Open attachment 140789 [details] (from bug 116539), and save it again as a PPT. => Save crashes (release build) or triggers the following assertion failure (debug build): https://opengrok.libreoffice.org/xref/core/vcl/source/app/scheduler.cxx#597 assert(nullptr == mpSchedulerData || utl::ConfigManager::IsFuzzing()); Observed using 6.2 master daily build & 6.0.0.3 / Windows 7. Since bug 116539 mentions no crash in Linux, I'm assuming this is Windows-specific. No crash with 5.4.0.3. => regression In 6.0.0.3 and onwards there's also a wide white rectangle covering part of the slide. Might be related to the crash. Crash report: https://crashreport.libreoffice.org/stats/crash_details/1c4233c0-626d-4e6d-9aea-53a87f100f88
Interestingly the debugger breaks elsewhere: setOriginURL(mpSwapFile->maOriginURL); https://opengrok.libreoffice.org/xref/core/vcl/source/gdi/impgraph.cxx#1579 Here mpSwapFile->maOriginURL is ??? (undefined I assume), even though it was empty until the bRet = ImplSwapIn( xIStm.get() ) call a bit above. Actually, ImplSwapIn( SvStream* xIStm ) calls ImplClear(), which does this: mpSwapFile.reset(); So if the swap file is thrown away, why is its maOriginURL taken later? Miklos, can you please take a look at what's supposed to be going on there?
Confirmed on windows 7 x64 with Version: 6.0.4.2 (x64) Build ID: 9b0d9b32d5dcda91d2f1a96dc04c645c450872bf CPU threads: 3; OS: Windows 6.1; UI render: default crashreport.libreoffice.org/stats/crash_details/91ef0460-5d32-472a-816a-0ed4da90d367 Unconfirmed on ubuntu 16.04 x64 with Version: 6.1.0.0.alpha1+ Build ID: 47dc3115f12ff16dc326b6edd12c46e6a6ef1843 CPU threads: 2; OS: Linux 4.4; UI render: default; VCL: gtk2; TinderBox: Linux-rpm_deb-x86_64@70-TDF, Branch:master, Time: 2018-05-17_00:32:17 Locale: en-US (en_US.UTF-8); Calc: Seems windows only then. Also the white rectangle isn't there in the linux build. But saving seems a lot slower than with v6.0.0.0
This seems to have begun at the below commit. Adding Cc: to Tomaž Vajngerl; Could you possibly take a look at this one? Thanks 0b17b35d7aff81ec5ba3790ff3544d18e5f6ae44 is the first bad commit commit 0b17b35d7aff81ec5ba3790ff3544d18e5f6ae44 Author: Norbert Thiebaud <nthiebaud@gmail.com> Date: Tue May 8 17:47:24 2018 -0700 source 711c2e49dd3c51877263148267344e2eb4ca7c0d author Tomaž Vajngerl <tomaz.vajngerl@collabora.co.uk> 2018-05-08 19:45:36 +0900 committer Tomaž Vajngerl <quikee@gmail.com> 2018-05-09 02:03:16 +0200 commit 711c2e49dd3c51877263148267344e2eb4ca7c0d (patch) tree f6b676ceb8a607922bdcff0b1d70567c03d5a26f parent f4e7d521aab79ecbb16c0518f14c7635902e9e10 (diff) tdf#116272 use correct property name when exporting to PPT
On pc Debian x86-64 with master sources updated yesterday + enable-dbgutil, I killed LO after about 30 seconds because it was too slow. I noticed these kinds of logs: warn:legacy.tools:5207:5207:sfx2/source/control/bindings.cxx:793: SfxBindings::Register while status-updating warn:legacy.tools:5207:5207:sfx2/source/control/bindings.cxx:1288: Reschedule in StateChanged => buff warn:vcl.gdi:5207:5207:vcl/source/graphic/Manager.cxx:135: Calculated size mismatch. Variable size is '-89853776' but calculated size is '114107674' warn:vcl.gdi:5207:5207:vcl/source/graphic/Manager.cxx:135: Calculated size mismatch. Variable size is '725992024' but calculated size is '409817674'
Work perfectly with Version: 6.1.0.0.alpha1+ Build ID: 1e2afc9bd3062cfba6b65b45c17a08f298014239 CPU threads: 4; OS: Linux 4.8; UI render: default; VCL: gtk3; Locale: fr-FR (fr_FR.UTF-8); Calc: group Linux mint 18.2 64 bits.
I guess the crash report and the most recent crash are different bugs, considering one was there in 6.0, and the other was introduce during 6.1 development. Additionally, my observations in comment 1 again refer to a different bug, only observable when debugging a Windows debug build with OpenGL disabled (maybe Linux, too, haven't checked). That was introduced in the following commit: https://cgit.freedesktop.org/libreoffice/core/commit/?id=075e2d8b8d98b1c6daa430e9b9a396b15ba22837 author Tomaž Vajngerl <tomaz.vajngerl@collabora.co.uk> 2018-02-13 21:26:40 +0900 committer Tomaž Vajngerl <quikee@gmail.com> 2018-02-13 20:29:24 +0100 graphic: Remember the origin URL after swap out - swap in
I can't reproduce the crash in Version: 6.2.0.0.alpha0+ Build ID: 48b49937fed5e50d299a94063eb325799ff672e9 CPU threads: 1; OS: Windows 6.1; UI render: default; TinderBox: Win-x86@42, Branch:master, Time: 2018-06-05_01:47:06 Locale: es-ES (es_ES); Calc: group threaded @MM, @Aron, @Raal, @Julien, Could you please retry with a daily build?
Not reproduced in Version: 6.2.0.0.alpha0+ Build ID: 48b49937fed5e50d299a94063eb325799ff672e9 CPU threads: 1; OS: Windows 6.1; UI render: default; TinderBox: Win-x86@42, Branch:master, Time: 2018-06-05_01:47:06 Locale: es-ES (es_ES); Calc: group threaded either
Not reproduced in Version: 6.1.0.0.beta1+ Build ID: 91d8af2c5cf4e8ec0f1ce0e532e0c896de77750b CPU threads: 16; OS: Windows 6.3; UI render: default; Locale: en-GB (en_GB); Calc: group threaded either
Still getting it with 565340d457f41197474a75ba1b036bdc3d569041 (from 06-03), only with default rendering.
Adding an if guard checking if mpSwapFile isn't empty around the following lines gets rid of the debugger break and the assertion failure: https://opengrok.libreoffice.org/xref/core/vcl/source/gdi/impgraph.cxx#1592 ====== setOriginURL(mpSwapFile->maOriginURL); mpSwapFile.reset(); ====== However, I'm not sure if that makes the code consistent, and would rather leave the change to Tomaz.
Tomaž Vajngerl committed a patch related to this issue. It has been pushed to "master": http://cgit.freedesktop.org/libreoffice/core/commit/?id=3cd264bb09c6c14bf9eb769d508d7015c4bdc83b tdf#117797 guard access to mpSwapFile as it may not be set It will be available in 6.2.0. 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.
Tomaž Vajngerl committed a patch related to this issue. It has been pushed to "libreoffice-6-1": http://cgit.freedesktop.org/libreoffice/core/commit/?id=22dbd5904f4f69d98530fecdd9e4383dd088f76f&h=libreoffice-6-1 tdf#117797 guard access to mpSwapFile as it may not be set It will be available in 6.1.1. 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.
This bug report mentioned different issues I encountered after following the same steps. Let it only concern the issue Tomaž fixed, thanks for that! Closing as FIXED. I'm not getting an assert anymore, perhaps that's gone as well, but saving still crashes after consuming a lot of memory, I will check how that behaves sometimes, and open a different bug report if needed.
> Let it only concern the issue Tomaž fixed, thanks for that! > Closing as FIXED. Setting to VERIFIED!