Summary: | assertion, DocumentRedlineManager.cxx:246 | ||
---|---|---|---|
Product: | LibreOffice | Reporter: | Terrence Enger <lo_bugs> |
Component: | Writer | Assignee: | László Németh <nemeth> |
Status: | RESOLVED FIXED | ||
Severity: | normal | CC: | serval2412, telesto |
Priority: | medium | Keywords: | haveBacktrace |
Version: | 7.1.0.0.alpha0+ | ||
Hardware: | All | ||
OS: | All | ||
Whiteboard: | |||
Crash report or crash signature: | Regression By: | ||
Bug Depends on: | |||
Bug Blocks: | 105537, 137770 | ||
Attachments: |
bt with debug symbols
backtrace after bug 138771 is fixed |
Description
Terrence Enger
2020-10-26 15:04:44 UTC
To be sure this doesn't get lost Bug 137684 comment 4 From: László Németh @Terence: thanks for your detailed bug report and analysis! Missing margin data and backspace crash are different bugs, and they will handled by the upcoming patches. I will be able to continue on only in the next month, but my plan is to finish these in December. Please, file the detected problems in different bug reports, this could speed up my work because that is based on a ticket system, or you can wait a few weeks for a better solution (because otherwise I must work on other smaller change tracking/Writer problems in parallel). @Julien Would you mind to confirm. Saves me the trouble of downloading a debug build :P Created attachment 166784 [details]
bt with debug symbols
On pc Debian x86-64 with master sources updated today, I could reproduce this.
Just a note that the backtraces in bug description and in comment 3 show the same function signatures. Created attachment 167368 [details] backtrace after bug 138771 is fixed The backtrace is from commit b80c2745 (2020-11-17) built and running on debian-buster. After bug 137771 is fixed, I get a different failure. Using the step numbers from bug description: (4) Works as expected. That was bug 137771. (5) Brings entire word "brown" back into the document and leaves left margin empty. (6) Program crashes with different assertion (one line,rewrapped): soffice.bin: /home/terry/lo_hacking/git/libo6/sw/source/core/bastyp/bparr.cxx:84: BigPtrEntry* BigPtrArray::operator[](sal_uLong) const: Assertion `idx < m_nSize' failed. Step (3) is still necessary to the crash. (In reply to Terrence Enger from comment #5) > (6) Program crashes with different assertion (one line,rewrapped): > > soffice.bin: > /home/terry/lo_hacking/git/libo6/sw/source/core/bastyp/bparr.cxx:84: > BigPtrEntry* BigPtrArray::operator[](sal_uLong) const: > Assertion `idx < m_nSize' failed. Would say another example of bug 135167 (or one of the others in the metabug) I see that the assertion here is raised at the same line as the assertion in bug 137167 and that the overwhelming majority of the function signatures in the backtraces are the same. I assume that the assertion I originally reported was fixed along with bug 13772 by https://git.libreoffice.org/core/commit/d129e1785e541f217d4f6547faec87b9fa772019, and I am setting status RESOLVED FIXED. Thank you, Telesto, for catching this. |