https://gcc.gnu.org/bugzilla/show_bug.cgi?id=96428
--- Comment #8 from Tobias Burnus <burnus at gcc dot gnu.org> --- (In reply to Thomas Schwinge from comment #7) > These attachment 48988 [details] testcases got included in commit > 344f09a756ebd50510cc1eb3db111fd61c527702. (part of this PR, commit g:344f09a756ebd50510cc1eb3db111fd61c527702 ) > I don't understand ... > module m2 > real*8 :: mysum > !$acc declare device_resident(mysum) ... That's the disadvantage of turning a full program, which is ICEing during compilation, into a reduced testcase: it can easily happen that the reduced program is no longer sensible or valid. > So, is this really testing what it means to be testing? At least partially: yes – the compiler is no longer ICEing. But it probably should be fixed to be valid code and ... > Should the testcase get some 'target openacc_nvidia_accel_selected' > 'scan-offload-rtl-dump' added to make sure that we're actually generating > the expected PTX instructions? > Also, the testcase files should be renamed 'libgomp.oacc-fortran/pr96428-*' > to match the PR ID. ... does not harm, either. — Are you volunteering to do this?
