From: Tom Tromey <tromey@redhat.com>
To: Jan Kratochvil <jan.kratochvil@redhat.com>
Cc: gdb-patches@sources.redhat.com
Subject: Re: RFC: next/finish/etc -vs- exceptions
Date: Fri, 10 Dec 2010 20:07:00 -0000 [thread overview]
Message-ID: <m3ipz1xuhl.fsf@fleche.redhat.com> (raw)
In-Reply-To: <20101210045230.GA29751@host0.dyn.jankratochvil.net> (Jan Kratochvil's message of "Fri, 10 Dec 2010 05:52:30 +0100")
>>>>> "Jan" == Jan Kratochvil <jan.kratochvil@redhat.com> writes:
Jan> The new field `tp->control.exception_resume_breakpoint' should be
Jan> probably handled by existing save_infcall_control_state,
Jan> restore_infcall_control_state and discard_infcall_control_state
Jan> somehow like step_resume_breakpoint is.
Thanks for the test case.
Here is what I am testing.
It just mirrors the existing step_resume_breakpoint code.
I think that this is the right thing to do -- stash any
exception-related breakpoint while doing the infcall, then restore it
if the infcall succeeds.
Tom
diff --git a/gdb/infrun.c b/gdb/infrun.c
index 1bc00a4..f72a18d 100644
--- a/gdb/infrun.c
+++ b/gdb/infrun.c
@@ -6379,6 +6379,7 @@ save_infcall_control_state (void)
inf_status->inferior_control = inf->control;
tp->control.step_resume_breakpoint = NULL;
+ tp->control.exception_resume_breakpoint = NULL;
/* Save original bpstat chain to INF_STATUS; replace it in TP with copy of
chain. If caller's caller is walking the chain, they'll be happier if we
@@ -6428,6 +6429,10 @@ restore_infcall_control_state (struct infcall_control_state *inf_status)
if (tp->control.step_resume_breakpoint)
tp->control.step_resume_breakpoint->disposition = disp_del_at_next_stop;
+ if (tp->control.exception_resume_breakpoint)
+ tp->control.exception_resume_breakpoint->disposition
+ = disp_del_at_next_stop;
+
/* Handle the bpstat_copy of the chain. */
bpstat_clear (&tp->control.stop_bpstat);
@@ -6476,6 +6481,10 @@ discard_infcall_control_state (struct infcall_control_state *inf_status)
inf_status->thread_control.step_resume_breakpoint->disposition
= disp_del_at_next_stop;
+ if (inf_status->thread_control.exception_resume_breakpoint)
+ inf_status->thread_control.exception_resume_breakpoint->disposition
+ = disp_del_at_next_stop;
+
/* See save_infcall_control_state for info on stop_bpstat. */
bpstat_clear (&inf_status->thread_control.stop_bpstat);
diff --git a/gdb/testsuite/gdb.cp/nextoverthrow.cc b/gdb/testsuite/gdb.cp/nextoverthrow.cc
index b25cb34..c3da516 100644
--- a/gdb/testsuite/gdb.cp/nextoverthrow.cc
+++ b/gdb/testsuite/gdb.cp/nextoverthrow.cc
@@ -19,8 +19,9 @@
using namespace std;
-void dummy ()
+int dummy ()
{
+ return 0;
}
class NextOverThrowDerivates
@@ -89,6 +90,18 @@ public:
function1 (val); // until here
}
+ void resumebpt (int val)
+ {
+ try
+ {
+ throw val;
+ }
+ catch (int x)
+ {
+ dummy ();
+ }
+ }
+
};
NextOverThrowDerivates next_cases;
@@ -198,6 +211,18 @@ int main ()
testval = val; // End: advance
}
+ // Test of "resumebpt".
+ try
+ {
+ next_cases.resumebpt (10); // Start: resumebpt
+ next_cases.resumebpt (11); // Second: resumebpt
+ }
+ catch (int val)
+ {
+ dummy ();
+ testval = val; // End: resumebpt
+ }
+
testval = 32; // done
}
diff --git a/gdb/testsuite/gdb.cp/nextoverthrow.exp b/gdb/testsuite/gdb.cp/nextoverthrow.exp
index 960ea0d..c19674f 100644
--- a/gdb/testsuite/gdb.cp/nextoverthrow.exp
+++ b/gdb/testsuite/gdb.cp/nextoverthrow.exp
@@ -149,5 +149,14 @@ gdb_test "advance $line" ".*" "advance-over-throw"
tbreak_and_cont "End: advance"
verify_testval "pre-check - advance" 8
+tbreak_and_cont "Start: resumebpt"
+gdb_test "tbreak _Unwind_RaiseException"
+gdb_test "continue" "Temporary breakpoint.*" "continuing to _Unwind_RaiseException"
+gdb_test "finish" "Run till exit .*"
+gdb_test {set $retpc=$pc}
+gdb_test {break *$retpc if dummy ()}
+tbreak_and_cont "Second: resumebpt"
+gdb_test "next"
+
tbreak_and_cont "done"
verify_testval "post-check - advance" 9
next prev parent reply other threads:[~2010-12-10 20:07 UTC|newest]
Thread overview: 40+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-10-07 1:37 Tom Tromey
2010-11-24 17:53 ` Joel Brobecker
2010-11-24 18:24 ` Tom Tromey
2010-11-25 7:59 ` Jan Kratochvil
2010-11-27 17:25 ` Doug Evans
2010-11-28 8:29 ` Joel Brobecker
2010-11-30 16:43 ` Tom Tromey
2010-11-30 17:02 ` Jan Kratochvil
2010-11-30 17:15 ` Phil Muldoon
2010-11-30 20:15 ` Tom Tromey
2010-12-01 13:42 ` Jan Kratochvil
2010-12-01 21:40 ` Tom Tromey
2010-11-30 18:23 ` Tom Tromey
2010-11-30 18:55 ` Tom Tromey
2010-12-02 15:32 ` Tom Tromey
2010-12-09 16:37 ` Tom Tromey
2010-12-10 4:52 ` Jan Kratochvil
2010-12-10 20:07 ` Tom Tromey [this message]
2010-12-11 5:27 ` Jan Kratochvil
2010-12-15 21:18 ` Tom Tromey
-- strict thread matches above, loose matches on Subject: below --
2009-05-29 22:32 Tom Tromey
2009-05-30 21:08 ` Doug Evans
2009-05-30 23:18 ` Tom Tromey
2009-06-10 16:13 ` Joel Brobecker
2009-06-10 16:50 ` Tom Tromey
2009-06-10 17:05 ` Pedro Alves
2009-06-10 17:13 ` Daniel Jacobowitz
2009-06-10 17:47 ` Pedro Alves
2009-06-10 18:32 ` Tom Tromey
2009-06-10 18:49 ` Pedro Alves
2009-06-12 20:49 ` Tom Tromey
2009-06-12 21:51 ` Pedro Alves
2009-06-12 22:07 ` Pedro Alves
2010-11-25 4:54 ` Jan Kratochvil
2009-06-11 14:45 ` Joel Brobecker
2009-06-11 15:47 ` Tom Tromey
2009-07-24 17:38 ` Tom Tromey
2009-07-24 18:54 ` Pedro Alves
2009-07-24 19:18 ` Tom Tromey
2009-09-09 18:56 ` Tom Tromey
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=m3ipz1xuhl.fsf@fleche.redhat.com \
--to=tromey@redhat.com \
--cc=gdb-patches@sources.redhat.com \
--cc=jan.kratochvil@redhat.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