[PATCH] D141307: Add -f[no-]loop-versioning option

2023-04-18 Thread Mats Petersson via Phabricator via cfe-commits
This revision was landed with ongoing or failed builds.
This revision was automatically updated to reflect the committed changes.
Closed by commit rGefae695d52cc: Add -f[no-]loop-versioning option (authored by 
MatsPetersson).

Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D141307/new/

https://reviews.llvm.org/D141307

Files:
  clang/include/clang/Driver/Options.td
  clang/lib/Driver/ToolChains/Flang.cpp
  clang/lib/Driver/ToolChains/Flang.h
  flang/include/flang/Frontend/CodeGenOptions.def
  flang/include/flang/Tools/CLOptions.inc
  flang/lib/Frontend/CompilerInvocation.cpp
  flang/lib/Frontend/FrontendActions.cpp
  flang/test/Driver/driver-help-hidden.f90
  flang/test/Driver/driver-help.f90
  flang/test/Driver/frontend-forwarding.f90
  flang/test/Driver/version-loops.f90

Index: flang/test/Driver/version-loops.f90
===
--- /dev/null
+++ flang/test/Driver/version-loops.f90
@@ -0,0 +1,54 @@
+! Test that flang-new forwards the -f{no-,}version-loops-for-stride 
+! options correctly to flang-new -fc1 for different variants of optimisation
+! and explicit flags.
+
+! RUN: %flang -### %s -o %t 2>&1   -O3 \
+! RUN:   | FileCheck %s
+  
+! RUN: %flang -### %s -o %t 2>&1 -O2 \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O2
+
+! RUN: %flang -### %s -o %t 2>&1  -O2 -fversion-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O2-with
+  
+! RUN: %flang -### %s -o %t 2>&1  -O4 \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O4
+  
+! RUN: %flang -### %s -o %t 2>&1  -Ofast \
+! RUN:   | FileCheck %s --check-prefix=CHECK-Ofast
+  
+! RUN: %flang -### %s -o %t 2>&1 -Ofast -fno-version-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-Ofast-no
+
+! RUN: %flang -### %s -o %t 2>&1 -O3 -fno-version-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O3-no
+
+! CHECK: "{{.*}}flang-new" "-fc1"
+! CHECK-SAME: "-fversion-loops-for-stride"
+! CHECK-SAME: "-O3"
+
+! CHECK-O2: "{{.*}}flang-new" "-fc1"
+! CHECK-O2-NOT: "-fversion-loops-for-stride"
+! CHECK-O2-SAME: "-O2"  
+
+! CHECK-O2-with: "{{.*}}flang-new" "-fc1"
+! CHECK-O2-with-SAME: "-fversion-loops-for-stride"
+! CHECK-O2-with-SAME: "-O2"  
+  
+! CHECK-O4: "{{.*}}flang-new" "-fc1"
+! CHECK-O4-SAME: "-fversion-loops-for-stride"
+! CHECK-O4-SAME: "-O3"
+
+! CHECK-Ofast: "{{.*}}flang-new" "-fc1"
+! CHECK-Ofast-SAME: "-ffast-math"
+! CHECK-Ofast-SAME: "-fversion-loops-for-stride"
+! CHECK-Ofast-SAME: "-O3"
+
+! CHECK-Ofast-no: "{{.*}}flang-new" "-fc1"
+! CHECK-Ofast-no-SAME: "-ffast-math"
+! CHECK-Ofast-no-NOT: "-fversion-loops-for-stride"
+! CHECK-Ofast-no-SAME: "-O3"
+
+! CHECK-O3-no: "{{.*}}flang-new" "-fc1"
+! CHECK-O3-no-NOT: "-fversion-loops-for-stride"
+! CHECK-O3-no-SAME: "-O3"
Index: flang/test/Driver/frontend-forwarding.f90
===
--- flang/test/Driver/frontend-forwarding.f90
+++ flang/test/Driver/frontend-forwarding.f90
@@ -15,6 +15,8 @@
 ! RUN: -fassociative-math \
 ! RUN: -freciprocal-math \
 ! RUN: -fpass-plugin=Bye%pluginext \
+! RUN: -fversion-loops-for-stride \
+! RUN: -mllvm -print-before-all\
 ! RUN: -mllvm -print-before-all \
 ! RUN: -save-temps=obj \
 ! RUN: -P \
@@ -34,5 +36,6 @@
 ! CHECK: "-freciprocal-math"
 ! CHECK: "-fconvert=little-endian"
 ! CHECK: "-fpass-plugin=Bye
+! CHECK: "-fversion-loops-for-stride"  
 ! CHECK: "-mllvm" "-print-before-all"
 ! CHECK: "-save-temps=obj"
Index: flang/test/Driver/driver-help.f90
===
--- flang/test/Driver/driver-help.f90
+++ flang/test/Driver/driver-help.f90
@@ -48,6 +48,8 @@
 ! HELP-NEXT: -fno-integrated-as  Disable the integrated assembler
 ! HELP-NEXT: -fno-signed-zeros  Allow optimizations that ignore the sign of floating point zeros
 ! HELP-NEXT: -fno-stack-arrays  Allocate array temporaries on the heap (default)
+! HELP-NEXT: -fno-version-loops-for-stride
+! HELP-NEXT:Do not create unit-strided loops (default)
 ! HELP-NEXT: -fopenacc  Enable OpenACC
 ! HELP-NEXT: -fopenmp   Parse OpenMP pragmas and generate parallel code.
 ! HELP-NEXT: -fpass-plugin= Load pass plugin from a dynamic shared object file (only with new pass manager).
@@ -55,6 +57,8 @@
 ! HELP-NEXT: -fstack-arrays Attempt to allocate array temporaries on the stack, no matter their size
 ! HELP-NEXT: -fsyntax-only  Run the preprocessor, parser and semantic analysis stages
 ! HELP-NEXT: -funderscoring Appends one trailing underscore to external names
+! HELP-NEXT: -fversion-loops-for-stride
+! HELP-NEXT:Create unit-strided versions of loops
 ! HELP-NEXT: -fxor-operator Enable .XOR. as a synonym of .NEQV.
 ! HELP-NEXT: -gline-tables-only Emit debug line number tables only
 ! HELP-NEXT: -g Generate 

[PATCH] D141307: Add -f[no-]loop-versioning option

2023-04-18 Thread Mats Petersson via Phabricator via cfe-commits
Leporacanthicus updated this revision to Diff 514525.
Leporacanthicus added a comment.

Rebase only


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D141307/new/

https://reviews.llvm.org/D141307

Files:
  clang/include/clang/Driver/Options.td
  clang/lib/Driver/ToolChains/Flang.cpp
  clang/lib/Driver/ToolChains/Flang.h
  flang/include/flang/Frontend/CodeGenOptions.def
  flang/include/flang/Tools/CLOptions.inc
  flang/lib/Frontend/CompilerInvocation.cpp
  flang/lib/Frontend/FrontendActions.cpp
  flang/test/Driver/driver-help-hidden.f90
  flang/test/Driver/driver-help.f90
  flang/test/Driver/frontend-forwarding.f90
  flang/test/Driver/version-loops.f90

Index: flang/test/Driver/version-loops.f90
===
--- /dev/null
+++ flang/test/Driver/version-loops.f90
@@ -0,0 +1,54 @@
+! Test that flang-new forwards the -f{no-,}version-loops-for-stride 
+! options correctly to flang-new -fc1 for different variants of optimisation
+! and explicit flags.
+
+! RUN: %flang -### %s -o %t 2>&1   -O3 \
+! RUN:   | FileCheck %s
+  
+! RUN: %flang -### %s -o %t 2>&1 -O2 \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O2
+
+! RUN: %flang -### %s -o %t 2>&1  -O2 -fversion-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O2-with
+  
+! RUN: %flang -### %s -o %t 2>&1  -O4 \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O4
+  
+! RUN: %flang -### %s -o %t 2>&1  -Ofast \
+! RUN:   | FileCheck %s --check-prefix=CHECK-Ofast
+  
+! RUN: %flang -### %s -o %t 2>&1 -Ofast -fno-version-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-Ofast-no
+
+! RUN: %flang -### %s -o %t 2>&1 -O3 -fno-version-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O3-no
+
+! CHECK: "{{.*}}flang-new" "-fc1"
+! CHECK-SAME: "-fversion-loops-for-stride"
+! CHECK-SAME: "-O3"
+
+! CHECK-O2: "{{.*}}flang-new" "-fc1"
+! CHECK-O2-NOT: "-fversion-loops-for-stride"
+! CHECK-O2-SAME: "-O2"  
+
+! CHECK-O2-with: "{{.*}}flang-new" "-fc1"
+! CHECK-O2-with-SAME: "-fversion-loops-for-stride"
+! CHECK-O2-with-SAME: "-O2"  
+  
+! CHECK-O4: "{{.*}}flang-new" "-fc1"
+! CHECK-O4-SAME: "-fversion-loops-for-stride"
+! CHECK-O4-SAME: "-O3"
+
+! CHECK-Ofast: "{{.*}}flang-new" "-fc1"
+! CHECK-Ofast-SAME: "-ffast-math"
+! CHECK-Ofast-SAME: "-fversion-loops-for-stride"
+! CHECK-Ofast-SAME: "-O3"
+
+! CHECK-Ofast-no: "{{.*}}flang-new" "-fc1"
+! CHECK-Ofast-no-SAME: "-ffast-math"
+! CHECK-Ofast-no-NOT: "-fversion-loops-for-stride"
+! CHECK-Ofast-no-SAME: "-O3"
+
+! CHECK-O3-no: "{{.*}}flang-new" "-fc1"
+! CHECK-O3-no-NOT: "-fversion-loops-for-stride"
+! CHECK-O3-no-SAME: "-O3"
Index: flang/test/Driver/frontend-forwarding.f90
===
--- flang/test/Driver/frontend-forwarding.f90
+++ flang/test/Driver/frontend-forwarding.f90
@@ -15,6 +15,8 @@
 ! RUN: -fassociative-math \
 ! RUN: -freciprocal-math \
 ! RUN: -fpass-plugin=Bye%pluginext \
+! RUN: -fversion-loops-for-stride \
+! RUN: -mllvm -print-before-all\
 ! RUN: -mllvm -print-before-all \
 ! RUN: -save-temps=obj \
 ! RUN: -P \
@@ -34,5 +36,6 @@
 ! CHECK: "-freciprocal-math"
 ! CHECK: "-fconvert=little-endian"
 ! CHECK: "-fpass-plugin=Bye
+! CHECK: "-fversion-loops-for-stride"  
 ! CHECK: "-mllvm" "-print-before-all"
 ! CHECK: "-save-temps=obj"
Index: flang/test/Driver/driver-help.f90
===
--- flang/test/Driver/driver-help.f90
+++ flang/test/Driver/driver-help.f90
@@ -48,6 +48,8 @@
 ! HELP-NEXT: -fno-integrated-as  Disable the integrated assembler
 ! HELP-NEXT: -fno-signed-zeros  Allow optimizations that ignore the sign of floating point zeros
 ! HELP-NEXT: -fno-stack-arrays  Allocate array temporaries on the heap (default)
+! HELP-NEXT: -fno-version-loops-for-stride
+! HELP-NEXT:Do not create unit-strided loops (default)
 ! HELP-NEXT: -fopenacc  Enable OpenACC
 ! HELP-NEXT: -fopenmp   Parse OpenMP pragmas and generate parallel code.
 ! HELP-NEXT: -fpass-plugin= Load pass plugin from a dynamic shared object file (only with new pass manager).
@@ -55,6 +57,8 @@
 ! HELP-NEXT: -fstack-arrays Attempt to allocate array temporaries on the stack, no matter their size
 ! HELP-NEXT: -fsyntax-only  Run the preprocessor, parser and semantic analysis stages
 ! HELP-NEXT: -funderscoring Appends one trailing underscore to external names
+! HELP-NEXT: -fversion-loops-for-stride
+! HELP-NEXT:Create unit-strided versions of loops
 ! HELP-NEXT: -fxor-operator Enable .XOR. as a synonym of .NEQV.
 ! HELP-NEXT: -gline-tables-only Emit debug line number tables only
 ! HELP-NEXT: -g Generate source-level debug information
@@ -146,6 +150,8 @@
 ! HELP-FC1-NEXT: -fno-reformat  Dump the cooked character stream in -E mode
 

[PATCH] D141307: Add -f[no-]loop-versioning option

2023-04-17 Thread Mats Petersson via Phabricator via cfe-commits
Leporacanthicus updated this revision to Diff 514403.
Leporacanthicus added a comment.

Rebase only


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D141307/new/

https://reviews.llvm.org/D141307

Files:
  clang/include/clang/Driver/Options.td
  clang/lib/Driver/ToolChains/Flang.cpp
  clang/lib/Driver/ToolChains/Flang.h
  flang/include/flang/Frontend/CodeGenOptions.def
  flang/include/flang/Tools/CLOptions.inc
  flang/lib/Frontend/CompilerInvocation.cpp
  flang/lib/Frontend/FrontendActions.cpp
  flang/test/Driver/driver-help-hidden.f90
  flang/test/Driver/driver-help.f90
  flang/test/Driver/frontend-forwarding.f90
  flang/test/Driver/version-loops.f90

Index: flang/test/Driver/version-loops.f90
===
--- /dev/null
+++ flang/test/Driver/version-loops.f90
@@ -0,0 +1,54 @@
+! Test that flang-new forwards the -f{no-,}version-loops-for-stride 
+! options correctly to flang-new -fc1 for different variants of optimisation
+! and explicit flags.
+
+! RUN: %flang -### %s -o %t 2>&1   -O3 \
+! RUN:   | FileCheck %s
+  
+! RUN: %flang -### %s -o %t 2>&1 -O2 \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O2
+
+! RUN: %flang -### %s -o %t 2>&1  -O2 -fversion-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O2-with
+  
+! RUN: %flang -### %s -o %t 2>&1  -O4 \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O4
+  
+! RUN: %flang -### %s -o %t 2>&1  -Ofast \
+! RUN:   | FileCheck %s --check-prefix=CHECK-Ofast
+  
+! RUN: %flang -### %s -o %t 2>&1 -Ofast -fno-version-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-Ofast-no
+
+! RUN: %flang -### %s -o %t 2>&1 -O3 -fno-version-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O3-no
+
+! CHECK: "{{.*}}flang-new" "-fc1"
+! CHECK-SAME: "-fversion-loops-for-stride"
+! CHECK-SAME: "-O3"
+
+! CHECK-O2: "{{.*}}flang-new" "-fc1"
+! CHECK-O2-NOT: "-fversion-loops-for-stride"
+! CHECK-O2-SAME: "-O2"  
+
+! CHECK-O2-with: "{{.*}}flang-new" "-fc1"
+! CHECK-O2-with-SAME: "-fversion-loops-for-stride"
+! CHECK-O2-with-SAME: "-O2"  
+  
+! CHECK-O4: "{{.*}}flang-new" "-fc1"
+! CHECK-O4-SAME: "-fversion-loops-for-stride"
+! CHECK-O4-SAME: "-O3"
+
+! CHECK-Ofast: "{{.*}}flang-new" "-fc1"
+! CHECK-Ofast-SAME: "-ffast-math"
+! CHECK-Ofast-SAME: "-fversion-loops-for-stride"
+! CHECK-Ofast-SAME: "-O3"
+
+! CHECK-Ofast-no: "{{.*}}flang-new" "-fc1"
+! CHECK-Ofast-no-SAME: "-ffast-math"
+! CHECK-Ofast-no-NOT: "-fversion-loops-for-stride"
+! CHECK-Ofast-no-SAME: "-O3"
+
+! CHECK-O3-no: "{{.*}}flang-new" "-fc1"
+! CHECK-O3-no-NOT: "-fversion-loops-for-stride"
+! CHECK-O3-no-SAME: "-O3"
Index: flang/test/Driver/frontend-forwarding.f90
===
--- flang/test/Driver/frontend-forwarding.f90
+++ flang/test/Driver/frontend-forwarding.f90
@@ -15,6 +15,8 @@
 ! RUN: -fassociative-math \
 ! RUN: -freciprocal-math \
 ! RUN: -fpass-plugin=Bye%pluginext \
+! RUN: -fversion-loops-for-stride \
+! RUN: -mllvm -print-before-all\
 ! RUN: -mllvm -print-before-all \
 ! RUN: -save-temps=obj \
 ! RUN: -P \
@@ -34,5 +36,6 @@
 ! CHECK: "-freciprocal-math"
 ! CHECK: "-fconvert=little-endian"
 ! CHECK: "-fpass-plugin=Bye
+! CHECK: "-fversion-loops-for-stride"  
 ! CHECK: "-mllvm" "-print-before-all"
 ! CHECK: "-save-temps=obj"
Index: flang/test/Driver/driver-help.f90
===
--- flang/test/Driver/driver-help.f90
+++ flang/test/Driver/driver-help.f90
@@ -48,6 +48,8 @@
 ! HELP-NEXT: -fno-integrated-as  Disable the integrated assembler
 ! HELP-NEXT: -fno-signed-zeros  Allow optimizations that ignore the sign of floating point zeros
 ! HELP-NEXT: -fno-stack-arrays  Allocate array temporaries on the heap (default)
+! HELP-NEXT: -fno-version-loops-for-stride
+! HELP-NEXT:Do not create unit-strided loops (default)
 ! HELP-NEXT: -fopenacc  Enable OpenACC
 ! HELP-NEXT: -fopenmp   Parse OpenMP pragmas and generate parallel code.
 ! HELP-NEXT: -fpass-plugin= Load pass plugin from a dynamic shared object file (only with new pass manager).
@@ -55,6 +57,8 @@
 ! HELP-NEXT: -fstack-arrays Attempt to allocate array temporaries on the stack, no matter their size
 ! HELP-NEXT: -fsyntax-only  Run the preprocessor, parser and semantic analysis stages
 ! HELP-NEXT: -funderscoring Appends one trailing underscore to external names
+! HELP-NEXT: -fversion-loops-for-stride
+! HELP-NEXT:Create unit-strided versions of loops
 ! HELP-NEXT: -fxor-operator Enable .XOR. as a synonym of .NEQV.
 ! HELP-NEXT: -gline-tables-only Emit debug line number tables only
 ! HELP-NEXT: -g Generate source-level debug information
@@ -146,6 +150,8 @@
 ! HELP-FC1-NEXT: -fno-reformat  Dump the cooked character stream in -E mode
 

[PATCH] D141307: Add -f[no-]loop-versioning option

2023-04-14 Thread Mats Petersson via Phabricator via cfe-commits
Leporacanthicus updated this revision to Diff 513672.
Leporacanthicus added a comment.

Remove unnecessary braces in if-statement.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D141307/new/

https://reviews.llvm.org/D141307

Files:
  clang/include/clang/Driver/Options.td
  clang/lib/Driver/ToolChains/Flang.cpp
  clang/lib/Driver/ToolChains/Flang.h
  flang/include/flang/Frontend/CodeGenOptions.def
  flang/include/flang/Tools/CLOptions.inc
  flang/lib/Frontend/CompilerInvocation.cpp
  flang/lib/Frontend/FrontendActions.cpp
  flang/test/Driver/driver-help-hidden.f90
  flang/test/Driver/driver-help.f90
  flang/test/Driver/frontend-forwarding.f90
  flang/test/Driver/version-loops.f90

Index: flang/test/Driver/version-loops.f90
===
--- /dev/null
+++ flang/test/Driver/version-loops.f90
@@ -0,0 +1,54 @@
+! Test that flang-new forwards the -f{no-,}version-loops-for-stride 
+! options correctly to flang-new -fc1 for different variants of optimisation
+! and explicit flags.
+
+! RUN: %flang -### %s -o %t 2>&1   -O3 \
+! RUN:   | FileCheck %s
+  
+! RUN: %flang -### %s -o %t 2>&1 -O2 \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O2
+
+! RUN: %flang -### %s -o %t 2>&1  -O2 -fversion-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O2-with
+  
+! RUN: %flang -### %s -o %t 2>&1  -O4 \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O4
+  
+! RUN: %flang -### %s -o %t 2>&1  -Ofast \
+! RUN:   | FileCheck %s --check-prefix=CHECK-Ofast
+  
+! RUN: %flang -### %s -o %t 2>&1 -Ofast -fno-version-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-Ofast-no
+
+! RUN: %flang -### %s -o %t 2>&1 -O3 -fno-version-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O3-no
+
+! CHECK: "{{.*}}flang-new" "-fc1"
+! CHECK-SAME: "-fversion-loops-for-stride"
+! CHECK-SAME: "-O3"
+
+! CHECK-O2: "{{.*}}flang-new" "-fc1"
+! CHECK-O2-NOT: "-fversion-loops-for-stride"
+! CHECK-O2-SAME: "-O2"  
+
+! CHECK-O2-with: "{{.*}}flang-new" "-fc1"
+! CHECK-O2-with-SAME: "-fversion-loops-for-stride"
+! CHECK-O2-with-SAME: "-O2"  
+  
+! CHECK-O4: "{{.*}}flang-new" "-fc1"
+! CHECK-O4-SAME: "-fversion-loops-for-stride"
+! CHECK-O4-SAME: "-O3"
+
+! CHECK-Ofast: "{{.*}}flang-new" "-fc1"
+! CHECK-Ofast-SAME: "-ffast-math"
+! CHECK-Ofast-SAME: "-fversion-loops-for-stride"
+! CHECK-Ofast-SAME: "-O3"
+
+! CHECK-Ofast-no: "{{.*}}flang-new" "-fc1"
+! CHECK-Ofast-no-SAME: "-ffast-math"
+! CHECK-Ofast-no-NOT: "-fversion-loops-for-stride"
+! CHECK-Ofast-no-SAME: "-O3"
+
+! CHECK-O3-no: "{{.*}}flang-new" "-fc1"
+! CHECK-O3-no-NOT: "-fversion-loops-for-stride"
+! CHECK-O3-no-SAME: "-O3"
Index: flang/test/Driver/frontend-forwarding.f90
===
--- flang/test/Driver/frontend-forwarding.f90
+++ flang/test/Driver/frontend-forwarding.f90
@@ -15,6 +15,8 @@
 ! RUN: -fassociative-math \
 ! RUN: -freciprocal-math \
 ! RUN: -fpass-plugin=Bye%pluginext \
+! RUN: -fversion-loops-for-stride \
+! RUN: -mllvm -print-before-all\
 ! RUN: -mllvm -print-before-all \
 ! RUN: -save-temps=obj \
 ! RUN: -P \
@@ -34,5 +36,6 @@
 ! CHECK: "-freciprocal-math"
 ! CHECK: "-fconvert=little-endian"
 ! CHECK: "-fpass-plugin=Bye
+! CHECK: "-fversion-loops-for-stride"  
 ! CHECK: "-mllvm" "-print-before-all"
 ! CHECK: "-save-temps=obj"
Index: flang/test/Driver/driver-help.f90
===
--- flang/test/Driver/driver-help.f90
+++ flang/test/Driver/driver-help.f90
@@ -48,6 +48,8 @@
 ! HELP-NEXT: -fno-integrated-as  Disable the integrated assembler
 ! HELP-NEXT: -fno-signed-zeros  Allow optimizations that ignore the sign of floating point zeros
 ! HELP-NEXT: -fno-stack-arrays  Allocate array temporaries on the heap (default)
+! HELP-NEXT: -fno-version-loops-for-stride
+! HELP-NEXT:Do not create unit-strided loops (default)
 ! HELP-NEXT: -fopenacc  Enable OpenACC
 ! HELP-NEXT: -fopenmp   Parse OpenMP pragmas and generate parallel code.
 ! HELP-NEXT: -fpass-plugin= Load pass plugin from a dynamic shared object file (only with new pass manager).
@@ -55,6 +57,8 @@
 ! HELP-NEXT: -fstack-arrays Attempt to allocate array temporaries on the stack, no matter their size
 ! HELP-NEXT: -fsyntax-only  Run the preprocessor, parser and semantic analysis stages
 ! HELP-NEXT: -funderscoring Appends one trailing underscore to external names
+! HELP-NEXT: -fversion-loops-for-stride
+! HELP-NEXT:Create unit-strided versions of loops
 ! HELP-NEXT: -fxor-operator Enable .XOR. as a synonym of .NEQV.
 ! HELP-NEXT: -gline-tables-only Emit debug line number tables only
 ! HELP-NEXT: -g Generate source-level debug information
@@ -146,6 +150,8 @@
 ! HELP-FC1-NEXT: -fno-reformat  Dump the 

[PATCH] D141307: Add -f[no-]loop-versioning option

2023-04-12 Thread Andrzej Warzynski via Phabricator via cfe-commits
awarzynski accepted this revision.
awarzynski added a comment.

LGTM, thanks!


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D141307/new/

https://reviews.llvm.org/D141307

___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits


[PATCH] D141307: Add -f[no-]loop-versioning option

2023-04-12 Thread Mats Petersson via Phabricator via cfe-commits
Leporacanthicus updated this revision to Diff 512929.
Leporacanthicus marked 3 inline comments as done.
Leporacanthicus added a comment.

- Add new function to help clarify the purpose of the options being used.
- Add descriptive comment to make function use clearer.
- Update tests to make them a more explicit.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D141307/new/

https://reviews.llvm.org/D141307

Files:
  clang/include/clang/Driver/Options.td
  clang/lib/Driver/ToolChains/Flang.cpp
  clang/lib/Driver/ToolChains/Flang.h
  flang/include/flang/Frontend/CodeGenOptions.def
  flang/include/flang/Tools/CLOptions.inc
  flang/lib/Frontend/CompilerInvocation.cpp
  flang/lib/Frontend/FrontendActions.cpp
  flang/test/Driver/driver-help-hidden.f90
  flang/test/Driver/driver-help.f90
  flang/test/Driver/frontend-forwarding.f90
  flang/test/Driver/version-loops.f90

Index: flang/test/Driver/version-loops.f90
===
--- /dev/null
+++ flang/test/Driver/version-loops.f90
@@ -0,0 +1,54 @@
+! Test that flang-new forwards the -f{no-,}version-loops-for-stride 
+! options correctly to flang-new -fc1 for different variants of optimisation
+! and explicit flags.
+
+! RUN: %flang -### %s -o %t 2>&1   -O3 \
+! RUN:   | FileCheck %s
+  
+! RUN: %flang -### %s -o %t 2>&1 -O2 \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O2
+
+! RUN: %flang -### %s -o %t 2>&1  -O2 -fversion-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O2-with
+  
+! RUN: %flang -### %s -o %t 2>&1  -O4 \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O4
+  
+! RUN: %flang -### %s -o %t 2>&1  -Ofast \
+! RUN:   | FileCheck %s --check-prefix=CHECK-Ofast
+  
+! RUN: %flang -### %s -o %t 2>&1 -Ofast -fno-version-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-Ofast-no
+
+! RUN: %flang -### %s -o %t 2>&1 -O3 -fno-version-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O3-no
+
+! CHECK: "{{.*}}flang-new" "-fc1"
+! CHECK-SAME: "-fversion-loops-for-stride"
+! CHECK-SAME: "-O3"
+
+! CHECK-O2: "{{.*}}flang-new" "-fc1"
+! CHECK-O2-NOT: "-fversion-loops-for-stride"
+! CHECK-O2-SAME: "-O2"  
+
+! CHECK-O2-with: "{{.*}}flang-new" "-fc1"
+! CHECK-O2-with-SAME: "-fversion-loops-for-stride"
+! CHECK-O2-with-SAME: "-O2"  
+  
+! CHECK-O4: "{{.*}}flang-new" "-fc1"
+! CHECK-O4-SAME: "-fversion-loops-for-stride"
+! CHECK-O4-SAME: "-O3"
+
+! CHECK-Ofast: "{{.*}}flang-new" "-fc1"
+! CHECK-Ofast-SAME: "-ffast-math"
+! CHECK-Ofast-SAME: "-fversion-loops-for-stride"
+! CHECK-Ofast-SAME: "-O3"
+
+! CHECK-Ofast-no: "{{.*}}flang-new" "-fc1"
+! CHECK-Ofast-no-SAME: "-ffast-math"
+! CHECK-Ofast-no-NOT: "-fversion-loops-for-stride"
+! CHECK-Ofast-no-SAME: "-O3"
+
+! CHECK-O3-no: "{{.*}}flang-new" "-fc1"
+! CHECK-O3-no-NOT: "-fversion-loops-for-stride"
+! CHECK-O3-no-SAME: "-O3"
Index: flang/test/Driver/frontend-forwarding.f90
===
--- flang/test/Driver/frontend-forwarding.f90
+++ flang/test/Driver/frontend-forwarding.f90
@@ -15,6 +15,8 @@
 ! RUN: -fassociative-math \
 ! RUN: -freciprocal-math \
 ! RUN: -fpass-plugin=Bye%pluginext \
+! RUN: -fversion-loops-for-stride \
+! RUN: -mllvm -print-before-all\
 ! RUN: -mllvm -print-before-all \
 ! RUN: -save-temps=obj \
 ! RUN: -P \
@@ -34,5 +36,6 @@
 ! CHECK: "-freciprocal-math"
 ! CHECK: "-fconvert=little-endian"
 ! CHECK: "-fpass-plugin=Bye
+! CHECK: "-fversion-loops-for-stride"  
 ! CHECK: "-mllvm" "-print-before-all"
 ! CHECK: "-save-temps=obj"
Index: flang/test/Driver/driver-help.f90
===
--- flang/test/Driver/driver-help.f90
+++ flang/test/Driver/driver-help.f90
@@ -48,6 +48,8 @@
 ! HELP-NEXT: -fno-integrated-as  Disable the integrated assembler
 ! HELP-NEXT: -fno-signed-zeros  Allow optimizations that ignore the sign of floating point zeros
 ! HELP-NEXT: -fno-stack-arrays  Allocate array temporaries on the heap (default)
+! HELP-NEXT: -fno-version-loops-for-stride
+! HELP-NEXT:Do not create unit-strided loops (default)
 ! HELP-NEXT: -fopenacc  Enable OpenACC
 ! HELP-NEXT: -fopenmp   Parse OpenMP pragmas and generate parallel code.
 ! HELP-NEXT: -fpass-plugin= Load pass plugin from a dynamic shared object file (only with new pass manager).
@@ -55,6 +57,8 @@
 ! HELP-NEXT: -fstack-arrays Attempt to allocate array temporaries on the stack, no matter their size
 ! HELP-NEXT: -fsyntax-only  Run the preprocessor, parser and semantic analysis stages
 ! HELP-NEXT: -funderscoring Appends one trailing underscore to external names
+! HELP-NEXT: -fversion-loops-for-stride
+! HELP-NEXT:Create unit-strided versions of loops
 ! HELP-NEXT: -fxor-operator Enable .XOR. as a synonym of .NEQV.
 ! HELP-NEXT: -gline-tables-only Emit 

[PATCH] D141307: Add -f[no-]loop-versioning option

2023-04-11 Thread Mats Petersson via Phabricator via cfe-commits
Leporacanthicus updated this revision to Diff 512532.
Leporacanthicus added a comment.

Rebase only.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D141307/new/

https://reviews.llvm.org/D141307

Files:
  clang/include/clang/Driver/Options.td
  clang/lib/Driver/ToolChains/Flang.cpp
  flang/include/flang/Frontend/CodeGenOptions.def
  flang/include/flang/Tools/CLOptions.inc
  flang/lib/Frontend/CompilerInvocation.cpp
  flang/lib/Frontend/FrontendActions.cpp
  flang/test/Driver/driver-help-hidden.f90
  flang/test/Driver/driver-help.f90
  flang/test/Driver/frontend-forwarding.f90
  flang/test/Driver/version-loops.f90

Index: flang/test/Driver/version-loops.f90
===
--- /dev/null
+++ flang/test/Driver/version-loops.f90
@@ -0,0 +1,54 @@
+! Test that flang-new forwards the -f{no-,}version-loops-for-stride 
+! options corredly to flang-new -fc1 for different variants of optimisation
+! and explicit flags.
+
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O3 \
+! RUN:   | FileCheck %s
+  
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O2 \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O2
+
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O2 -fversion-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O2-with
+  
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O4 \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O4
+  
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -Ofast \
+! RUN:   | FileCheck %s --check-prefix=CHECK-Ofast
+  
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -Ofast -fno-version-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-Ofast-no
+
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O3 -fno-version-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O3-no
+  
+! CHECK: "-fversion-loops-for-stride"
+! CHECK: "-O3"
+
+! CHECK-O2-NOT: "-fversion-loops-for-stride"
+! CHECK-O2: "-O2"  
+
+! CHECK-O2-with: "-fversion-loops-for-stride"
+! CHECK-O2-with: "-O2"  
+  
+! CHECK-O4: "-fversion-loops-for-stride"
+! CHECK-O4: "-O3"
+
+! CHECK-Ofast: "-ffast-math"
+! CHECK-Ofast: "-fversion-loops-for-stride"
+! CHECK-Ofast: "-O3"
+
+! CHECK-Ofast-no: "-ffast-math"
+! CHECK-Ofast-no-NOT: "-fversion-loops-for-stride"
+! CHECK-Ofast-no: "-O3"
+
+! CHECK-O3-no-NOT: "-fversion-loops-for-stride"
+! CHECK-O3-no: "-O3"
Index: flang/test/Driver/frontend-forwarding.f90
===
--- flang/test/Driver/frontend-forwarding.f90
+++ flang/test/Driver/frontend-forwarding.f90
@@ -15,6 +15,8 @@
 ! RUN: -fassociative-math \
 ! RUN: -freciprocal-math \
 ! RUN: -fpass-plugin=Bye%pluginext \
+! RUN: -fversion-loops-for-stride \
+! RUN: -mllvm -print-before-all\
 ! RUN: -mllvm -print-before-all \
 ! RUN: -save-temps=obj \
 ! RUN: -P \
@@ -34,5 +36,6 @@
 ! CHECK: "-freciprocal-math"
 ! CHECK: "-fconvert=little-endian"
 ! CHECK: "-fpass-plugin=Bye
+! CHECK: "-fversion-loops-for-stride"  
 ! CHECK: "-mllvm" "-print-before-all"
 ! CHECK: "-save-temps=obj"
Index: flang/test/Driver/driver-help.f90
===
--- flang/test/Driver/driver-help.f90
+++ flang/test/Driver/driver-help.f90
@@ -48,6 +48,8 @@
 ! HELP-NEXT: -fno-integrated-as  Disable the integrated assembler
 ! HELP-NEXT: -fno-signed-zeros  Allow optimizations that ignore the sign of floating point zeros
 ! HELP-NEXT: -fno-stack-arrays  Allocate array temporaries on the heap (default)
+! HELP-NEXT: -fno-version-loops-for-stride
+! HELP-NEXT:Do not create unit-strided loops (default)
 ! HELP-NEXT: -fopenacc  Enable OpenACC
 ! HELP-NEXT: -fopenmp   Parse OpenMP pragmas and generate parallel code.
 ! HELP-NEXT: -fpass-plugin= Load pass plugin from a dynamic shared object file (only with new pass manager).
@@ -55,6 +57,8 @@
 ! HELP-NEXT: -fstack-arrays Attempt to allocate array temporaries on the stack, no matter their size
 ! HELP-NEXT: -fsyntax-only  Run the preprocessor, parser and semantic analysis stages
 ! HELP-NEXT: -funderscoring Appends one trailing underscore to external names
+! HELP-NEXT: -fversion-loops-for-stride
+! HELP-NEXT:Create unit-strided versions of loops
 ! HELP-NEXT: -fxor-operator Enable .XOR. as a synonym of .NEQV.
 ! HELP-NEXT: -gline-tables-only Emit debug line number tables only
 ! HELP-NEXT: -g Generate source-level debug information
@@ -143,6 +147,8 @@
 ! HELP-FC1-NEXT: -fno-reformat  Dump the cooked character stream in -E mode
 ! HELP-FC1-NEXT: -fno-signed-zeros  Allow optimizations that ignore the sign of floating point zeros
 ! HELP-FC1-NEXT: -fno-stack-arrays  Allocate array temporaries on the heap (default)
+! 

[PATCH] D141307: Add -f[no-]loop-versioning option

2023-04-07 Thread Andrzej Warzynski via Phabricator via cfe-commits
awarzynski added a comment.

Mats, thanks for working on this! Just a few minor suggestions from me.




Comment at: clang/lib/Driver/ToolChains/Flang.cpp:55
 
+static bool shouldLoopVersion(const ArgList ) {
+  if (Arg *A = Args.getLastArg(options::OPT_Ofast, options::OPT_O,

Could you add a short Docstring, pls? In particular, is the logic implemented 
by this method consistent with GFortran? Are there any external docs that could 
be referred to here?



Comment at: clang/lib/Driver/ToolChains/Flang.cpp:56
+static bool shouldLoopVersion(const ArgList ) {
+  if (Arg *A = Args.getLastArg(options::OPT_Ofast, options::OPT_O,
+   options::OPT_O4, options::OPT_floop_versioning,

Please rewrite this to use [[ 
https://llvm.org/docs/CodingStandards.html#use-early-exits-and-continue-to-simplify-code
 | early exit ]], e.g.:
```
  Arg *LoopVersionOption = Args.getLastArg(options::OPT_Ofast, options::OPT_O,
   options::OPT_O4, options::OPT_floop_versioning,
   options::OPT_fno_loop_versioning)
  if !(LoopVersionOption)
return false;

  // The remaining logic HERE
```



Comment at: clang/lib/Driver/ToolChains/Flang.cpp:109-110
   addDebugInfoKind(CmdArgs, DebugInfoKind);
+  if (shouldLoopVersion(Args))
+CmdArgs.push_back("-fversion-loops-for-stride");
 }

Would you classify this as a code-gen option? Alongside `-fstack-arrays` and 
`-flang-experimental-hlfir`? It sound like we could introduce another hook 
here, `adddCodeGenOptions`?



Comment at: flang/test/Driver/version-loops.f90:2
+! Test that flang-new forwards the -f{no-,}version-loops-for-stride 
+! options corredly to flang-new -fc1 for different variants of optimisation
+! and explicit flags.





Comment at: flang/test/Driver/version-loops.f90:5
+
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O3 \

[nit] If you are using `-###`, then you can just skip `-fsyntax-only` (it 
doesn't really matter what "action" is requested from the frontend driver).



Comment at: flang/test/Driver/version-loops.f90:33-34
+  
+! CHECK: "-fversion-loops-for-stride"
+! CHECK: "-O3"
+

Similar suggestion for other `CHECK` lines - this will make the test a bit more 
explicit about the expected output.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D141307/new/

https://reviews.llvm.org/D141307

___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits


[PATCH] D141307: Add -f[no-]loop-versioning option

2023-04-06 Thread Mats Petersson via Phabricator via cfe-commits
Leporacanthicus updated this revision to Diff 511416.
Leporacanthicus added a comment.

Rebase only


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D141307/new/

https://reviews.llvm.org/D141307

Files:
  clang/include/clang/Driver/Options.td
  clang/lib/Driver/ToolChains/Flang.cpp
  flang/include/flang/Frontend/CodeGenOptions.def
  flang/include/flang/Tools/CLOptions.inc
  flang/lib/Frontend/CompilerInvocation.cpp
  flang/lib/Frontend/FrontendActions.cpp
  flang/test/Driver/driver-help-hidden.f90
  flang/test/Driver/driver-help.f90
  flang/test/Driver/frontend-forwarding.f90
  flang/test/Driver/version-loops.f90

Index: flang/test/Driver/version-loops.f90
===
--- /dev/null
+++ flang/test/Driver/version-loops.f90
@@ -0,0 +1,54 @@
+! Test that flang-new forwards the -f{no-,}version-loops-for-stride 
+! options corredly to flang-new -fc1 for different variants of optimisation
+! and explicit flags.
+
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O3 \
+! RUN:   | FileCheck %s
+  
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O2 \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O2
+
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O2 -fversion-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O2-with
+  
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O4 \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O4
+  
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -Ofast \
+! RUN:   | FileCheck %s --check-prefix=CHECK-Ofast
+  
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -Ofast -fno-version-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-Ofast-no
+
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O3 -fno-version-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O3-no
+  
+! CHECK: "-fversion-loops-for-stride"
+! CHECK: "-O3"
+
+! CHECK-O2-NOT: "-fversion-loops-for-stride"
+! CHECK-O2: "-O2"  
+
+! CHECK-O2-with: "-fversion-loops-for-stride"
+! CHECK-O2-with: "-O2"  
+  
+! CHECK-O4: "-fversion-loops-for-stride"
+! CHECK-O4: "-O3"
+
+! CHECK-Ofast: "-ffast-math"
+! CHECK-Ofast: "-fversion-loops-for-stride"
+! CHECK-Ofast: "-O3"
+
+! CHECK-Ofast-no: "-ffast-math"
+! CHECK-Ofast-no-NOT: "-fversion-loops-for-stride"
+! CHECK-Ofast-no: "-O3"
+
+! CHECK-O3-no-NOT: "-fversion-loops-for-stride"
+! CHECK-O3-no: "-O3"
Index: flang/test/Driver/frontend-forwarding.f90
===
--- flang/test/Driver/frontend-forwarding.f90
+++ flang/test/Driver/frontend-forwarding.f90
@@ -15,6 +15,8 @@
 ! RUN: -fassociative-math \
 ! RUN: -freciprocal-math \
 ! RUN: -fpass-plugin=Bye%pluginext \
+! RUN: -fversion-loops-for-stride \
+! RUN: -mllvm -print-before-all\
 ! RUN: -mllvm -print-before-all \
 ! RUN: -save-temps=obj \
 ! RUN: -P \
@@ -34,5 +36,6 @@
 ! CHECK: "-freciprocal-math"
 ! CHECK: "-fconvert=little-endian"
 ! CHECK: "-fpass-plugin=Bye
+! CHECK: "-fversion-loops-for-stride"  
 ! CHECK: "-mllvm" "-print-before-all"
 ! CHECK: "-save-temps=obj"
Index: flang/test/Driver/driver-help.f90
===
--- flang/test/Driver/driver-help.f90
+++ flang/test/Driver/driver-help.f90
@@ -48,6 +48,8 @@
 ! HELP-NEXT: -fno-integrated-as  Disable the integrated assembler
 ! HELP-NEXT: -fno-signed-zeros  Allow optimizations that ignore the sign of floating point zeros
 ! HELP-NEXT: -fno-stack-arrays  Allocate array temporaries on the heap (default)
+! HELP-NEXT: -fno-version-loops-for-stride
+! HELP-NEXT:Do not create unit-strided loops (default)
 ! HELP-NEXT: -fopenacc  Enable OpenACC
 ! HELP-NEXT: -fopenmp   Parse OpenMP pragmas and generate parallel code.
 ! HELP-NEXT: -fpass-plugin= Load pass plugin from a dynamic shared object file (only with new pass manager).
@@ -55,6 +57,8 @@
 ! HELP-NEXT: -fstack-arrays Attempt to allocate array temporaries on the stack, no matter their size
 ! HELP-NEXT: -fsyntax-only  Run the preprocessor, parser and semantic analysis stages
 ! HELP-NEXT: -funderscoring Appends one trailing underscore to external names
+! HELP-NEXT: -fversion-loops-for-stride
+! HELP-NEXT:Create unit-strided versions of loops
 ! HELP-NEXT: -fxor-operator Enable .XOR. as a synonym of .NEQV.
 ! HELP-NEXT: -gline-tables-only Emit debug line number tables only
 ! HELP-NEXT: -g Generate source-level debug information
@@ -143,6 +147,8 @@
 ! HELP-FC1-NEXT: -fno-reformat  Dump the cooked character stream in -E mode
 ! HELP-FC1-NEXT: -fno-signed-zeros  Allow optimizations that ignore the sign of floating point zeros
 ! HELP-FC1-NEXT: -fno-stack-arrays  Allocate array temporaries on the heap (default)
+! 

[PATCH] D141307: Add -f[no-]loop-versioning option

2023-04-04 Thread Mats Petersson via Phabricator via cfe-commits
Leporacanthicus updated this revision to Diff 510756.
Leporacanthicus added a comment.

Rebase.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D141307/new/

https://reviews.llvm.org/D141307

Files:
  clang/include/clang/Driver/Options.td
  clang/lib/Driver/ToolChains/Flang.cpp
  flang/include/flang/Frontend/CodeGenOptions.def
  flang/include/flang/Tools/CLOptions.inc
  flang/lib/Frontend/CompilerInvocation.cpp
  flang/lib/Frontend/FrontendActions.cpp
  flang/test/Driver/driver-help-hidden.f90
  flang/test/Driver/driver-help.f90
  flang/test/Driver/frontend-forwarding.f90
  flang/test/Driver/version-loops.f90

Index: flang/test/Driver/version-loops.f90
===
--- /dev/null
+++ flang/test/Driver/version-loops.f90
@@ -0,0 +1,54 @@
+! Test that flang-new forwards the -f{no-,}version-loops-for-stride 
+! options corredly to flang-new -fc1 for different variants of optimisation
+! and explicit flags.
+
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O3 \
+! RUN:   | FileCheck %s
+  
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O2 \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O2
+
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O2 -fversion-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O2-with
+  
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O4 \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O4
+  
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -Ofast \
+! RUN:   | FileCheck %s --check-prefix=CHECK-Ofast
+  
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -Ofast -fno-version-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-Ofast-no
+
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O3 -fno-version-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O3-no
+  
+! CHECK: "-fversion-loops-for-stride"
+! CHECK: "-O3"
+
+! CHECK-O2-NOT: "-fversion-loops-for-stride"
+! CHECK-O2: "-O2"  
+
+! CHECK-O2-with: "-fversion-loops-for-stride"
+! CHECK-O2-with: "-O2"  
+  
+! CHECK-O4: "-fversion-loops-for-stride"
+! CHECK-O4: "-O3"
+
+! CHECK-Ofast: "-ffast-math"
+! CHECK-Ofast: "-fversion-loops-for-stride"
+! CHECK-Ofast: "-O3"
+
+! CHECK-Ofast-no: "-ffast-math"
+! CHECK-Ofast-no-NOT: "-fversion-loops-for-stride"
+! CHECK-Ofast-no: "-O3"
+
+! CHECK-O3-no-NOT: "-fversion-loops-for-stride"
+! CHECK-O3-no: "-O3"
Index: flang/test/Driver/frontend-forwarding.f90
===
--- flang/test/Driver/frontend-forwarding.f90
+++ flang/test/Driver/frontend-forwarding.f90
@@ -15,6 +15,8 @@
 ! RUN: -fassociative-math \
 ! RUN: -freciprocal-math \
 ! RUN: -fpass-plugin=Bye%pluginext \
+! RUN: -fversion-loops-for-stride \
+! RUN: -mllvm -print-before-all\
 ! RUN: -mllvm -print-before-all \
 ! RUN: -save-temps=obj \
 ! RUN: -P \
@@ -34,5 +36,6 @@
 ! CHECK: "-freciprocal-math"
 ! CHECK: "-fconvert=little-endian"
 ! CHECK: "-fpass-plugin=Bye
+! CHECK: "-fversion-loops-for-stride"  
 ! CHECK: "-mllvm" "-print-before-all"
 ! CHECK: "-save-temps=obj"
Index: flang/test/Driver/driver-help.f90
===
--- flang/test/Driver/driver-help.f90
+++ flang/test/Driver/driver-help.f90
@@ -48,6 +48,8 @@
 ! HELP-NEXT: -fno-integrated-as  Disable the integrated assembler
 ! HELP-NEXT: -fno-signed-zeros  Allow optimizations that ignore the sign of floating point zeros
 ! HELP-NEXT: -fno-stack-arrays  Allocate array temporaries on the heap (default)
+! HELP-NEXT: -fno-version-loops-for-stride
+! HELP-NEXT:Do not create unit-strided loops (default)
 ! HELP-NEXT: -fopenacc  Enable OpenACC
 ! HELP-NEXT: -fopenmp   Parse OpenMP pragmas and generate parallel code.
 ! HELP-NEXT: -fpass-plugin= Load pass plugin from a dynamic shared object file (only with new pass manager).
@@ -55,6 +57,8 @@
 ! HELP-NEXT: -fstack-arrays Attempt to allocate array temporaries on the stack, no matter their size
 ! HELP-NEXT: -fsyntax-only  Run the preprocessor, parser and semantic analysis stages
 ! HELP-NEXT: -funderscoring Appends one trailing underscore to external names
+! HELP-NEXT: -fversion-loops-for-stride
+! HELP-NEXT:Create unit-strided versions of loops
 ! HELP-NEXT: -fxor-operator Enable .XOR. as a synonym of .NEQV.
 ! HELP-NEXT: -gline-tables-only Emit debug line number tables only
 ! HELP-NEXT: -g Generate source-level debug information
@@ -143,6 +147,8 @@
 ! HELP-FC1-NEXT: -fno-reformat  Dump the cooked character stream in -E mode
 ! HELP-FC1-NEXT: -fno-signed-zeros  Allow optimizations that ignore the sign of floating point zeros
 ! HELP-FC1-NEXT: -fno-stack-arrays  Allocate array temporaries on the heap (default)
+! 

[PATCH] D141307: Add -f[no-]loop-versioning option

2023-03-17 Thread Mats Petersson via Phabricator via cfe-commits
Leporacanthicus updated this revision to Diff 506125.
Leporacanthicus added a comment.

Rebase and fix conflicts


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D141307/new/

https://reviews.llvm.org/D141307

Files:
  clang/include/clang/Driver/Options.td
  clang/lib/Driver/ToolChains/Flang.cpp
  flang/include/flang/Frontend/CodeGenOptions.def
  flang/include/flang/Tools/CLOptions.inc
  flang/lib/Frontend/CompilerInvocation.cpp
  flang/lib/Frontend/FrontendActions.cpp
  flang/test/Driver/driver-help-hidden.f90
  flang/test/Driver/driver-help.f90
  flang/test/Driver/frontend-forwarding.f90
  flang/test/Driver/version-loops.f90

Index: flang/test/Driver/version-loops.f90
===
--- /dev/null
+++ flang/test/Driver/version-loops.f90
@@ -0,0 +1,54 @@
+! Test that flang-new forwards the -f{no-,}version-loops-for-stride 
+! options corredly to flang-new -fc1 for different variants of optimisation
+! and explicit flags.
+
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O3 \
+! RUN:   | FileCheck %s
+  
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O2 \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O2
+
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O2 -fversion-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O2-with
+  
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O4 \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O4
+  
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -Ofast \
+! RUN:   | FileCheck %s --check-prefix=CHECK-Ofast
+  
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -Ofast -fno-version-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-Ofast-no
+
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O3 -fno-version-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O3-no
+  
+! CHECK: "-fversion-loops-for-stride"
+! CHECK: "-O3"
+
+! CHECK-O2-NOT: "-fversion-loops-for-stride"
+! CHECK-O2: "-O2"  
+
+! CHECK-O2-with: "-fversion-loops-for-stride"
+! CHECK-O2-with: "-O2"  
+  
+! CHECK-O4: "-fversion-loops-for-stride"
+! CHECK-O4: "-O3"
+
+! CHECK-Ofast: "-ffast-math"
+! CHECK-Ofast: "-fversion-loops-for-stride"
+! CHECK-Ofast: "-O3"
+
+! CHECK-Ofast-no: "-ffast-math"
+! CHECK-Ofast-no-NOT: "-fversion-loops-for-stride"
+! CHECK-Ofast-no: "-O3"
+
+! CHECK-O3-no-NOT: "-fversion-loops-for-stride"
+! CHECK-O3-no: "-O3"
Index: flang/test/Driver/frontend-forwarding.f90
===
--- flang/test/Driver/frontend-forwarding.f90
+++ flang/test/Driver/frontend-forwarding.f90
@@ -15,6 +15,7 @@
 ! RUN: -fassociative-math \
 ! RUN: -freciprocal-math \
 ! RUN: -fpass-plugin=Bye%pluginext \
+! RUN: -fversion-loops-for-stride \
 ! RUN: -mllvm -print-before-all\
 ! RUN: -P \
 ! RUN:   | FileCheck %s
@@ -33,4 +34,5 @@
 ! CHECK: "-freciprocal-math"
 ! CHECK: "-fconvert=little-endian"
 ! CHECK: "-fpass-plugin=Bye
+! CHECK: "-fversion-loops-for-stride"  
 ! CHECK: "-mllvm" "-print-before-all"
Index: flang/test/Driver/driver-help.f90
===
--- flang/test/Driver/driver-help.f90
+++ flang/test/Driver/driver-help.f90
@@ -48,6 +48,8 @@
 ! HELP-NEXT: -fno-integrated-as  Disable the integrated assembler
 ! HELP-NEXT: -fno-signed-zeros  Allow optimizations that ignore the sign of floating point zeros
 ! HELP-NEXT: -fno-stack-arrays  Allocate array temporaries on the heap (default)
+! HELP-NEXT: -fno-version-loops-for-stride
+! HELP-NEXT:Do not create unit-strided loops (default)
 ! HELP-NEXT: -fopenacc  Enable OpenACC
 ! HELP-NEXT: -fopenmp   Parse OpenMP pragmas and generate parallel code.
 ! HELP-NEXT: -fpass-plugin= Load pass plugin from a dynamic shared object file (only with new pass manager).
@@ -55,6 +57,8 @@
 ! HELP-NEXT: -fstack-arrays Attempt to allocate array temporaries on the stack, no matter their size
 ! HELP-NEXT: -fsyntax-only  Run the preprocessor, parser and semantic analysis stages
 ! HELP-NEXT: -funderscoring Appends one trailing underscore to external names
+! HELP-NEXT: -fversion-loops-for-stride
+! HELP-NEXT:Create unit-strided versions of loops
 ! HELP-NEXT: -fxor-operator Enable .XOR. as a synonym of .NEQV.
 ! HELP-NEXT: -help  Display available options
 ! HELP-NEXT: -IAdd directory to the end of the list of include search paths
@@ -141,6 +145,8 @@
 ! HELP-FC1-NEXT: -fno-reformat  Dump the cooked character stream in -E mode
 ! HELP-FC1-NEXT: -fno-signed-zeros  Allow optimizations that ignore the sign of floating point zeros
 ! HELP-FC1-NEXT: -fno-stack-arrays  Allocate array temporaries on the heap (default)
+! HELP-FC1-NEXT: -fno-version-loops-for-stride
+! 

[PATCH] D141307: Add -f[no-]loop-versioning option

2023-03-09 Thread Tom Eccles via Phabricator via cfe-commits
tblah accepted this revision.
tblah added a comment.
This revision is now accepted and ready to land.

Looks good to me


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D141307/new/

https://reviews.llvm.org/D141307

___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits


[PATCH] D141307: Add -f[no-]loop-versioning option

2023-03-08 Thread Mats Petersson via Phabricator via cfe-commits
Leporacanthicus updated this revision to Diff 503459.
Leporacanthicus marked an inline comment as done.
Leporacanthicus added a comment.

Fix copied comment to reflect the new content.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D141307/new/

https://reviews.llvm.org/D141307

Files:
  clang/include/clang/Driver/Options.td
  clang/lib/Driver/ToolChains/Flang.cpp
  flang/include/flang/Frontend/CodeGenOptions.def
  flang/include/flang/Tools/CLOptions.inc
  flang/lib/Frontend/CompilerInvocation.cpp
  flang/lib/Frontend/FrontendActions.cpp
  flang/test/Driver/driver-help-hidden.f90
  flang/test/Driver/driver-help.f90
  flang/test/Driver/frontend-forwarding.f90
  flang/test/Driver/version-loops.f90

Index: flang/test/Driver/version-loops.f90
===
--- /dev/null
+++ flang/test/Driver/version-loops.f90
@@ -0,0 +1,54 @@
+! Test that flang-new forwards the -f{no-,}version-loops-for-stride 
+! options corredly to flang-new -fc1 for different variants of optimisation
+! and explicit flags.
+
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O3 \
+! RUN:   | FileCheck %s
+  
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O2 \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O2
+
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O2 -fversion-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O2-with
+  
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O4 \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O4
+  
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -Ofast \
+! RUN:   | FileCheck %s --check-prefix=CHECK-Ofast
+  
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -Ofast -fno-version-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-Ofast-no
+
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O3 -fno-version-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O3-no
+  
+! CHECK: "-fversion-loops-for-stride"
+! CHECK: "-O3"
+
+! CHECK-O2-NOT: "-fversion-loops-for-stride"
+! CHECK-O2: "-O2"  
+
+! CHECK-O2-with: "-fversion-loops-for-stride"
+! CHECK-O2-with: "-O2"  
+  
+! CHECK-O4: "-fversion-loops-for-stride"
+! CHECK-O4: "-O3"
+
+! CHECK-Ofast: "-ffast-math"
+! CHECK-Ofast: "-fversion-loops-for-stride"
+! CHECK-Ofast: "-O3"
+
+! CHECK-Ofast-no: "-ffast-math"
+! CHECK-Ofast-no-NOT: "-fversion-loops-for-stride"
+! CHECK-Ofast-no: "-O3"
+
+! CHECK-O3-no-NOT: "-fversion-loops-for-stride"
+! CHECK-O3-no: "-O3"
Index: flang/test/Driver/frontend-forwarding.f90
===
--- flang/test/Driver/frontend-forwarding.f90
+++ flang/test/Driver/frontend-forwarding.f90
@@ -15,6 +15,7 @@
 ! RUN: -fassociative-math \
 ! RUN: -freciprocal-math \
 ! RUN: -fpass-plugin=Bye%pluginext \
+! RUN: -fversion-loops-for-stride \
 ! RUN: -mllvm -print-before-all\
 ! RUN: -P \
 ! RUN:   | FileCheck %s
@@ -33,4 +34,5 @@
 ! CHECK: "-freciprocal-math"
 ! CHECK: "-fconvert=little-endian"
 ! CHECK: "-fpass-plugin=Bye
+! CHECK: "-fversion-loops-for-stride"  
 ! CHECK: "-mllvm" "-print-before-all"
Index: flang/test/Driver/driver-help.f90
===
--- flang/test/Driver/driver-help.f90
+++ flang/test/Driver/driver-help.f90
@@ -46,6 +46,8 @@
 ! HELP-NEXT: -fno-integrated-as  Disable the integrated assembler
 ! HELP-NEXT: -fno-signed-zeros  Allow optimizations that ignore the sign of floating point zeros
 ! HELP-NEXT: -fno-stack-arrays  Allocate array temporaries on the heap (default)
+! HELP-NEXT: -fno-version-loops-for-stride
+! HELP-NEXT:Do not create unit-strided loops (default)
 ! HELP-NEXT: -fopenacc  Enable OpenACC
 ! HELP-NEXT: -fopenmp   Parse OpenMP pragmas and generate parallel code.
 ! HELP-NEXT: -fpass-plugin= Load pass plugin from a dynamic shared object file (only with new pass manager).
@@ -53,6 +55,8 @@
 ! HELP-NEXT: -fstack-arrays Attempt to allocate array temporaries on the stack, no matter their size
 ! HELP-NEXT: -fsyntax-only  Run the preprocessor, parser and semantic analysis stages
 ! HELP-NEXT: -funderscoring Appends one trailing underscore to external names
+! HELP-NEXT: -fversion-loops-for-stride
+! HELP-NEXT:Create unit-strided versions of loops
 ! HELP-NEXT: -fxor-operator Enable .XOR. as a synonym of .NEQV.
 ! HELP-NEXT: -help  Display available options
 ! HELP-NEXT: -IAdd directory to the end of the list of include search paths
@@ -137,6 +141,8 @@
 ! HELP-FC1-NEXT: -fno-reformat  Dump the cooked character stream in -E mode
 ! HELP-FC1-NEXT: -fno-signed-zeros  Allow optimizations that ignore the sign of floating point zeros
 ! HELP-FC1-NEXT: -fno-stack-arrays  Allocate array temporaries on the heap 

[PATCH] D141307: Add -f[no-]loop-versioning option

2023-03-08 Thread Mats Petersson via Phabricator via cfe-commits
Leporacanthicus marked an inline comment as done.
Leporacanthicus added inline comments.



Comment at: clang/lib/Driver/ToolChains/Flang.cpp:69
+  Arg *loopVersioning =
+  Args.getLastArg(options::OPT_Ofast, options::OPT_floop_versioning,
+  options::OPT_fno_loop_versioning);

tblah wrote:
> Does GFortran only enable this with -Ofast? I would have thought this would 
> be an -On thing because it doesn't impact precision.
No, good point.



Comment at: flang/include/flang/Frontend/CodeGenOptions.def:27
 
+CODEGENOPT(LoopVersioning, 1, 0) ///< PIC level of the LLVM module.
 CODEGENOPT(PICLevel, 2, 0) ///< PIC level of the LLVM module.

tblah wrote:
> nit: comment
Darn, missed fixing this.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D141307/new/

https://reviews.llvm.org/D141307

___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits


[PATCH] D141307: Add -f[no-]loop-versioning option

2023-03-08 Thread Mats Petersson via Phabricator via cfe-commits
Leporacanthicus updated this revision to Diff 503360.
Leporacanthicus added a comment.
Herald added a reviewer: sscalpone.

Updates based on review comments:

- Add tests.
- Enable on -O3

Also changed the name to match gfortran.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D141307/new/

https://reviews.llvm.org/D141307

Files:
  clang/include/clang/Driver/Options.td
  clang/lib/Driver/ToolChains/Flang.cpp
  flang/include/flang/Frontend/CodeGenOptions.def
  flang/include/flang/Tools/CLOptions.inc
  flang/lib/Frontend/CompilerInvocation.cpp
  flang/lib/Frontend/FrontendActions.cpp
  flang/test/Driver/driver-help-hidden.f90
  flang/test/Driver/driver-help.f90
  flang/test/Driver/frontend-forwarding.f90
  flang/test/Driver/version-loops.f90

Index: flang/test/Driver/version-loops.f90
===
--- /dev/null
+++ flang/test/Driver/version-loops.f90
@@ -0,0 +1,54 @@
+! Test that flang-new forwards the -f{no-,}version-loops-for-stride 
+! options corredly to flang-new -fc1 for different variants of optimisation
+! and explicit flags.
+
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O3 \
+! RUN:   | FileCheck %s
+  
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O2 \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O2
+
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O2 -fversion-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O2-with
+  
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O4 \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O4
+  
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -Ofast \
+! RUN:   | FileCheck %s --check-prefix=CHECK-Ofast
+  
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -Ofast -fno-version-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-Ofast-no
+
+! RUN: %flang -fsyntax-only -### %s -o %t 2>&1 \
+! RUN:   -O3 -fno-version-loops-for-stride \
+! RUN:   | FileCheck %s --check-prefix=CHECK-O3-no
+  
+! CHECK: "-fversion-loops-for-stride"
+! CHECK: "-O3"
+
+! CHECK-O2-NOT: "-fversion-loops-for-stride"
+! CHECK-O2: "-O2"  
+
+! CHECK-O2-with: "-fversion-loops-for-stride"
+! CHECK-O2-with: "-O2"  
+  
+! CHECK-O4: "-fversion-loops-for-stride"
+! CHECK-O4: "-O3"
+
+! CHECK-Ofast: "-ffast-math"
+! CHECK-Ofast: "-fversion-loops-for-stride"
+! CHECK-Ofast: "-O3"
+
+! CHECK-Ofast-no: "-ffast-math"
+! CHECK-Ofast-no-NOT: "-fversion-loops-for-stride"
+! CHECK-Ofast-no: "-O3"
+
+! CHECK-O3-no-NOT: "-fversion-loops-for-stride"
+! CHECK-O3-no: "-O3"
Index: flang/test/Driver/frontend-forwarding.f90
===
--- flang/test/Driver/frontend-forwarding.f90
+++ flang/test/Driver/frontend-forwarding.f90
@@ -15,6 +15,7 @@
 ! RUN: -fassociative-math \
 ! RUN: -freciprocal-math \
 ! RUN: -fpass-plugin=Bye%pluginext \
+! RUN: -fversion-loops-for-stride \
 ! RUN: -mllvm -print-before-all\
 ! RUN: -P \
 ! RUN:   | FileCheck %s
@@ -33,4 +34,5 @@
 ! CHECK: "-freciprocal-math"
 ! CHECK: "-fconvert=little-endian"
 ! CHECK: "-fpass-plugin=Bye
+! CHECK: "-fversion-loops-for-stride"  
 ! CHECK: "-mllvm" "-print-before-all"
Index: flang/test/Driver/driver-help.f90
===
--- flang/test/Driver/driver-help.f90
+++ flang/test/Driver/driver-help.f90
@@ -46,6 +46,8 @@
 ! HELP-NEXT: -fno-integrated-as  Disable the integrated assembler
 ! HELP-NEXT: -fno-signed-zeros  Allow optimizations that ignore the sign of floating point zeros
 ! HELP-NEXT: -fno-stack-arrays  Allocate array temporaries on the heap (default)
+! HELP-NEXT: -fno-version-loops-for-stride
+! HELP-NEXT:Do not create unit-strided loops (default)
 ! HELP-NEXT: -fopenacc  Enable OpenACC
 ! HELP-NEXT: -fopenmp   Parse OpenMP pragmas and generate parallel code.
 ! HELP-NEXT: -fpass-plugin= Load pass plugin from a dynamic shared object file (only with new pass manager).
@@ -53,6 +55,8 @@
 ! HELP-NEXT: -fstack-arrays Attempt to allocate array temporaries on the stack, no matter their size
 ! HELP-NEXT: -fsyntax-only  Run the preprocessor, parser and semantic analysis stages
 ! HELP-NEXT: -funderscoring Appends one trailing underscore to external names
+! HELP-NEXT: -fversion-loops-for-stride
+! HELP-NEXT:Create unit-strided versions of loops
 ! HELP-NEXT: -fxor-operator Enable .XOR. as a synonym of .NEQV.
 ! HELP-NEXT: -help  Display available options
 ! HELP-NEXT: -IAdd directory to the end of the list of include search paths
@@ -137,6 +141,8 @@
 ! HELP-FC1-NEXT: -fno-reformat  Dump the cooked character stream in -E mode
 ! HELP-FC1-NEXT: -fno-signed-zeros  Allow optimizations that ignore the sign of floating point zeros
 ! HELP-FC1-NEXT: -fno-stack-arrays 

[PATCH] D141307: Add -f[no-]loop-versioning option

2023-03-06 Thread Tom Eccles via Phabricator via cfe-commits
tblah added a comment.

Please could you add tests for the flag forwarding logic in 
flang/tests/Driver/frontend-forwarding.f90. For example see 
https://reviews.llvm.org/rGd0d4b635786d510cd919cadbeb7e5e19983242cf




Comment at: clang/lib/Driver/ToolChains/Flang.cpp:69
+  Arg *loopVersioning =
+  Args.getLastArg(options::OPT_Ofast, options::OPT_floop_versioning,
+  options::OPT_fno_loop_versioning);

Does GFortran only enable this with -Ofast? I would have thought this would be 
an -On thing because it doesn't impact precision.



Comment at: flang/include/flang/Frontend/CodeGenOptions.def:27
 
+CODEGENOPT(LoopVersioning, 1, 0) ///< PIC level of the LLVM module.
 CODEGENOPT(PICLevel, 2, 0) ///< PIC level of the LLVM module.

nit: comment


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D141307/new/

https://reviews.llvm.org/D141307

___
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits