On 11/11/25 06:58, Andreas Schwab wrote:
On Nov 11 2025, Jan Dubiec wrote:
@@ -54,8 +54,8 @@ set_board_info gdb,noresults 1
set_board_info gcc,stack_size 4096
set_board_info gcc,text_size 60000
-# There's no long long support on this target
-set_board_info no_long_long 1
+# There is long long support on this target
+set_board_info no_long_long 0
The gcc and gdb testsuites only check for the existence of the
no_long_long setting, not its particular value.
I had suspected as much, but I am not familiar with the intricacies of
the toolchain testsuites; this is why I asked about this.
Do I correctly understand that {set_board_info no_long_long X} should be
omitted entirely for the GCC testsuite to exercise long long support?
(Obviously, those tests will fail if GCC does not actually support long
long on the target.)
-- Jacob
