Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Enable CEC for strings_udf #11884

Merged
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
17 changes: 5 additions & 12 deletions ci/gpu/build.sh
Original file line number Diff line number Diff line change
Expand Up @@ -295,22 +295,15 @@ py.test -n 8 --cache-clear --basetemp="$WORKSPACE/custreamz-cuda-tmp" --junitxml
gpuci_logger "Installing strings_udf"
gpuci_mamba_retry install strings_udf -c "${CONDA_BLD_DIR}" -c "${CONDA_ARTIFACT_PATH}"

# only install strings_udf after cuDF is finished testing without its presence
cd "$WORKSPACE/python/strings_udf/strings_udf"
gpuci_logger "Python py.test for strings_udf"
py.test -n 8 --cache-clear --basetemp="$WORKSPACE/strings-udf-cuda-tmp" --junitxml="$WORKSPACE/junit-strings-udf.xml" -v --cov-config=.coveragerc --cov=strings_udf --cov-report=xml:"$WORKSPACE/python/strings_udf/strings-udf-coverage.xml" --cov-report term tests

STRINGS_UDF_PYTEST_RETCODE=0
py.test -n 8 --cache-clear --basetemp="$WORKSPACE/strings-udf-cuda-tmp" --junitxml="$WORKSPACE/junit-strings-udf.xml" -v --cov-config=.coveragerc --cov=strings_udf --cov-report=xml:"$WORKSPACE/python/strings_udf/strings-udf-coverage.xml" --cov-report term tests || STRINGS_UDF_PYTEST_RETCODE=$?
# retest cuDF UDFs
cd "$WORKSPACE/python/cudf/cudf"
gpuci_logger "Python py.test retest cuDF UDFs"
py.test -n 8 --cache-clear --basetemp="$WORKSPACE/cudf-cuda-strings-udf-tmp" --ignore="$WORKSPACE/python/cudf/cudf/benchmarks" --junitxml="$WORKSPACE/junit-cudf-strings-udf.xml" -v --cov-config="$WORKSPACE/python/cudf/.coveragerc" --cov=cudf --cov-report=xml:"$WORKSPACE/python/cudf/cudf-strings-udf-coverage.xml" --cov-report term --dist=loadscope tests/test_udf_masked_ops.py

if [ ${STRINGS_UDF_PYTEST_RETCODE} -eq 5 ]; then
echo "No strings UDF tests were run, but this script will continue to execute."
elif [ ${STRINGS_UDF_PYTEST_RETCODE} -ne 0 ]; then
exit ${STRINGS_UDF_PYTEST_RETCODE}
else
cd "$WORKSPACE/python/cudf/cudf"
gpuci_logger "Python py.test retest cuDF UDFs"
py.test -n 8 --cache-clear --basetemp="$WORKSPACE/cudf-cuda-strings-udf-tmp" --ignore="$WORKSPACE/python/cudf/cudf/benchmarks" --junitxml="$WORKSPACE/junit-cudf-strings-udf.xml" -v --cov-config="$WORKSPACE/python/cudf/.coveragerc" --cov=cudf --cov-report=xml:"$WORKSPACE/python/cudf/cudf-strings-udf-coverage.xml" --cov-report term --dist=loadscope tests
fi

# Run benchmarks with both cudf and pandas to ensure compatibility is maintained.
# Benchmarks are run in DEBUG_ONLY mode, meaning that only small data sizes are used.
Expand Down
2 changes: 1 addition & 1 deletion conda/recipes/strings_udf/meta.yaml
Original file line number Diff line number Diff line change
Expand Up @@ -50,7 +50,7 @@ requirements:
- cudf ={{ version }}
- {{ pin_compatible('cudatoolkit', max_pin='x', min_pin='x') }}
- cachetools
- ptxcompiler # [linux64] # CUDA enhanced compatibility. See https://github.com/rapidsai/ptxcompiler
- ptxcompiler >=0.7.0 # [linux64] # CUDA enhanced compatibility. See https://github.com/rapidsai/ptxcompiler
test: # [linux64]
requires: # [linux64]
- cudatoolkit {{ cuda_version }}.* # [linux64]
Expand Down
17 changes: 9 additions & 8 deletions python/cudf/cudf/core/udf/__init__.py
Original file line number Diff line number Diff line change
Expand Up @@ -23,17 +23,20 @@
| {types.boolean}
)
_STRING_UDFS_ENABLED = False

try:
import strings_udf
from strings_udf import ptxpath

if ptxpath:
utils.ptx_files.append(ptxpath)
brandon-b-miller marked this conversation as resolved.
Show resolved Hide resolved

if strings_udf.ENABLED:
wence- marked this conversation as resolved.
Show resolved Hide resolved
from . import strings_typing # isort: skip
from . import strings_lowering # isort: skip
from strings_udf import ptxpath
from strings_udf._lib.cudf_jit_udf import to_string_view_array
from strings_udf._typing import str_view_arg_handler, string_view

# add an overload of MaskedType.__init__(string_view, bool)
from . import strings_typing # isort: skip
from . import strings_lowering # isort: skip

cuda_lower(api.Masked, strings_typing.string_view, types.boolean)(
masked_lowering.masked_constructor
)
Expand All @@ -47,13 +50,11 @@
utils.launch_arg_getters[dtype("O")] = to_string_view_array
utils.masked_array_types[dtype("O")] = string_view
utils.JIT_SUPPORTED_TYPES |= STRING_TYPES
utils.ptx_files.append(ptxpath)

utils.arg_handlers.append(str_view_arg_handler)
row_function.itemsizes[dtype("O")] = string_view.size_bytes

_STRING_UDFS_ENABLED = True
else:
del strings_udf

except ImportError as e:
# allow cuDF to work without strings_udf
Expand Down
Empty file.
137 changes: 66 additions & 71 deletions python/strings_udf/strings_udf/__init__.py
Original file line number Diff line number Diff line change
@@ -1,36 +1,20 @@
# Copyright (c) 2022, NVIDIA CORPORATION.
import glob
import os
import re
import subprocess
import sys

from cubinlinker.patch import _numba_version_ok, get_logger, new_patched_linker
from numba import cuda
from ptxcompiler.patch import CMD
from numba.cuda.cudadrv.driver import Linker
from ptxcompiler.patch import NO_DRIVER, safe_get_versions

from . import _version

__version__ = _version.get_versions()["version"]

ENABLED = False
logger = get_logger()


def compiler_from_ptx_file(path):
"""Parse a PTX file header and extract the CUDA version used to compile it.
Here is an example PTX header that this function should parse:
// Generated by NVIDIA NVVM Compiler
//
// Compiler Build ID: CL-30672275
// Cuda compilation tools, release 11.5, V11.5.119
// Based on NVVM 7
"""
file = open(path).read()
major, minor = (
re.search(r"Cuda compilation tools, release ([0-9\.]+)", file)
.group(1)
.split(".")
)
return int(major), int(minor)
# tracks the version of CUDA used to build the c++ and PTX components
STRINGS_UDF_PTX_VERSION = (11, 5)


def _get_appropriate_file(sms, cc):
Expand All @@ -41,59 +25,70 @@ def _get_appropriate_file(sms, cc):
return None


# adapted from PTXCompiler
cp = subprocess.run([sys.executable, "-c", CMD], capture_output=True)
if cp.returncode == 0:
# must have a driver to proceed
versions = [int(s) for s in cp.stdout.strip().split()]
driver_version = tuple(versions[:2])
runtime_version = tuple(versions[2:])
def maybe_patch_numba_linker(driver_version):
# Numba thinks cubinlinker is only needed if the driver is older than the ctk
# but when strings_udf is present, it might also need to patch because the PTX
# file strings_udf relies on may be newer than the driver as well
if driver_version < STRINGS_UDF_PTX_VERSION:
logger.debug(
"Driver version %s.%s needs patching due to strings_udf"
% driver_version
)
if _numba_version_ok:
logger.debug("Patching Numba Linker")
Linker.new = new_patched_linker
else:
logger.debug("Cannot patch Numba Linker - unsupported version")


def _get_ptx_file():
if "RAPIDS_NO_INITIALIZE" in os.environ:
cc = int(os.environ.get("STRINGS_UDF_CC", "52"))
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The idea here is that downstream consumers (like dask-cuda) that set RAPIDS_NO_INITIALIZE have a way to advertise the compute capability of the device they are running on (without creating a context). Do we care about bikeshedding this name?

else:
dev = cuda.get_current_device()

# CUDA enhanced compatibility not yet enabled
if driver_version >= runtime_version:
# Load the highest compute capability file available that is less than
# the current device's.
dev = cuda.get_current_device()
cc = int("".join(str(x) for x in dev.compute_capability))
files = glob.glob(
os.path.join(os.path.dirname(__file__), "shim_*.ptx")
files = glob.glob(os.path.join(os.path.dirname(__file__), "shim_*.ptx"))
if len(files) == 0:
raise RuntimeError(
"This strings_udf installation is missing the necessary PTX "
f"files for compute capability {cc}. "
"Please file an issue reporting this error and how you "
"installed cudf and strings_udf."
wence- marked this conversation as resolved.
Show resolved Hide resolved
"https://github.com/rapidsai/cudf/issues"
)
if len(files) == 0:
raise RuntimeError(
"This strings_udf installation is missing the necessary PTX "
"files. Please file an issue reporting this error and how you "
"installed cudf and strings_udf."
)

suffix_a_sm = None
regular_sms = []

for f in files:
file_name = os.path.basename(f)
sm_number = file_name.rstrip(".ptx").lstrip("shim_")
if sm_number.endswith("a"):
processed_sm_number = int(sm_number.rstrip("a"))
if processed_sm_number == cc:
suffix_a_sm = (processed_sm_number, f)
else:
regular_sms.append((int(sm_number), f))

regular_result = None

if regular_sms:
regular_result = _get_appropriate_file(regular_sms, cc)

if suffix_a_sm is None and regular_result is None:
raise RuntimeError(
"This strings_udf installation is missing the necessary PTX "
f"files that are <={cc}."
)
elif suffix_a_sm is not None:
ptxpath = suffix_a_sm[1]
else:
ptxpath = regular_result[1]

if driver_version >= compiler_from_ptx_file(ptxpath):
ENABLED = True
regular_sms = []

for f in files:
file_name = os.path.basename(f)
sm_number = file_name.rstrip(".ptx").lstrip("shim_")
if sm_number.endswith("a"):
processed_sm_number = int(sm_number.rstrip("a"))
if processed_sm_number == cc:
return f
else:
del ptxpath
regular_sms.append((int(sm_number), f))

regular_result = None

if regular_sms:
regular_result = _get_appropriate_file(regular_sms, cc)

if regular_result is None:
raise RuntimeError(
"This strings_udf installation is missing the necessary PTX "
f"files that are <={cc}."
)
else:
return regular_result[1]


ptxpath = None
versions = safe_get_versions()
if versions != NO_DRIVER:
driver_version, runtime_version = versions
maybe_patch_numba_linker(driver_version)
ptxpath = _get_ptx_file()
3 changes: 0 additions & 3 deletions python/strings_udf/strings_udf/tests/test_string_udfs.py
Original file line number Diff line number Diff line change
Expand Up @@ -15,9 +15,6 @@
from strings_udf._lib.cudf_jit_udf import to_string_view_array
from strings_udf._typing import str_view_arg_handler, string_view

if not strings_udf.ENABLED:
pytest.skip("Strings UDF not enabled.", allow_module_level=True)


def get_kernel(func, dtype):
"""
Expand Down