commit 9a546dd9f4e89e4e31a29fadf36714d2cc329ee3
author Shilei Tian <tianshilei1992@gmail.com> Mon Jul 26 22:45:52 2021 -0400
committer Copybara-Service <copybara-worker@google.com> Thu Jul 29 19:31:15 2021 -0700
tree 7ccffc298137063d948b1ca3e7dfbd6950ca57b3
parent 44bea3be52422c54def538d59ad2781b7c3aae79

[AbstractAttributor] Fold __kmpc_parallel_level if possible

Similar to D105787, this patch tries to fold `__kmpc_parallel_level` if possible.
Note that `__kmpc_parallel_level` doesn't take activeness into consideration,
based on current `deviceRTLs`, its return value can be such as 0, 1, 2, instead
of 0, 129, 130, etc. that also indicate activeness.

Reviewed By: jdoerfert

Differential Revision: https://reviews.llvm.org/D106154

GitOrigin-RevId: e97e0a4fad091474131ad84d6f46009bf84c5b60

libomptarget/deviceRTLs/common/src/parallel.cu

diff --git a/libomptarget/deviceRTLs/common/src/parallel.cu b/libomptarget/deviceRTLs/common/src/parallel.cu
index 2656f3e..b12d5cc 100644
--- a/libomptarget/deviceRTLs/common/src/parallel.cu
+++ b/libomptarget/deviceRTLs/common/src/parallel.cu
@@ -239,7 +239,7 @@
   currTaskDescr->RestoreLoopData();
 }
 
-EXTERN uint8_t __kmpc_parallel_level() {
+NOINLINE EXTERN uint8_t __kmpc_parallel_level() {
   return parallelLevel[GetWarpId()] & (OMP_ACTIVE_PARALLEL_LEVEL - 1);
 }
 
@@ -282,11 +282,11 @@
 // parallel interface
 ////////////////////////////////////////////////////////////////////////////////
 
-EXTERN void __kmpc_parallel_51(kmp_Ident *ident, kmp_int32 global_tid,
-                               kmp_int32 if_expr, kmp_int32 num_threads,
-                               int proc_bind, void *fn, void *wrapper_fn,
-                               void **args, size_t nargs) {
-
+NOINLINE EXTERN void __kmpc_parallel_51(kmp_Ident *ident, kmp_int32 global_tid,
+                                        kmp_int32 if_expr,
+                                        kmp_int32 num_threads, int proc_bind,
+                                        void *fn, void *wrapper_fn, void **args,
+                                        size_t nargs) {
   // Handle the serialized case first, same for SPMD/non-SPMD except that in
   // SPMD mode we already incremented the parallel level counter, account for
   // that.

libomptarget/deviceRTLs/interface.h

diff --git a/libomptarget/deviceRTLs/interface.h b/libomptarget/deviceRTLs/interface.h
index ade1bfe..2e80dc3 100644
--- a/libomptarget/deviceRTLs/interface.h
+++ b/libomptarget/deviceRTLs/interface.h
@@ -222,7 +222,7 @@
                                     int32_t num_threads);
 EXTERN void __kmpc_serialized_parallel(kmp_Ident *loc, uint32_t global_tid);
 EXTERN void __kmpc_end_serialized_parallel(kmp_Ident *loc, uint32_t global_tid);
-EXTERN uint8_t __kmpc_parallel_level();
+NOINLINE EXTERN uint8_t __kmpc_parallel_level();
 
 // proc bind
 EXTERN void __kmpc_push_proc_bind(kmp_Ident *loc, uint32_t global_tid,
@@ -441,10 +441,11 @@
 /// \param wrapper_fn  The worker wrapper function of fn.
 /// \param args        The pointer array of arguments to fn.
 /// \param nargs       The number of arguments to fn.
-EXTERN void __kmpc_parallel_51(ident_t *ident, kmp_int32 global_tid,
-                               kmp_int32 if_expr, kmp_int32 num_threads,
-                               int proc_bind, void *fn, void *wrapper_fn,
-                               void **args, size_t nargs);
+NOINLINE EXTERN void __kmpc_parallel_51(ident_t *ident, kmp_int32 global_tid,
+                                        kmp_int32 if_expr,
+                                        kmp_int32 num_threads, int proc_bind,
+                                        void *fn, void *wrapper_fn, void **args,
+                                        size_t nargs);
 
 // SPMD execution mode interrogation function.
 EXTERN int8_t __kmpc_is_spmd_exec_mode();