Quantcast

Re: Bug#859660: Bug#859660: artemis running issue

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
7 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Bug#859660: Bug#859660: artemis running issue

Andreas Tille-5
Hi Java developers,

is there any known issue with running JARs via symlink to /usr/bin?

I suspect this is the problem with this bug but I wonder what might be
the recommended way from Java team about this way to start Java programs
that used to work but is not always relieable.

Kind regards

      Andreas.

On Fri, Apr 07, 2017 at 12:43:42PM +0200, Andreas Tille wrote:

> Hi,
>
> On Fri, Apr 07, 2017 at 12:17:08PM +0200, Sascha Steinbiss wrote:
> > Hi,
> >
> > > على الأربعاء  5 نيسـان 2017 ‫12:32، كتب Jerome:
> > >> When running artemis package, get this issue :
> > >>
> > >> $ art
> > >> bash: /usr/bin/art: cannot execute binary file: Exec format error
> >
> > Probably something to do with jarwrapper? I remember running into
> > something similar quite some time ago.
>
> I'd assume the same.  I can confirm that I can not reproduce here as
> well but the problem sounds like jarwrapper connected.  I do not
> remember what package I had in the past with a similar issue but
> providing a shell wrapper calling java with some options and the jar
> file solved the issue.
>
> It seems that jarwrapper is not really reliable - as far as I know
> specifically if users are using a self compiled kernel where this
> feature can be switched of.
>
> Hope this helps
>
>      Andreas.

--
http://fam-tille.de

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Bug#859660: Bug#859660: artemis running issue

tony mancill
Hi Andreas,

This does sound like an issue we had in the past with jarwrapper and binfmt-support.  IIRC, Colin Watson was able to track down the source of this, but at the moment I cannot locate the details of that exchange.  It was something along the lines of there being multiple interpreters registered for the given binfmt.

If we know that the user's kernel supports binfmt_misc, then we should be able to figure out what's happening.  The output of "sudo update-binfmts --display jarwrapper" (should point to jarwrapper) and "sudo update-binfmts --display jar" (should point to jexec) might be informative.  I'll try to reproduce locally and report back.

Cheers,
tony




On Sun, Apr 23, 2017 at 11:36 PM, Andreas Tille <[hidden email]> wrote:
Hi Java developers,

is there any known issue with running JARs via symlink to /usr/bin?

I suspect this is the problem with this bug but I wonder what might be
the recommended way from Java team about this way to start Java programs
that used to work but is not always relieable.

Kind regards

      Andreas.

On Fri, Apr 07, 2017 at 12:43:42PM +0200, Andreas Tille wrote:
> Hi,
>
> On Fri, Apr 07, 2017 at 12:17:08PM +0200, Sascha Steinbiss wrote:
> > Hi,
> >
> > > على الأربعاء  5 نيسـان 2017 ‫12:32، كتب Jerome:
> > >> When running artemis package, get this issue :
> > >>
> > >> $ art
> > >> bash: /usr/bin/art: cannot execute binary file: Exec format error
> >
> > Probably something to do with jarwrapper? I remember running into
> > something similar quite some time ago.
>
> I'd assume the same.  I can confirm that I can not reproduce here as
> well but the problem sounds like jarwrapper connected.  I do not
> remember what package I had in the past with a similar issue but
> providing a shell wrapper calling java with some options and the jar
> file solved the issue.
>
> It seems that jarwrapper is not really reliable - as far as I know
> specifically if users are using a self compiled kernel where this
> feature can be switched of.
>
> Hope this helps
>
>      Andreas.

--
http://fam-tille.de


Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Bug#859660: Bug#859660: artemis running issue

Andreas Tille-5
Control: severity -1 important
Control: tags -1 moreinfo

Hi Tony,

On Fri, Apr 28, 2017 at 08:11:52AM -0700, tony mancill wrote:
> This does sound like an issue we had in the past with jarwrapper and
> binfmt-support.  IIRC, Colin Watson was able to track down the source of
> this, but at the moment I cannot locate the details of that exchange.  It
> was something along the lines of there being multiple interpreters
> registered for the given binfmt.

Thanks for ths info.
 
> If we know that the user's kernel supports binfmt_misc, then we should be
> able to figure out what's happening.  The output of "sudo update-binfmts
> --display jarwrapper" (should point to jarwrapper) and "sudo update-binfmts
> --display jar" (should point to jexec) might be informative.  I'll try to
> reproduce locally and report back.

I've reduced the severity of this bug from grave to important and have
added the tag moreinfo.  Jerome, could you please provide the said info?

Kind regards

        Andreas.

--
http://fam-tille.de

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Bug#859660: Bug#859660: artemis running issue

Jerome-11
Dear Andreas.

I put here the results of the commands. Home that helps...

$ sudo update-binfmts --display jarwrapper
update-binfmts: warning: jarwrapper not in database of installed binary
formats.
update-binfmts: exiting due to previous errors
$ sudo update-binfmts --display jar
update-binfmts: warning: jar not in database of installed binary formats.
update-binfmts: exiting due to previous errors


Thanks !

Regards

Le 06/05/2017 à 00:16, Andreas Tille a écrit :

> Control: severity -1 important
> Control: tags -1 moreinfo
>
> Hi Tony,
>
> On Fri, Apr 28, 2017 at 08:11:52AM -0700, tony mancill wrote:
>> This does sound like an issue we had in the past with jarwrapper and
>> binfmt-support.  IIRC, Colin Watson was able to track down the source of
>> this, but at the moment I cannot locate the details of that exchange.  It
>> was something along the lines of there being multiple interpreters
>> registered for the given binfmt.
>
> Thanks for ths info.
>
>> If we know that the user's kernel supports binfmt_misc, then we should be
>> able to figure out what's happening.  The output of "sudo update-binfmts
>> --display jarwrapper" (should point to jarwrapper) and "sudo update-binfmts
>> --display jar" (should point to jexec) might be informative.  I'll try to
>> reproduce locally and report back.
>
> I've reduced the severity of this bug from grave to important and have
> added the tag moreinfo.  Jerome, could you please provide the said info?
>
> Kind regards
>
>         Andreas.
>


--
-- Jérôme
L'oisiveté est, dit-on, la mère de tous les vices,
mais l'excès de travail est le père de toutes les soumissions.
      (Albert Jacquard)

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Bug#859660: Bug#859660: artemis running issue

Andreas Tille-5
Hi Jerome,

On Mon, May 08, 2017 at 01:42:40PM -0500, Jerome wrote:
> I put here the results of the commands. Home that helps...

I think so since it confirms the expected reason for the issue you
observed.
 
> $ sudo update-binfmts --display jarwrapper
> update-binfmts: warning: jarwrapper not in database of installed binary
> formats.
> update-binfmts: exiting due to previous errors
> $ sudo update-binfmts --display jar
> update-binfmts: warning: jar not in database of installed binary formats.
> update-binfmts: exiting due to previous errors

I have access to two different boxes, one running Jessie (stable):

$ sudo update-binfmts --display jarwrapper
[sudo] password for tillea:
jarwrapper (enabled):
     package = <local>
        type = magic
      offset = 0
       magic = PK\x03\x04
        mask =
 interpreter = /usr/bin/jarwrapper
    detector = /usr/bin/jardetector

$ sudo update-binfmts --display jar
jar (enabled):
     package = openjdk-7
        type = magic
      offset = 0
       magic = PK\x03\x04
        mask =
 interpreter = /usr/bin/jexec
    detector =


and one running Stretch (testing):


$ sudo update-binfmts --display jarwrapper
jarwrapper (enabled):
     package = jarwrapper
        type = magic
      offset = 0
       magic = PK\x03\x04
        mask =
 interpreter = /usr/bin/jarwrapper
    detector = /usr/bin/jardetector
$ sudo update-binfmts --display jar
jar (enabled):
     package = openjdk-7
        type = magic
      offset = 0
       magic = PK\x03\x04
        mask =
 interpreter = /usr/bin/jexec
    detector =


Could you please post the result of

   apt-cache policy jarwrapper

make sure it is installed

   sudo apt-get install jarwrapper

and try again?   If artemis works afterwards the solution would be to
add jarwrapper to the dependencies (hmmm, I somehow assumed that this
would be the case but its not. :-()

Thanks for your patience

       Andreas.


--
http://fam-tille.de

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Bug#859660: Bug#859660: artemis running issue

Jerome-11
Dear Andreas.

Thank's a lot for your questions and remark. This will help me to detect
where was the problem.
And i apologize that teh error is generated during the creation of a
live debian iso file. As the error wa detected in the live itself, i
send the report as this.. Forgetting that i'm in a live session. You're
bug system is too perfect, as it permit report bugs in a live session!

The problem was detected and corrected in the binfmt-support bug (#750245).

So i'm declaring this issue corrected, as this from another problem.

Regards

Le 09/05/2017 à 05:19, Andreas Tille a écrit :

> Hi Jerome,
>
> On Mon, May 08, 2017 at 01:42:40PM -0500, Jerome wrote:
>> I put here the results of the commands. Home that helps...
>
> I think so since it confirms the expected reason for the issue you
> observed.
>
>> $ sudo update-binfmts --display jarwrapper
>> update-binfmts: warning: jarwrapper not in database of installed binary
>> formats.
>> update-binfmts: exiting due to previous errors
>> $ sudo update-binfmts --display jar
>> update-binfmts: warning: jar not in database of installed binary formats.
>> update-binfmts: exiting due to previous errors
>
> I have access to two different boxes, one running Jessie (stable):
>
> $ sudo update-binfmts --display jarwrapper
> [sudo] password for tillea:
> jarwrapper (enabled):
>      package = <local>
>         type = magic
>       offset = 0
>        magic = PK\x03\x04
>         mask =
>  interpreter = /usr/bin/jarwrapper
>     detector = /usr/bin/jardetector
>
> $ sudo update-binfmts --display jar
> jar (enabled):
>      package = openjdk-7
>         type = magic
>       offset = 0
>        magic = PK\x03\x04
>         mask =
>  interpreter = /usr/bin/jexec
>     detector =
>
>
> and one running Stretch (testing):
>
>
> $ sudo update-binfmts --display jarwrapper
> jarwrapper (enabled):
>      package = jarwrapper
>         type = magic
>       offset = 0
>        magic = PK\x03\x04
>         mask =
>  interpreter = /usr/bin/jarwrapper
>     detector = /usr/bin/jardetector
> $ sudo update-binfmts --display jar
> jar (enabled):
>      package = openjdk-7
>         type = magic
>       offset = 0
>        magic = PK\x03\x04
>         mask =
>  interpreter = /usr/bin/jexec
>     detector =
>
>
> Could you please post the result of
>
>    apt-cache policy jarwrapper
>
> make sure it is installed
>
>    sudo apt-get install jarwrapper
>
> and try again?   If artemis works afterwards the solution would be to
> add jarwrapper to the dependencies (hmmm, I somehow assumed that this
> would be the case but its not. :-()
>
> Thanks for your patience
>
>        Andreas.
>
>


--
-- Jérôme
L'homme, cet insecte invisible,
rampant dans les sillons d'un globe imperceptible
        (Lamartine, L'infini dans les cieux)

Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Bug#859660: Bug#859660: artemis running issue

tony mancill
On Thu, May 11, 2017 at 12:33:36PM -0500, Jerome wrote:

> Thank's a lot for your questions and remark. This will help me to detect
> where was the problem.
> And i apologize that teh error is generated during the creation of a live
> debian iso file. As the error wa detected in the live itself, i send the
> report as this.. Forgetting that i'm in a live session. You're bug system is
> too perfect, as it permit report bugs in a live session!
>
> The problem was detected and corrected in the binfmt-support bug (#750245).
>
> So i'm declaring this issue corrected, as this from another problem.
Hello Jerome,

Thank you for the follow-up and for tracing the problem back to an issue
with the Live CD.  It's always useful to know where these issues remain.

While I was poking around with this, I was able to reproduce the problem
by installing the jessie-backports version of artemis into a jessie
chroot on a stretch (testing) system.  (Which is nothing new, because
it's just another mhandanifestation of #750245.)  The root cause of the
issue in the older chroot can be seen by trying to run the
jarwrapper.postinst directly:

# /var/lib/dpkg/info/jarwrapper.postinst
update-binfmts: warning: found manually created entry for jarwrapper in /proc/sys/fs/binfmt_misc; leaving it alone

Installing a artemis into a stretch or sid chroot on the same system
works fine.

What's curious to me (and perhaps helpful to others?), is that if I
create a jessie chroot and install jarwrapper *before* installing
artemis, and more to the point, before installing the JDK and its
dependencies, the jarwrapper binfmt functions correctly.

Cheers,
tony

signature.asc (849 bytes) Download Attachment
Loading...