Hi Jakub,
I can confirm that this fixes the problem for arm-none-eabi.
Kind regards,
Torbjörn
On 2026-08-04 17:34, Jakub Jelinek wrote:
On Tue, Aug 04, 2026 at 02:13:19PM +0200, Torbjorn SVENSSON wrote:
This causes regression on most (all?) arm-none-eabi targets.
/build/gcc_src/gcc/testsuite/gcc.dg/pr126464.c:56:3: warning: floating constant
exceeds range of 'long double' [-Woverflow]
This patch guards it with preprocessor tests whether such floating point
constants are representable.
Tested on x86_64-linux, ok for trunk?
2026-08-04 Jakub Jelinek <[email protected]>
PR tree-optimization/126464
* gcc.dg/pr126464.c: Guard uses of 1e300 with
__DBL_MAX_10_EXP__ >= 301 and uses of 1e4000L with
__LDBL_MAX_10_EXP__ >= 4001.
--- gcc/testsuite/gcc.dg/pr126464.c.jj 2026-07-30 09:56:33.400729346 +0200
+++ gcc/testsuite/gcc.dg/pr126464.c 2026-08-04 17:31:18.458149999 +0200
@@ -47,15 +47,19 @@ qux (long double x)
int
main ()
{
+#if __DBL_MAX_10_EXP__ >= 301
if (!__builtin_isinf ((double) 1e300)
&& __builtin_isinf ((float) 1e300)
&& (foo (-1e300) != -5e299
|| baz (1e300) != 5e299))
__builtin_abort ();
+#endif
+#if __LDBL_MAX_10_EXP__ >= 4001
if (!__builtin_isinf ((long double) 1e4000L)
&& __builtin_isinf ((double) 1e4000L)
&& (bar (1e4000L) != 5e3999L
|| qux (-1e4000L) != -5e3999L))
__builtin_abort ();
+#endif
}
Jakub
