Wed, Dec 2
@Meinersbur Are the changes fine for you?
Tue, Dec 1
Mon, Nov 30
LGTM. Please wait for the approval from someone from the OpenMP side.
Just have a small comment. Otherwise I think it looks good to me. Please wait for the approval of someone on the OpenMP side.
Wed, Nov 25
Allow wait clause multiple times
Push correct patch version
Close this patch since clauses have changed For OpenACC 3.1 and these tests will be added to a new patch where clauses will be changes as well.
Tue, Nov 24
I'm fine with leaving it there for now, looking forward to see everything upstream!
If this can't be tested / is dead code, I'm not super comfortable having this in-tree. What is the plan for fir-dev development to move upstream, I'm still quite puzzled and don't understand why the development does not happen in tree, I think I asked on the mailing list but never got an answer.
I'd like to see IR -> IR unit-test (mlir-opt style) provided here. I don't think it is reasonable to upstream code without proper testing.
Rebase + add couple of test cases
Some small comments
I guess you need to rebase this now that the memory-order-clause is in.
Mon, Nov 23
Sat, Nov 21
Fine for me too
Fri, Nov 20
Remove blank lines added by mistake
IIUC, the rational for adding these tests was to just add clauses which might be missing from the list of AccClause ?
Fix spec references in comment
Thu, Nov 19
Remove commented line
Wed, Nov 18
Small style comment
Tue, Nov 17
@Meinersbur Thanks for the review. Just updated the patch.
Address review comments
I see. We should add some unittests for the functions in that file.
Fri, Nov 13
I think it is very unusual to have a library in LLVM that isn't used in LLVM and so does not come with any testing, so yes it'd be better to do so.
I'm not actually sure if llvm-config just relies implicitly on this?
@clementval : it isn't clear why this is in llvm/lib/Frontend/? Should this be a top-level frontend/lib kind of structure? There isn't any user of this inside LLVM right now other than flang and it isn't clear to me how it connects to LLVM itself.
Are there any OpenACC specific tests?
AFAICT after a quick look through the monorepo, there doesn't seem to be any llvm-specific tests for OpenACC.
Thu, Nov 12
Mon, Nov 9
Thu, Nov 5
Rebase + address review comment
Rebase + address review comments
Ok from my side. Maybe @kiranchandramohan wants to double check the latest changes.
Thanks for moving those to TableGen and common infra. This makes lots of sense. Small comment about the list with comma.
Wed, Nov 4