| Message ID | 20241108114848.2650646-1-torbjorn.svensson@foss.st.com |
|---|---|
| State | New |
| Headers | show |
| Series | [v2] testsuite: arm: Use effective-target for pr84556.cc test | expand |
On 08/11/2024 11:48, Torbjörn SVENSSON wrote: > Changes since v1: > > - Clarified the commit message to include where the descision is taken > and why it's a bad idea to use "dg-do run" in a test case. > Note: This does not only fix it for arm-none-eabi. I see the same > kind of construct used by for example sparc. > > Sorry for the confusion Richard, I hope it's more clear why this is > needed now. :) > > Ok for trunk and releases/gcc-14? > > -- > > Using "dg-do run" with a selector overrides the default selector set by > vect.exp that picks between "dg-do run" and "dg-do compile" based on the > target's support for simd operations for Arm targets. > The actual selection of default operation is performed in > check_vect_support_and_set_flags. > > gcc/testsuite/ChangeLog: > > * g++.dg/vect/pr84556.cc: Change from "dg-do run" with selector > to instead use dg-require-effective-target with the same > selector. > > Signed-off-by: Torbjörn SVENSSON <torbjorn.svensson@foss.st.com> OK R.
On 2024-11-08 12:57, Richard Earnshaw (lists) wrote: > On 08/11/2024 11:48, Torbjörn SVENSSON wrote: >> Changes since v1: >> >> - Clarified the commit message to include where the descision is taken >> and why it's a bad idea to use "dg-do run" in a test case. >> Note: This does not only fix it for arm-none-eabi. I see the same >> kind of construct used by for example sparc. >> >> Sorry for the confusion Richard, I hope it's more clear why this is >> needed now. :) >> >> Ok for trunk and releases/gcc-14? >> >> -- >> >> Using "dg-do run" with a selector overrides the default selector set by >> vect.exp that picks between "dg-do run" and "dg-do compile" based on the >> target's support for simd operations for Arm targets. >> The actual selection of default operation is performed in >> check_vect_support_and_set_flags. >> >> gcc/testsuite/ChangeLog: >> >> * g++.dg/vect/pr84556.cc: Change from "dg-do run" with selector >> to instead use dg-require-effective-target with the same >> selector. >> >> Signed-off-by: Torbjörn SVENSSON <torbjorn.svensson@foss.st.com> > > OK Pushed as r15-5039-g85c3d944800 and r14.2.0-373-g8cf9b265704. Kind regards, Torbjörn > > R. >
diff --git a/gcc/testsuite/g++.dg/vect/pr84556.cc b/gcc/testsuite/g++.dg/vect/pr84556.cc index 6b1c9cec515..c7e331628a8 100644 --- a/gcc/testsuite/g++.dg/vect/pr84556.cc +++ b/gcc/testsuite/g++.dg/vect/pr84556.cc @@ -1,5 +1,5 @@ // PR c++/84556 -// { dg-do run { target c++11 } } +// { dg-require-effective-target c++11 } // { dg-additional-options "-O2 -fopenmp-simd" } // { dg-additional-options "-mavx" { target avx_runtime } }
Changes since v1: - Clarified the commit message to include where the descision is taken and why it's a bad idea to use "dg-do run" in a test case. Note: This does not only fix it for arm-none-eabi. I see the same kind of construct used by for example sparc. Sorry for the confusion Richard, I hope it's more clear why this is needed now. :) Ok for trunk and releases/gcc-14? -- Using "dg-do run" with a selector overrides the default selector set by vect.exp that picks between "dg-do run" and "dg-do compile" based on the target's support for simd operations for Arm targets. The actual selection of default operation is performed in check_vect_support_and_set_flags. gcc/testsuite/ChangeLog: * g++.dg/vect/pr84556.cc: Change from "dg-do run" with selector to instead use dg-require-effective-target with the same selector. Signed-off-by: Torbjörn SVENSSON <torbjorn.svensson@foss.st.com> --- gcc/testsuite/g++.dg/vect/pr84556.cc | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)