This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
Re: java/3803: buggy code generation
- To: gcc-gnats at gcc dot gnu dot org, turms at web dot de
- Subject: Re: java/3803: buggy code generation
- From: Alexandre Petit-Bianco <apbianco at cygnus dot com>
- Date: Wed, 15 Aug 2001 11:23:47 -0700
- Cc: gcc-prs at gcc dot gnu dot org, java-prs at gcc dot gnu dot org, gcc-bugs at gcc dot gnu dot org
- Reply-to: apbianco at cygnus dot com
> Happens of course only if I compile the java-files. Not with the
> class files.
OK. It seems to be related to java/3096 and java/3965:
http://gcc.gnu.org/ml/java-prs/2001-q2/msg00070.html
So I rewrote the only loop that uses a pre-incremented array access
and the test started to work:
RC5 Java Demo (gregh@lightspeed.net)
Keys: 256 keys
Time: 1 ms
Rate: 256000 keys/sec
Keys: 512 keys
Time: 4 ms
Rate: 128000 keys/sec
Keys: 768 keys
Time: 6 ms
Rate: 128000 keys/sec
java/3096 and java/3965 are next on my list of things to do. I'll keep
you posted.
./A
--- RC5test.java~ Wed Aug 15 10:57:52 2001
+++ RC5test.java Wed Aug 15 11:07:58 2001
@@ -36,7 +36,7 @@ class RC5test implements Runnable {
//break;
}
int i = 0;
- while (++key[i] == 0) {
+ for (key[i]++; key[i] == 0; key[i]++) {
i++;
if (i >= 1) {
long now = System.currentTimeMillis();