[Bug mi/28077] New: -exec-interrupt in non-stop mode causes internal-error / assertion in target_stop

[Bug mi/28077] New: -exec-interrupt in non-stop mode causes internal-error / assertion in target_stop

[jonah at kichwacoders dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=28077

            Bug ID: 28077
           Summary: -exec-interrupt in non-stop mode causes internal-error
                    / assertion in target_stop
           Product: gdb
           Version: 11.1
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: mi
          Assignee: unassigned at sourceware dot org
          Reporter: jonah at kichwacoders dot com
  Target Milestone: ---

This is a regression in GDB 11 pre-release and current HEAD that worked in GDB
10.

For a simple program like this loop.c.

volatile int i;
int main() {
  while (1) {
    i += 2;
    i *= 2;
  }
}

gcc -g -o loop loop.c

causes an internal error like this when doing -exec-interrupt in MI mode, it
does not fail in CLI mode.

target.c:3768: internal-error: void target_stop(ptid_t): Assertion
`!proc_target->commit_resumed_state' failed.

MI commands to reproduce:

$ gdb --interpreter=mi3 # or mi2
-gdb-set non-stop on
-gdb-set mi-async on
-file-exec-and-symbols --thread-group i1  loop
-exec-run
-exec-interrupt --thread-group i1

This was done on Ubuntu 20.04 on x86_64 hardware:
$ lsb_release 
LSB Version:    core-11.1.0ubuntu2-noarch:security-11.1.0ubuntu2-noarch
$ uname -a
Linux ditto 5.4.0-77-generic #86-Ubuntu SMP Thu Jun 17 02:35:03 UTC 2021 x86_64
x86_64 x86_64 GNU/Linux
$ gcc --version
gcc (Ubuntu 10.3.0-1ubuntu1~20.04) 10.3.0

I am running a git bisect now and will report back if that provides something
useful.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug mi/28077] -exec-interrupt in non-stop mode causes internal-error / assertion in target_stop

[jonah at kichwacoders dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=28077

--- Comment #1 from Jonah Graham <jonah at kichwacoders dot com> ---
The commit that caused the regression is
1192f124a308601f5fef7a35715ccd6f904e7b17 

Note that "interrupt" or "-exec-interrupt" without the --thread-group argument
does not cause this to fail.

Here is my oneliner to check (part of my git bisect script):

(sleep .5s ; echo '-gdb-set non-stop on' ; echo '-gdb-set mi-async on' ;  echo
'-file-exec-and-symbols --thread-group i1 loop' ; echo '-exec-run' ; sleep .5s
; echo '-exec-interrupt --thread-group i1') | gdb/gdb --interpreter=mi2 

The above asserts, but the below doesn't (without the --thread-group):

(sleep .5s ; echo '-gdb-set non-stop on' ; echo '-gdb-set mi-async on' ;  echo
'-file-exec-and-symbols --thread-group i1 loop' ; echo '-exec-run' ; sleep .5s
; echo '-exec-interrupt') | gdb/gdb --interpreter=mi2

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug mi/28077] -exec-interrupt in non-stop mode causes internal-error / assertion in target_stop

[simark at simark dot ca]

https://sourceware.org/bugzilla/show_bug.cgi?id=28077

Simon Marchi <simark at simark dot ca> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |11.1
                 CC|                            |simark at simark dot ca

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug mi/28077] -exec-interrupt in non-stop mode causes internal-error / assertion in target_stop

[simark at simark dot ca]

https://sourceware.org/bugzilla/show_bug.cgi?id=28077

--- Comment #2 from Simon Marchi <simark at simark dot ca> ---
Thanks for the report.

This should fix it (it fixes it here):

diff --git a/gdb/mi/mi-main.c b/gdb/mi/mi-main.c
index e293dddc08d8..44008d1c0a86 100644
--- a/gdb/mi/mi-main.c
+++ b/gdb/mi/mi-main.c
@@ -383,6 +383,9 @@ mi_cmd_exec_interrupt (const char *command, char **argv,
int argc)
     {
       struct inferior *inf = find_inferior_id (current_context->thread_group);

+      scoped_disable_commit_resumed disable_commit_resumed
+       ("interrupting all threads of thread group");
+
       iterate_over_threads (interrupt_thread_callback, &inf->pid);
     }
   else


I need to write a test though.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug mi/28077] -exec-interrupt in non-stop mode causes internal-error / assertion in target_stop

[jonah at kichwacoders dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=28077

--- Comment #3 from Jonah Graham <jonah at kichwacoders dot com> ---
(In reply to Simon Marchi from comment #2)
> This should fix it (it fixes it here):

It does for Comment 0 reduced test case. I am running the Eclipse CDT testsuite
now.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug mi/28077] -exec-interrupt in non-stop mode causes internal-error / assertion in target_stop

[jonah at kichwacoders dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=28077

--- Comment #4 from Jonah Graham <jonah at kichwacoders dot com> ---
(In reply to Jonah Graham from comment #3)
> (In reply to Simon Marchi from comment #2)
> > This should fix it (it fixes it here):
> 
> It does for Comment 0 reduced test case. I am running the Eclipse CDT
> testsuite now.

Eclipse CDT DSF-GDB testsuites, both local and remote debugging looks good. I
have a remaining failure[1] that I don't know the cause of yet (passes in GDB
10, fails in GDB HEAD). I'll narrow it down and submit it tomorrow hopefully.

[1]
https://ci.eclipse.org/cdt/view/Main%20(debug%20tests)/job/debug-tests-master-gdb-master/lastCompletedBuild/testReport/org.eclipse.cdt.tests.dsf.gdb.tests/OperationsWhileTargetIsRunningTest/detachWhileTargetRunningGDBAlive_gdbserver_/


Thanks Simon!

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug mi/28077] -exec-interrupt in non-stop mode causes internal-error / assertion in target_stop

[simark at simark dot ca]

https://sourceware.org/bugzilla/show_bug.cgi?id=28077

--- Comment #5 from Simon Marchi <simark at simark dot ca> ---
(In reply to Jonah Graham from comment #4)
> (In reply to Jonah Graham from comment #3)
> > (In reply to Simon Marchi from comment #2)
> > > This should fix it (it fixes it here):
> > 
> > It does for Comment 0 reduced test case. I am running the Eclipse CDT
> > testsuite now.
> 
> Eclipse CDT DSF-GDB testsuites, both local and remote debugging looks good.
> I have a remaining failure[1] that I don't know the cause of yet (passes in
> GDB 10, fails in GDB HEAD). I'll narrow it down and submit it tomorrow
> hopefully.
> 
> [1]
> https://ci.eclipse.org/cdt/view/Main%20(debug%20tests)/job/debug-tests-
> master-gdb-master/lastCompletedBuild/testReport/org.eclipse.cdt.tests.dsf.
> gdb.tests/OperationsWhileTargetIsRunningTest/
> detachWhileTargetRunningGDBAlive_gdbserver_/
> 
> 
> Thanks Simon!

Thanks, I'm almost done writing the test, so I should send a proper patch soon.

For the detachWhileTargetRunning issue, this series may be possible culprit:

https://sourceware.org/pipermail/gdb-patches/2021-January/175040.html

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug mi/28077] -exec-interrupt in non-stop mode causes internal-error / assertion in target_stop

[simark at simark dot ca]

https://sourceware.org/bugzilla/show_bug.cgi?id=28077

Simon Marchi <simark at simark dot ca> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|                            |2021-07-12
     Ever confirmed|0                           |1
             Status|UNCONFIRMED                 |NEW

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug mi/28077] -exec-interrupt in non-stop mode causes internal-error / assertion in target_stop

[simark at simark dot ca]

https://sourceware.org/bugzilla/show_bug.cgi?id=28077

Simon Marchi <simark at simark dot ca> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug mi/28077] -exec-interrupt in non-stop mode causes internal-error / assertion in target_stop

[simark at simark dot ca]

https://sourceware.org/bugzilla/show_bug.cgi?id=28077

Simon Marchi <simark at simark dot ca> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Assignee|unassigned at sourceware dot org   |simark at simark dot ca

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug mi/28077] -exec-interrupt in non-stop mode causes internal-error / assertion in target_stop

[cvs-commit at gcc dot gnu.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=28077

--- Comment #6 from cvs-commit at gcc dot gnu.org <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Simon Marchi <simark@sourceware.org>:

https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;h=5c046e0e631dc98faec53d99c76f9d3ef44da75c

commit 5c046e0e631dc98faec53d99c76f9d3ef44da75c
Author: Simon Marchi <simon.marchi@polymtl.ca>
Date:   Sun Jul 11 22:13:06 2021 -0400

    gdb: disable commit-resumed on -exec-interrupt --thread-group

    As reported in PR gdb/28077, we hit an internal error when using
    -exec-interrupt with --thread-group:

        info threads
        &"info threads\n"
        ~"  Id   Target Id             Frame \n"
        ~"* 1    process 403312 \"loop\" (running)\n"
        ^done
        (gdb)
        -exec-interrupt --thread-group i1
        ~"/home/simark/src/binutils-gdb/gdb/target.c:3768: internal-error: void
target_stop(ptid_t): Assertion `!proc_target->commit_resumed_state' failed.\nA
problem internal to GDB has been detected,\nfurther debugging may prove
unreliable.\nQuit this debugging session? (y or n) "

    This is because this code path never disables commit-resumed (a
    requirement for calling target_stop, as documented in
    process_stratum_target::»commit_resumed_state) before calling
    target_stop.

    The other 3 code paths in mi_cmd_exec_interrupt use interrupt_target_1,
    which does it.  But the --thread-group code path uses its own thing
    which doesn't do it.  Fix this by adding a scoped_disable_commit_resumed
    in this code path.

    Calling -exec-interrupt with --thread-group is apparently not tested at
    the moment (which is why this bug could creep in).  Add a new test for
    that.  The test runs two inferiors and tries to interrupt them with
    "-exec-interrupt --thread-group X".

    This will need to be merged in the gdb-11-branch, so here are ChangeLog
    entries:

    gdb/ChangeLog:

            * mi/mi-main.c (mi_cmd_exec_interrupt): Use
            scoped_disable_commit_resumed in the --thread-group case.

    gdb/testsuite/ChangeLog:

            * gdb.mi/interrupt-thread-group.c: New.
            * gdb.mi/interrupt-thread-group.exp: New.

    Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=28077
    Change-Id: I615efefcbcaf2c15d47caf5e4b9d82854b2a2fcb

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug mi/28077] -exec-interrupt in non-stop mode causes internal-error / assertion in target_stop

[cvs-commit at gcc dot gnu.org]

https://sourceware.org/bugzilla/show_bug.cgi?id=28077

--- Comment #7 from cvs-commit at gcc dot gnu.org <cvs-commit at gcc dot gnu.org> ---
The gdb-11-branch branch has been updated by Simon Marchi
<simark@sourceware.org>:

https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;h=1bdfa2b1a8ab4b40372552e3f6ceb3d2c7e621e4

commit 1bdfa2b1a8ab4b40372552e3f6ceb3d2c7e621e4
Author: Simon Marchi <simon.marchi@polymtl.ca>
Date:   Tue Jul 13 09:26:50 2021 -0400

    gdb: disable commit-resumed on -exec-interrupt --thread-group

    As reported in PR gdb/28077, we hit an internal error when using
    -exec-interrupt with --thread-group:

        info threads
        &"info threads\n"
        ~"  Id   Target Id             Frame \n"
        ~"* 1    process 403312 \"loop\" (running)\n"
        ^done
        (gdb)
        -exec-interrupt --thread-group i1
        ~"/home/simark/src/binutils-gdb/gdb/target.c:3768: internal-error: void
target_stop(ptid_t): Assertion `!proc_target->commit_resumed_state' failed.\nA
problem internal to GDB has been detected,\nfurther debugging may prove
unreliable.\nQuit this debugging session? (y or n) "

    This is because this code path never disables commit-resumed (a
    requirement for calling target_stop, as documented in
    process_stratum_target::»commit_resumed_state) before calling
    target_stop.

    The other 3 code paths in mi_cmd_exec_interrupt use interrupt_target_1,
    which does it.  But the --thread-group code path uses its own thing
    which doesn't do it.  Fix this by adding a scoped_disable_commit_resumed
    in this code path.

    Calling -exec-interrupt with --thread-group is apparently not tested at
    the moment (which is why this bug could creep in).  Add a new test for
    that.  The test runs two inferiors and tries to interrupt them with
    "-exec-interrupt --thread-group X".

    This will need to be merged in the gdb-11-branch, so here are ChangeLog
    entries:

    gdb/ChangeLog:

            * mi/mi-main.c (mi_cmd_exec_interrupt): Use
            scoped_disable_commit_resumed in the --thread-group case.

    gdb/testsuite/ChangeLog:

            * gdb.mi/interrupt-thread-group.c: New.
            * gdb.mi/interrupt-thread-group.exp: New.

    Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=28077
    Change-Id: I615efefcbcaf2c15d47caf5e4b9d82854b2a2fcb

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug mi/28077] -exec-interrupt in non-stop mode causes internal-error / assertion in target_stop

[simark at simark dot ca]

https://sourceware.org/bugzilla/show_bug.cgi?id=28077

Simon Marchi <simark at simark dot ca> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |FIXED
             Status|ASSIGNED                    |RESOLVED

--- Comment #8 from Simon Marchi <simark at simark dot ca> ---
Fixed.

-- 
You are receiving this mail because:
You are on the CC list for the bug.