optimize optimization flags

Move the optimization pragma before the #includes so that we don't
prevent functions from inlining because they were compiled with
different optimization flags than the caller.

Eliminate the flags for loop unrolling, tree transformations,
etc. because they either had no effect or were slightly detrimental.

Yields a speedup of 7% to 18% with four threads, depending on which
version of color science (V3-V7) is used.

src/iop/filmicrgb.c

@@ -15,6 +15,15 @@
    You should have received a copy of the GNU General Public License
    along with darktable.  If not, see <http://www.gnu.org/licenses/>.
 */
+
+/** Note :
+ * we use finite-math-only and fast-math because divisions by zero are manually avoided in the code
+ * fp-contract=fast enables hardware-accelerated Fused Multiply-Add
+**/
+#if defined(__GNUC__)
+#pragma GCC optimize("finite-math-only", "fp-contract=fast", "fast-math", "no-math-errno")
+#endif
+
 #ifdef HAVE_CONFIG_H
 #include "config.h"
 #endif
@@ -93,19 +102,6 @@ DT_MODULE_INTROSPECTION(6, dt_iop_filmicrgb_params_t)
  * */
 
 
-/** Note :
- * we use finite-math-only and fast-math because divisions by zero are manually avoided in the code
- * fp-contract=fast enables hardware-accelerated Fused Multiply-Add
- * the rest is loop reorganization and vectorization optimization
- **/
-#if defined(__GNUC__)
-#pragma GCC optimize("unroll-loops", "tree-loop-if-convert", "tree-loop-distribution", "no-strict-aliasing",      \
-                     "loop-interchange", "loop-nest-optimize", "tree-loop-im", "unswitch-loops",                  \
-                     "tree-loop-ivcanon", "ira-loop-pressure", "split-ivs-in-unroller",                           \
-                     "variable-expansion-in-unroller", "split-loops", "ivopts", "predictive-commoning",           \
-                     "tree-loop-linear", "loop-block", "loop-strip-mine", "finite-math-only", "fp-contract=fast", \
-                     "fast-math", "no-math-errno")
-#endif
 
 
 typedef enum dt_iop_filmicrgb_methods_type_t