Enable compiling with system cub. #7232
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Eventually, we will have to move on to using system cub, with ctk 11.4 and custom cub there are lots of warnings. I think this might be the last version we can workaround it. |
trivialfis
force-pushed
the
limit-cub-sort-size
branch
from
be6dbc0 to
ab7d37d
Compare
trivialfis
commented
Sep 16, 2021
| }); | ||
| // shrink down to pair | ||
| auto fptp_it_out = thrust::make_transform_output_iterator( | ||
| dh::tbegin(d_fptp), [=] __device__(Triple const &t) { | ||
| return thrust::make_pair(thrust::get<1>(t), thrust::get<2>(t)); | ||
| dh::TypedDiscard<Triple>{}, [d_fptp] __device__(Triple const &t) { |
There was a problem hiding this comment.
Originally this reduces the triple to pair, cub in CUDA 11.0~11.2 doesn't handle changed type. So here I just write the result in lambda and use discard iterator to keep the output type unchanged.
hcho3
approved these changes
Sep 17, 2021
* For now we need to limit the number of items to INT32_MAX This is caused by latest RMM brings cub into its include path, so when XGBoost is compiled with RMM the paths conflict.
trivialfis
force-pushed
the
limit-cub-sort-size
branch
from
ab7d37d to
5a3a1ae
Compare
|
@trivialfis Could you cherry-pick this PR to 1.4.0 branch? |
|
No, I'm planning for a new release |
NvTimLiu
pushed a commit
to NvTimLiu/spark-xgboost
that referenced
this pull request
Nov 1, 2021
NvTimLiu
added a commit
to NvTimLiu/spark-xgboost
that referenced
this pull request
Nov 1, 2021
Enable compiling with system cub. (dmlc#7232) See merge request nvspark/xgboost!394
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
- This works only with CUDA 11.4, for CUDA 11.2 there are some obscure errors with inclusive scan and custom iterators.