[TensorIR] Cross-Thread Reduction #9360
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
MasterJH5574
requested review from
areusch,
comaniac,
Hzfengsy,
jroesch,
junrushao,
kparzysz-quic,
masahi,
merrymercy,
tqchen,
vinx13,
yzhliu,
zhiics and
ZihengJiang
as code owners
Hzfengsy
reviewed
Oct 25, 2021
|
I will do another round of review next week! |
|
Will do the review tomorrow |
Hzfengsy
reviewed
Nov 11, 2021
|
Finally got some time for a detailed code review! Will take over this PR and try to get it merged! |
junrushao
force-pushed
the
m2/cross-thread-reduction
branch
from
3ca0e56 to
a560a9b
Compare
|
Did a pass over analysis and misc changes |
|
Love the PR and very comprehensively tested implementation ❤️ |
junrushao
force-pushed
the
m2/cross-thread-reduction
branch
from
4ac6c65 to
6a162d7
Compare
|
@Hzfengsy @MasterJH5574 Should be good to go. Please take another look :-) |
MasterJH5574
commented
Nov 14, 2021
|
@Hzfengsy Could you take another look? Junru's polishing looks very good, but I myself as the author cannot approve this PR 😅. |
Hzfengsy
approved these changes
Nov 14, 2021
There was a problem hiding this comment.
LGTM. Thanks @MasterJH5574 for such a great effort on this PR.
mehrdadh
pushed a commit
to mehrdadh/tvm
that referenced
this pull request
Dec 1, 2021
* [TensorIR] Cross-Thread Reduction * Code revision on analysis and misc * Refactor TransformReductionBlock * Refactor code organization * Address comment * Use `std::make_tuple` Co-authored-by: Junru Shao <[email protected]>
mehrdadh
pushed a commit
to mehrdadh/tvm
that referenced
this pull request
Dec 1, 2021
* [TensorIR] Cross-Thread Reduction * Code revision on analysis and misc * Refactor TransformReductionBlock * Refactor code organization * Address comment * Use `std::make_tuple` Co-authored-by: Junru Shao <[email protected]>
ylc
pushed a commit
to ylc/tvm
that referenced
this pull request
Jan 7, 2022
* [TensorIR] Cross-Thread Reduction * Code revision on analysis and misc * Refactor TransformReductionBlock * Refactor code organization * Address comment * Use `std::make_tuple` Co-authored-by: Junru Shao <[email protected]>
yangulei
pushed a commit
to yangulei/tvm
that referenced
this pull request
Jan 11, 2022
* [TensorIR] Cross-Thread Reduction * Code revision on analysis and misc * Refactor TransformReductionBlock * Refactor code organization * Address comment * Use `std::make_tuple` Co-authored-by: Junru Shao <[email protected]>
ylc
pushed a commit
to ylc/tvm
that referenced
this pull request
Jan 13, 2022
* [TensorIR] Cross-Thread Reduction * Code revision on analysis and misc * Refactor TransformReductionBlock * Refactor code organization * Address comment * Use `std::make_tuple` Co-authored-by: Junru Shao <[email protected]>
|
hi @MasterJH5574 , do we currently only support block reduction with warp level |
|
I've found the related implementation, thanks. |
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.
Hi community! This PR adds cross-thread reduction support for TensorIR. After this PR, cross-thread reduction patterns in TIR can be successfully lowered.
cc @Hzfengsy @vinx13 @comaniac @junrushao1994 @jcf94 @jinhongyii @spectrometerHBH @tqchen
Co-authored-by: Wuwei Lin [email protected]
Co-authored-by: Junru Shao [email protected]
Co-authored-by: Siyuan Feng [email protected]
Co-authored-by: Hongyi Jin [email protected]
Co-authored-by: Bohan Hou [email protected]