[cli/trampolines] Clean up definitions, fix win32 exporting issue #39543
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
vtjnash
reviewed
Feb 5, 2021
|
|
||
| // aarch64 on mac requires some special assembler syntax for both calculating memory | ||
| // offsets and even just the assembler statement separator token | ||
| #if defined(__aarch64__) |
There was a problem hiding this comment.
why wouldn't this go in trampolines_aarch64.S, when it is only defined and used there?
There was a problem hiding this comment.
because it's nice to have the .S files be as free of preprocessor stuff as possible; it makes it much easier to, with your eyeball, compare the trampolines across platforms.
|
Adding @vchuravy because I changed the ppc64le trampoline again; it clobbers one less register now and passes libblastrampoline tests on power now, so I figure we probably want that change here as well. |
b8e70ec to
2b38b20
Compare
vchuravy
approved these changes
Feb 5, 2021
f35914c to
c4def51
Compare
Our win32 trampolines weren't getting exported properly. Also take the opportunity to properly organize after debugging libblastrampoline and finding the current organizational strategy lacking. In the future, if we ever want to mangle the symbols from these trampolines, we now have a convenient `CNAME()` that is properly invoked from all trampolines and can consistently mangle names.
According to the examples on page 138-139 of the ELV v2 OpenPOWER ABI [0], we should avoid clobbering r9 and only touch r2 and r12. This fixed a test failure in libblastrampoline, so it's probably needed here as well. [0]: https://members.openpowerfoundation.org/document/dl/576
c4def51 to
f0b0cd8
Compare
ElOceanografo
pushed a commit
to ElOceanografo/julia
that referenced
this pull request
May 4, 2021
antoine-levitt
pushed a commit
to antoine-levitt/julia
that referenced
this pull request
May 9, 2021
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.
Our win32 trampolines weren't getting exported properly. Also take the
opportunity to properly organize after debugging libblastrampoline and
finding the current organizational strategy lacking.
In the future, if we ever want to mangle the symbols from these
trampolines, we now have a convenient
CNAME()that is properly invokedfrom all trampolines and can consistently mangle names.