arsenm created this revision.
arsenm added reviewers: JonChesterfield, yaxunl, jhuber6.
Herald added a project: All.
arsenm requested review of this revision.
Herald added a subscriber: wdng.

Pretty sure these lround->round cases are just wrong


https://reviews.llvm.org/D154000

Files:
  clang/lib/Headers/__clang_hip_math.h
  clang/test/Headers/__clang_hip_math.hip

Index: clang/test/Headers/__clang_hip_math.hip
===================================================================
--- clang/test/Headers/__clang_hip_math.hip
+++ clang/test/Headers/__clang_hip_math.hip
@@ -1515,14 +1515,14 @@
 
 // DEFAULT-LABEL: @test_llroundf(
 // DEFAULT-NEXT:  entry:
-// DEFAULT-NEXT:    [[CALL_I:%.*]] = tail call contract float @__ocml_round_f32(float noundef [[X:%.*]]) #[[ATTR14]]
-// DEFAULT-NEXT:    [[CONV_I:%.*]] = fptosi float [[CALL_I]] to i64
+// DEFAULT-NEXT:    [[TMP0:%.*]] = tail call contract float @llvm.round.f32(float [[X:%.*]])
+// DEFAULT-NEXT:    [[CONV_I:%.*]] = fptosi float [[TMP0]] to i64
 // DEFAULT-NEXT:    ret i64 [[CONV_I]]
 //
 // FINITEONLY-LABEL: @test_llroundf(
 // FINITEONLY-NEXT:  entry:
-// FINITEONLY-NEXT:    [[CALL_I:%.*]] = tail call nnan ninf contract nofpclass(nan inf) float @__ocml_round_f32(float noundef nofpclass(nan inf) [[X:%.*]]) #[[ATTR14]]
-// FINITEONLY-NEXT:    [[CONV_I:%.*]] = fptosi float [[CALL_I]] to i64
+// FINITEONLY-NEXT:    [[TMP0:%.*]] = tail call nnan ninf contract float @llvm.round.f32(float [[X:%.*]])
+// FINITEONLY-NEXT:    [[CONV_I:%.*]] = fptosi float [[TMP0]] to i64
 // FINITEONLY-NEXT:    ret i64 [[CONV_I]]
 //
 extern "C" __device__ long long int test_llroundf(float x) {
@@ -1531,14 +1531,14 @@
 
 // DEFAULT-LABEL: @test_llround(
 // DEFAULT-NEXT:  entry:
-// DEFAULT-NEXT:    [[CALL_I:%.*]] = tail call contract double @__ocml_round_f64(double noundef [[X:%.*]]) #[[ATTR14]]
-// DEFAULT-NEXT:    [[CONV_I:%.*]] = fptosi double [[CALL_I]] to i64
+// DEFAULT-NEXT:    [[TMP0:%.*]] = tail call contract double @llvm.round.f64(double [[X:%.*]])
+// DEFAULT-NEXT:    [[CONV_I:%.*]] = fptosi double [[TMP0]] to i64
 // DEFAULT-NEXT:    ret i64 [[CONV_I]]
 //
 // FINITEONLY-LABEL: @test_llround(
 // FINITEONLY-NEXT:  entry:
-// FINITEONLY-NEXT:    [[CALL_I:%.*]] = tail call nnan ninf contract nofpclass(nan inf) double @__ocml_round_f64(double noundef nofpclass(nan inf) [[X:%.*]]) #[[ATTR14]]
-// FINITEONLY-NEXT:    [[CONV_I:%.*]] = fptosi double [[CALL_I]] to i64
+// FINITEONLY-NEXT:    [[TMP0:%.*]] = tail call nnan ninf contract double @llvm.round.f64(double [[X:%.*]])
+// FINITEONLY-NEXT:    [[CONV_I:%.*]] = fptosi double [[TMP0]] to i64
 // FINITEONLY-NEXT:    ret i64 [[CONV_I]]
 //
 extern "C" __device__ long long int test_llround(double x) {
@@ -1691,14 +1691,14 @@
 
 // DEFAULT-LABEL: @test_lroundf(
 // DEFAULT-NEXT:  entry:
-// DEFAULT-NEXT:    [[CALL_I:%.*]] = tail call contract float @__ocml_round_f32(float noundef [[X:%.*]]) #[[ATTR14]]
-// DEFAULT-NEXT:    [[CONV_I:%.*]] = fptosi float [[CALL_I]] to i64
+// DEFAULT-NEXT:    [[TMP0:%.*]] = tail call contract float @llvm.round.f32(float [[X:%.*]])
+// DEFAULT-NEXT:    [[CONV_I:%.*]] = fptosi float [[TMP0]] to i64
 // DEFAULT-NEXT:    ret i64 [[CONV_I]]
 //
 // FINITEONLY-LABEL: @test_lroundf(
 // FINITEONLY-NEXT:  entry:
-// FINITEONLY-NEXT:    [[CALL_I:%.*]] = tail call nnan ninf contract nofpclass(nan inf) float @__ocml_round_f32(float noundef nofpclass(nan inf) [[X:%.*]]) #[[ATTR14]]
-// FINITEONLY-NEXT:    [[CONV_I:%.*]] = fptosi float [[CALL_I]] to i64
+// FINITEONLY-NEXT:    [[TMP0:%.*]] = tail call nnan ninf contract float @llvm.round.f32(float [[X:%.*]])
+// FINITEONLY-NEXT:    [[CONV_I:%.*]] = fptosi float [[TMP0]] to i64
 // FINITEONLY-NEXT:    ret i64 [[CONV_I]]
 //
 extern "C" __device__ long int test_lroundf(float x) {
@@ -1707,14 +1707,14 @@
 
 // DEFAULT-LABEL: @test_lround(
 // DEFAULT-NEXT:  entry:
-// DEFAULT-NEXT:    [[CALL_I:%.*]] = tail call contract double @__ocml_round_f64(double noundef [[X:%.*]]) #[[ATTR14]]
-// DEFAULT-NEXT:    [[CONV_I:%.*]] = fptosi double [[CALL_I]] to i64
+// DEFAULT-NEXT:    [[TMP0:%.*]] = tail call contract double @llvm.round.f64(double [[X:%.*]])
+// DEFAULT-NEXT:    [[CONV_I:%.*]] = fptosi double [[TMP0]] to i64
 // DEFAULT-NEXT:    ret i64 [[CONV_I]]
 //
 // FINITEONLY-LABEL: @test_lround(
 // FINITEONLY-NEXT:  entry:
-// FINITEONLY-NEXT:    [[CALL_I:%.*]] = tail call nnan ninf contract nofpclass(nan inf) double @__ocml_round_f64(double noundef nofpclass(nan inf) [[X:%.*]]) #[[ATTR14]]
-// FINITEONLY-NEXT:    [[CONV_I:%.*]] = fptosi double [[CALL_I]] to i64
+// FINITEONLY-NEXT:    [[TMP0:%.*]] = tail call nnan ninf contract double @llvm.round.f64(double [[X:%.*]])
+// FINITEONLY-NEXT:    [[CONV_I:%.*]] = fptosi double [[TMP0]] to i64
 // FINITEONLY-NEXT:    ret i64 [[CONV_I]]
 //
 extern "C" __device__ long int test_lround(double x) {
@@ -2616,13 +2616,13 @@
 
 // DEFAULT-LABEL: @test_roundf(
 // DEFAULT-NEXT:  entry:
-// DEFAULT-NEXT:    [[CALL_I:%.*]] = tail call contract float @__ocml_round_f32(float noundef [[X:%.*]]) #[[ATTR14]]
-// DEFAULT-NEXT:    ret float [[CALL_I]]
+// DEFAULT-NEXT:    [[TMP0:%.*]] = tail call contract float @llvm.round.f32(float [[X:%.*]])
+// DEFAULT-NEXT:    ret float [[TMP0]]
 //
 // FINITEONLY-LABEL: @test_roundf(
 // FINITEONLY-NEXT:  entry:
-// FINITEONLY-NEXT:    [[CALL_I:%.*]] = tail call nnan ninf contract nofpclass(nan inf) float @__ocml_round_f32(float noundef nofpclass(nan inf) [[X:%.*]]) #[[ATTR14]]
-// FINITEONLY-NEXT:    ret float [[CALL_I]]
+// FINITEONLY-NEXT:    [[TMP0:%.*]] = tail call nnan ninf contract float @llvm.round.f32(float [[X:%.*]])
+// FINITEONLY-NEXT:    ret float [[TMP0]]
 //
 extern "C" __device__ float test_roundf(float x) {
   return roundf(x);
@@ -2630,13 +2630,13 @@
 
 // DEFAULT-LABEL: @test_round(
 // DEFAULT-NEXT:  entry:
-// DEFAULT-NEXT:    [[CALL_I:%.*]] = tail call contract double @__ocml_round_f64(double noundef [[X:%.*]]) #[[ATTR14]]
-// DEFAULT-NEXT:    ret double [[CALL_I]]
+// DEFAULT-NEXT:    [[TMP0:%.*]] = tail call contract double @llvm.round.f64(double [[X:%.*]])
+// DEFAULT-NEXT:    ret double [[TMP0]]
 //
 // FINITEONLY-LABEL: @test_round(
 // FINITEONLY-NEXT:  entry:
-// FINITEONLY-NEXT:    [[CALL_I:%.*]] = tail call nnan ninf contract nofpclass(nan inf) double @__ocml_round_f64(double noundef nofpclass(nan inf) [[X:%.*]]) #[[ATTR14]]
-// FINITEONLY-NEXT:    ret double [[CALL_I]]
+// FINITEONLY-NEXT:    [[TMP0:%.*]] = tail call nnan ninf contract double @llvm.round.f64(double [[X:%.*]])
+// FINITEONLY-NEXT:    ret double [[TMP0]]
 //
 extern "C" __device__ double test_round(double x) {
   return round(x);
Index: clang/lib/Headers/__clang_hip_math.h
===================================================================
--- clang/lib/Headers/__clang_hip_math.h
+++ clang/lib/Headers/__clang_hip_math.h
@@ -312,7 +312,7 @@
 long long int llrintf(float __x) { return __ocml_rint_f32(__x); }
 
 __DEVICE__
-long long int llroundf(float __x) { return __ocml_round_f32(__x); }
+long long int llroundf(float __x) { return __builtin_roundf(__x); }
 
 __DEVICE__
 float log10f(float __x) { return __ocml_log10_f32(__x); }
@@ -333,7 +333,7 @@
 long int lrintf(float __x) { return __ocml_rint_f32(__x); }
 
 __DEVICE__
-long int lroundf(float __x) { return __ocml_round_f32(__x); }
+long int lroundf(float __x) { return __builtin_roundf(__x); }
 
 __DEVICE__
 float modff(float __x, float *__iptr) {
@@ -460,7 +460,7 @@
 }
 
 __DEVICE__
-float roundf(float __x) { return __ocml_round_f32(__x); }
+float roundf(float __x) { return __builtin_roundf(__x); }
 
 __DEVICE__
 float rsqrtf(float __x) { return __ocml_rsqrt_f32(__x); }
@@ -860,7 +860,7 @@
 long long int llrint(double __x) { return __ocml_rint_f64(__x); }
 
 __DEVICE__
-long long int llround(double __x) { return __ocml_round_f64(__x); }
+long long int llround(double __x) { return __builtin_round(__x); }
 
 __DEVICE__
 double log(double __x) { return __ocml_log_f64(__x); }
@@ -881,7 +881,7 @@
 long int lrint(double __x) { return __ocml_rint_f64(__x); }
 
 __DEVICE__
-long int lround(double __x) { return __ocml_round_f64(__x); }
+long int lround(double __x) { return __builtin_round(__x); }
 
 __DEVICE__
 double modf(double __x, double *__iptr) {
@@ -1016,7 +1016,7 @@
 }
 
 __DEVICE__
-double round(double __x) { return __ocml_round_f64(__x); }
+double round(double __x) { return __builtin_round(__x); }
 
 __DEVICE__
 double rsqrt(double __x) { return __ocml_rsqrt_f64(__x); }
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
  • [PATCH] D154000: HIP: Direc... Matt Arsenault via Phabricator via cfe-commits

Reply via email to