[mono][interp] Align simd types to 16 bytes by default #81219
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
|
Tagging subscribers to this area: @BrzVlad Issue DetailsAll simd interp vars (args, IL locals and other allocated vars) are now aligned to 16 byte offsets, instead of the 8 byte default alignment. We use the general
|
BrzVlad
force-pushed
the
feature-interp-align-v128
branch
from
0b8d5f2 to
b4472bc
Compare
This was referenced Jan 27, 2023
BrzVlad
force-pushed
the
feature-interp-align-v128
branch
from
b4472bc to
215d2c3
Compare
|
/azp run runtime-wasm |
|
Azure Pipelines successfully started running 1 pipeline(s). |
|
/azp run runtime-wasm-libtests |
|
Azure Pipelines successfully started running 1 pipeline(s). |
BrzVlad
force-pushed
the
feature-interp-align-v128
branch
from
215d2c3 to
d07cb1a
Compare
kotlarmilos
approved these changes
Jan 30, 2023
| @@ -4167,8 +4187,13 @@ interp_method_compute_offsets (TransformData *td, InterpMethod *imethod, MonoMet | |||
| return; | |||
| } | |||
| } | |||
| offset += align - 1; | |||
| offset &= ~(align - 1); | |||
| MonoClass *klass = mono_class_from_mono_type_internal (header->locals [i]); | |||
There was a problem hiding this comment.
Could be encapsulated if used on multiple occasions
BrzVlad
force-pushed
the
feature-interp-align-v128
branch
from
d07cb1a to
51043ee
Compare
All interp vars (args, il locals and other allocated vars) are now aligned to 16 byte offsets.
Assumption that return offset is identical to location of param offset for this opcode is no longer true. Set the param_offset explicitly, separate from the return, similar to dotnet#81017
BrzVlad
force-pushed
the
feature-interp-align-v128
branch
from
51043ee to
5491767
Compare
ghost
locked as resolved and limited conversation to collaborators
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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.
All simd interp vars (args, IL locals and other allocated vars) are now aligned to 16 byte offsets, instead of the 8 byte default alignment.
We use the general
MonoClass->simd_typefor now, which encapsulates vectors of all sizes, to determine if we align them. This PR has no benefit by itself, it is just prerequisite for addition of intrisics for vectorized operations, and submitted by itself for independent testing purposes.