Skip to content

Instantly share code, notes, and snippets.

@hughperkins

hughperkins/gist:09110acfde7bcf550ee7c2f8dc56ebed

Created May 29, 2017 19:43
Show Gist options
  • Save hughperkins/09110acfde7bcf550ee7c2f8dc56ebed to your computer and use it in GitHub Desktop.
Save hughperkins/09110acfde7bcf550ee7c2f8dc56ebed to your computer and use it in GitHub Desktop.
Download ZIP
Raw
gistfile1.txt
b [[ 1.27999997 1.96000004 1.88 0.36000001 1.27999997 1.84000003
1.27999997 1.88 1. 0.75999999]]
expected [[ 3.07999992 2.03999996 3. 1.72000003 2.79999995 2.51999998
2.03999996 3.55999994 1.88 2.07999992]]
gpu [[ 3.07999992 2.03999996 3. 1.72000003 2.79999995 2.51999998
2.03999996 3.55999994 1.88 2.07999992]]
diff 0.0
func add float32
a [[ 1.79999995 0.08 1.12 1.36000001 1.51999998 0.68000001
0.75999999 1.67999995 0.88 1.32000005]]
b [[ 1.27999997 1.96000004 1.88 0.36000001 1.27999997 1.84000003
1.27999997 1.88 1. 0.75999999]]
expected [[ 3.07999992 2.03999996 3. 1.72000003 2.79999995 2.51999998
2.03999996 3.55999994 1.88 2.07999992]]
gpu [[ 3.07999992 2.03999996 3. 1.72000003 2.79999995 2.51999998
2.03999996 3.55999994 1.88 2.07999992]]
diff 0.0
func add float32
a [[ 1.79999995 0.08 1.12 1.36000001 1.51999998 0.68000001
0.75999999 1.67999995 0.88 1.32000005]]
b [[ 1.27999997 1.96000004 1.88 0.36000001 1.27999997 1.84000003
1.27999997 1.88 1. 0.75999999]]
expected [[ 3.07999992 2.03999996 3. 1.72000003 2.79999995 2.51999998
2.03999996 3.55999994 1.88 2.07999992]]
gpu [[ 3.07999992 2.03999996 3. 1.72000003 2.79999995 2.51999998
2.03999996 3.55999994 1.88 2.07999992]]
diff 0.0
func add float32
a [[ 1.79999995 0.08 1.12 1.36000001 1.51999998 0.68000001
0.75999999 1.67999995 0.88 1.32000005]]
b [[ 1.27999997 1.96000004 1.88 0.36000001 1.27999997 1.84000003
1.27999997 1.88 1. 0.75999999]]
expected [[ 3.07999992 2.03999996 3. 1.72000003 2.79999995 2.51999998
2.03999996 3.55999994 1.88 2.07999992]]
gpu [[ 3.07999992 2.03999996 3. 1.72000003 2.79999995 2.51999998
2.03999996 3.55999994 1.88 2.07999992]]
diff 0.0
func add float32
a [[ 1.79999995 0.08 1.12 1.36000001 1.51999998 0.68000001
0.75999999 1.67999995 0.88 1.32000005]]
b [[ 1.27999997 1.96000004 1.88 0.36000001 1.27999997 1.84000003
1.27999997 1.88 1. 0.75999999]]
expected [[ 3.07999992 2.03999996 3. 1.72000003 2.79999995 2.51999998
2.03999996 3.55999994 1.88 2.07999992]]
gpu [[ 3.07999992 2.03999996 3. 1.72000003 2.79999995 2.51999998
2.03999996 3.55999994 1.88 2.07999992]]
diff 0.0
func add float32
a [[ 1.79999995 0.08 1.12 1.36000001 1.51999998 0.68000001
0.75999999 1.67999995 0.88 1.32000005]]
b [[ 1.27999997 1.96000004 1.88 0.36000001 1.27999997 1.84000003
1.27999997 1.88 1. 0.75999999]]
expected [[ 3.07999992 2.03999996 3. 1.72000003 2.79999995 2.51999998
2.03999996 3.55999994 1.88 2.07999992]]
gpu [[ 3.07999992 2.03999996 3. 1.72000003 2.79999995 2.51999998
2.03999996 3.55999994 1.88 2.07999992]]
diff 0.0
func add float32
a [[ 1.79999995 0.08 1.12 1.36000001 1.51999998 0.68000001
0.75999999 1.67999995 0.88 1.32000005]]
b [[ 1.27999997 1.96000004 1.88 0.36000001 1.27999997 1.84000003
1.27999997 1.88 1. 0.75999999]]
expected [[ 3.07999992 2.03999996 3. 1.72000003 2.79999995 2.51999998
2.03999996 3.55999994 1.88 2.07999992]]
gpu [[ 3.07999992 2.03999996 3. 1.72000003 2.79999995 2.51999998
2.03999996 3.55999994 1.88 2.07999992]]
diff 0.0
func add float32
a [[ 1.79999995 0.08 1.12 1.36000001 1.51999998 0.68000001
0.75999999 1.67999995 0.88 1.32000005]]
b [[ 1.27999997 1.96000004 1.88 0.36000001 1.27999997 1.84000003
1.27999997 1.88 1. 0.75999999]]
expected [[ 3.07999992 2.03999996 3. 1.72000003 2.79999995 2.51999998
2.03999996 3.55999994 1.88 2.07999992]]
gpu [[ 3.07999992 2.03999996 3. 1.72000003 2.79999995 2.51999998
2.03999996 3.55999994 1.88 2.07999992]]
diff 0.0
func add float32
a [[ 1.79999995 0.08 1.12 1.36000001 1.51999998 0.68000001
0.75999999 1.67999995 0.88 1.32000005]]
b [[ 1.27999997 1.96000004 1.88 0.36000001 1.27999997 1.84000003
1.27999997 1.88 1. 0.75999999]]
expected [[ 3.07999992 2.03999996 3. 1.72000003 2.79999995 2.51999998
2.03999996 3.55999994 1.88 2.07999992]]
gpu [[ 3.07999992 2.03999996 3. 1.72000003 2.79999995 2.51999998
2.03999996 3.55999994 1.88 2.07999992]]
diff 0.0
func add float32
a [[ 1.79999995 0.08 1.12 1.36000001 1.51999998 0.68000001
0.75999999 1.67999995 0.88 1.32000005]]
b [[ 1.27999997 1.96000004 1.88 0.36000001 1.27999997 1.84000003
1.27999997 1.88 1. 0.75999999]]
expected [[ 3.07999992 2.03999996 3. 1.72000003 2.79999995 2.51999998
2.03999996 3.55999994 1.88 2.07999992]]
gpu [[ 3.07999992 2.03999996 3. 1.72000003 2.79999995 2.51999998
2.03999996 3.55999994 1.88 2.07999992]]
diff 0.0
func add float32
a [[ 1.79999995 0.08 1.12 1.36000001 1.51999998 0.68000001
0.75999999 1.67999995 0.88 1.32000005]]
b [[ 1.27999997 1.96000004 1.88 0.36000001 1.27999997 1.84000003
1.27999997 1.88 1. 0.75999999]]
expected [[ 3.07999992 2.03999996 3. 1.72000003 2.79999995 2.51999998
2.03999996 3.55999994 1.88 2.07999992]]
gpu [[ 3.07999992 2.03999996 3. 1.72000003 2.79999995 2.51999998
2.03999996 3.55999994 1.88 2.07999992]]cudaStreamSynchronize queue=0x192f800
[COCL] cuEventDestroy CoclEvent=0x2e3f320
[COCL] ~CoclEvent() this=0x2e3f320
[COCL] ~CoclEvent() releasing underlying clevent 0x7f01e0000a00
cudaStreamSynchronize queue=0x2e359c0
[COCL] cuEventDestroy CoclEvent=0x18821e0
[COCL] ~CoclEvent() this=0x18821e0
cudaStreamSynchronize queue=0x1880720
[COCL] cuEventDestroy CoclEvent=0x1880150
[COCL] ~CoclEvent() this=0x1880150
cudaStreamSynchronize queue=0x2e3fbc0
[COCL] cuEventDestroy CoclEvent=0x186d8a0
[COCL] ~CoclEvent() this=0x186d8a0
PollLoop iteration lock mu_
PollLoop iteration call PollEVents
Pollevents()
PollLoop iteration after PollEVents
core/common_runtime/gpu/gpu_event_mgr.h FreeMemory()
FreeMemory done
PollLoop() calling notify
PollLoop() done
[COCL] cuEventDestroy CoclEvent=0x7f01ec000fc0
[COCL] ~CoclEvent() this=0x7f01ec000fc0
[COCL] ~CoclEvent() releasing underlying clevent 0x7f01e0001d30
cudaGetDeviceCount count=0x1
cudaGetDeviceCount count=0x1
cudaGetDeviceCount count=0x1
[MEM] cuMemGetInfo redirected
I tensorflow/core/common_runtime/gpu/gpu_device.cc:1083] Creating TensorFlow device (/gpu:0) -> (device: 0, name: GRID K520, pci bus id: 0000.0000)
cudaGetDeviceCount count=0x1
PollLoop()
PollLoop iteration start
PollLoop iteration not queue_empty
[COCL] CoclEvent() this=0x186d8a0
[COCL] cuEventCreate flags=0x1 new CoclEvent=0x186d8a0
[COCL] CoclEvent() this=0x188c140
[COCL] cuEventCreate flags=0x1 new CoclEvent=0x188c140
[COCL] CoclEvent() this=0x1870210
[COCL] cuEventCreate flags=0x1 new CoclEvent=0x1870210
[COCL] CoclEvent() this=0x2e393e0
[COCL] cuEventCreate flags=0x1 new CoclEvent=0x2e393e0
[MEM] cuMemsetD32 redirected value 0 count=0x101 location=0x33980 memory=0x2cdeec0
PollLoop iteration lock mu_
PollLoop iteration call PollEVents
Pollevents()
PollLoop iteration after PollEVents
core/common_runtime/gpu/gpu_event_mgr.h FreeMemory()
FreeMemory done
PollLoop iteration start
PollLoop iteration queue_empty
PollLoop iteration lock mu_
PollLoop iteration call PollEVents
Pollevents()
PollLoop iteration after PollEVents
core/common_runtime/gpu/gpu_event_mgr.h FreeMemory()
FreeMemory done
PollLoop iteration start
PollLoop iteration queue_empty
PollLoop iteration lock mu_
PollLoop iteration call PollEVents
Pollevents()
PollLoop iteration after PollEVents
core/common_runtime/gpu/gpu_event_mgr.h FreeMemory()
FreeMemory done
PollLoop iteration start
PollLoop iteration queue_empty
PollLoop iteration lock mu_
PollLoop iteration call PollEVents
Pollevents()
PollLoop iteration after PollEVents
core/common_runtime/gpu/gpu_event_mgr.h FreeMemory()
FreeMemory done
PollLoop iteration start
PollLoop iteration queue_empty
PollLoop iteration lock mu_
PollLoop iteration call PollEVents
Pollevents()
PollLoop iteration after PollEVents
core/common_runtime/gpu/gpu_event_mgr.h FreeMemory()
FreeMemory done
PollLoop iteration start
PollLoop iteration queue_empty
PollLoop iteration lock mu_
PollLoop iteration call PollEVents
Pollevents()
PollLoop iteration after PollEVents
core/common_runtime/gpu/gpu_event_mgr.h FreeMemory()
FreeMemory done
PollLoop iteration start
PollLoop iteration queue_empty
PollLoop iteration lock mu_
PollLoop iteration call PollEVents
Pollevents()
PollLoop iteration after PollEVents
core/common_runtime/gpu/gpu_event_mgr.h FreeMemory()
FreeMemory done
PollLoop iteration start
PollLoop iteration queue_empty
PollLoop iteration lock mu_
PollLoop iteration call PollEVents
Pollevents()
PollLoop iteration after PollEVents
core/common_runtime/gpu/gpu_event_mgr.h FreeMemory()
FreeMemory done
PollLoop iteration start
PollLoop iteration queue_empty
PollLoop iteration lock mu_
PollLoop iteration call PollEVents
Pollevents()
PollLoop iteration after PollEVents
core/common_runtime/gpu/gpu_event_mgr.h FreeMemory()
FreeMemory done
PollLoop iteration start
PollLoop iteration queue_empty
[COCL] cuEventRecord CoclEvent=0x186d8a0 queue=0x19b4a30
[COCL] cuEventRecord CoclEvent=0x186d8a0 created clevent=0x7f01ec001280
[COCL] [MEM] cuMemcpyHtoDAsync dst=cuEventRecord CoclEvent=0x33e80 src=0x18832e0 bytes=0x28
0x186d8a0 PollLoop iteration lock mu_
PollLoop iteration call PollEVents
Pollevents()
PollLoop iteration after PollEVents
core/common_runtime/gpu/gpu_event_mgr.h FreeMemory()
FreeMemory done
PollLoop iteration start
PollLoop iteration queue_empty
queue=0x19b4a30
[MEM] ... done cuMemcpyHtoDAsync dst=0x33e80 src=0x18832e0 bytes=0x28
[COCL] cuEventRecord releasing existing clevent 0x7f01ec001280
debugui iu=0x7f0221353e60
debugui origfn=0x7f01ec000f60
debugui pre=0x7b
debugui post=0x7b
debugui &iu.func=0x7f02213541a0
debugui *(long *)(char *)&iu.func=0x7f01ec000f60
QueueInUse() iu=0x7f0221353e60 origfn=139646231121760 pre=123 func=139646231121760 post=123
queueInUse no free events: creating new one
[COCL] cuEventRecord CoclEvent=0x186d8a0 created clevent=0x7f01e4001470
[COCL] CoclEvent() this=0x7f01ec000b90
[COCL] cuEventCreate flags=0x1 new CoclEvent=0x7f01ec000b90
[COCL] cuEventRecord CoclEvent=0x7f01ec000b90 queue=0x1880720
[MEM] cuMemcpyHtoDAsync dst=0x33f80 src=0x18fd9d0 bytes=0x28
[COCL] cuEventRecord CoclEvent=0x7f01ec000b90 created clevent=0x7f01ec001280
debugui iu=0x7f0221353e60
debugui origfn=0x7f01ec000f60
debugui pre=0x7b
debugui post=0x7b
debugui &iu.func=0x7f02213541a0
debugui *(long *)(char *)&iu.func=0x7f01ec000f60
queueInUse event=0x7f01ec0018d0 iu=0x7f0221353e60 origfn=139646231121760 pre=123 func=139646231121760 post=123
debugui iu=0x17d38a0
debugui origfn=0x7f01ec000f60
debugui pre=0x7b
debugui post=0x7b
debugui &iu.func=0x7f02213541a0
debugui *(long *)(char *)&iu.func=0x7f01ec000f60
queueInUse queued iu used_events[used_events.size() - 1] iu=0x17d38a0 origfn=139646231121760 pre=123 func=139646231121760 post=123 used_events_.size() 0x1
queueInUse after notify_all(): used_events_.size() 0x1
Pollevents()
iterate iu:0x17d38a0
pollevents pollfortatus
debugui iu=0x17d38a0
debugui origfn=0x7f01ec000f60
debugui pre=0x7b
debugui post=0x7b
debugui &iu.func=0x7f02213541a0
debugui *(long *)(char *)&iu.func=0x7f01ec000f60
ui: iu=0x17d38a0 origfn=139646231121760 pre=123 func=139646231121760 post=123
[COCL] cuEventQuery CoclEvent=0x7f01ec000b90 clevent=0x7f01ec001280
[COCL] clGetEventInfo: 0x3
[COCL] cuEventQuery, event not finished yet
pollevents pollforstatus s=0x2
status is kpending
core/common_runtime/gpu/gpu_event_mgr.h FreeMemory()
FreeMemory done
PollLoop iteration lock mu_
PollLoop iteration call PollEVents
[MEM] Pollevents()
iterate iu:0x17d38a0
pollevents pollfortatus
debugui iu=0x17d38a0
debugui origfn=0x7f01ec000f60
debugui pre=0x7b
debugui post=0x7b
... done cuMemcpyHtoDAsync dst= debugui &iu.func=0x7f02213541a0
debugui *(long *)(char *)&iu.func=0x33f800x7f02213541c0
ui: iu=0x17d38a0 origfn=139646231121760 pre=123 func=139647123800512 post=123
src=0x18fd9d0 bytes=0x28
[COCL] cuEventQuery CoclEvent=0x7f01ec000b90 clevent=0x7f01ec001280
[COCL] clGetEventInfo: 0
[COCL] cuEventQuery, event completed
pollevents pollforstatus s=0x3
status is kcomplete
debugui iu=0x17d38a0
debugui origfn=0x7f01ec000f60
debugui pre=0x7b
debugui post=0x7b
debugui &iu.func=0x7f02213541a0
debugui *(long *)(char *)&iu.func=0x7f02213541c0
ui: iu=0x17d38a0 origfn=139646231121760 pre=123 func=139647123800512 post=123
debugui iu=0x7f02227fbae0
debugui origfn=0x7f01ec000f60
debugui pre=0x7b
debugui post=0x7b
debugui &iu.func=0x7f02213541a0
debugui *(long *)(char *)&iu.func=0x7f02213541c0
to_free last: iu=0x7f02227fbae0 origfn=139646231121760 pre=123 func=139647123800512 post=123
pushed back iu to to_free
debugui iu=0x17d38a0
debugui origfn=0x7f01ec000f60
debugui pre=0x7b
debugui post=0x7b
debugui &iu.func=0x7f02213541a0
debugui *(long *)(char *)&iu.func=0x7f02213541c0
ui: iu=0x17d38a0 origfn=139646231121760 pre=123 func=139647123800512 post=123
free_events_ last: 0x7f01ec0018d0
pushed iu.event 0x7f01ec0018d0 to free_events_
PollLoop iteration after PollEVents
core/common_runtime/gpu/gpu_event_mgr.h FreeMemory()
FreeMemory iteration
debugui iu=0x7f02227fbae0
debugui origfn=0x7f01ec000f60
debugui pre=0x7b
debugui post=0x7b
debugui &iu.func=0x7f02213541a0
debugui *(long *)(char *)&iu.func=0x7f02213541c0
Freememory iu=iu=0x7f02227fbae0 origfn=139646231121760 pre=123 func=139647123800512 post=123
debugui iu=0x7f02227fbae0
debugui origfn=0x7f01ec000f60
debugui pre=0x7b
debugui post=0x7b
debugui &iu.func=0x7f02213541a0
debugui *(long *)(char *)&iu.func=0x7f02213541c0
debugui iu= Freememory 2 iu=iu=0x7f02227fbae0 origfn=139646231121760 pre=123 func=139647123800512 post=123
debugui iu=0x7f02227fbae0
debugui origfn=0x7f01ec000f60
debugui pre=0x7b
debugui post=0x7b
debugui &iu.func=0x7f02213541a0
debugui *(long *)(char *)&iu.func=0x7f02213541c0
Freememory 3 iu=iu=0x7f02227fbae0 origfn=139646231121760 pre=123 func=139647123800512 post=123
0x7f01f27f9e60
debugui origfn=0x7f01e4000d40
debugui pre=0x7b
debugui post=0x7b
debugui &iu.func=0x7f01f27fa1a0
debugui *(long *)(char *)&iu.func=0x7f01e4000d40
QueueInUse() iu=0x7f01f27f9e60 origfn=139646096903488 pre=123 func=139646096903488 post=123
[COCL] cuEventRecord CoclEvent=0x7f01ec000b90 queue=0x1880720
[COCL] cuEventRecord releasing existing clevent 0x7f01ec001280
[COCL] cuEventRecord CoclEvent=0x7f01ec000b90 created clevent=0x7f01e4000ac0
debugui iu=0x7f01f27f9e60
debugui origfn=0x7f01e4000d40
debugui pre=0x7b
debugui post=0x7b
debugui &iu.func=0x7f01f27fa1a0
debugui *(long *)(char *)&iu.func=0x7f01e4000d40
queueInUse event=0x7f01ec0018d0 iu=0x7f01f27f9e60 origfn=139646096903488 pre=123 func=139646096903488 post=123
debugui iu=0x17d3908
debugui origfn=0x7f01e4000d40
debugui pre=0x7b
debugui post=0x7b
debugui &iu.func=0x7f01f27fa1a0
debugui *(long *)(char *)&iu.func=0x7f01e4000d40
queueInUse queued iu used_events[used_events.size() - 1] iu=0x17d3908 origfn=139646096903488 pre=123 func=139646096903488 post=123 used_events_.size() 0x1
queueInUse after notify_all(): used_events_.size() 0x1
Pollevents()
iterate iu:0x17d3908
/home/ubuntu/git/tensorflow-cl/myrun: line 6: 4949 Segmentation fault (core dumped) python tensorflow/stream_executor/cl/test/test_binary_ops.py
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment