Re: [PATCH][libgomp, nvptx] Fix hang in gomp_team_barrier_wait_end

[Tom de Vries <tdevries@suse.de>]

[-- Attachment #1: Type: text/plain, Size: 1320 bytes --]

On 4/21/21 7:02 PM, Alexander Monakov wrote:
> On Wed, 21 Apr 2021, Tom de Vries wrote:
> 
>>> I don't think implementing futex_wait is possible on nvptx.
>>>
>>
>> Well, I gave it a try, attached below.  Can you explain why you think
>> it's not possible, or pinpoint a problem in the implementation?
> 
> Responding only to this for now. When I said futex_wait I really meant
> Linux futex wait, where the API is tied to a 32-bit futex control word
> and nothing else. Your implementation works with a gomp_barrier_t that
> includes more than one field. It would be confusing to call it a
> "futex wait", it is not a 1:1 replacement.
> 
> (i.e. unlike a proper futex, it can work only for gomp_barrier_t objects)

Ah, I see, agreed, that makes sense.  I was afraid there was some
fundamental problem that I overlooked.

Here's an updated version.  I've tried to make it clear that the
futex_wait/wake are locally used versions, not generic functionality.

The main change in structure is that I'm now using the
generation_to_barrier trick from the rtems port, allowing linux/bar.c to
be included rather than copied (because the barrier argument is now
implicit).

Furthermore, I've reviewed the MEMMODELs used for the atomic accesses,
and updated a few.

Also now the cpu_relax from doacross.h is used.

Thanks,
- Tom

[-- Attachment #2: 0002-libgomp-nvptx-Fix-hang-in-gomp_team_barrier_wait_end.patch --]
[-- Type: text/x-patch, Size: 17212 bytes --]

[libgomp, nvptx] Fix hang in gomp_team_barrier_wait_end

Consider the following omp fragment.
...
  #pragma omp target
  #pragma omp parallel num_threads (2)
  #pragma omp task
    ;
...

This hangs at -O0 for nvptx.

Investigating the behaviour gives us the following trace of events:
- both threads execute GOMP_task, where they:
  - deposit a task, and
  - execute gomp_team_barrier_wake
- thread 1 executes gomp_team_barrier_wait_end and, not being the last thread,
  proceeds to wait at the team barrier
- thread 0 executes gomp_team_barrier_wait_end and, being the last thread, it
  calls gomp_barrier_handle_tasks, where it:
  - executes both tasks and marks the team barrier done
  - executes a gomp_team_barrier_wake which wakes up thread 1
- thread 1 exits the team barrier
- thread 0 returns from gomp_barrier_handle_tasks and goes to wait at
  the team barrier.
- thread 0 hangs.

To understand why there is a hang here, it's good to understand how things
are setup for nvptx.  The libgomp/config/nvptx/bar.c implementation is
a copy of the libgomp/config/linux/bar.c implementation, with uses of both
futex_wake and do_wait replaced with uses of ptx insn bar.sync:
...
  if (bar->total > 1)
    asm ("bar.sync 1, %0;" : : "r" (32 * bar->total));
...

The point where thread 0 goes to wait at the team barrier, corresponds in
the linux implementation with a do_wait.  In the linux case, the call to
do_wait doesn't hang, because it's waiting for bar->generation to become
a certain value, and if bar->generation already has that value, it just
proceeds, without any need for coordination with other threads.

In the nvtpx case, the bar.sync waits until thread 1 joins it in the same
logical barrier, which never happens: thread 1 is lingering in the
thread pool at the thread pool barrier (using a different logical barrier),
waiting to join a new team.

The easiest way to fix this is to revert to the posix implementation for
bar.{c,h}.  That however falls back on a busy-waiting approach, and
does not take advantage of the ptx bar.sync insn.

Instead, we revert to the linux implementation for bar.c,
and implement bar.c local functions futex_wait and futex_wake using the
bar.sync insn.

This is a WIP version that does not yet take performance into consideration,
but instead focuses on copying a working version as completely as possible,
and isolating the machine-specific changes to as few functions as
possible.

The bar.sync insn takes an argument specifying how many threads are
participating, and that doesn't play well with the futex syntax where it's
not clear in advance how many threads will be woken up.

This is solved by waking up all waiting threads each time a futex_wait or
futex_wake happens, and possibly going back to sleep with an updated thread
count.

Tested libgomp on x86_64 with nvptx accelerator, both as-is and with
do_spin hardcoded to 1.

libgomp/ChangeLog:

2021-04-20  Tom de Vries  <tdevries@suse.de>

	PR target/99555
	* config/nvptx/bar.c (generation_to_barrier): New function, copied
	from config/rtems/bar.c.
	(futex_wait, futex_wake): New function.
	(do_spin, do_wait): New function, copied from config/linux/wait.h.
	(gomp_barrier_wait_end, gomp_barrier_wait_last)
	(gomp_team_barrier_wake, gomp_team_barrier_wait_end):
	(gomp_team_barrier_wait_cancel_end, gomp_team_barrier_cancel): Remove
	and replace with include of config/linux/bar.c.
	* config/nvptx/bar.h (gomp_barrier_t): Add fields waiters and lock.
	(gomp_barrier_init): Init new fields.
	* testsuite/libgomp.c-c++-common/task-detach-6.c: Remove nvptx-specific
	workarounds.
	* testsuite/libgomp.c/pr99555-1.c: Same.
	* testsuite/libgomp.fortran/task-detach-6.f90: Same.

---
 libgomp/config/nvptx/bar.c                         | 258 +++++++++------------
 libgomp/config/nvptx/bar.h                         |   4 +
 .../testsuite/libgomp.c-c++-common/task-detach-6.c |   8 -
 libgomp/testsuite/libgomp.c/pr99555-1.c            |   8 -
 .../testsuite/libgomp.fortran/task-detach-6.f90    |  12 -
 5 files changed, 115 insertions(+), 175 deletions(-)

diff --git a/libgomp/config/nvptx/bar.c b/libgomp/config/nvptx/bar.c
index c5c2fa8829b..e0e6e5ed839 100644
--- a/libgomp/config/nvptx/bar.c
+++ b/libgomp/config/nvptx/bar.c
@@ -30,183 +30,147 @@
 #include <limits.h>
 #include "libgomp.h"
 
+/* For cpu_relax.  */
+#include "doacross.h"
 
-void
-gomp_barrier_wait_end (gomp_barrier_t *bar, gomp_barrier_state_t state)
-{
-  if (__builtin_expect (state & BAR_WAS_LAST, 0))
-    {
-      /* Next time we'll be awaiting TOTAL threads again.  */
-      bar->awaited = bar->total;
-      __atomic_store_n (&bar->generation, bar->generation + BAR_INCR,
-			MEMMODEL_RELEASE);
-    }
-  if (bar->total > 1)
-    asm ("bar.sync 1, %0;" : : "r" (32 * bar->total));
-}
+/* Assuming ADDR is &bar->generation, return bar.  Copied from
+   rtems/bar.c.  */
 
-void
-gomp_barrier_wait (gomp_barrier_t *bar)
+static gomp_barrier_t *
+generation_to_barrier (int *addr)
 {
-  gomp_barrier_wait_end (bar, gomp_barrier_wait_start (bar));
+  char *bar
+    = (char *) addr - __builtin_offsetof (gomp_barrier_t, generation);
+  return (gomp_barrier_t *)bar;
 }
 
-/* Like gomp_barrier_wait, except that if the encountering thread
-   is not the last one to hit the barrier, it returns immediately.
-   The intended usage is that a thread which intends to gomp_barrier_destroy
-   this barrier calls gomp_barrier_wait, while all other threads
-   call gomp_barrier_wait_last.  When gomp_barrier_wait returns,
-   the barrier can be safely destroyed.  */
+/* Implement futex_wait-like behaviour to plug into the linux/bar.c
+   implementation.  Assumes ADDR is &bar->generation.   */
 
-void
-gomp_barrier_wait_last (gomp_barrier_t *bar)
+static inline void
+futex_wait (int *addr, int val)
 {
-  /* Deferring to gomp_barrier_wait does not use the optimization opportunity
-     allowed by the interface contract for all-but-last participants.  The
-     original implementation in config/linux/bar.c handles this better.  */
-  gomp_barrier_wait (bar);
-}
+  gomp_barrier_t *bar = generation_to_barrier (addr);
 
-void
-gomp_team_barrier_wake (gomp_barrier_t *bar, int count)
-{
-  if (bar->total > 1)
-    asm ("bar.sync 1, %0;" : : "r" (32 * bar->total));
-}
+  if (bar->total < 2)
+    /* A barrier with less than two threads, nop.  */
+    return;
 
-void
-gomp_team_barrier_wait_end (gomp_barrier_t *bar, gomp_barrier_state_t state)
-{
-  unsigned int generation, gen;
+  gomp_mutex_lock (&bar->lock);
 
-  if (__builtin_expect (state & BAR_WAS_LAST, 0))
+  /* Futex semantics: only go to sleep if *addr == val.  */
+  if (__builtin_expect (__atomic_load_n (addr, MEMMODEL_ACQUIRE) != val, 0))
     {
-      /* Next time we'll be awaiting TOTAL threads again.  */
-      struct gomp_thread *thr = gomp_thread ();
-      struct gomp_team *team = thr->ts.team;
-
-      bar->awaited = bar->total;
-      team->work_share_cancelled = 0;
-      if (__builtin_expect (team->task_count, 0))
-	{
-	  gomp_barrier_handle_tasks (state);
-	  state &= ~BAR_WAS_LAST;
-	}
-      else
-	{
-	  state &= ~BAR_CANCELLED;
-	  state += BAR_INCR - BAR_WAS_LAST;
-	  __atomic_store_n (&bar->generation, state, MEMMODEL_RELEASE);
-	  if (bar->total > 1)
-	    asm ("bar.sync 1, %0;" : : "r" (32 * bar->total));
-	  return;
-	}
+      gomp_mutex_unlock (&bar->lock);
+      return;
     }
 
-  generation = state;
-  state &= ~BAR_CANCELLED;
-  do
+  /* Register as waiter.  */
+  unsigned int waiters
+    = __atomic_add_fetch (&bar->waiters, 1, MEMMODEL_ACQ_REL);
+  if (waiters == 0)
+    __builtin_abort ();
+  unsigned int waiter_id = waiters;
+
+  if (waiters > 1)
     {
-      if (bar->total > 1)
-	asm ("bar.sync 1, %0;" : : "r" (32 * bar->total));
-      gen = __atomic_load_n (&bar->generation, MEMMODEL_ACQUIRE);
-      if (__builtin_expect (gen & BAR_TASK_PENDING, 0))
-	{
-	  gomp_barrier_handle_tasks (state);
-	  gen = __atomic_load_n (&bar->generation, MEMMODEL_ACQUIRE);
-	}
-      generation |= gen & BAR_WAITING_FOR_TASK;
+      /* Wake other threads in bar.sync.  */
+      asm volatile ("bar.sync 1, %0;" : : "r" (32 * waiters));
+
+      /* Ensure that they have updated waiters.  */
+      asm volatile ("bar.sync 1, %0;" : : "r" (32 * waiters));
     }
-  while (gen != state + BAR_INCR);
-}
 
-void
-gomp_team_barrier_wait (gomp_barrier_t *bar)
-{
-  gomp_team_barrier_wait_end (bar, gomp_barrier_wait_start (bar));
-}
+  gomp_mutex_unlock (&bar->lock);
 
-void
-gomp_team_barrier_wait_final (gomp_barrier_t *bar)
-{
-  gomp_barrier_state_t state = gomp_barrier_wait_final_start (bar);
-  if (__builtin_expect (state & BAR_WAS_LAST, 0))
-    bar->awaited_final = bar->total;
-  gomp_team_barrier_wait_end (bar, state);
+  while (1)
+    {
+      /* Wait for next thread in barrier.  */
+      asm volatile ("bar.sync 1, %0;" : : "r" (32 * (waiters + 1)));
+
+      /* Get updated waiters.  */
+      unsigned int updated_waiters
+	= __atomic_load_n (&bar->waiters, MEMMODEL_ACQUIRE);
+
+      /* Notify that we have updated waiters.  */
+      asm volatile ("bar.sync 1, %0;" : : "r" (32 * (waiters + 1)));
+
+      waiters = updated_waiters;
+
+      if (waiter_id > waiters)
+	/* A wake happened, and we're in the group of woken threads.  */
+	break;
+
+      /* Continue waiting.  */
+    }
 }
 
-bool
-gomp_team_barrier_wait_cancel_end (gomp_barrier_t *bar,
-				   gomp_barrier_state_t state)
+/* Implement futex_wake-like behaviour to plug into the linux/bar.c
+   implementation.  Assumes ADDR is &bar->generation.  */
+
+static inline void
+futex_wake (int *addr, int count)
 {
-  unsigned int generation, gen;
+  gomp_barrier_t *bar = generation_to_barrier (addr);
+
+  if (bar->total < 2)
+    /* A barrier with less than two threads, nop.  */
+    return;
 
-  if (__builtin_expect (state & BAR_WAS_LAST, 0))
+  gomp_mutex_lock (&bar->lock);
+  unsigned int waiters = __atomic_load_n (&bar->waiters, MEMMODEL_ACQUIRE);
+  if (waiters == 0)
     {
-      /* Next time we'll be awaiting TOTAL threads again.  */
-      /* BAR_CANCELLED should never be set in state here, because
-	 cancellation means that at least one of the threads has been
-	 cancelled, thus on a cancellable barrier we should never see
-	 all threads to arrive.  */
-      struct gomp_thread *thr = gomp_thread ();
-      struct gomp_team *team = thr->ts.team;
-
-      bar->awaited = bar->total;
-      team->work_share_cancelled = 0;
-      if (__builtin_expect (team->task_count, 0))
-	{
-	  gomp_barrier_handle_tasks (state);
-	  state &= ~BAR_WAS_LAST;
-	}
-      else
-	{
-	  state += BAR_INCR - BAR_WAS_LAST;
-	  __atomic_store_n (&bar->generation, state, MEMMODEL_RELEASE);
-	  if (bar->total > 1)
-	    asm ("bar.sync 1, %0;" : : "r" (32 * bar->total));
-	  return false;
-	}
+      /* No threads to wake.  */
+      gomp_mutex_unlock (&bar->lock);
+      return;
     }
 
-  if (__builtin_expect (state & BAR_CANCELLED, 0))
-    return true;
+  if (count == INT_MAX)
+    /* Release all threads.  */
+    __atomic_store_n (&bar->waiters, 0, MEMMODEL_RELEASE);
+  else if (count < bar->total)
+    /* Release count threads.  */
+    __atomic_add_fetch (&bar->waiters, -count, MEMMODEL_ACQ_REL);
+  else
+    /* Count has an illegal value.  */
+    __builtin_abort ();
 
-  generation = state;
-  do
-    {
-      if (bar->total > 1)
-	asm ("bar.sync 1, %0;" : : "r" (32 * bar->total));
-      gen = __atomic_load_n (&bar->generation, MEMMODEL_ACQUIRE);
-      if (__builtin_expect (gen & BAR_CANCELLED, 0))
-	return true;
-      if (__builtin_expect (gen & BAR_TASK_PENDING, 0))
-	{
-	  gomp_barrier_handle_tasks (state);
-	  gen = __atomic_load_n (&bar->generation, MEMMODEL_ACQUIRE);
-	}
-      generation |= gen & BAR_WAITING_FOR_TASK;
-    }
-  while (gen != state + BAR_INCR);
+  /* Wake other threads in bar.sync.  */
+  asm volatile ("bar.sync 1, %0;" : : "r" (32 * (waiters + 1)));
+
+  /* Let them get the updated waiters.  */
+  asm volatile ("bar.sync 1, %0;" : : "r" (32 * (waiters + 1)));
 
-  return false;
+  gomp_mutex_unlock (&bar->lock);
 }
 
-bool
-gomp_team_barrier_wait_cancel (gomp_barrier_t *bar)
+/* Copied from linux/wait.h.  */
+
+static inline int do_spin (int *addr, int val)
 {
-  return gomp_team_barrier_wait_cancel_end (bar, gomp_barrier_wait_start (bar));
+  unsigned long long i, count = gomp_spin_count_var;
+
+  if (__builtin_expect (__atomic_load_n (&gomp_managed_threads,
+					 MEMMODEL_RELAXED)
+			> gomp_available_cpus, 0))
+    count = gomp_throttled_spin_count_var;
+  for (i = 0; i < count; i++)
+    if (__builtin_expect (__atomic_load_n (addr, MEMMODEL_RELAXED) != val, 0))
+      return 0;
+    else
+      cpu_relax ();
+  return 1;
 }
 
-void
-gomp_team_barrier_cancel (struct gomp_team *team)
+/* Copied from linux/wait.h.  */
+
+static inline void do_wait (int *addr, int val)
 {
-  gomp_mutex_lock (&team->task_lock);
-  if (team->barrier.generation & BAR_CANCELLED)
-    {
-      gomp_mutex_unlock (&team->task_lock);
-      return;
-    }
-  team->barrier.generation |= BAR_CANCELLED;
-  gomp_mutex_unlock (&team->task_lock);
-  gomp_team_barrier_wake (&team->barrier, INT_MAX);
+  if (do_spin (addr, val))
+    futex_wait (addr, val);
 }
+
+/* Reuse the linux implementation.  */
+#define GOMP_WAIT_H 1
+#include "../linux/bar.c"
diff --git a/libgomp/config/nvptx/bar.h b/libgomp/config/nvptx/bar.h
index 9bf3d914a02..c69426e1629 100644
--- a/libgomp/config/nvptx/bar.h
+++ b/libgomp/config/nvptx/bar.h
@@ -38,6 +38,8 @@ typedef struct
   unsigned generation;
   unsigned awaited;
   unsigned awaited_final;
+  unsigned waiters;
+  gomp_mutex_t lock;
 } gomp_barrier_t;
 
 typedef unsigned int gomp_barrier_state_t;
@@ -57,6 +59,8 @@ static inline void gomp_barrier_init (gomp_barrier_t *bar, unsigned count)
   bar->awaited = count;
   bar->awaited_final = count;
   bar->generation = 0;
+  bar->waiters = 0;
+  gomp_mutex_init (&bar->lock);
 }
 
 static inline void gomp_barrier_reinit (gomp_barrier_t *bar, unsigned count)
diff --git a/libgomp/testsuite/libgomp.c-c++-common/task-detach-6.c b/libgomp/testsuite/libgomp.c-c++-common/task-detach-6.c
index f18b57bf047..e5c2291e6ff 100644
--- a/libgomp/testsuite/libgomp.c-c++-common/task-detach-6.c
+++ b/libgomp/testsuite/libgomp.c-c++-common/task-detach-6.c
@@ -2,9 +2,6 @@
 
 #include <omp.h>
 #include <assert.h>
-#include <unistd.h> // For 'alarm'.
-
-#include "on_device_arch.h"
 
 /* Test tasks with detach clause on an offload device.  Each device
    thread spawns off a chain of tasks, that can then be executed by
@@ -12,11 +9,6 @@
 
 int main (void)
 {
-  //TODO See '../libgomp.c/pr99555-1.c'.
-  if (on_device_arch_nvptx ())
-    alarm (4); /*TODO Until resolved, make sure that we exit quickly, with error status.
-		 { dg-xfail-run-if "PR99555" { offload_device_nvptx } } */
-
   int x = 0, y = 0, z = 0;
   int thread_count;
   omp_event_handle_t detach_event1, detach_event2;
diff --git a/libgomp/testsuite/libgomp.c/pr99555-1.c b/libgomp/testsuite/libgomp.c/pr99555-1.c
index bd33b93716b..7386e016fd2 100644
--- a/libgomp/testsuite/libgomp.c/pr99555-1.c
+++ b/libgomp/testsuite/libgomp.c/pr99555-1.c
@@ -2,16 +2,8 @@
 
 // { dg-additional-options "-O0" }
 
-#include <unistd.h> // For 'alarm'.
-
-#include "../libgomp.c-c++-common/on_device_arch.h"
-
 int main (void)
 {
-  if (on_device_arch_nvptx ())
-    alarm (4); /*TODO Until resolved, make sure that we exit quickly, with error status.
-		 { dg-xfail-run-if "PR99555" { offload_device_nvptx } } */
-
 #pragma omp target
 #pragma omp parallel // num_threads(1)
 #pragma omp task
diff --git a/libgomp/testsuite/libgomp.fortran/task-detach-6.f90 b/libgomp/testsuite/libgomp.fortran/task-detach-6.f90
index e4373b4c6f1..03a3b61540d 100644
--- a/libgomp/testsuite/libgomp.fortran/task-detach-6.f90
+++ b/libgomp/testsuite/libgomp.fortran/task-detach-6.f90
@@ -1,6 +1,5 @@
 ! { dg-do run }
 
-! { dg-additional-sources on_device_arch.c }
   ! { dg-prune-output "command-line option '-fintrinsic-modules-path=.*' is valid for Fortran but not for C" }
 
 ! Test tasks with detach clause on an offload device.  Each device
@@ -14,17 +13,6 @@ program task_detach_6
   integer :: x = 0, y = 0, z = 0
   integer :: thread_count
 
-  interface
-    integer function on_device_arch_nvptx() bind(C)
-    end function on_device_arch_nvptx
-  end interface
-
-  !TODO See '../libgomp.c/pr99555-1.c'.
-  if (on_device_arch_nvptx () /= 0) then
-     call alarm (4, 0); !TODO Until resolved, make sure that we exit quickly, with error status.
-     ! { dg-xfail-run-if "PR99555" { offload_device_nvptx } }
-  end if
-
   !$omp target map (tofrom: x, y, z) map (from: thread_count)
     !$omp parallel private (detach_event1, detach_event2)
       !$omp single