You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@harmony.apache.org by Alexey Petrenko <al...@gmail.com> on 2008/02/20 18:10:18 UTC
[APPROVAL] revert r620461
Guys,
I'm going to revert r620461 because it looks like it causes a
functional suite failure...
Any objections?
SY, Alexey
2008/2/19, Andrey Pavlenko (JIRA) <ji...@apache.org>:
>
> [ https://issues.apache.org/jira/browse/HARMONY-5473?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12570167#action_12570167 ]
>
> Andrey Pavlenko commented on HARMONY-5473:
> ------------------------------------------
>
> It looks like this fix caused a regression - the test http://people.apache.org/~smishura/r620734/Linux_x86_64/func/junit/functional/org/apache/harmony/test/func/api/java/util/jar/Manifest/package-summary.html failed after the commit. As mentioned in HARMONY-4419, IOEx should be thrown for \n\r character sequence.
>
> > [classlib][archive] Manifest file with empty line proves IOException
> > --------------------------------------------------------------------
> >
> > Key: HARMONY-5473
> > URL: https://issues.apache.org/jira/browse/HARMONY-5473
> > Project: Harmony
> > Issue Type: Bug
> > Components: Classlib
> > Reporter: Evgeniya Maenkova
> > Assignee: Alexey Petrenko
> > Priority: Critical
> > Attachments: archive.patch, archive.patch, hyts_manifest1.mf, MANIFEST.MF, MANIFEST2.MF
> >
> >
> > I cannot start some of my apps, because their jar-files contain manifests with empty lines, proving IOException on Harmony.
> > The situation can be reproduced by manifest file with empty line creation and invoking new Manifest(InputStream). Manifest files which proves this will be attached.
> > Test case output:
> > $ /cygdrive/c/Harmony_to_run/trunk/working_vm/build/win_ia32_msvc_release/deploy/jdk/jre/bin/java Test MANIFEST2.MF
> > Uncaught exception in main:
> > java.io.IOException: Invalid attribute
> > at java.util.jar.InitManifest.addAttribute(InitManifest.java:282)
> > at java.util.jar.InitManifest.<init>(InitManifest.java:71)
> > at java.util.jar.Manifest.read(Manifest.java:173)
> > at java.util.jar.Manifest.<init>(Manifest.java:76)
> > at Test.main(Test.java:8)
> > $ /cygdrive/w/UBS/Builds/jdk1.6.0_win32/bin/java Test MANIFEST2.MF
> > end
> > Test case is:
> > import java.io.FileInputStream;
> > import java.util.jar.Manifest;
> > public class Test {
> > public static void main(String[] args) throws Exception {
> > String fileName = args[0];
> > FileInputStream fs = new FileInputStream(fileName);
> > Manifest m = new Manifest(fs);
> > fs.close();
> > System.out.println("end");
> > }
> > }
>
>
> --
> This message is automatically generated by JIRA.
> -
> You can reply to this email to add a comment to the issue online.
>
>
Re: [APPROVAL] revert r620461
Posted by Alexei Fedotov <al...@gmail.com>.
I will look into the issue in two hours. There is one more problem
about this patch: it just violate specification. We may choose to
violate Sun's specification in a way Sun's VM does, or revert the
patch.
On Thu, Feb 21, 2008 at 12:26 PM, Alexey Petrenko
<al...@gmail.com> wrote:
> 2008/2/21, Stepan Mishura <st...@gmail.com>:
>
> > On 2/20/08, Alexey Petrenko <al...@gmail.com> wrote:
> > > Guys,
> > >
> > > I'm going to revert r620461 because it looks like it causes a
> > > functional suite failure...
> > > Any objections?
> >
> >
> > As JIRA description says: "cannot start some of my apps, because their
> > jar-files contain manifests with empty lines".
> >
> > So with the patch applied we have 'some apps' running and 1 failing
> > func test. With reverted patch we will have 'some apps' failing and 1
> > passing test. I thought that apps have higher priority so I'm not sure
> > that the patch should be reverted.
> >
> > Is there any chance to fix the issue entirely?
> This would be the best solution. But unfortunately I have no time at
> the moment and no other volunteers to fix it.
>
> So we got two options: to revert or not revert...
>
> Other opinions? Volunteers?
>
> SY, Alexey
>
--
With best regards,
Alexei
Re: [APPROVAL] revert r620461
Posted by Alexey Petrenko <al...@gmail.com>.
Ok, if nobody objects we will keep the patch.
However, Alexei is welcome with the fix :)
SY, Alexey
2008/2/21, Andrey Pavlenko <an...@googlemail.com>:
> It looks like the applied patch is not quite correct because it introduced a
> bug violating the spec. But, from the other side, the bug is not very
> critical and can be fixed later. IMO it's ok to fix this issue after M5.
>
> On Thu, Feb 21, 2008 at 12:26 PM, Alexey Petrenko <
>
> alexey.a.petrenko@gmail.com> wrote:
>
> > 2008/2/21, Stepan Mishura <st...@gmail.com>:
> > > On 2/20/08, Alexey Petrenko <al...@gmail.com> wrote:
> > > > Guys,
> > > >
> > > > I'm going to revert r620461 because it looks like it causes a
> > > > functional suite failure...
> > > > Any objections?
> > >
> > >
> > > As JIRA description says: "cannot start some of my apps, because their
> > > jar-files contain manifests with empty lines".
> > >
> > > So with the patch applied we have 'some apps' running and 1 failing
> > > func test. With reverted patch we will have 'some apps' failing and 1
> > > passing test. I thought that apps have higher priority so I'm not sure
> > > that the patch should be reverted.
> > >
> > > Is there any chance to fix the issue entirely?
> > This would be the best solution. But unfortunately I have no time at
> > the moment and no other volunteers to fix it.
> >
> > So we got two options: to revert or not revert...
> >
> > Other opinions? Volunteers?
> >
> > SY, Alexey
> >
>
Re: [APPROVAL] revert r620461
Posted by Andrey Pavlenko <an...@googlemail.com>.
It looks like the applied patch is not quite correct because it introduced a
bug violating the spec. But, from the other side, the bug is not very
critical and can be fixed later. IMO it's ok to fix this issue after M5.
On Thu, Feb 21, 2008 at 12:26 PM, Alexey Petrenko <
alexey.a.petrenko@gmail.com> wrote:
> 2008/2/21, Stepan Mishura <st...@gmail.com>:
> > On 2/20/08, Alexey Petrenko <al...@gmail.com> wrote:
> > > Guys,
> > >
> > > I'm going to revert r620461 because it looks like it causes a
> > > functional suite failure...
> > > Any objections?
> >
> >
> > As JIRA description says: "cannot start some of my apps, because their
> > jar-files contain manifests with empty lines".
> >
> > So with the patch applied we have 'some apps' running and 1 failing
> > func test. With reverted patch we will have 'some apps' failing and 1
> > passing test. I thought that apps have higher priority so I'm not sure
> > that the patch should be reverted.
> >
> > Is there any chance to fix the issue entirely?
> This would be the best solution. But unfortunately I have no time at
> the moment and no other volunteers to fix it.
>
> So we got two options: to revert or not revert...
>
> Other opinions? Volunteers?
>
> SY, Alexey
>
Re: [APPROVAL] revert r620461
Posted by Alexey Petrenko <al...@gmail.com>.
2008/2/21, Stepan Mishura <st...@gmail.com>:
> On 2/20/08, Alexey Petrenko <al...@gmail.com> wrote:
> > Guys,
> >
> > I'm going to revert r620461 because it looks like it causes a
> > functional suite failure...
> > Any objections?
>
>
> As JIRA description says: "cannot start some of my apps, because their
> jar-files contain manifests with empty lines".
>
> So with the patch applied we have 'some apps' running and 1 failing
> func test. With reverted patch we will have 'some apps' failing and 1
> passing test. I thought that apps have higher priority so I'm not sure
> that the patch should be reverted.
>
> Is there any chance to fix the issue entirely?
This would be the best solution. But unfortunately I have no time at
the moment and no other volunteers to fix it.
So we got two options: to revert or not revert...
Other opinions? Volunteers?
SY, Alexey
Re: [APPROVAL] revert r620461
Posted by Stepan Mishura <st...@gmail.com>.
On 2/20/08, Alexey Petrenko <al...@gmail.com> wrote:
> Guys,
>
> I'm going to revert r620461 because it looks like it causes a
> functional suite failure...
> Any objections?
As JIRA description says: "cannot start some of my apps, because their
jar-files contain manifests with empty lines".
So with the patch applied we have 'some apps' running and 1 failing
func test. With reverted patch we will have 'some apps' failing and 1
passing test. I thought that apps have higher priority so I'm not sure
that the patch should be reverted.
Is there any chance to fix the issue entirely?
Thanks,
Stepan.
>
> SY, Alexey
>
> 2008/2/19, Andrey Pavlenko (JIRA) <ji...@apache.org>:
> >
> > [ https://issues.apache.org/jira/browse/HARMONY-5473?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12570167#action_12570167 ]
> >
> > Andrey Pavlenko commented on HARMONY-5473:
> > ------------------------------------------
> >
> > It looks like this fix caused a regression - the test http://people.apache.org/~smishura/r620734/Linux_x86_64/func/junit/functional/org/apache/harmony/test/func/api/java/util/jar/Manifest/package-summary.html failed after the commit. As mentioned in HARMONY-4419, IOEx should be thrown for \n\r character sequence.
> >
> > > [classlib][archive] Manifest file with empty line proves IOException
> > > --------------------------------------------------------------------
> > >
> > > Key: HARMONY-5473
> > > URL: https://issues.apache.org/jira/browse/HARMONY-5473
> > > Project: Harmony
> > > Issue Type: Bug
> > > Components: Classlib
> > > Reporter: Evgeniya Maenkova
> > > Assignee: Alexey Petrenko
> > > Priority: Critical
> > > Attachments: archive.patch, archive.patch, hyts_manifest1.mf, MANIFEST.MF, MANIFEST2.MF
> > >
> > >
> > > I cannot start some of my apps, because their jar-files contain manifests with empty lines, proving IOException on Harmony.
> > > The situation can be reproduced by manifest file with empty line creation and invoking new Manifest(InputStream). Manifest files which proves this will be attached.
> > > Test case output:
> > > $ /cygdrive/c/Harmony_to_run/trunk/working_vm/build/win_ia32_msvc_release/deploy/jdk/jre/bin/java Test MANIFEST2.MF
> > > Uncaught exception in main:
> > > java.io.IOException: Invalid attribute
> > > at java.util.jar.InitManifest.addAttribute(InitManifest.java:282)
> > > at java.util.jar.InitManifest.<init>(InitManifest.java:71)
> > > at java.util.jar.Manifest.read(Manifest.java:173)
> > > at java.util.jar.Manifest.<init>(Manifest.java:76)
> > > at Test.main(Test.java:8)
> > > $ /cygdrive/w/UBS/Builds/jdk1.6.0_win32/bin/java Test MANIFEST2.MF
> > > end
> > > Test case is:
> > > import java.io.FileInputStream;
> > > import java.util.jar.Manifest;
> > > public class Test {
> > > public static void main(String[] args) throws Exception {
> > > String fileName = args[0];
> > > FileInputStream fs = new FileInputStream(fileName);
> > > Manifest m = new Manifest(fs);
> > > fs.close();
> > > System.out.println("end");
> > > }
> > > }
> >
> >
> > --
> > This message is automatically generated by JIRA.
> > -
> > You can reply to this email to add a comment to the issue online.
> >
> >
>
RE: [APPROVAL] revert r620461
Posted by "Maenkova, Evgeniya G" <ev...@intel.com>.
I propose roll back as the best solution now.
There is no clarity about whether my apps uses valid manifest or not.
I'm trying to clarify now whether RI contradicts with specs or not.
Thanks,
Evgeniya
-----Original Message-----
From: Stepan Mishura [mailto:stepan.mishura@gmail.com]
Sent: Friday, February 22, 2008 11:32 AM
To: dev@harmony.apache.org
Subject: Re: [APPROVAL] revert r620461
On 2/20/08, Alexey Petrenko <al...@gmail.com> wrote:
> Guys,
>
> I'm going to revert r620461 because it looks like it causes a
> functional suite failure...
> Any objections?
>
Alexey,
I've checked the reproducer test from HARMONY-5473 with
MANIFEST.MF.txt against the last M5 candidate r629320 and I see:
Uncaught exception in main:
java.io.IOException: Invalid attribute
at
java.util.jar.InitManifest.addAttribute(InitManifest.java:304)
at java.util.jar.InitManifest.<init>(InitManifest.java:71)
at java.util.jar.Manifest.read(Manifest.java:173)
at java.util.jar.Manifest.<init>(Manifest.java:76)
at test.main(test.java:10)
So for me the patch you applied is invalid - it didn't fix the issue,
moreover it introduced regression (failed func test).
BTW, this is a good example that it worth to add a regression test for a
fix [1]
http://mail-archives.apache.org/mod_mbox/harmony-dev/200802.mbox/%3c6e47
b64f0802111942r4f5753e4m29a8535d1532b461@mail.gmail.com%3e
Thanks,
Stepan.
> SY, Alexey
>
> 2008/2/19, Andrey Pavlenko (JIRA) <ji...@apache.org>:
> >
> > [
https://issues.apache.org/jira/browse/HARMONY-5473?page=com.atlassian.ji
ra.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=125701
67#action_12570167 ]
> >
> > Andrey Pavlenko commented on HARMONY-5473:
> > ------------------------------------------
> >
> > It looks like this fix caused a regression - the test
http://people.apache.org/~smishura/r620734/Linux_x86_64/func/junit/funct
ional/org/apache/harmony/test/func/api/java/util/jar/Manifest/package-su
mmary.html failed after the commit. As mentioned in HARMONY-4419, IOEx
should be thrown for \n\r character sequence.
> >
> > > [classlib][archive] Manifest file with empty line proves
IOException
> > >
--------------------------------------------------------------------
> > >
> > > Key: HARMONY-5473
> > > URL:
https://issues.apache.org/jira/browse/HARMONY-5473
> > > Project: Harmony
> > > Issue Type: Bug
> > > Components: Classlib
> > > Reporter: Evgeniya Maenkova
> > > Assignee: Alexey Petrenko
> > > Priority: Critical
> > > Attachments: archive.patch, archive.patch,
hyts_manifest1.mf, MANIFEST.MF, MANIFEST2.MF
> > >
> > >
> > > I cannot start some of my apps, because their jar-files contain
manifests with empty lines, proving IOException on Harmony.
> > > The situation can be reproduced by manifest file with empty line
creation and invoking new Manifest(InputStream). Manifest files which
proves this will be attached.
> > > Test case output:
> > > $
/cygdrive/c/Harmony_to_run/trunk/working_vm/build/win_ia32_msvc_release/
deploy/jdk/jre/bin/java Test MANIFEST2.MF
> > > Uncaught exception in main:
> > > java.io.IOException: Invalid attribute
> > > at
java.util.jar.InitManifest.addAttribute(InitManifest.java:282)
> > > at
java.util.jar.InitManifest.<init>(InitManifest.java:71)
> > > at java.util.jar.Manifest.read(Manifest.java:173)
> > > at java.util.jar.Manifest.<init>(Manifest.java:76)
> > > at Test.main(Test.java:8)
> > > $ /cygdrive/w/UBS/Builds/jdk1.6.0_win32/bin/java Test
MANIFEST2.MF
> > > end
> > > Test case is:
> > > import java.io.FileInputStream;
> > > import java.util.jar.Manifest;
> > > public class Test {
> > > public static void main(String[] args) throws Exception {
> > > String fileName = args[0];
> > > FileInputStream fs = new FileInputStream(fileName);
> > > Manifest m = new Manifest(fs);
> > > fs.close();
> > > System.out.println("end");
> > > }
> > > }
> >
> >
> > --
> > This message is automatically generated by JIRA.
> > -
> > You can reply to this email to add a comment to the issue online.
> >
> >
>
--------------------------------------------------------------------
Closed Joint Stock Company Intel A/O
Registered legal address: Krylatsky Hills Business Park,
17 Krylatskaya Str., Bldg 4, Moscow 121614,
Russia Federation
This e-mail and any attachments may contain confidential material for
the sole use of the intended recipient(s). Any review or distribution
by others is strictly prohibited. If you are not the intended
recipient, please contact the sender and delete all copies.
Re: [APPROVAL] revert r620461
Posted by Stepan Mishura <st...@gmail.com>.
On 2/20/08, Alexey Petrenko <al...@gmail.com> wrote:
> Guys,
>
> I'm going to revert r620461 because it looks like it causes a
> functional suite failure...
> Any objections?
>
Alexey,
I've checked the reproducer test from HARMONY-5473 with
MANIFEST.MF.txt against the last M5 candidate r629320 and I see:
Uncaught exception in main:
java.io.IOException: Invalid attribute
at java.util.jar.InitManifest.addAttribute(InitManifest.java:304)
at java.util.jar.InitManifest.<init>(InitManifest.java:71)
at java.util.jar.Manifest.read(Manifest.java:173)
at java.util.jar.Manifest.<init>(Manifest.java:76)
at test.main(test.java:10)
So for me the patch you applied is invalid - it didn't fix the issue,
moreover it introduced regression (failed func test).
BTW, this is a good example that it worth to add a regression test for a fix [1]
http://mail-archives.apache.org/mod_mbox/harmony-dev/200802.mbox/%3c6e47b64f0802111942r4f5753e4m29a8535d1532b461@mail.gmail.com%3e
Thanks,
Stepan.
> SY, Alexey
>
> 2008/2/19, Andrey Pavlenko (JIRA) <ji...@apache.org>:
> >
> > [ https://issues.apache.org/jira/browse/HARMONY-5473?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12570167#action_12570167 ]
> >
> > Andrey Pavlenko commented on HARMONY-5473:
> > ------------------------------------------
> >
> > It looks like this fix caused a regression - the test http://people.apache.org/~smishura/r620734/Linux_x86_64/func/junit/functional/org/apache/harmony/test/func/api/java/util/jar/Manifest/package-summary.html failed after the commit. As mentioned in HARMONY-4419, IOEx should be thrown for \n\r character sequence.
> >
> > > [classlib][archive] Manifest file with empty line proves IOException
> > > --------------------------------------------------------------------
> > >
> > > Key: HARMONY-5473
> > > URL: https://issues.apache.org/jira/browse/HARMONY-5473
> > > Project: Harmony
> > > Issue Type: Bug
> > > Components: Classlib
> > > Reporter: Evgeniya Maenkova
> > > Assignee: Alexey Petrenko
> > > Priority: Critical
> > > Attachments: archive.patch, archive.patch, hyts_manifest1.mf, MANIFEST.MF, MANIFEST2.MF
> > >
> > >
> > > I cannot start some of my apps, because their jar-files contain manifests with empty lines, proving IOException on Harmony.
> > > The situation can be reproduced by manifest file with empty line creation and invoking new Manifest(InputStream). Manifest files which proves this will be attached.
> > > Test case output:
> > > $ /cygdrive/c/Harmony_to_run/trunk/working_vm/build/win_ia32_msvc_release/deploy/jdk/jre/bin/java Test MANIFEST2.MF
> > > Uncaught exception in main:
> > > java.io.IOException: Invalid attribute
> > > at java.util.jar.InitManifest.addAttribute(InitManifest.java:282)
> > > at java.util.jar.InitManifest.<init>(InitManifest.java:71)
> > > at java.util.jar.Manifest.read(Manifest.java:173)
> > > at java.util.jar.Manifest.<init>(Manifest.java:76)
> > > at Test.main(Test.java:8)
> > > $ /cygdrive/w/UBS/Builds/jdk1.6.0_win32/bin/java Test MANIFEST2.MF
> > > end
> > > Test case is:
> > > import java.io.FileInputStream;
> > > import java.util.jar.Manifest;
> > > public class Test {
> > > public static void main(String[] args) throws Exception {
> > > String fileName = args[0];
> > > FileInputStream fs = new FileInputStream(fileName);
> > > Manifest m = new Manifest(fs);
> > > fs.close();
> > > System.out.println("end");
> > > }
> > > }
> >
> >
> > --
> > This message is automatically generated by JIRA.
> > -
> > You can reply to this email to add a comment to the issue online.
> >
> >
>