1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
|
# Copyright (c) 2022, NVIDIA CORPORATION. All rights reserved.
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions
# are met:
# * Redistributions of source code must retain the above copyright
# notice, this list of conditions and the following disclaimer.
# * Redistributions in binary form must reproduce the above copyright
# notice, this list of conditions and the following disclaimer in the
# documentation and/or other materials provided with the distribution.
# * Neither the name of NVIDIA CORPORATION nor the names of its
# contributors may be used to endorse or promote products derived
# from this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS ``AS IS'' AND ANY
# EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
# PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
# CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
# EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
# PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
# PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
# OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
import NvRules
def get_identifier():
return "ThreadDivergence"
def get_name():
return "Thread Divergence"
def get_description():
return "Warp and thread control flow analysis"
def get_section_identifier():
return "WarpStateStats"
def apply(handle):
ctx = NvRules.get_context(handle)
action = ctx.range_by_idx(0).action_by_idx(0)
fe = ctx.frontend()
threadInstExecutedName = "smsp__thread_inst_executed_per_inst_executed.ratio"
threadInstExecutedNPOName = "smsp__thread_inst_executed_pred_on_per_inst_executed.ratio"
threadInstExecuted = action.metric_by_name(threadInstExecutedName).as_double()
threadInstExecutedNPO = action.metric_by_name(threadInstExecutedNPOName).as_double()
fms = []
threshold = 24
if threadInstExecuted < threshold or threadInstExecutedNPO < threshold:
if threadInstExecuted < threshold:
fms.append((threadInstExecutedName, threadInstExecuted, NvRules.IFrontend.Severity_SEVERITY_LOW, "{} < {}".format(threadInstExecuted, threshold)))
if threadInstExecutedNPO < threshold:
fms.append((threadInstExecutedNPOName, threadInstExecutedNPO, NvRules.IFrontend.Severity_SEVERITY_LOW, "{} < {}".format(threadInstExecutedNPO, threshold)))
message = "Instructions are executed in warps, which are groups of 32 threads. Optimal instruction throughput is achieved if all 32 threads of a warp execute the same instruction. The chosen launch configuration, early thread completion, and divergent flow control can significantly lower the number of active threads in a warp per cycle. This kernel achieves an average of {0:.1f} threads being active per cycle.".format(threadInstExecuted)
if threadInstExecutedNPO < threadInstExecuted:
fms.append((threadInstExecutedNPOName, threadInstExecutedNPO, NvRules.IFrontend.Severity_SEVERITY_HIGH, "{} < {}".format(threadInstExecutedNPO, threadInstExecuted)))
message += " This is further reduced to {0:.1f} threads per warp due to predication. The compiler may use predication to avoid an actual branch. Instead, all instructions are scheduled, but a per-thread condition code or predicate controls which threads execute the instructions. Try to avoid different execution paths within a warp when possible.".format(threadInstExecutedNPO)
message += " In addition, ensure your kernel makes use of Independent Thread Scheduling, which allows a warp to reconverge after a data-dependent conditional block by explicitly calling __syncwarp()."
msg_id = fe.message(NvRules.IFrontend.MsgType_MSG_WARNING, message)
for fm in fms:
fe.focus_metric(msg_id, fm[0], fm[1], fm[2], fm[3])
|
