Hi Stuart,
On 5/14/15 3:44 AM, Stuart Marks wrote:
Hi all,
Please review this change to optimize a test. Basically the test did
string formatting for every assertion, but the string was thrown away
if the assertion passed -- the most common case. The change is to do
the string formatting only when an assertion fails and information
needs to be printed out.
Thanks to Andrey Zakharov for discovering and investigating this.
Bug report:
https://bugs.openjdk.java.net/browse/JDK-8078463
Webrev:
http://cr.openjdk.java.net/~smarks/reviews/8078463/webrev.0/
On my (new, fast) laptop, with JVM options -Xcomp -XX:+DeoptimizeALot
-client, the unmodified test takes about 21.4 seconds to run. The
modified test takes only 12.3 seconds.
Note that I have added several overloads of check() with different
arguments. I tried an alternative, which is a varargs version of check():
static void check(boolean cond, String fmt, Object... args) {
if (cond) {
pass();
} else {
fail(String.format(fmt, args));
}
}
This of course is much simpler code, but it took 14.2 seconds, about
15% slower than the proposed version. Is the simpler code worth the
slowdown? I could go either way.
I'm curious: have you tried with using a lambda instead? changing:
394 static void check(String desc, boolean cond) {
395 if (cond) {
396 pass();
397 } else {
398 fail(desc);
399 }
400 }
into
static void check(Supplier<String> descSupplier, boolean cond) {
if (cond) {
pass();
} else {
fail(descSupplier.get())
}
}
I wonder how the performance would compare...
best regards,
-- daniel
Thanks.
s'marks
