diff --git a/doc/reference.rst b/doc/reference.rst
index fab13029c167891df4e19894ebe4db7c1aa5de6d..f7347c118fda00fa7a6ba9e9f36a26bc83c3a077 100644
--- a/doc/reference.rst
+++ b/doc/reference.rst
@@ -467,4 +467,11 @@ Options
.. autofunction:: set_options
+Controlling caching
+-------------------
+
+.. autofunction:: set_caching_enabled
+
+.. autoclass:: CacheMode
+
.. vim: tw=75:spell
diff --git a/loopy/__init__.py b/loopy/__init__.py
index 8934aebc633d53b80df7d013306e3e73552c19d5..d16d0159706d6ad09b817b327489ec7b153868a6 100644
--- a/loopy/__init__.py
+++ b/loopy/__init__.py
@@ -1325,4 +1325,38 @@ def register_function_manglers(kernel, manglers):
# }}}
+# {{{ cache control
+
+CACHING_ENABLED = True
+
+
+def set_caching_enabled(flag):
+ """Set whether :mod:`loopy` is allowed to use disk caching for its various
+ code generation stages.
+ """
+ global CACHING_ENABLED
+ CACHING_ENABLED = flag
+
+
+class CacheMode(object):
+ """A context manager for setting whether :mod:`loopy` is allowed to use
+ disk caches.
+ """
+
+ def __init__(self, new_flag):
+ self.new_flag = new_flag
+
+ def __enter__(self):
+ global CACHING_ENABLED
+ self.previous_mode = CACHING_ENABLED
+ CACHING_ENABLED = self.new_flag
+
+ def __exit__(self, exc_type, exc_val, exc_tb):
+ global CACHING_ENABLED
+ CACHING_ENABLED = self.previous_mode
+ del self.previous_mode
+
+# }}}
+
+
# vim: foldmethod=marker
diff --git a/loopy/codegen/__init__.py b/loopy/codegen/__init__.py
index 1c9eb386aa2f98c6793e624ef5393a7c07b7211c..870c9611e0c161034652d898eddcb29933a56236 100644
--- a/loopy/codegen/__init__.py
+++ b/loopy/codegen/__init__.py
@@ -341,18 +341,25 @@ def generate_code(kernel, device=None):
raise LoopyError("cannot generate code for a kernel that has not been "
"scheduled")
- if device is not None:
- device_id = device.persistent_unique_id
- else:
- device_id = None
-
- code_gen_cache_key = (kernel, device_id)
- try:
- result = code_gen_cache[code_gen_cache_key]
- logger.info("%s: code generation cache hit" % kernel.name)
- return result
- except KeyError:
- pass
+ # {{{ cache retrieval
+
+ from loopy import CACHING_ENABLED
+
+ if CACHING_ENABLED:
+ if device is not None:
+ device_id = device.persistent_unique_id
+ else:
+ device_id = None
+
+ code_gen_cache_key = (kernel, device_id)
+ try:
+ result = code_gen_cache[code_gen_cache_key]
+ logger.info("%s: code generation cache hit" % kernel.name)
+ return result
+ except KeyError:
+ pass
+
+ # }}}
from loopy.preprocess import infer_unknown_types
kernel = infer_unknown_types(kernel, expect_completion=True)
diff --git a/loopy/preprocess.py b/loopy/preprocess.py
index ab4a0734b37da4cb4b15532fa2ffbf4310b93bf9..1ca0676a1686a4ab551a8f343c558944b476376c 100644
--- a/loopy/preprocess.py
+++ b/loopy/preprocess.py
@@ -1076,18 +1076,24 @@ def preprocess_kernel(kernel, device=None):
raise LoopyError("cannot re-preprocess an already preprocessed "
"kernel")
- if device is not None:
- device_id = device.persistent_unique_id
- else:
- device_id = None
+ # {{{ cache retrieval
+
+ from loopy import CACHING_ENABLED
+ if CACHING_ENABLED:
+ if device is not None:
+ device_id = device.persistent_unique_id
+ else:
+ device_id = None
- pp_cache_key = (kernel, device_id)
- try:
- result = preprocess_cache[pp_cache_key]
- logger.info("%s: preprocess cache hit" % kernel.name)
- return result
- except KeyError:
- pass
+ pp_cache_key = (kernel, device_id)
+ try:
+ result = preprocess_cache[pp_cache_key]
+ logger.info("%s: preprocess cache hit" % kernel.name)
+ return result
+ except KeyError:
+ pass
+
+ # }}}
logger.info("%s: preprocess start" % kernel.name)
@@ -1135,7 +1141,8 @@ def preprocess_kernel(kernel, device=None):
kernel = kernel.copy(
state=kernel_state.PREPROCESSED)
- preprocess_cache[pp_cache_key] = kernel
+ if CACHING_ENABLED:
+ preprocess_cache[pp_cache_key] = kernel
return kernel
diff --git a/loopy/schedule.py b/loopy/schedule.py
index fa81ae756d62e2f5f35b943e8bdf0bb13e6dcf26..d9253dea3b10edaccf2e880f74e1cbe927c97afd 100644
--- a/loopy/schedule.py
+++ b/loopy/schedule.py
@@ -1111,15 +1111,21 @@ schedule_cache = PersistentDict("loopy-schedule-cache-v2-"+VERSION_TEXT,
def get_one_scheduled_kernel(kernel):
+ from loopy import CACHING_ENABLED
sched_cache_key = kernel
- try:
- result, ambiguous = schedule_cache[sched_cache_key]
+ from_cache = False
- logger.info("%s: schedule cache hit" % kernel.name)
- from_cache = True
- except KeyError:
- from_cache = False
+ if CACHING_ENABLED:
+ try:
+ result, ambiguous = schedule_cache[sched_cache_key]
+
+ logger.info("%s: schedule cache hit" % kernel.name)
+ from_cache = True
+ except KeyError:
+ pass
+
+ if not from_cache:
ambiguous = False
kernel_count = 0
@@ -1142,7 +1148,7 @@ def get_one_scheduled_kernel(kernel):
"schedule found, ignoring", LoopyWarning,
stacklevel=2)
- if not from_cache:
+ if CACHING_ENABLED and not from_cache:
schedule_cache[sched_cache_key] = result, ambiguous
return result
diff --git a/test/test_linalg.py b/test/test_linalg.py
index 47c7600d6dc74c4468f9ca2010801589be777789..cfc3df7da4ef30a5ffb511aa8e966e1b19a690f7 100644
--- a/test/test_linalg.py
+++ b/test/test_linalg.py
@@ -530,14 +530,15 @@ def test_ilp_race_matmul(ctx_factory):
knl = lp.split_iname(knl, "k", 2)
knl = lp.add_prefetch(knl, 'b', ["k_inner"])
- from loopy.diagnostic import WriteRaceConditionWarning
- from warnings import catch_warnings
- with catch_warnings(record=True) as warn_list:
- knl = lp.preprocess_kernel(knl)
- list(lp.generate_loop_schedules(knl))
-
- assert any(isinstance(w.message, WriteRaceConditionWarning)
- for w in warn_list)
+ with lp.CacheMode(False):
+ from loopy.diagnostic import WriteRaceConditionWarning
+ from warnings import catch_warnings
+ with catch_warnings(record=True) as warn_list:
+ knl = lp.preprocess_kernel(knl)
+ list(lp.generate_loop_schedules(knl))
+
+ assert any(isinstance(w.message, WriteRaceConditionWarning)
+ for w in warn_list)
def test_fancy_matrix_mul(ctx_factory):
diff --git a/test/test_loopy.py b/test/test_loopy.py
index a3f07f299bff6c9b3121e963dd1a612bd6e89725..94bd89acac7cc9e8b4f31fcebe87b3cab393b656 100644
--- a/test/test_loopy.py
+++ b/test/test_loopy.py
@@ -933,13 +933,14 @@ def test_ilp_write_race_detection_global(ctx_factory):
knl = lp.preprocess_kernel(knl, ctx.devices[0])
- from loopy.diagnostic import WriteRaceConditionWarning
- from warnings import catch_warnings
- with catch_warnings(record=True) as warn_list:
- list(lp.generate_loop_schedules(knl))
-
- assert any(isinstance(w.message, WriteRaceConditionWarning)
- for w in warn_list)
+ with lp.CacheMode(False):
+ from loopy.diagnostic import WriteRaceConditionWarning
+ from warnings import catch_warnings
+ with catch_warnings(record=True) as warn_list:
+ list(lp.generate_loop_schedules(knl))
+
+ assert any(isinstance(w.message, WriteRaceConditionWarning)
+ for w in warn_list)
def test_ilp_write_race_avoidance_local(ctx_factory):