Implement AArch64 ABI #14519
Merged
Implement AArch64 ABI #14519
+308
−10
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
yuyichao
added
system:arm
yuyichao
force-pushed
the
yyc/aarch64-abi
branch
2 times, most recently
from
b72c604 to
4ecabcc
Compare
yuyichao
force-pushed
the
yyc/aarch64-abi
branch
from
4ecabcc to
d0fbd71
Compare
|
i believe tests will be covered by #14215 since this is a strict improvement, i'm going to merge this now. |
|
Nice!! |
|
Can we backport this to support it on 0.4? |
|
Putting the label on for now. |
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Outdated suggestions cannot be applied.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 fixes a few SegFault in the test (similar to the ones on ARM) and make the ccall test pass.
A few notes about this PR,
va_listand we don't have that yet.abi_aarch64.cppare comments copied from the calling convension doc, which I use when writing this PR. I think keeping those lines can help reviewing and other people reading the code.__fp16(Float16) is allowed as C argument on AArch64 and is passed in a SIMD and floating-point register if possible (different fromInt16).byvalattribute to abyRefargument breaks the code generation for that argument. See repro here. Note that in the working case themov x0, spcorrectly set the pass-by-pointer argument while in the broken case this instruction is missing. This is on LLVM 3.7.0 and I couldn't find any docs about how thebyvalattribute affects ABI or how is it processes in a target-dependent way so I'm not sure if this is an LLVM bug or if we are using it incorrectly (on AArch64 at least) (@Keno @vtjnash ). Clang doesn't emit such attribute for these arguments so I just leave it out.__int128in C is 16 bytes on AArch64 so I increase theMAX_ALIGNfor AArch64. This also seems to be the case on x64 although I haven't find any document for this.