From: Pedro Alves <pedro@codesourcery.com>
To: gdb-patches@sourceware.org
Cc: Daniel Jacobowitz <drow@false.org>,
Roland McGrath <roland@redhat.com>,
Jan Kratochvil <jan.kratochvil@redhat.com>,
Doug Evans <dje@google.com>,
mark.kettenis@xs4all.nl
Subject: Re: [patch] Fix Linux attach to signalled/stopped processes
Date: Sun, 13 Apr 2008 09:35:00 -0000 [thread overview]
Message-ID: <200804121737.49116.pedro@codesourcery.com> (raw)
In-Reply-To: <200804120021.34444.pedro@codesourcery.com>
[-- Attachment #1: Type: text/plain, Size: 806 bytes --]
A Saturday 12 April 2008 00:21:33, Pedro Alves wrote:
> A Friday 11 April 2008 23:54:48, Pedro Alves wrote:
> > A Friday 11 April 2008 23:19:58, Daniel Jacobowitz wrote:
> > > Any idea how to do this?
> >
> Actually, get_last_target_status should give you the ptid
> you want, I think.
Like so ?
The only way I could find to pass args != NULL, was to
attach, and then "quit $signo".
"detach $SIGNO" doesn't work:
(gdb) detach 21
Undefined detach command: "21". Try "help detach".
-- and was glad it didn't. I'm was hoping we can use
"detach $pid" in the future, to mirror "attach $pid",
but now I'm not sure we can...
Anyway, "q 19" did send a SIGSTOP, and the process
was stopped.
And I was able to attach to stopped processes, and
the new tests passed, which is nice. Hurray!
--
Pedro Alves
[-- Attachment #2: attach_stopped_pend_sig.diff --]
[-- Type: text/x-diff, Size: 3404 bytes --]
2008-04-12 Pedro Alves <pedro@codesourcery.com>
* linux-nat.c (get_pending_status): New.
(detach_callback): Pass pending signal.
(linux_nat_detach): Pass pending signal.
---
gdb/linux-nat.c | 50 ++++++++++++++++++++++++++++++++++++++++----------
1 file changed, 40 insertions(+), 10 deletions(-)
Index: src/gdb/linux-nat.c
===================================================================
--- src.orig/gdb/linux-nat.c 2008-04-12 16:20:27.000000000 +0100
+++ src/gdb/linux-nat.c 2008-04-12 17:36:54.000000000 +0100
@@ -1289,6 +1289,36 @@ linux_nat_attach (char *args, int from_t
}
}
+/* Get pending status of LP. */
+static int
+get_pending_status (struct lwp_info *lp, int *status)
+{
+ struct target_waitstatus last;
+ ptid_t last_ptid;
+
+ get_last_target_status (&last_ptid, &last);
+
+ /* If this lwp is the ptid that GDB is processing an event from, the
+ signal will be in stop_signal. Otherwise, in all-stop + sync
+ mode, we may cache pending events in lp->status while trying to
+ stop all threads (see stop_wait_callback). In async mode, the
+ events are always cached in waitpid_queue. */
+
+ *status = 0;
+ if (GET_LWP (lp->ptid) == GET_LWP (last_ptid))
+ {
+ if (stop_signal != TARGET_SIGNAL_0
+ && signal_pass_state (stop_signal))
+ *status = W_STOPCODE (target_signal_to_host (stop_signal));
+ }
+ else if (target_can_async_p ())
+ queued_waitpid (GET_LWP (lp->ptid), status, __WALL);
+ else
+ *status = lp->status;
+
+ return 0;
+}
+
static int
detach_callback (struct lwp_info *lp, void *data)
{
@@ -1311,18 +1341,18 @@ detach_callback (struct lwp_info *lp, vo
lp->signalled = 0;
}
- /* Pass on the last signal, if appropriate. */
- if (lp->status == 0 && GET_LWP (lp->ptid) == GET_LWP (inferior_ptid)
- && stop_signal != TARGET_SIGNAL_0 && signal_pass_state (stop_signal))
- lp->status = W_STOPCODE (target_signal_to_host (stop_signal));
-
/* We don't actually detach from the LWP that has an id equal to the
overall process id just yet. */
if (GET_LWP (lp->ptid) != GET_PID (lp->ptid))
{
+ int status = 0;
+
+ /* Pass on any pending signal for this LWP. */
+ get_pending_status (lwp_list, &status);
+
errno = 0;
if (ptrace (PTRACE_DETACH, GET_LWP (lp->ptid), 0,
- WSTOPSIG (lp->status)) < 0)
+ WSTOPSIG (status)) < 0)
error (_("Can't detach %s: %s"), target_pid_to_str (lp->ptid),
safe_strerror (errno));
@@ -1332,7 +1362,6 @@ detach_callback (struct lwp_info *lp, vo
target_pid_to_str (lp->ptid),
strsignal (WSTOPSIG (lp->status)));
- drain_queued_events (GET_LWP (lp->ptid));
delete_lwp (lp->ptid);
}
@@ -1349,14 +1378,15 @@ linux_nat_detach (char *args, int from_t
if (target_can_async_p ())
linux_nat_async (NULL, 0);
- iterate_over_lwps (detach_callback, &status);
+ iterate_over_lwps (detach_callback, NULL);
/* Only the initial process should be left right now. */
gdb_assert (num_lwps == 1);
/* Pass on any pending signal for the last LWP. */
- status = lwp_list->status;
- if (WIFSTOPPED (status) && (args == NULL || *args == '\0'))
+ if ((args == NULL || *args == '\0')
+ && get_pending_status (lwp_list, &status) != -1
+ && WIFSTOPPED (status))
{
args = alloca (8);
sprintf (args, "%d", (int) WSTOPSIG (status));
next prev parent reply other threads:[~2008-04-12 16:38 UTC|newest]
Thread overview: 52+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-03-31 22:07 Doug Evans
2008-04-02 0:01 ` Jan Kratochvil
2008-04-02 0:07 ` Roland McGrath
2008-04-10 15:30 ` Daniel Jacobowitz
2008-04-10 15:37 ` Jan Kratochvil
2008-04-10 15:49 ` Daniel Jacobowitz
2008-04-10 16:00 ` Jan Kratochvil
2008-04-10 19:59 ` Daniel Jacobowitz
2008-04-10 15:39 ` Daniel Jacobowitz
2008-04-10 16:00 ` Jan Kratochvil
2008-04-10 19:48 ` Daniel Jacobowitz
2008-04-11 8:46 ` Roland McGrath
2008-04-11 17:46 ` Jan Kratochvil
2008-04-11 19:01 ` Daniel Jacobowitz
2008-04-12 7:58 ` Roland McGrath
2008-04-14 15:09 ` Daniel Jacobowitz
2008-04-14 15:31 ` Daniel Jacobowitz
2008-04-15 22:14 ` Jan Kratochvil
2008-05-01 18:50 ` Daniel Jacobowitz
2008-07-05 8:48 ` Jan Kratochvil
2008-07-05 13:48 ` Daniel Jacobowitz
2008-09-24 12:46 ` Andreas Schwab
2008-09-26 3:52 ` Jan Kratochvil
2008-09-26 13:00 ` Daniel Jacobowitz
2008-09-28 11:43 ` Jan Kratochvil
2008-09-28 14:58 ` Daniel Jacobowitz
2008-04-15 8:14 ` Roland McGrath
2008-04-15 13:02 ` Daniel Jacobowitz
2008-04-16 7:01 ` Roland McGrath
2008-04-11 22:20 ` Daniel Jacobowitz
2008-04-11 22:21 ` Pedro Alves
2008-04-11 22:25 ` Daniel Jacobowitz
2008-04-12 0:02 ` Pedro Alves
2008-04-12 0:19 ` Pedro Alves
2008-04-13 9:35 ` Pedro Alves [this message]
2008-04-13 13:40 ` Pedro Alves
2008-04-12 16:38 ` Roland McGrath
2008-04-12 16:43 ` Eli Zaretskii
-- strict thread matches above, loose matches on Subject: below --
2007-06-06 14:34 Jan Kratochvil
2007-06-11 13:44 ` Jan Kratochvil
2007-06-15 18:02 ` Mark Kettenis
2007-06-26 22:40 ` Jan Kratochvil
2007-06-27 0:13 ` Mark Kettenis
2007-06-27 11:59 ` Jan Kratochvil
2007-06-27 18:30 ` Mark Kettenis
2007-06-30 11:45 ` Jan Kratochvil
2007-06-30 11:57 ` Eli Zaretskii
2007-06-30 17:15 ` Jan Kratochvil
2007-06-30 18:52 ` Eli Zaretskii
[not found] ` <200706301852.l5UIq8ek010536@brahms.sibelius.xs4all.nl>
2007-07-01 3:17 ` Eli Zaretskii
2007-07-01 9:34 ` Mark Kettenis
2007-07-01 10:03 ` Jan Kratochvil
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=200804121737.49116.pedro@codesourcery.com \
--to=pedro@codesourcery.com \
--cc=dje@google.com \
--cc=drow@false.org \
--cc=gdb-patches@sourceware.org \
--cc=jan.kratochvil@redhat.com \
--cc=mark.kettenis@xs4all.nl \
--cc=roland@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