https://gcc.gnu.org/bugzilla/show_bug.cgi?id=126940
Bug ID: 126940
Summary: [17 Regression] testsuite fails for
libgomp.oacc-fortran/host_data-5.F90
Product: gcc
Version: 17.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: fortran
Assignee: unassigned at gcc dot gnu.org
Reporter: burnus at gcc dot gnu.org
CC: tschwinge at gcc dot gnu.org
Target Milestone: ---
Between last Friday→Saturday (~midnight UTC) and last night (Tue→Wed, ~4h
before midnight UTC) the following new fails started showing up:
FAIL: libgomp.oacc-fortran/host_data-5.F90 -DACC_DEVICE_TYPE_radeon=1
-DACC_MEM_SHARED=0 -foffload=amdgcn-amdhsa -O0 at line 46 (test for
warnings, line 41)
FAIL: libgomp.oacc-fortran/host_data-5.F90 -DACC_DEVICE_TYPE_radeon=1
-DACC_MEM_SHARED=0 -foffload=amdgcn-amdhsa -O0 (test for excess errors)
(Likewise for the other optimization levels.)
The excess errors are of the form:
libgomp.oacc-fortran/host_data-5.F90:41:38: note: variable 'iftmp.75' declared
in block isn't candidate for adjusting OpenACC privatization level: not
addressable
libgomp.oacc-fortran/host_data-5.F90:41:38: note: variable 'iftmp.77' declared
in block isn't candidate for adjusting OpenACC privatization level: not
addressable
libgomp.oacc-fortran/host_data-5.F90:41:38: note: variable 'iftmp.86' declared
in block isn't candidate for adjusting OpenACC privatization level: not
addressable
(and a bunch more of the same kind)
Line 41 is:
!$acc data copyin(host_p, host_parr)
and the failing line 46 is
! { dg-note {variable 'parm\.[0-9]+' declared in block isn't candidate for
adjusting OpenACC privatization level: artificial} "" { target { !
openacc_host_selected } } .-5 }
