On Fri, Dec 09, 2016 at 10:44:17AM -0700, Martin Sebor wrote: > On 12/09/2016 06:26 AM, Andreas Schwab wrote: > >FAIL: gcc.dg/attr-alloc_size-3.c (test for warnings, line 445) > >FAIL: gcc.dg/attr-alloc_size-3.c (test for excess errors) > >Excess errors: > >/daten/aranym/gcc/gcc-20161209/gcc/testsuite/gcc.dg/attr-alloc_size-3.c:244:3: > > warning: product '65535 * 65535' of arguments 1 and 2 exceeds maximum > >object size 2147483647 [-Walloc-size-larger-than=] > >/daten/aranym/gcc/gcc-20161209/gcc/testsuite/gcc.dg/attr-alloc_size-3.c:298:3: > > warning: argument 1 value '4294967294' exceeds maximum object size > >2147483647 [-Walloc-size-larger-than=] > >/daten/aranym/gcc/gcc-20161209/gcc/testsuite/gcc.dg/attr-alloc_size-3.c:299:3: > > warning: argument 1 value '4294967295' exceeds maximum object size > >2147483647 [-Walloc-size-larger-than=] > > I assume this was on an ILP32 target like i686. It should be fixed > by r243497 along with the failures on arm* targets that Christophe > pointed out earlier. Please let me know if that doesn't clear it > up for you (and what target).
on arm-unknown-linux-gnueabi (and likely other ILP32) targets. doesn't look like valid ChangeLog entry (r243497). Also, especially when adding new tests related to gimple-ssa-sprintf.c that keep failing on non-x86_64 a lot, it would be nice if you at least tested them through make check-gcc RUNTESTFLAGS="--target_board=unix\{-m32,-m64\} dg.exp=name_of_test*" first. Jakub