAT_WHAT_SET_LONGJMP_RESUME (
return;
}
- /* Set a breakpoint at callee's return address (the address at
- which the caller will resume). */
- insert_step_resume_breakpoint_at_caller (get_current_frame ());
+ if (target_get_execution_direction () == EXEC_REVERSE)
+ {
+ /* Set a breakpoint at callee's start address.
+ From there we can step once and be back in the caller. */
+ struct symtab_and_line sr_sal;
+ init_sal (&sr_sal);
+ sr_sal.pc = ecs->stop_func_start;
+ insert_step_resume_breakpoint_at_sal (sr_sal, null_frame_id);
+ }
+ else
+ {
+ /* Set a breakpoint at callee's return address (the address
+ at which the caller will resume). */
+ insert_step_resume_breakpoint_at_caller (get_current_frame ());
+ }
keep_going (ecs);
return;
}
@@ -3344,6 +3414,28 @@ step_into_function (struct execution_con
ecs->stop_func_start = gdbarch_skip_prologue
(current_gdbarch, ecs->stop_func_start);
+ if (target_get_execution_direction () == EXEC_REVERSE)
+ {
+ stop_func_sal = find_pc_line (stop_pc, 0);
+
+ /* OK, we're just gonna keep stepping here. */
+ if (stop_func_sal.pc == stop_pc)
+ {
+ /* We're there already. Just stop stepping now. */
+ ecs->event_thread->stop_step = 1;
+ print_stop_reason (END_STEPPING_RANGE, 0);
+ stop_stepping (ecs);
+ return;
+ }
+ /* Else just reset the step range and keep going.
+ No step-resume breakpoint, they don't work for
+ epilogues, which can have multiple entry paths. */
+ ecs->event_thread->step_range_start = stop_func_sal.pc;
+ ecs->event_thread->step_range_end = stop_func_sal.end;
+ keep_going (ecs);
+ return;
+ }
+ /* else... */
stop_func_sal = find_pc_line (ecs->stop_func_start, 0);
/* Use the step_resume_break to step until the end of the prologue,
even if that involves jumps (as it seems to on the vax under
@@ -3712,6 +3804,10 @@ print_stop_reason (enum inferior_stop_re
annotate_signal_string_end ();
ui_out_text (uiout, ".\n");
break;
+ case NO_HISTORY:
+ /* Reverse execution: target ran out of history info. */
+ ui_out_text (uiout, "\nNo more reverse-execution history.\n");
+ break;
default:
internal_error (__FILE__, __LINE__,
_("print_stop_reason: unrecognized enum value"));
next reply other threads:[~2008-10-01 19:20 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-10-01 19:20 Michael Snyder [this message]
2008-10-06 20:09 ` Pedro Alves
2008-10-06 20:54 ` Michael Snyder
2008-10-06 21:14 ` Pedro Alves
2008-10-06 21:22 ` Michael Snyder
2008-10-06 21:26 ` Joel Brobecker
2008-10-06 21:47 ` Michael Snyder
2008-10-06 21:22 ` Joel Brobecker
[not found] ` <48EA83AD.9040004@vmware.com>
2008-10-06 21:43 ` Joel Brobecker
2008-10-06 23:46 ` Michael Snyder
2008-10-07 4:07 ` Joel Brobecker
2008-10-07 18:46 ` Michael Snyder
2008-10-08 0:31 ` Joel Brobecker
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=48E3CD0B.8020003@vmware.com \
--to=msnyder@vmware.com \
--cc=drow@false.org \
--cc=gdb-patches@sourceware.org \
--cc=pedro@codesourcery.com \
--cc=teawater@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox