diff mbox

[review] Set names of worker threads

Message ID gerrit.1571543710000.I60473d65ae9ae14d8c56ddde39684240c16aaf35@gnutoolchain-gerrit.osci.io
State New
Headers show

Commit Message

Simon Marchi (Code Review) Oct. 20, 2019, 3:55 a.m. UTC
Change URL: https://gnutoolchain-gerrit.osci.io/r/c/binutils-gdb/+/176
......................................................................

Set names of worker threads

This adds some configury so that gdb can set the names of worker
threads.  This makes them show up more nicely when debugging gdb
itself.

gdb/ChangeLog
2019-10-19  Tom Tromey  <tom@tromey.com>

	* gdbsupport/thread-pool.c (thread_pool::set_thread_count): Set
	name of worker thread.
	* gdbsupport/common.m4 (GDB_AC_COMMON): Check for
	pthread_setname_np.
	* configure, config.in: Rebuild.

gdb/gdbserver/ChangeLog
2019-10-19  Tom Tromey  <tom@tromey.com>

	* configure, config.in: Rebuild.

Change-Id: I60473d65ae9ae14d8c56ddde39684240c16aaf35
---
M gdb/ChangeLog
M gdb/config.in
M gdb/configure
M gdb/gdbserver/ChangeLog
M gdb/gdbserver/config.in
M gdb/gdbserver/configure
M gdb/gdbsupport/common.m4
M gdb/gdbsupport/thread-pool.c
8 files changed, 43 insertions(+), 9 deletions(-)

Comments

Simon Marchi (Code Review) Nov. 6, 2019, 12:49 a.m. UTC | #1
Pedro Alves has posted comments on this change.

Change URL: https://gnutoolchain-gerrit.osci.io/r/c/binutils-gdb/+/176
......................................................................


Patch Set 3:

I'll try to take a look at this tomorrow.
Simon Marchi (Code Review) Nov. 26, 2019, 7:38 p.m. UTC | #2
Pedro Alves has posted comments on this change.

Change URL: https://gnutoolchain-gerrit.osci.io/r/c/binutils-gdb/+/176
......................................................................


Patch Set 4:

(3 comments)

Some comments below.  Nothing major.

| --- gdb/gdbsupport/thread-pool.c
| +++ gdb/gdbsupport/thread-pool.c
| @@ -22,15 +22,24 @@ #include "common-defs.h"
|  #if CXX_STD_THREAD
|  
|  #include "gdbsupport/thread-pool.h"
|  #include "gdbsupport/alt-stack.h"
|  #include "gdbsupport/block-signals.h"
|  #include <algorithm>
|  
| +/* On the off chance that we have the pthread library on a Windows
| +   host, but std::thread is not using it, avoid calling
| +   pthread_setname_np on Windows.  */

PS4, Line 31:

Meh.  Judging from
<https://stackoverflow.com/questions/19250008/mingw-stdthread-with-
windows-api>, I think this could be solved.  But yeah, we really don't
have to bend backwards in this initial patch.

| +#ifndef _WIN32
| +#ifdef HAVE_PTHREAD_SETNAME_NP
| +#include <pthread.h>
| +#endif
| +#endif

PS4, Line 36:

I'd rather this was:

#ifndef _WIN32
# ifdef HAVE_PTHREAD_SETNAME_NP
#  define USE_PTHREAD_SETNAME_NP
# endif
#endif

#ifdef USE_PTHREAD_SETNAME_NP
# include <pthread.h>
#endif

so that below we don't duplicate the checks, and we'd have a single
place to edit in the future if the checks evolve.  See below...

| +
|  namespace gdb
|  {
|  
|  /* The thread pool detach()s its threads, so that the threads will not
|     prevent the process from exiting.  However, it was discovered that
|     if any detached threads were still waiting on a condition variable,
|     then the condition variable's destructor would wait for the threads
|     to exit -- defeating the purpose.

 ...

| @@ -57,14 +66,19 @@ thread_pool::set_thread_count (size_t num_threads)
|    if (m_thread_count < num_threads)
|      {
|        /* Ensure that signals used by gdb are blocked in the new
|  	 threads.  */
|        block_signals blocker;
|        for (size_t i = m_thread_count; i < num_threads; ++i)
|  	{
|  	  std::thread thread (&thread_pool::thread_function, this);
| +#ifndef _WIN32 /* See the comment at the top of the file.  */
| +#ifdef HAVE_PTHREAD_SETNAME_NP

PS4, Line 75:

... here, we'd do:

#ifdef USE_PTHREAD_SETNAME_NP

| +	  pthread_setname_np (thread.native_handle (), "gdb worker");
| +#endif
| +#endif
|  	  thread.detach ();
|  	}
|      }
|    /* If the new size is smaller, terminate some existing threads.  */
|    if (num_threads < m_thread_count)
|      {
Simon Marchi (Code Review) Nov. 26, 2019, 7:54 p.m. UTC | #3
Tom Tromey has posted comments on this change.

Change URL: https://gnutoolchain-gerrit.osci.io/r/c/binutils-gdb/+/176
......................................................................


Patch Set 4:

(1 comment)

| --- gdb/gdbsupport/thread-pool.c
| +++ gdb/gdbsupport/thread-pool.c
| @@ -23,14 +23,23 @@ #if CXX_STD_THREAD
|  
|  #include "gdbsupport/thread-pool.h"
|  #include "gdbsupport/alt-stack.h"
|  #include "gdbsupport/block-signals.h"
|  #include <algorithm>
|  
| +/* On the off chance that we have the pthread library on a Windows
| +   host, but std::thread is not using it, avoid calling
| +   pthread_setname_np on Windows.  */
| +#ifndef _WIN32
| +#ifdef HAVE_PTHREAD_SETNAME_NP
| +#include <pthread.h>
| +#endif
| +#endif

PS4, Line 36:

> I'd rather this was:
...

I made this change.

| +
|  namespace gdb
|  {
|  
|  /* The thread pool detach()s its threads, so that the threads will not
|     prevent the process from exiting.  However, it was discovered that
|     if any detached threads were still waiting on a condition variable,
|     then the condition variable's destructor would wait for the threads
|     to exit -- defeating the purpose.
diff mbox

Patch

diff --git a/gdb/ChangeLog b/gdb/ChangeLog
index ec8f644..17ec8e5 100644
--- a/gdb/ChangeLog
+++ b/gdb/ChangeLog
@@ -1,5 +1,13 @@ 
 2019-10-19  Tom Tromey  <tom@tromey.com>
 
+	* gdbsupport/thread-pool.c (thread_pool::set_thread_count): Set
+	name of worker thread.
+	* gdbsupport/common.m4 (GDB_AC_COMMON): Check for
+	pthread_setname_np.
+	* configure, config.in: Rebuild.
+
+2019-10-19  Tom Tromey  <tom@tromey.com>
+
 	* python/python.c (class gdbpy_gil): New.
 	(struct gdbpy_event): Add constructor, destructor, operator().
 	(gdbpy_post_event): Use run_on_main_thread.
diff --git a/gdb/config.in b/gdb/config.in
index ff2fd12..54f1dd1 100644
--- a/gdb/config.in
+++ b/gdb/config.in
@@ -345,6 +345,9 @@ 
 /* Have PTHREAD_PRIO_INHERIT. */
 #undef HAVE_PTHREAD_PRIO_INHERIT
 
+/* Define to 1 if you have the `pthread_setname_np' function. */
+#undef HAVE_PTHREAD_SETNAME_NP
+
 /* Define to 1 if you have the `pthread_sigmask' function. */
 #undef HAVE_PTHREAD_SIGMASK
 
diff --git a/gdb/configure b/gdb/configure
index 5571cf0..2a967bf 100755
--- a/gdb/configure
+++ b/gdb/configure
@@ -14387,12 +14387,13 @@ 
 
     # This check must be here, while LIBS includes any necessary
     # threading library.
-    for ac_func in pthread_sigmask
+    for ac_func in pthread_sigmask pthread_setname_np
 do :
-  ac_fn_cxx_check_func "$LINENO" "pthread_sigmask" "ac_cv_func_pthread_sigmask"
-if test "x$ac_cv_func_pthread_sigmask" = xyes; then :
+  as_ac_var=`$as_echo "ac_cv_func_$ac_func" | $as_tr_sh`
+ac_fn_cxx_check_func "$LINENO" "$ac_func" "$as_ac_var"
+if eval test \"x\$"$as_ac_var"\" = x"yes"; then :
   cat >>confdefs.h <<_ACEOF
-#define HAVE_PTHREAD_SIGMASK 1
+#define `$as_echo "HAVE_$ac_func" | $as_tr_cpp` 1
 _ACEOF
 
 fi
diff --git a/gdb/gdbserver/ChangeLog b/gdb/gdbserver/ChangeLog
index 28c3608..e5a5829 100644
--- a/gdb/gdbserver/ChangeLog
+++ b/gdb/gdbserver/ChangeLog
@@ -1,5 +1,9 @@ 
 2019-10-19  Tom Tromey  <tom@tromey.com>
 
+	* configure, config.in: Rebuild.
+
+2019-10-19  Tom Tromey  <tom@tromey.com>
+
 	* remote-utils.c (block_unblock_async_io): Use gdb_sigmask.
 	* linux-low.c (linux_wait_for_event_filtered, linux_async): Use
 	gdb_sigmask.
diff --git a/gdb/gdbserver/config.in b/gdb/gdbserver/config.in
index 3027ffa..14cf709 100644
--- a/gdb/gdbserver/config.in
+++ b/gdb/gdbserver/config.in
@@ -195,6 +195,9 @@ 
 /* Have PTHREAD_PRIO_INHERIT. */
 #undef HAVE_PTHREAD_PRIO_INHERIT
 
+/* Define to 1 if you have the `pthread_setname_np' function. */
+#undef HAVE_PTHREAD_SETNAME_NP
+
 /* Define to 1 if you have the `pthread_sigmask' function. */
 #undef HAVE_PTHREAD_SIGMASK
 
diff --git a/gdb/gdbserver/configure b/gdb/gdbserver/configure
index 260e8c0..c5144c4 100755
--- a/gdb/gdbserver/configure
+++ b/gdb/gdbserver/configure
@@ -7725,12 +7725,13 @@ 
 
     # This check must be here, while LIBS includes any necessary
     # threading library.
-    for ac_func in pthread_sigmask
+    for ac_func in pthread_sigmask pthread_setname_np
 do :
-  ac_fn_cxx_check_func "$LINENO" "pthread_sigmask" "ac_cv_func_pthread_sigmask"
-if test "x$ac_cv_func_pthread_sigmask" = xyes; then :
+  as_ac_var=`$as_echo "ac_cv_func_$ac_func" | $as_tr_sh`
+ac_fn_cxx_check_func "$LINENO" "$ac_func" "$as_ac_var"
+if eval test \"x\$"$as_ac_var"\" = x"yes"; then :
   cat >>confdefs.h <<_ACEOF
-#define HAVE_PTHREAD_SIGMASK 1
+#define `$as_echo "HAVE_$ac_func" | $as_tr_cpp` 1
 _ACEOF
 
 fi
diff --git a/gdb/gdbsupport/common.m4 b/gdb/gdbsupport/common.m4
index 7da765b..0adcf26 100644
--- a/gdb/gdbsupport/common.m4
+++ b/gdb/gdbsupport/common.m4
@@ -57,7 +57,7 @@ 
 
     # This check must be here, while LIBS includes any necessary
     # threading library.
-    AC_CHECK_FUNCS([pthread_sigmask])
+    AC_CHECK_FUNCS([pthread_sigmask pthread_setname_np])
 
     LIBS="$save_LIBS"
     CXXFLAGS="$save_CXXFLAGS"
diff --git a/gdb/gdbsupport/thread-pool.c b/gdb/gdbsupport/thread-pool.c
index 993e097..0c3c1ec 100644
--- a/gdb/gdbsupport/thread-pool.c
+++ b/gdb/gdbsupport/thread-pool.c
@@ -23,6 +23,15 @@ 
 #include "gdbsupport/block-signals.h"
 #include <algorithm>
 
+/* On the off chance that we have the pthread library on a Windows
+   host, but std::thread is not using it, avoid calling
+   pthread_setname_np on Windows.  */
+#ifndef _WIN32
+#ifdef HAVE_PTHREAD_SETNAME_NP
+#include <pthread.h>
+#endif
+#endif
+
 namespace gdb
 {
 
@@ -55,6 +64,11 @@ 
       for (size_t i = m_count; i < num_threads; ++i)
 	{
 	  std::thread thread (&thread_pool::thread_function, this);
+#ifndef _WIN32 /* See the comment at the top of the file.  */
+#ifdef HAVE_PTHREAD_SETNAME_NP
+	  pthread_setname_np (thread.native_handle (), "gdb worker");
+#endif
+#endif
 	  thread.detach ();
 	}
     }