Re: [PATCH] gdb, infrun: do not discard a step-completed pending waitstatus

[Guinevere Larsen <guinevere@redhat.com>]

On 9/15/26 11:14 AM, Markus Metzger wrote:
> Consider a scenario with two breakpoints on adjacent instructions:
>
>      bp1 at 0xf00
>      bp2 at 0xf01
>
> as well as two threads in all-stop-on-top-of-non-stop mode.
>
> Assume that threads A hits bp1 and we report the breakpoint hit to the
> user.  When the user continues, we start a step-over for thread A at 0xf00.
>
> Assume that thread B now hits bp1 and we report the breakpoint hit to the
> user.  We stop all threads to report the event.  Meanwhile, the step-over
> of thread A completes, so we save the pending waitstatus (stop_pc=0xf01,
> currently_stepping=1) of thread A.
>
> When the user continues, clear_proceed_status_thread() discards the
> pending step completed waitstatus of thread A, and proceed() starts
> another step-over for thread A at 0xf01.
>
> We skip bp2 for thread A.
>
> Remove the code in clear_proceed_status_thread() that discards a step
> completed waitstatus and let it get handled normally.

Hi! Thanks for working on this.

I have a question about the situation in general. From what I 
understand, the original intent of the code is that, if a user has asked 
for a step, but we hit something more important (like watchpoint or 
breakpoint), then we shouldn't mention the step. With this change, a 
user will now see the step once they resume from the breakpoint, if I 
understand the code correctly.

Should we restrict this change to only announcing breakpoints, or is 
this behavior change for stepping ok?

I don't have an opinion on what is better, but I think if we think that 
the new behavior is better, it should be called out in the commit message.

-- 
Cheers,
Guinevere Larsen
it/its
she/her (deprecated)

> ---
>   gdb/infrun.c                              | 25 ++-----
>   gdb/testsuite/gdb.threads/adjacent-bp.c   | 48 +++++++++++++
>   gdb/testsuite/gdb.threads/adjacent-bp.exp | 86 +++++++++++++++++++++++
>   3 files changed, 139 insertions(+), 20 deletions(-)
>   create mode 100644 gdb/testsuite/gdb.threads/adjacent-bp.c
>   create mode 100644 gdb/testsuite/gdb.threads/adjacent-bp.exp
>
> diff --git a/gdb/infrun.c b/gdb/infrun.c
> index b9618fb6422..4ab9f4aaf64 100644
> --- a/gdb/infrun.c
> +++ b/gdb/infrun.c
> @@ -3100,28 +3100,13 @@ clear_proceed_status_thread (struct thread_info *tp)
>     infrun_debug_printf ("%s", tp->ptid.to_string ().c_str ());
>     gdb_assert (tp->internal_state () != THREAD_INT_RUNNING);
>   
> -  /* If we're starting a new sequence, then the previous finished
> -     single-step is no longer relevant.  */
>     if (tp->has_pending_waitstatus ())
>       {
> -      if (tp->stop_reason () == TARGET_STOPPED_BY_SINGLE_STEP)
> -	{
> -	  infrun_debug_printf ("pending event of %s was a finished step. "
> -			       "Discarding.",
> -			       tp->ptid.to_string ().c_str ());
> -
> -	  tp->set_internal_state (THREAD_INT_STOPPED);
> -	  tp->clear_pending_waitstatus ();
> -	  tp->set_stop_reason (TARGET_STOPPED_BY_NO_REASON);
> -	}
> -      else
> -	{
> -	  infrun_debug_printf
> -	    ("thread %s has pending wait status %s (currently_stepping=%d).",
> -	     tp->ptid.to_string ().c_str (),
> -	     tp->pending_waitstatus ().to_string ().c_str (),
> -	     tp->control.currently_stepping);
> -	}
> +      infrun_debug_printf
> +	("thread %s has pending wait status %s (currently_stepping=%d).",
> +	 tp->ptid.to_string ().c_str (),
> +	 tp->pending_waitstatus ().to_string ().c_str (),
> +	 tp->control.currently_stepping);
>       }
>   
>     /* If this signal should not be seen by program, give it zero.
> diff --git a/gdb/testsuite/gdb.threads/adjacent-bp.c b/gdb/testsuite/gdb.threads/adjacent-bp.c
> new file mode 100644
> index 00000000000..e72e674f52a
> --- /dev/null
> +++ b/gdb/testsuite/gdb.threads/adjacent-bp.c
> @@ -0,0 +1,48 @@
> +/* Copyright 2026 Free Software Foundation, Inc.
> +
> +   This file is part of GDB.
> +
> +   This program is free software; you can redistribute it and/or modify
> +   it under the terms of the GNU General Public License as published by
> +   the Free Software Foundation; either version 3 of the License, or
> +   (at your option) any later version.
> +
> +   This program is distributed in the hope that it will be useful,
> +   but WITHOUT ANY WARRANTY; without even the implied warranty of
> +   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> +   GNU General Public License for more details.
> +
> +   You should have received a copy of the GNU General Public License
> +   along with this program.  If not, see <http://www.gnu.org/licenses/>.  */
> +
> +#include <pthread.h>
> +#include <unistd.h>
> +
> +static  pthread_barrier_t barrier;
> +
> +static void *
> +test (void *arg)
> +{
> +  pthread_barrier_wait (&barrier);
> +  int a = 0;  /* break here.  */
> +  int b = 0;
> +  int c = 0;
> +  return arg;
> +}
> +
> +int
> +main ()
> +{
> +  pthread_t th;
> +
> +  alarm (500);
> +
> +  pthread_barrier_init (&barrier, NULL, 2);
> +  pthread_create (&th, NULL, test, NULL);
> +  test (NULL);
> +
> +  pthread_join (th, NULL);
> +  pthread_barrier_destroy (&barrier);
> +
> +  return 0;
> +}
> diff --git a/gdb/testsuite/gdb.threads/adjacent-bp.exp b/gdb/testsuite/gdb.threads/adjacent-bp.exp
> new file mode 100644
> index 00000000000..9e7782c01c2
> --- /dev/null
> +++ b/gdb/testsuite/gdb.threads/adjacent-bp.exp
> @@ -0,0 +1,86 @@
> +# Copyright 2026 Free Software Foundation, Inc.
> +
> +# This program is free software; you can redistribute it and/or modify
> +# it under the terms of the GNU General Public License as published by
> +# the Free Software Foundation; either version 3 of the License, or
> +# (at your option) any later version.
> +#
> +# This program is distributed in the hope that it will be useful,
> +# but WITHOUT ANY WARRANTY; without even the implied warranty of
> +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> +# GNU General Public License for more details.
> +#
> +# You should have received a copy of the GNU General Public License
> +# along with this program.  If not, see <http://www.gnu.org/licenses/>.
> +
> +# Test that GDB does not skip a breakpoint when a step-over completes
> +# while another event leads to a stop.
> +
> +standard_testfile
> +
> +if {[prepare_for_testing "failed to prepare" ${testfile} ${srcfile} \
> +	  {debug pthreads}]} {
> +    return
> +}
> +
> +if {![runto_main]} {
> +    return
> +}
> +
> +# Find a sequence of adjacent instructions.
> +set bp_line [gdb_get_line_number "break here"]
> +set pcs {}
> +gdb_test_multiple "info line $bp_line" "" {
> +    -re -wrap "starts at address ($hex).*" {
> +	pass $gdb_test_name
> +
> +	set line "\\s+($hex) \[^\r\n\]+"
> +	gdb_test_multiple "x/3i $expect_out(1,string)" "disassemble" {
> +	    -re -wrap "$line\r\n$line\r\n$line.*" {
> +		pass $gdb_test_name
> +
> +		lappend pcs $expect_out(1,string)
> +		lappend pcs $expect_out(2,string)
> +		lappend pcs $expect_out(3,string)
> +	    }
> +	    -re -wrap "" {
> +		fail $gdb_test_name
> +	    }
> +	}
> +    }
> +    -re -wrap "" {
> +	fail $gdb_test_name
> +    }
> +}
> +
> +# Set breakpoints on adjacent instructions.
> +foreach pc $pcs {
> +    gdb_breakpoint "\*$pc"
> +}
> +
> +# Continue from breakpoint to breakpoint.
> +set hits [dict create]
> +set iter 0
> +gdb_test_multiple "continue" "" {
> +    -re -wrap "hit Breakpoint.*" {
> +	dict incr hits [get_hexadecimal_valueof "\$pc" invalid "stop $iter"]
> +	incr iter
> +	send_gdb "continue\n"
> +	exp_continue
> +    }
> +    -re -wrap "$inferior_exited_re normally.*" {
> +	pass "$gdb_test_name"
> +    }
> +}
> +
> +# We expect all breakpoints to be hit by both threads.
> +foreach pc $pcs {
> +    if {[dict exists $hits $pc]} {
> +	gdb_assert {[dict get $hits $pc] eq 2} "breakpoint at $pc"
> +	dict unset hits $pc
> +    } else {
> +	fail "breakpoint at $pc"
> +    }
> +}
> +# And no unrelated stops.
> +gdb_assert {[dict size $hits] eq 0} "no extra stops"