RFA (threads testsuite): More thread tests

RFA (threads testsuite): More thread tests

[Daniel Jacobowitz]

Here's two tests I had lying around from when I developed the gdbserver
threads support.  Gdbserver passes them with flying colors (if you use my
other patch which lets gdbserver run tests properly).  GDB shows a couple of
problems, unpredictably (not always repeatable).

OK to add these?

-- 
Daniel Jacobowitz                           Carnegie Mellon University
MontaVista Software                         Debian GNU/Linux Developer

2002-07-09  Daniel Jacobowitz  <drow@mvista.com>

	* gdb.threads/print-threads.c: New file.
	* gdb.threads/print-threads.exp: New file.
	* gdb.threads/schedlock.c: New file.
	* gdb.threads/schedlock.exp: New file.

--- /dev/null	Wed Dec 31 19:00:00 1969
+++ gdb.threads/print-threads.c	Tue Jun 11 14:56:44 2002
@@ -0,0 +1,58 @@
+#include <stdio.h>
+#include <unistd.h>
+#include <stdlib.h>
+#include <pthread.h>
+
+void *thread_function(void *arg); /* Pointer to function executed by each thread */
+
+int slow = 0;
+
+#define NUM 5
+
+int main() {
+    int res;
+    pthread_t threads[NUM];
+    void *thread_result;
+    int args[NUM];
+    int i;
+
+    for (i = 0; i < NUM; i++)
+      {
+	args[i] = i;
+	res = pthread_create(&threads[i], NULL, thread_function, (void *)&args[i]);
+      }
+
+    for (i = 0; i < NUM; i++)
+      res = pthread_join(threads[i], &thread_result);
+
+    printf ("Done\n");
+
+    if (slow)
+      sleep (4);
+
+    exit(EXIT_SUCCESS);
+}
+
+void *thread_function(void *arg) {
+    int my_number = *(int *)arg;
+    int rand_num;
+
+    printf ("Print 1, thread %d\n", my_number);
+    sleep (1);
+
+    if (slow)
+      {
+	printf ("Print 2, thread %d\n", my_number);
+	sleep (1);
+	printf ("Print 3, thread %d\n", my_number);
+	sleep (1);
+	printf ("Print 4, thread %d\n", my_number);
+	sleep (1);
+	printf ("Print 5, thread %d\n", my_number);
+	sleep (1);
+      }
+
+    printf("Bye from %d\n", my_number);
+    pthread_exit(NULL);
+}
+
--- /dev/null	Wed Dec 31 19:00:00 1969
+++ gdb.threads/print-threads.exp	Tue Jun 11 14:56:44 2002
@@ -0,0 +1,158 @@
+# Copyright (C) 1996, 1997, 2002 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 2 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, write to the Free Software
+# Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.  
+
+# Please email any bugs, comments, and/or additions to this file to:
+# bug-gdb@prep.ai.mit.edu
+
+# This file was written by Daniel Jacobowitz <drow@mvista.com>
+# (parts based on pthreads.exp by Fred Fish (fnf@cygnus.com).
+
+if $tracelevel then {
+	strace $tracelevel
+}
+
+set prms_id 0
+set bug_id 0
+
+# This only works with native configurations
+if ![isnative] then {
+    return
+}
+
+set testfile "print-threads"
+set srcfile ${testfile}.c
+set binfile ${objdir}/${subdir}/${testfile}
+
+# regexp for "horizontal" text (i.e. doesn't include newline or
+# carriage return)
+set horiz "\[^\n\r\]*"
+
+set built_binfile 0
+
+# Default to the usual (only?) -lpthread on GNU/Linux to quiet noise
+if [istarget "*-*-linux*"] then {
+    set possible_libs "-lpthread -lpthreads -lthread"
+} else {
+    set possible_libs "-lpthreads -lpthread -lthread"
+}
+
+set why_msg "unrecognized error"
+foreach lib $possible_libs {
+    set options "debug"
+    lappend options "incdir=${objdir}/${subdir}"
+    lappend options "libs=$lib"
+    set ccout [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable $options]
+    switch -regexp -- $ccout {
+	".*no posix threads support.*" {
+	    set why_msg "missing threads include file"
+	    break
+	}
+	".*cannot open -lpthread.*" {
+	    set why_msg "missing runtime threads library"
+	}
+	".*Can't find library for -lpthread.*" {
+	    set why_msg "missing runtime threads library"
+	}
+	{^$} {
+	    pass "successfully compiled posix threads test case"
+	    set built_binfile 1
+	    break
+	}
+    }
+}
+if {$built_binfile == "0"} {
+    unsupported "Couldn't compile ${srcfile}, ${why_msg}"
+    return -1
+}
+
+# Now we can proceed with the real testing.
+
+# Start with a fresh gdb.
+
+gdb_exit
+gdb_start
+gdb_reinitialize_dir $srcdir/$subdir
+gdb_load ${binfile}
+
+gdb_test "set print sevenbit-strings" ""
+#gdb_test "set print address off" ""
+gdb_test "set width 0" ""
+
+# We'll need this when we send_gdb a ^C to GDB.  Need to do it before we
+# run the program and gdb starts saving and restoring tty states.
+# On Ultrix, we don't need it and it is really slow (because shell_escape
+# doesn't use vfork).
+if ![istarget "*-*-ultrix*"] then {
+    gdb_test "shell stty intr '^C'" ""
+}
+
+proc test_all_threads { name kill } {
+    global gdb_prompt
+
+    set i 0
+    set j 0
+    send_gdb "continue\n"
+    gdb_expect {
+	-re "Breakpoint \[0-9\]+, thread_function \\(arg=.*\\) at .*print-threads.c:\[0-9\]+.*$gdb_prompt" {
+	    set i [expr $i + 1]
+	    pass "Hit thread_function breakpoint, $i ($name)"
+	    send_gdb "continue\n"
+	    exp_continue
+	}
+	-re "Breakpoint \[0-9\]+, .* kill \\(.*\\) .*$gdb_prompt" {
+	    set j [expr $j + 1]
+	    if { $kill == 1 } {
+		pass "Hit kill breakpoint, $j ($name)"
+	    } else {
+		fail "Hit kill breakpoint, $j ($name) (unexpected)"
+	    }
+	    send_gdb "continue\n"
+	    exp_continue
+	}
+	-re "Program exited normally\\.\[\r\n\]+$gdb_prompt" {
+	    pass "program exited normally"
+	    if {$i == 5} {
+		pass "all threads ran once ($name)"
+	    } else {
+		fail "all threads ran once ($name) (total $i threads ran)"
+	    }
+	}
+	-re "$gdb_prompt" {
+	    fail "Running threads ($name) (unknown output)"
+	}
+	timeout {
+	    fail "Running threads ($name) (timeout)"
+	}
+    }
+}
+
+runto_main
+gdb_test "break thread_function" "Breakpoint \[0-9\]+ at 0x\[0-9a-f\]+: file .*print-threads.c, line \[0-9\]*\\."
+gdb_test "set var slow = 0" ""
+test_all_threads "fast" 0
+
+runto_main
+gdb_test "break thread_function" "Breakpoint \[0-9\]+ at 0x\[0-9a-f\]+: file .*print-threads.c, line \[0-9\]*\\." "break thread_function (2)"
+gdb_test "set var slow = 1" ""
+test_all_threads "slow" 0
+
+runto_main
+gdb_test "break thread_function" "Breakpoint \[0-9\]+ at 0x\[0-9a-f\]+: file .*print-threads.c, line \[0-9\]*\\." "break thread_function (3)"
+gdb_test "set var slow = 1" "" "set var slow = 1 (2)"
+gdb_test "break kill" "Breakpoint \[0-9\]+ at 0x\[0-9a-f\]+"
+test_all_threads "slow with kill breakpoint" 1
+
+return 0
--- /dev/null	Wed Dec 31 19:00:00 1969
+++ gdb.threads/schedlock.c	Tue Jun 11 14:56:44 2002
@@ -0,0 +1,44 @@
+#include <stdio.h>
+#include <unistd.h>
+#include <stdlib.h>
+#include <pthread.h>
+
+void *thread_function(void *arg); /* Pointer to function executed by each thread */
+
+#define NUM 5
+
+int args[NUM+1];
+
+int main() {
+    int res;
+    pthread_t threads[NUM];
+    void *thread_result;
+    int i;
+
+    for (i = 0; i < NUM; i++)
+      {
+	args[i] = 1;
+	res = pthread_create(&threads[i], NULL, thread_function, (void *)i);
+      }
+
+    /* schedlock.exp: last thread start.  */
+    args[i] = 1;
+    thread_function ((void *) i);
+
+    exit(EXIT_SUCCESS);
+}
+
+void *thread_function(void *arg) {
+    int my_number = (int) arg;
+    int *myp = &args[my_number];
+
+    /* Don't run forever.  Run just short of it :)  */
+    while (*myp > 0)
+      {
+	/* schedlock.exp: main loop.  */
+	(*myp) ++;
+      }
+
+    pthread_exit(NULL);
+}
+
--- /dev/null	Wed Dec 31 19:00:00 1969
+++ gdb.threads/schedlock.exp	Tue Jun 11 14:56:44 2002
@@ -0,0 +1,374 @@
+# Copyright (C) 1996, 1997, 2002 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 2 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, write to the Free Software
+# Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.  
+
+# Please email any bugs, comments, and/or additions to this file to:
+# bug-gdb@prep.ai.mit.edu
+
+# This file was written by Daniel Jacobowitz <drow@mvista.com>
+# (parts based on pthreads.exp by Fred Fish (fnf@cygnus.com).
+
+if $tracelevel then {
+	strace $tracelevel
+}
+
+set prms_id 0
+set bug_id 0
+
+# This only works with native configurations
+if ![isnative] then {
+    return
+}
+
+set testfile "schedlock"
+set srcfile ${testfile}.c
+set binfile ${objdir}/${subdir}/${testfile}
+
+set built_binfile 0
+
+# Default to the usual (only?) -lpthread on GNU/Linux to quiet noise
+if [istarget "*-*-linux*"] then {
+    set possible_libs "-lpthread -lpthreads -lthread"
+} else {
+    set possible_libs "-lpthreads -lpthread -lthread"
+}
+
+set why_msg "unrecognized error"
+foreach lib $possible_libs {
+    set options "debug"
+    lappend options "incdir=${objdir}/${subdir}"
+    lappend options "libs=$lib"
+    set ccout [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable $options]
+    switch -regexp -- $ccout {
+	".*no posix threads support.*" {
+	    set why_msg "missing threads include file"
+	    break
+	}
+	".*cannot open -lpthread.*" {
+	    set why_msg "missing runtime threads library"
+	}
+	".*Can't find library for -lpthread.*" {
+	    set why_msg "missing runtime threads library"
+	}
+	{^$} {
+	    pass "successfully compiled posix threads test case"
+	    set built_binfile 1
+	    break
+	}
+    }
+}
+if {$built_binfile == "0"} {
+    unsupported "Couldn't compile ${srcfile}, ${why_msg}"
+    return -1
+}
+
+# Now we can proceed with the real testing.
+
+proc get_args { } {
+  global list_count
+  global gdb_prompt
+
+  send_gdb "print args\n"
+  gdb_expect {
+    -re "\\\$\[0-9\]+ = {(\[0-9\]+), (\[0-9\]+), (\[0-9\]+), (\[0-9\]+), (\[0-9\]+), (\[0-9\]+)}.*$gdb_prompt"
+      {
+	set list_count [expr $list_count + 1]
+	pass "listed args ($list_count)"
+	return [list $expect_out(1,string) $expect_out(2,string) $expect_out(3,string) $expect_out(4,string) $expect_out(5,string) $expect_out(6,string)]
+      }
+    -re "$gdb_prompt"
+      {
+	fail "listed args ($list_count) (unknown output)"
+      }
+    timeout
+      {
+	fail "listed args ($list_count) (timeout)"
+      }
+  }
+}
+
+proc stop_process { description } {
+  global gdb_prompt
+
+  # For this to work we must be sure to consume the "Continuing."
+  # message first, or GDB's signal handler may not be in place.
+  after 1000 {send_gdb "\003"}
+  gdb_expect {
+    -re "Program received signal SIGINT.*$gdb_prompt $"
+      {
+	pass $description
+      }
+    timeout
+      {
+	fail "$description (timeout)"
+      }
+  }
+}
+
+proc get_current_thread { description } {
+  global gdb_prompt
+
+  send_gdb "bt\n"
+  gdb_expect {
+    -re "thread_function \\(arg=0x(\[0-9\])\\).*$gdb_prompt $"
+      {
+	pass $description
+	return $expect_out(1,string)
+      }
+    -re "$gdb_prompt $"
+      {
+	fail "$description (unknown output)"
+      }
+    timeout
+      {
+	fail "$description (timeout)"
+      }
+  }
+
+  return
+  send_gdb "info threads\n"
+  gdb_expect {
+    -re "\\* (\[0-9\]+) .*$gdb_prompt $"
+      {
+	pass $description
+	return $expect_out(1,string)
+      }
+    -re "$gdb_prompt $"
+      {
+	fail "$description (unknown output)"
+      }
+    timeout
+      {
+	fail "$description (timeout)"
+      }
+  }
+}
+
+proc my_continue { msg } {
+  send_gdb "continue\n"
+  gdb_expect {
+    -re "Continuing"
+      { pass "continue ($msg)" }
+    timeout
+      { fail "continue ($msg) (timeout)" }
+  }
+
+  stop_process "stop all threads ($msg)"
+
+  # Make sure we're in one of the looping threads.
+  gdb_breakpoint [gdb_get_line_number "schedlock.exp: main loop"]
+  gdb_continue_to_breakpoint "return to loop ($msg)"
+  delete_breakpoints
+}
+
+proc step_ten_loops { msg } {
+    global gdb_prompt
+
+    for {set i 0} {[expr $i < 10]} {set i [expr $i + 1]} {
+	send_gdb "step\n"
+	set other_step 0
+	gdb_expect {
+	    -re ".*myp\\) \\+\\+;\[\r\n\]+$gdb_prompt $" {
+		pass "step to increment ($msg $i)"
+	    }
+	    -re "$gdb_prompt $" {
+		if {$other_step == 0} {
+		    set other_step 1
+		    send_gdb "step\n"
+		    exp_continue
+		} else {
+		    fail "step to increment ($msg $i)"
+		    # FIXME cascade?
+		}
+	    }
+	    timeout {
+		fail "step to increment ($msg $i) (timeout)"
+	    }
+	}
+    }
+}
+
+# Start with a fresh gdb.
+
+gdb_exit
+gdb_start
+gdb_reinitialize_dir $srcdir/$subdir
+
+# We'll need this when we send_gdb a ^C to GDB.  Need to do it before we
+# run the program and gdb starts saving and restoring tty states.
+# On Ultrix, we don't need it and it is really slow (because shell_escape
+# doesn't use vfork).
+if ![istarget "*-*-ultrix*"] then {
+    gdb_test "shell stty intr '^C'" ""
+}
+
+gdb_load ${binfile}
+
+gdb_test "set print sevenbit-strings" ""
+gdb_test "set width 0" ""
+
+runto_main
+
+# See if scheduler locking is available on this target.
+send_gdb "set scheduler-locking off\n"
+global gdb_prompt
+gdb_expect {
+  -re "Target .* cannot support this command"
+    {
+      unsupported "target does not support scheduler locking"
+      return
+    }
+  -re "$gdb_prompt $"
+    {
+      pass "scheduler locking set to none"
+    }
+  timeout
+    {
+      unsupported "target does not support scheduler locking (timeout)"
+      return
+    }
+}
+
+gdb_breakpoint [gdb_get_line_number "schedlock.exp: last thread start"]
+gdb_continue_to_breakpoint "all threads started"
+
+global list_count
+set list_count 0
+
+set start_args [get_args]
+
+# First make sure that all threads are alive.
+my_continue "initial"
+
+set cont_args [get_args]
+
+for {set i 0} {[expr $i < 6]} {set i [expr $i + 1]} {
+  if {[lindex $start_args $i] == [lindex $cont_args $i]} {
+    fail "thread $i ran (didn't run)"
+  } else {
+    pass "thread $i ran"
+  }
+}
+
+# We can't change threads, unfortunately, in current GDB.  Use
+# whichever we stopped in.
+set curthread [get_current_thread "find current thread (1)"]
+
+
+
+
+# Test stepping without scheduler locking.
+gdb_test "set scheduler-locking off"  ""
+
+step_ten_loops "unlocked"
+
+# Make sure we're still in the same thread.
+set newthread [get_current_thread "find current thread (2)"]
+if {$curthread == $newthread} {
+    pass "step without lock does not change thread"
+} else {
+    fail "step without lock does not change thread (switched to thread $newthread)"
+}
+
+set start_args $cont_args
+set cont_args [get_args]
+
+for {set i 0} {[expr $i < 6]} {set i [expr $i + 1]} {
+  if {[lindex $start_args $i] == [lindex $cont_args $i]} {
+    if {$i == $curthread} {
+      fail "current thread stepped (didn't run)"
+    } else {
+      fail "other thread $i ran (didn't run) (1)"
+    }
+  } else {
+    if {$i == $curthread} {
+	if {[lindex $start_args $i] == [expr [lindex $cont_args $i] - 10]} {
+	    pass "current thread stepped"
+	} else {
+	    fail "current thread stepped (wrong amount)"
+	}
+    } else {
+      pass "other thread $i ran (1)"
+    }
+  }
+}
+
+# Test continue with scheduler locking
+gdb_test "set scheduler-locking on"  ""
+
+my_continue "with lock"
+
+# Make sure we're still in the same thread.
+set newthread [get_current_thread "find current thread (3)"]
+if {$curthread == $newthread} {
+    pass "continue with lock does not change thread"
+} else {
+    fail "continue with lock does not change thread (switched to thread $newthread)"
+}
+
+set start_args $cont_args
+set cont_args [get_args]
+
+for {set i 0} {[expr $i < 6]} {set i [expr $i + 1]} {
+  if {[lindex $start_args $i] == [lindex $cont_args $i]} {
+    if {$i == $curthread} {
+      fail "current thread ran (didn't run)"
+    } else {
+      pass "other thread $i didn't run"
+    }
+  } else {
+    if {$i == $curthread} {
+      pass "current thread ran"
+    } else {
+      fail "other thread $i didn't run (ran)"
+    }
+  }
+}
+
+# Test stepping with scheduler locking
+step_ten_loops "locked"
+
+# Make sure we're still in the same thread.
+set newthread [get_current_thread "find current thread (2)"]
+if {$curthread == $newthread} {
+    pass "step with lock does not change thread"
+} else {
+    fail "step with lock does not change thread (switched to thread $newthread)"
+}
+
+set start_args $cont_args
+set cont_args [get_args]
+
+for {set i 0} {[expr $i < 6]} {set i [expr $i + 1]} {
+  if {[lindex $start_args $i] == [lindex $cont_args $i]} {
+    if {$i == $curthread} {
+      fail "current thread stepped locked (didn't run)"
+    } else {
+      pass "other thread $i didn't run (stepping)"
+    }
+  } else {
+    if {$i == $curthread} {
+	if {[lindex $start_args $i] == [expr [lindex $cont_args $i] - 10]} {
+	    pass "current thread stepped locked"
+	} else {
+	    fail "current thread stepped locked (wrong amount)"
+	}
+    } else {
+      fail "other thread $i didn't run (stepping) (ran)"
+    }
+  }
+}
+
+return 0

Re: RFA (threads testsuite): More thread tests

[Fernando Nasser]

Daniel Jacobowitz wrote:
> 
> Here's two tests I had lying around from when I developed the gdbserver
> threads support.  Gdbserver passes them with flying colors (if you use my
> other patch which lets gdbserver run tests properly).  GDB shows a couple of
> problems, unpredictably (not always repeatable).
> 

Can you elaborate on what these problems are and on which host you've
seen them?


> OK to add these?
> 

Michael Snyder, have you seen these?


Daniel, in the meanwhile please fix your CL entries.  
After the "New file." please briefly say what the new files are for.

Also, we need some comments in the new testfile headers saying
what the tests are testing.  So far we have to infer from the file names
;-)


Are there some lines missing in your  get_current_thread  proc?
The second part is after a return that seem orphan in there.


AFAICT the rest looks very nice.  I am really concerned with the
intermittent failures you've mentioned though.

Regards,
Fernando


> --
> Daniel Jacobowitz                           Carnegie Mellon University
> MontaVista Software                         Debian GNU/Linux Developer
> 
> 2002-07-09  Daniel Jacobowitz  <drow@mvista.com>
> 
>         * gdb.threads/print-threads.c: New file.
>         * gdb.threads/print-threads.exp: New file.
>         * gdb.threads/schedlock.c: New file.
>         * gdb.threads/schedlock.exp: New file.
> 
> --- /dev/null   Wed Dec 31 19:00:00 1969
> +++ gdb.threads/print-threads.c Tue Jun 11 14:56:44 2002
> @@ -0,0 +1,58 @@
> +#include <stdio.h>
> +#include <unistd.h>
> +#include <stdlib.h>
> +#include <pthread.h>
> +
> +void *thread_function(void *arg); /* Pointer to function executed by each thread */
> +
> +int slow = 0;
> +
> +#define NUM 5
> +
> +int main() {
> +    int res;
> +    pthread_t threads[NUM];
> +    void *thread_result;
> +    int args[NUM];
> +    int i;
> +
> +    for (i = 0; i < NUM; i++)
> +      {
> +       args[i] = i;
> +       res = pthread_create(&threads[i], NULL, thread_function, (void *)&args[i]);
> +      }
> +
> +    for (i = 0; i < NUM; i++)
> +      res = pthread_join(threads[i], &thread_result);
> +
> +    printf ("Done\n");
> +
> +    if (slow)
> +      sleep (4);
> +
> +    exit(EXIT_SUCCESS);
> +}
> +
> +void *thread_function(void *arg) {
> +    int my_number = *(int *)arg;
> +    int rand_num;
> +
> +    printf ("Print 1, thread %d\n", my_number);
> +    sleep (1);
> +
> +    if (slow)
> +      {
> +       printf ("Print 2, thread %d\n", my_number);
> +       sleep (1);
> +       printf ("Print 3, thread %d\n", my_number);
> +       sleep (1);
> +       printf ("Print 4, thread %d\n", my_number);
> +       sleep (1);
> +       printf ("Print 5, thread %d\n", my_number);
> +       sleep (1);
> +      }
> +
> +    printf("Bye from %d\n", my_number);
> +    pthread_exit(NULL);
> +}
> +
> --- /dev/null   Wed Dec 31 19:00:00 1969
> +++ gdb.threads/print-threads.exp       Tue Jun 11 14:56:44 2002
> @@ -0,0 +1,158 @@
> +# Copyright (C) 1996, 1997, 2002 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 2 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, write to the Free Software
> +# Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
> +
> +# Please email any bugs, comments, and/or additions to this file to:
> +# bug-gdb@prep.ai.mit.edu
> +
> +# This file was written by Daniel Jacobowitz <drow@mvista.com>
> +# (parts based on pthreads.exp by Fred Fish (fnf@cygnus.com).
> +
> +if $tracelevel then {
> +       strace $tracelevel
> +}
> +
> +set prms_id 0
> +set bug_id 0
> +
> +# This only works with native configurations
> +if ![isnative] then {
> +    return
> +}
> +
> +set testfile "print-threads"
> +set srcfile ${testfile}.c
> +set binfile ${objdir}/${subdir}/${testfile}
> +
> +# regexp for "horizontal" text (i.e. doesn't include newline or
> +# carriage return)
> +set horiz "\[^\n\r\]*"
> +
> +set built_binfile 0
> +
> +# Default to the usual (only?) -lpthread on GNU/Linux to quiet noise
> +if [istarget "*-*-linux*"] then {
> +    set possible_libs "-lpthread -lpthreads -lthread"
> +} else {
> +    set possible_libs "-lpthreads -lpthread -lthread"
> +}
> +
> +set why_msg "unrecognized error"
> +foreach lib $possible_libs {
> +    set options "debug"
> +    lappend options "incdir=${objdir}/${subdir}"
> +    lappend options "libs=$lib"
> +    set ccout [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable $options]
> +    switch -regexp -- $ccout {
> +       ".*no posix threads support.*" {
> +           set why_msg "missing threads include file"
> +           break
> +       }
> +       ".*cannot open -lpthread.*" {
> +           set why_msg "missing runtime threads library"
> +       }
> +       ".*Can't find library for -lpthread.*" {
> +           set why_msg "missing runtime threads library"
> +       }
> +       {^$} {
> +           pass "successfully compiled posix threads test case"
> +           set built_binfile 1
> +           break
> +       }
> +    }
> +}
> +if {$built_binfile == "0"} {
> +    unsupported "Couldn't compile ${srcfile}, ${why_msg}"
> +    return -1
> +}
> +
> +# Now we can proceed with the real testing.
> +
> +# Start with a fresh gdb.
> +
> +gdb_exit
> +gdb_start
> +gdb_reinitialize_dir $srcdir/$subdir
> +gdb_load ${binfile}
> +
> +gdb_test "set print sevenbit-strings" ""
> +#gdb_test "set print address off" ""
> +gdb_test "set width 0" ""
> +
> +# We'll need this when we send_gdb a ^C to GDB.  Need to do it before we
> +# run the program and gdb starts saving and restoring tty states.
> +# On Ultrix, we don't need it and it is really slow (because shell_escape
> +# doesn't use vfork).
> +if ![istarget "*-*-ultrix*"] then {
> +    gdb_test "shell stty intr '^C'" ""
> +}
> +
> +proc test_all_threads { name kill } {
> +    global gdb_prompt
> +
> +    set i 0
> +    set j 0
> +    send_gdb "continue\n"
> +    gdb_expect {
> +       -re "Breakpoint \[0-9\]+, thread_function \\(arg=.*\\) at .*print-threads.c:\[0-9\]+.*$gdb_prompt" {
> +           set i [expr $i + 1]
> +           pass "Hit thread_function breakpoint, $i ($name)"
> +           send_gdb "continue\n"
> +           exp_continue
> +       }
> +       -re "Breakpoint \[0-9\]+, .* kill \\(.*\\) .*$gdb_prompt" {
> +           set j [expr $j + 1]
> +           if { $kill == 1 } {
> +               pass "Hit kill breakpoint, $j ($name)"
> +           } else {
> +               fail "Hit kill breakpoint, $j ($name) (unexpected)"
> +           }
> +           send_gdb "continue\n"
> +           exp_continue
> +       }
> +       -re "Program exited normally\\.\[\r\n\]+$gdb_prompt" {
> +           pass "program exited normally"
> +           if {$i == 5} {
> +               pass "all threads ran once ($name)"
> +           } else {
> +               fail "all threads ran once ($name) (total $i threads ran)"
> +           }
> +       }
> +       -re "$gdb_prompt" {
> +           fail "Running threads ($name) (unknown output)"
> +       }
> +       timeout {
> +           fail "Running threads ($name) (timeout)"
> +       }
> +    }
> +}
> +
> +runto_main
> +gdb_test "break thread_function" "Breakpoint \[0-9\]+ at 0x\[0-9a-f\]+: file .*print-threads.c, line \[0-9\]*\\."
> +gdb_test "set var slow = 0" ""
> +test_all_threads "fast" 0
> +
> +runto_main
> +gdb_test "break thread_function" "Breakpoint \[0-9\]+ at 0x\[0-9a-f\]+: file .*print-threads.c, line \[0-9\]*\\." "break thread_function (2)"
> +gdb_test "set var slow = 1" ""
> +test_all_threads "slow" 0
> +
> +runto_main
> +gdb_test "break thread_function" "Breakpoint \[0-9\]+ at 0x\[0-9a-f\]+: file .*print-threads.c, line \[0-9\]*\\." "break thread_function (3)"
> +gdb_test "set var slow = 1" "" "set var slow = 1 (2)"
> +gdb_test "break kill" "Breakpoint \[0-9\]+ at 0x\[0-9a-f\]+"
> +test_all_threads "slow with kill breakpoint" 1
> +
> +return 0
> --- /dev/null   Wed Dec 31 19:00:00 1969
> +++ gdb.threads/schedlock.c     Tue Jun 11 14:56:44 2002
> @@ -0,0 +1,44 @@
> +#include <stdio.h>
> +#include <unistd.h>
> +#include <stdlib.h>
> +#include <pthread.h>
> +
> +void *thread_function(void *arg); /* Pointer to function executed by each thread */
> +
> +#define NUM 5
> +
> +int args[NUM+1];
> +
> +int main() {
> +    int res;
> +    pthread_t threads[NUM];
> +    void *thread_result;
> +    int i;
> +
> +    for (i = 0; i < NUM; i++)
> +      {
> +       args[i] = 1;
> +       res = pthread_create(&threads[i], NULL, thread_function, (void *)i);
> +      }
> +
> +    /* schedlock.exp: last thread start.  */
> +    args[i] = 1;
> +    thread_function ((void *) i);
> +
> +    exit(EXIT_SUCCESS);
> +}
> +
> +void *thread_function(void *arg) {
> +    int my_number = (int) arg;
> +    int *myp = &args[my_number];
> +
> +    /* Don't run forever.  Run just short of it :)  */
> +    while (*myp > 0)
> +      {
> +       /* schedlock.exp: main loop.  */
> +       (*myp) ++;
> +      }
> +
> +    pthread_exit(NULL);
> +}
> +
> --- /dev/null   Wed Dec 31 19:00:00 1969
> +++ gdb.threads/schedlock.exp   Tue Jun 11 14:56:44 2002
> @@ -0,0 +1,374 @@
> +# Copyright (C) 1996, 1997, 2002 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 2 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, write to the Free Software
> +# Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
> +
> +# Please email any bugs, comments, and/or additions to this file to:
> +# bug-gdb@prep.ai.mit.edu
> +
> +# This file was written by Daniel Jacobowitz <drow@mvista.com>
> +# (parts based on pthreads.exp by Fred Fish (fnf@cygnus.com).
> +
> +if $tracelevel then {
> +       strace $tracelevel
> +}
> +
> +set prms_id 0
> +set bug_id 0
> +
> +# This only works with native configurations
> +if ![isnative] then {
> +    return
> +}
> +
> +set testfile "schedlock"
> +set srcfile ${testfile}.c
> +set binfile ${objdir}/${subdir}/${testfile}
> +
> +set built_binfile 0
> +
> +# Default to the usual (only?) -lpthread on GNU/Linux to quiet noise
> +if [istarget "*-*-linux*"] then {
> +    set possible_libs "-lpthread -lpthreads -lthread"
> +} else {
> +    set possible_libs "-lpthreads -lpthread -lthread"
> +}
> +
> +set why_msg "unrecognized error"
> +foreach lib $possible_libs {
> +    set options "debug"
> +    lappend options "incdir=${objdir}/${subdir}"
> +    lappend options "libs=$lib"
> +    set ccout [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable $options]
> +    switch -regexp -- $ccout {
> +       ".*no posix threads support.*" {
> +           set why_msg "missing threads include file"
> +           break
> +       }
> +       ".*cannot open -lpthread.*" {
> +           set why_msg "missing runtime threads library"
> +       }
> +       ".*Can't find library for -lpthread.*" {
> +           set why_msg "missing runtime threads library"
> +       }
> +       {^$} {
> +           pass "successfully compiled posix threads test case"
> +           set built_binfile 1
> +           break
> +       }
> +    }
> +}
> +if {$built_binfile == "0"} {
> +    unsupported "Couldn't compile ${srcfile}, ${why_msg}"
> +    return -1
> +}
> +
> +# Now we can proceed with the real testing.
> +
> +proc get_args { } {
> +  global list_count
> +  global gdb_prompt
> +
> +  send_gdb "print args\n"
> +  gdb_expect {
> +    -re "\\\$\[0-9\]+ = {(\[0-9\]+), (\[0-9\]+), (\[0-9\]+), (\[0-9\]+), (\[0-9\]+), (\[0-9\]+)}.*$gdb_prompt"
> +      {
> +       set list_count [expr $list_count + 1]
> +       pass "listed args ($list_count)"
> +       return [list $expect_out(1,string) $expect_out(2,string) $expect_out(3,string) $expect_out(4,string) $expect_out(5,string) $expect_out(6,string)]
> +      }
> +    -re "$gdb_prompt"
> +      {
> +       fail "listed args ($list_count) (unknown output)"
> +      }
> +    timeout
> +      {
> +       fail "listed args ($list_count) (timeout)"
> +      }
> +  }
> +}
> +
> +proc stop_process { description } {
> +  global gdb_prompt
> +
> +  # For this to work we must be sure to consume the "Continuing."
> +  # message first, or GDB's signal handler may not be in place.
> +  after 1000 {send_gdb "\003"}
> +  gdb_expect {
> +    -re "Program received signal SIGINT.*$gdb_prompt $"
> +      {
> +       pass $description
> +      }
> +    timeout
> +      {
> +       fail "$description (timeout)"
> +      }
> +  }
> +}
> +
> +proc get_current_thread { description } {
> +  global gdb_prompt
> +
> +  send_gdb "bt\n"
> +  gdb_expect {
> +    -re "thread_function \\(arg=0x(\[0-9\])\\).*$gdb_prompt $"
> +      {
> +       pass $description
> +       return $expect_out(1,string)
> +      }
> +    -re "$gdb_prompt $"
> +      {
> +       fail "$description (unknown output)"
> +      }
> +    timeout
> +      {
> +       fail "$description (timeout)"
> +      }
> +  }
> +
> +  return
> +  send_gdb "info threads\n"
> +  gdb_expect {
> +    -re "\\* (\[0-9\]+) .*$gdb_prompt $"
> +      {
> +       pass $description
> +       return $expect_out(1,string)
> +      }
> +    -re "$gdb_prompt $"
> +      {
> +       fail "$description (unknown output)"
> +      }
> +    timeout
> +      {
> +       fail "$description (timeout)"
> +      }
> +  }
> +}
> +
> +proc my_continue { msg } {
> +  send_gdb "continue\n"
> +  gdb_expect {
> +    -re "Continuing"
> +      { pass "continue ($msg)" }
> +    timeout
> +      { fail "continue ($msg) (timeout)" }
> +  }
> +
> +  stop_process "stop all threads ($msg)"
> +
> +  # Make sure we're in one of the looping threads.
> +  gdb_breakpoint [gdb_get_line_number "schedlock.exp: main loop"]
> +  gdb_continue_to_breakpoint "return to loop ($msg)"
> +  delete_breakpoints
> +}
> +
> +proc step_ten_loops { msg } {
> +    global gdb_prompt
> +
> +    for {set i 0} {[expr $i < 10]} {set i [expr $i + 1]} {
> +       send_gdb "step\n"
> +       set other_step 0
> +       gdb_expect {
> +           -re ".*myp\\) \\+\\+;\[\r\n\]+$gdb_prompt $" {
> +               pass "step to increment ($msg $i)"
> +           }
> +           -re "$gdb_prompt $" {
> +               if {$other_step == 0} {
> +                   set other_step 1
> +                   send_gdb "step\n"
> +                   exp_continue
> +               } else {
> +                   fail "step to increment ($msg $i)"
> +                   # FIXME cascade?
> +               }
> +           }
> +           timeout {
> +               fail "step to increment ($msg $i) (timeout)"
> +           }
> +       }
> +    }
> +}
> +
> +# Start with a fresh gdb.
> +
> +gdb_exit
> +gdb_start
> +gdb_reinitialize_dir $srcdir/$subdir
> +
> +# We'll need this when we send_gdb a ^C to GDB.  Need to do it before we
> +# run the program and gdb starts saving and restoring tty states.
> +# On Ultrix, we don't need it and it is really slow (because shell_escape
> +# doesn't use vfork).
> +if ![istarget "*-*-ultrix*"] then {
> +    gdb_test "shell stty intr '^C'" ""
> +}
> +
> +gdb_load ${binfile}
> +
> +gdb_test "set print sevenbit-strings" ""
> +gdb_test "set width 0" ""
> +
> +runto_main
> +
> +# See if scheduler locking is available on this target.
> +send_gdb "set scheduler-locking off\n"
> +global gdb_prompt
> +gdb_expect {
> +  -re "Target .* cannot support this command"
> +    {
> +      unsupported "target does not support scheduler locking"
> +      return
> +    }
> +  -re "$gdb_prompt $"
> +    {
> +      pass "scheduler locking set to none"
> +    }
> +  timeout
> +    {
> +      unsupported "target does not support scheduler locking (timeout)"
> +      return
> +    }
> +}
> +
> +gdb_breakpoint [gdb_get_line_number "schedlock.exp: last thread start"]
> +gdb_continue_to_breakpoint "all threads started"
> +
> +global list_count
> +set list_count 0
> +
> +set start_args [get_args]
> +
> +# First make sure that all threads are alive.
> +my_continue "initial"
> +
> +set cont_args [get_args]
> +
> +for {set i 0} {[expr $i < 6]} {set i [expr $i + 1]} {
> +  if {[lindex $start_args $i] == [lindex $cont_args $i]} {
> +    fail "thread $i ran (didn't run)"
> +  } else {
> +    pass "thread $i ran"
> +  }
> +}
> +
> +# We can't change threads, unfortunately, in current GDB.  Use
> +# whichever we stopped in.
> +set curthread [get_current_thread "find current thread (1)"]
> +
> +
> +
> +
> +# Test stepping without scheduler locking.
> +gdb_test "set scheduler-locking off"  ""
> +
> +step_ten_loops "unlocked"
> +
> +# Make sure we're still in the same thread.
> +set newthread [get_current_thread "find current thread (2)"]
> +if {$curthread == $newthread} {
> +    pass "step without lock does not change thread"
> +} else {
> +    fail "step without lock does not change thread (switched to thread $newthread)"
> +}
> +
> +set start_args $cont_args
> +set cont_args [get_args]
> +
> +for {set i 0} {[expr $i < 6]} {set i [expr $i + 1]} {
> +  if {[lindex $start_args $i] == [lindex $cont_args $i]} {
> +    if {$i == $curthread} {
> +      fail "current thread stepped (didn't run)"
> +    } else {
> +      fail "other thread $i ran (didn't run) (1)"
> +    }
> +  } else {
> +    if {$i == $curthread} {
> +       if {[lindex $start_args $i] == [expr [lindex $cont_args $i] - 10]} {
> +           pass "current thread stepped"
> +       } else {
> +           fail "current thread stepped (wrong amount)"
> +       }
> +    } else {
> +      pass "other thread $i ran (1)"
> +    }
> +  }
> +}
> +
> +# Test continue with scheduler locking
> +gdb_test "set scheduler-locking on"  ""
> +
> +my_continue "with lock"
> +
> +# Make sure we're still in the same thread.
> +set newthread [get_current_thread "find current thread (3)"]
> +if {$curthread == $newthread} {
> +    pass "continue with lock does not change thread"
> +} else {
> +    fail "continue with lock does not change thread (switched to thread $newthread)"
> +}
> +
> +set start_args $cont_args
> +set cont_args [get_args]
> +
> +for {set i 0} {[expr $i < 6]} {set i [expr $i + 1]} {
> +  if {[lindex $start_args $i] == [lindex $cont_args $i]} {
> +    if {$i == $curthread} {
> +      fail "current thread ran (didn't run)"
> +    } else {
> +      pass "other thread $i didn't run"
> +    }
> +  } else {
> +    if {$i == $curthread} {
> +      pass "current thread ran"
> +    } else {
> +      fail "other thread $i didn't run (ran)"
> +    }
> +  }
> +}
> +
> +# Test stepping with scheduler locking
> +step_ten_loops "locked"
> +
> +# Make sure we're still in the same thread.
> +set newthread [get_current_thread "find current thread (2)"]
> +if {$curthread == $newthread} {
> +    pass "step with lock does not change thread"
> +} else {
> +    fail "step with lock does not change thread (switched to thread $newthread)"
> +}
> +
> +set start_args $cont_args
> +set cont_args [get_args]
> +
> +for {set i 0} {[expr $i < 6]} {set i [expr $i + 1]} {
> +  if {[lindex $start_args $i] == [lindex $cont_args $i]} {
> +    if {$i == $curthread} {
> +      fail "current thread stepped locked (didn't run)"
> +    } else {
> +      pass "other thread $i didn't run (stepping)"
> +    }
> +  } else {
> +    if {$i == $curthread} {
> +       if {[lindex $start_args $i] == [expr [lindex $cont_args $i] - 10]} {
> +           pass "current thread stepped locked"
> +       } else {
> +           fail "current thread stepped locked (wrong amount)"
> +       }
> +    } else {
> +      fail "other thread $i didn't run (stepping) (ran)"
> +    }
> +  }
> +}
> +
> +return 0

-- 
Fernando Nasser
Red Hat Canada Ltd.                     E-Mail:  fnasser@redhat.com
2323 Yonge Street, Suite #300
Toronto, Ontario   M4P 2C9

Re: RFA (threads testsuite): More thread tests

[Daniel Jacobowitz]

On Tue, Aug 13, 2002 at 06:41:25PM -0400, Fernando Nasser wrote:
> Daniel Jacobowitz wrote:
> > 
> > Here's two tests I had lying around from when I developed the gdbserver
> > threads support.  Gdbserver passes them with flying colors (if you use my
> > other patch which lets gdbserver run tests properly).  GDB shows a couple of
> > problems, unpredictably (not always repeatable).
> > 
> 
> Can you elaborate on what these problems are and on which host you've
> seen them?

They vary greatly, unfortunately.  They're all reproducible but not
necessarily repeatable; run the tests a half-dozen times and you'll
probably see them.  My system is SMP which makes things even more
timing-dependant.

From one run just now:
FAIL: gdb.threads/print-threads.exp: Running threads (slow with kill breakpoint) (unknown output)

FAIL: gdb.threads/schedlock.exp: continue with lock does not change thread (switched to thread 4)
FAIL: gdb.threads/schedlock.exp: other thread 0 didn't run (ran)
FAIL: gdb.threads/schedlock.exp: other thread 2 didn't run (ran)
FAIL: gdb.threads/schedlock.exp: other thread 3 didn't run (ran)
FAIL: gdb.threads/schedlock.exp: other thread 4 didn't run (ran)
FAIL: gdb.threads/schedlock.exp: other thread 5 didn't run (ran)

FAIL: gdb.threads/schedlock.exp: step with lock does not change thread (switched to thread 4)
FAIL: gdb.threads/schedlock.exp: current thread stepped locked (didn't run)
FAIL: gdb.threads/schedlock.exp: other thread 4 didn't run (stepping) (ran)

The first one is:
Cannot find thread 4101: no thread to satisfy query

The second set are some problem with set scheduler-locking that causes
GDB to lose control of some threads.  I don't remember why this
happens, if I ever knew.  The third set involve stopping in the wrong
thread.


You may have seen that, about three months ago now (?) I committed a
multi-thread support package for gdbserver.  These tests were written
while I was developing it.  If you run the testsuite against gdbserver
(which requires a little fiddling with the test harness still...) all
of these tests will pass.  They sure didn't when I was working on it,
though, which is why I wrote them :)

> > OK to add these?
> > 
> 
> Michael Snyder, have you seen these?
> 
> 
> Daniel, in the meanwhile please fix your CL entries.  
> After the "New file." please briefly say what the new files are for.
> 
> Also, we need some comments in the new testfile headers saying
> what the tests are testing.  So far we have to infer from the file names
> ;-)

In my defense, the first point is against GNU ChangeLog policy - you
aren't supposed to explain the "why" of things in ChangeLogs.  I don't
really like that policy, so I'll be glad to elaborate :)  I'll take
care of the comments too.

> Are there some lines missing in your  get_current_thread  proc?
> The second part is after a return that seem orphan in there.

Extra lines, actually.  Only the part before the return is supposed to
be there, thanks for spotting that.

Thanks for the comments; I'll repost after Michael gets a chance to
comment on the tests themselves.

-- 
Daniel Jacobowitz
MontaVista Software                         Debian GNU/Linux Developer

Re: RFA (threads testsuite): More thread tests

[Michael Snyder]

Daniel Jacobowitz wrote:
> 
> Here's two tests I had lying around from when I developed the gdbserver
> threads support.  Gdbserver passes them with flying colors (if you use my
> other patch which lets gdbserver run tests properly).  GDB shows a couple of
> problems, unpredictably (not always repeatable).
> 
> OK to add these?

Hi Joel, 

I understand the point of schedlock.exp, but what's the point of
print-threads.exp?  What is it that you're testing?

Michael

> 
> --
> Daniel Jacobowitz                           Carnegie Mellon University
> MontaVista Software                         Debian GNU/Linux Developer
> 
> 2002-07-09  Daniel Jacobowitz  <drow@mvista.com>
> 
>         * gdb.threads/print-threads.c: New file.
>         * gdb.threads/print-threads.exp: New file.
>         * gdb.threads/schedlock.c: New file.
>         * gdb.threads/schedlock.exp: New file.
> 
> --- /dev/null   Wed Dec 31 19:00:00 1969
> +++ gdb.threads/print-threads.c Tue Jun 11 14:56:44 2002
> @@ -0,0 +1,58 @@
> +#include <stdio.h>
> +#include <unistd.h>
> +#include <stdlib.h>
> +#include <pthread.h>
> +
> +void *thread_function(void *arg); /* Pointer to function executed by each thread */
> +
> +int slow = 0;
> +
> +#define NUM 5
> +
> +int main() {
> +    int res;
> +    pthread_t threads[NUM];
> +    void *thread_result;
> +    int args[NUM];
> +    int i;
> +
> +    for (i = 0; i < NUM; i++)
> +      {
> +       args[i] = i;
> +       res = pthread_create(&threads[i], NULL, thread_function, (void *)&args[i]);
> +      }
> +
> +    for (i = 0; i < NUM; i++)
> +      res = pthread_join(threads[i], &thread_result);
> +
> +    printf ("Done\n");
> +
> +    if (slow)
> +      sleep (4);
> +
> +    exit(EXIT_SUCCESS);
> +}
> +
> +void *thread_function(void *arg) {
> +    int my_number = *(int *)arg;
> +    int rand_num;
> +
> +    printf ("Print 1, thread %d\n", my_number);
> +    sleep (1);
> +
> +    if (slow)
> +      {
> +       printf ("Print 2, thread %d\n", my_number);
> +       sleep (1);
> +       printf ("Print 3, thread %d\n", my_number);
> +       sleep (1);
> +       printf ("Print 4, thread %d\n", my_number);
> +       sleep (1);
> +       printf ("Print 5, thread %d\n", my_number);
> +       sleep (1);
> +      }
> +
> +    printf("Bye from %d\n", my_number);
> +    pthread_exit(NULL);
> +}
> +
> --- /dev/null   Wed Dec 31 19:00:00 1969
> +++ gdb.threads/print-threads.exp       Tue Jun 11 14:56:44 2002
> @@ -0,0 +1,158 @@
> +# Copyright (C) 1996, 1997, 2002 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 2 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, write to the Free Software
> +# Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
> +
> +# Please email any bugs, comments, and/or additions to this file to:
> +# bug-gdb@prep.ai.mit.edu
> +
> +# This file was written by Daniel Jacobowitz <drow@mvista.com>
> +# (parts based on pthreads.exp by Fred Fish (fnf@cygnus.com).
> +
> +if $tracelevel then {
> +       strace $tracelevel
> +}
> +
> +set prms_id 0
> +set bug_id 0
> +
> +# This only works with native configurations
> +if ![isnative] then {
> +    return
> +}
> +
> +set testfile "print-threads"
> +set srcfile ${testfile}.c
> +set binfile ${objdir}/${subdir}/${testfile}
> +
> +# regexp for "horizontal" text (i.e. doesn't include newline or
> +# carriage return)
> +set horiz "\[^\n\r\]*"
> +
> +set built_binfile 0
> +
> +# Default to the usual (only?) -lpthread on GNU/Linux to quiet noise
> +if [istarget "*-*-linux*"] then {
> +    set possible_libs "-lpthread -lpthreads -lthread"
> +} else {
> +    set possible_libs "-lpthreads -lpthread -lthread"
> +}
> +
> +set why_msg "unrecognized error"
> +foreach lib $possible_libs {
> +    set options "debug"
> +    lappend options "incdir=${objdir}/${subdir}"
> +    lappend options "libs=$lib"
> +    set ccout [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable $options]
> +    switch -regexp -- $ccout {
> +       ".*no posix threads support.*" {
> +           set why_msg "missing threads include file"
> +           break
> +       }
> +       ".*cannot open -lpthread.*" {
> +           set why_msg "missing runtime threads library"
> +       }
> +       ".*Can't find library for -lpthread.*" {
> +           set why_msg "missing runtime threads library"
> +       }
> +       {^$} {
> +           pass "successfully compiled posix threads test case"
> +           set built_binfile 1
> +           break
> +       }
> +    }
> +}
> +if {$built_binfile == "0"} {
> +    unsupported "Couldn't compile ${srcfile}, ${why_msg}"
> +    return -1
> +}
> +
> +# Now we can proceed with the real testing.
> +
> +# Start with a fresh gdb.
> +
> +gdb_exit
> +gdb_start
> +gdb_reinitialize_dir $srcdir/$subdir
> +gdb_load ${binfile}
> +
> +gdb_test "set print sevenbit-strings" ""
> +#gdb_test "set print address off" ""
> +gdb_test "set width 0" ""
> +
> +# We'll need this when we send_gdb a ^C to GDB.  Need to do it before we
> +# run the program and gdb starts saving and restoring tty states.
> +# On Ultrix, we don't need it and it is really slow (because shell_escape
> +# doesn't use vfork).
> +if ![istarget "*-*-ultrix*"] then {
> +    gdb_test "shell stty intr '^C'" ""
> +}
> +
> +proc test_all_threads { name kill } {
> +    global gdb_prompt
> +
> +    set i 0
> +    set j 0
> +    send_gdb "continue\n"
> +    gdb_expect {
> +       -re "Breakpoint \[0-9\]+, thread_function \\(arg=.*\\) at .*print-threads.c:\[0-9\]+.*$gdb_prompt" {
> +           set i [expr $i + 1]
> +           pass "Hit thread_function breakpoint, $i ($name)"
> +           send_gdb "continue\n"
> +           exp_continue
> +       }
> +       -re "Breakpoint \[0-9\]+, .* kill \\(.*\\) .*$gdb_prompt" {
> +           set j [expr $j + 1]
> +           if { $kill == 1 } {
> +               pass "Hit kill breakpoint, $j ($name)"
> +           } else {
> +               fail "Hit kill breakpoint, $j ($name) (unexpected)"
> +           }
> +           send_gdb "continue\n"
> +           exp_continue
> +       }
> +       -re "Program exited normally\\.\[\r\n\]+$gdb_prompt" {
> +           pass "program exited normally"
> +           if {$i == 5} {
> +               pass "all threads ran once ($name)"
> +           } else {
> +               fail "all threads ran once ($name) (total $i threads ran)"
> +           }
> +       }
> +       -re "$gdb_prompt" {
> +           fail "Running threads ($name) (unknown output)"
> +       }
> +       timeout {
> +           fail "Running threads ($name) (timeout)"
> +       }
> +    }
> +}
> +
> +runto_main
> +gdb_test "break thread_function" "Breakpoint \[0-9\]+ at 0x\[0-9a-f\]+: file .*print-threads.c, line \[0-9\]*\\."
> +gdb_test "set var slow = 0" ""
> +test_all_threads "fast" 0
> +
> +runto_main
> +gdb_test "break thread_function" "Breakpoint \[0-9\]+ at 0x\[0-9a-f\]+: file .*print-threads.c, line \[0-9\]*\\." "break thread_function (2)"
> +gdb_test "set var slow = 1" ""
> +test_all_threads "slow" 0
> +
> +runto_main
> +gdb_test "break thread_function" "Breakpoint \[0-9\]+ at 0x\[0-9a-f\]+: file .*print-threads.c, line \[0-9\]*\\." "break thread_function (3)"
> +gdb_test "set var slow = 1" "" "set var slow = 1 (2)"
> +gdb_test "break kill" "Breakpoint \[0-9\]+ at 0x\[0-9a-f\]+"
> +test_all_threads "slow with kill breakpoint" 1
> +
> +return 0
> --- /dev/null   Wed Dec 31 19:00:00 1969
> +++ gdb.threads/schedlock.c     Tue Jun 11 14:56:44 2002
> @@ -0,0 +1,44 @@
> +#include <stdio.h>
> +#include <unistd.h>
> +#include <stdlib.h>
> +#include <pthread.h>
> +
> +void *thread_function(void *arg); /* Pointer to function executed by each thread */
> +
> +#define NUM 5
> +
> +int args[NUM+1];
> +
> +int main() {
> +    int res;
> +    pthread_t threads[NUM];
> +    void *thread_result;
> +    int i;
> +
> +    for (i = 0; i < NUM; i++)
> +      {
> +       args[i] = 1;
> +       res = pthread_create(&threads[i], NULL, thread_function, (void *)i);
> +      }
> +
> +    /* schedlock.exp: last thread start.  */
> +    args[i] = 1;
> +    thread_function ((void *) i);
> +
> +    exit(EXIT_SUCCESS);
> +}
> +
> +void *thread_function(void *arg) {
> +    int my_number = (int) arg;
> +    int *myp = &args[my_number];
> +
> +    /* Don't run forever.  Run just short of it :)  */
> +    while (*myp > 0)
> +      {
> +       /* schedlock.exp: main loop.  */
> +       (*myp) ++;
> +      }
> +
> +    pthread_exit(NULL);
> +}
> +
> --- /dev/null   Wed Dec 31 19:00:00 1969
> +++ gdb.threads/schedlock.exp   Tue Jun 11 14:56:44 2002
> @@ -0,0 +1,374 @@
> +# Copyright (C) 1996, 1997, 2002 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 2 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, write to the Free Software
> +# Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
> +
> +# Please email any bugs, comments, and/or additions to this file to:
> +# bug-gdb@prep.ai.mit.edu
> +
> +# This file was written by Daniel Jacobowitz <drow@mvista.com>
> +# (parts based on pthreads.exp by Fred Fish (fnf@cygnus.com).
> +
> +if $tracelevel then {
> +       strace $tracelevel
> +}
> +
> +set prms_id 0
> +set bug_id 0
> +
> +# This only works with native configurations
> +if ![isnative] then {
> +    return
> +}
> +
> +set testfile "schedlock"
> +set srcfile ${testfile}.c
> +set binfile ${objdir}/${subdir}/${testfile}
> +
> +set built_binfile 0
> +
> +# Default to the usual (only?) -lpthread on GNU/Linux to quiet noise
> +if [istarget "*-*-linux*"] then {
> +    set possible_libs "-lpthread -lpthreads -lthread"
> +} else {
> +    set possible_libs "-lpthreads -lpthread -lthread"
> +}
> +
> +set why_msg "unrecognized error"
> +foreach lib $possible_libs {
> +    set options "debug"
> +    lappend options "incdir=${objdir}/${subdir}"
> +    lappend options "libs=$lib"
> +    set ccout [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable $options]
> +    switch -regexp -- $ccout {
> +       ".*no posix threads support.*" {
> +           set why_msg "missing threads include file"
> +           break
> +       }
> +       ".*cannot open -lpthread.*" {
> +           set why_msg "missing runtime threads library"
> +       }
> +       ".*Can't find library for -lpthread.*" {
> +           set why_msg "missing runtime threads library"
> +       }
> +       {^$} {
> +           pass "successfully compiled posix threads test case"
> +           set built_binfile 1
> +           break
> +       }
> +    }
> +}
> +if {$built_binfile == "0"} {
> +    unsupported "Couldn't compile ${srcfile}, ${why_msg}"
> +    return -1
> +}
> +
> +# Now we can proceed with the real testing.
> +
> +proc get_args { } {
> +  global list_count
> +  global gdb_prompt
> +
> +  send_gdb "print args\n"
> +  gdb_expect {
> +    -re "\\\$\[0-9\]+ = {(\[0-9\]+), (\[0-9\]+), (\[0-9\]+), (\[0-9\]+), (\[0-9\]+), (\[0-9\]+)}.*$gdb_prompt"
> +      {
> +       set list_count [expr $list_count + 1]
> +       pass "listed args ($list_count)"
> +       return [list $expect_out(1,string) $expect_out(2,string) $expect_out(3,string) $expect_out(4,string) $expect_out(5,string) $expect_out(6,string)]
> +      }
> +    -re "$gdb_prompt"
> +      {
> +       fail "listed args ($list_count) (unknown output)"
> +      }
> +    timeout
> +      {
> +       fail "listed args ($list_count) (timeout)"
> +      }
> +  }
> +}
> +
> +proc stop_process { description } {
> +  global gdb_prompt
> +
> +  # For this to work we must be sure to consume the "Continuing."
> +  # message first, or GDB's signal handler may not be in place.
> +  after 1000 {send_gdb "\003"}
> +  gdb_expect {
> +    -re "Program received signal SIGINT.*$gdb_prompt $"
> +      {
> +       pass $description
> +      }
> +    timeout
> +      {
> +       fail "$description (timeout)"
> +      }
> +  }
> +}
> +
> +proc get_current_thread { description } {
> +  global gdb_prompt
> +
> +  send_gdb "bt\n"
> +  gdb_expect {
> +    -re "thread_function \\(arg=0x(\[0-9\])\\).*$gdb_prompt $"
> +      {
> +       pass $description
> +       return $expect_out(1,string)
> +      }
> +    -re "$gdb_prompt $"
> +      {
> +       fail "$description (unknown output)"
> +      }
> +    timeout
> +      {
> +       fail "$description (timeout)"
> +      }
> +  }
> +
> +  return
> +  send_gdb "info threads\n"
> +  gdb_expect {
> +    -re "\\* (\[0-9\]+) .*$gdb_prompt $"
> +      {
> +       pass $description
> +       return $expect_out(1,string)
> +      }
> +    -re "$gdb_prompt $"
> +      {
> +       fail "$description (unknown output)"
> +      }
> +    timeout
> +      {
> +       fail "$description (timeout)"
> +      }
> +  }
> +}
> +
> +proc my_continue { msg } {
> +  send_gdb "continue\n"
> +  gdb_expect {
> +    -re "Continuing"
> +      { pass "continue ($msg)" }
> +    timeout
> +      { fail "continue ($msg) (timeout)" }
> +  }
> +
> +  stop_process "stop all threads ($msg)"
> +
> +  # Make sure we're in one of the looping threads.
> +  gdb_breakpoint [gdb_get_line_number "schedlock.exp: main loop"]
> +  gdb_continue_to_breakpoint "return to loop ($msg)"
> +  delete_breakpoints
> +}
> +
> +proc step_ten_loops { msg } {
> +    global gdb_prompt
> +
> +    for {set i 0} {[expr $i < 10]} {set i [expr $i + 1]} {
> +       send_gdb "step\n"
> +       set other_step 0
> +       gdb_expect {
> +           -re ".*myp\\) \\+\\+;\[\r\n\]+$gdb_prompt $" {
> +               pass "step to increment ($msg $i)"
> +           }
> +           -re "$gdb_prompt $" {
> +               if {$other_step == 0} {
> +                   set other_step 1
> +                   send_gdb "step\n"
> +                   exp_continue
> +               } else {
> +                   fail "step to increment ($msg $i)"
> +                   # FIXME cascade?
> +               }
> +           }
> +           timeout {
> +               fail "step to increment ($msg $i) (timeout)"
> +           }
> +       }
> +    }
> +}
> +
> +# Start with a fresh gdb.
> +
> +gdb_exit
> +gdb_start
> +gdb_reinitialize_dir $srcdir/$subdir
> +
> +# We'll need this when we send_gdb a ^C to GDB.  Need to do it before we
> +# run the program and gdb starts saving and restoring tty states.
> +# On Ultrix, we don't need it and it is really slow (because shell_escape
> +# doesn't use vfork).
> +if ![istarget "*-*-ultrix*"] then {
> +    gdb_test "shell stty intr '^C'" ""
> +}
> +
> +gdb_load ${binfile}
> +
> +gdb_test "set print sevenbit-strings" ""
> +gdb_test "set width 0" ""
> +
> +runto_main
> +
> +# See if scheduler locking is available on this target.
> +send_gdb "set scheduler-locking off\n"
> +global gdb_prompt
> +gdb_expect {
> +  -re "Target .* cannot support this command"
> +    {
> +      unsupported "target does not support scheduler locking"
> +      return
> +    }
> +  -re "$gdb_prompt $"
> +    {
> +      pass "scheduler locking set to none"
> +    }
> +  timeout
> +    {
> +      unsupported "target does not support scheduler locking (timeout)"
> +      return
> +    }
> +}
> +
> +gdb_breakpoint [gdb_get_line_number "schedlock.exp: last thread start"]
> +gdb_continue_to_breakpoint "all threads started"
> +
> +global list_count
> +set list_count 0
> +
> +set start_args [get_args]
> +
> +# First make sure that all threads are alive.
> +my_continue "initial"
> +
> +set cont_args [get_args]
> +
> +for {set i 0} {[expr $i < 6]} {set i [expr $i + 1]} {
> +  if {[lindex $start_args $i] == [lindex $cont_args $i]} {
> +    fail "thread $i ran (didn't run)"
> +  } else {
> +    pass "thread $i ran"
> +  }
> +}
> +
> +# We can't change threads, unfortunately, in current GDB.  Use
> +# whichever we stopped in.
> +set curthread [get_current_thread "find current thread (1)"]
> +
> +
> +
> +
> +# Test stepping without scheduler locking.
> +gdb_test "set scheduler-locking off"  ""
> +
> +step_ten_loops "unlocked"
> +
> +# Make sure we're still in the same thread.
> +set newthread [get_current_thread "find current thread (2)"]
> +if {$curthread == $newthread} {
> +    pass "step without lock does not change thread"
> +} else {
> +    fail "step without lock does not change thread (switched to thread $newthread)"
> +}
> +
> +set start_args $cont_args
> +set cont_args [get_args]
> +
> +for {set i 0} {[expr $i < 6]} {set i [expr $i + 1]} {
> +  if {[lindex $start_args $i] == [lindex $cont_args $i]} {
> +    if {$i == $curthread} {
> +      fail "current thread stepped (didn't run)"
> +    } else {
> +      fail "other thread $i ran (didn't run) (1)"
> +    }
> +  } else {
> +    if {$i == $curthread} {
> +       if {[lindex $start_args $i] == [expr [lindex $cont_args $i] - 10]} {
> +           pass "current thread stepped"
> +       } else {
> +           fail "current thread stepped (wrong amount)"
> +       }
> +    } else {
> +      pass "other thread $i ran (1)"
> +    }
> +  }
> +}
> +
> +# Test continue with scheduler locking
> +gdb_test "set scheduler-locking on"  ""
> +
> +my_continue "with lock"
> +
> +# Make sure we're still in the same thread.
> +set newthread [get_current_thread "find current thread (3)"]
> +if {$curthread == $newthread} {
> +    pass "continue with lock does not change thread"
> +} else {
> +    fail "continue with lock does not change thread (switched to thread $newthread)"
> +}
> +
> +set start_args $cont_args
> +set cont_args [get_args]
> +
> +for {set i 0} {[expr $i < 6]} {set i [expr $i + 1]} {
> +  if {[lindex $start_args $i] == [lindex $cont_args $i]} {
> +    if {$i == $curthread} {
> +      fail "current thread ran (didn't run)"
> +    } else {
> +      pass "other thread $i didn't run"
> +    }
> +  } else {
> +    if {$i == $curthread} {
> +      pass "current thread ran"
> +    } else {
> +      fail "other thread $i didn't run (ran)"
> +    }
> +  }
> +}
> +
> +# Test stepping with scheduler locking
> +step_ten_loops "locked"
> +
> +# Make sure we're still in the same thread.
> +set newthread [get_current_thread "find current thread (2)"]
> +if {$curthread == $newthread} {
> +    pass "step with lock does not change thread"
> +} else {
> +    fail "step with lock does not change thread (switched to thread $newthread)"
> +}
> +
> +set start_args $cont_args
> +set cont_args [get_args]
> +
> +for {set i 0} {[expr $i < 6]} {set i [expr $i + 1]} {
> +  if {[lindex $start_args $i] == [lindex $cont_args $i]} {
> +    if {$i == $curthread} {
> +      fail "current thread stepped locked (didn't run)"
> +    } else {
> +      pass "other thread $i didn't run (stepping)"
> +    }
> +  } else {
> +    if {$i == $curthread} {
> +       if {[lindex $start_args $i] == [expr [lindex $cont_args $i] - 10]} {
> +           pass "current thread stepped locked"
> +       } else {
> +           fail "current thread stepped locked (wrong amount)"
> +       }
> +    } else {
> +      fail "other thread $i didn't run (stepping) (ran)"
> +    }
> +  }
> +}
> +
> +return 0

Re: RFA (threads testsuite): More thread tests

[Daniel Jacobowitz]

On Thu, Aug 22, 2002 at 04:34:51PM -0700, Michael Snyder wrote:
> Daniel Jacobowitz wrote:
> > 
> > Here's two tests I had lying around from when I developed the gdbserver
> > threads support.  Gdbserver passes them with flying colors (if you use my
> > other patch which lets gdbserver run tests properly).  GDB shows a couple of
> > problems, unpredictably (not always repeatable).
> > 
> > OK to add these?
> 
> Hi Joel, 
> 
> I understand the point of schedlock.exp, but what's the point of
> print-threads.exp?  What is it that you're testing?

No specific feature - just the general ability to handle threads doing
things.  It has a slightly different behavior pattern than the other
threads testscases, and triggered different problems.  Oh, and I
remember - there is no other testcase in the testsuite with
pthread_join in it; no threads ever actually exit.  I found some
problems there while I was implementing the gdbserver threads support.

-- 
Daniel Jacobowitz
MontaVista Software                         Debian GNU/Linux Developer

Re: RFA (threads testsuite): More thread tests

[Michael Snyder]

Daniel Jacobowitz wrote:
> 
> On Thu, Aug 22, 2002 at 04:34:51PM -0700, Michael Snyder wrote:
> > Daniel Jacobowitz wrote:
> > >
> > > Here's two tests I had lying around from when I developed the gdbserver
> > > threads support.  Gdbserver passes them with flying colors (if you use my
> > > other patch which lets gdbserver run tests properly).  GDB shows a couple of
> > > problems, unpredictably (not always repeatable).
> > >
> > > OK to add these?
> >
> > Hi Joel,
> >
> > I understand the point of schedlock.exp, but what's the point of
> > print-threads.exp?  What is it that you're testing?
> 
> No specific feature - just the general ability to handle threads doing
> things.  It has a slightly different behavior pattern than the other
> threads testscases, and triggered different problems.  Oh, and I
> remember - there is no other testcase in the testsuite with
> pthread_join in it; no threads ever actually exit.  I found some
> problems there while I was implementing the gdbserver threads support.

OK, then, since we definitely need more thread testing, 
let's accept these.  I like the way you verified that 
schedlock was implemented for the target before testing it.

BTW, do these really have to be native-only?  They should
work for embedded pthread targets, shouldn't they?

Re: RFA (threads testsuite): More thread tests

[Daniel Jacobowitz]

On Fri, Aug 23, 2002 at 11:51:41AM -0700, Michael Snyder wrote:
> Daniel Jacobowitz wrote:
> > 
> > On Thu, Aug 22, 2002 at 04:34:51PM -0700, Michael Snyder wrote:
> > > Daniel Jacobowitz wrote:
> > > >
> > > > Here's two tests I had lying around from when I developed the gdbserver
> > > > threads support.  Gdbserver passes them with flying colors (if you use my
> > > > other patch which lets gdbserver run tests properly).  GDB shows a couple of
> > > > problems, unpredictably (not always repeatable).
> > > >
> > > > OK to add these?
> > >
> > > Hi Joel,
> > >
> > > I understand the point of schedlock.exp, but what's the point of
> > > print-threads.exp?  What is it that you're testing?
> > 
> > No specific feature - just the general ability to handle threads doing
> > things.  It has a slightly different behavior pattern than the other
> > threads testscases, and triggered different problems.  Oh, and I
> > remember - there is no other testcase in the testsuite with
> > pthread_join in it; no threads ever actually exit.  I found some
> > problems there while I was implementing the gdbserver threads support.
> 
> OK, then, since we definitely need more thread testing, 
> let's accept these.  I like the way you verified that 
> schedlock was implemented for the target before testing it.
> 
> BTW, do these really have to be native-only?  They should
> work for embedded pthread targets, shouldn't they?

(And I'll make them use the new compile command too!)

Sure.  But which ones will it work on, that's the question.  I'm
tempted to have a gdb_skip_threads_test, but not sure what to key it
off of...

-- 
Daniel Jacobowitz
MontaVista Software                         Debian GNU/Linux Developer

Re: RFA (threads testsuite): More thread tests

[Daniel Jacobowitz]

On Fri, Aug 23, 2002 at 03:02:54PM -0400, Daniel Jacobowitz wrote:
> On Fri, Aug 23, 2002 at 11:51:41AM -0700, Michael Snyder wrote:
> > Daniel Jacobowitz wrote:
> > > 
> > > On Thu, Aug 22, 2002 at 04:34:51PM -0700, Michael Snyder wrote:
> > > > Daniel Jacobowitz wrote:
> > > > >
> > > > > Here's two tests I had lying around from when I developed the gdbserver
> > > > > threads support.  Gdbserver passes them with flying colors (if you use my
> > > > > other patch which lets gdbserver run tests properly).  GDB shows a couple of
> > > > > problems, unpredictably (not always repeatable).
> > > > >
> > > > > OK to add these?
> > > >
> > > > Hi Joel,
> > > >
> > > > I understand the point of schedlock.exp, but what's the point of
> > > > print-threads.exp?  What is it that you're testing?
> > > 
> > > No specific feature - just the general ability to handle threads doing
> > > things.  It has a slightly different behavior pattern than the other
> > > threads testscases, and triggered different problems.  Oh, and I
> > > remember - there is no other testcase in the testsuite with
> > > pthread_join in it; no threads ever actually exit.  I found some
> > > problems there while I was implementing the gdbserver threads support.
> > 
> > OK, then, since we definitely need more thread testing, 
> > let's accept these.  I like the way you verified that 
> > schedlock was implemented for the target before testing it.
> > 
> > BTW, do these really have to be native-only?  They should
> > work for embedded pthread targets, shouldn't they?
> 
> (And I'll make them use the new compile command too!)

No, I won't, it isn't committed yet :).

> Sure.  But which ones will it work on, that's the question.  I'm
> tempted to have a gdb_skip_threads_test, but not sure what to key it
> off of...

I've committed them without the isnative check.  Now they'll run
anywhere they link.  For the cases where GDB doesn't handle that...
we'll have to find some way to suppress them.

-- 
Daniel Jacobowitz
MontaVista Software                         Debian GNU/Linux Developer

Re: RFA (threads testsuite): More thread tests

[Michael Snyder]

Daniel Jacobowitz wrote:
> 
> On Fri, Aug 23, 2002 at 11:51:41AM -0700, Michael Snyder wrote:
> > Daniel Jacobowitz wrote:
> > >
> > > On Thu, Aug 22, 2002 at 04:34:51PM -0700, Michael Snyder wrote:
> > > > Daniel Jacobowitz wrote:
> > > > >
> > > > > Here's two tests I had lying around from when I developed the gdbserver
> > > > > threads support.  Gdbserver passes them with flying colors (if you use my
> > > > > other patch which lets gdbserver run tests properly).  GDB shows a couple of
> > > > > problems, unpredictably (not always repeatable).
> > > > >
> > > > > OK to add these?
> > > >
> > > > Hi Joel,
> > > >
> > > > I understand the point of schedlock.exp, but what's the point of
> > > > print-threads.exp?  What is it that you're testing?
> > >
> > > No specific feature - just the general ability to handle threads doing
> > > things.  It has a slightly different behavior pattern than the other
> > > threads testscases, and triggered different problems.  Oh, and I
> > > remember - there is no other testcase in the testsuite with
> > > pthread_join in it; no threads ever actually exit.  I found some
> > > problems there while I was implementing the gdbserver threads support.
> >
> > OK, then, since we definitely need more thread testing,
> > let's accept these.  I like the way you verified that
> > schedlock was implemented for the target before testing it.
> >
> > BTW, do these really have to be native-only?  They should
> > work for embedded pthread targets, shouldn't they?
> 
> (And I'll make them use the new compile command too!)
> 
> Sure.  But which ones will it work on, that's the question.  I'm
> tempted to have a gdb_skip_threads_test, but not sure what to key it
> off of...

Well -- if they won't build (eg. because there's no threads library), 
it's a pretty sure bet they won't run.  If they do build...