An embarrassing bug in the equal(String,String) private implementation
method. The Strings compare equal when there is a % in one string, a
different char in the same place in the other string, but the two chars
following are the same.
http://cr.openjdk.java.net/~chegar/7041800/webrev.00/web
Chris Hegarty wrote:
An embarrassing bug in the equal(String,String) private implementation
method. The Strings compare equal when there is a % in one string, a
different char in the same place in the other string, but the two
chars following are the same.
http://cr.openjdk.java.net/~chegar
On 01/09/11 09:36, Chris Hegarty wrote:
An embarrassing bug in the equal(String,String) private implementation
method. The Strings compare equal when there is a % in one string, a
different char in the same place in the other string, but the two
chars following are the same.
http://cr.openj
On 09/ 1/11 11:40 AM, Michael McMahon wrote:
On 01/09/11 09:36, Chris Hegarty wrote:
...
Change looks fine. I just wonder if the test should go into the existing
mega-test, rather than creating a second file just for this bug?
Agreed, I added it to the mega-test.
Updated webrev:
http://cr.
Chris Hegarty wrote:
Agreed, I added it to the mega-test.
Updated webrev:
http://cr.openjdk.java.net/~chegar/7041800/webrev.01/webrev/
Thanks for putting it in the unit test. Looks fine now.
-Alan.
Changeset: fcb33500b325
Author:chegar
Date: 2011-09-01 13:53 +0100
URL: http://hg.openjdk.java.net/jdk8/tl/jdk/rev/fcb33500b325
7041800: URI.equals may incorrectly return true with escaped octets
Reviewed-by: alanb, michaelm
! src/share/classes/java/net/URI.java
! test/java/net/URI
Hi,
test/java/net/Inet6Address/B6558853.java tests if the address obtained from
getHostAddress() on connections
using IPv6 link-local addresses contains the zone id. For
Inet6Address.getHostAddress() to return the zone id,
Inet6Address.scope_id_set needs to be set to true in addition to set
Thomas,
Thanks for pointing this out. We overlooked this in the recent change in
this
area. One thing though, in the second change to DatagramSocket we can't
just check for isBound() since the socket might legitimately be unbound
(bindaddr is null). All I can think is that we catch the excepti
Maybe I posted a bad patch, but my intent was to do the try-finally after
checking for a non-null bind address.
public DatagramSocket(SocketAddress bindaddr) throws SocketException {
// create a datagram socket.
createImpl();
if (bindaddr != null) {
try {
Right. That would work assuming createImpl() doesn't leave the socket open
if it throws an exception, which seems to be the case.
On 01/09/11 18:05, Salter, Thomas A wrote:
Maybe I posted a bad patch, but my intent was to do the try-finally after
checking for a non-null bind address.
publ
Changeset: a45d78d26450
Author:jjh
Date: 2011-09-01 14:35 -0700
URL: http://hg.openjdk.java.net/jdk8/tl/langtools/rev/a45d78d26450
7086071: tools/javac/7079713/TestCircularClassfile.java fails on windows
Summary: delete file before renaming another file to it
Reviewed-by: jjg
! tes
Hi guys,
In some linuxes, when you bind on a loopback network interface, sendto
will be fail and errono will be set to EHOSTUNREACH.
In this situation, EHOSTUNREACH maybe need treat as INVAL: return false,
not throw an exception.
Patch is attached.
Does anyone interested in this issue?
--
On 09/02/2011 12:50 PM, Charles Lee wrote:
Hi guys,
In some linuxes, when you bind on a loopback network interface, sendto
will be fail and errono will be set to EHOSTUNREACH.
In this situation, EHOSTUNREACH maybe need treat as INVAL: return
false, not throw an exception.
Patch is attached.
Changeset: ffada2ce20e5
Author:darcy
Date: 2011-09-01 23:00 -0700
URL: http://hg.openjdk.java.net/jdk8/tl/jdk/rev/ffada2ce20e5
7082971: More performance tuning of BigDecimal and other java.math classes
Reviewed-by: darcy
Contributed-by: sergey.kukse...@oracle.com
! src/share/classe
14 matches
Mail list logo
