Bug 58451

You haven't chosen any version in this report.
I could not confirm this behavior. Just tested with LO Version 4.0.0.0.beta1+,
TinderBox: Linux-x86_10-Release-Configuration, Branch:libreoffice-4-0, Time: 2012-12-15_16:11:06.
If I want to create a report without any table a popup appears:
The report can not be executed unless it is bound to content.
When I create a table (what isn't described in the report) and connect to the report, I can execute it. There doesn't appear any content in the report, because there isn't any content in the table.

Hi Robert!

Sorry, I'm using LO 4 Master.

But, the crash happens here. This is the error:

# A fatal error has been detected by the Java Runtime Environment:
#
#  SIGSEGV (0xb) at pc=0xad79c6ff, pid=4749, tid=3036846528
#
# JRE version: 7.0_09
# Java VM: OpenJDK Client VM (23.2-b09 mixed mode linux-x86 )
# Problematic frame:
# C  [libdbulo.so+0x856ff]  com::sun::star::uno::cpp_release(void*)+0xc
#
# Core dump written. Default location: /mnt/dados/gitroot/core/solver/unxlngi6.pro/installation/opt/program/core or core.4749
#
# An error report file with more information is saved as:
# /tmp/hs_err_pid4749.log
#
# If you would like to submit a bug report, please include
# instructions on how to reproduce the bug and visit:
#   http://icedtea.classpath.org/bugzilla
#

[error occurred during error reporting , id 0xb]

This info give more help?

Thanks a lot!

Hello

(In reply to comment #1)
> If I want to create a report without any table a popup appears:
> The report can not be executed unless it is bound to content.
> When I create a table (what isn't described in the report) and connect to
> the report, I can execute it. There doesn't appear any content in the
> report, because there isn't any content in the table.

I confirm this comment : I can not reproduce with windows 7 64bits &
Version 4.0.0.0.beta1+ (Build ID: b0eb7231a9643d71be3125be7248c91242339ab)

I also tried after creating a table.
Execute the report displays another message: 
"The report can not be executed unless at least on object has been inserted".

Note: my JRE is 1.6.0_38

Regards
Pierre-Yves

It could help if you attach /tmp/hs_err_pid4749.log but I don't think you still have it since it's in /tmp and you must have already rebooted.

Could you retrieve a backtrace by following this link (https://wiki.documentfoundation.org/BugReport#How_to_get_a_backtrace_.28on_Linux.29)?

Can not confirm with :

Version 4.1.0.0.alpha0+ (Build ID: 74b960ef3cd96f0f0e170234bc0947bfb51723f)

on Linux Mint Maya 32bit.

When I click on Create Report in Design the Report Template opens. No crash.


Alex

Created attachment 71989 [details]
1

After compile the LO with head 8e25844811ebcf313f4ae9d0caae1f97fd230a7d and open Base and creating a new database

Created attachment 71990 [details]
The parameters of the new database

Created attachment 71991 [details]
3 - Name of new database

Created attachment 71992 [details]
After create the new database, click on Create Report in Design View

Created attachment 71993 [details]
After crick in Create Report in Design View, just click Execute Report

Created attachment 71994 [details]
Backtrace with --backtrace and crash

Created attachment 71995 [details]
valgrind log

Guys I hope this information could help us to fix this issue..

Maybe more useful information:
[marcos@jedi program]$ java -version
java version "1.7.0_09-icedtea"
OpenJDK Runtime Environment (fedora-2.3.3.fc17.1-i386)
OpenJDK Server VM (build 23.2-b09, mixed mode)

[marcos@jedi program]$ uname -a
Linux jedi 3.5.4-2.fc17.i686 #1 SMP Wed Sep 26 22:32:49 UTC 2012 i686 i686 i386 GNU/Linux

I'm using Fedora 17 with Xfce

Created attachment 71996 [details]
bt + console logs on master

On pc Debian x86-64 with master sources updated today (commit bb97ecdbcc8d8dafd39e728b21bc68efee4eccbc), I reproduced the problem

Marcos: great for the detailed steps. It could be interesting you add the option "--enable-symbols" for having a even more useful backtrace. Anyway, I did it (I've got --enable-dbgutil to have symbols + debug logs)
Nonetheless, thank you for your bt and Valgrind.

Marcos: should have read the content of Valgrind log, what you attached is not at a Valgrind log. see this link (https://wiki.documentfoundation.org/BugReport#How_to_get_a_valgrind_log_.28on_Linux.29)

(Update tracker to New)

in a gdb session, I've got this:
Breakpoint 2, dbtools::showError (_rInfo=..., _xParent=
    uno::Reference to {<com::sun::star::lang::XComponent> = {<com::sun::star::uno::XInterface> = {_vptr.XInterface = 0x7f5ecc5b5f10}, <No data fields>}, <No data fields>}, 
    _rxContext=uno::Reference to {<com::sun::star::uno::XInterface> = {_vptr.XInterface = 0x7f5ece853b30}, <No data fields>})
    at /home/julien/compile-libreoffice/libo/connectivity/source/commontools/dbtools.cxx:1485
1485	    if (_rInfo.isValid())
(gdb) n
1489	            Reference< XExecutableDialog > xErrorDialog = ErrorMessageDialog::create( _rxContext, "", _xParent, _rInfo.get() );
(gdb) p _rxContext
$2 = uno::Reference to {<com::sun::star::uno::XInterface> = {_vptr.XInterface = 0x7f5ece853b30}, <No data fields>}
(gdb) p _xParent
$3 = uno::Reference to {<com::sun::star::lang::XComponent> = {<com::sun::star::uno::XInterface> = {_vptr.XInterface = 0x7f5ecc5b5f10}, <No data fields>}, <No data fields>}
(gdb) p _rInfo.get()
$4 = 
  uno::Any {<com::sun::star::sdbc::SQLWarning> = {<com::sun::star::sdbc::SQLException> = {<com::sun::star::uno::Exception> = {Message = "The report can not be executed unless it is bound to content.", Context = empty uno::Reference}, SQLState = "", ErrorCode = 0, NextException = empty uno::Any}, <No data fields>}, Details = ""}
(gdb) n

Program received signal SIGSEGV, Segmentation fault.
0x0000000001a89580 in ?? ()

I wonder if it could be due to: 19854e653be1dc01f5b655fca851a6e3a651a940  fdo#46808, Adapt sdb::ErrorMessageDialog UNO service to new style

The problem could be the fact that getORB() function in dbaccess/inc/genericcontroller.hxx returns ::com::sun::star::lang::XMultiServiceFactory not ::com::sun::star::uno::XComponentContext

Noel: would you have some time to take a look at this?
The problem to change there are 3 files in DBAccess which use XMultiServiceFactory:
source/ui/app/AppController.cxx
source/ui/browser/sbagrid.cxx
source/ui/browser/genericcontroller.cxx

I'll be happy to look at this, but I'll only be back at my hacking machine in a week.

Julien, you are correct, I need to propogate the 
  XMultiServiceFactory->XComponentContext
change up the hierarchy to 
   OGenericUnoController
and dump the unnecessary
  m_xContext
field in 
  OReportController.

That being said, the only way I can see this happening is if showError is being called after OGenericUnoController::disposing() has cleared it's m_xServiceFactory field. (why it is doing that is another question).

Probably just removing the clearing of m_xServiceFactory in OGenericUnoController will make this crash go away.

(In reply to comment #18)
> I'll be happy to look at this, but I'll only be back at my hacking machine
> in a week.
> 
> Julien, you are correct, I need to propogate the 
>   XMultiServiceFactory->XComponentContext
> change up the hierarchy to 
>    OGenericUnoController
> and dump the unnecessary
>   m_xContext
> field in 
>   OReportController.
> 
> That being said, the only way I can see this happening is if showError is
> being called after OGenericUnoController::disposing() has cleared it's
> m_xServiceFactory field. (why it is doing that is another question).
> 
> Probably just removing the clearing of m_xServiceFactory in
> OGenericUnoController will make this crash go away.

Thank you Noel for your feedback. I sent a post on dev mailing list (see http://nabble.documentfoundation.org/About-fdo-58451-Base-crash-when-Execute-report-with-a-empty-database-td4025905.html). Perhaps someone may help before you come back :-)

Created attachment 72027 [details]
valgrind log - now with valgrind installed!!

I tought that valgrind was installed here, but it doesn't..

I belive this is the right log message!

Created attachment 72033 [details]
vagrind report

When i follow https://wiki.documentfoundation.org/BugReport#How_to_get_a_valgrind_log_.28on_Linux.29 it seems that an error occurs when trying to get the summary of icedtea...

So, attaching this...

Noel Grandin committed a patch related to this issue.
It has been pushed to "master":

http://cgit.freedesktop.org/libreoffice/core/commit/?id=63894cb4fed44aed01092cdee1fe30463eb5223e

fix fdo#58451



The patch should be included in the daily builds available at
http://dev-builds.libreoffice.org/daily/ in the next 24-48 hours. More
information about daily builds can be found at:
http://wiki.documentfoundation.org/Testing_Daily_Builds
Affected users are encouraged to test the fix and report feedback.

Created attachment 72352 [details]
Now it works!

Thanks a lot Noel!

Now it works! See this print: now it gives a message of empty report when testing.

Closing the issue. Thanks again!