Re: [Xenomai] Xenomai-forge pSOS app hangs on t_suspend + t_delete

2014-01-11 Thread Kim De Mey 
On Jan 10, 2014 2:52 PM, Philippe Gerum r...@xenomai.org wrote:

 On 01/10/2014 12:35 PM, Kim De Mey wrote:

 2014/1/10 Philippe Gerum r...@xenomai.org:

 On 01/10/2014 10:25 AM, Philippe Gerum wrote:


 On 01/09/2014 11:29 AM, Kim De Mey wrote:


 2014/1/9 Philippe Gerum r...@xenomai.org:


 On 01/08/2014 01:23 PM, Kim De Mey wrote:


 2014/1/8 Philippe Gerum r...@xenomai.org:


 On 01/08/2014 10:25 AM, Kim De Mey wrote:



 Here are the backtraces:

 main thread:
 (gdb) bt
 #0  clock_nanosleep (clock_id=optimized out, flags=optimized
out,
 req=optimized out, rem=optimized out)
at ../sysdeps/unix/sysv/linux/clock_nanosleep.c:51
 #1  0x00b45e10 in threadobj_sleep () from


/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0

 #2  0x00b2b618 in tm_wkafter () from


/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0

 #3  0x1b3c in main (argc=1, argv=0x104120e0) at
 suspend_delete_easy.c:42

 first psos task:
 (gdb) bt
 #0  0x00b6b5c8 in __old_sem_wait (sem=optimized out) at
 ../nptl/sysdeps/unix/sysv/linux/sem_wait.c:105
 #1  0x00b45c84 in threadobj_cancel () from


/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0

 #2  0x00b2ac44 in t_delete () from


/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0

 #3  0x1a70 in test (a=0, b=0, c=0, d=0) at
suspend_delete_easy.c:22
 #4  0x00b2a658 in task_trampoline () from


/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0

 #5  0x00b63824 in start_thread (arg=optimized out) at
 pthread_create.c:299
 #6  0x00c7df4c in __thread_start () from
 output/host/usr/mips64-buildroot-linux-gnu/sysroot/lib32/libc.so.6

 second psos task:
 (gdb) bt
 #0  0x00b6cbbc in read () from


output/host/usr/mips64-buildroot-linux-gnu/sysroot/lib32/libpthread.so.0

 #1  0x00b47bd8 in notifier_wait () from


/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0

 #2  0x00b460b0 in notifier_callback () from


/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0

 #3  0x00b47d30 in notifier_sighandler () from


/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0

 #4  signal handler called
 #5  sigcancel_handler (sig=32, si=0x1e03aa8, ctx=0x1e03b28) at
 init.c:136
 #6  signal handler called
 #7  clock_nanosleep (clock_id=optimized out, flags=optimized
out,
 req=optimized out, rem=optimized out)
at ../sysdeps/unix/sysv/linux/clock_nanosleep.c:51
 #8  0x00b45e10 in threadobj_sleep () from


/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0

 #9  0x00b2b618 in tm_wkafter () from


/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0

 #10 0x19bc in idle_task (a=0, b=0, c=0, d=0) at
 suspend_delete_easy.c:9
 #11 0x00b2a658 in task_trampoline () from


/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0

 #12 0x00b63824 in start_thread (arg=optimized out) at
 pthread_create.c:299
 #13 0x00c7df4c in __thread_start () from
 output/host/usr/mips64-buildroot-linux-gnu/sysroot/lib32/libc.so.6


 Invocation command line was

  $ ./configure --target=mips64-buildroot-linux-gnu
 --host=mips64-buildroot-linux-gnu --build=x86_64-unknown-linux-gnu
 --prefix=/usr --exec-prefix=/usr --sysconfdir=/etc --program-prefix=
 --disable-gtk-doc --disable-doc --disable-docs
--disable-documentation
 --with-xmlto=no --with-fop=no --enable-ipv6 --enable-static
 --enable-shared --with-core=mercury
 --includedir=/usr/include/xenomai-forge --disable-doc-install
 --enable-lores-clock


 Is this what you were asking?



 Yes, thanks.



 By the way, I forgot the mention the following test that I did:

 When the application is blocked I've done an echo 1 
 /proc/pid/fd/xx,
 where xx is the file descriptor that is being read from in the
 notifier_wait().
 This kinda emulates a t_resume.

 After doing this the t_delete gets finalized and the application
 unblocks.


 This should fix the race, until the notifier is reworked.

 diff --git a/include/copperplate/threadobj.h
 b/include/copperplate/threadobj.h
 index 184c711..218c274 100644
 --- a/include/copperplate/threadobj.h
 +++ b/include/copperplate/threadobj.h
 @@ -124,6 +124,7 @@ void threadobj_save_timeout(struct
 threadobj_corespec *corespec,
#define __THREAD_S_ACTIVE  (1  5)/* Running user
 code. */
#define __THREAD_S_SUSPENDED   (1  6)/* Suspended via
 threadobj_suspend(). */
#define __THREAD_S_SAFE(1  7)/* TCB
 release deferred. */
 +#define __THREAD_S_ZOMBIE  (1  8)/* Deletion process
 ongoing. */
#define __THREAD_S_DEBUG   (1  31)   /* Debug mode

Re: [Xenomai] Xenomai-forge pSOS app hangs on t_suspend + t_delete

2014-01-10 Thread Philippe Gerum 

On 01/09/2014 11:29 AM, Kim De Mey wrote:

2014/1/9 Philippe Gerum r...@xenomai.org:

On 01/08/2014 01:23 PM, Kim De Mey wrote:

2014/1/8 Philippe Gerum r...@xenomai.org:

On 01/08/2014 10:25 AM, Kim De Mey wrote:


Here are the backtraces:

main thread:
(gdb) bt
#0  clock_nanosleep (clock_id=optimized out, flags=optimized out,
req=optimized out, rem=optimized out)
  at ../sysdeps/unix/sysv/linux/clock_nanosleep.c:51
#1  0x00b45e10 in threadobj_sleep () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0
#2  0x00b2b618 in tm_wkafter () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0
#3  0x1b3c in main (argc=1, argv=0x104120e0) at suspend_delete_easy.c:42

first psos task:
(gdb) bt
#0  0x00b6b5c8 in __old_sem_wait (sem=optimized out) at
../nptl/sysdeps/unix/sysv/linux/sem_wait.c:105
#1  0x00b45c84 in threadobj_cancel () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0
#2  0x00b2ac44 in t_delete () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0
#3  0x1a70 in test (a=0, b=0, c=0, d=0) at suspend_delete_easy.c:22
#4  0x00b2a658 in task_trampoline () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0
#5  0x00b63824 in start_thread (arg=optimized out) at pthread_create.c:299
#6  0x00c7df4c in __thread_start () from
output/host/usr/mips64-buildroot-linux-gnu/sysroot/lib32/libc.so.6

second psos task:
(gdb) bt
#0  0x00b6cbbc in read () from
output/host/usr/mips64-buildroot-linux-gnu/sysroot/lib32/libpthread.so.0
#1  0x00b47bd8 in notifier_wait () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0
#2  0x00b460b0 in notifier_callback () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0
#3  0x00b47d30 in notifier_sighandler () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0
#4  signal handler called
#5  sigcancel_handler (sig=32, si=0x1e03aa8, ctx=0x1e03b28) at init.c:136
#6  signal handler called
#7  clock_nanosleep (clock_id=optimized out, flags=optimized out,
req=optimized out, rem=optimized out)
  at ../sysdeps/unix/sysv/linux/clock_nanosleep.c:51
#8  0x00b45e10 in threadobj_sleep () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0
#9  0x00b2b618 in tm_wkafter () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0
#10 0x19bc in idle_task (a=0, b=0, c=0, d=0) at suspend_delete_easy.c:9
#11 0x00b2a658 in task_trampoline () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0
#12 0x00b63824 in start_thread (arg=optimized out) at pthread_create.c:299
#13 0x00c7df4c in __thread_start () from
output/host/usr/mips64-buildroot-linux-gnu/sysroot/lib32/libc.so.6


Invocation command line was

$ ./configure --target=mips64-buildroot-linux-gnu
--host=mips64-buildroot-linux-gnu --build=x86_64-unknown-linux-gnu
--prefix=/usr --exec-prefix=/usr --sysconfdir=/etc --program-prefix=
--disable-gtk-doc --disable-doc --disable-docs --disable-documentation
--with-xmlto=no --with-fop=no --enable-ipv6 --enable-static
--enable-shared --with-core=mercury
--includedir=/usr/include/xenomai-forge --disable-doc-install
--enable-lores-clock


Is this what you were asking?


Yes, thanks.




By the way, I forgot the mention the following test that I did:

When the application is blocked I've done an echo 1  /proc/pid/fd/xx,
where xx is the file descriptor that is being read from in the notifier_wait().
This kinda emulates a t_resume.

After doing this the t_delete gets finalized and the application unblocks.



This should fix the race, until the notifier is reworked.

diff --git a/include/copperplate/threadobj.h b/include/copperplate/threadobj.h
index 184c711..218c274 100644
--- a/include/copperplate/threadobj.h
+++ b/include/copperplate/threadobj.h
@@ -124,6 +124,7 @@ void threadobj_save_timeout(struct threadobj_corespec 
*corespec,
  #define __THREAD_S_ACTIVE  (1  5)/* Running user code. */
  #define __THREAD_S_SUSPENDED   (1  6)/* Suspended via 
threadobj_suspend(). */
  #define __THREAD_S_SAFE(1  7)/* TCB release 
deferred. */
+#define __THREAD_S_ZOMBIE  (1  8)/* Deletion process ongoing. */
  #define __THREAD_S_DEBUG   (1  31)   /* Debug mode enabled. */
  /*
   * threadobj-run_state, locklessly updated by current, merged
diff --git a/lib/copperplate/threadobj.c b/lib/copperplate/threadobj.c
index a31479c..c54cdf8 100644
--- a/lib/copperplate/threadobj.c
+++ b/lib/copperplate/threadobj.c
@@ -420,11 +420,13 @@ static void notifier_callback(const struct notifier *nf)

Re: [Xenomai] Xenomai-forge pSOS app hangs on t_suspend + t_delete

2014-01-10 Thread Philippe Gerum 

On 01/10/2014 10:25 AM, Philippe Gerum wrote:

On 01/09/2014 11:29 AM, Kim De Mey wrote:

2014/1/9 Philippe Gerum r...@xenomai.org:

On 01/08/2014 01:23 PM, Kim De Mey wrote:

2014/1/8 Philippe Gerum r...@xenomai.org:

On 01/08/2014 10:25 AM, Kim De Mey wrote:


Here are the backtraces:

main thread:
(gdb) bt
#0  clock_nanosleep (clock_id=optimized out, flags=optimized out,
req=optimized out, rem=optimized out)
  at ../sysdeps/unix/sysv/linux/clock_nanosleep.c:51
#1  0x00b45e10 in threadobj_sleep () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0

#2  0x00b2b618 in tm_wkafter () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0

#3  0x1b3c in main (argc=1, argv=0x104120e0) at
suspend_delete_easy.c:42

first psos task:
(gdb) bt
#0  0x00b6b5c8 in __old_sem_wait (sem=optimized out) at
../nptl/sysdeps/unix/sysv/linux/sem_wait.c:105
#1  0x00b45c84 in threadobj_cancel () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0

#2  0x00b2ac44 in t_delete () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0

#3  0x1a70 in test (a=0, b=0, c=0, d=0) at suspend_delete_easy.c:22
#4  0x00b2a658 in task_trampoline () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0

#5  0x00b63824 in start_thread (arg=optimized out) at
pthread_create.c:299
#6  0x00c7df4c in __thread_start () from
output/host/usr/mips64-buildroot-linux-gnu/sysroot/lib32/libc.so.6

second psos task:
(gdb) bt
#0  0x00b6cbbc in read () from
output/host/usr/mips64-buildroot-linux-gnu/sysroot/lib32/libpthread.so.0

#1  0x00b47bd8 in notifier_wait () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0

#2  0x00b460b0 in notifier_callback () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0

#3  0x00b47d30 in notifier_sighandler () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0

#4  signal handler called
#5  sigcancel_handler (sig=32, si=0x1e03aa8, ctx=0x1e03b28) at
init.c:136
#6  signal handler called
#7  clock_nanosleep (clock_id=optimized out, flags=optimized out,
req=optimized out, rem=optimized out)
  at ../sysdeps/unix/sysv/linux/clock_nanosleep.c:51
#8  0x00b45e10 in threadobj_sleep () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0

#9  0x00b2b618 in tm_wkafter () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0

#10 0x19bc in idle_task (a=0, b=0, c=0, d=0) at
suspend_delete_easy.c:9
#11 0x00b2a658 in task_trampoline () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0

#12 0x00b63824 in start_thread (arg=optimized out) at
pthread_create.c:299
#13 0x00c7df4c in __thread_start () from
output/host/usr/mips64-buildroot-linux-gnu/sysroot/lib32/libc.so.6


Invocation command line was

$ ./configure --target=mips64-buildroot-linux-gnu
--host=mips64-buildroot-linux-gnu --build=x86_64-unknown-linux-gnu
--prefix=/usr --exec-prefix=/usr --sysconfdir=/etc --program-prefix=
--disable-gtk-doc --disable-doc --disable-docs --disable-documentation
--with-xmlto=no --with-fop=no --enable-ipv6 --enable-static
--enable-shared --with-core=mercury
--includedir=/usr/include/xenomai-forge --disable-doc-install
--enable-lores-clock


Is this what you were asking?


Yes, thanks.




By the way, I forgot the mention the following test that I did:

When the application is blocked I've done an echo 1 
/proc/pid/fd/xx,
where xx is the file descriptor that is being read from in the
notifier_wait().
This kinda emulates a t_resume.

After doing this the t_delete gets finalized and the application
unblocks.



This should fix the race, until the notifier is reworked.

diff --git a/include/copperplate/threadobj.h
b/include/copperplate/threadobj.h
index 184c711..218c274 100644
--- a/include/copperplate/threadobj.h
+++ b/include/copperplate/threadobj.h
@@ -124,6 +124,7 @@ void threadobj_save_timeout(struct
threadobj_corespec *corespec,
  #define __THREAD_S_ACTIVE  (1  5)/* Running user
code. */
  #define __THREAD_S_SUSPENDED   (1  6)/* Suspended via
threadobj_suspend(). */
  #define __THREAD_S_SAFE(1  7)/* TCB
release deferred. */
+#define __THREAD_S_ZOMBIE  (1  8)/* Deletion process
ongoing. */
  #define __THREAD_S_DEBUG   (1  31)   /* Debug mode
enabled. */
  /*
   * threadobj-run_state, locklessly updated by current, merged
diff --git a/lib/copperplate/threadobj.c b/lib/copperplate/threadobj.c
index a31479c..c54cdf8 100644
--- a/lib/copperplate/threadobj.c
+++ b/lib/copperplate/threadobj.c
@@ -420,11 +420,13 @@

Re: [Xenomai] Xenomai-forge pSOS app hangs on t_suspend + t_delete

2014-01-10 Thread Kim De Mey 
2014/1/10 Philippe Gerum r...@xenomai.org:
 On 01/10/2014 10:25 AM, Philippe Gerum wrote:

 On 01/09/2014 11:29 AM, Kim De Mey wrote:

 2014/1/9 Philippe Gerum r...@xenomai.org:

 On 01/08/2014 01:23 PM, Kim De Mey wrote:

 2014/1/8 Philippe Gerum r...@xenomai.org:

 On 01/08/2014 10:25 AM, Kim De Mey wrote:


 Here are the backtraces:

 main thread:
 (gdb) bt
 #0  clock_nanosleep (clock_id=optimized out, flags=optimized out,
 req=optimized out, rem=optimized out)
   at ../sysdeps/unix/sysv/linux/clock_nanosleep.c:51
 #1  0x00b45e10 in threadobj_sleep () from

 /repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0

 #2  0x00b2b618 in tm_wkafter () from

 /repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0

 #3  0x1b3c in main (argc=1, argv=0x104120e0) at
 suspend_delete_easy.c:42

 first psos task:
 (gdb) bt
 #0  0x00b6b5c8 in __old_sem_wait (sem=optimized out) at
 ../nptl/sysdeps/unix/sysv/linux/sem_wait.c:105
 #1  0x00b45c84 in threadobj_cancel () from

 /repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0

 #2  0x00b2ac44 in t_delete () from

 /repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0

 #3  0x1a70 in test (a=0, b=0, c=0, d=0) at suspend_delete_easy.c:22
 #4  0x00b2a658 in task_trampoline () from

 /repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0

 #5  0x00b63824 in start_thread (arg=optimized out) at
 pthread_create.c:299
 #6  0x00c7df4c in __thread_start () from
 output/host/usr/mips64-buildroot-linux-gnu/sysroot/lib32/libc.so.6

 second psos task:
 (gdb) bt
 #0  0x00b6cbbc in read () from

 output/host/usr/mips64-buildroot-linux-gnu/sysroot/lib32/libpthread.so.0

 #1  0x00b47bd8 in notifier_wait () from

 /repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0

 #2  0x00b460b0 in notifier_callback () from

 /repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0

 #3  0x00b47d30 in notifier_sighandler () from

 /repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0

 #4  signal handler called
 #5  sigcancel_handler (sig=32, si=0x1e03aa8, ctx=0x1e03b28) at
 init.c:136
 #6  signal handler called
 #7  clock_nanosleep (clock_id=optimized out, flags=optimized out,
 req=optimized out, rem=optimized out)
   at ../sysdeps/unix/sysv/linux/clock_nanosleep.c:51
 #8  0x00b45e10 in threadobj_sleep () from

 /repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0

 #9  0x00b2b618 in tm_wkafter () from

 /repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0

 #10 0x19bc in idle_task (a=0, b=0, c=0, d=0) at
 suspend_delete_easy.c:9
 #11 0x00b2a658 in task_trampoline () from

 /repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0

 #12 0x00b63824 in start_thread (arg=optimized out) at
 pthread_create.c:299
 #13 0x00c7df4c in __thread_start () from
 output/host/usr/mips64-buildroot-linux-gnu/sysroot/lib32/libc.so.6


 Invocation command line was

 $ ./configure --target=mips64-buildroot-linux-gnu
 --host=mips64-buildroot-linux-gnu --build=x86_64-unknown-linux-gnu
 --prefix=/usr --exec-prefix=/usr --sysconfdir=/etc --program-prefix=
 --disable-gtk-doc --disable-doc --disable-docs --disable-documentation
 --with-xmlto=no --with-fop=no --enable-ipv6 --enable-static
 --enable-shared --with-core=mercury
 --includedir=/usr/include/xenomai-forge --disable-doc-install
 --enable-lores-clock


 Is this what you were asking?


 Yes, thanks.



 By the way, I forgot the mention the following test that I did:

 When the application is blocked I've done an echo 1 
 /proc/pid/fd/xx,
 where xx is the file descriptor that is being read from in the
 notifier_wait().
 This kinda emulates a t_resume.

 After doing this the t_delete gets finalized and the application
 unblocks.


 This should fix the race, until the notifier is reworked.

 diff --git a/include/copperplate/threadobj.h
 b/include/copperplate/threadobj.h
 index 184c711..218c274 100644
 --- a/include/copperplate/threadobj.h
 +++ b/include/copperplate/threadobj.h
 @@ -124,6 +124,7 @@ void threadobj_save_timeout(struct
 threadobj_corespec *corespec,
   #define __THREAD_S_ACTIVE  (1  5)/* Running user
 code. */
   #define __THREAD_S_SUSPENDED   (1  6)/* Suspended via
 threadobj_suspend(). */
   #define __THREAD_S_SAFE(1  7)/* TCB
 release deferred. */
 +#define __THREAD_S_ZOMBIE  (1  8)/* Deletion process
 ongoing. */
   #define __THREAD_S_DEBUG   (1  31)   /* Debug mode
 enabled. */
   /*
* threadobj-run_state, locklessly updated by current, merged
 diff --git a/lib/copperplate/threadobj.c

Re: [Xenomai] Xenomai-forge pSOS app hangs on t_suspend + t_delete

2014-01-09 Thread Kim De Mey 
2014/1/9 Philippe Gerum r...@xenomai.org:
 On 01/08/2014 01:23 PM, Kim De Mey wrote:
 2014/1/8 Philippe Gerum r...@xenomai.org:
 On 01/08/2014 10:25 AM, Kim De Mey wrote:

 Here are the backtraces:

 main thread:
 (gdb) bt
 #0  clock_nanosleep (clock_id=optimized out, flags=optimized out,
 req=optimized out, rem=optimized out)
  at ../sysdeps/unix/sysv/linux/clock_nanosleep.c:51
 #1  0x00b45e10 in threadobj_sleep () from
 /repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0
 #2  0x00b2b618 in tm_wkafter () from
 /repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0
 #3  0x1b3c in main (argc=1, argv=0x104120e0) at suspend_delete_easy.c:42

 first psos task:
 (gdb) bt
 #0  0x00b6b5c8 in __old_sem_wait (sem=optimized out) at
 ../nptl/sysdeps/unix/sysv/linux/sem_wait.c:105
 #1  0x00b45c84 in threadobj_cancel () from
 /repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0
 #2  0x00b2ac44 in t_delete () from
 /repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0
 #3  0x1a70 in test (a=0, b=0, c=0, d=0) at suspend_delete_easy.c:22
 #4  0x00b2a658 in task_trampoline () from
 /repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0
 #5  0x00b63824 in start_thread (arg=optimized out) at pthread_create.c:299
 #6  0x00c7df4c in __thread_start () from
 output/host/usr/mips64-buildroot-linux-gnu/sysroot/lib32/libc.so.6

 second psos task:
 (gdb) bt
 #0  0x00b6cbbc in read () from
 output/host/usr/mips64-buildroot-linux-gnu/sysroot/lib32/libpthread.so.0
 #1  0x00b47bd8 in notifier_wait () from
 /repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0
 #2  0x00b460b0 in notifier_callback () from
 /repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0
 #3  0x00b47d30 in notifier_sighandler () from
 /repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0
 #4  signal handler called
 #5  sigcancel_handler (sig=32, si=0x1e03aa8, ctx=0x1e03b28) at init.c:136
 #6  signal handler called
 #7  clock_nanosleep (clock_id=optimized out, flags=optimized out,
 req=optimized out, rem=optimized out)
  at ../sysdeps/unix/sysv/linux/clock_nanosleep.c:51
 #8  0x00b45e10 in threadobj_sleep () from
 /repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0
 #9  0x00b2b618 in tm_wkafter () from
 /repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0
 #10 0x19bc in idle_task (a=0, b=0, c=0, d=0) at suspend_delete_easy.c:9
 #11 0x00b2a658 in task_trampoline () from
 /repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0
 #12 0x00b63824 in start_thread (arg=optimized out) at pthread_create.c:299
 #13 0x00c7df4c in __thread_start () from
 output/host/usr/mips64-buildroot-linux-gnu/sysroot/lib32/libc.so.6


 Invocation command line was

$ ./configure --target=mips64-buildroot-linux-gnu
 --host=mips64-buildroot-linux-gnu --build=x86_64-unknown-linux-gnu
 --prefix=/usr --exec-prefix=/usr --sysconfdir=/etc --program-prefix=
 --disable-gtk-doc --disable-doc --disable-docs --disable-documentation
 --with-xmlto=no --with-fop=no --enable-ipv6 --enable-static
 --enable-shared --with-core=mercury
 --includedir=/usr/include/xenomai-forge --disable-doc-install
 --enable-lores-clock


 Is this what you were asking?

 Yes, thanks.



 By the way, I forgot the mention the following test that I did:

 When the application is blocked I've done an echo 1  /proc/pid/fd/xx,
 where xx is the file descriptor that is being read from in the 
 notifier_wait().
 This kinda emulates a t_resume.

 After doing this the t_delete gets finalized and the application unblocks.


 This should fix the race, until the notifier is reworked.

 diff --git a/include/copperplate/threadobj.h b/include/copperplate/threadobj.h
 index 184c711..218c274 100644
 --- a/include/copperplate/threadobj.h
 +++ b/include/copperplate/threadobj.h
 @@ -124,6 +124,7 @@ void threadobj_save_timeout(struct threadobj_corespec 
 *corespec,
  #define __THREAD_S_ACTIVE  (1  5)/* Running user code. */
  #define __THREAD_S_SUSPENDED   (1  6)/* Suspended via 
 threadobj_suspend(). */
  #define __THREAD_S_SAFE(1  7)/* TCB release 
 deferred. */
 +#define __THREAD_S_ZOMBIE  (1  8)/* Deletion process ongoing. 
 */
  #define __THREAD_S_DEBUG   (1  31)   /* Debug mode enabled. */
  /*
   * threadobj-run_state, locklessly updated by current, merged
 diff --git a/lib/copperplate/threadobj.c b/lib/copperplate/threadobj.c
 index a31479c..c54cdf8 100644
 --- a/lib/copperplate/threadobj.c
 +++ b/lib/copperplate/threadobj.c
 @@ -420,11 +420,13 @@ static void

[Xenomai] Xenomai-forge pSOS app hangs on t_suspend + t_delete

2014-01-08 Thread Kim De Mey 
Hi,

I have an issue with a pSOS application that hangs after doing t_delete.
We are using Xenomai-forge with Mercury core.
The issue still occurs with the latest update.

I can reproduce the issue with a very simple test application.
The test application has a pSOS task that creates and starts another task.
Then it suspends and deletes this task.
The t_suspend() and t_delete() functions are called right after each other.
This is the crucial part for the issue to happen.

Example code:

static void idle_task(u_long a,u_long b,u_long c,u_long d)
{
  while (1) tm_wkafter(100);
}

static void test(u_long a,u_long b,u_long c,u_long d)
{
  u_long tid,args[4] = {0,0,0,0};

  t_create(IDLE,10,0,0,0,tid);
  t_start(tid,0,idle_task, args);

  tm_wkafter(1000);

  t_suspend(tid);
  t_delete(tid);
  printf(After t_delete of suspended task\n);

  while (1) tm_wkafter(1000);
}

What I notice is that the cancel_sync() call in threadobj.c remains stuck
at the sem_wait() call.
The sem_post in finalize_thread() does not happen.
It even looks like the finalize_thread() is never called.
So it seems that for some reason the thread does not get cancelled (?).

The issue does not always occur in my simple test.
If I run it on more than 1 core it does not happen.
So it looks like something racy.
I notice the following difference in call order:
The issue occurs when notifier_callback(), the suspend callback,
happens after the sem_wait() (and thus after pthread_cancel()).
The issue does not occur when the callback is started before that the
sem_wait() call (or pthread_cancel()) occurs.
I am not sure if that has anything to do with it.

Can somebody have a look at this or give me some pointers because
I had a look at the code and do not understand what could be causing this.

Regards,
Kim

___
Xenomai mailing list
Xenomai@xenomai.org
http://www.xenomai.org/mailman/listinfo/xenomai

Re: [Xenomai] Xenomai-forge pSOS app hangs on t_suspend + t_delete

2014-01-08 Thread Philippe Gerum 

On 01/08/2014 10:25 AM, Kim De Mey wrote:

Hi,

I have an issue with a pSOS application that hangs after doing t_delete.
We are using Xenomai-forge with Mercury core.
The issue still occurs with the latest update.

I can reproduce the issue with a very simple test application.
The test application has a pSOS task that creates and starts another task.
Then it suspends and deletes this task.
The t_suspend() and t_delete() functions are called right after each other.
This is the crucial part for the issue to happen.

Example code:

static void idle_task(u_long a,u_long b,u_long c,u_long d)
{
   while (1) tm_wkafter(100);
}

static void test(u_long a,u_long b,u_long c,u_long d)
{
   u_long tid,args[4] = {0,0,0,0};

   t_create(IDLE,10,0,0,0,tid);
   t_start(tid,0,idle_task, args);

   tm_wkafter(1000);

   t_suspend(tid);
   t_delete(tid);
   printf(After t_delete of suspended task\n);

   while (1) tm_wkafter(1000);
}

What I notice is that the cancel_sync() call in threadobj.c remains stuck
at the sem_wait() call.
The sem_post in finalize_thread() does not happen.
It even looks like the finalize_thread() is never called.
So it seems that for some reason the thread does not get cancelled (?).

The issue does not always occur in my simple test.
If I run it on more than 1 core it does not happen.
So it looks like something racy.
I notice the following difference in call order:
The issue occurs when notifier_callback(), the suspend callback,
happens after the sem_wait() (and thus after pthread_cancel()).
The issue does not occur when the callback is started before that the
sem_wait() call (or pthread_cancel()) occurs.
I am not sure if that has anything to do with it.

Can somebody have a look at this or give me some pointers because
I had a look at the code and do not understand what could be causing this.



When your test hangs, could you attach gdb to the process, then dump the 
backtraces for all threads which are still present?


Also, the exact invocation command line of the build configuration 
script as available from config.log, would help.


TIA,

--
Philippe.

___
Xenomai mailing list
Xenomai@xenomai.org
http://www.xenomai.org/mailman/listinfo/xenomai

Re: [Xenomai] Xenomai-forge pSOS app hangs on t_suspend + t_delete

2014-01-08 Thread Kim De Mey 
2014/1/8 Philippe Gerum r...@xenomai.org:
 On 01/08/2014 10:25 AM, Kim De Mey wrote:

 Hi,

 I have an issue with a pSOS application that hangs after doing t_delete.
 We are using Xenomai-forge with Mercury core.
 The issue still occurs with the latest update.

 I can reproduce the issue with a very simple test application.
 The test application has a pSOS task that creates and starts another task.
 Then it suspends and deletes this task.
 The t_suspend() and t_delete() functions are called right after each
 other.
 This is the crucial part for the issue to happen.

 Example code:

 static void idle_task(u_long a,u_long b,u_long c,u_long d)
 {
while (1) tm_wkafter(100);
 }

 static void test(u_long a,u_long b,u_long c,u_long d)
 {
u_long tid,args[4] = {0,0,0,0};

t_create(IDLE,10,0,0,0,tid);
t_start(tid,0,idle_task, args);

tm_wkafter(1000);

t_suspend(tid);
t_delete(tid);
printf(After t_delete of suspended task\n);

while (1) tm_wkafter(1000);
 }

 What I notice is that the cancel_sync() call in threadobj.c remains stuck
 at the sem_wait() call.
 The sem_post in finalize_thread() does not happen.
 It even looks like the finalize_thread() is never called.
 So it seems that for some reason the thread does not get cancelled (?).

 The issue does not always occur in my simple test.
 If I run it on more than 1 core it does not happen.
 So it looks like something racy.
 I notice the following difference in call order:
 The issue occurs when notifier_callback(), the suspend callback,
 happens after the sem_wait() (and thus after pthread_cancel()).
 The issue does not occur when the callback is started before that the
 sem_wait() call (or pthread_cancel()) occurs.
 I am not sure if that has anything to do with it.

 Can somebody have a look at this or give me some pointers because
 I had a look at the code and do not understand what could be causing this.


 When your test hangs, could you attach gdb to the process, then dump the
 backtraces for all threads which are still present?

 Also, the exact invocation command line of the build configuration script as
 available from config.log, would help.

 TIA,

 --
 Philippe.


Here are the backtraces:

main thread:
(gdb) bt
#0  clock_nanosleep (clock_id=optimized out, flags=optimized out,
req=optimized out, rem=optimized out)
at ../sysdeps/unix/sysv/linux/clock_nanosleep.c:51
#1  0x00b45e10 in threadobj_sleep () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0
#2  0x00b2b618 in tm_wkafter () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0
#3  0x1b3c in main (argc=1, argv=0x104120e0) at suspend_delete_easy.c:42

first psos task:
(gdb) bt
#0  0x00b6b5c8 in __old_sem_wait (sem=optimized out) at
../nptl/sysdeps/unix/sysv/linux/sem_wait.c:105
#1  0x00b45c84 in threadobj_cancel () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0
#2  0x00b2ac44 in t_delete () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0
#3  0x1a70 in test (a=0, b=0, c=0, d=0) at suspend_delete_easy.c:22
#4  0x00b2a658 in task_trampoline () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0
#5  0x00b63824 in start_thread (arg=optimized out) at pthread_create.c:299
#6  0x00c7df4c in __thread_start () from
output/host/usr/mips64-buildroot-linux-gnu/sysroot/lib32/libc.so.6

second psos task:
(gdb) bt
#0  0x00b6cbbc in read () from
output/host/usr/mips64-buildroot-linux-gnu/sysroot/lib32/libpthread.so.0
#1  0x00b47bd8 in notifier_wait () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0
#2  0x00b460b0 in notifier_callback () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0
#3  0x00b47d30 in notifier_sighandler () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0
#4  signal handler called
#5  sigcancel_handler (sig=32, si=0x1e03aa8, ctx=0x1e03b28) at init.c:136
#6  signal handler called
#7  clock_nanosleep (clock_id=optimized out, flags=optimized out,
req=optimized out, rem=optimized out)
at ../sysdeps/unix/sysv/linux/clock_nanosleep.c:51
#8  0x00b45e10 in threadobj_sleep () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libcopperplate.so.0
#9  0x00b2b618 in tm_wkafter () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0
#10 0x19bc in idle_task (a=0, b=0, c=0, d=0) at suspend_delete_easy.c:9
#11 0x00b2a658 in task_trampoline () from
/repo/kdemey/buildroot/output/host/usr/mips64-buildroot-linux-gnu/sysroot/usr/lib/libpsos.so.0
#12 0x00b63824 in start_thread (arg=optimized out) at pthread_create.c:299
#13 0x00c7df4c in __thread_start