[SVE] Add support for the SVE target.#180
Merged
fpetrogalli merged 1 commit intomasterfrom Mar 14, 2018
Merged
Conversation
This commit adds support for the AArch64 Scalable Vector Extension (SVE) [1]. The vector functions are provided to target Vector Length Agnostic (VLA) execution [2]. To build SLEEF with SVE support, a compiler that support the SVE Arm C Language Extensions (ACLE) [2] must be used. At the time of publishing this patch, the only compiler with SVE ACLE support is Arm Compiler for HPC [3]. The Cmake configuration expectes Arm Instruction Emulator (ArmIE) [4] to execute the tests on native AArch64 hardware without SVE support. The SVE target is build without taking advantage of the native masking capabilities of SVE. This will be targeted in a upcoming release of SLEEF, together with the AVX512F native masking capabilities [5]. Additional changes introduced in this patch are: 1. The mkrename* script have been modified to support VLA names in the functions. In particular, 'x' is used to represent the vector length of the SVE symbols. 2. '__sizeless_struct' is a prototype language extension only implemented by Arm Compiler For HPC [3] to allow the declaration of SVE tuple types as described in section 3.4 of Arm C Language Extensions for SVE [2]. 3. A new 'iutsve' executable is generated to test the SVE functions. [1] https://developer.arm.com/products/software-development-tools/hpc/sve [2] https://developer.arm.com/docs/100987/0000 [3] https://developer.arm.com/products/software-development-tools/hpc/arm-compiler-for-hpc [4] https://developer.arm.com/products/software-development-tools/hpc/arm-instruction-emulator [5] #142
shibatch
approved these changes
Mar 14, 2018
Owner
shibatch
left a comment
There was a problem hiding this comment.
Looks good to me.
I confirmed that the test passes with SVE on my environment.
We need to think of a way to enable testing on a CI environment, though.
Author
|
Did you use |
Owner
|
yes |
Author
|
Might be a good idea to report the |
Owner
|
Here you are. |
Author
|
Thanks. What system are you using? |
Owner
|
This one is Firefly-RK3399. |
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
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 commit adds support for the AArch64 Scalable Vector Extension
(SVE) [1]. The vector functions are provided to target Vector Length
Agnostic (VLA) execution [2].
To build SLEEF with SVE support, a compiler that support the SVE Arm C
Language Extensions (ACLE) [2] must be used.
At the time of publishing this patch, the only compiler with SVE ACLE
support is Arm Compiler for HPC [3].
The Cmake configuration expectes Arm Instruction Emulator (ArmIE) [4]
to execute the tests on native AArch64 hardware without SVE support.
The SVE target is build without taking advantage of the native masking
capabilities of SVE. This will be targeted in a upcoming release of
SLEEF, together with the AVX512F native masking capabilities [5].
Additional changes introduced in this patch are:
The mkrename* script have been modified to support VLA names in the
functions. In particular, 'x' is used to represent the vector
length of the SVE symbols.
'__sizeless_struct' is a prototype language extension only
implemented by Arm Compiler For HPC [3] to allow the declaration of
SVE tuple types as described in section 3.4 of Arm C Language
Extensions for SVE [2].
A new 'iutsve' executable is generated to test the SVE functions.
[1] https://developer.arm.com/products/software-development-tools/hpc/sve
[2] https://developer.arm.com/docs/100987/0000
[3] https://developer.arm.com/products/software-development-tools/hpc/arm-compiler-for-hpc
[4] https://developer.arm.com/products/software-development-tools/hpc/arm-instruction-emulator
[5] #142