Re: [PATCH 4/5] testsuite: [arm/mve] Use dg-add-options arm_v8_1m_mve in MVE tests
On Tue, 14 Apr 2020 at 10:38, Andre Vieira (lists)
wrote:
>
> On 10/04/2020 14:55, Christophe Lyon via Gcc-patches wrote:
> > Several ARM/MVE tests can be compiled even if the toolchain does not
> > support -mfloat-abi=hard (softfp is OK).
> >
> > Use dg-add-options arm_v8_1m_mve or arm_v8_1m_mve_fp instead of using
> > dg-additional-options.
> Hi Christophe,
>
> I think a bunch of these tests were initially meant to test hard float
> abi with vectors, especially in the MVE integer cases, this is what the
> scan-assemblers are meant to test. However, it seems to pass for
> float-abi=softfp too ... which means these tests don't really mean
> anything. I think it would be good to remove the scan-assembler tests
> for now and improve these tests later with run tests, or some smarter
> function body testing.
>
> I suggest we apply your changes and I can follow up with a patch to
> remove the scan-assemblers for now, if a maintainer agrees with me that is.
Not sure why I missed that, but mve_vector_float2.c
fails on arm-linux-gnueabi, because it's scan-assembler works only
with mfloat-abi=hard:
gcc.target/arm/mve/intrinsics/mve_vector_float2.c scan-assembler
vmov\\tq[0-7], q[0-7]
I checked that
diff --git a/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float2.c
b/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float2.c
index e3cf8f8..10fe904 100644
--- a/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float2.c
+++ b/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float2.c
@@ -1,6 +1,7 @@
+/* { dg-require-effective-target arm_hard_ok } */
/* { dg-require-effective-target arm_v8_1m_mve_fp_ok } */
/* { dg-add-options arm_v8_1m_mve_fp } */
-/* { dg-skip-if "Incompatible float ABI" { *-*-* } {
"-mfloat-abi=soft" } {""} } */
+/* { dg-additional-options "-mfloat-abi=hard" } */
#include "arm_mve.h"
does the trick, but I also tried the check-function-bodies route. I
hoped it would
be easy, but... it wasn't because the testcase is also executed in C++
mode, so the function label is no longer foo32: but _Z5foo32v:
and check-function-bodies wants exact matching here.
Sigh
Christophe
>
> Cheers,
> Andre
> > 2020-04-10 Christophe Lyon
> >
> > gcc/testsuite/
> > * gcc.target/arm/mve/intrinsics/mve_vector_float.c: Use
> > arm_v8_1m_mve_fp.
> > * gcc.target/arm/mve/intrinsics/mve_vector_float1.c: Likewise.
> > * gcc.target/arm/mve/intrinsics/mve_vector_float2.c: Likewise.
> > * gcc.target/arm/mve/intrinsics/mve_vector_int.c: Use
> > arm_v8_1m_mve.
> > * gcc.target/arm/mve/intrinsics/mve_vector_int1.c: Likewise.
> > * gcc.target/arm/mve/intrinsics/mve_vector_int2.c: Likewise.
> > * gcc.target/arm/mve/intrinsics/mve_vector_uint.c: Likewise.
> > * gcc.target/arm/mve/intrinsics/mve_vector_uint1.c: Likewise.
> > * gcc.target/arm/mve/intrinsics/mve_vector_uint2.c: Likewise.
> > ---
> > gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float.c | 2 +-
> > gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float1.c | 2 +-
> > gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float2.c | 2 +-
> > gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_int.c| 2 +-
> > gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_int1.c | 2 +-
> > gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_int2.c | 2 +-
> > gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_uint.c | 2 +-
> > gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_uint1.c | 2 +-
> > gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_uint2.c | 2 +-
> > 9 files changed, 9 insertions(+), 9 deletions(-)
> >
> > diff --git a/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float.c
> > b/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float.c
> > index 881157f..6519b81 100644
> > --- a/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float.c
> > +++ b/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float.c
> > @@ -1,6 +1,6 @@
> > /* { dg-require-effective-target arm_v8_1m_mve_fp_ok } */
> > +/* { dg-add-options arm_v8_1m_mve_fp } */
> > /* { dg-skip-if "Incompatible float ABI" { *-*-* } { "-mfloat-abi=soft" }
> > {""} } */
> > -/* { dg-additional-options "-march=armv8.1-m.main+mve.fp -mfpu=auto
> > -mfloat-abi=hard -mthumb --save-temps" } */
> >
> > #include "arm_mve.h"
> >
> > diff --git
> > a/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float1.c
> > b/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float1.c
> > index 9515ed6..855e3b8 100644
> > --- a/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float1.c
> > +++ b/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float1.c
> > @@ -1,6 +1,6 @@
> > /* { dg-require-effective-target arm_v8_1m_mve_fp_ok } */
> > +/* { dg-add-options arm_v8_1m_mve_fp } */
> > /* { dg-skip-if "Incompatible float ABI" { *-*-* } { "-mfloat-abi=soft" }
> > {""} } */
> > -/* { dg-additional-options
RE: [PATCH 4/5] testsuite: [arm/mve] Use dg-add-options arm_v8_1m_mve in MVE tests
> -Original Message-
> From: Gcc-patches On Behalf Of Andre
> Vieira (lists)
> Sent: 14 April 2020 09:38
> To: Christophe Lyon
> Cc: gcc-patches@gcc.gnu.org
> Subject: Re: [PATCH 4/5] testsuite: [arm/mve] Use dg-add-options
> arm_v8_1m_mve in MVE tests
>
> On 10/04/2020 14:55, Christophe Lyon via Gcc-patches wrote:
> > Several ARM/MVE tests can be compiled even if the toolchain does not
> > support -mfloat-abi=hard (softfp is OK).
> >
> > Use dg-add-options arm_v8_1m_mve or arm_v8_1m_mve_fp instead of
> using
> > dg-additional-options.
> Hi Christophe,
>
> I think a bunch of these tests were initially meant to test hard float
> abi with vectors, especially in the MVE integer cases, this is what the
> scan-assemblers are meant to test. However, it seems to pass for
> float-abi=softfp too ... which means these tests don't really mean
> anything. I think it would be good to remove the scan-assembler tests
> for now and improve these tests later with run tests, or some smarter
> function body testing.
>
> I suggest we apply your changes and I can follow up with a patch to
> remove the scan-assemblers for now, if a maintainer agrees with me that is.
>
Sure, thanks.
Ok.
Kyrill
> Cheers,
> Andre
> > 2020-04-10 Christophe Lyon
> >
> > gcc/testsuite/
> > * gcc.target/arm/mve/intrinsics/mve_vector_float.c: Use
> > arm_v8_1m_mve_fp.
> > * gcc.target/arm/mve/intrinsics/mve_vector_float1.c: Likewise.
> > * gcc.target/arm/mve/intrinsics/mve_vector_float2.c: Likewise.
> > * gcc.target/arm/mve/intrinsics/mve_vector_int.c: Use
> > arm_v8_1m_mve.
> > * gcc.target/arm/mve/intrinsics/mve_vector_int1.c: Likewise.
> > * gcc.target/arm/mve/intrinsics/mve_vector_int2.c: Likewise.
> > * gcc.target/arm/mve/intrinsics/mve_vector_uint.c: Likewise.
> > * gcc.target/arm/mve/intrinsics/mve_vector_uint1.c: Likewise.
> > * gcc.target/arm/mve/intrinsics/mve_vector_uint2.c: Likewise.
> > ---
> > gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float.c | 2 +-
> > gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float1.c | 2 +-
> > gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float2.c | 2 +-
> > gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_int.c| 2 +-
> > gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_int1.c | 2 +-
> > gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_int2.c | 2 +-
> > gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_uint.c | 2 +-
> > gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_uint1.c | 2 +-
> > gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_uint2.c | 2 +-
> > 9 files changed, 9 insertions(+), 9 deletions(-)
> >
> > diff --git a/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float.c
> b/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float.c
> > index 881157f..6519b81 100644
> > --- a/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float.c
> > +++ b/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float.c
> > @@ -1,6 +1,6 @@
> > /* { dg-require-effective-target arm_v8_1m_mve_fp_ok } */
> > +/* { dg-add-options arm_v8_1m_mve_fp } */
> > /* { dg-skip-if "Incompatible float ABI" { *-*-* } { "-mfloat-abi=soft" }
> > {""} }
> */
> > -/* { dg-additional-options "-march=armv8.1-m.main+mve.fp -mfpu=auto -
> mfloat-abi=hard -mthumb --save-temps" } */
> >
> > #include "arm_mve.h"
> >
> > diff --git a/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float1.c
> b/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float1.c
> > index 9515ed6..855e3b8 100644
> > --- a/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float1.c
> > +++ b/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float1.c
> > @@ -1,6 +1,6 @@
> > /* { dg-require-effective-target arm_v8_1m_mve_fp_ok } */
> > +/* { dg-add-options arm_v8_1m_mve_fp } */
> > /* { dg-skip-if "Incompatible float ABI" { *-*-* } { "-mfloat-abi=soft" }
> > {""} }
> */
> > -/* { dg-additional-options "-march=armv8.1-m.main+mve.fp -mfpu=auto -
> mfloat-abi=hard -mthumb --save-temps" } */
> >
> > #include "arm_mve.h"
> >
> > diff --git a/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float2.c
> b/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float2.c
> > index 3ce8ea3..e3cf8f8 100644
> > --- a/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float2.c
> > +++ b/gcc/testsuite/gcc.target
Re: [PATCH 4/5] testsuite: [arm/mve] Use dg-add-options arm_v8_1m_mve in MVE tests
On 10/04/2020 14:55, Christophe Lyon via Gcc-patches wrote:
Several ARM/MVE tests can be compiled even if the toolchain does not
support -mfloat-abi=hard (softfp is OK).
Use dg-add-options arm_v8_1m_mve or arm_v8_1m_mve_fp instead of using
dg-additional-options.
Hi Christophe,
I think a bunch of these tests were initially meant to test hard float
abi with vectors, especially in the MVE integer cases, this is what the
scan-assemblers are meant to test. However, it seems to pass for
float-abi=softfp too ... which means these tests don't really mean
anything. I think it would be good to remove the scan-assembler tests
for now and improve these tests later with run tests, or some smarter
function body testing.
I suggest we apply your changes and I can follow up with a patch to
remove the scan-assemblers for now, if a maintainer agrees with me that is.
Cheers,
Andre
2020-04-10 Christophe Lyon
gcc/testsuite/
* gcc.target/arm/mve/intrinsics/mve_vector_float.c: Use
arm_v8_1m_mve_fp.
* gcc.target/arm/mve/intrinsics/mve_vector_float1.c: Likewise.
* gcc.target/arm/mve/intrinsics/mve_vector_float2.c: Likewise.
* gcc.target/arm/mve/intrinsics/mve_vector_int.c: Use
arm_v8_1m_mve.
* gcc.target/arm/mve/intrinsics/mve_vector_int1.c: Likewise.
* gcc.target/arm/mve/intrinsics/mve_vector_int2.c: Likewise.
* gcc.target/arm/mve/intrinsics/mve_vector_uint.c: Likewise.
* gcc.target/arm/mve/intrinsics/mve_vector_uint1.c: Likewise.
* gcc.target/arm/mve/intrinsics/mve_vector_uint2.c: Likewise.
---
gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float.c | 2 +-
gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float1.c | 2 +-
gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float2.c | 2 +-
gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_int.c| 2 +-
gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_int1.c | 2 +-
gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_int2.c | 2 +-
gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_uint.c | 2 +-
gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_uint1.c | 2 +-
gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_uint2.c | 2 +-
9 files changed, 9 insertions(+), 9 deletions(-)
diff --git a/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float.c
b/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float.c
index 881157f..6519b81 100644
--- a/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float.c
+++ b/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float.c
@@ -1,6 +1,6 @@
/* { dg-require-effective-target arm_v8_1m_mve_fp_ok } */
+/* { dg-add-options arm_v8_1m_mve_fp } */
/* { dg-skip-if "Incompatible float ABI" { *-*-* } { "-mfloat-abi=soft" }
{""} } */
-/* { dg-additional-options "-march=armv8.1-m.main+mve.fp -mfpu=auto
-mfloat-abi=hard -mthumb --save-temps" } */
#include "arm_mve.h"
diff --git a/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float1.c b/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float1.c
index 9515ed6..855e3b8 100644
--- a/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float1.c
+++ b/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float1.c
@@ -1,6 +1,6 @@
/* { dg-require-effective-target arm_v8_1m_mve_fp_ok } */
+/* { dg-add-options arm_v8_1m_mve_fp } */
/* { dg-skip-if "Incompatible float ABI" { *-*-* } { "-mfloat-abi=soft" }
{""} } */
-/* { dg-additional-options "-march=armv8.1-m.main+mve.fp -mfpu=auto
-mfloat-abi=hard -mthumb --save-temps" } */
#include "arm_mve.h"
diff --git a/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float2.c b/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float2.c
index 3ce8ea3..e3cf8f8 100644
--- a/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float2.c
+++ b/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_float2.c
@@ -1,6 +1,6 @@
/* { dg-require-effective-target arm_v8_1m_mve_fp_ok } */
+/* { dg-add-options arm_v8_1m_mve_fp } */
/* { dg-skip-if "Incompatible float ABI" { *-*-* } { "-mfloat-abi=soft" }
{""} } */
-/* { dg-additional-options "-march=armv8.1-m.main+mve.fp -mfpu=auto
-mfloat-abi=hard -mthumb --save-temps" } */
#include "arm_mve.h"
diff --git a/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_int.c b/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_int.c
index dab0705..e70cbc1 100644
--- a/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_int.c
+++ b/gcc/testsuite/gcc.target/arm/mve/intrinsics/mve_vector_int.c
@@ -1,6 +1,6 @@
/* { dg-require-effective-target arm_v8_1m_mve_ok } */
+/* { dg-add-options arm_v8_1m_mve } */
/* { dg-skip-if "Incompatible float ABI" { *-*-* } { "-mfloat-abi=soft" }
{""} } */
-/* { dg-additional-options "-march=armv8.1-m.main+mve -mfpu=auto -mfloat-abi=hard
-mthumb --save-temps" } */
#include "arm_mve.h"
diff --git
