public inbox for gdb-patches@sourceware.org
 help / color / mirror / Atom feed
* [PATCH 0/2] Fix "thread find" with multi inferior/target (PR gdb/26631)
@ 2020-09-17 18:00 Pedro Alves
  2020-09-17 18:00 ` [PATCH 1/2] Split gdb.multi/multi-target.exp into separate testcases Pedro Alves
  2020-09-17 18:00 ` [PATCH 2/2] Fix "thread find" with multiple inferiors/targets (PR gdb/26631) Pedro Alves
  0 siblings, 2 replies; 8+ messages in thread
From: Pedro Alves @ 2020-09-17 18:00 UTC (permalink / raw)
  To: gdb-patches

"thread find" with multiple inferiors got broken with the multi-target
work.  Patch #2 fixes it.  Patch #1 is a preparatory patch for the
testcase in patch #2.

Pedro Alves (2):
  Split gdb.multi/multi-target.exp into separate testcases
  Fix "thread find" with multiple inferiors/targets (PR gdb/26631)

 gdb/testsuite/gdb.multi/mtarg-continue.exp       | 105 +++++
 gdb/testsuite/gdb.multi/mtarg-info-inferiors.exp | 110 +++++
 gdb/testsuite/gdb.multi/mtarg-interrupt.exp      |  79 ++++
 gdb/testsuite/gdb.multi/mtarg-no-resumed.exp     |  90 ++++
 gdb/testsuite/gdb.multi/mtarg-ping-pong-next.exp |  85 ++++
 gdb/testsuite/gdb.multi/mtarg-thread-find.exp    | 106 +++++
 gdb/testsuite/gdb.multi/multi-target.exp         | 546 -----------------------
 gdb/testsuite/gdb.multi/multi-target.exp.tcl     | 185 ++++++++
 gdb/thread.c                                     |   6 +
 9 files changed, 766 insertions(+), 546 deletions(-)
 create mode 100644 gdb/testsuite/gdb.multi/mtarg-continue.exp
 create mode 100644 gdb/testsuite/gdb.multi/mtarg-info-inferiors.exp
 create mode 100644 gdb/testsuite/gdb.multi/mtarg-interrupt.exp
 create mode 100644 gdb/testsuite/gdb.multi/mtarg-no-resumed.exp
 create mode 100644 gdb/testsuite/gdb.multi/mtarg-ping-pong-next.exp
 create mode 100644 gdb/testsuite/gdb.multi/mtarg-thread-find.exp
 delete mode 100644 gdb/testsuite/gdb.multi/multi-target.exp
 create mode 100644 gdb/testsuite/gdb.multi/multi-target.exp.tcl


base-commit: 9519b2eea05637faa380a68fd679a950176de6db
-- 
2.14.5


^ permalink raw reply	[flat|nested] 8+ messages in thread

* [PATCH 1/2] Split gdb.multi/multi-target.exp into separate testcases
  2020-09-17 18:00 [PATCH 0/2] Fix "thread find" with multi inferior/target (PR gdb/26631) Pedro Alves
@ 2020-09-17 18:00 ` Pedro Alves
  2020-09-17 19:16   ` Simon Marchi
  2020-09-17 19:41   ` Simon Marchi
  2020-09-17 18:00 ` [PATCH 2/2] Fix "thread find" with multiple inferiors/targets (PR gdb/26631) Pedro Alves
  1 sibling, 2 replies; 8+ messages in thread
From: Pedro Alves @ 2020-09-17 18:00 UTC (permalink / raw)
  To: gdb-patches

gdb.multi/multi-target.exp sets up a debug environment with multiple
gdbservers, multiple native processes, and multiple cores, which has
proved useful for exercising a number of multi-target scenarios.

But, as we add more tests to gdb.base/multi-target.exp, it is growing
a bit too large (making a bit cumbersome to debug) and too slow to run
(if you have glibc debug info).

This commit thus splits the multi-target.exp into several testcases,
one per use case.  The common setup code is moved to a new
multi-target.exp.tcl file that is included by all the resulting
multi-target testcases.

gdb/testsuite/ChangeLog:

	* gdb.multi/mtarg-continue.exp: New file, factored out from
	multi-target.exp.
	* gdb.multi/mtarg-info-inferiors.exp: New file, factored out from
	multi-target.exp.
	* gdb.multi/mtarg-interrupt.exp: New file, factored out from
	multi-target.exp.
	* gdb.multi/mtarg-no-resumed.exp: New file, factored out from
	multi-target.exp.
	* gdb.multi/mtarg-ping-pong-next.exp: New file, factored out from
	multi-target.exp.
	* gdb.multi/multi-target.exp.tcl: New file, factored out from
	multi-target.exp.
	* gdb.multi/multi-target.exp: Delete.
---
 gdb/testsuite/gdb.multi/mtarg-continue.exp       | 105 +++++
 gdb/testsuite/gdb.multi/mtarg-info-inferiors.exp | 110 +++++
 gdb/testsuite/gdb.multi/mtarg-interrupt.exp      |  79 ++++
 gdb/testsuite/gdb.multi/mtarg-no-resumed.exp     |  90 ++++
 gdb/testsuite/gdb.multi/mtarg-ping-pong-next.exp |  85 ++++
 gdb/testsuite/gdb.multi/multi-target.exp         | 546 -----------------------
 gdb/testsuite/gdb.multi/multi-target.exp.tcl     | 185 ++++++++
 7 files changed, 654 insertions(+), 546 deletions(-)
 create mode 100644 gdb/testsuite/gdb.multi/mtarg-continue.exp
 create mode 100644 gdb/testsuite/gdb.multi/mtarg-info-inferiors.exp
 create mode 100644 gdb/testsuite/gdb.multi/mtarg-interrupt.exp
 create mode 100644 gdb/testsuite/gdb.multi/mtarg-no-resumed.exp
 create mode 100644 gdb/testsuite/gdb.multi/mtarg-ping-pong-next.exp
 delete mode 100644 gdb/testsuite/gdb.multi/multi-target.exp
 create mode 100644 gdb/testsuite/gdb.multi/multi-target.exp.tcl

diff --git a/gdb/testsuite/gdb.multi/mtarg-continue.exp b/gdb/testsuite/gdb.multi/mtarg-continue.exp
new file mode 100644
index 00000000000..15a77b8ea85
--- /dev/null
+++ b/gdb/testsuite/gdb.multi/mtarg-continue.exp
@@ -0,0 +1,105 @@
+# Copyright 2017-2020 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 "continue" to breakpoints in different targets.  In non-stop
+# mode, also tests "interrupt -a".
+
+source $srcdir/$subdir/multi-target.exp.tcl
+
+if {![mtarg_prepare]} {
+    return
+}
+
+proc test_continue {non-stop} {
+    if {![setup ${non-stop}]} {
+	untested "setup failed"
+	return
+    }
+
+    proc set_break {inf} {
+	gdb_test "break function${inf} thread ${inf}.1" \
+	"Breakpoint .* function${inf}\\..*"
+    }
+
+    # Select inferior INF, and then run to a breakpoint on inferior
+    # INF+1.
+    proc test_continue_inf {inf} {
+	upvar 1 non-stop non-stop
+
+	global gdb_prompt
+	delete_breakpoints
+
+	set next_inf [next_live_inferior $inf]
+
+	gdb_test "inferior $inf" "Switching to inferior $inf.*"
+	set_break $next_inf
+
+	if {${non-stop} == "off"} {
+	    gdb_test "continue" "hit Breakpoint .* function${next_inf}.*"
+	} else {
+	    set msg "continue"
+	    gdb_test_multiple "continue -a&" $msg {
+		-re "Continuing.*$gdb_prompt " {
+		    pass $msg
+		}
+	    }
+
+	    set msg "hit bp"
+	    gdb_test_multiple "" $msg {
+		-re "hit Breakpoint .* function${next_inf}" {
+		    pass $msg
+		}
+	    }
+
+	    set msg "stop all threads"
+	    gdb_test_multiple "interrupt -a" $msg {
+		-re "$gdb_prompt " {
+		    for {set i 0} {$i < 7} {incr i} {
+			set ok 0
+			gdb_test_multiple "" $msg {
+			    -re "Thread\[^\r\n\]*stopped\\." {
+				set ok 1
+			    }
+			}
+			if {!$ok} {
+			    break
+			}
+		    }
+		    gdb_assert $ok $msg
+		}
+	    }
+	}
+    }
+
+    for {set i 1} {$i <= 5} {incr i} {
+	if {$i == 3} {
+	    # This is a core inferior.
+	    continue
+	}
+
+	with_test_prefix "inf$i" {
+	    test_continue_inf $i
+	}
+    }
+}
+
+# Some basic "continue" + breakpoints tests.
+with_test_prefix "continue" {
+    foreach_with_prefix non-stop {"off" "on"} {
+	test_continue ${non-stop}
+    }
+}
+
+mtarg_cleanup
diff --git a/gdb/testsuite/gdb.multi/mtarg-info-inferiors.exp b/gdb/testsuite/gdb.multi/mtarg-info-inferiors.exp
new file mode 100644
index 00000000000..8b5a6227761
--- /dev/null
+++ b/gdb/testsuite/gdb.multi/mtarg-info-inferiors.exp
@@ -0,0 +1,110 @@
+# Copyright 2017-2020 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 "info inferiors" and "info connections" with multiple targets.
+
+source $srcdir/$subdir/multi-target.exp.tcl
+
+if {![mtarg_prepare]} {
+    return
+}
+
+# Test "info inferiors" and "info connections".  MULTI_PROCESS
+# indicates whether the multi-process feature of remote targets is
+# turned off or on.
+proc test_info_inferiors {multi_process} {
+    setup "off"
+
+    gdb_test_no_output \
+	"set remote multiprocess-feature-packet $multi_process"
+
+    # Get the description for inferior INF for when the current
+    # inferior id is CURRENT.
+    proc inf_desc {inf current} {
+	set ws "\[ \t\]+"
+	global decimal
+	upvar multi_process multi_process
+
+	if {($multi_process == "off") && ($inf == 2 || $inf == 5)} {
+	    set desc "Remote target"
+	} else {
+	    set desc "process ${decimal}"
+	}
+
+	set desc "${inf}${ws}${desc}${ws}"
+	if {$inf == $current} {
+	    return "\\* $desc"
+	} else {
+	    return "  $desc"
+	}
+    }
+
+    # Get the "Num" column for CONNECTION for when the current
+    # inferior id is CURRENT_INF.
+    proc connection_num {connection current_inf} {
+	switch $current_inf {
+	    "4" { set current_connection "1"}
+	    "5" { set current_connection "4"}
+	    "6" { set current_connection "5"}
+	    default { set current_connection $current_inf}
+	}
+	if {$connection == $current_connection} {
+	    return "\\* $connection"
+	} else {
+	    return "  $connection"
+	}
+    }
+
+    set ws "\[ \t\]+"
+    global decimal binfile
+
+    # Test "info connections" and "info inferior" by switching to each
+    # inferior one by one.
+    for {set inf 1} {$inf <= 6} {incr inf} {
+	with_test_prefix "inferior $inf" {
+	    gdb_test "inferior $inf" "Switching to inferior $inf.*"
+
+	    gdb_test "info connections" \
+		[multi_line \
+		     "Num${ws}What${ws}Description${ws}" \
+		     "[connection_num 1 $inf]${ws}native${ws}Native process${ws}" \
+		     "[connection_num 2 $inf]${ws}extended-remote localhost:$decimal${ws}Extended remote serial target in gdb-specific protocol${ws}" \
+		     "[connection_num 3 $inf]${ws}core${ws}Local core dump file${ws}" \
+		     "[connection_num 4 $inf]${ws}extended-remote localhost:$decimal${ws}Extended remote serial target in gdb-specific protocol${ws}" \
+		     "[connection_num 5 $inf]${ws}core${ws}Local core dump file${ws}" \
+		    ]
+
+	    gdb_test "info inferiors" \
+		[multi_line \
+		     "Num${ws}Description${ws}Connection${ws}Executable${ws}" \
+		     "[inf_desc 1 $inf]1 \\(native\\)${ws}${binfile}${ws}" \
+		     "[inf_desc 2 $inf]2 \\(extended-remote localhost:$decimal\\)${ws}${binfile}${ws}" \
+		     "[inf_desc 3 $inf]3 \\(core\\)${ws}${binfile}${ws}" \
+		     "[inf_desc 4 $inf]1 \\(native\\)${ws}${binfile}${ws}" \
+		     "[inf_desc 5 $inf]4 \\(extended-remote localhost:$decimal\\)${ws}${binfile}${ws}" \
+		     "[inf_desc 6 $inf]5 \\(core\\)${ws}${binfile}${ws}" \
+		    ]
+	}
+    }
+}
+
+# Test "info inferiors" and "info connections" commands.
+with_test_prefix "info-inferiors" {
+    foreach_with_prefix multi_process {"on" "off"} {
+	test_info_inferiors $multi_process
+    }
+}
+
+mtarg_cleanup
diff --git a/gdb/testsuite/gdb.multi/mtarg-interrupt.exp b/gdb/testsuite/gdb.multi/mtarg-interrupt.exp
new file mode 100644
index 00000000000..32f01947401
--- /dev/null
+++ b/gdb/testsuite/gdb.multi/mtarg-interrupt.exp
@@ -0,0 +1,79 @@
+# Copyright 2017-2020 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 interrupting multiple targets with Ctrl-C.
+
+source $srcdir/$subdir/multi-target.exp.tcl
+
+if {![mtarg_prepare]} {
+    return
+}
+
+proc test_ctrlc {} {
+    if {![setup "off"]} {
+	untested "setup failed"
+	return
+    }
+
+    delete_breakpoints
+
+    # Select inferior INF, continue all inferiors, and then Ctrl-C.
+    proc test_ctrlc_inf {inf} {
+	global gdb_prompt
+
+	gdb_test "inferior $inf" "Switching to inferior $inf.*"
+
+	set msg "continue"
+	gdb_test_multiple "continue" $msg {
+	    -re "Continuing" {
+		pass $msg
+	    }
+	}
+
+	after 200 { send_gdb "\003" }
+
+	set msg "send_gdb control C"
+	gdb_test_multiple "" $msg {
+	    -re "received signal SIGINT.*$gdb_prompt $" {
+		pass $msg
+	    }
+	}
+
+	set msg "all threads stopped"
+	gdb_test_multiple "info threads" "$msg" {
+	    -re "\\\(running\\\).*$gdb_prompt $" {
+		fail $msg
+	    }
+	    -re "$gdb_prompt $" {
+		pass $msg
+	    }
+	}
+    }
+
+    for {set i 1} {$i <= 5} {incr i} {
+	if {$i == 3} {
+	    # This is a core inferior.
+	    continue
+	}
+
+	with_test_prefix "inf$i" {
+	    test_ctrlc_inf $i
+	}
+    }
+}
+
+test_ctrlc
+
+mtarg_cleanup
diff --git a/gdb/testsuite/gdb.multi/mtarg-no-resumed.exp b/gdb/testsuite/gdb.multi/mtarg-no-resumed.exp
new file mode 100644
index 00000000000..262a7fd891e
--- /dev/null
+++ b/gdb/testsuite/gdb.multi/mtarg-no-resumed.exp
@@ -0,0 +1,90 @@
+# Copyright 2017-2020 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 when there's a foreground execution command in progress, a
+# TARGET_WAITKIND_NO_RESUMED for a particular target is ignored when
+# other targets are still resumed.
+
+source $srcdir/$subdir/multi-target.exp.tcl
+
+if {![mtarg_prepare]} {
+    return
+}
+
+proc test_no_resumed_infs {inf_A inf_B} {
+    global gdb_prompt
+
+    if {![setup "off"]} {
+	untested "setup failed"
+	return
+    }
+
+    gdb_test "thread $inf_A.2" "Switching to thread $inf_A\.2 .*" \
+	"select thread of target A"
+
+    gdb_test_no_output "set scheduler-locking on"
+
+    gdb_test_multiple "continue &" "" {
+	-re "Continuing.*$gdb_prompt " {
+	    pass $gdb_test_name
+	}
+    }
+
+    gdb_test "thread $inf_B.2" "Switching to thread $inf_B\.2 .*" \
+	"select thread of target B"
+    gdb_test "p exit_thread = 1" " = 1" \
+	"set the thread to exit on resumption"
+
+    # Wait 3 seconds.  If we see any response from GDB, such as
+    # "No unwaited-for children left." it's a bug.
+    gdb_test_multiple "continue" "continue" {
+	-timeout 3
+	timeout {
+	    pass $gdb_test_name
+	}
+    }
+
+    # Now stop the program (all targets).
+    send_gdb "\003"
+    gdb_test_multiple "" "send_gdb control C" {
+	-re "received signal SIGINT.*$gdb_prompt $" {
+	    pass $gdb_test_name
+	}
+    }
+
+    gdb_test_multiple "info threads" "all threads stopped" {
+	-re "\\\(running\\\).*$gdb_prompt $" {
+	    fail $gdb_test_name
+	}
+	-re "$gdb_prompt $" {
+	    pass $gdb_test_name
+	}
+    }
+}
+
+# inferior 1 -> native
+# inferior 2 -> extended-remote 1
+# inferior 5 -> extended-remote 2
+set inferiors {1 2 5}
+foreach_with_prefix inf_A $inferiors {
+    foreach_with_prefix inf_B $inferiors {
+	if {$inf_A == $inf_B} {
+	    continue
+	}
+	test_no_resumed_infs $inf_A $inf_B
+    }
+}
+
+mtarg_cleanup
diff --git a/gdb/testsuite/gdb.multi/mtarg-ping-pong-next.exp b/gdb/testsuite/gdb.multi/mtarg-ping-pong-next.exp
new file mode 100644
index 00000000000..551e383b6d1
--- /dev/null
+++ b/gdb/testsuite/gdb.multi/mtarg-ping-pong-next.exp
@@ -0,0 +1,85 @@
+# Copyright 2017-2020 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 "next" bouncing between two breakpoints in two threads running
+# in different targets.
+
+source $srcdir/$subdir/multi-target.exp.tcl
+
+if {![mtarg_prepare]} {
+    return
+}
+
+proc test_ping_pong_next {} {
+    global srcfile
+
+    if {![setup "off"]} {
+	untested "setup failed"
+	return
+    }
+
+    # block/unblock inferiors 1 and 2 according to INF1 and INF2.
+    proc block {inf1 inf2} {
+	gdb_test "thread apply 1.1 p wait_for_gdb = $inf1" " = $inf1"
+	gdb_test "thread apply 2.1 p wait_for_gdb = $inf2" " = $inf2"
+    }
+
+    # We're use inferiors 1 and 2.  Make sure they're really connected
+    # to different targets.
+    gdb_test "thread apply 1.1 maint print target-stack" \
+	"- native.*"
+    gdb_test "thread apply 2.1 maint print target-stack" \
+	"- extended-remote.*"
+
+    # Set two breakpoints, one for each of inferior 1 and 2.  Inferior
+    # 1 is running on the native target, and inferior 2 is running on
+    # extended-gdbserver.  Run to breakpoint 1 to gets things started.
+    set line1 [gdb_get_line_number "set break 1 here"]
+    set line2 [gdb_get_line_number "set break 2 here"]
+
+    gdb_test "thread 1.1" "Switching to thread 1.1 .*"
+
+    gdb_test "break $srcfile:$line1 thread 1.1" \
+	"Breakpoint .*$srcfile:$line1\\..*"
+
+    gdb_test "continue" "hit Breakpoint .*"
+
+    gdb_test "break $srcfile:$line2 thread 2.1" \
+	"Breakpoint .*$srcfile:$line2\\..*"
+
+    # Now block inferior 1 and issue "next".  We should stop at the
+    # breakpoint for inferior 2, given schedlock off.
+    with_test_prefix "next inf 1" {
+	block 1 0
+	gdb_test "next" "Thread 2.1 .*hit Breakpoint .*$srcfile:$line2.*"
+    }
+
+    # Now unblock inferior 2 and block inferior 1.  "next" should run
+    # into the breakpoint in inferior 1.
+    with_test_prefix "next inf 2" {
+	block 0 1
+	gdb_test "next" "Thread 1.1 .*hit Breakpoint .*$srcfile:$line1.*"
+    }
+
+    # Try nexting inferior 1 again.
+    with_test_prefix "next inf 1 again" {
+	block 1 0
+	gdb_test "next" "Thread 2.1 .*hit Breakpoint .*$srcfile:$line2.*"
+    }
+}
+
+test_ping_pong_next
+
+mtarg_cleanup
diff --git a/gdb/testsuite/gdb.multi/multi-target.exp b/gdb/testsuite/gdb.multi/multi-target.exp
deleted file mode 100644
index d19cee6595a..00000000000
--- a/gdb/testsuite/gdb.multi/multi-target.exp
+++ /dev/null
@@ -1,546 +0,0 @@
-# Copyright 2017-2020 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 multi-target features.
-
-load_lib gdbserver-support.exp
-
-if { [skip_gdbserver_tests] } {
-    return 0
-}
-
-standard_testfile
-
-# The plain remote target can't do multiple inferiors.
-if {[target_info gdb_protocol] != ""} {
-    return
-}
-
-if { [prepare_for_testing "failed to prepare" ${binfile} "${srcfile}" \
-	  {debug pthreads}] } {
-    return
-}
-
-# Keep a list of (inferior ID, spawn ID).
-set server_spawn_ids [list]
-
-proc connect_target_extended_remote {binfile num} {
-    set res [gdbserver_start "--multi" ""]
-    global server_spawn_ids server_spawn_id
-    lappend server_spawn_ids $num $server_spawn_id
-    set gdbserver_gdbport [lindex $res 1]
-    return [gdb_target_cmd "extended-remote" $gdbserver_gdbport]
-}
-
-# Add and start inferior number NUM.  Returns true on success, false
-# otherwise.
-proc add_inferior {num target binfile {gcorefile ""}} {
-    # Start another inferior.
-    gdb_test "add-inferior -no-connection" "Added inferior $num" \
-	"add empty inferior $num"
-    gdb_test "inferior $num" "Switching to inferior $num.*" \
-	"switch to inferior $num"
-    gdb_test "file ${binfile}" ".*" "load file in inferior $num"
-    gdb_test_no_output "set remote exec-file ${binfile}" \
-	"set remote-exec file in inferior $num"
-
-    if {$target == "core"} {
-	gdb_test "core $gcorefile" "Core was generated by.*" \
-	    "core [file tail $gcorefile], inf $num"
-	return 1
-    }
-
-    if {$target == "extended-remote"} {
-	if {[connect_target_extended_remote $binfile $num]} {
-	    return 0
-	}
-    }
-    if ![runto "all_started"] then {
-	return 0
-    }
-    delete_breakpoints
-
-    return 1
-}
-
-proc prepare_core {} {
-    global gcorefile gcore_created
-    global binfile
-
-    clean_restart ${binfile}
-
-    if ![runto all_started] then {
-	return -1
-    }
-
-    global testfile
-    set gcorefile [standard_output_file $testfile.gcore]
-    set gcore_created [gdb_gcore_cmd $gcorefile "save a core file"]
-}
-
-proc next_live_inferior {inf} {
-    incr inf
-    if {$inf == 3} {
-	# 3 is a core.
-	return 4
-    }
-    if {$inf > 5} {
-	# 6 is a core.
-	return 1
-    }
-
-    return $inf
-}
-
-# Clean up the server_spawn_ids.
-proc cleanup_gdbservers { } {
-    global server_spawn_id
-    global server_spawn_ids
-    foreach { inferior_id spawn_id } $server_spawn_ids {
-	set server_spawn_id $spawn_id
-	gdb_test "inferior $inferior_id"
-	gdbserver_exit 0
-    }
-    set server_spawn_ids [list]
-}
-
-# Return true on success, false otherwise.
-
-proc setup {non-stop} {
-    global gcorefile gcore_created
-    global binfile
-
-    cleanup_gdbservers
-    clean_restart ${binfile}
-
-    # multi-target depends on target running in non-stop mode.  Force
-    # it on for remote targets, until this is the default.
-    gdb_test_no_output "maint set target-non-stop on"
-
-    gdb_test_no_output "set non-stop ${non-stop}"
-
-    if ![runto all_started] then {
-	return 0
-    }
-
-    delete_breakpoints
-
-    # inferior 1 -> native
-    # inferior 2 -> extended-remote
-    # inferior 3 -> core
-    # inferior 4 -> native
-    # inferior 5 -> extended-remote
-    # inferior 6 -> core
-    if {![add_inferior 2 "extended-remote" $binfile]} {
-	return 0
-    }
-    if {![add_inferior 3 "core" $binfile $gcorefile]} {
-	return 0
-    }
-    if {![add_inferior 4 "native" $binfile]} {
-	return 0
-    }
-    if {![add_inferior 5 "extended-remote" $binfile]} {
-	return 0
-    }
-    if {![add_inferior 6 "core" $binfile $gcorefile]} {
-	return 0
-    }
-
-    # For debugging.
-    gdb_test "info threads" ".*"
-
-    # Make "continue" resume all inferiors.
-    if {${non-stop} == "off"} {
-	gdb_test_no_output "set schedule-multiple on"
-    }
-
-    return 1
-}
-
-# Test "continue" to breakpoints in different targets.  In non-stop
-# mode, also tests "interrupt -a".
-proc test_continue {non-stop} {
-    if {![setup ${non-stop}]} {
-	untested "setup failed"
-	return
-    }
-
-    proc set_break {inf} {
-	gdb_test "break function${inf} thread ${inf}.1" \
-	"Breakpoint .* function${inf}\\..*"
-    }
-
-    # Select inferior INF, and then run to a breakpoint on inferior
-    # INF+1.
-    proc test_continue_inf {inf} {
-	upvar 1 non-stop non-stop
-
-	global gdb_prompt
-	delete_breakpoints
-
-	set next_inf [next_live_inferior $inf]
-
-	gdb_test "inferior $inf" "Switching to inferior $inf.*"
-	set_break $next_inf
-
-	if {${non-stop} == "off"} {
-	    gdb_test "continue" "hit Breakpoint .* function${next_inf}.*"
-	} else {
-	    set msg "continue"
-	    gdb_test_multiple "continue -a&" $msg {
-		-re "Continuing.*$gdb_prompt " {
-		    pass $msg
-		}
-	    }
-
-	    set msg "hit bp"
-	    gdb_test_multiple "" $msg {
-		-re "hit Breakpoint .* function${next_inf}" {
-		    pass $msg
-		}
-	    }
-
-	    set msg "stop all threads"
-	    gdb_test_multiple "interrupt -a" $msg {
-		-re "$gdb_prompt " {
-		    for {set i 0} {$i < 7} {incr i} {
-			set ok 0
-			gdb_test_multiple "" $msg {
-			    -re "Thread\[^\r\n\]*stopped\\." {
-				set ok 1
-			    }
-			}
-			if {!$ok} {
-			    break
-			}
-		    }
-		    gdb_assert $ok $msg
-		}
-	    }
-	}
-    }
-
-    for {set i 1} {$i <= 5} {incr i} {
-	if {$i == 3} {
-	    # This is a core inferior.
-	    continue
-	}
-
-	with_test_prefix "inf$i" {
-	    test_continue_inf $i
-	}
-    }
-}
-
-# Test interrupting multiple targets with Ctrl-C.
-
-proc test_ctrlc {} {
-    if {![setup "off"]} {
-	untested "setup failed"
-	return
-    }
-
-    delete_breakpoints
-
-    # Select inferior INF, continue all inferiors, and then Ctrl-C.
-    proc test_ctrlc_inf {inf} {
-	global gdb_prompt
-
-	gdb_test "inferior $inf" "Switching to inferior $inf.*"
-
-	set msg "continue"
-	gdb_test_multiple "continue" $msg {
-	    -re "Continuing" {
-		pass $msg
-	    }
-	}
-
-	after 200 { send_gdb "\003" }
-
-	set msg "send_gdb control C"
-	gdb_test_multiple "" $msg {
-	    -re "received signal SIGINT.*$gdb_prompt $" {
-		pass $msg
-	    }
-	}
-
-	set msg "all threads stopped"
-	gdb_test_multiple "info threads" "$msg" {
-	    -re "\\\(running\\\).*$gdb_prompt $" {
-		fail $msg
-	    }
-	    -re "$gdb_prompt $" {
-		pass $msg
-	    }
-	}
-    }
-
-    for {set i 1} {$i <= 5} {incr i} {
-	if {$i == 3} {
-	    # This is a core inferior.
-	    continue
-	}
-
-	with_test_prefix "inf$i" {
-	    test_ctrlc_inf $i
-	}
-    }
-}
-
-# Test "next" bouncing between two breakpoints in two threads running
-# in different targets.
-proc test_ping_pong_next {} {
-    global srcfile
-
-    if {![setup "off"]} {
-	untested "setup failed"
-	return
-    }
-
-    # block/unblock inferiors 1 and 2 according to INF1 and INF2.
-    proc block {inf1 inf2} {
-	gdb_test "thread apply 1.1 p wait_for_gdb = $inf1" " = $inf1"
-	gdb_test "thread apply 2.1 p wait_for_gdb = $inf2" " = $inf2"
-    }
-
-    # We're use inferiors 1 and 2.  Make sure they're really connected
-    # to different targets.
-    gdb_test "thread apply 1.1 maint print target-stack" \
-	"- native.*"
-    gdb_test "thread apply 2.1 maint print target-stack" \
-	"- extended-remote.*"
-
-    # Set two breakpoints, one for each of inferior 1 and 2.  Inferior
-    # 1 is running on the native target, and inferior 2 is running on
-    # extended-gdbserver.  Run to breakpoint 1 to gets things started.
-    set line1 [gdb_get_line_number "set break 1 here"]
-    set line2 [gdb_get_line_number "set break 2 here"]
-
-    gdb_test "thread 1.1" "Switching to thread 1.1 .*"
-
-    gdb_test "break $srcfile:$line1 thread 1.1" \
-	"Breakpoint .*$srcfile:$line1\\..*"
-
-    gdb_test "continue" "hit Breakpoint .*"
-
-    gdb_test "break $srcfile:$line2 thread 2.1" \
-	"Breakpoint .*$srcfile:$line2\\..*"
-
-    # Now block inferior 1 and issue "next".  We should stop at the
-    # breakpoint for inferior 2, given schedlock off.
-    with_test_prefix "next inf 1" {
-	block 1 0
-	gdb_test "next" "Thread 2.1 .*hit Breakpoint .*$srcfile:$line2.*"
-    }
-
-    # Now unblock inferior 2 and block inferior 1.  "next" should run
-    # into the breakpoint in inferior 1.
-    with_test_prefix "next inf 2" {
-	block 0 1
-	gdb_test "next" "Thread 1.1 .*hit Breakpoint .*$srcfile:$line1.*"
-    }
-
-    # Try nexting inferior 1 again.
-    with_test_prefix "next inf 1 again" {
-	block 1 0
-	gdb_test "next" "Thread 2.1 .*hit Breakpoint .*$srcfile:$line2.*"
-    }
-}
-
-# Test "info inferiors" and "info connections".  MULTI_PROCESS
-# indicates whether the multi-process feature of remote targets is
-# turned off or on.
-proc test_info_inferiors {multi_process} {
-    setup "off"
-
-    gdb_test_no_output \
-	"set remote multiprocess-feature-packet $multi_process"
-
-    # Get the description for inferior INF for when the current
-    # inferior id is CURRENT.
-    proc inf_desc {inf current} {
-	set ws "\[ \t\]+"
-	global decimal
-	upvar multi_process multi_process
-
-	if {($multi_process == "off") && ($inf == 2 || $inf == 5)} {
-	    set desc "Remote target"
-	} else {
-	    set desc "process ${decimal}"
-	}
-
-	set desc "${inf}${ws}${desc}${ws}"
-	if {$inf == $current} {
-	    return "\\* $desc"
-	} else {
-	    return "  $desc"
-	}
-    }
-
-    # Get the "Num" column for CONNECTION for when the current
-    # inferior id is CURRENT_INF.
-    proc connection_num {connection current_inf} {
-	switch $current_inf {
-	    "4" { set current_connection "1"}
-	    "5" { set current_connection "4"}
-	    "6" { set current_connection "5"}
-	    default { set current_connection $current_inf}
-	}
-	if {$connection == $current_connection} {
-	    return "\\* $connection"
-	} else {
-	    return "  $connection"
-	}
-    }
-
-    set ws "\[ \t\]+"
-    global decimal binfile
-
-    # Test "info connections" and "info inferior" by switching to each
-    # inferior one by one.
-    for {set inf 1} {$inf <= 6} {incr inf} {
-	with_test_prefix "inferior $inf" {
-	    gdb_test "inferior $inf" "Switching to inferior $inf.*"
-
-	    gdb_test "info connections" \
-		[multi_line \
-		     "Num${ws}What${ws}Description${ws}" \
-		     "[connection_num 1 $inf]${ws}native${ws}Native process${ws}" \
-		     "[connection_num 2 $inf]${ws}extended-remote localhost:$decimal${ws}Extended remote serial target in gdb-specific protocol${ws}" \
-		     "[connection_num 3 $inf]${ws}core${ws}Local core dump file${ws}" \
-		     "[connection_num 4 $inf]${ws}extended-remote localhost:$decimal${ws}Extended remote serial target in gdb-specific protocol${ws}" \
-		     "[connection_num 5 $inf]${ws}core${ws}Local core dump file${ws}" \
-		    ]
-
-	    gdb_test "info inferiors" \
-		[multi_line \
-		     "Num${ws}Description${ws}Connection${ws}Executable${ws}" \
-		     "[inf_desc 1 $inf]1 \\(native\\)${ws}${binfile}${ws}" \
-		     "[inf_desc 2 $inf]2 \\(extended-remote localhost:$decimal\\)${ws}${binfile}${ws}" \
-		     "[inf_desc 3 $inf]3 \\(core\\)${ws}${binfile}${ws}" \
-		     "[inf_desc 4 $inf]1 \\(native\\)${ws}${binfile}${ws}" \
-		     "[inf_desc 5 $inf]4 \\(extended-remote localhost:$decimal\\)${ws}${binfile}${ws}" \
-		     "[inf_desc 6 $inf]5 \\(core\\)${ws}${binfile}${ws}" \
-		    ]
-	}
-    }
-}
-
-# Test that when there's a foreground execution command in progress, a
-# TARGET_WAITKIND_NO_RESUMED for a particular target is ignored when
-# other targets are still resumed.
-
-proc test_no_resumed {} {
-    proc test_no_resumed_infs {inf_A inf_B} {
-	global gdb_prompt
-
-	if {![setup "off"]} {
-	    untested "setup failed"
-	    return
-	}
-
-	gdb_test "thread $inf_A.2" "Switching to thread $inf_A\.2 .*" \
-	    "select thread of target A"
-
-	gdb_test_no_output "set scheduler-locking on"
-
-	gdb_test_multiple "continue &" "" {
-	    -re "Continuing.*$gdb_prompt " {
-		pass $gdb_test_name
-	    }
-	}
-
-	gdb_test "thread $inf_B.2" "Switching to thread $inf_B\.2 .*" \
-	    "select thread of target B"
-	gdb_test "p exit_thread = 1" " = 1" \
-	    "set the thread to exit on resumption"
-
-	# Wait 3 seconds.  If we see any response from GDB, such as
-	# "No unwaited-for children left." it's a bug.
-	gdb_test_multiple "continue" "continue" {
-	    -timeout 3
-	    timeout {
-		pass $gdb_test_name
-	    }
-	}
-
-	# Now stop the program (all targets).
-	send_gdb "\003"
-	gdb_test_multiple "" "send_gdb control C" {
-	    -re "received signal SIGINT.*$gdb_prompt $" {
-		pass $gdb_test_name
-	    }
-	}
-
-	gdb_test_multiple "info threads" "all threads stopped" {
-	    -re "\\\(running\\\).*$gdb_prompt $" {
-		fail $gdb_test_name
-	    }
-	    -re "$gdb_prompt $" {
-		pass $gdb_test_name
-	    }
-	}
-    }
-
-    # inferior 1 -> native
-    # inferior 2 -> extended-remote 1
-    # inferior 5 -> extended-remote 2
-    set inferiors {1 2 5}
-    foreach_with_prefix inf_A $inferiors {
-	foreach_with_prefix inf_B $inferiors {
-	    if {$inf_A == $inf_B} {
-		continue
-	    }
-	    test_no_resumed_infs $inf_A $inf_B
-	}
-    }
-}
-
-
-# Make a core file with two threads upfront.  Several tests load the
-# same core file.
-prepare_core
-
-# Some basic "continue" + breakpoints tests.
-with_test_prefix "continue" {
-    foreach_with_prefix non-stop {"off" "on"} {
-	test_continue ${non-stop}
-    }
-}
-
-# Some basic all-stop Ctrl-C tests.
-with_test_prefix "interrupt" {
-    test_ctrlc
-}
-
-# Test ping-ponging between two targets with "next".
-with_test_prefix "ping-pong" {
-    test_ping_pong_next
-}
-
-# Test "info inferiors" and "info connections" commands.
-with_test_prefix "info-inferiors" {
-    foreach_with_prefix multi_process {"on" "off"} {
-	test_info_inferiors $multi_process
-    }
-}
-
-# Test TARGET_WAITKIND_NO_RESUMED handling with multiple targets.
-with_test_prefix "no-resumed" {
-    test_no_resumed
-}
-
-cleanup_gdbservers
diff --git a/gdb/testsuite/gdb.multi/multi-target.exp.tcl b/gdb/testsuite/gdb.multi/multi-target.exp.tcl
new file mode 100644
index 00000000000..f098f075991
--- /dev/null
+++ b/gdb/testsuite/gdb.multi/multi-target.exp.tcl
@@ -0,0 +1,185 @@
+# Copyright 2017-2020 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/>.
+
+# Common routines for testing multi-target features.
+
+load_lib gdbserver-support.exp
+
+standard_testfile multi-target.c
+
+# Keep a list of (inferior ID, spawn ID).
+set server_spawn_ids [list]
+
+proc connect_target_extended_remote {binfile num} {
+    set res [gdbserver_start "--multi" ""]
+    global server_spawn_ids server_spawn_id
+    lappend server_spawn_ids $num $server_spawn_id
+    set gdbserver_gdbport [lindex $res 1]
+    return [gdb_target_cmd "extended-remote" $gdbserver_gdbport]
+}
+
+# Add and start inferior number NUM.  Returns true on success, false
+# otherwise.
+proc add_inferior {num target binfile {gcorefile ""}} {
+    # Start another inferior.
+    gdb_test "add-inferior -no-connection" "Added inferior $num" \
+	"add empty inferior $num"
+    gdb_test "inferior $num" "Switching to inferior $num.*" \
+	"switch to inferior $num"
+    gdb_test "file ${binfile}" ".*" "load file in inferior $num"
+    gdb_test_no_output "set remote exec-file ${binfile}" \
+	"set remote-exec file in inferior $num"
+
+    if {$target == "core"} {
+	gdb_test "core $gcorefile" "Core was generated by.*" \
+	    "core [file tail $gcorefile], inf $num"
+	return 1
+    }
+
+    if {$target == "extended-remote"} {
+	if {[connect_target_extended_remote $binfile $num]} {
+	    return 0
+	}
+    }
+    if ![runto "all_started"] then {
+	return 0
+    }
+    delete_breakpoints
+
+    return 1
+}
+
+proc prepare_core {} {
+    global gcorefile gcore_created
+    global binfile
+
+    clean_restart ${binfile}
+
+    if ![runto all_started] then {
+	return -1
+    }
+
+    global testfile
+    set gcorefile [standard_output_file $testfile.gcore]
+    set gcore_created [gdb_gcore_cmd $gcorefile "save a core file"]
+}
+
+proc next_live_inferior {inf} {
+    incr inf
+    if {$inf == 3} {
+	# 3 is a core.
+	return 4
+    }
+    if {$inf > 5} {
+	# 6 is a core.
+	return 1
+    }
+
+    return $inf
+}
+
+# Clean up the server_spawn_ids.
+proc cleanup_gdbservers { } {
+    global server_spawn_id
+    global server_spawn_ids
+    foreach { inferior_id spawn_id } $server_spawn_ids {
+	set server_spawn_id $spawn_id
+	gdb_test "inferior $inferior_id"
+	gdbserver_exit 0
+    }
+    set server_spawn_ids [list]
+}
+
+# Return true on success, false otherwise.
+
+proc setup {non-stop} {
+    global gcorefile gcore_created
+    global binfile
+
+    cleanup_gdbservers
+    clean_restart ${binfile}
+
+    # multi-target depends on target running in non-stop mode.  Force
+    # it on for remote targets, until this is the default.
+    gdb_test_no_output "maint set target-non-stop on"
+
+    gdb_test_no_output "set non-stop ${non-stop}"
+
+    if ![runto all_started] then {
+	return 0
+    }
+
+    delete_breakpoints
+
+    # inferior 1 -> native
+    # inferior 2 -> extended-remote
+    # inferior 3 -> core
+    # inferior 4 -> native
+    # inferior 5 -> extended-remote
+    # inferior 6 -> core
+    if {![add_inferior 2 "extended-remote" $binfile]} {
+	return 0
+    }
+    if {![add_inferior 3 "core" $binfile $gcorefile]} {
+	return 0
+    }
+    if {![add_inferior 4 "native" $binfile]} {
+	return 0
+    }
+    if {![add_inferior 5 "extended-remote" $binfile]} {
+	return 0
+    }
+    if {![add_inferior 6 "core" $binfile $gcorefile]} {
+	return 0
+    }
+
+    # For debugging.
+    gdb_test "info threads" ".*"
+
+    # Make "continue" resume all inferiors.
+    if {${non-stop} == "off"} {
+	gdb_test_no_output "set schedule-multiple on"
+    }
+
+    return 1
+}
+
+proc mtarg_prepare {} {
+    global binfile srcfile
+
+    if { [skip_gdbserver_tests] } {
+	return 0
+    }
+
+    # The plain remote target can't do multiple inferiors.
+    if {[target_info gdb_protocol] != ""} {
+	return 0
+    }
+
+    if { [prepare_for_testing "failed to prepare" ${binfile} "${srcfile}" \
+	      {debug pthreads}] } {
+	return 0
+    }
+
+    # Make a core file with two threads upfront.  Several tests load
+    # the same core file.
+    prepare_core
+
+    return 1
+}
+
+proc mtarg_cleanup {} {
+    cleanup_gdbservers
+}
-- 
2.14.5


^ permalink raw reply	[flat|nested] 8+ messages in thread

* [PATCH 2/2] Fix "thread find" with multiple inferiors/targets (PR gdb/26631)
  2020-09-17 18:00 [PATCH 0/2] Fix "thread find" with multi inferior/target (PR gdb/26631) Pedro Alves
  2020-09-17 18:00 ` [PATCH 1/2] Split gdb.multi/multi-target.exp into separate testcases Pedro Alves
@ 2020-09-17 18:00 ` Pedro Alves
  2020-09-17 19:37   ` Simon Marchi
  1 sibling, 1 reply; 8+ messages in thread
From: Pedro Alves @ 2020-09-17 18:00 UTC (permalink / raw)
  To: gdb-patches

"thread find" with multiple inferiors got broken with the multi-target
work:

 Thread 1 "gdb" hit Breakpoint 1, internal_error (...) at ../../src/gdbsupport/errors.cc:51
 51      {
 (top-gdb) bt
 #0  internal_error (file=0xffffd4d0 <error: Cannot access memory at address 0xffffd4d0>, line=0, fmt=0x555556330320 "en_US.UTF-8") at ../../src/gdbsupport/errors.cc:51
 #1  0x0000555555bca4c7 in target_thread_name (info=0x555556801290) at ../../src/gdb/target.c:2035
 #2  0x0000555555beb07a in thread_find_command (arg=0x7fffffffe08e "1", from_tty=0) at ../../src/gdb/thread.c:1959
 #3  0x000055555572ec49 in do_const_cfunc (c=0x555556786bc0, args=0x7fffffffe08e "1", from_tty=0) at ../../src/gdb/cli/cli-decode.c:95
 #4  0x0000555555732abd in cmd_func (cmd=0x555556786bc0, args=0x7fffffffe08e "1", from_tty=0) at ../../src/gdb/cli/cli-decode.c:2181
 #5  0x0000555555bf1245 in execute_command (p=0x7fffffffe08e "1", from_tty=0) at ../../src/gdb/top.c:664
 #6  0x00005555559cad10 in catch_command_errors (command=0x555555bf0c31 <execute_command(char const*, int)>, arg=0x7fffffffe082 "thread find 1", from_tty=0) at ../../src/gdb/main.c:457
 #7  0x00005555559cc33d in captured_main_1 (context=0x7fffffffdb60) at ../../src/gdb/main.c:1218
 #8  0x00005555559cc571 in captured_main (data=0x7fffffffdb60) at ../../src/gdb/main.c:1243
 #9  0x00005555559cc5e8 in gdb_main (args=0x7fffffffdb60) at ../../src/gdb/main.c:1268
 #10 0x0000555555623816 in main (argc=17, argv=0x7fffffffdc78) at ../../src/gdb/gdb.c:32

The problem is that we're not switching to the inferior/target before
calling target methods, which trips on an assertion put in place
exactly to catch this sort of problem.

gdb/testsuite/ChangeLog:

	* gdb.multi/mtarg-thread-find.exp: New file.

gdb/ChangeLog:

	* thread.c (thread_find_command): Switch inferior before
	calling target methods.
---
 gdb/testsuite/gdb.multi/mtarg-thread-find.exp | 106 ++++++++++++++++++++++++++
 gdb/thread.c                                  |   6 ++
 2 files changed, 112 insertions(+)
 create mode 100644 gdb/testsuite/gdb.multi/mtarg-thread-find.exp

diff --git a/gdb/testsuite/gdb.multi/mtarg-thread-find.exp b/gdb/testsuite/gdb.multi/mtarg-thread-find.exp
new file mode 100644
index 00000000000..faca16e7aef
--- /dev/null
+++ b/gdb/testsuite/gdb.multi/mtarg-thread-find.exp
@@ -0,0 +1,106 @@
+# Copyright 2020 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 "thread find" with multiple inferiors/targets.  Regression test
+# for PR gdb/26631.
+
+source $srcdir/$subdir/multi-target.exp.tcl
+
+if {![mtarg_prepare]} {
+    return
+}
+
+proc test_thread_find {} {
+    global decimal gdb_prompt
+
+    set NUM_INFS 6
+
+    if {![setup "off"]} {
+	untested "setup failed"
+	return
+    }
+
+    # This makes sure we don't crash.  See PR gdb/26631.
+    gdb_test "thread find xxxyyyzzz" \
+	"No threads match 'xxxyyyzzz'"
+
+    # Create thread names.
+    for {set i 1} {$i <= $NUM_INFS} {incr i} {
+	gdb_test "thread apply $i.1 thread name threadname_$i" \
+	    "Thread $i.1 .*" \
+	    "name thread $i"
+    }
+
+    # Collect target ids.
+
+    for {set i 1} {$i <= $NUM_INFS} {incr i} {
+	set target_id($i) ""
+    }
+    set any "\[^\r\n\]*"
+    gdb_test_multiple "info threads" "collect thread id" {
+	-re ". ($decimal).$decimal  (Thread ${any}) \"threadname_\[0-9\]+\" $any" {
+	    set thr_num $expect_out(1,string)
+	    set target_id($thr_num) $expect_out(2,string)
+	    exp_continue
+	}
+	-re ".*$gdb_prompt $" {
+	    pass "collect target id"
+	}
+    }
+
+    # Find the threads by name.  Note we repeat the search with each
+    # inferior selected, so that we're sure that GDB doesn't get
+    # confused with which target stack to consult.
+
+    with_test_prefix "find by name" {
+	for {set sel_inf 1} {$sel_inf <= $NUM_INFS} {incr sel_inf} {
+	    with_test_prefix "inf $sel_inf" {
+
+		gdb_test "inferior $sel_inf" \
+		    "Switching to inferior $sel_inf .*"
+
+		for {set find_inf 1} {$find_inf <= $NUM_INFS} {incr find_inf} {
+		    gdb_test "thread find threadname_$find_inf" \
+			"Thread $find_inf.1 has name 'threadname_$find_inf'" \
+			"find thread name $find_inf"
+		}
+	    }
+	}
+    }
+
+    # Find the threads by target id.  Likewise we repeat the search
+    # with each inferior selected.
+
+    with_test_prefix "find by target id" {
+	for {set sel_inf 1} {$sel_inf <= $NUM_INFS} {incr sel_inf} {
+	    with_test_prefix "inf $sel_inf" {
+
+		gdb_test "inferior $sel_inf" \
+		    "Switching to inferior $sel_inf .*"
+
+		for {set find_inf 1} {$find_inf <= $NUM_INFS} {incr find_inf} {
+		    set target_id_re [string_to_regexp $target_id($find_inf)]
+		    gdb_test "thread find $target_id($find_inf)" \
+			"Thread $find_inf.1 has target id '$target_id_re'.*" \
+			"find thread target id $find_inf"
+		}
+	    }
+	}
+    }
+}
+
+test_thread_find
+
+mtarg_cleanup
diff --git a/gdb/thread.c b/gdb/thread.c
index c915407581f..0217f3b54f7 100644
--- a/gdb/thread.c
+++ b/gdb/thread.c
@@ -1946,9 +1946,15 @@ thread_find_command (const char *arg, int from_tty)
   if (tmp != 0)
     error (_("Invalid regexp (%s): %s"), tmp, arg);
 
+  /* We're going to be switching threads.  */
+  scoped_restore_current_thread restore_thread;
+
   update_thread_list ();
+
   for (thread_info *tp : all_threads ())
     {
+      switch_to_inferior_no_thread (tp->inf);
+
       if (tp->name != NULL && re_exec (tp->name))
 	{
 	  printf_filtered (_("Thread %s has name '%s'\n"),
-- 
2.14.5


^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [PATCH 1/2] Split gdb.multi/multi-target.exp into separate testcases
  2020-09-17 18:00 ` [PATCH 1/2] Split gdb.multi/multi-target.exp into separate testcases Pedro Alves
@ 2020-09-17 19:16   ` Simon Marchi
  2020-09-18 12:34     ` Pedro Alves
  2020-09-17 19:41   ` Simon Marchi
  1 sibling, 1 reply; 8+ messages in thread
From: Simon Marchi @ 2020-09-17 19:16 UTC (permalink / raw)
  To: Pedro Alves, gdb-patches

On 2020-09-17 2:00 p.m., Pedro Alves wrote:
> gdb.multi/multi-target.exp sets up a debug environment with multiple
> gdbservers, multiple native processes, and multiple cores, which has
> proved useful for exercising a number of multi-target scenarios.
>
> But, as we add more tests to gdb.base/multi-target.exp, it is growing
> a bit too large (making a bit cumbersome to debug) and too slow to run
> (if you have glibc debug info).
>
> This commit thus splits the multi-target.exp into several testcases,
> one per use case.  The common setup code is moved to a new
> multi-target.exp.tcl file that is included by all the resulting
> multi-target testcases.

Nice, that makes the test cases much more readable too.

> diff --git a/gdb/testsuite/gdb.multi/mtarg-ping-pong-next.exp b/gdb/testsuite/gdb.multi/mtarg-ping-pong-next.exp
> new file mode 100644
> index 00000000000..551e383b6d1
> --- /dev/null
> +++ b/gdb/testsuite/gdb.multi/mtarg-ping-pong-next.exp
> @@ -0,0 +1,85 @@
> +# Copyright 2017-2020 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 "next" bouncing between two breakpoints in two threads running
> +# in different targets.
> +
> +source $srcdir/$subdir/multi-target.exp.tcl
> +
> +if {![mtarg_prepare]} {
> +    return
> +}
> +
> +proc test_ping_pong_next {} {
> +    global srcfile
> +
> +    if {![setup "off"]} {
> +	untested "setup failed"
> +	return
> +    }
> +
> +    # block/unblock inferiors 1 and 2 according to INF1 and INF2.
> +    proc block {inf1 inf2} {
> +	gdb_test "thread apply 1.1 p wait_for_gdb = $inf1" " = $inf1"
> +	gdb_test "thread apply 2.1 p wait_for_gdb = $inf2" " = $inf2"
> +    }
> +
> +    # We're use inferiors 1 and 2.  Make sure they're really connected

I just spotted this typo: "use" -> "using".

Simon

^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [PATCH 2/2] Fix "thread find" with multiple inferiors/targets (PR gdb/26631)
  2020-09-17 18:00 ` [PATCH 2/2] Fix "thread find" with multiple inferiors/targets (PR gdb/26631) Pedro Alves
@ 2020-09-17 19:37   ` Simon Marchi
  0 siblings, 0 replies; 8+ messages in thread
From: Simon Marchi @ 2020-09-17 19:37 UTC (permalink / raw)
  To: Pedro Alves, gdb-patches

On 2020-09-17 2:00 p.m., Pedro Alves wrote:
> "thread find" with multiple inferiors got broken with the multi-target
> work:
>
>  Thread 1 "gdb" hit Breakpoint 1, internal_error (...) at ../../src/gdbsupport/errors.cc:51
>  51      {
>  (top-gdb) bt
>  #0  internal_error (file=0xffffd4d0 <error: Cannot access memory at address 0xffffd4d0>, line=0, fmt=0x555556330320 "en_US.UTF-8") at ../../src/gdbsupport/errors.cc:51
>  #1  0x0000555555bca4c7 in target_thread_name (info=0x555556801290) at ../../src/gdb/target.c:2035
>  #2  0x0000555555beb07a in thread_find_command (arg=0x7fffffffe08e "1", from_tty=0) at ../../src/gdb/thread.c:1959
>  #3  0x000055555572ec49 in do_const_cfunc (c=0x555556786bc0, args=0x7fffffffe08e "1", from_tty=0) at ../../src/gdb/cli/cli-decode.c:95
>  #4  0x0000555555732abd in cmd_func (cmd=0x555556786bc0, args=0x7fffffffe08e "1", from_tty=0) at ../../src/gdb/cli/cli-decode.c:2181
>  #5  0x0000555555bf1245 in execute_command (p=0x7fffffffe08e "1", from_tty=0) at ../../src/gdb/top.c:664
>  #6  0x00005555559cad10 in catch_command_errors (command=0x555555bf0c31 <execute_command(char const*, int)>, arg=0x7fffffffe082 "thread find 1", from_tty=0) at ../../src/gdb/main.c:457
>  #7  0x00005555559cc33d in captured_main_1 (context=0x7fffffffdb60) at ../../src/gdb/main.c:1218
>  #8  0x00005555559cc571 in captured_main (data=0x7fffffffdb60) at ../../src/gdb/main.c:1243
>  #9  0x00005555559cc5e8 in gdb_main (args=0x7fffffffdb60) at ../../src/gdb/main.c:1268
>  #10 0x0000555555623816 in main (argc=17, argv=0x7fffffffdc78) at ../../src/gdb/gdb.c:32
>
> The problem is that we're not switching to the inferior/target before
> calling target methods, which trips on an assertion put in place
> exactly to catch this sort of problem.

This LGTM.  And I can't wait until we make all target methods
independent of the current target/inferior/thread :).

Simon

^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [PATCH 1/2] Split gdb.multi/multi-target.exp into separate testcases
  2020-09-17 18:00 ` [PATCH 1/2] Split gdb.multi/multi-target.exp into separate testcases Pedro Alves
  2020-09-17 19:16   ` Simon Marchi
@ 2020-09-17 19:41   ` Simon Marchi
  2020-09-18 13:01     ` Pedro Alves
  1 sibling, 1 reply; 8+ messages in thread
From: Simon Marchi @ 2020-09-17 19:41 UTC (permalink / raw)
  To: Pedro Alves, gdb-patches

On 2020-09-17 2:00 p.m., Pedro Alves wrote:
> gdb.multi/multi-target.exp sets up a debug environment with multiple
> gdbservers, multiple native processes, and multiple cores, which has
> proved useful for exercising a number of multi-target scenarios.
>
> But, as we add more tests to gdb.base/multi-target.exp, it is growing
> a bit too large (making a bit cumbersome to debug) and too slow to run
> (if you have glibc debug info).
>
> This commit thus splits the multi-target.exp into several testcases,
> one per use case.  The common setup code is moved to a new
> multi-target.exp.tcl file that is included by all the resulting
> multi-target testcases.
>
> gdb/testsuite/ChangeLog:
>
> 	* gdb.multi/mtarg-continue.exp: New file, factored out from
> 	multi-target.exp.
> 	* gdb.multi/mtarg-info-inferiors.exp: New file, factored out from
> 	multi-target.exp.
> 	* gdb.multi/mtarg-interrupt.exp: New file, factored out from
> 	multi-target.exp.
> 	* gdb.multi/mtarg-no-resumed.exp: New file, factored out from
> 	multi-target.exp.
> 	* gdb.multi/mtarg-ping-pong-next.exp: New file, factored out from
> 	multi-target.exp.
> 	* gdb.multi/multi-target.exp.tcl: New file, factored out from
> 	multi-target.exp.
> 	* gdb.multi/multi-target.exp: Delete.

Maybe a small comment here: I think it would be clearer to not
abbreviate "multi-target" to "mtarg", so to name the files
"multi-target-foo.exp".  Abbreviations like that are a bit intimidating
when you are not familiar with them.

Simon

^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [PATCH 1/2] Split gdb.multi/multi-target.exp into separate testcases
  2020-09-17 19:16   ` Simon Marchi
@ 2020-09-18 12:34     ` Pedro Alves
  0 siblings, 0 replies; 8+ messages in thread
From: Pedro Alves @ 2020-09-18 12:34 UTC (permalink / raw)
  To: Simon Marchi, gdb-patches

On 9/17/20 8:16 PM, Simon Marchi wrote:
> On 2020-09-17 2:00 p.m., Pedro Alves wrote:
>> gdb.multi/multi-target.exp sets up a debug environment with multiple
>> gdbservers, multiple native processes, and multiple cores, which has
>> proved useful for exercising a number of multi-target scenarios.
>>
>> But, as we add more tests to gdb.base/multi-target.exp, it is growing
>> a bit too large (making a bit cumbersome to debug) and too slow to run
>> (if you have glibc debug info).
>>
>> This commit thus splits the multi-target.exp into several testcases,
>> one per use case.  The common setup code is moved to a new
>> multi-target.exp.tcl file that is included by all the resulting
>> multi-target testcases.
> 
> Nice, that makes the test cases much more readable too.

I agree.

>> +    # We're use inferiors 1 and 2.  Make sure they're really connected
> 
> I just spotted this typo: "use" -> "using".

I think I originally meant to say "We'll use".  I fixed it to use that.

^ permalink raw reply	[flat|nested] 8+ messages in thread

* Re: [PATCH 1/2] Split gdb.multi/multi-target.exp into separate testcases
  2020-09-17 19:41   ` Simon Marchi
@ 2020-09-18 13:01     ` Pedro Alves
  0 siblings, 0 replies; 8+ messages in thread
From: Pedro Alves @ 2020-09-18 13:01 UTC (permalink / raw)
  To: Simon Marchi, gdb-patches

On 9/17/20 8:41 PM, Simon Marchi wrote:
> On 2020-09-17 2:00 p.m., Pedro Alves wrote:
>> gdb.multi/multi-target.exp sets up a debug environment with multiple
>> gdbservers, multiple native processes, and multiple cores, which has
>> proved useful for exercising a number of multi-target scenarios.
>>
>> But, as we add more tests to gdb.base/multi-target.exp, it is growing
>> a bit too large (making a bit cumbersome to debug) and too slow to run
>> (if you have glibc debug info).
>>
>> This commit thus splits the multi-target.exp into several testcases,
>> one per use case.  The common setup code is moved to a new
>> multi-target.exp.tcl file that is included by all the resulting
>> multi-target testcases.
>>
>> gdb/testsuite/ChangeLog:
>>
>> 	* gdb.multi/mtarg-continue.exp: New file, factored out from
>> 	multi-target.exp.
>> 	* gdb.multi/mtarg-info-inferiors.exp: New file, factored out from
>> 	multi-target.exp.
>> 	* gdb.multi/mtarg-interrupt.exp: New file, factored out from
>> 	multi-target.exp.
>> 	* gdb.multi/mtarg-no-resumed.exp: New file, factored out from
>> 	multi-target.exp.
>> 	* gdb.multi/mtarg-ping-pong-next.exp: New file, factored out from
>> 	multi-target.exp.
>> 	* gdb.multi/multi-target.exp.tcl: New file, factored out from
>> 	multi-target.exp.
>> 	* gdb.multi/multi-target.exp: Delete.
> 
> Maybe a small comment here: I think it would be clearer to not
> abbreviate "multi-target" to "mtarg", so to name the files
> "multi-target-foo.exp".  Abbreviations like that are a bit intimidating
> when you are not familiar with them.

I made that change, here and in patch #2.  I agree that it does indeed
look better.

I also renamed the mtarg_prepare and mtarg_cleanup routines to 
multi_target_prepare and multi_target_cleanup.

I merged this to the master and gdb 10 branches.

Thanks for the review!

Pedro Alves

^ permalink raw reply	[flat|nested] 8+ messages in thread

end of thread, other threads:[~2020-09-18 13:01 UTC | newest]

Thread overview: 8+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-09-17 18:00 [PATCH 0/2] Fix "thread find" with multi inferior/target (PR gdb/26631) Pedro Alves
2020-09-17 18:00 ` [PATCH 1/2] Split gdb.multi/multi-target.exp into separate testcases Pedro Alves
2020-09-17 19:16   ` Simon Marchi
2020-09-18 12:34     ` Pedro Alves
2020-09-17 19:41   ` Simon Marchi
2020-09-18 13:01     ` Pedro Alves
2020-09-17 18:00 ` [PATCH 2/2] Fix "thread find" with multiple inferiors/targets (PR gdb/26631) Pedro Alves
2020-09-17 19:37   ` Simon Marchi

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).