[SYCL-MLIR] Improve CodeGen for Add and Sub expressions #7421
+991
−94
Conversation
This file contains hidden or 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
victor-eds
commented
Nov 16, 2022
Contributor
Author
victor-eds
left a comment
victor-eds
left a comment
There was a problem hiding this comment.
NOTE: We cannot generate NUW, NSW, exact or inbounds op flags. Not supported in MLIR yet.
Contributor
Author
|
First of two PRs addressing binary arithmetic operations. |
victor-eds
force-pushed
the
visit-add-sub
branch
from
df208f3 to
79562cc
Compare
whitneywhtsang
approved these changes
Nov 18, 2022
Enable addition/subtraction of floating point vectors. Make integer and pointer addition commutative. Instead of generating an incorrect `getelementptr` instruction with a `null` pointer argument, generate an `inttoptr` using the index as an argument. Use polygeist.subindex instead of polygeist.memref2pointer+llvm.getelementptr+polygeist.pointer2memref when subtracting an index to a pointer. Cast pointers to void and functions to i8 before operating. Signed-off-by: Victor Perez <[email protected]>
victor-eds
force-pushed
the
visit-add-sub
branch
from
65f2fa2 to
b0b8e45
Compare
etiotto
approved these changes
Nov 21, 2022
etiotto
left a comment
etiotto
left a comment
There was a problem hiding this comment.
LGTM - Added some comments on the lit tests and other minor comments.
etiotto
pushed a commit
to etiotto/intel-llvm
that referenced
this pull request
Nov 22, 2022
Enable addition/subtraction of floating point vectors. Make integer and pointer addition commutative. Instead of generating an incorrect `getelementptr` instruction with a `null` pointer argument, generate an `inttoptr` using the index as an argument. Use polygeist.subindex instead of polygeist.memref2pointer+llvm.getelementptr+polygeist.pointer2memref when subtracting an index to a pointer. Cast pointers to void and functions to i8 before operating. Signed-off-by: Victor Perez <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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.
Enable addition/subtraction of floating point vectors.
Make integer and pointer addition commutative.
Instead of generating an incorrect
getelementptrinstruction with anullpointer argument, generate aninttoptrusing the index as an argument.Use polygeist.subindex instead of
polygeist.memref2pointer+llvm.getelementptr+polygeist.pointer2memref when subtracting an index to a pointer.
Cast pointers to void and functions to i8 before operating.
Signed-off-by: Victor Perez [email protected]