[Encoding][Codegen] Add initial pad encoding layout attrs #19865
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
kuhar
requested review from
bjacob,
hanhanW,
antiagainst,
qedawkins,
MaheshRavishankar and
benvanik
as code owners
kuhar
force-pushed
the
pad-encoding-attrs-only
branch
from
ad3c070 to
f66bba0
Compare
kuhar
changed the title
hanhanW
approved these changes
Feb 3, 2025
Comment on lines
+159
to
+169
| if (ArrayAttr layoutsAttr = getLayouts()) { | ||
| if (!llvm::all_of(layoutsAttr.getValue(), | ||
| llvm::IsaPred<SerializedEncodingLayoutAttrInterface>)) { | ||
| return nullptr; | ||
| } | ||
|
|
||
| auto layoutsAttrArray = | ||
| llvm::to_vector_of<IREE::Encoding::EncodingLayoutAttrInterface>( | ||
| layoutsAttr.getValue()); | ||
| Value res; | ||
| for (auto attr : layoutsAttrArray) { | ||
| for (auto attr : | ||
| layoutsAttr.getAsRange<SerializedEncodingLayoutAttrInterface>()) { |
Comment on lines
+37
to
+40
| #include <cassert> | ||
| #include <cfloat> | ||
| #include <cstdint> | ||
| #include <numeric> |
There was a problem hiding this comment.
I don't remember which include style we are following. LLVM puts the system headers at the end, while Google puts them in the middle. IREE is following the mixed styles. It looks okay to me because our Codegen codebase is very close to MLIR codebase.
@benvanik @MaheshRavishankar which style should we follow?
https://google.github.io/styleguide/cppguide.html#Names_and_Order_of_Includes
https://llvm.org/docs/CodingStandards.html#include-style
There was a problem hiding this comment.
We should write it down somewhere on iree.dev. I remember the include order is something that we explicitly tweaked when switching to the llvm clang-format style: https://github.com/iree-org/iree/blob/main/compiler/.clang-format
|
I need to rebase this on top of #19879 |
These allow us to pad allocations without changing the logical tensor sizes or data layouts. Split the encoding layout attribute interface into two: * One with target-specific information that allows us to decide layouts. * One with serialized target-agnostic padding information. Signed-off-by: Jakub Kuderski <jakub@nod-labs.com>
kuhar
force-pushed
the
pad-encoding-attrs-only
branch
from
91a09e7 to
b7af694
Compare
kuhar
changed the title
ita9naiwa
pushed a commit
to ita9naiwa/iree
that referenced
this pull request
Feb 4, 2025
…9865) These allow us to pad allocations without changing the logical tensor sizes or data layouts. Split the encoding layout attribute interface into two: * One with target-specific information that allows us to decide layouts. * One with serialized target-agnostic padding information. Signed-off-by: Jakub Kuderski <jakub@nod-labs.com> Signed-off-by: Hyunsung Lee <ita9naiwa@gmail.com>
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.
These allow us to pad allocations without changing the logical tensor sizes or data layouts.
Split the encoding layout attribute interface into two: