Hey folks, with some help from Claes Redestad I've filed an issue with
analysis and reproduction of the legendary "zip file closed" error.
https://bugs.openjdk.java.net/browse/JDK-8246714
It turns out that URLClassLoader is just plain broken when used to
access jar files (which means this bug
> On Jun 5, 2020, at 7:57 PM, Magnus Ihse Bursie
> wrote:
>
> On 2020-06-05 09:52, Kim Barrett wrote:
>> [Changes are only to the build system, but since the changes have jdk-wide
>> effect I’ve cc’d what I think are the relevant dev lists.]
>>
>> This change is part of JEP 347; the intent is
> On Jun 5, 2020, at 7:57 PM, Magnus Ihse Bursie
> wrote:
>
> On 2020-06-05 13:59, Jim Laskey wrote:
>> I know there was a discussion about this elsewhere but I would like to take
>> the opportunity to correct this now
>>
>> make//autoconf/flags-cflags.m4:241
>>
>> […]
>> MacOSX has been
On 2020-06-05 09:52, Kim Barrett wrote:
[Changes are only to the build system, but since the changes have jdk-wide
effect I’ve cc’d what I think are the relevant dev lists.]
This change is part of JEP 347; the intent is to target JDK 16.
Please review this change to the building of C++ code in
On 2020-06-05 13:59, Jim Laskey wrote:
I know there was a discussion about this elsewhere but I would like to take the
opportunity to correct this now
make//autoconf/flags-cflags.m4:241
elif test "x$TOOLCHAIN_TYPE" = xclang; then
if test "x$OPENJDK_TARGET_OS" = xmacosx; then
#
In SigningBase.java:
---
KEYCHAIN = (value == null) ? "jpackagerTest.keychain" : value;
---
Is "jpackagerTest.keychain" some legacy string? Should not it be
"jpackageTest.keychain"?
- Alexey
On 6/5/2020 4:23 PM, Andy Herrick wrote:
Please review fix to issue [1] at [2].
The fix enables
I’m fixing the two in java.base. The other two are in different modules and
would require changes to export. So you can file on those.
> On Jun 5, 2020, at 5:36 PM, Stuart Marks wrote:
>
>
>
>> On 6/3/20 10:36 AM, Stuart Marks wrote:
>> 3) Integer wraparound/overflow during size
On 6/3/20 10:36 AM, Stuart Marks wrote:
3) Integer wraparound/overflow during size computation. AS.newLength generates
this:
OutOfMemoryError: Required array length too large
(3) is the only case generated by the library. In fact, AS.hugeLength() has
oldLength and minGrowth
Please review fix to issue [1] at [2].
The fix enables running jpackage signing tests on macos with user
supplied certificates.
[1] https://bugs.openjdk.java.net/browse/JDK-8246706
[2] http://cr.openjdk.java.net/~herrick/8246706/webrev.01/index.html
/Andy
Thank you.
> On Jun 5, 2020, at 3:21 PM, Igor Ignatyev wrote:
>
> LGTM
>
> -- Igor
>
>> On Jun 5, 2020, at 11:19 AM, Jim Laskey wrote:
>>
>> Updated diff
>>
>>
>> diff -r 71ec718a0bd0 test/jdk/ProblemList.txt
>> --- a/test/jdk/ProblemList.txt Fri Jun 05 11:46:58 2020 -0300
>> +++
LGTM
-- Igor
> On Jun 5, 2020, at 11:19 AM, Jim Laskey wrote:
>
> Updated diff
>
>
> diff -r 71ec718a0bd0 test/jdk/ProblemList.txt
> --- a/test/jdk/ProblemList.txtFri Jun 05 11:46:58 2020 -0300
> +++ b/test/jdk/ProblemList.txtFri Jun 05 15:18:36 2020 -0300
> @@ -883,7 +883,6
Updated diff
diff -r 71ec718a0bd0 test/jdk/ProblemList.txt
--- a/test/jdk/ProblemList.txt Fri Jun 05 11:46:58 2020 -0300
+++ b/test/jdk/ProblemList.txt Fri Jun 05 15:18:36 2020 -0300
@@ -883,7 +883,6 @@
# jdk_util
java/util/ServiceLoader/ReloadTest.java 8242935
Thank you.
> On Jun 5, 2020, at 2:51 PM, Igor Ignatyev wrote:
>
> Hi Jim,
>
> I'd also recommend you to add `@requires os.maxMemory > 4G` so the tests
> don't get executed on hosts which don't have enough memory for requested heap
> size.
>
> -- Igor
>
>> On Jun 5, 2020, at 10:46 AM, Jim
Hi Jim,
I'd also recommend you to add `@requires os.maxMemory > 4G` so the tests don't
get executed on hosts which don't have enough memory for requested heap size.
-- Igor
> On Jun 5, 2020, at 10:46 AM, Jim Laskey wrote:
>
> Followup: can I get a quick review for the following. Both tests
Followup: can I get a quick review for the following. Both tests were sensitive
to heap size.
jbs: https://bugs.openjdk.java.net/browse/JDK-8246697
diff -r 71ec718a0bd0 test/jdk/ProblemList.txt
--- a/test/jdk/ProblemList.txt Fri Jun 05 11:46:58 2020 -0300
+++ b/test/jdk/ProblemList.txt Fri
Together these two changes might be a lot this late in the cycle, but it
does make everything cleaner, and passes all test I have tried to run
with them.
I approve.
/Andy
On 6/4/2020 4:46 PM, Alexey Semenyuk wrote:
Please review fix [2] for jpackage bug [1].
Move duplicated functionality
Hi Alexey,
On 05/06/2020 17:33, Alexey Bakhtin wrote:
Hi Daniel,
Thank you for review
Yes, I can move TlsChannelBinding class into the com.sun.jndi.ldap.sasl package
and LdapClient related changes into the LdapSasl.saslBind method.
Also, you are right with exceptions. I will rename them to
ObjectInputStream.resolveClass(ObjectStreamClass) is specified to find "the
first class loader on the current thread's stack (starting from the currently
executing method) that is neither the platform class loader nor its ancestor."
Finding that class loader is done with a virtual frame walk
On 6/5/20 6:42 AM, Christoph Dreis wrote:
Hi,
forgive me if this was discussed already, but I found that the implementation
of hidden classes causes a change in behavior for Class.getCanonicalName().
A hidden class has no canonical name [1] and it can't be nominally
referenced (like an
Hi Daniel,
Thank you for review
Yes, I can move TlsChannelBinding class into the com.sun.jndi.ldap.sasl package
and LdapClient related changes into the LdapSasl.saslBind method.
Also, you are right with exceptions. I will rename them to the NamingException.
However, I’d like to parse TLS
Extra points would be to substitute the resulting snippet for the actual
implementation to see if it compiles and satisfies the tests. For example, like
this:
default boolean replace(K key, V oldValue, V newValue) {
Map map = this; // synthetic line
if (map.containsKey(key) &&
Looks good to me
On Fri, Jun 5, 2020 at 8:20 AM Conor Cleary wrote:
>
> Hi,
>
> Could someone please take a look at my webrev for JDK-8243655
> 'Map.replace javadoc code snippet typo'?
>
> This fix addresses a typo in the javadoc for Map.replace(K, V, V) in the
> default implementation code. The
+1
> On Jun 5, 2020, at 12:20 PM, Conor Cleary wrote:
>
> Hi,
>
> Could someone please take a look at my webrev for JDK-8243655 'Map.replace
> javadoc code snippet typo'?
>
> This fix addresses a typo in the javadoc for Map.replace(K, V, V) in the
> default implementation code. The
Hi,
Could someone please take a look at my webrev for JDK-8243655
'Map.replace javadoc code snippet typo'?
This fix addresses a typo in the javadoc for Map.replace(K, V, V) in the
default implementation code. The parameter 'value' in the conditional
block is changed to 'oldValue' to reflect
Hi Alexey,
Could we move the new code from LdapClient.java and LdapCtxt.java
into the com.sun.jndi.ldap.sasl package, and possibly delay
its execution until the com.sun.jndi.ldap.sasl.LdapSasl.saslBind
method is called?
The new TlsChannelBinding class should also be preferably moved
to
Hello Max,
Thank you a lot for review.
Could you check the new version of the patch :
http://cr.openjdk.java.net/~abakhtin/8245527/webrev.v4/
I’ve made the following changes:
- TlsChannelBinding class is moved to java.naming module.
java.security.sasl module is not affected any more
- I pass
Right. The request and the content is right the subject of the e-mail was
wrong. Sorry about that.
> On Jun 5, 2020, at 11:48 AM, Igor Ignatyev wrote:
>
> Hi Jim,
>
> you should use another bug id, one which you are going to use to fix the
> problem, in the problem list.
>
> -- Igor
>
>>
Hi Jim,
you should use another bug id, one which you are going to use to fix the
problem, in the problem list.
-- Igor
> On Jun 5, 2020, at 7:41 AM, Jim Laskey wrote:
>
> Can I get a quick review for the following. Getting some unexpected OOM from
> test. Pulling until I figure it out.
>
>
Looks good.
-Sundar
On 05/06/20 8:11 pm, Jim Laskey wrote:
Can I get a quick review for the following. Getting some unexpected OOM from
test. Pulling until I figure it out.
jbs: https://bugs.openjdk.java.net/browse/JDK-8246696
diff -r 345693999a92 test/jdk/ProblemList.txt
---
Can I get a quick review for the following. Getting some unexpected OOM from
test. Pulling until I figure it out.
jbs: https://bugs.openjdk.java.net/browse/JDK-8246696
diff -r 345693999a92 test/jdk/ProblemList.txt
--- a/test/jdk/ProblemList.txt Fri Jun 05 11:35:06 2020 -0300
+++
Vladimir, Leonid,
Thank you for your reviews! pushed.
-- Igor
> On Jun 4, 2020, at 3:55 PM, Leonid Mesnik wrote:
>
> Looks good.
>
> Leonid
> On Jun 5, 2020, at 1:02 AM, Vladimir Ivanov
> wrote:
>
> Looks good.
>
> Best regards,
> Vladimir Ivanov
>
> On 27.05.2020 20:25, Igor Ignatyev
Hi,
forgive me if this was discussed already, but I found that the implementation
of hidden classes causes a change in behavior for Class.getCanonicalName().
When using the following example code:
interface Function {
void doSomething();
}
public class Example {
public static
looks OK.
/Andy
On 6/4/2020 4:22 PM, Alexey Semenyuk wrote:
Please review fix [2] for jpackage bug [1].
Move functionality to collect data about app (main class name, main
jar, module name, etc) from JLinkBundlerHelper and
StandardBundlerParam classes in dedicated LauncherData class.
Remove
Thanks, Roger!
On 2020-06-05 14:18, Roger Riggs wrote:
Looks good.
Looks good.
Thanks for the extra attention.
Roger
On 6/5/20 6:09 AM, Claes Redestad wrote:
Hi Roger,
yes, things were a bit inconsistent whether this is checking all or one.
I went with all (even though all usage only asks for one) and updated
to use plural consistently.
/Claes
On
I know there was a discussion about this elsewhere but I would like to take the
opportunity to correct this now
make//autoconf/flags-cflags.m4:241
elif test "x$TOOLCHAIN_TYPE" = xclang; then
if test "x$OPENJDK_TARGET_OS" = xmacosx; then
# On MacOSX we optimize for size, something
Hi Roger,
yes, things were a bit inconsistent whether this is checking all or one.
I went with all (even though all usage only asks for one) and updated
to use plural consistently.
/Claes
On 2020-06-05 11:23, Roger Riggs wrote:
Hi Claes,
Also, since the requested attribute is a bit mask the
Hi Claes,
Also, since the requested attribute is a bit mask the test should be
that all of the attributes are present, not just some, iIt might prevent
a mistake down the line.
Probably the method name and argument names should be plural.
Line: 120: should be
119 public boolean
> On 4 Jun 2020, at 21:16, Mandy Chung wrote:
>
> ...
> I finalized an updated CSR per all feedbacks which includes the deprecation
> `sun.misc.Unsafe::shouldBeInitialized`.
>https://bugs.openjdk.java.net/browse/JDK-8245871
>
> updated webrev:
>
Looks good.
Best regards,
Vladimir Ivanov
On 27.05.2020 20:25, Igor Ignatyev wrote:
http://cr.openjdk.java.net/~iignatyev//8245874/webrev.00
8 lines changed: 2 ins; 0 del; 6 mod;
Hi all,
could you please review this small and trivial cleanup in TEST.ROOT files of
test/jdk/ and
[Changes are only to the build system, but since the changes have jdk-wide
effect I’ve cc’d what I think are the relevant dev lists.]
This change is part of JEP 347; the intent is to target JDK 16.
Please review this change to the building of C++ code in the JDK to
enable the use of C++14
41 matches
Mail list logo