configure: add GCC option -fexcess-precision=standard

This option forces GCC to follow the ISO C standards concerning
rounding of intermediate FP results.  It avoids some FP issues
with the x86 32 bits ports of OCaml, which can run into
excess precision problems due to the x87 FP unit.

Closes: #7917

configure

@@ -12506,7 +12506,8 @@ case $host in #(
   gcc-*) :
     internal_cflags="-Wno-unused $gcc_warnings"
         # TODO: see whether the code can be fixed to avoid -Wno-unused
-        common_cflags="-O2 -fno-strict-aliasing -fwrapv -mms-bitfields"
+        common_cflags="-O2 -fno-strict-aliasing -fwrapv \
+-fexcess-precision=standard -mms-bitfields"
         internal_cppflags='-DUNICODE -D_UNICODE'
         internal_cppflags="$internal_cppflags -DWINDOWS_UNICODE="
         internal_cppflags="${internal_cppflags}\$(WINDOWS_UNICODE)" ;; #(
@@ -12534,12 +12535,18 @@ $as_echo "$as_me: WARNING: This version of GCC is rather old.
 $as_echo "$as_me: WARNING: Consider using GCC version 4.2 or above." >&2;};
       common_cflags="-std=gnu99 -O";
       internal_cflags="$gcc_warnings" ;; #(
-  gcc-4-*) :
-    common_cflags="-std=gnu99 -O2 -fno-strict-aliasing -fwrapv \
+  gcc-4-234) :
+    # No -fexcess-precision option before GCC 4.5
+      common_cflags="-std=gnu99 -O2 -fno-strict-aliasing -fwrapv \
 -fno-builtin-memcmp";
       internal_cflags="$gcc_warnings" ;; #(
+  gcc-4-*) :
+    common_cflags="-std=gnu99 -O2 -fno-strict-aliasing -fwrapv \
+-fno-builtin-memcmp -fexcess-precision=standard";
+      internal_cflags="$gcc_warnings" ;; #(
   gcc-*) :
-    common_cflags="-O2 -fno-strict-aliasing -fwrapv";
+    common_cflags="-O2 -fno-strict-aliasing -fwrapv \
+-fexcess-precision=standard";
       internal_cflags="$gcc_warnings -fno-common" ;; #(
   msvc-*) :
     common_cflags="-nologo -O2 -Gy- -MD"

configure.ac

@@ -561,7 +561,8 @@ AS_CASE([$host],
       [gcc-*],
         [internal_cflags="-Wno-unused $gcc_warnings"
         # TODO: see whether the code can be fixed to avoid -Wno-unused
-        common_cflags="-O2 -fno-strict-aliasing -fwrapv -mms-bitfields"
+        common_cflags="-O2 -fno-strict-aliasing -fwrapv \
+-fexcess-precision=standard -mms-bitfields"
         internal_cppflags='-DUNICODE -D_UNICODE'
         internal_cppflags="$internal_cppflags -DWINDOWS_UNICODE="
         internal_cppflags="${internal_cppflags}\$(WINDOWS_UNICODE)"],
@@ -583,12 +584,18 @@ AS_CASE([$host],
       AC_MSG_WARN([Consider using GCC version 4.2 or above.]);
       common_cflags="-std=gnu99 -O";
       internal_cflags="$gcc_warnings"],
-    [gcc-4-*],
+    [gcc-4-[234]],
+      # No -fexcess-precision option before GCC 4.5
       [common_cflags="-std=gnu99 -O2 -fno-strict-aliasing -fwrapv \
 -fno-builtin-memcmp";
       internal_cflags="$gcc_warnings"],
+    [gcc-4-*],
+      [common_cflags="-std=gnu99 -O2 -fno-strict-aliasing -fwrapv \
+-fno-builtin-memcmp -fexcess-precision=standard";
+      internal_cflags="$gcc_warnings"],
     [gcc-*],
-      [common_cflags="-O2 -fno-strict-aliasing -fwrapv";
+      [common_cflags="-O2 -fno-strict-aliasing -fwrapv \
+-fexcess-precision=standard";
       internal_cflags="$gcc_warnings -fno-common"],
     [msvc-*],
       [common_cflags="-nologo -O2 -Gy- -MD"