Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: Yao Qi <qiyaoltc@gmail.com>
To: gdb-patches@sourceware.org
Subject: [PATCH 2/8] Delete reinsert breakpoints from forked child
Date: Fri, 20 May 2016 15:13:00 -0000	[thread overview]
Message-ID: <1463757161-25850-3-git-send-email-yao.qi@linaro.org> (raw)
In-Reply-To: <1463757161-25850-1-git-send-email-yao.qi@linaro.org>

When a thread is stepping over a syscall instruction with software
single step, GDBserver inserts reinsert breakpoints at the next pcs.
If the syscall call is fork, the forked child has reinsert breakpoint
in its space, and GDBserver clones parent's breakpoint list to child's.
When GDBserver resumes the child, it hits the reinsert breakpoint.  Since
it is a GDBserver internal breakpoint, GDBserver will do step-over again,
and remove this reinsert breakpoint on step-over is finished, so the
reinsert breakpoint left in the child process doesn't make any trouble.
However, GDBserver still need to remove the reinsert breakpoints from
the child, in order to avoid the unnecessary breakpoint hit and
step-over.

gdb/gdbserver:

2016-05-20  Yao Qi  <yao.qi@linaro.org>

	* linux-low.c (handle_extended_wait): If the parent is doing
	step-over, remove the reinsert breakpoints from the child.
---
 gdb/gdbserver/linux-low.c | 19 +++++++++++++++++++
 1 file changed, 19 insertions(+)

diff --git a/gdb/gdbserver/linux-low.c b/gdb/gdbserver/linux-low.c
index 8e8f710..a63cc7a 100644
--- a/gdb/gdbserver/linux-low.c
+++ b/gdb/gdbserver/linux-low.c
@@ -568,6 +568,25 @@ handle_extended_wait (struct lwp_info **orig_event_lwp, int wstat)
 	  event_lwp->status_pending_p = 1;
 	  event_lwp->status_pending = wstat;
 
+	  /* If the parent thread is doing step-over with reinsert
+	     breakpoints, the reinsert breakpoints are still in child's
+	     process space and cloned to its breakpoint list from the
+	     parent's.  Remove them.  */
+	  if (event_lwp->bp_reinsert != 0 && can_software_single_step ())
+	    {
+	      struct thread_info *saved_thread = current_thread;
+
+	      /* The child process is forked and stopped, so it is safe
+		 to access its memory without stopping all other threads
+		 from other processes.  */
+	      current_thread = child_thr;
+	      delete_reinsert_breakpoints ();
+	      current_thread = saved_thread;
+
+	      gdb_assert (has_reinsert_breakpoints (parent_proc));
+	      gdb_assert (!has_reinsert_breakpoints (child_proc));
+	    }
+
 	  /* Report the event.  */
 	  return 0;
 	}
-- 
1.9.1


  parent reply	other threads:[~2016-05-20 15:13 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-05-20 15:12 [PATCH 0/8] Use reinsert breakpoint for vCont;s Yao Qi
2016-05-20 15:13 ` [PATCH 6/8] Make reinsert_breakpoint thread specific Yao Qi
2016-05-24 16:39   ` Pedro Alves
2016-05-20 15:13 ` [PATCH 3/8] Pass breakpoint type in set_breakpoint_at Yao Qi
2016-05-24 16:10   ` Pedro Alves
2016-05-20 15:13 ` [PATCH 4/8] Create sub classes of 'struct breakpoint' Yao Qi
2016-05-24 16:23   ` Pedro Alves
2016-05-20 15:13 ` Yao Qi [this message]
2016-05-24 16:06   ` [PATCH 2/8] Delete reinsert breakpoints from forked child Pedro Alves
2016-05-20 15:13 ` [PATCH 7/8] Use reinsert_breakpoint for vCont;s Yao Qi
2016-05-24 17:39   ` Pedro Alves
2016-05-26 16:52     ` Yao Qi
2016-05-26 17:00       ` Pedro Alves
2016-05-20 15:13 ` [PATCH 1/8] Switch to current thread before finish_step_over Yao Qi
2016-05-24 15:59   ` Pedro Alves
2016-05-20 15:13 ` [PATCH 5/8] Refactor clone_all_breakpoints Yao Qi
2016-05-24 16:27   ` Pedro Alves
2016-05-20 15:13 ` [PATCH 8/8] [GDBserver] Support vCont s and S actions with software single step Yao Qi
2016-05-24 17:41   ` Pedro Alves
2016-05-24 15:08 ` [PATCH 0/8] Use reinsert breakpoint for vCont;s Antoine Tremblay

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=1463757161-25850-3-git-send-email-yao.qi@linaro.org \
    --to=qiyaoltc@gmail.com \
    --cc=gdb-patches@sourceware.org \
    /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