jhuber6 added a comment.
In D87946#2283049 <https://reviews.llvm.org/D87946#2283049>, @jdoerfert wrote:
> I guess you can try to use `sed` to update the tests.
That's what I did for all the clang tests, they all pass as far as I know and I
can use the built clang to correctly build offloaded applications for CUDA at
least. The tests that are problematic are the OpenMPOpt tests because I'm
assuming they had a hard coded argument index somewhere into one of the runtime
functions that needs to be incremented by one now.
I also didn't add the an ident_t argument to all of the runtime functions, only
the ones that I could see being generated by `CGOpenMPRuntime.cpp`. I just made
those functions pass in a `nullptr` if they call into one of the functions that
requires an ident_t.
================
Comment at: openmp/libomptarget/src/interface.cpp:149
-EXTERN void __tgt_target_data_begin_nowait_mapper(int64_t device_id,
+EXTERN void __tgt_target_data_begin_nowait_mapper(const ident_t *loc, int64_t
device_id,
int32_t arg_num, void **args_base, void **args, int64_t *arg_sizes,
----------------
I should probably remove `const` here so we can pass `loc` to
`__kmpc_omp_taskwait`.
Repository:
rG LLVM Github Monorepo
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D87946/new/
https://reviews.llvm.org/D87946
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
