This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

[RFA:] No timeout for ext/pb_ds/regression/hash_data_map_rand.cc


For background, see
<URL:http://gcc.gnu.org/ml/gcc-patches/2007-02/msg00450.html>.
This test is one of those that are close to the 5 min timeout
for "a modern system", causing this to be flagged as a
regression whenever there's any load on the system.  As with
testsuite/23_containers/set/modifiers/16728.cc there didn't seem
to be anything magical about the numbers.  Reducing the number
of iterations seemed better than reducing both, and "as usual"
by a factor of 1000.

Ok to commit?

libstdc++-v3:
	* testsuite/ext/pb_ds/regression/hash_data_map_rand.cc: Use only 5
        iterations for simulator targets.

Index: testsuite/ext/pb_ds/regression/hash_data_map_rand.cc
===================================================================
--- testsuite/ext/pb_ds/regression/hash_data_map_rand.cc	(revision 121581)
+++ testsuite/ext/pb_ds/regression/hash_data_map_rand.cc	(working copy)
@@ -48,13 +48,22 @@
 #include <regression/rand/assoc/rand_regression_test.hpp>
 #include <regression/assoc/common_type.hpp>
 
+// This can take long on simulators, timing out the test.
+// { dg-options "-DPB_DS_REGRESSION -DITERATIONS=5" { target simulator } }
+#ifndef ITERATIONS
+#define ITERATIONS 5000
+#endif
+#ifndef KEYS
+#define KEYS 10000
+#endif
+
 int
 main(int argc, char* a_p_argv[])
 {
   using namespace pb_ds::test;
   typedef hash_map_tl_t map_tl_t;
 
-  return rand_regression_test(5000, 10000,
+  return rand_regression_test(ITERATIONS, KEYS,
 			      "hash_data_map_rand_regression_test",
 			      map_tl_t());
 }

brgds, H-P


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]