Bug 157042 - [Linux only] Calc crashes when closed in rtl::str::release<_rtl_uString>(_rtl_uString*) (steps in comment 27)
Summary: [Linux only] Calc crashes when closed in rtl::str::release<_rtl_uString>(_rtl...
Status: RESOLVED FIXED
Alias: None
Product: LibreOffice
Classification: Unclassified
Component: Calc (show other bugs)
Version:
(earliest affected)
7.6.0.3 release
Hardware: x86-64 (AMD64) Linux (All)
: medium critical
Assignee: Not Assigned
URL:
Whiteboard: target:7.6.5 target:24.2.1 target:24.8.0
Keywords: bibisectRequest, regression
: 158228 159088 159831 (view as bug list)
Depends on:
Blocks: KDE, KF5 Crash Exit
  Show dependency treegraph
 
Reported: 2023-08-31 18:19 UTC by Eric
Modified: 2024-02-22 04:10 UTC (History)
11 users (show)

See Also:
Crash report or crash signature: ["void rtl::str::release<_rtl_uString>(_rtl_uString*)"]


Attachments
system log from last crash exiting calc (11.08 KB, text/plain)
2023-09-03 18:15 UTC, John Kizer
Details
Another crash and debug data (665 bytes, text/vnd.kde.kcrash-report)
2023-09-22 13:26 UTC, Eric
Details
chart of number of crashes with this signature per version (15.14 KB, image/png)
2023-11-16 16:07 UTC, Stéphane Guillou (stragu)
Details
minidump with 7.6.5 daily build (415.27 KB, application/vnd.tcpdump.pcap)
2024-01-31 04:11 UTC, Stéphane Guillou (stragu)
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Eric 2023-08-31 18:19:44 UTC
Description:
Version: 7.6.0.3 (X86_64) / LibreOffice Community
Build ID: 60(Build:3)
CPU threads: 12; OS: Linux 6.4; UI render: default; VCL: kf5 (cairo+xcb)
Locale: en-US (en_US.UTF-8); UI: en-US
Calc: CL threaded


openSUSE Tumbleweed 20230823 (x86_64) on kernel 6.4.11-1-default
Installed Memory 16G
KDE 5.109.0 Plasma 5.27.7
Qt: 5.15.10
Display Type:    x11

Several times now, when i close calc it segfaults. Here is the debug info:

Application: soffice (soffice), signal: Segmentation fault

[KCrash Handler]
#4  rtl::str::release<_rtl_uString>(_rtl_uString*) () at /usr/src/debug/libreoffice-7.6.0.3/sal/rtl/strtmpl.hxx:745
#5  rtl_uString_release(_rtl_uString*) (pThis=0x7fac60757fc0) at /usr/src/debug/libreoffice-7.6.0.3/sal/rtl/ustring.cxx:1178
#6  0x00007facc04bb95b in rtl::OUString::~OUString() () at /usr/src/debug/libreoffice-7.6.0.3/include/rtl/ustring.hxx:549
#7  std::pair<rtl::OUString const, com::sun::star::uno::Any>::~pair() () at /usr/include/c++/13/bits/stl_pair.h:187
#8  std::__new_allocator<std::__detail::_Hash_node<std::pair<rtl::OUString const, com::sun::star::uno::Any>, true> >::destroy<std::pair<rtl::OUString const, com::sun::star::uno::Any> >(std::pair<rtl::OUString const, com::sun::star::uno::Any>*) () at /usr/include/c++/13/bits/new_allocator.h:194
#9  std::allocator_traits<std::allocator<std::__detail::_Hash_node<std::pair<rtl::OUString const, com::sun::star::uno::Any>, true> > >::destroy<std::pair<rtl::OUString const, com::sun::star::uno::Any> >(std::allocator<std::__detail::_Hash_node<std::pair<rtl::OUString const, com::sun::star::uno::Any>, true> >&, std::pair<rtl::OUString const, com::sun::star::uno::Any>*) () at /usr/include/c++/13/bits/alloc_traits.h:557
#10 std::__detail::_Hashtable_alloc<std::allocator<std::__detail::_Hash_node<std::pair<rtl::OUString const, com::sun::star::uno::Any>, true> > >::_M_deallocate_node(std::__detail::_Hash_node<std::pair<rtl::OUString const, com::sun::star::uno::Any>, true>*) () at /usr/include/c++/13/bits/hashtable_policy.h:2020
#11 std::__detail::_Hashtable_alloc<std::allocator<std::__detail::_Hash_node<std::pair<rtl::OUString const, com::sun::star::uno::Any>, true> > >::_M_deallocate_nodes(std::__detail::_Hash_node<std::pair<rtl::OUString const, com::sun::star::uno::Any>, true>*) () at /usr/include/c++/13/bits/hashtable_policy.h:2042
#12 std::_Hashtable<rtl::OUString, std::pair<rtl::OUString const, com::sun::star::uno::Any>, std::allocator<std::pair<rtl::OUString const, com::sun::star::uno::Any> >, std::__detail::_Select1st, std::equal_to<rtl::OUString>, std::hash<rtl::OUString>, std::__detail::_Mod_range_hashing, std::__detail::_Default_ranged_hash, std::__detail::_Prime_rehash_policy, std::__detail::_Hashtable_traits<true, false, true> >::clear() (this=0x7facc370a440 <_ZZN10comphelper6detail20ConfigurationWrapper3getEvE7WRAPPER.lto_priv.0+64>) at /usr/include/c++/13/bits/hashtable.h:2509
#13 0x00007facc060cae3 in std::unordered_map<rtl::OUString, com::sun::star::uno::Any, std::hash<rtl::OUString>, std::equal_to<rtl::OUString>, std::allocator<std::pair<rtl::OUString const, com::sun::star::uno::Any> > >::clear() () at /usr/include/c++/13/bits/unordered_map.h:798
#14 comphelper::detail::ConfigurationWrapper::~ConfigurationWrapper() () at /usr/src/debug/libreoffice-7.6.0.3/comphelper/source/misc/configuration.cxx:167
#15 0x00007facbee41b66 in __run_exit_handlers () at /lib64/libc.so.6
#16 0x00007facbee41cb0 in  () at /lib64/libc.so.6
#17 0x00007facbee281f7 in __libc_start_call_main () at /lib64/libc.so.6
#18 0x00007facbee282b9 in __libc_start_main_impl () at /lib64/libc.so.6
#19 0x000055a5f4df30c5 in _start () at ../sysdeps/x86_64/start.S:115
[Inferior 1 (process 21830) detached]

Steps to Reproduce:
1.After working on calc with multiple sheets, I save my work then close it. Calc crashes at that point
2.
3.

Actual Results:
Calc crashes with a segfault

Expected Results:
Should close without a segfault


Reproducible: Sometimes


User Profile Reset: No

Additional Info:
Version: 7.6.0.3 (X86_64) / LibreOffice Community
Build ID: 60(Build:3)
CPU threads: 12; OS: Linux 6.4; UI render: default; VCL: kf5 (cairo+xcb)
Locale: en-US (en_US.UTF-8); UI: en-US
Calc: CL threaded
Comment 1 m_a_riosv 2023-08-31 19:54:07 UTC
Please test in safe mode, Menu/Help/Restart in Safe Mode
Comment 2 m_a_riosv 2023-08-31 19:55:53 UTC
And updating the version could be a good chance. Earlier versions are used to be the more problematic.
Comment 3 Eric 2023-09-01 06:33:44 UTC
As far as I can tell 7.6 is the latest version - i don't understand you comment on updating.
I'll test in safe mode but the issue is intermittent so it may take awhile. I'll try speeding things up by just repeating the opening and closing
Comment 4 Julien Nabet 2023-09-01 06:46:18 UTC
Just for the test, could you:
- open a term/console
- type:
export SAL_USE_VCLPLUGIN=gen
- try to reproduce the crash
?

The goal is to know if it's rendering dependent or not.

Also did you get LO 7.6.0.3 from openSUSE's repository or from another way?
(I've just read about Tumbleweed being a rolling release version).
I mean, when installing a software, there are dependencies to take into account.
Comment 5 m_a_riosv 2023-09-01 14:09:50 UTC
Yes, but there are subversion, specially with bug fixes. 
https://downloadarchive.documentfoundation.org/libreoffice/old/7.6.0.3/
Comment 6 Eric 2023-09-01 19:15:57 UTC
In order to really test in a timely fashion I need a script that will open  calc in  safe mode, open my ods calc file and then close it.
Calc does not stay in safe mode, you have to select it every time you open it
As I said his is intermittent but can someone determine from the code dump above whats going wrong? I can program in straight C all day but C++ is greek to me except on a very basic level. I was wondering if it was some kind of race condition or a bad pointer in a rarely used code path that could be found by code reading
BTW: Why doesn't the [KCrash Handler] show #1,#2 and #3?
Comment 7 Julien Nabet 2023-09-02 07:10:53 UTC
(In reply to Eric from comment #6)
> ...
> Calc does not stay in safe mode, you have to select it every time you open it
> ...
In this case, apply https://wiki.documentfoundation.org/QA/FirstSteps#Corrupted_user_profile

> As I said his is intermittent but can someone determine from the code dump
> above whats going wrong? I can program in straight C all day but C++ is
> greek to me except on a very basic level. 
Ok if you know well C, it's a good basis. Then you need to understand object-oriented programming part, perhaps you already did some Java?
Anyway, if you want to try contributing, a good start page is:
https://wiki.documentfoundation.org/Development/GetInvolved
Comment 8 John Kizer 2023-09-03 18:15:15 UTC
Created attachment 189321 [details]
system log from last crash exiting calc
Comment 9 Xisco Faulí 2023-09-11 12:08:26 UTC
it looks like a duplicate of bug 156683

*** This bug has been marked as a duplicate of bug 156683 ***
Comment 10 Eric 2023-09-22 13:26:22 UTC
Created attachment 189760 [details]
Another crash and debug data

calc crashed on me again on exit, this time the backtrace is a bit different
Comment 11 Stéphane Guillou (stragu) 2023-09-30 20:24:06 UTC
(In reply to Eric from comment #10)
> Created attachment 189760 [details]
> Another crash and debug data
> 
> calc crashed on me again on exit, this time the backtrace is a bit different

Eric, which version was that with?
Please see with version 7.6.2.1 if you can still reproduce.
Comment 12 Eric 2023-10-01 17:08:57 UTC
Well, I updated since i posted that and I am now on 7.6.1.1
I don't know what the version was on 9/22 but probably not the same.
I can't force it, i tried many times.
Comment 13 Stéphane Guillou (stragu) 2023-10-01 20:08:12 UTC
Yes, since the 22nd, version 7.6.2.1 was released, which includes another fix for bug 156683, which is why it would be great if you could update to see if it does inded resolve what you witnessed.
Much appreciated!
Comment 14 Eric 2023-10-18 01:00:17 UTC
7.6.1.1 on openSuse Tumbleweed. Crashed on close.
I am sure TW will release a newer version soon. I wont report any more crashes on 7.6.1.1

Application: soffice (soffice), signal: Segmentation fault

[KCrash Handler]
#4  0x00007f8587a72ac0 in rtl_uString_release () at /usr/lib64/libreoffice/program/libuno_sal.so.3
#5  0x00007f85846beeab in  () at /usr/lib64/libreoffice/program/libmergedlo.so
#6  0x00007f8584810083 in  () at /usr/lib64/libreoffice/program/libmergedlo.so
#7  0x00007f8583041b26 in __run_exit_handlers () at /lib64/libc.so.6
#8  0x00007f8583041c70 in  () at /lib64/libc.so.6
#9  0x00007f85830281b7 in __libc_start_call_main () at /lib64/libc.so.6
#10 0x00007f8583028279 in __libc_start_main_impl () at /lib64/libc.so.6
#11 0x0000558c2b9410c5 in  ()
[Inferior 1 (process 20941) detached]
Comment 15 Xisco Faulí 2023-10-18 07:28:46 UTC
(In reply to Eric from comment #14)
> 7.6.1.1 on openSuse Tumbleweed. Crashed on close.
> I am sure TW will release a newer version soon. I wont report any more
> crashes on 7.6.1.1
> 
> Application: soffice (soffice), signal: Segmentation fault
> 
> [KCrash Handler]
> #4  0x00007f8587a72ac0 in rtl_uString_release () at
> /usr/lib64/libreoffice/program/libuno_sal.so.3
> #5  0x00007f85846beeab in  () at
> /usr/lib64/libreoffice/program/libmergedlo.so
> #6  0x00007f8584810083 in  () at
> /usr/lib64/libreoffice/program/libmergedlo.so
> #7  0x00007f8583041b26 in __run_exit_handlers () at /lib64/libc.so.6
> #8  0x00007f8583041c70 in  () at /lib64/libc.so.6
> #9  0x00007f85830281b7 in __libc_start_call_main () at /lib64/libc.so.6
> #10 0x00007f8583028279 in __libc_start_main_impl () at /lib64/libc.so.6
> #11 0x0000558c2b9410c5 in  ()
> [Inferior 1 (process 20941) detached]

Hello Eric,
you should upgrade to LibreOffice 7.6.2.1
Comment 16 Eric 2023-10-18 14:01:16 UTC
I just did last night. So now lets see if this crash still happens on close.
Comment 17 Stéphane Guillou (stragu) 2023-11-16 16:01:11 UTC
Looks like the same crash in bug 158228 for 7.6.2.1, let's mark as new.
Eric also reproduced in 7.6.2.1, see bug 156683 comment 28.
Comment 18 Stéphane Guillou (stragu) 2023-11-16 16:01:51 UTC
*** Bug 158228 has been marked as a duplicate of this bug. ***
Comment 19 Stéphane Guillou (stragu) 2023-11-16 16:07:34 UTC
Created attachment 190876 [details]
chart of number of crashes with this signature per version

Looks like the signature is back in force in 7.6: https://crashreport.libreoffice.org/stats/signature/void%20rtl::str::release%3C_rtl_uString%3E(_rtl_uString*)
Comment 20 Eric 2023-11-16 16:20:59 UTC
I just get a "Not Found" page when i click on your link:
https://crashreport.libreoffice.org/stats/signature/void%20rtl::str::release%3C_rtl_uString%3E(_rtl_uString*)
Comment 21 Julien Nabet 2023-11-16 19:38:49 UTC
I forgot to talk about https://wiki.documentfoundation.org/QA/FirstSteps, could you give it a try?
Comment 22 Eric 2023-11-16 20:23:55 UTC
Yes, that link works.
Why did you post it though? It points to a page on how to submit bugs.
Comment 23 Julien Nabet 2023-11-16 20:28:05 UTC
(In reply to Eric from comment #22)
> Yes, that link works.
> Why did you post it though? It points to a page on how to submit bugs.

It provides more than this, it gives common root causes like corrupted LO profile, Skia related pb,etc and provide some solution or workarounds.
Comment 24 Matt K 2023-12-12 00:47:19 UTC
I'm unable to repro using the following:

Version: 24.8.0.0.alpha0+ (X86_64) / LibreOffice Community
Build ID: 1f45453152a34beea5ec1b858c45a585ead3a73e
CPU threads: 16; OS: Windows 10.0 Build 22631; UI render: default; VCL: win
Locale: en-US (en_US); UI: en-US
Calc: threaded

Can you please provide a sample document on which the crash occurs, with some repro steps of how you interacted with document, particularly if you can repro every time?
Comment 25 Eric 2023-12-12 01:08:24 UTC
The crash is intermittent and doesn't seem to be able to be forced. Also, one thing i seem to notice, if i have a few spreadsheets open (instances not tabs)(say 2 or 3 instances of calc) and then after awhile i close one then the other then the last ( not rapidly and sometimes a minute or more from close to close), it's always the last sheet closing that causes it. Actual content doesn't seem t matter. Each sheet had a few rows of data, not a big crowded sheet with tons of data and hundreds of rows. I think there might be some interaction between the sheets in the calc code?
Comment 26 m_a_riosv 2024-01-09 15:21:38 UTC
*** Bug 159088 has been marked as a duplicate of this bug. ***
Comment 27 John Stamp 2024-01-31 00:27:04 UTC
I think that I found a reliable way to trigger the crash.  At least, I can reliably trigger it with 24.2.0-rc2 in Debian testing.

  * Start Calc with a new spreadsheet.

  * Enter some text into a cell and hit return.

  * Copy the cell with ctrl-c or right click.

  * Close Calc.  This should trigger the crash.

I could save the file or not before closing, and even as its own separate step, but this rtl_uString_release crash only happens when I close Calc.
Comment 28 Stéphane Guillou (stragu) 2024-01-31 03:14:54 UTC
(In reply to John Stamp from comment #27)
> I think that I found a reliable way to trigger the crash.  At least, I can
> reliably trigger it with 24.2.0-rc2 in Debian testing.
> 
>   * Start Calc with a new spreadsheet.
> 
>   * Enter some text into a cell and hit return.
> 
>   * Copy the cell with ctrl-c or right click.
> 
>   * Close Calc.  This should trigger the crash.
> 
> I could save the file or not before closing, and even as its own separate
> step, but this rtl_uString_release crash only happens when I close Calc.
Thank you! Using these steps, I could crash it with:

Version: 7.6.4.1 (X86_64) / LibreOffice Community
Build ID: e19e193f88cd6c0525a17fb7a176ed8e6a3e2aa1
CPU threads: 8; OS: Linux 5.15; UI render: default; VCL: kf5 (cairo+xcb)
Locale: en-AU (en_AU.UTF-8); UI: en-US
Calc: threaded

Crash report: https://crashreport.libreoffice.org/stats/crash_details/39f7ddbb-7fa1-40fc-9b40-f36d280aedde

No crash with gtk3 VCL plugin, no crash in 7.5.9.2.
Comment 29 Stéphane Guillou (stragu) 2024-01-31 03:49:45 UTC
Tried bibisecting, but libreoffice-64-7.6 master does not crash with those steps. I also tried adding a dictionary to have auto spellcheck active.
(Repository is currently at 30ee116ee498160a366a5a618649f832ae371f7d)
Comment 30 Stéphane Guillou (stragu) 2024-01-31 04:11:00 UTC
Created attachment 192277 [details]
minidump with 7.6.5 daily build

Version: 7.6.5.0.0+ (X86_64) / LibreOffice Community
Build ID: 5db313902110c95c440bb0e96f46acc97092d7ca
CPU threads: 8; OS: Linux 5.15; UI render: default; VCL: kf5 (cairo+xcb)
Locale: en-AU (en_AU.UTF-8); UI: en-US
Calc: threaded

In console:

warn:sfx.control:458070:458070:sfx2/source/control/dispatch.cxx:1211: Childwindow slot missing: 25917
warn:sfx.control:458070:458070:sfx2/source/control/dispatch.cxx:1211: Childwindow slot missing: 26189
warn:sfx.control:458070:458070:sfx2/source/control/dispatch.cxx:1211: Childwindow slot missing: 26190
warn:xmloff:458070:458070:xmloff/source/text/XMLTextListAutoStylePool.cxx:163: getStyleFamilies() from XModel failed for export!
warn:editeng:458070:458070:editeng/source/uno/unofield.cxx:375: Id service unknown: 5
warn:editeng:458070:458070:editeng/source/uno/unofield.cxx:375: Id service unknown: 2
warn:editeng:458070:458070:editeng/source/uno/unofield.cxx:375: Id service unknown: 5
warn:editeng:458070:458070:editeng/source/uno/unofield.cxx:375: Id service unknown: 10
warn:editeng:458070:458070:editeng/source/uno/unofield.cxx:375: Id service unknown: 2
warn:editeng:458070:458070:editeng/source/uno/unofield.cxx:375: Id service unknown: 3
warn:editeng:458070:458070:editeng/source/uno/unofield.cxx:375: Id service unknown: 5
warn:editeng:458070:458070:editeng/source/uno/unofield.cxx:375: Id service unknown: 2
warn:editeng:458070:458070:editeng/source/uno/unofield.cxx:375: Id service unknown: 5
warn:editeng:458070:458070:editeng/source/uno/unofield.cxx:375: Id service unknown: 10
warn:editeng:458070:458070:editeng/source/uno/unofield.cxx:375: Id service unknown: 2
warn:editeng:458070:458070:editeng/source/uno/unofield.cxx:375: Id service unknown: 3
warn:xmloff:458070:458070:xmloff/source/text/XMLTextListAutoStylePool.cxx:163: getStyleFamilies() from XModel failed for export!
warn:xmloff:458070:458070:xmloff/source/text/XMLTextListAutoStylePool.cxx:163: getStyleFamilies() from XModel failed for export!
warn:legacy.osl:458070:458070:xmloff/source/core/SettingsExportHelper.cxx:176: this type is not implemented now
warn:legacy.tools:458070:458070:sfx2/source/bastyp/progress.cxx:200: Can't make progress bar!
warn:legacy.tools:458070:458070:sfx2/source/bastyp/progress.cxx:200: Can't make progress bar!
warn:sfx.appl:458070:458070:sfx2/source/appl/app.cxx:166: Memory leak: some object shells were not removed!
warn:unotools.config:458070:458070:unotools/source/config/configmgr.cxx:147: ConfigManager not empty
Comment 31 Xisco Faulí 2024-01-31 08:46:38 UTC
*** Bug 152477 has been marked as a duplicate of this bug. ***
Comment 32 Xisco Faulí 2024-01-31 08:56:55 UTC
Looking at other reports of the same crash, the issue started in branch 7.4 with 7df433cdc33b4d6ba38eafad9282d015571433ef "optimize ConfigurationProperty::get()" which was later reverted. See https://cgit.freedesktop.org/libreoffice/core/log/?h=libreoffice-7-4&qt=grep&q=optimize+ConfigurationProperty%3A%3Aget%28%29
In branch 7.5 the same happened: https://cgit.freedesktop.org/libreoffice/core/log/?h=libreoffice-7-5&qt=grep&q=optimize+ConfigurationProperty%3A%3Aget%28%29 while in branch 7.6 the issue was reverted and later reapplied, https://cgit.freedesktop.org/libreoffice/core/log/?h=libreoffice-7-6&qt=grep&q=optimize+ConfigurationProperty%3A%3Aget()
Comment 33 Commit Notification 2024-01-31 11:53:51 UTC
Xisco Fauli committed a patch related to this issue.
It has been pushed to "libreoffice-7-6":

https://git.libreoffice.org/core/commit/7b46c77366fb3effd2de9bf5ba11ebd3c064974a

tdf#157042: Revert "re-apply "optimize ConfigurationProperty::get()""

It will be available in 7.6.5.

The patch should be included in the daily builds available at
https://dev-builds.libreoffice.org/daily/ in the next 24-48 hours. More
information about daily builds can be found at:
https://wiki.documentfoundation.org/Testing_Daily_Builds

Affected users are encouraged to test the fix and report feedback.
Comment 34 John Stamp 2024-02-01 03:33:44 UTC
Hi Xisco,

I tried the 2024-02-01 daily build for 7.6.5 and couldn't trigger a crash anymore while using the kf5 vcl.

The 2024-02-01 daily build for 24.2 still triggers a crash though.
Comment 35 Xisco Faulí 2024-02-05 10:59:37 UTC
(In reply to John Stamp from comment #34)
> Hi Xisco,
> 
> I tried the 2024-02-01 daily build for 7.6.5 and couldn't trigger a crash
> anymore while using the kf5 vcl.
> 
> The 2024-02-01 daily build for 24.2 still triggers a crash though.

Hi John,
Looking at https://crashreport.libreoffice.org/stats/signature/void%20rtl::str::release%3C_rtl_uString%3E(_rtl_uString*)#summary I see the issue is not being reported for 24.2 version. Do you still reproduce the issue with LibreOffice 24.2.0.3 from https://www.libreoffice.org/download/download-libreoffice/ ?
Comment 36 John Stamp 2024-02-05 22:07:39 UTC
Yes, after installing all the 24.2.0.3 packages from LibreOffice, I can still crash Calc on close, but only with the kde vcl.

The "Send crash reports" option was enabled, so hopefully you got one.

The KCrash Handler output wasn't very helpful, but here's the relevant bit:

  [KCrash Handler]
  #4  0x00007fef7a7c3210 in rtl_uString_release () from /opt/libreoffice24.2/program/libuno_sal.so.3
  #5  0x00007fef76e0f40b in ?? () from /opt/libreoffice24.2/program/libmergedlo.so
  #6  0x00007fef76f61ee3 in ?? () from /opt/libreoffice24.2/program/libmergedlo.so

  
Finally, I was poking around in the crash reports and came across a number of <name omitted> signatures for Linux & 24.2.0.3

  https://crashreport.libreoffice.org/stats/signature/%3Cname%20omitted%3E

A number of them are crashing at rtl::str::release():

  https://crashreport.libreoffice.org/stats/crash_details/07a51006-082e-46dd-808d-ed04faf6d930

  https://crashreport.libreoffice.org/stats/crash_details/55f2f1d8-a7f8-4b0d-9f95-d328ce581e36

  https://crashreport.libreoffice.org/stats/crash_details/3d5e7d71-f5da-4dac-a357-95ca0071bb8f

  etc...

Same issue?
Comment 37 Michael Weghorn 2024-02-06 09:52:39 UTC
(In reply to John Stamp from comment #27)
> I think that I found a reliable way to trigger the crash.  At least, I can
> reliably trigger it with 24.2.0-rc2 in Debian testing.
> 
>   * Start Calc with a new spreadsheet.
> 
>   * Enter some text into a cell and hit return.
> 
>   * Copy the cell with ctrl-c or right click.
> 
>   * Close Calc.  This should trigger the crash.
> 
> I could save the file or not before closing, and even as its own separate
> step, but this rtl_uString_release crash only happens when I close Calc.

I can reproduce with a current daily build:

Version: 24.8.0.0.alpha0+ (X86_64) / LibreOffice Community
Build ID: 6b6849107562b258aa8858e94ff3c07160f07062
CPU threads: 12; OS: Linux 6.5; UI render: default; VCL: kf5 (cairo+xcb)
Locale: en-GB (en_GB.UTF-8); UI: en-US
Calc: threaded

but interestingly not with a local debug build:

Version: 24.8.0.0.alpha0+ (X86_64) / LibreOffice Community
Build ID: 4d381b54d1c598c181b4a21a8bf0db86eb4668d1
CPU threads: 12; OS: Linux 6.5; UI render: default; VCL: kf5 (cairo+xcb)
Locale: en-GB (en_GB.UTF-8); UI: en-US
Calc: threaded
Comment 38 Commit Notification 2024-02-07 07:24:16 UTC
Xisco Fauli committed a patch related to this issue.
It has been pushed to "libreoffice-24-2":

https://git.libreoffice.org/core/commit/dd7a69068b8f92052b45a60014e6f6ed689c2947

tdf#157042: Revert "re-apply "optimize ConfigurationProperty::get()""

It will be available in 24.2.1.

The patch should be included in the daily builds available at
https://dev-builds.libreoffice.org/daily/ in the next 24-48 hours. More
information about daily builds can be found at:
https://wiki.documentfoundation.org/Testing_Daily_Builds

Affected users are encouraged to test the fix and report feedback.
Comment 39 John Stamp 2024-02-08 16:44:35 UTC
Yes, that seems to fix it.  I can't make it crash anymore.

Version: 24.2.1.0.0+ (X86_64) / LibreOffice Community
Build ID: c7e9e3a35c00cae8cea13cbef1de9c696a881cdb
CPU threads: 1; OS: Linux 6.6; UI render: default; VCL: kf5 (cairo+xcb)
Locale: en-US (en_US.UTF-8); UI: en-US
Calc: threaded
Comment 40 Commit Notification 2024-02-08 20:33:40 UTC
Xisco Fauli committed a patch related to this issue.
It has been pushed to "master":

https://git.libreoffice.org/core/commit/cfcd6b25575087366924e39cf352bf3bb3853d6a

tdf#157042: Revert "re-apply "optimize ConfigurationProperty::get()""

It will be available in 24.8.0.

The patch should be included in the daily builds available at
https://dev-builds.libreoffice.org/daily/ in the next 24-48 hours. More
information about daily builds can be found at:
https://wiki.documentfoundation.org/Testing_Daily_Builds

Affected users are encouraged to test the fix and report feedback.
Comment 41 Xisco Faulí 2024-02-08 20:34:54 UTC
The patch has been reverted in all branches ( master, libreoffice-24-2 and libreoffice-7-6 ) so this can be considered as fixed. Please retest it to make it sure it has been fixed in all branches
Comment 42 Stéphane Guillou (stragu) 2024-02-22 04:10:02 UTC
*** Bug 159831 has been marked as a duplicate of this bug. ***