gccbuiltins: Add AMD GCN and Nvidia PTX #3411
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
|
Shouldn't. LLVM handles the intrinsics. |
|
Yeah that's what I thought too. - I've only checked the generated NVPTX file, which contains ~550 builtins, and this works: // ldc2 -c bla.d -betterC -mtriple=nvptx64 -mcpu=sm_50
import ldc.gccbuiltins_nvvm;
double foo(double a, double b) { return __nvvm_add_rm_d(a, b); }My LLVM doesn't include the AMDGPU backend (which AFAICT features 2 targets, |
| set(target_arch "AArch64;ARM;Mips;PowerPC;SystemZ;X86") | ||
| set(target_name "aarch64;arm;mips;ppc;s390;x86") | ||
| set(target_arch "AArch64;AMDGPU;ARM;Mips;NVPTX;PowerPC;SystemZ;X86") | ||
| set(target_name "aarch64;amdgcn;arm;mips;nvvm;ppc;s390;x86") |
There was a problem hiding this comment.
Why is this done as a whitelist anyway? Are there any targets we want to exclude builtins for?
There was a problem hiding this comment.
- IIRC, there are targets like BPF and Hexagon (no idea what those are) with a handful of GCC builtins.
- AFAICT, you need to know the target prefix (
nvvm) for the tblgen invocation. - We additionally filter this list and only keep those for which the LLVM backend is available. That restriction can probably be lifted for a uniform
ldc.gccbuiltins_*imports set, independent from the LLVM build config.
There was a problem hiding this comment.
AFAICT, you need to know the target prefix (nvvm) for the tblgen invocation.
Thats annoying.
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Outdated suggestions cannot be applied.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 is sort of a reboot of #2556. @thewilsonator: The generated files are .di headers, do those really need any special dcompute handling?