[NFC][OpenMP] Use clang_cc1 to driver tests

[jsji]

Summary

The test driver-fopenmp-extensions.c is failing on platforms that does
not use integrated-as. It can be reproduced using -fno-integrated-as on
Linux too.

bin/clang -c -Xclang -verify=omp -fopenmp -fopenmp-extensions
-fno-openmp-extensions
../llvm-project/clang/test/OpenMP/driver-fopenmp-extensions.c
-fno-integrated-as
Assembler messages:
Error: can't open /tmp/driver-fopenmp-extensions-8fafe8.s for reading:
No such file or directory
clang-14: error: assembler command failed with exit code 1 (use -v to
see invocation)

The goal of this test is to verify syntax diags only,
so we should use clang_cc1 to test.

[ABataev]

I think we can use just clang_cc1 here instead of %clang -c

[jsji]

Use clang_cc1 instead.

[jsji]

In D109255#2983002, @ABataev wrote:

I think we can use just clang_cc1 here instead of %clang -c

Good idea.

[ABataev]

LG

[jdenny]

This test is meant to check that the driver processes these options correctly. -cc1 isn't the driver and often ignores either the positive or negative version of an option.

[jsji]

In D109255#2983017, @jdenny wrote:

This test is meant to check that the driver processes these options correctly. -cc1 isn't the driver and often ignores either the positive or negative version of an option.

OK, I can revert to driver + fsyntax-only if that is preferred?

[jdenny]

In D109255#2983018, @jsji wrote:

In D109255#2983017, @jdenny wrote:

This test is meant to check that the driver processes these options correctly. -cc1 isn't the driver and often ignores either the positive or negative version of an option.

OK, I can revert to driver + fsyntax-only if that is preferred?

That would be my preference. Otherwise, we never test the interface exposed to the user, and the test is misnamed.

[jsji]

Use driver + -fsyntax-only.

[ABataev]

In D109255#2983017, @jdenny wrote:

This test is meant to check that the driver processes these options correctly. -cc1 isn't the driver and often ignores either the positive or negative version of an option.

Better not to do such checks, just need to check that driver correctly translates driver options to the frontend options (using -###), better to use -verify only with the frontend tests. Suggest to rework this test.

[jsji]

@ABataev Is this still OK for you?

[jsji]

In D109255#2983052, @ABataev wrote:

In D109255#2983017, @jdenny wrote:

This test is meant to check that the driver processes these options correctly. -cc1 isn't the driver and often ignores either the positive or negative version of an option.

Better not to do such checks, just need to check that driver correctly translates driver options to the frontend options (using -###), better to use -verify only with the frontend tests. Suggest to rework this test.

Make sense to me. How about we use _cc1 as Alexey suggested here (and rename the file to remove driver- prefix). Maybe Joel can add another test to test the driver correctly translates driver options to the frontend options (using -###) in test/Driver dir?

[jsji]

use _cc1 for now.

[jdenny]

In D109255#2983063, @jsji wrote:

In D109255#2983052, @ABataev wrote:

In D109255#2983017, @jdenny wrote:

This test is meant to check that the driver processes these options correctly. -cc1 isn't the driver and often ignores either the positive or negative version of an option.

Better not to do such checks, just need to check that driver correctly translates driver options to the frontend options (using -###), better to use -verify only with the frontend tests. Suggest to rework this test.

Make sense to me. How about we use _cc1 as Alexey suggested here (and rename the file to remove driver- prefix).

OK.

Maybe Joel can add another test to test the driver correctly translates driver options to the frontend options (using -###) in test/Driver dir?

I probably won't get to it today, but I can work on it later.

[jsji]

I probably won't get to it today, but I can work on it later.

Sure, take your time. Thank you!

[jdenny]

LGTM. Thanks for the fix.

clang/test/OpenMP/driver-fopenmp-extensions.c

-// RUN: %clang -c -Xclang -verify=ompx -fopenmp %s
-// RUN: %clang -c -Xclang -verify=ompx -fopenmp-simd %s
-
-// RUN: %clang -c -Xclang -verify=ompx -fopenmp -fopenmp-extensions %s
-// RUN: %clang -c -Xclang -verify=ompx -fopenmp-simd -fopenmp-extensions %s
-
-// RUN: %clang -c -Xclang -verify=omp -fopenmp -fno-openmp-extensions %s
-// RUN: %clang -c -Xclang -verify=omp -fopenmp-simd -fno-openmp-extensions %s
-
-// RUN: %clang -c -Xclang -verify=omp -fopenmp \
-// RUN:     -fopenmp-extensions -fno-openmp-extensions %s
-// RUN: %clang -c -Xclang -verify=omp -fopenmp-simd \
-// RUN:     -fopenmp-extensions -fno-openmp-extensions %s
-
-// RUN: %clang -c -Xclang -verify=ompx -fopenmp \
-// RUN:     -fno-openmp-extensions -fopenmp-extensions %s
-// RUN: %clang -c -Xclang -verify=ompx -fopenmp-simd \
-// RUN:     -fno-openmp-extensions -fopenmp-extensions %s
-
-void foo() {
-  int x;
-  // ompx-no-diagnostics
-  // omp-error@+1 {{incorrect map type modifier}}
-  #pragma omp target map(ompx_hold, alloc: x)
-  ;
-}

clang/test/OpenMP/fopenmp-extensions.c

+// RUN: %clang_cc1 -verify=ompx -fopenmp %s
+// RUN: %clang_cc1 -verify=ompx -fopenmp-simd %s
+
+// RUN: %clang_cc1 -verify=ompx -fopenmp -fopenmp-extensions %s
+// RUN: %clang_cc1 -verify=ompx -fopenmp-simd -fopenmp-extensions %s
+
+// RUN: %clang_cc1 -verify=omp -fopenmp -fno-openmp-extensions %s
+// RUN: %clang_cc1 -verify=omp -fopenmp-simd -fno-openmp-extensions %s
+
+// RUN: %clang_cc1 -verify=omp -fopenmp \
+// RUN:     -fopenmp-extensions -fno-openmp-extensions %s
+// RUN: %clang_cc1 -verify=omp -fopenmp-simd \
+// RUN:     -fopenmp-extensions -fno-openmp-extensions %s
+
+// RUN: %clang_cc1 -verify=ompx -fopenmp \
+// RUN:     -fno-openmp-extensions -fopenmp-extensions %s
+// RUN: %clang_cc1 -verify=ompx -fopenmp-simd \
+// RUN:     -fno-openmp-extensions -fopenmp-extensions %s
+
+void foo() {
+  int x;
+  // ompx-no-diagnostics
+  // omp-error@+1 {{incorrect map type modifier}}
+  #pragma omp target map(ompx_hold, alloc: x)
+  ;
+}