SUGGESTED FIX
The problem was that _is_default_XXX was not being correctly
set in the argument processing. This caused the CL option
settings for these parameters to be ignored when we used
the _is_default_XX test.
------- arguments.cpp -------
1560a1561,1567
> } else if (match_option(option, "-XX:MaxNewSize", &tail)) {
> _is_default_max_new_size = false;
> // Back up so we can use normal argument processing
> match_option(option, "-XX:", &tail);
> if (!process_argument(tail, args->ignoreUnrecognized)) {
> return JNI_EINVAL;
> }
1571a1579,1592
> if (!process_argument(tail, args->ignoreUnrecognized)) {
> return JNI_EINVAL;
> }
> } else if (match_option(option, "-XX:MaxTenuringThreshold", &tail)) {
> _is_default_max_tenuring_threshold = false;
> // Back up so we can use normal argument processing
> match_option(option, "-XX:", &tail);
> if (!process_argument(tail, args->ignoreUnrecognized)) {
> return JNI_EINVAL;
> }
> } else if (match_option(option, "-XX:SurvivorRatio", &tail)) {
> _is_default_survivor_ratio = false;
> // Back up so we can use normal argument processing
> match_option(option, "-XX:", &tail);
###@###.### 2003-06-06: fix putback to [Tiger's] gc_baseline.
|
Java