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.

Sign up for GitHub

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

Replace custom CUDA bindings previously provided by RMM with official CUDA Python bindings #10008

Merged
merged 3 commits into from
Jan 20, 2022
Merged

Replace custom CUDA bindings previously provided by RMM with official CUDA Python bindings #10008

Changes from 2 commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
86f00a3
Replace custom CUDA bindings previously provided by RMM with official…
shwina Jan 10, 2022
2454435
Unused import
shwina Jan 11, 2022
206535d
Add cuda-python as explicit dependency.
shwina Jan 12, 2022
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
50 changes: 22 additions & 28 deletions python/cudf/cudf/utils/gpu_utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,10 +15,10 @@ def validate_setup():

import warnings

from cuda.cudart import cudaError_t

from rmm._cuda.gpu import (
CUDARuntimeError,
cudaDeviceAttr,
cudaError,
deviceGetName,
driverGetVersion,
getDeviceAttribute,
Expand All @@ -30,30 +30,30 @@ def _try_get_old_or_new_symbols():
try:
# CUDA 10.2+ symbols
return [
cudaError.cudaErrorDeviceUninitialized,
cudaError.cudaErrorTimeout,
cudaError_t.cudaErrorDeviceUninitialized,
cudaError_t.cudaErrorTimeout,
]
except AttributeError:
# CUDA 10.1 symbols
return [cudaError.cudaErrorDeviceUninitilialized]
return [cudaError_t.cudaErrorDeviceUninitilialized]

notify_caller_errors = {
cudaError.cudaErrorInitializationError,
cudaError.cudaErrorInsufficientDriver,
cudaError.cudaErrorInvalidDeviceFunction,
cudaError.cudaErrorInvalidDevice,
cudaError.cudaErrorStartupFailure,
cudaError.cudaErrorInvalidKernelImage,
cudaError.cudaErrorAlreadyAcquired,
cudaError.cudaErrorOperatingSystem,
cudaError.cudaErrorNotPermitted,
cudaError.cudaErrorNotSupported,
cudaError.cudaErrorSystemNotReady,
cudaError.cudaErrorSystemDriverMismatch,
cudaError.cudaErrorCompatNotSupportedOnDevice,
cudaError_t.cudaErrorInitializationError,
cudaError_t.cudaErrorInsufficientDriver,
cudaError_t.cudaErrorInvalidDeviceFunction,
cudaError_t.cudaErrorInvalidDevice,
cudaError_t.cudaErrorStartupFailure,
cudaError_t.cudaErrorInvalidKernelImage,
cudaError_t.cudaErrorAlreadyAcquired,
cudaError_t.cudaErrorOperatingSystem,
cudaError_t.cudaErrorNotPermitted,
cudaError_t.cudaErrorNotSupported,
cudaError_t.cudaErrorSystemNotReady,
cudaError_t.cudaErrorSystemDriverMismatch,
cudaError_t.cudaErrorCompatNotSupportedOnDevice,
*_try_get_old_or_new_symbols(),
vyasr marked this conversation as resolved.
Show resolved Hide resolved
cudaError.cudaErrorUnknown,
cudaError.cudaErrorApiFailureBase,
cudaError_t.cudaErrorUnknown,
cudaError_t.cudaErrorApiFailureBase,
}

try:
Expand All @@ -68,11 +68,7 @@ def _try_get_old_or_new_symbols():
# Cupy throws RunTimeException to get GPU count,
# hence obtaining GPU count by in-house cpp api above

# 75 - Indicates to get "cudaDevAttrComputeCapabilityMajor" attribute
# 0 - Get GPU 0
major_version = getDeviceAttribute(
cudaDeviceAttr.cudaDevAttrComputeCapabilityMajor, 0
)
major_version = getDeviceAttribute("ComputeCapabilityMajor", 0)

if major_version >= 6:
# You have a GPU with NVIDIA Pascal™ architecture or better
Expand All @@ -86,9 +82,7 @@ def _try_get_old_or_new_symbols():
pass
else:
device_name = deviceGetName(0)
minor_version = getDeviceAttribute(
cudaDeviceAttr.cudaDevAttrComputeCapabilityMinor, 0
)
minor_version = getDeviceAttribute("ComputeCapabilityMinor", 0)
warnings.warn(
f"You will need a GPU with NVIDIA Pascal™ or "
f"newer architecture"
Expand Down