mirror-github-bspeice/cuburn
1
0
Fork 0
mirror of https://github.com/stevenrobertson/cuburn.git synced 2025-02-05 03:30:05 -05:00
Code Issues Actions Packages Projects Releases Wiki Activity

Grow launch sizes and synchronize if they pile up

Browse Source
This commit is contained in:
Steven Robertson 2017-05-15 00:43:10 -07:00
parent 5402838a74
commit 7dc58a0e1c
1 changed files with 14 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
cuburn/render.py
View File

@ -335,14 +335,25 @@ class RenderManager(ClsMod):
# GPUs from locking up and to give us a chance to flush some stuff. # GPUs from locking up and to give us a chance to flush some stuff.
hidden_stream = cuda.Stream() hidden_stream = cuda.Stream()
iter_stream_left, iter_stream_right = self.stream_a, hidden_stream iter_stream_left, iter_stream_right = self.stream_a, hidden_stream
BLOCK_SIZE = 4 block_size = 4
while nrounds: while nrounds:
n = min(nrounds, BLOCK_SIZE) n = min(nrounds, block_size)
now = time.time()
launch('iter', rdr.mod, iter_stream_left, (32, 8, 1), (nts, n), launch('iter', rdr.mod, iter_stream_left, (32, 8, 1), (nts, n),
self.fb.d_front, self.fb.d_left, self.fb.d_front, self.fb.d_left,
self.fb.d_rb, self.fb.d_seeds, self.fb.d_points, self.fb.d_rb, self.fb.d_seeds, self.fb.d_points,
self.fb.d_uleft, self.info_a.d_params) self.fb.d_uleft, self.info_a.d_params)
delta = time.time() - now
if delta > 0.1:
# More than 100ms passed attempting to launch. The GPU is likely
# out of queued execution resources on a long render, and scheduling
# additional work will just keep spinning the CPU at 100%.
# Do a blocking sync to free up resources. This may slightly reduce
# parallelism but makes it a whole heck of a lot easier to keep
# using the computer while things render.
print 'Launches became blocking, synchronizing'
iter_stream_right.synchronize()
# Make sure the other stream is done flushing before we start # Make sure the other stream is done flushing before we start
iter_stream_left.wait_for_event(cuda.Event().record(iter_stream_right)) iter_stream_left.wait_for_event(cuda.Event().record(iter_stream_right))
@ -355,6 +366,7 @@ class RenderManager(ClsMod):
self.fb.flip_side() self.fb.flip_side()
iter_stream_left, iter_stream_right = iter_stream_right, iter_stream_left iter_stream_left, iter_stream_right = iter_stream_right, iter_stream_left
nrounds -= n nrounds -= n
block_size += block_size / 2
# Always wait on all events in the hidden stream before continuing on A # Always wait on all events in the hidden stream before continuing on A
self.stream_a.wait_for_event(cuda.Event().record(hidden_stream)) self.stream_a.wait_for_event(cuda.Event().record(hidden_stream))